Mr. S, age 47, weighs 209 lb and has a history of seizure disorder, bipolar disorder not otherwise specified, hypertension, and type 2 diabetes mellitus. He presents to the emergency department after not taking his medications for 2 days while on vacation. He has increased energy, decreased sleep, and pressured speech, and insists on walking for up to 10 hours per day “in preparation for a marathon,” even though he has a 4-cm foot ulcer. His family reports that he had been compliant with his medications until the present incident.

 

Mr. S has no known drug allergies. His medications include oral divalproex sodium delayed release (valproic acid [VPA]), 1,000 mg twice a day, oral lisinopril, 20 mg every morning, and insulin glargine, 22 units subcutaneously every evening.

A complete blood count, basic metabolic panel, creatine kinase level, VPA level, and urine drug screen are ordered. Relevant results include a serum creatinine level of 1.4 mg/dL (normal range: 0.6 to 1.2 mg/dL), a glucose serum level of 188 mg/dL (normal range: 70 to 100 mg/dL), and a VPA level of 23 mcg/mL (therapeutic range: 50 to 125 mcg/mL). A liver function panel is within normal limits: albumin level of 3.9 g/dL, aspartate aminotransferase level of 18 IU/L, and alanine aminotransferase level of 14 IU/L. In light of Mr. S’s seizure history, neurology is consulted and the decision is made to continue treating him with VPA because he has been seizure-free for 4.5 years and this medication has also helped with his bipolar disorder.

Mr. S is admitted to the hospital and his home medications are resumed at the current doses. On hospital Day 3, Mr. S’s VPA level is 62 mcg/mL, his obsession with a marathon has remitted, and his sleep pattern has normalized. Infectious disease and podiatry services are consulted for his diabetic foot infection, which has ulcerated down to the bone. IV ertapenem, 1,000 mg/d, is initiated with plans for debridement the following week. Two days later, Mr. S has a witnessed seizure; his VPA level is 9 mcg/mL.

A common question asked of pharmacists is, “Will protein binding changes affect drug dosages?” In this article, I describe how protein binding changes may occur, and the complexity of the dynamic. Being highly bound to a protein typically does not mean all medications will interact, but some interactions can be important. This article does not cover medications that bind to hormones.

Why is protein binding important? When a medication is bound to plasma protein, it is not free to act. There can be a delay in therapeutic effect (because no drug is available to react), delayed elimination, or possibly displacement of another protein-bound medication. Additionally, medications tend not to cross the blood-brain barrier or be eliminated when bound. For example, if a drug is 99% bound (leaving 1% free) and displacement now leaves 2% of the drug free, this event has doubled the amount of free drug. As the unbound medication is eliminated, the drug that is bound to the protein can act as a reservoir. A dynamic relationship exists between bound drug, unbound drug, and rate of elimination.

Which proteins do drugs commonly bind to? The proteins often associated with binding include albumin, alpha-1-acid glycoprotein (AAG), and lipoproteins. Albumin comprises 60% of total plasma protein in the plasma. Lipoproteins include very high-density lipoprotein (VHDL), high-density lipoprotein (HDL), very low-density lipoprotein (VLDL), and low-density lipoprotein (LDL).¹ Medications that bind to lipoproteins include cyclosporine, tacrolimus, and propofol.²

Continued to: What common disease states can cause hypoalbuminemia?

 

What common disease states can cause hypoalbuminemia? Many disease states can result in low albumin levels. The most common ones are malnutrition, malignancies, stress, injury, burns, pregnancy, and diabetes.³ When there is less albumin to bind to, free drug levels may be increased.

Can AAG levels change with disease states as well? Because AAG accounts for a lower percentage of total plasma protein than albumin, there may be less clinical concern regarding AAG. AAG levels usually do not drop, but instead can become elevated during times of trauma, inflammation, and acute myocardial infarction. This could result in increased binding of the free drug.⁴Which medications bind to red blood cells (RBCs)? There are several locations for drugs to bind to RBCs, including to hemoglobin and the plasma membrane. Medications that commonly bind to RBCs include barbiturates, chlorpromazine, imipramine, and phenytoin.⁵

What are common highly-bound medications? The Table¹ provides examples of medications that are >90% protein-bound. However, this information may be misleading because many medications are highly bound. Zhang et al¹ compiled binding data for 222 drugs, half of which bind 90% to 100%. However, the literature does not indicate that they all have clinically significant interactions. Benet and Hoener⁶ discuss how factors other than protein binding affect potential drug interactions, and the complexity of the body’s ability to compensate for increased free drug. Medication characteristics that may contribute to producing a significant interaction include, but are not limited to:

• free vs protein-bound drug in the plasma or tissue
• volume of distribution
• organs affected
• hepatic bioavailability
• drug clearance.

For example, VPA is 93% protein-bound and phenytoin is 91% protein-bound.¹ However, this interaction is affected by more than just protein binding. VPA not only displaces the protein-bound phenytoin, but also inhibits its metabolism, which together result in increased free phenytoin levels.

Continued to: Another area of concern is a critically ill patient...

 

Another area of concern is a critically ill patient who has a change in his or her pH. Medications that are highly bound and have high clearance rates may be affected. This is of particular concern when prescribing antibiotics that are time-dependent, such as beta-lactams.³

What happened to Mr. S? Mr. S likely experienced a drug–drug interaction that resulted in a subtherapeutic VPA level and subsequent seizure. Case reports have shown evidence that the carbapenem class of antibiotics, which includes ertapenem, interacts with VPA.⁷ Proposed mechanisms include a lowering of VPA serum levels due to a redistribution of the VPA onto the RBCs due to carbapenem. Other theories include the possibility that carbapenems may limit oral VPA absorption, decrease VPA enterohepatic recirculation, and increase VPA metabolism.⁷ Using VPA and ertapenem together is discouraged because seizures have been reported among patients receiving this combination. If it is medically necessary to administer VPA and ertapenem, closely monitor VPA levels. In Mr. S’s case, another broad-spectrum antibiotic, such as piperacillin-tazobactam, could have been used, for his diabetic foot infection.

While many medications may have high protein binding, there are few clinically important known interactions. However, our understanding of the relationship between protein binding and drug interactions may improve with additional research.

CASE CONTINUED

Under neurology’s care, lacosamide is added for treatment of Mr. S’s seizures. No more seizures are noted during the remainder of his hospitalization. Infectious disease services change his antibiotic to piperacillin-tazobactam. Mr. S continues to progress well and is discharged to a rehabilitation center 2 days later.

Related Resource

• DrugBank. www.drugbank.ca. Canadian Institutes of Health Research.

Drug Brand Names
Amiodarone • Cordarone, Pacerone
Bumetanide • Bumex
Bupivacaine • Marcaine, Sensorcaine
Buprenorphine • Belbuca, Subutex
Ceftriaxone • Rocephin
Chlordiazepoxide • Librium
Chlorpromazine • Thorazine
Clozapine • Clozaril
Cyclosporine • Gengraf, Neoral
Diazepam • Valium
Doxycycline • Acticlate, Doryx
Duloxetine • Cymbalta
Ertapenem • Invanz
Fluoxetine • Prozac, Sarafem
Furosemide • Lasix
Glargine (Insulin) • Lantus, Toujeo
Glipizide • Glucotrol
Haloperidol • Haldol
Ibuprofen • Advil, Motrin
Imipramine • Tofranil
Lacosamide • Vimpat
Lisinopril • Prinivil, Zestril
Lorazepam • Ativan
Nicardipine • Cardene
Nortriptyline • Pamelor
Paclitaxel • Abraxane, Taxol
Phenytoin • Dilantin, Phenytek
Piperacillin-tazobactam • Zosyn
Propofol • Diprivan
Sertraline • Zoloft
Tacrolimus • Prograf
Tamoxifen • Soltamox
Valproic acid • Depakene, Depakote
Verapamil • Calan, Verelan
Warfarin • Coumadin, Jantoven

Mr. S, age 47, weighs 209 lb and has a history of seizure disorder, bipolar disorder not otherwise specified, hypertension, and type 2 diabetes mellitus. He presents to the emergency department after not taking his medications for 2 days while on vacation. He has increased energy, decreased sleep, and pressured speech, and insists on walking for up to 10 hours per day “in preparation for a marathon,” even though he has a 4-cm foot ulcer. His family reports that he had been compliant with his medications until the present incident.

 

Mr. S has no known drug allergies. His medications include oral divalproex sodium delayed release (valproic acid [VPA]), 1,000 mg twice a day, oral lisinopril, 20 mg every morning, and insulin glargine, 22 units subcutaneously every evening.

A complete blood count, basic metabolic panel, creatine kinase level, VPA level, and urine drug screen are ordered. Relevant results include a serum creatinine level of 1.4 mg/dL (normal range: 0.6 to 1.2 mg/dL), a glucose serum level of 188 mg/dL (normal range: 70 to 100 mg/dL), and a VPA level of 23 mcg/mL (therapeutic range: 50 to 125 mcg/mL). A liver function panel is within normal limits: albumin level of 3.9 g/dL, aspartate aminotransferase level of 18 IU/L, and alanine aminotransferase level of 14 IU/L. In light of Mr. S’s seizure history, neurology is consulted and the decision is made to continue treating him with VPA because he has been seizure-free for 4.5 years and this medication has also helped with his bipolar disorder.

Mr. S is admitted to the hospital and his home medications are resumed at the current doses. On hospital Day 3, Mr. S’s VPA level is 62 mcg/mL, his obsession with a marathon has remitted, and his sleep pattern has normalized. Infectious disease and podiatry services are consulted for his diabetic foot infection, which has ulcerated down to the bone. IV ertapenem, 1,000 mg/d, is initiated with plans for debridement the following week. Two days later, Mr. S has a witnessed seizure; his VPA level is 9 mcg/mL.

A common question asked of pharmacists is, “Will protein binding changes affect drug dosages?” In this article, I describe how protein binding changes may occur, and the complexity of the dynamic. Being highly bound to a protein typically does not mean all medications will interact, but some interactions can be important. This article does not cover medications that bind to hormones.

Why is protein binding important? When a medication is bound to plasma protein, it is not free to act. There can be a delay in therapeutic effect (because no drug is available to react), delayed elimination, or possibly displacement of another protein-bound medication. Additionally, medications tend not to cross the blood-brain barrier or be eliminated when bound. For example, if a drug is 99% bound (leaving 1% free) and displacement now leaves 2% of the drug free, this event has doubled the amount of free drug. As the unbound medication is eliminated, the drug that is bound to the protein can act as a reservoir. A dynamic relationship exists between bound drug, unbound drug, and rate of elimination.

Which proteins do drugs commonly bind to? The proteins often associated with binding include albumin, alpha-1-acid glycoprotein (AAG), and lipoproteins. Albumin comprises 60% of total plasma protein in the plasma. Lipoproteins include very high-density lipoprotein (VHDL), high-density lipoprotein (HDL), very low-density lipoprotein (VLDL), and low-density lipoprotein (LDL).¹ Medications that bind to lipoproteins include cyclosporine, tacrolimus, and propofol.²

Continued to: What common disease states can cause hypoalbuminemia?

 

What common disease states can cause hypoalbuminemia? Many disease states can result in low albumin levels. The most common ones are malnutrition, malignancies, stress, injury, burns, pregnancy, and diabetes.³ When there is less albumin to bind to, free drug levels may be increased.

Can AAG levels change with disease states as well? Because AAG accounts for a lower percentage of total plasma protein than albumin, there may be less clinical concern regarding AAG. AAG levels usually do not drop, but instead can become elevated during times of trauma, inflammation, and acute myocardial infarction. This could result in increased binding of the free drug.⁴Which medications bind to red blood cells (RBCs)? There are several locations for drugs to bind to RBCs, including to hemoglobin and the plasma membrane. Medications that commonly bind to RBCs include barbiturates, chlorpromazine, imipramine, and phenytoin.⁵

What are common highly-bound medications? The Table¹ provides examples of medications that are >90% protein-bound. However, this information may be misleading because many medications are highly bound. Zhang et al¹ compiled binding data for 222 drugs, half of which bind 90% to 100%. However, the literature does not indicate that they all have clinically significant interactions. Benet and Hoener⁶ discuss how factors other than protein binding affect potential drug interactions, and the complexity of the body’s ability to compensate for increased free drug. Medication characteristics that may contribute to producing a significant interaction include, but are not limited to:

• free vs protein-bound drug in the plasma or tissue
• volume of distribution
• organs affected
• hepatic bioavailability
• drug clearance.

For example, VPA is 93% protein-bound and phenytoin is 91% protein-bound.¹ However, this interaction is affected by more than just protein binding. VPA not only displaces the protein-bound phenytoin, but also inhibits its metabolism, which together result in increased free phenytoin levels.

Continued to: Another area of concern is a critically ill patient...

 

Another area of concern is a critically ill patient who has a change in his or her pH. Medications that are highly bound and have high clearance rates may be affected. This is of particular concern when prescribing antibiotics that are time-dependent, such as beta-lactams.³

What happened to Mr. S? Mr. S likely experienced a drug–drug interaction that resulted in a subtherapeutic VPA level and subsequent seizure. Case reports have shown evidence that the carbapenem class of antibiotics, which includes ertapenem, interacts with VPA.⁷ Proposed mechanisms include a lowering of VPA serum levels due to a redistribution of the VPA onto the RBCs due to carbapenem. Other theories include the possibility that carbapenems may limit oral VPA absorption, decrease VPA enterohepatic recirculation, and increase VPA metabolism.⁷ Using VPA and ertapenem together is discouraged because seizures have been reported among patients receiving this combination. If it is medically necessary to administer VPA and ertapenem, closely monitor VPA levels. In Mr. S’s case, another broad-spectrum antibiotic, such as piperacillin-tazobactam, could have been used, for his diabetic foot infection.

While many medications may have high protein binding, there are few clinically important known interactions. However, our understanding of the relationship between protein binding and drug interactions may improve with additional research.

CASE CONTINUED

Under neurology’s care, lacosamide is added for treatment of Mr. S’s seizures. No more seizures are noted during the remainder of his hospitalization. Infectious disease services change his antibiotic to piperacillin-tazobactam. Mr. S continues to progress well and is discharged to a rehabilitation center 2 days later.

Related Resource

• DrugBank. www.drugbank.ca. Canadian Institutes of Health Research.

Drug Brand Names
Amiodarone • Cordarone, Pacerone
Bumetanide • Bumex
Bupivacaine • Marcaine, Sensorcaine
Buprenorphine • Belbuca, Subutex
Ceftriaxone • Rocephin
Chlordiazepoxide • Librium
Chlorpromazine • Thorazine
Clozapine • Clozaril
Cyclosporine • Gengraf, Neoral
Diazepam • Valium
Doxycycline • Acticlate, Doryx
Duloxetine • Cymbalta
Ertapenem • Invanz
Fluoxetine • Prozac, Sarafem
Furosemide • Lasix
Glargine (Insulin) • Lantus, Toujeo
Glipizide • Glucotrol
Haloperidol • Haldol
Ibuprofen • Advil, Motrin
Imipramine • Tofranil
Lacosamide • Vimpat
Lisinopril • Prinivil, Zestril
Lorazepam • Ativan
Nicardipine • Cardene
Nortriptyline • Pamelor
Paclitaxel • Abraxane, Taxol
Phenytoin • Dilantin, Phenytek
Piperacillin-tazobactam • Zosyn
Propofol • Diprivan
Sertraline • Zoloft
Tacrolimus • Prograf
Tamoxifen • Soltamox
Valproic acid • Depakene, Depakote
Verapamil • Calan, Verelan
Warfarin • Coumadin, Jantoven

Mr. S, age 47, weighs 209 lb and has a history of seizure disorder, bipolar disorder not otherwise specified, hypertension, and type 2 diabetes mellitus. He presents to the emergency department after not taking his medications for 2 days while on vacation. He has increased energy, decreased sleep, and pressured speech, and insists on walking for up to 10 hours per day “in preparation for a marathon,” even though he has a 4-cm foot ulcer. His family reports that he had been compliant with his medications until the present incident.

 

Mr. S has no known drug allergies. His medications include oral divalproex sodium delayed release (valproic acid [VPA]), 1,000 mg twice a day, oral lisinopril, 20 mg every morning, and insulin glargine, 22 units subcutaneously every evening.

A complete blood count, basic metabolic panel, creatine kinase level, VPA level, and urine drug screen are ordered. Relevant results include a serum creatinine level of 1.4 mg/dL (normal range: 0.6 to 1.2 mg/dL), a glucose serum level of 188 mg/dL (normal range: 70 to 100 mg/dL), and a VPA level of 23 mcg/mL (therapeutic range: 50 to 125 mcg/mL). A liver function panel is within normal limits: albumin level of 3.9 g/dL, aspartate aminotransferase level of 18 IU/L, and alanine aminotransferase level of 14 IU/L. In light of Mr. S’s seizure history, neurology is consulted and the decision is made to continue treating him with VPA because he has been seizure-free for 4.5 years and this medication has also helped with his bipolar disorder.

Mr. S is admitted to the hospital and his home medications are resumed at the current doses. On hospital Day 3, Mr. S’s VPA level is 62 mcg/mL, his obsession with a marathon has remitted, and his sleep pattern has normalized. Infectious disease and podiatry services are consulted for his diabetic foot infection, which has ulcerated down to the bone. IV ertapenem, 1,000 mg/d, is initiated with plans for debridement the following week. Two days later, Mr. S has a witnessed seizure; his VPA level is 9 mcg/mL.

A common question asked of pharmacists is, “Will protein binding changes affect drug dosages?” In this article, I describe how protein binding changes may occur, and the complexity of the dynamic. Being highly bound to a protein typically does not mean all medications will interact, but some interactions can be important. This article does not cover medications that bind to hormones.

Why is protein binding important? When a medication is bound to plasma protein, it is not free to act. There can be a delay in therapeutic effect (because no drug is available to react), delayed elimination, or possibly displacement of another protein-bound medication. Additionally, medications tend not to cross the blood-brain barrier or be eliminated when bound. For example, if a drug is 99% bound (leaving 1% free) and displacement now leaves 2% of the drug free, this event has doubled the amount of free drug. As the unbound medication is eliminated, the drug that is bound to the protein can act as a reservoir. A dynamic relationship exists between bound drug, unbound drug, and rate of elimination.

Which proteins do drugs commonly bind to? The proteins often associated with binding include albumin, alpha-1-acid glycoprotein (AAG), and lipoproteins. Albumin comprises 60% of total plasma protein in the plasma. Lipoproteins include very high-density lipoprotein (VHDL), high-density lipoprotein (HDL), very low-density lipoprotein (VLDL), and low-density lipoprotein (LDL).¹ Medications that bind to lipoproteins include cyclosporine, tacrolimus, and propofol.²

Continued to: What common disease states can cause hypoalbuminemia?

 

What common disease states can cause hypoalbuminemia? Many disease states can result in low albumin levels. The most common ones are malnutrition, malignancies, stress, injury, burns, pregnancy, and diabetes.³ When there is less albumin to bind to, free drug levels may be increased.

Can AAG levels change with disease states as well? Because AAG accounts for a lower percentage of total plasma protein than albumin, there may be less clinical concern regarding AAG. AAG levels usually do not drop, but instead can become elevated during times of trauma, inflammation, and acute myocardial infarction. This could result in increased binding of the free drug.⁴Which medications bind to red blood cells (RBCs)? There are several locations for drugs to bind to RBCs, including to hemoglobin and the plasma membrane. Medications that commonly bind to RBCs include barbiturates, chlorpromazine, imipramine, and phenytoin.⁵

What are common highly-bound medications? The Table¹ provides examples of medications that are >90% protein-bound. However, this information may be misleading because many medications are highly bound. Zhang et al¹ compiled binding data for 222 drugs, half of which bind 90% to 100%. However, the literature does not indicate that they all have clinically significant interactions. Benet and Hoener⁶ discuss how factors other than protein binding affect potential drug interactions, and the complexity of the body’s ability to compensate for increased free drug. Medication characteristics that may contribute to producing a significant interaction include, but are not limited to:

• free vs protein-bound drug in the plasma or tissue
• volume of distribution
• organs affected
• hepatic bioavailability
• drug clearance.

For example, VPA is 93% protein-bound and phenytoin is 91% protein-bound.¹ However, this interaction is affected by more than just protein binding. VPA not only displaces the protein-bound phenytoin, but also inhibits its metabolism, which together result in increased free phenytoin levels.

Continued to: Another area of concern is a critically ill patient...

 

Another area of concern is a critically ill patient who has a change in his or her pH. Medications that are highly bound and have high clearance rates may be affected. This is of particular concern when prescribing antibiotics that are time-dependent, such as beta-lactams.³

What happened to Mr. S? Mr. S likely experienced a drug–drug interaction that resulted in a subtherapeutic VPA level and subsequent seizure. Case reports have shown evidence that the carbapenem class of antibiotics, which includes ertapenem, interacts with VPA.⁷ Proposed mechanisms include a lowering of VPA serum levels due to a redistribution of the VPA onto the RBCs due to carbapenem. Other theories include the possibility that carbapenems may limit oral VPA absorption, decrease VPA enterohepatic recirculation, and increase VPA metabolism.⁷ Using VPA and ertapenem together is discouraged because seizures have been reported among patients receiving this combination. If it is medically necessary to administer VPA and ertapenem, closely monitor VPA levels. In Mr. S’s case, another broad-spectrum antibiotic, such as piperacillin-tazobactam, could have been used, for his diabetic foot infection.

While many medications may have high protein binding, there are few clinically important known interactions. However, our understanding of the relationship between protein binding and drug interactions may improve with additional research.

CASE CONTINUED

Under neurology’s care, lacosamide is added for treatment of Mr. S’s seizures. No more seizures are noted during the remainder of his hospitalization. Infectious disease services change his antibiotic to piperacillin-tazobactam. Mr. S continues to progress well and is discharged to a rehabilitation center 2 days later.

Related Resource

• DrugBank. www.drugbank.ca. Canadian Institutes of Health Research.

Drug Brand Names
Amiodarone • Cordarone, Pacerone
Bumetanide • Bumex
Bupivacaine • Marcaine, Sensorcaine
Buprenorphine • Belbuca, Subutex
Ceftriaxone • Rocephin
Chlordiazepoxide • Librium
Chlorpromazine • Thorazine
Clozapine • Clozaril
Cyclosporine • Gengraf, Neoral
Diazepam • Valium
Doxycycline • Acticlate, Doryx
Duloxetine • Cymbalta
Ertapenem • Invanz
Fluoxetine • Prozac, Sarafem
Furosemide • Lasix
Glargine (Insulin) • Lantus, Toujeo
Glipizide • Glucotrol
Haloperidol • Haldol
Ibuprofen • Advil, Motrin
Imipramine • Tofranil
Lacosamide • Vimpat
Lisinopril • Prinivil, Zestril
Lorazepam • Ativan
Nicardipine • Cardene
Nortriptyline • Pamelor
Paclitaxel • Abraxane, Taxol
Phenytoin • Dilantin, Phenytek
Piperacillin-tazobactam • Zosyn
Propofol • Diprivan
Sertraline • Zoloft
Tacrolimus • Prograf
Tamoxifen • Soltamox
Valproic acid • Depakene, Depakote
Verapamil • Calan, Verelan
Warfarin • Coumadin, Jantoven

CASE Suicidal ideation, flare-up of ulcerative colitis

Mr. J, age 56, who has a history of major depressive disorder (MDD), generalized anxiety disorder (GAD), and ulcerative colitis (UC), presents to the emergency department (ED) with suicidal ideation and a plan to overdose on his medications. He reports no current emotional or financial stressors in his personal life. Home medications documented at the time of his arrival to the ED include sertraline, 100 mg/d, bupropion, 150 mg/d, buspirone, 10 mg 3 times daily, diazepam 10 mg 3 times daily, as needed, adalimumab, 40 mg IM every 2 weeks, and diphenhydramine, 50 mg every night.

A recent flare-up of UC resulted in Mr. J being placed on a 15-week prednisone taper, beginning at 80 mg/d and decreasing by 5 mg weekly, which was completed 2 weeks before he presented to the ED. After completing the prednisone taper, Mr. J went to his primary care physician (PCP) on 3 separate occasions due to episodes of severe depression. Although the PCP prescribed multiple medications to target Mr. J’s depressive symptoms, he continued to decline.

Subsequently, Mr. J came to the ED and is admitted to the psychiatric unit for safety and stabilization. Upon admission, Mr. J becomes bedridden, and reports that his current depressive episode is the most severe that he has ever experienced in his more than 30 years of having MDD. He says that neither bupropion nor buspirone are helping with his depression, anxiety, or any related symptom.

[polldaddy:10120537]

The authors’ observations

At admission, all of Mr. J’s home medications, except sertraline and adalimumab, which had been prescribed to treat UC (Box^1,2), were discontinued. His diazepam was discontinued because the clinician felt it may have been contributing to Mr. J’s inability to walk or get out of bed. Diazepam was not tapered because it was initiated 7 days prior to admission and was thought to be exacerbating his depression and suicidal ideation. Bupropion and buspirone, which were initiated 2 weeks prior, were discontinued because Mr. J reported that neither medication was helping with his depression, anxiety, or any related symptom.

 

EVALUATION Poor appetite, anxiety, and continued suicidality

During evaluation, vital signs, laboratory findings, and diagnostic testing are found to be unremarkable. Mr. J’s presentation and complaints are entirely subjective, and include poor appetite, fatigue, difficulty sleeping, sorrow, anxiety, and continued suicidality. Mr. J reports that he feels miserable, which is reflected by his poor eye contact, soft speech, and body language.

Continued to: The authors' observations

 

 

The authors’ observations

MDD is a mood disorder characterized by depressed mood and/or loss of interest or pleasure for more than 2 weeks.³ First-line pharmacotherapy for MDD includes monotherapy with a selective serotonin reuptake inhibitor (SSRI), serotonin-norepinephrine reuptake inhibitor (SNRI), mirtazapine, or bupropion.⁴ Medication selection is typically based on patient-specific factors, adverse effect profile, drug–drug interactions, and cost. Other treatments include electroconvulsive therapy (ECT) or cognitive-behavioral therapy (CBT).^4,5 Augmentation agents, such as second-generation antipsychotics, lithium, thyroid hormone supplementation, buspirone, anticonvulsants, and combinations of antidepressants, may also be considered.⁴

TREATMENT Condition worsens

On Day 2 of hospitalization, Mr. J is started on aripiprazole, 5 mg/d, clonazepam, 1 mg twice daily, and melatonin, 5 mg, each night for sleep. Aripiprazole, 5 mg/d, is initiated as an adjunct to sertraline for MDD because Mr. J reports feeling much worse and continues to report that he would “rather die than feel this way.” Mr. J begins to believe that his current state is his new baseline, and that feeling better is no longer possible.

On Day 3 of hospitalization, records are obtained from a clinician at an outside facility who previously treated Mr. J; this clinician suspected Mr. J may have bipolar disorder. On Days 3 and 5 of hospitalization, aripiprazole is titrated to 10 mg/d, and then to 20 mg/d, respectively. On Day 6, sertraline is increased to 150 mg/d because Mr. J continues to report low mood and limited sleep and is less and less interactive during interviews. He remains suicidal, and because bipolar depression is suspected (although this is not a clear diagnosis in his records), a trial of divalproex sodium, 250 mg twice daily, is initiated on Day 6.

By Day 8 of hospitalization, there is no notable change in Mr. J’s depressive symptoms. On Day 9, sertraline is increased to 200 mg/d, with little improvement from Mr. J’s perspective. The multidisciplinary team evaluates him, and when directly asked, Mr. J cites his 4 greatest complaints to be poor sleep, fatigue, no appetite, and depressed mood. Once again, he states, “I would rather be dead than go on feeling this way.”

[polldaddy:10120587]

The authors’ observations

Due to Mr. J’s severe, unrelenting depressive episode, the treatment team obtained his informed consent to undergo ECT. On Day 9, before initiating ECT, the pharmacist recommended mirtazapine, even though the patient weighed almost 89 kg (196.21 lb) and had a body mass index of 27.8 kg/m². The treatment team thought that mirtazapine augmentation could potentially help the sertraline work more quickly while targeting Mr. J’s 4 greatest complaints.

Mirtazapine is a central alpha-2 antagonist or noradrenergic and specific serotonergic antidepressant (NaSSA) that works through antagonism of the presynaptic alpha-2 adrenergic receptors to indirectly regulate release of monoamines and increase the release of serotonin and norepinephrine.⁶ Additionally, mirtazapine has antagonist actions at 5HT2A, 5HT2C, 5HT3, and histamine-1 receptors.⁶ Potential adverse effects include drowsiness and increased appetite leading to weight gain.⁷ Mirtazapine’s therapeutic efficacy is similar to SSRIs for treating depression.⁴ Mirtazapine in combination with an SNRI has been referred to as “California rocket fuel” due to the theoretical pharmacologic synergy and resulting strong antidepressant action.⁶ It was hypothesized that similar effects could be seen by augmenting the SSRI sertraline with mirtazapine.

Continued to: The time to efficacy with mirtazapine...

 

The time to efficacy with mirtazapine is approximately 2 to 4 weeks, but anxiety symptoms and poor sleep or insomnia may improve in the first week.⁸ Studies have suggested the possibility of a more rapid onset of efficacy with mirtazapine than with SSRIs, as well as potential response acceleration in MDD and other psychiatric illnesses such as anxiety disorders or obsessive-compulsive disorder (OCD).^9,10 A review that included several double-blind studies and compared mirtazapine with SSRIs found the amount of responders with persistent improvement with onset in Week 1 was more pronounced with mirtazapine.⁹

Augmenting an SSRI with mirtazapine is a potential therapeutic option because it can help boost the efficacy of the prescribed SSRI while enhancing appetite and blunting the activating or anxiety-like effects of some SSRIs, which may help with relaxation and sleep.⁴ The combination of an SSRI plus mirtazapine has been studied in patients with MDD, posttraumatic stress disorder, and OCD; it was found to improve symptoms of those conditions due to the medications’ complementary mechanisms of action.^4,11-13 Also, mirtazapine has been shown to decrease the rates of relapse after an acute phase of depression.^4,14

OUTCOME Rapid improvement


On Day 9, Mr. J receives the first dose of mirtazapine, 7.5 mg at bedtime. On Day 10, when Mr. J wakes, his mood is notably improved. He is more interactive (sitting up in bed reading and making eye contact with the staff during an interview), and he reports improved sleep and eats most of his breakfast.

After receiving 3 doses of mirtazapine, Mr. J reports that he feels back to his normal self; he is interactive, alert, and eating well. Due to the rapid improvement in mood, ECT is discontinued, and he does not receive any ECT treatment during the remainder of his hospitalization.

On Day 11, divalproex is discontinued. Because Mr. J receives only 5 days of therapy with this agent, his divalproex level is not checked. At this point, the treatment team feels confident in ruling out bipolar disorder.

On Day 15, Mr. J is discharged with sertraline, 200 mg/d, mirtazapine, 7.5 mg/d at 7 pm, aripiprazole, 20 mg/d, clonazepam, 1 mg twice daily as needed for anxiety, melatonin 5 mg/d, and adalimumab, 40 mg IM every 2 weeks. Discharge instructions include a follow-up in 2 weeks to evaluate continuation strategies for the discharge medications.

Ten months after his depressive episode, Mr. J has had no further admissions at the hospital where he received the treatment described here.

 

Bottom Line

Evidence for the treatment of major depressive disorder induced by corticosteroid withdrawal is limited. Despite trials of agents from multiple medication classes, the depressive episode may not improve. Adding mirtazapine to a selective serotonin reuptake inhibitor or serotonin-norepinephrine reuptake inhibitor may prove successful.

Related Resources

• Watanabe N, Omori IM, Nakagawa A, et al. Mirtazapine versus other antidepressive agents for depression. Cochrane Database Syst Rev. 2011;(12):CD006528.
• Kenna HA, Poon AW, de los Angeles CP, et al. Psychiatric complications of treatment with corticosteroids: review with case report. Psychiatry Clin Neurosci. 2011;65(6):549-560.

Drug Brand Names
Adalimumab • Humira
Aripiprazole • Abilify
Bupropion • Wellbutrin, Zyban
Buspirone • Buspar
Clonazepam • Klonopin
Diazepam • Valium
Diphenhydramine • Benadryl
Divalproex • Depakote, Depakote ER
Lithium • Eskalith, Lithobid
Mirtazapine • Remeron
Prednisone • Deltasone
Sertraline • Zoloft

CASE Suicidal ideation, flare-up of ulcerative colitis

Mr. J, age 56, who has a history of major depressive disorder (MDD), generalized anxiety disorder (GAD), and ulcerative colitis (UC), presents to the emergency department (ED) with suicidal ideation and a plan to overdose on his medications. He reports no current emotional or financial stressors in his personal life. Home medications documented at the time of his arrival to the ED include sertraline, 100 mg/d, bupropion, 150 mg/d, buspirone, 10 mg 3 times daily, diazepam 10 mg 3 times daily, as needed, adalimumab, 40 mg IM every 2 weeks, and diphenhydramine, 50 mg every night.

A recent flare-up of UC resulted in Mr. J being placed on a 15-week prednisone taper, beginning at 80 mg/d and decreasing by 5 mg weekly, which was completed 2 weeks before he presented to the ED. After completing the prednisone taper, Mr. J went to his primary care physician (PCP) on 3 separate occasions due to episodes of severe depression. Although the PCP prescribed multiple medications to target Mr. J’s depressive symptoms, he continued to decline.

Subsequently, Mr. J came to the ED and is admitted to the psychiatric unit for safety and stabilization. Upon admission, Mr. J becomes bedridden, and reports that his current depressive episode is the most severe that he has ever experienced in his more than 30 years of having MDD. He says that neither bupropion nor buspirone are helping with his depression, anxiety, or any related symptom.

[polldaddy:10120537]

The authors’ observations

At admission, all of Mr. J’s home medications, except sertraline and adalimumab, which had been prescribed to treat UC (Box^1,2), were discontinued. His diazepam was discontinued because the clinician felt it may have been contributing to Mr. J’s inability to walk or get out of bed. Diazepam was not tapered because it was initiated 7 days prior to admission and was thought to be exacerbating his depression and suicidal ideation. Bupropion and buspirone, which were initiated 2 weeks prior, were discontinued because Mr. J reported that neither medication was helping with his depression, anxiety, or any related symptom.

 

EVALUATION Poor appetite, anxiety, and continued suicidality

During evaluation, vital signs, laboratory findings, and diagnostic testing are found to be unremarkable. Mr. J’s presentation and complaints are entirely subjective, and include poor appetite, fatigue, difficulty sleeping, sorrow, anxiety, and continued suicidality. Mr. J reports that he feels miserable, which is reflected by his poor eye contact, soft speech, and body language.

Continued to: The authors' observations

 

 

The authors’ observations

MDD is a mood disorder characterized by depressed mood and/or loss of interest or pleasure for more than 2 weeks.³ First-line pharmacotherapy for MDD includes monotherapy with a selective serotonin reuptake inhibitor (SSRI), serotonin-norepinephrine reuptake inhibitor (SNRI), mirtazapine, or bupropion.⁴ Medication selection is typically based on patient-specific factors, adverse effect profile, drug–drug interactions, and cost. Other treatments include electroconvulsive therapy (ECT) or cognitive-behavioral therapy (CBT).^4,5 Augmentation agents, such as second-generation antipsychotics, lithium, thyroid hormone supplementation, buspirone, anticonvulsants, and combinations of antidepressants, may also be considered.⁴

TREATMENT Condition worsens

On Day 2 of hospitalization, Mr. J is started on aripiprazole, 5 mg/d, clonazepam, 1 mg twice daily, and melatonin, 5 mg, each night for sleep. Aripiprazole, 5 mg/d, is initiated as an adjunct to sertraline for MDD because Mr. J reports feeling much worse and continues to report that he would “rather die than feel this way.” Mr. J begins to believe that his current state is his new baseline, and that feeling better is no longer possible.

On Day 3 of hospitalization, records are obtained from a clinician at an outside facility who previously treated Mr. J; this clinician suspected Mr. J may have bipolar disorder. On Days 3 and 5 of hospitalization, aripiprazole is titrated to 10 mg/d, and then to 20 mg/d, respectively. On Day 6, sertraline is increased to 150 mg/d because Mr. J continues to report low mood and limited sleep and is less and less interactive during interviews. He remains suicidal, and because bipolar depression is suspected (although this is not a clear diagnosis in his records), a trial of divalproex sodium, 250 mg twice daily, is initiated on Day 6.

By Day 8 of hospitalization, there is no notable change in Mr. J’s depressive symptoms. On Day 9, sertraline is increased to 200 mg/d, with little improvement from Mr. J’s perspective. The multidisciplinary team evaluates him, and when directly asked, Mr. J cites his 4 greatest complaints to be poor sleep, fatigue, no appetite, and depressed mood. Once again, he states, “I would rather be dead than go on feeling this way.”

[polldaddy:10120587]

The authors’ observations

Due to Mr. J’s severe, unrelenting depressive episode, the treatment team obtained his informed consent to undergo ECT. On Day 9, before initiating ECT, the pharmacist recommended mirtazapine, even though the patient weighed almost 89 kg (196.21 lb) and had a body mass index of 27.8 kg/m². The treatment team thought that mirtazapine augmentation could potentially help the sertraline work more quickly while targeting Mr. J’s 4 greatest complaints.

Mirtazapine is a central alpha-2 antagonist or noradrenergic and specific serotonergic antidepressant (NaSSA) that works through antagonism of the presynaptic alpha-2 adrenergic receptors to indirectly regulate release of monoamines and increase the release of serotonin and norepinephrine.⁶ Additionally, mirtazapine has antagonist actions at 5HT2A, 5HT2C, 5HT3, and histamine-1 receptors.⁶ Potential adverse effects include drowsiness and increased appetite leading to weight gain.⁷ Mirtazapine’s therapeutic efficacy is similar to SSRIs for treating depression.⁴ Mirtazapine in combination with an SNRI has been referred to as “California rocket fuel” due to the theoretical pharmacologic synergy and resulting strong antidepressant action.⁶ It was hypothesized that similar effects could be seen by augmenting the SSRI sertraline with mirtazapine.

Continued to: The time to efficacy with mirtazapine...

 

The time to efficacy with mirtazapine is approximately 2 to 4 weeks, but anxiety symptoms and poor sleep or insomnia may improve in the first week.⁸ Studies have suggested the possibility of a more rapid onset of efficacy with mirtazapine than with SSRIs, as well as potential response acceleration in MDD and other psychiatric illnesses such as anxiety disorders or obsessive-compulsive disorder (OCD).^9,10 A review that included several double-blind studies and compared mirtazapine with SSRIs found the amount of responders with persistent improvement with onset in Week 1 was more pronounced with mirtazapine.⁹

Augmenting an SSRI with mirtazapine is a potential therapeutic option because it can help boost the efficacy of the prescribed SSRI while enhancing appetite and blunting the activating or anxiety-like effects of some SSRIs, which may help with relaxation and sleep.⁴ The combination of an SSRI plus mirtazapine has been studied in patients with MDD, posttraumatic stress disorder, and OCD; it was found to improve symptoms of those conditions due to the medications’ complementary mechanisms of action.^4,11-13 Also, mirtazapine has been shown to decrease the rates of relapse after an acute phase of depression.^4,14

OUTCOME Rapid improvement


On Day 9, Mr. J receives the first dose of mirtazapine, 7.5 mg at bedtime. On Day 10, when Mr. J wakes, his mood is notably improved. He is more interactive (sitting up in bed reading and making eye contact with the staff during an interview), and he reports improved sleep and eats most of his breakfast.

After receiving 3 doses of mirtazapine, Mr. J reports that he feels back to his normal self; he is interactive, alert, and eating well. Due to the rapid improvement in mood, ECT is discontinued, and he does not receive any ECT treatment during the remainder of his hospitalization.

On Day 11, divalproex is discontinued. Because Mr. J receives only 5 days of therapy with this agent, his divalproex level is not checked. At this point, the treatment team feels confident in ruling out bipolar disorder.

On Day 15, Mr. J is discharged with sertraline, 200 mg/d, mirtazapine, 7.5 mg/d at 7 pm, aripiprazole, 20 mg/d, clonazepam, 1 mg twice daily as needed for anxiety, melatonin 5 mg/d, and adalimumab, 40 mg IM every 2 weeks. Discharge instructions include a follow-up in 2 weeks to evaluate continuation strategies for the discharge medications.

Ten months after his depressive episode, Mr. J has had no further admissions at the hospital where he received the treatment described here.

 

Bottom Line

Evidence for the treatment of major depressive disorder induced by corticosteroid withdrawal is limited. Despite trials of agents from multiple medication classes, the depressive episode may not improve. Adding mirtazapine to a selective serotonin reuptake inhibitor or serotonin-norepinephrine reuptake inhibitor may prove successful.

Related Resources

• Watanabe N, Omori IM, Nakagawa A, et al. Mirtazapine versus other antidepressive agents for depression. Cochrane Database Syst Rev. 2011;(12):CD006528.
• Kenna HA, Poon AW, de los Angeles CP, et al. Psychiatric complications of treatment with corticosteroids: review with case report. Psychiatry Clin Neurosci. 2011;65(6):549-560.

Drug Brand Names
Adalimumab • Humira
Aripiprazole • Abilify
Bupropion • Wellbutrin, Zyban
Buspirone • Buspar
Clonazepam • Klonopin
Diazepam • Valium
Diphenhydramine • Benadryl
Divalproex • Depakote, Depakote ER
Lithium • Eskalith, Lithobid
Mirtazapine • Remeron
Prednisone • Deltasone
Sertraline • Zoloft

CASE Suicidal ideation, flare-up of ulcerative colitis

Mr. J, age 56, who has a history of major depressive disorder (MDD), generalized anxiety disorder (GAD), and ulcerative colitis (UC), presents to the emergency department (ED) with suicidal ideation and a plan to overdose on his medications. He reports no current emotional or financial stressors in his personal life. Home medications documented at the time of his arrival to the ED include sertraline, 100 mg/d, bupropion, 150 mg/d, buspirone, 10 mg 3 times daily, diazepam 10 mg 3 times daily, as needed, adalimumab, 40 mg IM every 2 weeks, and diphenhydramine, 50 mg every night.

A recent flare-up of UC resulted in Mr. J being placed on a 15-week prednisone taper, beginning at 80 mg/d and decreasing by 5 mg weekly, which was completed 2 weeks before he presented to the ED. After completing the prednisone taper, Mr. J went to his primary care physician (PCP) on 3 separate occasions due to episodes of severe depression. Although the PCP prescribed multiple medications to target Mr. J’s depressive symptoms, he continued to decline.

Subsequently, Mr. J came to the ED and is admitted to the psychiatric unit for safety and stabilization. Upon admission, Mr. J becomes bedridden, and reports that his current depressive episode is the most severe that he has ever experienced in his more than 30 years of having MDD. He says that neither bupropion nor buspirone are helping with his depression, anxiety, or any related symptom.

[polldaddy:10120537]

The authors’ observations

At admission, all of Mr. J’s home medications, except sertraline and adalimumab, which had been prescribed to treat UC (Box^1,2), were discontinued. His diazepam was discontinued because the clinician felt it may have been contributing to Mr. J’s inability to walk or get out of bed. Diazepam was not tapered because it was initiated 7 days prior to admission and was thought to be exacerbating his depression and suicidal ideation. Bupropion and buspirone, which were initiated 2 weeks prior, were discontinued because Mr. J reported that neither medication was helping with his depression, anxiety, or any related symptom.

 

EVALUATION Poor appetite, anxiety, and continued suicidality

During evaluation, vital signs, laboratory findings, and diagnostic testing are found to be unremarkable. Mr. J’s presentation and complaints are entirely subjective, and include poor appetite, fatigue, difficulty sleeping, sorrow, anxiety, and continued suicidality. Mr. J reports that he feels miserable, which is reflected by his poor eye contact, soft speech, and body language.

Continued to: The authors' observations

 

 

The authors’ observations

MDD is a mood disorder characterized by depressed mood and/or loss of interest or pleasure for more than 2 weeks.³ First-line pharmacotherapy for MDD includes monotherapy with a selective serotonin reuptake inhibitor (SSRI), serotonin-norepinephrine reuptake inhibitor (SNRI), mirtazapine, or bupropion.⁴ Medication selection is typically based on patient-specific factors, adverse effect profile, drug–drug interactions, and cost. Other treatments include electroconvulsive therapy (ECT) or cognitive-behavioral therapy (CBT).^4,5 Augmentation agents, such as second-generation antipsychotics, lithium, thyroid hormone supplementation, buspirone, anticonvulsants, and combinations of antidepressants, may also be considered.⁴

TREATMENT Condition worsens

On Day 2 of hospitalization, Mr. J is started on aripiprazole, 5 mg/d, clonazepam, 1 mg twice daily, and melatonin, 5 mg, each night for sleep. Aripiprazole, 5 mg/d, is initiated as an adjunct to sertraline for MDD because Mr. J reports feeling much worse and continues to report that he would “rather die than feel this way.” Mr. J begins to believe that his current state is his new baseline, and that feeling better is no longer possible.

On Day 3 of hospitalization, records are obtained from a clinician at an outside facility who previously treated Mr. J; this clinician suspected Mr. J may have bipolar disorder. On Days 3 and 5 of hospitalization, aripiprazole is titrated to 10 mg/d, and then to 20 mg/d, respectively. On Day 6, sertraline is increased to 150 mg/d because Mr. J continues to report low mood and limited sleep and is less and less interactive during interviews. He remains suicidal, and because bipolar depression is suspected (although this is not a clear diagnosis in his records), a trial of divalproex sodium, 250 mg twice daily, is initiated on Day 6.

By Day 8 of hospitalization, there is no notable change in Mr. J’s depressive symptoms. On Day 9, sertraline is increased to 200 mg/d, with little improvement from Mr. J’s perspective. The multidisciplinary team evaluates him, and when directly asked, Mr. J cites his 4 greatest complaints to be poor sleep, fatigue, no appetite, and depressed mood. Once again, he states, “I would rather be dead than go on feeling this way.”

[polldaddy:10120587]

The authors’ observations

Due to Mr. J’s severe, unrelenting depressive episode, the treatment team obtained his informed consent to undergo ECT. On Day 9, before initiating ECT, the pharmacist recommended mirtazapine, even though the patient weighed almost 89 kg (196.21 lb) and had a body mass index of 27.8 kg/m². The treatment team thought that mirtazapine augmentation could potentially help the sertraline work more quickly while targeting Mr. J’s 4 greatest complaints.

Mirtazapine is a central alpha-2 antagonist or noradrenergic and specific serotonergic antidepressant (NaSSA) that works through antagonism of the presynaptic alpha-2 adrenergic receptors to indirectly regulate release of monoamines and increase the release of serotonin and norepinephrine.⁶ Additionally, mirtazapine has antagonist actions at 5HT2A, 5HT2C, 5HT3, and histamine-1 receptors.⁶ Potential adverse effects include drowsiness and increased appetite leading to weight gain.⁷ Mirtazapine’s therapeutic efficacy is similar to SSRIs for treating depression.⁴ Mirtazapine in combination with an SNRI has been referred to as “California rocket fuel” due to the theoretical pharmacologic synergy and resulting strong antidepressant action.⁶ It was hypothesized that similar effects could be seen by augmenting the SSRI sertraline with mirtazapine.

Continued to: The time to efficacy with mirtazapine...

 

The time to efficacy with mirtazapine is approximately 2 to 4 weeks, but anxiety symptoms and poor sleep or insomnia may improve in the first week.⁸ Studies have suggested the possibility of a more rapid onset of efficacy with mirtazapine than with SSRIs, as well as potential response acceleration in MDD and other psychiatric illnesses such as anxiety disorders or obsessive-compulsive disorder (OCD).^9,10 A review that included several double-blind studies and compared mirtazapine with SSRIs found the amount of responders with persistent improvement with onset in Week 1 was more pronounced with mirtazapine.⁹

Augmenting an SSRI with mirtazapine is a potential therapeutic option because it can help boost the efficacy of the prescribed SSRI while enhancing appetite and blunting the activating or anxiety-like effects of some SSRIs, which may help with relaxation and sleep.⁴ The combination of an SSRI plus mirtazapine has been studied in patients with MDD, posttraumatic stress disorder, and OCD; it was found to improve symptoms of those conditions due to the medications’ complementary mechanisms of action.^4,11-13 Also, mirtazapine has been shown to decrease the rates of relapse after an acute phase of depression.^4,14

OUTCOME Rapid improvement


On Day 9, Mr. J receives the first dose of mirtazapine, 7.5 mg at bedtime. On Day 10, when Mr. J wakes, his mood is notably improved. He is more interactive (sitting up in bed reading and making eye contact with the staff during an interview), and he reports improved sleep and eats most of his breakfast.

After receiving 3 doses of mirtazapine, Mr. J reports that he feels back to his normal self; he is interactive, alert, and eating well. Due to the rapid improvement in mood, ECT is discontinued, and he does not receive any ECT treatment during the remainder of his hospitalization.

On Day 11, divalproex is discontinued. Because Mr. J receives only 5 days of therapy with this agent, his divalproex level is not checked. At this point, the treatment team feels confident in ruling out bipolar disorder.

On Day 15, Mr. J is discharged with sertraline, 200 mg/d, mirtazapine, 7.5 mg/d at 7 pm, aripiprazole, 20 mg/d, clonazepam, 1 mg twice daily as needed for anxiety, melatonin 5 mg/d, and adalimumab, 40 mg IM every 2 weeks. Discharge instructions include a follow-up in 2 weeks to evaluate continuation strategies for the discharge medications.

Ten months after his depressive episode, Mr. J has had no further admissions at the hospital where he received the treatment described here.

 

Bottom Line

Evidence for the treatment of major depressive disorder induced by corticosteroid withdrawal is limited. Despite trials of agents from multiple medication classes, the depressive episode may not improve. Adding mirtazapine to a selective serotonin reuptake inhibitor or serotonin-norepinephrine reuptake inhibitor may prove successful.

Related Resources

• Watanabe N, Omori IM, Nakagawa A, et al. Mirtazapine versus other antidepressive agents for depression. Cochrane Database Syst Rev. 2011;(12):CD006528.
• Kenna HA, Poon AW, de los Angeles CP, et al. Psychiatric complications of treatment with corticosteroids: review with case report. Psychiatry Clin Neurosci. 2011;65(6):549-560.

Drug Brand Names
Adalimumab • Humira
Aripiprazole • Abilify
Bupropion • Wellbutrin, Zyban
Buspirone • Buspar
Clonazepam • Klonopin
Diazepam • Valium
Diphenhydramine • Benadryl
Divalproex • Depakote, Depakote ER
Lithium • Eskalith, Lithobid
Mirtazapine • Remeron
Prednisone • Deltasone
Sertraline • Zoloft

Although antipsychotics have revolutionized the treatment of severe mental illnesses, adverse effects often present a substantial obstacle to adherence. One of the most tenacious and difficult-to-treat adverse effects is tardive dyskinesia (TD), a neuromotor syndrome with characteristic involuntary repetitive movements, typically of the muscles of the jaw, lips, and tongue. In addition to spasms and grimacing, patients can have choreoathetoid movements of the neck. In more extreme presentations, some patients can have difficulty breathing. TD is a largely irreversible condition. It is often a disfiguring lifelong disability that can further stigmatize patients who already suffer scorn and derision. TD usually has a delayed onset after a patient is started on an antipsychotic.¹ The syndrome is more commonly associated with first-generation antipsychotics, but affects up to 20% of patients who are treated with second-generation antipsychotics.¹ In the United States, TD affects as many as 500,000 patients.¹

There are several palliative interventions for TD, but the evidence for a consistently reliable treatment is weak. Branched-chain amino acids, ginkgo biloba, melatonin, and vitamin E have been investigated as interventions. Other approaches include switching to an alternate antipsychotic such as clozapine, adjusting the antipsychotic dose, using anticholinergic medications, adjunctive amantadine, gamma aminobutyric acid agonists, or adding tetrabenazine.

The FDA recently approved two vesicular monoamine transporter 2 (VMAT2) inhibitors, deutetrabenazine and valbenazine, for addressing symptoms of TD. However, these medications can cost tens of thousands of dollars per year, and also carry the risk of adverse effects such as sedation, akathisia, urinary retention, constipation, and muscle pain.² When treating a patient who develops TD, one might consider other potentially effective therapies with low adverse effect profiles that may be more cost-effective than existing treatments. The bioactive form of vitamin B6 (pyridoxine), pyridoxal-5-phosphate, has been used to treat various antipsychotic-induced movement disorders. Preliminary evidence suggests that vitamin B6 may help reduce the symptoms of TD.

A recent Cochrane Database Review (2015)³ of pyridoxal-5-phosphate treatment for TD found a significant improvement in symptoms compared with placebo. Although the studies included in this review were limited by modest sample sizes and short follow-up periods, 2 of the investigations revealed improvements of >40% in extrapyramidal symptoms with vitamin B6 compared with placebo. Lerner et al (2001)⁴ conducted a randomized, double-blind, placebo-controlled crossover trial in which 15 inpatients with schizophrenia who met the criteria for TD were assigned to vitamin B6, 400 mg/d, or placebo for 4 weeks. After a 2-week washout period, the placebo group was given vitamin B6 and vice versa. Compared with placebo, mean scores on the parkinsonism and dyskinetic movement subscales of the Extrapyramidal Symptom Rating Scale were significantly better in the third week of treatment with vitamin B6.

Lerner et al (2007)⁵ later conducted a separate crossover study using the same design with a washout period. This trial included a larger sample size (50 inpatients with DSM-IV diagnoses of schizophrenia or schizoaffective disorder and TD) and the dosage of vitamin B6 was increased to 1,200 mg/d over 26 weeks. Patients who received vitamin B6 experienced a significantly greater decrease in Extrapyramidal Symptom Rating Scale scores compared with those in the placebo group.

Continued to: A 29-year-old woman with treatment-resistant schizophrenia...

 

Umar et al (2016)⁶ published a case review of a 29-year-old woman with treatment-resistant schizophrenia with TD who was treated with clozapine, 400 mg/d. She was started on vitamin B6, 450 mg/d, for 4 weeks, and then her dose was increased to 600 mg/d. At 6 months, she experienced a 78% reduction in the severity of her TD symptoms, as measured by the Abnormal Involuntary Movement Scale. The authors reported that this improvement was maintained for 1 year after vitamin B6 was stopped.

Miodownik et al (2008)⁷ reported in a study of 89 patients with schizophrenia that those with TD (n = 40) had diminished amounts of vitamin B6 in their plasma compared with patients without symptoms of motor disturbances (n = 49).

Vitamin B6 has been known to improve other psychotropic-induced movement disorders. In a study of lithium-induced tremors, treatment with pyridoxine, 900 to 1,200 mg/d, resulted in “impressive improvement until total disappearance of tremor.”⁸ Lerner et al (2004)⁹ also reported significant improvement for patients with neuroleptic-induced akathisia who were treated with vitamin B6.

Some proposed mechanisms of action

Pyridoxal-5-phosphate is a coenzyme in the synthesis of dopamine and other neuro­transmitters. This might explain in part the biochemical mechanism of vitamin B6 in attenuating motor symptoms following long-term dopamine blockade. Chronic neurotransmitter antagonism may result in an upregulation of dopamine receptors in response. This compensatory reaction might create a dopamine receptor super-sensitivity in the nigrostriatal pathways.¹⁰

Another potential mechanism of action might be vitamin B6’s potent antioxidant properties and its scavenging of free radicals. The neurotoxicity of oxidative stress has been implicated in various movement disorders and psychiatric conditions.

In all of the studies described here, patients continued to receive daily antipsychotic treatment. In these trials, the adverse effects of vitamin B6 were minimal or negligible. In one study, vitamin B6 was reported to have had a better adverse effect profile than placebo.⁴

Although antipsychotics have revolutionized the treatment of severe mental illnesses, adverse effects often present a substantial obstacle to adherence. One of the most tenacious and difficult-to-treat adverse effects is tardive dyskinesia (TD), a neuromotor syndrome with characteristic involuntary repetitive movements, typically of the muscles of the jaw, lips, and tongue. In addition to spasms and grimacing, patients can have choreoathetoid movements of the neck. In more extreme presentations, some patients can have difficulty breathing. TD is a largely irreversible condition. It is often a disfiguring lifelong disability that can further stigmatize patients who already suffer scorn and derision. TD usually has a delayed onset after a patient is started on an antipsychotic.¹ The syndrome is more commonly associated with first-generation antipsychotics, but affects up to 20% of patients who are treated with second-generation antipsychotics.¹ In the United States, TD affects as many as 500,000 patients.¹

There are several palliative interventions for TD, but the evidence for a consistently reliable treatment is weak. Branched-chain amino acids, ginkgo biloba, melatonin, and vitamin E have been investigated as interventions. Other approaches include switching to an alternate antipsychotic such as clozapine, adjusting the antipsychotic dose, using anticholinergic medications, adjunctive amantadine, gamma aminobutyric acid agonists, or adding tetrabenazine.

The FDA recently approved two vesicular monoamine transporter 2 (VMAT2) inhibitors, deutetrabenazine and valbenazine, for addressing symptoms of TD. However, these medications can cost tens of thousands of dollars per year, and also carry the risk of adverse effects such as sedation, akathisia, urinary retention, constipation, and muscle pain.² When treating a patient who develops TD, one might consider other potentially effective therapies with low adverse effect profiles that may be more cost-effective than existing treatments. The bioactive form of vitamin B6 (pyridoxine), pyridoxal-5-phosphate, has been used to treat various antipsychotic-induced movement disorders. Preliminary evidence suggests that vitamin B6 may help reduce the symptoms of TD.

A recent Cochrane Database Review (2015)³ of pyridoxal-5-phosphate treatment for TD found a significant improvement in symptoms compared with placebo. Although the studies included in this review were limited by modest sample sizes and short follow-up periods, 2 of the investigations revealed improvements of >40% in extrapyramidal symptoms with vitamin B6 compared with placebo. Lerner et al (2001)⁴ conducted a randomized, double-blind, placebo-controlled crossover trial in which 15 inpatients with schizophrenia who met the criteria for TD were assigned to vitamin B6, 400 mg/d, or placebo for 4 weeks. After a 2-week washout period, the placebo group was given vitamin B6 and vice versa. Compared with placebo, mean scores on the parkinsonism and dyskinetic movement subscales of the Extrapyramidal Symptom Rating Scale were significantly better in the third week of treatment with vitamin B6.

Lerner et al (2007)⁵ later conducted a separate crossover study using the same design with a washout period. This trial included a larger sample size (50 inpatients with DSM-IV diagnoses of schizophrenia or schizoaffective disorder and TD) and the dosage of vitamin B6 was increased to 1,200 mg/d over 26 weeks. Patients who received vitamin B6 experienced a significantly greater decrease in Extrapyramidal Symptom Rating Scale scores compared with those in the placebo group.

Continued to: A 29-year-old woman with treatment-resistant schizophrenia...

 

Umar et al (2016)⁶ published a case review of a 29-year-old woman with treatment-resistant schizophrenia with TD who was treated with clozapine, 400 mg/d. She was started on vitamin B6, 450 mg/d, for 4 weeks, and then her dose was increased to 600 mg/d. At 6 months, she experienced a 78% reduction in the severity of her TD symptoms, as measured by the Abnormal Involuntary Movement Scale. The authors reported that this improvement was maintained for 1 year after vitamin B6 was stopped.

Miodownik et al (2008)⁷ reported in a study of 89 patients with schizophrenia that those with TD (n = 40) had diminished amounts of vitamin B6 in their plasma compared with patients without symptoms of motor disturbances (n = 49).

Vitamin B6 has been known to improve other psychotropic-induced movement disorders. In a study of lithium-induced tremors, treatment with pyridoxine, 900 to 1,200 mg/d, resulted in “impressive improvement until total disappearance of tremor.”⁸ Lerner et al (2004)⁹ also reported significant improvement for patients with neuroleptic-induced akathisia who were treated with vitamin B6.

Some proposed mechanisms of action

Pyridoxal-5-phosphate is a coenzyme in the synthesis of dopamine and other neuro­transmitters. This might explain in part the biochemical mechanism of vitamin B6 in attenuating motor symptoms following long-term dopamine blockade. Chronic neurotransmitter antagonism may result in an upregulation of dopamine receptors in response. This compensatory reaction might create a dopamine receptor super-sensitivity in the nigrostriatal pathways.¹⁰

Another potential mechanism of action might be vitamin B6’s potent antioxidant properties and its scavenging of free radicals. The neurotoxicity of oxidative stress has been implicated in various movement disorders and psychiatric conditions.

In all of the studies described here, patients continued to receive daily antipsychotic treatment. In these trials, the adverse effects of vitamin B6 were minimal or negligible. In one study, vitamin B6 was reported to have had a better adverse effect profile than placebo.⁴

Although antipsychotics have revolutionized the treatment of severe mental illnesses, adverse effects often present a substantial obstacle to adherence. One of the most tenacious and difficult-to-treat adverse effects is tardive dyskinesia (TD), a neuromotor syndrome with characteristic involuntary repetitive movements, typically of the muscles of the jaw, lips, and tongue. In addition to spasms and grimacing, patients can have choreoathetoid movements of the neck. In more extreme presentations, some patients can have difficulty breathing. TD is a largely irreversible condition. It is often a disfiguring lifelong disability that can further stigmatize patients who already suffer scorn and derision. TD usually has a delayed onset after a patient is started on an antipsychotic.¹ The syndrome is more commonly associated with first-generation antipsychotics, but affects up to 20% of patients who are treated with second-generation antipsychotics.¹ In the United States, TD affects as many as 500,000 patients.¹

There are several palliative interventions for TD, but the evidence for a consistently reliable treatment is weak. Branched-chain amino acids, ginkgo biloba, melatonin, and vitamin E have been investigated as interventions. Other approaches include switching to an alternate antipsychotic such as clozapine, adjusting the antipsychotic dose, using anticholinergic medications, adjunctive amantadine, gamma aminobutyric acid agonists, or adding tetrabenazine.

The FDA recently approved two vesicular monoamine transporter 2 (VMAT2) inhibitors, deutetrabenazine and valbenazine, for addressing symptoms of TD. However, these medications can cost tens of thousands of dollars per year, and also carry the risk of adverse effects such as sedation, akathisia, urinary retention, constipation, and muscle pain.² When treating a patient who develops TD, one might consider other potentially effective therapies with low adverse effect profiles that may be more cost-effective than existing treatments. The bioactive form of vitamin B6 (pyridoxine), pyridoxal-5-phosphate, has been used to treat various antipsychotic-induced movement disorders. Preliminary evidence suggests that vitamin B6 may help reduce the symptoms of TD.

A recent Cochrane Database Review (2015)³ of pyridoxal-5-phosphate treatment for TD found a significant improvement in symptoms compared with placebo. Although the studies included in this review were limited by modest sample sizes and short follow-up periods, 2 of the investigations revealed improvements of >40% in extrapyramidal symptoms with vitamin B6 compared with placebo. Lerner et al (2001)⁴ conducted a randomized, double-blind, placebo-controlled crossover trial in which 15 inpatients with schizophrenia who met the criteria for TD were assigned to vitamin B6, 400 mg/d, or placebo for 4 weeks. After a 2-week washout period, the placebo group was given vitamin B6 and vice versa. Compared with placebo, mean scores on the parkinsonism and dyskinetic movement subscales of the Extrapyramidal Symptom Rating Scale were significantly better in the third week of treatment with vitamin B6.

Lerner et al (2007)⁵ later conducted a separate crossover study using the same design with a washout period. This trial included a larger sample size (50 inpatients with DSM-IV diagnoses of schizophrenia or schizoaffective disorder and TD) and the dosage of vitamin B6 was increased to 1,200 mg/d over 26 weeks. Patients who received vitamin B6 experienced a significantly greater decrease in Extrapyramidal Symptom Rating Scale scores compared with those in the placebo group.

Continued to: A 29-year-old woman with treatment-resistant schizophrenia...

 

Umar et al (2016)⁶ published a case review of a 29-year-old woman with treatment-resistant schizophrenia with TD who was treated with clozapine, 400 mg/d. She was started on vitamin B6, 450 mg/d, for 4 weeks, and then her dose was increased to 600 mg/d. At 6 months, she experienced a 78% reduction in the severity of her TD symptoms, as measured by the Abnormal Involuntary Movement Scale. The authors reported that this improvement was maintained for 1 year after vitamin B6 was stopped.

Miodownik et al (2008)⁷ reported in a study of 89 patients with schizophrenia that those with TD (n = 40) had diminished amounts of vitamin B6 in their plasma compared with patients without symptoms of motor disturbances (n = 49).

Vitamin B6 has been known to improve other psychotropic-induced movement disorders. In a study of lithium-induced tremors, treatment with pyridoxine, 900 to 1,200 mg/d, resulted in “impressive improvement until total disappearance of tremor.”⁸ Lerner et al (2004)⁹ also reported significant improvement for patients with neuroleptic-induced akathisia who were treated with vitamin B6.

Some proposed mechanisms of action

Pyridoxal-5-phosphate is a coenzyme in the synthesis of dopamine and other neuro­transmitters. This might explain in part the biochemical mechanism of vitamin B6 in attenuating motor symptoms following long-term dopamine blockade. Chronic neurotransmitter antagonism may result in an upregulation of dopamine receptors in response. This compensatory reaction might create a dopamine receptor super-sensitivity in the nigrostriatal pathways.¹⁰

Another potential mechanism of action might be vitamin B6’s potent antioxidant properties and its scavenging of free radicals. The neurotoxicity of oxidative stress has been implicated in various movement disorders and psychiatric conditions.

In all of the studies described here, patients continued to receive daily antipsychotic treatment. In these trials, the adverse effects of vitamin B6 were minimal or negligible. In one study, vitamin B6 was reported to have had a better adverse effect profile than placebo.⁴

During my residency training, I was trained in the standard “reactive” psychiatric consultation model. In this system, I would see consults placed by the primary team after they identified a behavioral issue in a patient. As a trainee, I experienced frequent frustrations working in this model: Consults that are discharge-dependent (“Can you see the patient before he is discharged this morning?”), consults for acute behavioral dysregulation (“The patient is near the elevator, can you come see him ASAP?”), or consults for consequences of poor management of alcohol/benzodiazepine withdrawal (“The patient is confused and trying to leave”).

As a fellow in consultation-liaison (C-L) psychiatry, I was introduced to the “proactive” consultation model, which avoids some of these issues. In this article, which is intended for residents who have not been exposed to this new approach, I explain how the proactive model changes our experience as C-L clinicians.

The Behavioral Intervention Team

At Yale New Haven Hospital, the Behavioral Intervention Team (BIT) is a proactive, multidisciplinary psychiatric consultation service that serves the internal medicine units at the hospital. The team consists of nurse practitioners, nurse liaison specialists, social workers, and psychiatrists. The team identifies and removes behavioral barriers in the care of hospitalized mentally ill patients.

The BIT collaborates closely with the medical team through formal and informal consultation; co-management of behavioral issues; education of medical, nursing, and social work staff; and direct care of complex patients with behavioral disorders. The BIT assists the medical team with transitions to appropriate outpatient and inpatient psychiatric care. The team also manages the relationship with the insurer when a patient requires a stay in a psychiatric unit.

This model has a critical financial benefit in reducing the length of stay, but it also has many other benefits. It focuses on early recognition and treatment, and helps mitigate the effects of mental or substance use disorders on patients’ recovery. BIT members educate their peers regarding management of a multitude of behavioral issues. This fosters extensive informal collaboration (“curbside consultation”), which helps patients who did not receive a formal consult. The model distributes work more rationally among different professional specialists. It yields a relationship with medical teams that is not only more effective, but also more enjoyable. In the BIT model, psychiatrists pick the cases where they feel they can have the most impact, and avoid the cases they feel they cannot have any.^1-3

CASE A better approach to alcohol withdrawal

Mr. X, age 56, has a history of alcohol use disorder, hypertension, and coronary artery disease. He’s had multiple past admissions for complicated alcohol withdrawal. He is transferred from a local community hospital, where he had presented with chest pain. His last drink was 2 days prior to admission, and his blood alcohol level is <10 mg/dL.

During Mr. X’s previous hospitalizations, psychiatric consults were performed in the standard reactive model. The primary team initially prescribed an ineffective dosage of benzodiazepines for his alcohol withdrawal. This escalated his withdrawal into delirium tremens, after which psychiatry was involved. Due to this early ineffective management, the patient had a prolonged medical ICU stay and overall stay, experienced increased medical complications, and required increased staff resources because he was extremely agitated.

Continued to: During this hospitalization...

 

During this hospitalization, Mr. X arrives with similar medical complaints. The nurse practitioner on the BIT service, who screened all admissions each day, examines the prior notes (she finds the team sign-outs to be particularly useful). She suggests a psychiatric consult on Day 1 of the admission, which the primary medical team orders. The BIT nurse practitioner gives apt recommendations of evidence-based management, including a benzodiazepine taper, high-dose thiamine, and psychopharmacologic approaches to severe agitation. The nurse liaison specialist on the service makes behavioral plans for managing agitation, which she communicates to the nurses caring for Mr. X.

Because his withdrawal is managed more promptly, Mr. X’s length of stay is shorter and he does not experience any medical complications. The BIT social worker helps find appropriate aftercare options, including residential treatment and Alcoholics Anonymous meetings, to which the patient agrees.

Participating in this case was highly educational for me as a trainee. This case is but one example among many where proactive consultation provided prompt care, lowered the rate of complications, reduced length of stay, and resulted in greater provider satisfaction. The Table⁴ contrasts the proactive and reactive consultation models. The following 5 factors are critical in the proactive consultation model^4,5:

1. Standardized and reliable procedure screening of all admissions, involving a mental health professional, through record review and staff contact. This screening should identify patients with issues who will benefit specifically from in-hospital services, rather than just patients with any psychiatric issue. An electronic medical record is essential to efficient screening, team communication, and progress monitoring. Truly integrated consultation would be impossible with a paper chart.

Continued to: 2. Rapid intervention...

 

2. Rapid intervention that anticipates impending problems before a cascade of complications starts.

3. Collaborative engagement with the primary medical team, sharing the burden of caring for the complex inpatient, and transmitting critical behavioral management skills to all caregivers, including the skill of recognizing patients who can benefit from a psychiatric consultation.

4. Daily and close contact between behavioral and medical teams, ensuring that treatment recommendations are understood, enacted, and reinforced, ineffective treatments are discontinued, and new problems are addressed before complicating consequences arise. Dedicating specific personnel to specific hospital units and placing them in rounds simplifies communication and speeds intervention implementation.

5. A multidisciplinary consultation team, offering a range of responses, including informal curbside consultation, consultation with an advanced practice registered nurse, social work interventions, advice to discharge planning teams, psychological services, and access to specialized providers, such as addiction teams, as well as traditional consultation with an experienced psychiatrist.

Research has shown the effectiveness of proactive, embedded, multidisciplinary approaches.^1-3,5 It was a gratifying experience to work in this model. I worked intimately with medical clinicians, and shared the burden of responsibilities leading to optimal patient outcomes. The proactive consultation model truly re-emphasizes the “liaison” component of C-L psychiatry, as it was originally envisioned.

During my residency training, I was trained in the standard “reactive” psychiatric consultation model. In this system, I would see consults placed by the primary team after they identified a behavioral issue in a patient. As a trainee, I experienced frequent frustrations working in this model: Consults that are discharge-dependent (“Can you see the patient before he is discharged this morning?”), consults for acute behavioral dysregulation (“The patient is near the elevator, can you come see him ASAP?”), or consults for consequences of poor management of alcohol/benzodiazepine withdrawal (“The patient is confused and trying to leave”).

As a fellow in consultation-liaison (C-L) psychiatry, I was introduced to the “proactive” consultation model, which avoids some of these issues. In this article, which is intended for residents who have not been exposed to this new approach, I explain how the proactive model changes our experience as C-L clinicians.

The Behavioral Intervention Team

At Yale New Haven Hospital, the Behavioral Intervention Team (BIT) is a proactive, multidisciplinary psychiatric consultation service that serves the internal medicine units at the hospital. The team consists of nurse practitioners, nurse liaison specialists, social workers, and psychiatrists. The team identifies and removes behavioral barriers in the care of hospitalized mentally ill patients.

The BIT collaborates closely with the medical team through formal and informal consultation; co-management of behavioral issues; education of medical, nursing, and social work staff; and direct care of complex patients with behavioral disorders. The BIT assists the medical team with transitions to appropriate outpatient and inpatient psychiatric care. The team also manages the relationship with the insurer when a patient requires a stay in a psychiatric unit.

This model has a critical financial benefit in reducing the length of stay, but it also has many other benefits. It focuses on early recognition and treatment, and helps mitigate the effects of mental or substance use disorders on patients’ recovery. BIT members educate their peers regarding management of a multitude of behavioral issues. This fosters extensive informal collaboration (“curbside consultation”), which helps patients who did not receive a formal consult. The model distributes work more rationally among different professional specialists. It yields a relationship with medical teams that is not only more effective, but also more enjoyable. In the BIT model, psychiatrists pick the cases where they feel they can have the most impact, and avoid the cases they feel they cannot have any.^1-3

CASE A better approach to alcohol withdrawal

Mr. X, age 56, has a history of alcohol use disorder, hypertension, and coronary artery disease. He’s had multiple past admissions for complicated alcohol withdrawal. He is transferred from a local community hospital, where he had presented with chest pain. His last drink was 2 days prior to admission, and his blood alcohol level is <10 mg/dL.

During Mr. X’s previous hospitalizations, psychiatric consults were performed in the standard reactive model. The primary team initially prescribed an ineffective dosage of benzodiazepines for his alcohol withdrawal. This escalated his withdrawal into delirium tremens, after which psychiatry was involved. Due to this early ineffective management, the patient had a prolonged medical ICU stay and overall stay, experienced increased medical complications, and required increased staff resources because he was extremely agitated.

Continued to: During this hospitalization...

 

During this hospitalization, Mr. X arrives with similar medical complaints. The nurse practitioner on the BIT service, who screened all admissions each day, examines the prior notes (she finds the team sign-outs to be particularly useful). She suggests a psychiatric consult on Day 1 of the admission, which the primary medical team orders. The BIT nurse practitioner gives apt recommendations of evidence-based management, including a benzodiazepine taper, high-dose thiamine, and psychopharmacologic approaches to severe agitation. The nurse liaison specialist on the service makes behavioral plans for managing agitation, which she communicates to the nurses caring for Mr. X.

Because his withdrawal is managed more promptly, Mr. X’s length of stay is shorter and he does not experience any medical complications. The BIT social worker helps find appropriate aftercare options, including residential treatment and Alcoholics Anonymous meetings, to which the patient agrees.

Participating in this case was highly educational for me as a trainee. This case is but one example among many where proactive consultation provided prompt care, lowered the rate of complications, reduced length of stay, and resulted in greater provider satisfaction. The Table⁴ contrasts the proactive and reactive consultation models. The following 5 factors are critical in the proactive consultation model^4,5:

1. Standardized and reliable procedure screening of all admissions, involving a mental health professional, through record review and staff contact. This screening should identify patients with issues who will benefit specifically from in-hospital services, rather than just patients with any psychiatric issue. An electronic medical record is essential to efficient screening, team communication, and progress monitoring. Truly integrated consultation would be impossible with a paper chart.

Continued to: 2. Rapid intervention...

 

2. Rapid intervention that anticipates impending problems before a cascade of complications starts.

3. Collaborative engagement with the primary medical team, sharing the burden of caring for the complex inpatient, and transmitting critical behavioral management skills to all caregivers, including the skill of recognizing patients who can benefit from a psychiatric consultation.

4. Daily and close contact between behavioral and medical teams, ensuring that treatment recommendations are understood, enacted, and reinforced, ineffective treatments are discontinued, and new problems are addressed before complicating consequences arise. Dedicating specific personnel to specific hospital units and placing them in rounds simplifies communication and speeds intervention implementation.

5. A multidisciplinary consultation team, offering a range of responses, including informal curbside consultation, consultation with an advanced practice registered nurse, social work interventions, advice to discharge planning teams, psychological services, and access to specialized providers, such as addiction teams, as well as traditional consultation with an experienced psychiatrist.

Research has shown the effectiveness of proactive, embedded, multidisciplinary approaches.^1-3,5 It was a gratifying experience to work in this model. I worked intimately with medical clinicians, and shared the burden of responsibilities leading to optimal patient outcomes. The proactive consultation model truly re-emphasizes the “liaison” component of C-L psychiatry, as it was originally envisioned.

During my residency training, I was trained in the standard “reactive” psychiatric consultation model. In this system, I would see consults placed by the primary team after they identified a behavioral issue in a patient. As a trainee, I experienced frequent frustrations working in this model: Consults that are discharge-dependent (“Can you see the patient before he is discharged this morning?”), consults for acute behavioral dysregulation (“The patient is near the elevator, can you come see him ASAP?”), or consults for consequences of poor management of alcohol/benzodiazepine withdrawal (“The patient is confused and trying to leave”).

As a fellow in consultation-liaison (C-L) psychiatry, I was introduced to the “proactive” consultation model, which avoids some of these issues. In this article, which is intended for residents who have not been exposed to this new approach, I explain how the proactive model changes our experience as C-L clinicians.

The Behavioral Intervention Team

At Yale New Haven Hospital, the Behavioral Intervention Team (BIT) is a proactive, multidisciplinary psychiatric consultation service that serves the internal medicine units at the hospital. The team consists of nurse practitioners, nurse liaison specialists, social workers, and psychiatrists. The team identifies and removes behavioral barriers in the care of hospitalized mentally ill patients.

The BIT collaborates closely with the medical team through formal and informal consultation; co-management of behavioral issues; education of medical, nursing, and social work staff; and direct care of complex patients with behavioral disorders. The BIT assists the medical team with transitions to appropriate outpatient and inpatient psychiatric care. The team also manages the relationship with the insurer when a patient requires a stay in a psychiatric unit.

This model has a critical financial benefit in reducing the length of stay, but it also has many other benefits. It focuses on early recognition and treatment, and helps mitigate the effects of mental or substance use disorders on patients’ recovery. BIT members educate their peers regarding management of a multitude of behavioral issues. This fosters extensive informal collaboration (“curbside consultation”), which helps patients who did not receive a formal consult. The model distributes work more rationally among different professional specialists. It yields a relationship with medical teams that is not only more effective, but also more enjoyable. In the BIT model, psychiatrists pick the cases where they feel they can have the most impact, and avoid the cases they feel they cannot have any.^1-3

CASE A better approach to alcohol withdrawal

Mr. X, age 56, has a history of alcohol use disorder, hypertension, and coronary artery disease. He’s had multiple past admissions for complicated alcohol withdrawal. He is transferred from a local community hospital, where he had presented with chest pain. His last drink was 2 days prior to admission, and his blood alcohol level is <10 mg/dL.

During Mr. X’s previous hospitalizations, psychiatric consults were performed in the standard reactive model. The primary team initially prescribed an ineffective dosage of benzodiazepines for his alcohol withdrawal. This escalated his withdrawal into delirium tremens, after which psychiatry was involved. Due to this early ineffective management, the patient had a prolonged medical ICU stay and overall stay, experienced increased medical complications, and required increased staff resources because he was extremely agitated.

Continued to: During this hospitalization...

 

During this hospitalization, Mr. X arrives with similar medical complaints. The nurse practitioner on the BIT service, who screened all admissions each day, examines the prior notes (she finds the team sign-outs to be particularly useful). She suggests a psychiatric consult on Day 1 of the admission, which the primary medical team orders. The BIT nurse practitioner gives apt recommendations of evidence-based management, including a benzodiazepine taper, high-dose thiamine, and psychopharmacologic approaches to severe agitation. The nurse liaison specialist on the service makes behavioral plans for managing agitation, which she communicates to the nurses caring for Mr. X.

Because his withdrawal is managed more promptly, Mr. X’s length of stay is shorter and he does not experience any medical complications. The BIT social worker helps find appropriate aftercare options, including residential treatment and Alcoholics Anonymous meetings, to which the patient agrees.

Participating in this case was highly educational for me as a trainee. This case is but one example among many where proactive consultation provided prompt care, lowered the rate of complications, reduced length of stay, and resulted in greater provider satisfaction. The Table⁴ contrasts the proactive and reactive consultation models. The following 5 factors are critical in the proactive consultation model^4,5:

1. Standardized and reliable procedure screening of all admissions, involving a mental health professional, through record review and staff contact. This screening should identify patients with issues who will benefit specifically from in-hospital services, rather than just patients with any psychiatric issue. An electronic medical record is essential to efficient screening, team communication, and progress monitoring. Truly integrated consultation would be impossible with a paper chart.

Continued to: 2. Rapid intervention...

 

2. Rapid intervention that anticipates impending problems before a cascade of complications starts.

3. Collaborative engagement with the primary medical team, sharing the burden of caring for the complex inpatient, and transmitting critical behavioral management skills to all caregivers, including the skill of recognizing patients who can benefit from a psychiatric consultation.

4. Daily and close contact between behavioral and medical teams, ensuring that treatment recommendations are understood, enacted, and reinforced, ineffective treatments are discontinued, and new problems are addressed before complicating consequences arise. Dedicating specific personnel to specific hospital units and placing them in rounds simplifies communication and speeds intervention implementation.

5. A multidisciplinary consultation team, offering a range of responses, including informal curbside consultation, consultation with an advanced practice registered nurse, social work interventions, advice to discharge planning teams, psychological services, and access to specialized providers, such as addiction teams, as well as traditional consultation with an experienced psychiatrist.

Research has shown the effectiveness of proactive, embedded, multidisciplinary approaches.^1-3,5 It was a gratifying experience to work in this model. I worked intimately with medical clinicians, and shared the burden of responsibilities leading to optimal patient outcomes. The proactive consultation model truly re-emphasizes the “liaison” component of C-L psychiatry, as it was originally envisioned.

Bone marrow aspirate showing MM

New York—Three novel treatment strategies that target B-cell maturation antigen (BCMA) are showing promise in recent multiple myeloma (MM) clinical trials, according to Shaji K. Kumar, MD, of Mayo Clinic Cancer Center in Rochester, Minnesota.

The strategies include B-cell maturation antigen (BCMA) antibody-drug conjugate, BCMA-specific chimeric antigen receptor (CAR) T-cell therapies, and bispecific T-cell engagers (BiTEs).

“Clearly, there are a lot of exciting drugs that are currently in clinical trials, but these 3 platforms appear to be much more advanced than the others, and hopefully we will see that in the clinic in the near future,” Dr. Kumar told attendees at the NCCN 13th Annual Congress: Hematologic Malignancies.

BCMA is required for plasma cell survival and is broadly expressed on malignant plasma cells.

BCMA antibody-drug conjugate

The antibody-drug conjugate, GSK2857916, is a humanized IgG1 anti-BCMA antibody conjugated to a microtubule-disrupting agent. It produced an overall response rate of 67% at the 2 highest dose levels in 9 MM patients who had previously received multiple standard-of-care agents.

“Some of the responses were quite durable, lasting several months,” he said.

Now, GSK2857916 is being evaluated in a variety of different combinations, he said, including in an upcoming phase 2 study of the antibody-drug conjugate in combination with lenalidomide plus dexamethasone or bortezomib plus dexamethasone in patients with relapsed or refractory disease.

BCMA-specific CAR T-cell therapy

Some of the most “exciting” data with anti-BCMA CAR T-cell therapy in myeloma, according to Dr. Kumar, involves bb2121. bb2121 showed durable clinical responses in heavily pretreated patients, according to an ASH 2017 presentation.

“The overall response rate is quite significant,” Dr. Kumar said. He related a 94% rate of overall response that was even higher in patients treated with doses of 150 x 10⁶ CAR+ T cells or more. Many of the responses were lasting, he said, with 5 patients in ongoing response for more than a year.

“The results are exciting enough that this is actually moving forward with registration trials,” Dr. Kumar added.

Another novel CAR T-cell product, LCAR-B38M, has demonstrated promising results. LCAR-B38M principally targets BCMA and has led to a significant number of patients achieving stringent complete response that lasted beyond 1 year.

Multiple BCMA-targeting CAR T-cell products that use different vectors and different costimulatory molecules are currently in clinical trials, Dr. Kumar said.

BiTEs

In contrast to CAR T-cell products that must be customized to each patient in a process that takes weeks, BiTEs are a ready-made approach to allow T cells to engage with tumor cells.

“In patients with advanced disease, a lot can change in that short timeframe, so having an approach that is off-the-shelf, which is not patient specific, is quite attractive,” Dr. Kumar said.

BCMA-directed BiTE therapies to watch that are under investigation include AMG 420 and PF-06863135, he said. 

Bone marrow aspirate showing MM

New York—Three novel treatment strategies that target B-cell maturation antigen (BCMA) are showing promise in recent multiple myeloma (MM) clinical trials, according to Shaji K. Kumar, MD, of Mayo Clinic Cancer Center in Rochester, Minnesota.

The strategies include B-cell maturation antigen (BCMA) antibody-drug conjugate, BCMA-specific chimeric antigen receptor (CAR) T-cell therapies, and bispecific T-cell engagers (BiTEs).

“Clearly, there are a lot of exciting drugs that are currently in clinical trials, but these 3 platforms appear to be much more advanced than the others, and hopefully we will see that in the clinic in the near future,” Dr. Kumar told attendees at the NCCN 13th Annual Congress: Hematologic Malignancies.

BCMA is required for plasma cell survival and is broadly expressed on malignant plasma cells.

BCMA antibody-drug conjugate

The antibody-drug conjugate, GSK2857916, is a humanized IgG1 anti-BCMA antibody conjugated to a microtubule-disrupting agent. It produced an overall response rate of 67% at the 2 highest dose levels in 9 MM patients who had previously received multiple standard-of-care agents.

“Some of the responses were quite durable, lasting several months,” he said.

Now, GSK2857916 is being evaluated in a variety of different combinations, he said, including in an upcoming phase 2 study of the antibody-drug conjugate in combination with lenalidomide plus dexamethasone or bortezomib plus dexamethasone in patients with relapsed or refractory disease.

BCMA-specific CAR T-cell therapy

Some of the most “exciting” data with anti-BCMA CAR T-cell therapy in myeloma, according to Dr. Kumar, involves bb2121. bb2121 showed durable clinical responses in heavily pretreated patients, according to an ASH 2017 presentation.

“The overall response rate is quite significant,” Dr. Kumar said. He related a 94% rate of overall response that was even higher in patients treated with doses of 150 x 10⁶ CAR+ T cells or more. Many of the responses were lasting, he said, with 5 patients in ongoing response for more than a year.

“The results are exciting enough that this is actually moving forward with registration trials,” Dr. Kumar added.

Another novel CAR T-cell product, LCAR-B38M, has demonstrated promising results. LCAR-B38M principally targets BCMA and has led to a significant number of patients achieving stringent complete response that lasted beyond 1 year.

Multiple BCMA-targeting CAR T-cell products that use different vectors and different costimulatory molecules are currently in clinical trials, Dr. Kumar said.

BiTEs

In contrast to CAR T-cell products that must be customized to each patient in a process that takes weeks, BiTEs are a ready-made approach to allow T cells to engage with tumor cells.

“In patients with advanced disease, a lot can change in that short timeframe, so having an approach that is off-the-shelf, which is not patient specific, is quite attractive,” Dr. Kumar said.

BCMA-directed BiTE therapies to watch that are under investigation include AMG 420 and PF-06863135, he said. 

Bone marrow aspirate showing MM

New York—Three novel treatment strategies that target B-cell maturation antigen (BCMA) are showing promise in recent multiple myeloma (MM) clinical trials, according to Shaji K. Kumar, MD, of Mayo Clinic Cancer Center in Rochester, Minnesota.

The strategies include B-cell maturation antigen (BCMA) antibody-drug conjugate, BCMA-specific chimeric antigen receptor (CAR) T-cell therapies, and bispecific T-cell engagers (BiTEs).

“Clearly, there are a lot of exciting drugs that are currently in clinical trials, but these 3 platforms appear to be much more advanced than the others, and hopefully we will see that in the clinic in the near future,” Dr. Kumar told attendees at the NCCN 13th Annual Congress: Hematologic Malignancies.

BCMA is required for plasma cell survival and is broadly expressed on malignant plasma cells.

BCMA antibody-drug conjugate

The antibody-drug conjugate, GSK2857916, is a humanized IgG1 anti-BCMA antibody conjugated to a microtubule-disrupting agent. It produced an overall response rate of 67% at the 2 highest dose levels in 9 MM patients who had previously received multiple standard-of-care agents.

“Some of the responses were quite durable, lasting several months,” he said.

Now, GSK2857916 is being evaluated in a variety of different combinations, he said, including in an upcoming phase 2 study of the antibody-drug conjugate in combination with lenalidomide plus dexamethasone or bortezomib plus dexamethasone in patients with relapsed or refractory disease.

BCMA-specific CAR T-cell therapy

Some of the most “exciting” data with anti-BCMA CAR T-cell therapy in myeloma, according to Dr. Kumar, involves bb2121. bb2121 showed durable clinical responses in heavily pretreated patients, according to an ASH 2017 presentation.

“The overall response rate is quite significant,” Dr. Kumar said. He related a 94% rate of overall response that was even higher in patients treated with doses of 150 x 10⁶ CAR+ T cells or more. Many of the responses were lasting, he said, with 5 patients in ongoing response for more than a year.

“The results are exciting enough that this is actually moving forward with registration trials,” Dr. Kumar added.

Another novel CAR T-cell product, LCAR-B38M, has demonstrated promising results. LCAR-B38M principally targets BCMA and has led to a significant number of patients achieving stringent complete response that lasted beyond 1 year.

Multiple BCMA-targeting CAR T-cell products that use different vectors and different costimulatory molecules are currently in clinical trials, Dr. Kumar said.

BiTEs

In contrast to CAR T-cell products that must be customized to each patient in a process that takes weeks, BiTEs are a ready-made approach to allow T cells to engage with tumor cells.

“In patients with advanced disease, a lot can change in that short timeframe, so having an approach that is off-the-shelf, which is not patient specific, is quite attractive,” Dr. Kumar said.

BCMA-directed BiTE therapies to watch that are under investigation include AMG 420 and PF-06863135, he said. 

Photo courtesy of NIH

Older patients with untreated Hodgkin lymphoma (HL) can achieve significantly improved survival by adding brentuximab vedotin to their treatment before and after standard chemotherapy, a recent study found.

In patients with low comorbidity scores, responses were even more robust, reported lead author Andrew M. Evens, DO, of the Rutgers Cancer Institute of New Jersey, and colleagues.

“Causes of poor outcomes for older patients with HL are not fully understood but have been attributed to a combination of factors, including presence of comorbidities, poorer performance status, disease and biological differences, inability to tolerate chemotherapy at the full dose, and increased treatment-related toxicities,” the authors wrote in the Journal of Clinical Oncology.

The primary goal of the study was to improve outcomes for untreated, older patients, a group that’s historically been a difficult-to-treat patient population.

The phase 2 trial included 48 HL patients with a median age of 69 (range, 60 – 88).

All patients underwent geriatric assessment for comorbidities and loss of activities of daily living.

Treatment consisted of two doses of brentuximab followed by six cycles of doxorubicin, vinblastine, and dacarbazine (AVD), then four more doses of brentuximab (consolidation doses).

The primary endpoint was complete remission at completion of AVD.

Secondary outcomes included overall response rate, 2-year progression-free survival, 2-year overall survival, and safety.

Just over half the patients (52%) completed all cycles of therapy, and almost three quarters (73%) received at least one consolidation dose of brentuximab.

Among the first 23 evaluable patients, both the complete remission rate and overall response rate were 96%. Intention-to-treat survival rates for all 48 patients were 84% for 2-year progression-free survival and 93% for 2-year overall survival.

Historical 2-year progression-free survival rates in similar older patients is poor, at 50%, so the progression-free survival rate of 84% in this study represents a significant improvement.

Of note, patients with fewer comorbidities and without loss of instrumental activities of daily living showed more robust responses.

Patients with Cumulative Illness Rating Scale for Geriatrics (CIRS-G) comorbidity scores of less than 10 had a 2-year progression-free survival rate of 100% versus 45% for those with higher scores.

Similarly, patients without loss of instrumental activities achieved a progression-free survival rate of 94% versus 25% for those who had lost some instrumental activities.

Grade 3 or 4 adverse events occurred in 42% of patients, with neutropenia being the most common (44%).

“This study represents among the best-reported outcomes to date for untreated older patients with HL,” the investigators concluded.

Seattle Genetics supported the investigator-initiated trial. 

Photo courtesy of NIH

Older patients with untreated Hodgkin lymphoma (HL) can achieve significantly improved survival by adding brentuximab vedotin to their treatment before and after standard chemotherapy, a recent study found.

In patients with low comorbidity scores, responses were even more robust, reported lead author Andrew M. Evens, DO, of the Rutgers Cancer Institute of New Jersey, and colleagues.

“Causes of poor outcomes for older patients with HL are not fully understood but have been attributed to a combination of factors, including presence of comorbidities, poorer performance status, disease and biological differences, inability to tolerate chemotherapy at the full dose, and increased treatment-related toxicities,” the authors wrote in the Journal of Clinical Oncology.

The primary goal of the study was to improve outcomes for untreated, older patients, a group that’s historically been a difficult-to-treat patient population.

The phase 2 trial included 48 HL patients with a median age of 69 (range, 60 – 88).

All patients underwent geriatric assessment for comorbidities and loss of activities of daily living.

Treatment consisted of two doses of brentuximab followed by six cycles of doxorubicin, vinblastine, and dacarbazine (AVD), then four more doses of brentuximab (consolidation doses).

The primary endpoint was complete remission at completion of AVD.

Secondary outcomes included overall response rate, 2-year progression-free survival, 2-year overall survival, and safety.

Just over half the patients (52%) completed all cycles of therapy, and almost three quarters (73%) received at least one consolidation dose of brentuximab.

Among the first 23 evaluable patients, both the complete remission rate and overall response rate were 96%. Intention-to-treat survival rates for all 48 patients were 84% for 2-year progression-free survival and 93% for 2-year overall survival.

Historical 2-year progression-free survival rates in similar older patients is poor, at 50%, so the progression-free survival rate of 84% in this study represents a significant improvement.

Of note, patients with fewer comorbidities and without loss of instrumental activities of daily living showed more robust responses.

Patients with Cumulative Illness Rating Scale for Geriatrics (CIRS-G) comorbidity scores of less than 10 had a 2-year progression-free survival rate of 100% versus 45% for those with higher scores.

Similarly, patients without loss of instrumental activities achieved a progression-free survival rate of 94% versus 25% for those who had lost some instrumental activities.

Grade 3 or 4 adverse events occurred in 42% of patients, with neutropenia being the most common (44%).

“This study represents among the best-reported outcomes to date for untreated older patients with HL,” the investigators concluded.

Seattle Genetics supported the investigator-initiated trial. 

Photo courtesy of NIH

Older patients with untreated Hodgkin lymphoma (HL) can achieve significantly improved survival by adding brentuximab vedotin to their treatment before and after standard chemotherapy, a recent study found.

In patients with low comorbidity scores, responses were even more robust, reported lead author Andrew M. Evens, DO, of the Rutgers Cancer Institute of New Jersey, and colleagues.

“Causes of poor outcomes for older patients with HL are not fully understood but have been attributed to a combination of factors, including presence of comorbidities, poorer performance status, disease and biological differences, inability to tolerate chemotherapy at the full dose, and increased treatment-related toxicities,” the authors wrote in the Journal of Clinical Oncology.

The primary goal of the study was to improve outcomes for untreated, older patients, a group that’s historically been a difficult-to-treat patient population.

The phase 2 trial included 48 HL patients with a median age of 69 (range, 60 – 88).

All patients underwent geriatric assessment for comorbidities and loss of activities of daily living.

Treatment consisted of two doses of brentuximab followed by six cycles of doxorubicin, vinblastine, and dacarbazine (AVD), then four more doses of brentuximab (consolidation doses).

The primary endpoint was complete remission at completion of AVD.

Secondary outcomes included overall response rate, 2-year progression-free survival, 2-year overall survival, and safety.

Just over half the patients (52%) completed all cycles of therapy, and almost three quarters (73%) received at least one consolidation dose of brentuximab.

Among the first 23 evaluable patients, both the complete remission rate and overall response rate were 96%. Intention-to-treat survival rates for all 48 patients were 84% for 2-year progression-free survival and 93% for 2-year overall survival.

Historical 2-year progression-free survival rates in similar older patients is poor, at 50%, so the progression-free survival rate of 84% in this study represents a significant improvement.

Of note, patients with fewer comorbidities and without loss of instrumental activities of daily living showed more robust responses.

Patients with Cumulative Illness Rating Scale for Geriatrics (CIRS-G) comorbidity scores of less than 10 had a 2-year progression-free survival rate of 100% versus 45% for those with higher scores.

Similarly, patients without loss of instrumental activities achieved a progression-free survival rate of 94% versus 25% for those who had lost some instrumental activities.

Grade 3 or 4 adverse events occurred in 42% of patients, with neutropenia being the most common (44%).

“This study represents among the best-reported outcomes to date for untreated older patients with HL,” the investigators concluded.

Seattle Genetics supported the investigator-initiated trial. 

INTRODUCTION

Small cell lung cancer (SCLC) is an aggressive cancer of neuroendocrine origin that accounts for approximately 15% of all lung cancer cases, with approximately 33,000 patients diagnosed annually.¹ The incidence of SCLC in the United States has steadily declined over the past 30 years, presumably because of a decrease in the number of smokers and a change to low-tar filter cigarettes.² Although the overall incidence of SCLC has been decreasing, the incidence in women is increasing and the male-to-female incidence ratio is now 1:1.³ Nearly all cases of SCLC are associated with heavy tobacco exposure, making it a heterogeneous disease with a complex genomic landscape consisting of thousands of mutations.^4,5 Despite recent advances in the treatment of non-small cell lung cancer, the therapeutic options for SCLC remain limited, with a median overall survival (OS) of 9 months in patients with advanced disease.

DIAGNOSIS AND STAGING

CASE PRESENTATION

A 61-year-old man presents to the emergency department with progressive shortness of breath and cough over the past 6 weeks. He also reports a 20-lb weight loss over the same period. He is a current smoker and has been smoking 1 pack of cigarettes per day since the age of 18 years. A chest radiograph obtained in the emergency department shows a right hilar mass. Computed tomography (CT) scan confirms the presence of a 4.5-cm right hilar mass and enlarged mediastinal lymph nodes bilaterally.

• What are the next steps in diagnosis?

SCLC is characterized by rapid growth and early hematogenous metastasis. Consequently, only 25% of patients have limited-stage disease at the time of diagnosis. According to the Veterans Administration Lung Study Group (VALSG) staging system, limited-stage disease is defined as tumor that is confined to 1 hemithorax and can be encompassed within 1 radiation field. This typically includes mediastinal lymph nodes and ipsilateral supraclavicular lymph nodes. Approximately 75% of patients present with extensive-stage disease, which is defined as disease that cannot be classified as limited, including disease that extends beyond 1 hemithorax. Extensive-stage disease includes the presence of malignant pleural effusion and/or distant metastasis.⁶ The VALSG classification and staging system is more commonly used in clinical practice than the American Joint Committee on Cancer TNM staging system because it is less complex and directs treatment decisions, as most of the literature on SCLC classifies patients based on the VALSG system.⁷

Given SCLC’s propensity to metastasize quickly, none of the currently available screening methods have proven successful in early detection of SCLC. In the National Lung Cancer Screening Trial, 86% of the 125 patients who were diagnosed with SCLC while undergoing annual low-dose chest CT scans had advanced disease at diagnosis.^8,9 These results highlight the fact that most cases of SCLC develop in the interval between annual screening imaging.

SCLC frequently presents with a large hilar mass that is symptomatic. Common symptoms include shortness of breath and cough. In addition, patients with SCLC usually have bulky mediastinal adenopathy at presentation. SCLC is commonly located submucosally in the bronchus, and therefore hemoptysis is not a very common symptom at the time of presentation. Patients may present with superior vena cava syndrome from local compression by the tumor. Not infrequently, SCLC is associated with paraneoplastic syndromes that arise due to ectopic secretion of hormones or antibodies by the tumor cells. The paraneoplastic syndromes can be broadly categorized as endocrine or neurologic (Table 1). The presence of a paraneoplastic syndrome is often a clue to the potential diagnosis of SCLC in the presence of a hilar mass. Additionally, some paraneoplastic syndromes, more specifically endocrine paraneoplastic syndromes, follow the pattern of disease response and relapse, and therefore can sometimes serve as an early marker of disease relapse or progression.

The common sites of metastases include brain, liver, and bone. Therefore, the staging workup should include fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT scan. Contrast-enhanced CT scan of the chest and abdomen and bone scan can be obtained for staging in lieu of PET scan. Due to the physiologic FDG uptake, cerebral metastases cannot be assessed with sufficient certainty using PET-CT.¹⁰ Therefore, brain imaging with contrast-enhanced CT or magnetic resonance imaging (MRI) is also necessary. Although the incidence of metastasis to bone marrow is less than 10%, bone marrow aspiration and biopsy are warranted in patients with unexplained cytopenias, especially when the cytopenia is associated with teardrop-shaped red cells or nucleated red cells on peripheral blood smear, findings indicative of a marrow infiltrative process.⁷ The tissue diagnosis is established by obtaining a biopsy of the primary tumor or 1 of the metastatic sites. In localized disease, bronchoscopy (with endobronchial ultrasound, if necessary) with biopsy of the centrally located tumor and/or lymph node is required. Histologically, SCLC consists of monomorphic cells, a high nuclear-cytoplasmic ratio, and confluent necrosis. The tumor cells are positive for chromogranin, synaptophysin, and CD56 by immunohistochemistry, and very frequently are also positive for thyroid transcription factor 1.¹¹ Although serum tumor markers, including neuron-specific enolase and progastrin-releasing peptide, are frequently elevated in patients with SCLC, these markers are of limited value in clinical practice because they lack sensitivity and specificity.¹²

 

MANAGEMENT OF LIMITED-STAGE DISEASE

CASE CONTINUED

The patient undergoes FDG PET scan, which shows the presence of a hypermetabolic right hilar mass in addition to enlarged and hypermetabolic bilateral mediastinal lymph nodes. There are no other areas of FDG avidity. Brain MRI does not show any evidence of brain metastasis. Thus, the patient is confirmed to have limited-stage SCLC.

• What is the standard of care treatment for limited-stage SCLC?

SCLC is exquisitely sensitive to both chemotherapy and radiation, especially at the time of initial presentation. The standard of care treatment of limited-stage SCLC is 4 cycles of platinum-based chemotherapy in combination with thoracic radiation started within the first 2 cycles of chemotherapy (Figure 1).

This regimen yields an overall response rate between 75% and 90% and a complete response rate of 50%, with a median OS of 18 to 24 months.¹³ The cure rate with this approach is approximately 25%, with the remaining 75% of patients experiencing disease relapse within the first 5 years after completion of treatment. Attempts to improve the cure rate in the limited-stage setting by combining standard of care therapies with novel agents have been unsuccessful.

CHOICE OF CHEMOTHERAPY

Etoposide and cisplatin is the most commonly used initial combination chemotherapy regimen in limited-stage SCLC.¹⁴ This combination has largely replaced anthracycline-based regimens given its favorable efficacy and toxicity profile.^15–17 Several small randomized trials have shown comparable efficacy of carboplatin and etoposide in extensive-stage SCLC.^18–20 A meta-analysis of 4 randomized trials comparing cisplatin-based versus carboplatin-based regimens in 663 patients with SCLC (32% had limited-stage disease and 68% had extensive-stage disease) showed no statistically significant difference in response rate, progression-free survival (PFS), or OS between the 2 regimens.²¹ Therefore, in clinical practice carboplatin is frequently used instead of cisplatin in patients with extensive-stage disease. In patients with limited-stage disease, cisplatin is still the drug of choice. However, the toxicity profile of the 2 regimens is different. Cisplatin-based regimens are more commonly associated with neuropathy, nephrotoxicity, and chemotherapy-induced nausea/vomiting,¹⁸ while carboplatin-based regimens are more myelosuppressive.²² In addition, the combination of thoracic radiation with either of these regimens is associated with a higher risk of esophagitis, pneumonitis, and myelosuppression.²³ The use of myeloid growth factors is not recommended in patients undergoing concurrent chemoradiation.²⁴ Of note, intravenous etoposide is always preferred over oral etoposide, especially in the curative setting given the unreliable absorption and bioavailability of oral formulations.

THORACIC RADIOTHERAPY

Adding thoracic radiotherapy to platinum-etoposide chemotherapy improves local control and OS. Two meta-analyses of 13 trials including more than 2000 patients have shown a 25% to 30% decrease in local failure and a 5% to 7% increase in 2-year OS with chemoradiation compared to chemotherapy alone in limited-stage SCLC.^25,26 Early (within the first 2 cycles) concurrent thoracic radiation is superior to delayed and/or sequential radiation in terms of local control and OS.^23,27,28 The dose and fractionation of thoracic radiation in limited-stage SCLC has remained a controversial issue. The Eastern Cooperative Oncology Group/Radiation Therapy Oncology Group randomized trial compared 45 Gy of radiotherapy delivered twice daily over a period of 3 weeks to 45 Gy once daily over 5 weeks concurrently with chemotherapy. The twice daily regimen led to a 10% improvement in 5-year OS (26% versus 16%), but a higher incidence of grade 3 and 4 adverse events.¹³ Despite the survival advantage demonstrated by hyperfractionated radiotherapy, the results need to be interpreted with caution because the radiation doses are not biologically equivalent. In addition, the difficult logistics of patients receiving radiation twice a day has limited the routine implementation of this strategy. Subsequently, another randomized phase 3 trial (CONVERT) compared 45 Gy radiotherapy twice daily with 66 Gy radiotherapy once daily in limited-stage SCLC.²⁹ This trial did not show any difference in OS. The patients in the twice daily arm had a higher incidence of grade 4 neutropenia. Considering the results of these trials, both strategies—45 Gy fractionated twice daily or 60 Gy fractionated once daily, delivered concurrently with chemotherapy—are acceptable in the setting of limited-stage SCLC. However, quite often a hyperfractionated regimen is not feasible for patients and many radiation oncology centers. Hopefully, the ongoing CALGB 30610 study will clarify the optimal radiation schedule for limited-stage disease.

 

PROPHYLACTIC CRANIAL IRRADIATION

Approximately 75% of patients with limited-stage disease experience disease recurrence, and brain is the site of recurrence in approximately half of these patients.³⁰ Prophylactic cranial irradiation (PCI) consisting of 25 Gy radiotherapy delivered in 10 fractions has been shown to be effective in decreasing the incidence of cerebral metastases.^30–32 Although individual small studies have not shown a survival benefit of PCI because of small sample size and limited power, a meta-analysis of these studies has shown a 25% decrease in the 3-year incidence of brain metastasis and 5.4% increase in 3-year OS.³⁰ Most patients included in these studies had limited-stage disease. Therefore, PCI is the standard of care for patients with limited-stage disease who attain a partial or complete response to chemoradiation.

ROLE OF SURGERY

Surgical resection may be an acceptable choice in a very limited subset of patients with peripherally located small (< 5 cm) tumors where mediastinal lymph nodes have been confirmed to be uninvolved with complete mediastinal staging.^33,34 Most of the data in this setting are derived from retrospective studies.^35,36 A 5-year OS between 40% and 60% has been reported with this strategy in patients with clinical stage I disease. In general, when surgery is considered, lobectomy with mediastinal lymph node dissection followed by chemotherapy (if there is no nodal involvement) or chemoradiation (if nodal involvement) is recommended.^37,38 Wedge or segmental resections are not considered to be optimal surgical options.

MANAGEMENT OF EXTENSIVE-STAGE DISEASE

CASE CONTINUED

The patient receives 4 cycles of cisplatin and etoposide along with 70 Gy radiotherapy concurrently with the first 2 cycles of chemotherapy. His post-treatment CT scans show a partial response. He undergoes PCI 6 weeks after completion of treatment. At routine follow-up 18 months later, he is doing generally well except for mildly decreased appetite and an unintentional weight loss of 5 lb. CT scans demonstrate multiple hypodense liver lesions ranging from 7 mm to 2 cm in size and a 2-cm left adrenal gland lesion highly concerning for metastasis. FDG PET scan confirms that the adrenal and liver lesions are hypermetabolic. In addition, the PET scan shows multiple FDG-avid bone lesions throughout the spine. Brain MRI is negative for brain metastasis.

• What is the standard of care for treatment of extensive-stage disease?

Chemotherapy is the mainstay of treatment for extensive-stage SCLC; the goals of treatment are prolongation of survival, prevention or alleviation of cancer-related symptoms, and improvement in quality of life. The combination of etoposide with a platinum agent (carboplatin or cisplatin) is the preferred first-line treatment option. Carboplatin is more commonly used in clinical practice in this setting because of its comparable efficacy and better tolerability compared to cisplatin (Figure 2).²¹ A Japanese phase 3 trial comparing cisplatin plus irinotecan with cisplatin plus etoposide in the first-line setting in extensive-stage SCLC showed improvement in median and 2-year OS with the cisplatin/irinotecan regimen; however, 2 subsequent phase 3 trials conducted in the United States comparing these 2 regimens did not show any difference in OS. In addition, the cisplatin/irinotecan regimen was more toxic than the etoposide-based regimen.^39,40 Therefore, 4 to 6 cycles of platinum/etoposide remains the standard of care first-line treatment for extensive-stage SCLC in the United States. The combination yields a 60% to 70% response rate, but the majority of patients invariably experience disease progression, with a median OS of 9 to 11 months.⁴¹ Maintenance chemotherapy beyond the initial 4 to 6 cycles does not improve survival and is associated with higher cumulative toxicity.⁴²

 

Multiple attempts at improving first-line chemotherapy in extensive-stage disease have failed to show any meaningful difference in OS. For example, the addition of ifosfamide, palifosfamide, cyclophosphamide, taxane, or anthracycline to platinum doublet failed to show improvement in OS and led to more toxicity.^43–46 Additionally, the use of alternating or cyclic chemotherapies in an attempt to curb drug resistance has also failed to show survival benefit.^47–49 The addition of the antiangiogenic agent bevacizumab to standard platinum-based doublet has not prolonged OS in SCLC and has led to an unacceptably higher rate of tracheoesophageal fistula when used in conjunction with chemoradiation in limited-stage disease.^50–55 Finally, the immune checkpoint inhibitor ipilimumab in combination with platinum plus etoposide failed to improve PFS or OS compared to platinum plus etoposide alone in a recent phase 3 trial, and maintenance pembrolizumab after completion of platinum-based chemotherapy did not improve PFS.^56,57

More recently, a phase 2 study of pembrolizumab in extensive-stage SCLC (KEYNOTE 158) reported an overall response rate of 35.7%, median PFS of 2.1 months, and median OS of 14.6 months in patients who tested positive for programmed death ligand-1 (PD-L1) expression (which was defined as a PD-L1 Combined Positive Score ≥ 1).⁵⁸ The median duration of response has not been reached in this study, indicating that pembrolizumab may be a promising approach in patients with extensive-stage SCLC, especially for those with PD-L1–positive tumors.

Patients with extensive-stage disease who have brain metastasis at the time of diagnosis can be treated with systemic chemotherapy first if the brain metastases are asymptomatic and there is significant extracranial disease burden. In that case, whole brain radiotherapy should be given after completion of systemic therapy.

SECOND-LINE CHEMOTHERAPY

Despite being exquisitely chemosensitive, SCLC is associated with a very poor prognosis largely because of invariable disease progression following first-line therapy and lack of effective second-line treatment options that can lead to appreciable disease control. The choice of second-line treatment is predominantly determined by the time of disease relapse after first-line platinum-based therapy. If this interval is 6 months or longer, re-treatment utilizing the same platinum doublet is appropriate. However, if the interval is 6 months or less, second-line systemic therapy options should be explored. Unfortunately, the response rate tends to be less than 10% with most of the second-line therapies in platinum-resistant disease (defined as disease progression within 3 months of receiving platinum-based therapy). If disease progression occurs between 3 and 6 months after completion of platinum-based therapy, the response rate with second-line chemotherapy is in the range of 25%.^59,60

A number of second-line chemotherapy options have been explored in small studies, including topotecan, irinotecan, paclitaxel, docetaxel, temozolomide, vinorelbine, oral etoposide, gemcitabine, bendamustine, and CAV (cyclophosphamide, doxorubicin, vincristine) (Table 2).^61–77 Of these, topotecan (oral or intravenous) is the only second-line agent approved by the US Food and Drug Administration (FDA) for extensive-stage SCLC; this approval is based on a phase 3 trial showing a survival benefit with topotecan compared to best supportive care. The duration of second-line chemotherapy is not defined and is largely driven by duration of response and occurrence of side effects. Given the lack of effective treatment options, patients with platinum-resistant disease should be considered for clinical trials whenever possible. In patients with disease progression after second-line therapy, a third-line treatment can be considered if their Eastern Cooperative Oncology Group (ECOG) performance status is 0 to 2, but it is unclear whether further treatments would improve survival.

 

 

IMMUNOTHERAPY

The role of immune checkpoint inhibitors in the treatment of SCLC is evolving, and currently there are no FDA-approved immunotherapy agents for treating SCLC. A recently conducted phase 1/2 trial (CheckMate 032) studied the anti-programmed death(PD)-1 antibody nivolumab with or without the anti-cytotoxic T-lymphocyte–associated antigen (CTLA) -4 antibody ipilimumab in patients with relapsed SCLC. The authors reported response rates of 10% with nivolumab 3 mg/kg and 21% with nivolumab 1 mg/kg plus ipilimumab 3 mg/kg.^78,79 The 2-year OS was 26% with the combination and 14% with single-agent nivolumab. Only 18% of patients had PD-L1 expression of ≥ 1%, and the response rate did not correlate with PD-L1 status. The rate of grade 3 or 4 adverse events was approximately 20%, and only 10% of patients discontinued treatment because of toxicity. Based on these data, nivolumab plus ipilimumab is now included in the National Comprehensive Cancer Network guidelines as an option for patients with SCLC who experience disease relapse within 6 months of receiving platinum-based therapy;⁷ however, it is questionable whether routine use of this combination is justified based on currently available data. The evidence for the combination of nivolumab and ipilimumab remains limited. The efficacy and toxicity data from both randomized and nonrandomized cohorts were presented together, making it hard to interpret the results.

Another phase 1b study (KEYNOTE-028) evaluated the anti-PD-1 antibody pembrolizumab (10 mg/kg intravenously every 2 weeks) in patients with relapsed SCLC who had received 1 or more prior lines of therapy and had PD-L1 expression of ≥ 1%. This study showed a response rate of 33%, with a median duration of response of 19 months and 1-year OS of 38%.⁸⁰ Although only 28% of screened patients had PD-L1 expression of ≥ 1%, these results indicated that at least a subset of SCLC patients are able to achieve durable responses with immune checkpoint inhibition. A number of clinical trials utilizing immune checkpoint inhibitors in various combinations and settings are currently underway.

ROLE OF PROPHYLACTIC CRANIAL IRRADIATION

The role of PCI in extensive-stage SCLC is not clearly defined. A randomized phase 3 trial conducted by the European Organization for Research and Treatment of Cancer (EORTC) comparing PCI with no PCI in patients with extensive-stage SCLC who had a partial or complete response to initial platinum-based chemotherapy showed a decrease in the incidence of symptomatic brain metastasis and improvement in 1-year OS with PCI.⁸¹ However, this trial did not require mandatory brain imaging prior to PCI, and thus it is unclear if some patients in the PCI group had asymptomatic brain metastasis prior to enrollment and therefore received therapeutic benefit from brain radiation. Additionally, the dose and fractionation of PCI was not standardized across patient groups.

A more recent phase 3 study conducted in Japan that compared PCI (25 Gy in 10 fractions) with no PCI reported no difference in survival between the 2 groups.⁸² As opposed to the EORTC study, the Japanese study did require baseline brain imaging to confirm the absence of brain metastasis prior to enrollment. In addition, the control patients underwent periodic brain MRI to allow early detection of brain metastasis. Given the emergence of the new data, the impact of PCI on survival in patients with extensive-stage SCLC is unproven, and PCI likely has a role in a highly selected small group of patients with extensive-stage SCLC. PCI is not recommended for patients with poor performance status (ECOG performance score of 3 or 4) or underlying neurocognitive disorders.^34,83

The NMDA-receptor antagonist memantine can be used in patients undergoing PCI to delay the occurrence of cognitive dysfunction.⁶¹ Memantine 20 mg daily delayed time to cognitive decline and reduced the rate of decline in memory, executive function, and processing speed compared to placebo in patients receiving whole brain radiotherapy.⁸⁴

 

 

ROLE OF RADIOTHERAPY

A subset of patients with extensive-stage SCLC may benefit from consolidative thoracic radiotherapy after completion of platinum-based chemotherapy. A randomized trial that enrolled patients who achieved complete or near complete response after 3 cycles of cisplatin plus etoposide compared thoracic radiotherapy in combination with continued chemotherapy versus chemotherapy alone.⁸⁵ The median OS was longer with the addition of thoracic radiotherapy compared to chemotherapy alone. Another phase 3 trial did not show improvement in 1-year OS with consolidative thoracic radiotherapy, but 2-year OS and 6-month PFS were longer.⁸⁶ In general, consolidative thoracic radiotherapy benefits patients who have residual thoracic disease and low-bulk extrathoracic disease that has responded to systemic therapy.⁸⁷ In addition, patients who initially presented with bulky symptomatic thoracic disease should also be considered for consolidative radiation.

Similar to other solid tumors, radiotherapy should be utilized for palliative purposes in patients with painful bone metastasis, spinal cord compression, or brain metastasis. Surgery is generally not recommended for spinal cord compression given the short life expectancy of patients with extensive-stage disease. Whole brain radiotherapy is preferred over stereotactic radiosurgery because micrometastasis is frequently present even in the setting of 1 or 2 radiographically evident brain metastasis.

NOVEL THERAPIES

The very complex genetic landscape of SCLC accounts for its resistance to conventional therapy and high recurrence rate; however, at the same time this complexity can form the basis for effective targeted therapy for the disease. One of the major factors hindering the development of targeted therapies in SCLC is limited availability of tissue due to small tissue samples and the frequent presence of significant necrosis in the samples. In recent years, several different therapeutic strategies and targeted agents have been investigated for their potential role in SCLC. Several of them, including EGFR tyrosine kinase inhibitors (TKIs), BCR-ABL TKIs, mTOR inhibitors, and VEGF inhibitors, have not been shown to provide a survival advantage in this disease. Several others, including PARP inhibitors, cellular developmental pathway inhibitors, and antibody-drug conjugates, are being tested. A phase 1 study of veliparib combined with cisplatin and etoposide in patients with previously untreated extensive-stage SCLC demonstrated a complete response in 14.3%, a partial response in 57.1%, and stable disease in 28.6% of patients with an acceptable safety profile.⁸⁸ So far, none of these agents are approved for use in SCLC, and the majority are in early- phase clinical trials.⁸⁹

One of the emerging targets in the treatment of SCLC is delta-like protein 3 (DLL3). DLL3 is expressed on more than 80% of SCLC tumor cells and cancer stem cells. Rovalpituzumab tesirine is an antibody-drug conjugate consisting of humanized anti-DLL3 monoclonal antibody linked to SC-DR002, a DNA-crosslinking agent. A phase 1 trial of rovalpituzumab in patients with relapsed SCLC after 1 or 2 prior lines of therapy reported a response rate of 31% in patients with DLL3 expression of ≥ 50%. The median duration of response and median PFS were both 4.6 months.⁹⁰ Rovalpituzumab is currently in later phases of clinical trials and has a potential to serve as an option for patients with extensive-stage disease after disease progression on platinum-based therapy.

SUMMARY

Four to 6 cycles of carboplatin and etoposide remain the standard of care first-line treatment for patients with extensive stage SCLC. The only FDA-approved second-line treatment option is topotecan. Re-treatment with the original platinum doublet is a reasonable option for patients who have disease progression 6 months or longer after completion of platinum-based therapy. The immune checkpoint inhibitors pembrolizumab and combination nivolumab and ipilimumab have shown promising results in the second-line setting and beyond. The role of PCI has become more controversial in recent years, and periodic brain MRI in lieu of PCI is now an acceptable approach.

 

 

RESPONSE ASSESSMENT/SURVEILLANCE

For patients undergoing treatment for limited-stage SCLC, response assessment with contrast-enhanced CT of the chest/abdomen should be performed after completion of 4 cycles of chemotherapy and thoracic radiation.⁷ The surveillance guidelines consist of history, physical exam, and imaging every 3 months during the first 2 years, every 6 months during the third year, and annually thereafter. If PCI is not performed, brain MRI or contrast-enhanced CT scan should be performed every 3 or 4 months during the first 2 years of follow up. For extensive-stage disease, response assessment should be performed after every 2 cycles of therapy. After completion of therapy, history, physical exam, and imaging should be done every 2 months during the first year, every 3 or 4 months during years 2 and 3, every 6 months during years 4 and 5, and annually thereafter. Routine use of PET scan for surveillance is not recommended. Any new pulmonary nodule should prompt evaluation for a second primary lung malignancy. Finally, smoking cessation counseling is an integral part of management of any patient with SCLC and should be included with every clinic visit.

CONCLUSION

SCLC is a heterogeneous and genetically complex disease with a very high mortality rate. The current standard of care includes concurrent chemoradiation with cisplatin and etoposide for limited-stage SCLC and the combination of platinum and etoposide for extensive SCLC. A number of novel treatment approaches, including immune checkpoint inhibitors and antibody-drug conjugates, have had promising results in early clinical trials. Given the limited treatment options and large unmet need for new treatment options, enrollment in clinical trials is strongly recommended for patients with SCLC.

INTRODUCTION

Small cell lung cancer (SCLC) is an aggressive cancer of neuroendocrine origin that accounts for approximately 15% of all lung cancer cases, with approximately 33,000 patients diagnosed annually.¹ The incidence of SCLC in the United States has steadily declined over the past 30 years, presumably because of a decrease in the number of smokers and a change to low-tar filter cigarettes.² Although the overall incidence of SCLC has been decreasing, the incidence in women is increasing and the male-to-female incidence ratio is now 1:1.³ Nearly all cases of SCLC are associated with heavy tobacco exposure, making it a heterogeneous disease with a complex genomic landscape consisting of thousands of mutations.^4,5 Despite recent advances in the treatment of non-small cell lung cancer, the therapeutic options for SCLC remain limited, with a median overall survival (OS) of 9 months in patients with advanced disease.

DIAGNOSIS AND STAGING

CASE PRESENTATION

A 61-year-old man presents to the emergency department with progressive shortness of breath and cough over the past 6 weeks. He also reports a 20-lb weight loss over the same period. He is a current smoker and has been smoking 1 pack of cigarettes per day since the age of 18 years. A chest radiograph obtained in the emergency department shows a right hilar mass. Computed tomography (CT) scan confirms the presence of a 4.5-cm right hilar mass and enlarged mediastinal lymph nodes bilaterally.

• What are the next steps in diagnosis?

SCLC is characterized by rapid growth and early hematogenous metastasis. Consequently, only 25% of patients have limited-stage disease at the time of diagnosis. According to the Veterans Administration Lung Study Group (VALSG) staging system, limited-stage disease is defined as tumor that is confined to 1 hemithorax and can be encompassed within 1 radiation field. This typically includes mediastinal lymph nodes and ipsilateral supraclavicular lymph nodes. Approximately 75% of patients present with extensive-stage disease, which is defined as disease that cannot be classified as limited, including disease that extends beyond 1 hemithorax. Extensive-stage disease includes the presence of malignant pleural effusion and/or distant metastasis.⁶ The VALSG classification and staging system is more commonly used in clinical practice than the American Joint Committee on Cancer TNM staging system because it is less complex and directs treatment decisions, as most of the literature on SCLC classifies patients based on the VALSG system.⁷

Given SCLC’s propensity to metastasize quickly, none of the currently available screening methods have proven successful in early detection of SCLC. In the National Lung Cancer Screening Trial, 86% of the 125 patients who were diagnosed with SCLC while undergoing annual low-dose chest CT scans had advanced disease at diagnosis.^8,9 These results highlight the fact that most cases of SCLC develop in the interval between annual screening imaging.

SCLC frequently presents with a large hilar mass that is symptomatic. Common symptoms include shortness of breath and cough. In addition, patients with SCLC usually have bulky mediastinal adenopathy at presentation. SCLC is commonly located submucosally in the bronchus, and therefore hemoptysis is not a very common symptom at the time of presentation. Patients may present with superior vena cava syndrome from local compression by the tumor. Not infrequently, SCLC is associated with paraneoplastic syndromes that arise due to ectopic secretion of hormones or antibodies by the tumor cells. The paraneoplastic syndromes can be broadly categorized as endocrine or neurologic (Table 1). The presence of a paraneoplastic syndrome is often a clue to the potential diagnosis of SCLC in the presence of a hilar mass. Additionally, some paraneoplastic syndromes, more specifically endocrine paraneoplastic syndromes, follow the pattern of disease response and relapse, and therefore can sometimes serve as an early marker of disease relapse or progression.

The common sites of metastases include brain, liver, and bone. Therefore, the staging workup should include fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT scan. Contrast-enhanced CT scan of the chest and abdomen and bone scan can be obtained for staging in lieu of PET scan. Due to the physiologic FDG uptake, cerebral metastases cannot be assessed with sufficient certainty using PET-CT.¹⁰ Therefore, brain imaging with contrast-enhanced CT or magnetic resonance imaging (MRI) is also necessary. Although the incidence of metastasis to bone marrow is less than 10%, bone marrow aspiration and biopsy are warranted in patients with unexplained cytopenias, especially when the cytopenia is associated with teardrop-shaped red cells or nucleated red cells on peripheral blood smear, findings indicative of a marrow infiltrative process.⁷ The tissue diagnosis is established by obtaining a biopsy of the primary tumor or 1 of the metastatic sites. In localized disease, bronchoscopy (with endobronchial ultrasound, if necessary) with biopsy of the centrally located tumor and/or lymph node is required. Histologically, SCLC consists of monomorphic cells, a high nuclear-cytoplasmic ratio, and confluent necrosis. The tumor cells are positive for chromogranin, synaptophysin, and CD56 by immunohistochemistry, and very frequently are also positive for thyroid transcription factor 1.¹¹ Although serum tumor markers, including neuron-specific enolase and progastrin-releasing peptide, are frequently elevated in patients with SCLC, these markers are of limited value in clinical practice because they lack sensitivity and specificity.¹²

 

MANAGEMENT OF LIMITED-STAGE DISEASE

CASE CONTINUED

The patient undergoes FDG PET scan, which shows the presence of a hypermetabolic right hilar mass in addition to enlarged and hypermetabolic bilateral mediastinal lymph nodes. There are no other areas of FDG avidity. Brain MRI does not show any evidence of brain metastasis. Thus, the patient is confirmed to have limited-stage SCLC.

• What is the standard of care treatment for limited-stage SCLC?

SCLC is exquisitely sensitive to both chemotherapy and radiation, especially at the time of initial presentation. The standard of care treatment of limited-stage SCLC is 4 cycles of platinum-based chemotherapy in combination with thoracic radiation started within the first 2 cycles of chemotherapy (Figure 1).

This regimen yields an overall response rate between 75% and 90% and a complete response rate of 50%, with a median OS of 18 to 24 months.¹³ The cure rate with this approach is approximately 25%, with the remaining 75% of patients experiencing disease relapse within the first 5 years after completion of treatment. Attempts to improve the cure rate in the limited-stage setting by combining standard of care therapies with novel agents have been unsuccessful.

CHOICE OF CHEMOTHERAPY

Etoposide and cisplatin is the most commonly used initial combination chemotherapy regimen in limited-stage SCLC.¹⁴ This combination has largely replaced anthracycline-based regimens given its favorable efficacy and toxicity profile.^15–17 Several small randomized trials have shown comparable efficacy of carboplatin and etoposide in extensive-stage SCLC.^18–20 A meta-analysis of 4 randomized trials comparing cisplatin-based versus carboplatin-based regimens in 663 patients with SCLC (32% had limited-stage disease and 68% had extensive-stage disease) showed no statistically significant difference in response rate, progression-free survival (PFS), or OS between the 2 regimens.²¹ Therefore, in clinical practice carboplatin is frequently used instead of cisplatin in patients with extensive-stage disease. In patients with limited-stage disease, cisplatin is still the drug of choice. However, the toxicity profile of the 2 regimens is different. Cisplatin-based regimens are more commonly associated with neuropathy, nephrotoxicity, and chemotherapy-induced nausea/vomiting,¹⁸ while carboplatin-based regimens are more myelosuppressive.²² In addition, the combination of thoracic radiation with either of these regimens is associated with a higher risk of esophagitis, pneumonitis, and myelosuppression.²³ The use of myeloid growth factors is not recommended in patients undergoing concurrent chemoradiation.²⁴ Of note, intravenous etoposide is always preferred over oral etoposide, especially in the curative setting given the unreliable absorption and bioavailability of oral formulations.

THORACIC RADIOTHERAPY

Adding thoracic radiotherapy to platinum-etoposide chemotherapy improves local control and OS. Two meta-analyses of 13 trials including more than 2000 patients have shown a 25% to 30% decrease in local failure and a 5% to 7% increase in 2-year OS with chemoradiation compared to chemotherapy alone in limited-stage SCLC.^25,26 Early (within the first 2 cycles) concurrent thoracic radiation is superior to delayed and/or sequential radiation in terms of local control and OS.^23,27,28 The dose and fractionation of thoracic radiation in limited-stage SCLC has remained a controversial issue. The Eastern Cooperative Oncology Group/Radiation Therapy Oncology Group randomized trial compared 45 Gy of radiotherapy delivered twice daily over a period of 3 weeks to 45 Gy once daily over 5 weeks concurrently with chemotherapy. The twice daily regimen led to a 10% improvement in 5-year OS (26% versus 16%), but a higher incidence of grade 3 and 4 adverse events.¹³ Despite the survival advantage demonstrated by hyperfractionated radiotherapy, the results need to be interpreted with caution because the radiation doses are not biologically equivalent. In addition, the difficult logistics of patients receiving radiation twice a day has limited the routine implementation of this strategy. Subsequently, another randomized phase 3 trial (CONVERT) compared 45 Gy radiotherapy twice daily with 66 Gy radiotherapy once daily in limited-stage SCLC.²⁹ This trial did not show any difference in OS. The patients in the twice daily arm had a higher incidence of grade 4 neutropenia. Considering the results of these trials, both strategies—45 Gy fractionated twice daily or 60 Gy fractionated once daily, delivered concurrently with chemotherapy—are acceptable in the setting of limited-stage SCLC. However, quite often a hyperfractionated regimen is not feasible for patients and many radiation oncology centers. Hopefully, the ongoing CALGB 30610 study will clarify the optimal radiation schedule for limited-stage disease.

 

PROPHYLACTIC CRANIAL IRRADIATION

Approximately 75% of patients with limited-stage disease experience disease recurrence, and brain is the site of recurrence in approximately half of these patients.³⁰ Prophylactic cranial irradiation (PCI) consisting of 25 Gy radiotherapy delivered in 10 fractions has been shown to be effective in decreasing the incidence of cerebral metastases.^30–32 Although individual small studies have not shown a survival benefit of PCI because of small sample size and limited power, a meta-analysis of these studies has shown a 25% decrease in the 3-year incidence of brain metastasis and 5.4% increase in 3-year OS.³⁰ Most patients included in these studies had limited-stage disease. Therefore, PCI is the standard of care for patients with limited-stage disease who attain a partial or complete response to chemoradiation.

ROLE OF SURGERY

Surgical resection may be an acceptable choice in a very limited subset of patients with peripherally located small (< 5 cm) tumors where mediastinal lymph nodes have been confirmed to be uninvolved with complete mediastinal staging.^33,34 Most of the data in this setting are derived from retrospective studies.^35,36 A 5-year OS between 40% and 60% has been reported with this strategy in patients with clinical stage I disease. In general, when surgery is considered, lobectomy with mediastinal lymph node dissection followed by chemotherapy (if there is no nodal involvement) or chemoradiation (if nodal involvement) is recommended.^37,38 Wedge or segmental resections are not considered to be optimal surgical options.

MANAGEMENT OF EXTENSIVE-STAGE DISEASE

CASE CONTINUED

The patient receives 4 cycles of cisplatin and etoposide along with 70 Gy radiotherapy concurrently with the first 2 cycles of chemotherapy. His post-treatment CT scans show a partial response. He undergoes PCI 6 weeks after completion of treatment. At routine follow-up 18 months later, he is doing generally well except for mildly decreased appetite and an unintentional weight loss of 5 lb. CT scans demonstrate multiple hypodense liver lesions ranging from 7 mm to 2 cm in size and a 2-cm left adrenal gland lesion highly concerning for metastasis. FDG PET scan confirms that the adrenal and liver lesions are hypermetabolic. In addition, the PET scan shows multiple FDG-avid bone lesions throughout the spine. Brain MRI is negative for brain metastasis.

• What is the standard of care for treatment of extensive-stage disease?

Chemotherapy is the mainstay of treatment for extensive-stage SCLC; the goals of treatment are prolongation of survival, prevention or alleviation of cancer-related symptoms, and improvement in quality of life. The combination of etoposide with a platinum agent (carboplatin or cisplatin) is the preferred first-line treatment option. Carboplatin is more commonly used in clinical practice in this setting because of its comparable efficacy and better tolerability compared to cisplatin (Figure 2).²¹ A Japanese phase 3 trial comparing cisplatin plus irinotecan with cisplatin plus etoposide in the first-line setting in extensive-stage SCLC showed improvement in median and 2-year OS with the cisplatin/irinotecan regimen; however, 2 subsequent phase 3 trials conducted in the United States comparing these 2 regimens did not show any difference in OS. In addition, the cisplatin/irinotecan regimen was more toxic than the etoposide-based regimen.^39,40 Therefore, 4 to 6 cycles of platinum/etoposide remains the standard of care first-line treatment for extensive-stage SCLC in the United States. The combination yields a 60% to 70% response rate, but the majority of patients invariably experience disease progression, with a median OS of 9 to 11 months.⁴¹ Maintenance chemotherapy beyond the initial 4 to 6 cycles does not improve survival and is associated with higher cumulative toxicity.⁴²

 

Multiple attempts at improving first-line chemotherapy in extensive-stage disease have failed to show any meaningful difference in OS. For example, the addition of ifosfamide, palifosfamide, cyclophosphamide, taxane, or anthracycline to platinum doublet failed to show improvement in OS and led to more toxicity.^43–46 Additionally, the use of alternating or cyclic chemotherapies in an attempt to curb drug resistance has also failed to show survival benefit.^47–49 The addition of the antiangiogenic agent bevacizumab to standard platinum-based doublet has not prolonged OS in SCLC and has led to an unacceptably higher rate of tracheoesophageal fistula when used in conjunction with chemoradiation in limited-stage disease.^50–55 Finally, the immune checkpoint inhibitor ipilimumab in combination with platinum plus etoposide failed to improve PFS or OS compared to platinum plus etoposide alone in a recent phase 3 trial, and maintenance pembrolizumab after completion of platinum-based chemotherapy did not improve PFS.^56,57

More recently, a phase 2 study of pembrolizumab in extensive-stage SCLC (KEYNOTE 158) reported an overall response rate of 35.7%, median PFS of 2.1 months, and median OS of 14.6 months in patients who tested positive for programmed death ligand-1 (PD-L1) expression (which was defined as a PD-L1 Combined Positive Score ≥ 1).⁵⁸ The median duration of response has not been reached in this study, indicating that pembrolizumab may be a promising approach in patients with extensive-stage SCLC, especially for those with PD-L1–positive tumors.

Patients with extensive-stage disease who have brain metastasis at the time of diagnosis can be treated with systemic chemotherapy first if the brain metastases are asymptomatic and there is significant extracranial disease burden. In that case, whole brain radiotherapy should be given after completion of systemic therapy.

SECOND-LINE CHEMOTHERAPY

Despite being exquisitely chemosensitive, SCLC is associated with a very poor prognosis largely because of invariable disease progression following first-line therapy and lack of effective second-line treatment options that can lead to appreciable disease control. The choice of second-line treatment is predominantly determined by the time of disease relapse after first-line platinum-based therapy. If this interval is 6 months or longer, re-treatment utilizing the same platinum doublet is appropriate. However, if the interval is 6 months or less, second-line systemic therapy options should be explored. Unfortunately, the response rate tends to be less than 10% with most of the second-line therapies in platinum-resistant disease (defined as disease progression within 3 months of receiving platinum-based therapy). If disease progression occurs between 3 and 6 months after completion of platinum-based therapy, the response rate with second-line chemotherapy is in the range of 25%.^59,60

A number of second-line chemotherapy options have been explored in small studies, including topotecan, irinotecan, paclitaxel, docetaxel, temozolomide, vinorelbine, oral etoposide, gemcitabine, bendamustine, and CAV (cyclophosphamide, doxorubicin, vincristine) (Table 2).^61–77 Of these, topotecan (oral or intravenous) is the only second-line agent approved by the US Food and Drug Administration (FDA) for extensive-stage SCLC; this approval is based on a phase 3 trial showing a survival benefit with topotecan compared to best supportive care. The duration of second-line chemotherapy is not defined and is largely driven by duration of response and occurrence of side effects. Given the lack of effective treatment options, patients with platinum-resistant disease should be considered for clinical trials whenever possible. In patients with disease progression after second-line therapy, a third-line treatment can be considered if their Eastern Cooperative Oncology Group (ECOG) performance status is 0 to 2, but it is unclear whether further treatments would improve survival.

 

 

IMMUNOTHERAPY

The role of immune checkpoint inhibitors in the treatment of SCLC is evolving, and currently there are no FDA-approved immunotherapy agents for treating SCLC. A recently conducted phase 1/2 trial (CheckMate 032) studied the anti-programmed death(PD)-1 antibody nivolumab with or without the anti-cytotoxic T-lymphocyte–associated antigen (CTLA) -4 antibody ipilimumab in patients with relapsed SCLC. The authors reported response rates of 10% with nivolumab 3 mg/kg and 21% with nivolumab 1 mg/kg plus ipilimumab 3 mg/kg.^78,79 The 2-year OS was 26% with the combination and 14% with single-agent nivolumab. Only 18% of patients had PD-L1 expression of ≥ 1%, and the response rate did not correlate with PD-L1 status. The rate of grade 3 or 4 adverse events was approximately 20%, and only 10% of patients discontinued treatment because of toxicity. Based on these data, nivolumab plus ipilimumab is now included in the National Comprehensive Cancer Network guidelines as an option for patients with SCLC who experience disease relapse within 6 months of receiving platinum-based therapy;⁷ however, it is questionable whether routine use of this combination is justified based on currently available data. The evidence for the combination of nivolumab and ipilimumab remains limited. The efficacy and toxicity data from both randomized and nonrandomized cohorts were presented together, making it hard to interpret the results.

Another phase 1b study (KEYNOTE-028) evaluated the anti-PD-1 antibody pembrolizumab (10 mg/kg intravenously every 2 weeks) in patients with relapsed SCLC who had received 1 or more prior lines of therapy and had PD-L1 expression of ≥ 1%. This study showed a response rate of 33%, with a median duration of response of 19 months and 1-year OS of 38%.⁸⁰ Although only 28% of screened patients had PD-L1 expression of ≥ 1%, these results indicated that at least a subset of SCLC patients are able to achieve durable responses with immune checkpoint inhibition. A number of clinical trials utilizing immune checkpoint inhibitors in various combinations and settings are currently underway.

ROLE OF PROPHYLACTIC CRANIAL IRRADIATION

The role of PCI in extensive-stage SCLC is not clearly defined. A randomized phase 3 trial conducted by the European Organization for Research and Treatment of Cancer (EORTC) comparing PCI with no PCI in patients with extensive-stage SCLC who had a partial or complete response to initial platinum-based chemotherapy showed a decrease in the incidence of symptomatic brain metastasis and improvement in 1-year OS with PCI.⁸¹ However, this trial did not require mandatory brain imaging prior to PCI, and thus it is unclear if some patients in the PCI group had asymptomatic brain metastasis prior to enrollment and therefore received therapeutic benefit from brain radiation. Additionally, the dose and fractionation of PCI was not standardized across patient groups.

A more recent phase 3 study conducted in Japan that compared PCI (25 Gy in 10 fractions) with no PCI reported no difference in survival between the 2 groups.⁸² As opposed to the EORTC study, the Japanese study did require baseline brain imaging to confirm the absence of brain metastasis prior to enrollment. In addition, the control patients underwent periodic brain MRI to allow early detection of brain metastasis. Given the emergence of the new data, the impact of PCI on survival in patients with extensive-stage SCLC is unproven, and PCI likely has a role in a highly selected small group of patients with extensive-stage SCLC. PCI is not recommended for patients with poor performance status (ECOG performance score of 3 or 4) or underlying neurocognitive disorders.^34,83

The NMDA-receptor antagonist memantine can be used in patients undergoing PCI to delay the occurrence of cognitive dysfunction.⁶¹ Memantine 20 mg daily delayed time to cognitive decline and reduced the rate of decline in memory, executive function, and processing speed compared to placebo in patients receiving whole brain radiotherapy.⁸⁴

 

 

ROLE OF RADIOTHERAPY

A subset of patients with extensive-stage SCLC may benefit from consolidative thoracic radiotherapy after completion of platinum-based chemotherapy. A randomized trial that enrolled patients who achieved complete or near complete response after 3 cycles of cisplatin plus etoposide compared thoracic radiotherapy in combination with continued chemotherapy versus chemotherapy alone.⁸⁵ The median OS was longer with the addition of thoracic radiotherapy compared to chemotherapy alone. Another phase 3 trial did not show improvement in 1-year OS with consolidative thoracic radiotherapy, but 2-year OS and 6-month PFS were longer.⁸⁶ In general, consolidative thoracic radiotherapy benefits patients who have residual thoracic disease and low-bulk extrathoracic disease that has responded to systemic therapy.⁸⁷ In addition, patients who initially presented with bulky symptomatic thoracic disease should also be considered for consolidative radiation.

Similar to other solid tumors, radiotherapy should be utilized for palliative purposes in patients with painful bone metastasis, spinal cord compression, or brain metastasis. Surgery is generally not recommended for spinal cord compression given the short life expectancy of patients with extensive-stage disease. Whole brain radiotherapy is preferred over stereotactic radiosurgery because micrometastasis is frequently present even in the setting of 1 or 2 radiographically evident brain metastasis.

NOVEL THERAPIES

The very complex genetic landscape of SCLC accounts for its resistance to conventional therapy and high recurrence rate; however, at the same time this complexity can form the basis for effective targeted therapy for the disease. One of the major factors hindering the development of targeted therapies in SCLC is limited availability of tissue due to small tissue samples and the frequent presence of significant necrosis in the samples. In recent years, several different therapeutic strategies and targeted agents have been investigated for their potential role in SCLC. Several of them, including EGFR tyrosine kinase inhibitors (TKIs), BCR-ABL TKIs, mTOR inhibitors, and VEGF inhibitors, have not been shown to provide a survival advantage in this disease. Several others, including PARP inhibitors, cellular developmental pathway inhibitors, and antibody-drug conjugates, are being tested. A phase 1 study of veliparib combined with cisplatin and etoposide in patients with previously untreated extensive-stage SCLC demonstrated a complete response in 14.3%, a partial response in 57.1%, and stable disease in 28.6% of patients with an acceptable safety profile.⁸⁸ So far, none of these agents are approved for use in SCLC, and the majority are in early- phase clinical trials.⁸⁹

One of the emerging targets in the treatment of SCLC is delta-like protein 3 (DLL3). DLL3 is expressed on more than 80% of SCLC tumor cells and cancer stem cells. Rovalpituzumab tesirine is an antibody-drug conjugate consisting of humanized anti-DLL3 monoclonal antibody linked to SC-DR002, a DNA-crosslinking agent. A phase 1 trial of rovalpituzumab in patients with relapsed SCLC after 1 or 2 prior lines of therapy reported a response rate of 31% in patients with DLL3 expression of ≥ 50%. The median duration of response and median PFS were both 4.6 months.⁹⁰ Rovalpituzumab is currently in later phases of clinical trials and has a potential to serve as an option for patients with extensive-stage disease after disease progression on platinum-based therapy.

SUMMARY

Four to 6 cycles of carboplatin and etoposide remain the standard of care first-line treatment for patients with extensive stage SCLC. The only FDA-approved second-line treatment option is topotecan. Re-treatment with the original platinum doublet is a reasonable option for patients who have disease progression 6 months or longer after completion of platinum-based therapy. The immune checkpoint inhibitors pembrolizumab and combination nivolumab and ipilimumab have shown promising results in the second-line setting and beyond. The role of PCI has become more controversial in recent years, and periodic brain MRI in lieu of PCI is now an acceptable approach.

 

 

RESPONSE ASSESSMENT/SURVEILLANCE

For patients undergoing treatment for limited-stage SCLC, response assessment with contrast-enhanced CT of the chest/abdomen should be performed after completion of 4 cycles of chemotherapy and thoracic radiation.⁷ The surveillance guidelines consist of history, physical exam, and imaging every 3 months during the first 2 years, every 6 months during the third year, and annually thereafter. If PCI is not performed, brain MRI or contrast-enhanced CT scan should be performed every 3 or 4 months during the first 2 years of follow up. For extensive-stage disease, response assessment should be performed after every 2 cycles of therapy. After completion of therapy, history, physical exam, and imaging should be done every 2 months during the first year, every 3 or 4 months during years 2 and 3, every 6 months during years 4 and 5, and annually thereafter. Routine use of PET scan for surveillance is not recommended. Any new pulmonary nodule should prompt evaluation for a second primary lung malignancy. Finally, smoking cessation counseling is an integral part of management of any patient with SCLC and should be included with every clinic visit.

CONCLUSION

SCLC is a heterogeneous and genetically complex disease with a very high mortality rate. The current standard of care includes concurrent chemoradiation with cisplatin and etoposide for limited-stage SCLC and the combination of platinum and etoposide for extensive SCLC. A number of novel treatment approaches, including immune checkpoint inhibitors and antibody-drug conjugates, have had promising results in early clinical trials. Given the limited treatment options and large unmet need for new treatment options, enrollment in clinical trials is strongly recommended for patients with SCLC.

INTRODUCTION

Small cell lung cancer (SCLC) is an aggressive cancer of neuroendocrine origin that accounts for approximately 15% of all lung cancer cases, with approximately 33,000 patients diagnosed annually.¹ The incidence of SCLC in the United States has steadily declined over the past 30 years, presumably because of a decrease in the number of smokers and a change to low-tar filter cigarettes.² Although the overall incidence of SCLC has been decreasing, the incidence in women is increasing and the male-to-female incidence ratio is now 1:1.³ Nearly all cases of SCLC are associated with heavy tobacco exposure, making it a heterogeneous disease with a complex genomic landscape consisting of thousands of mutations.^4,5 Despite recent advances in the treatment of non-small cell lung cancer, the therapeutic options for SCLC remain limited, with a median overall survival (OS) of 9 months in patients with advanced disease.

DIAGNOSIS AND STAGING

CASE PRESENTATION

A 61-year-old man presents to the emergency department with progressive shortness of breath and cough over the past 6 weeks. He also reports a 20-lb weight loss over the same period. He is a current smoker and has been smoking 1 pack of cigarettes per day since the age of 18 years. A chest radiograph obtained in the emergency department shows a right hilar mass. Computed tomography (CT) scan confirms the presence of a 4.5-cm right hilar mass and enlarged mediastinal lymph nodes bilaterally.

• What are the next steps in diagnosis?

SCLC is characterized by rapid growth and early hematogenous metastasis. Consequently, only 25% of patients have limited-stage disease at the time of diagnosis. According to the Veterans Administration Lung Study Group (VALSG) staging system, limited-stage disease is defined as tumor that is confined to 1 hemithorax and can be encompassed within 1 radiation field. This typically includes mediastinal lymph nodes and ipsilateral supraclavicular lymph nodes. Approximately 75% of patients present with extensive-stage disease, which is defined as disease that cannot be classified as limited, including disease that extends beyond 1 hemithorax. Extensive-stage disease includes the presence of malignant pleural effusion and/or distant metastasis.⁶ The VALSG classification and staging system is more commonly used in clinical practice than the American Joint Committee on Cancer TNM staging system because it is less complex and directs treatment decisions, as most of the literature on SCLC classifies patients based on the VALSG system.⁷

Given SCLC’s propensity to metastasize quickly, none of the currently available screening methods have proven successful in early detection of SCLC. In the National Lung Cancer Screening Trial, 86% of the 125 patients who were diagnosed with SCLC while undergoing annual low-dose chest CT scans had advanced disease at diagnosis.^8,9 These results highlight the fact that most cases of SCLC develop in the interval between annual screening imaging.

SCLC frequently presents with a large hilar mass that is symptomatic. Common symptoms include shortness of breath and cough. In addition, patients with SCLC usually have bulky mediastinal adenopathy at presentation. SCLC is commonly located submucosally in the bronchus, and therefore hemoptysis is not a very common symptom at the time of presentation. Patients may present with superior vena cava syndrome from local compression by the tumor. Not infrequently, SCLC is associated with paraneoplastic syndromes that arise due to ectopic secretion of hormones or antibodies by the tumor cells. The paraneoplastic syndromes can be broadly categorized as endocrine or neurologic (Table 1). The presence of a paraneoplastic syndrome is often a clue to the potential diagnosis of SCLC in the presence of a hilar mass. Additionally, some paraneoplastic syndromes, more specifically endocrine paraneoplastic syndromes, follow the pattern of disease response and relapse, and therefore can sometimes serve as an early marker of disease relapse or progression.

The common sites of metastases include brain, liver, and bone. Therefore, the staging workup should include fluorodeoxyglucose (FDG) positron emission tomography (PET)/CT scan. Contrast-enhanced CT scan of the chest and abdomen and bone scan can be obtained for staging in lieu of PET scan. Due to the physiologic FDG uptake, cerebral metastases cannot be assessed with sufficient certainty using PET-CT.¹⁰ Therefore, brain imaging with contrast-enhanced CT or magnetic resonance imaging (MRI) is also necessary. Although the incidence of metastasis to bone marrow is less than 10%, bone marrow aspiration and biopsy are warranted in patients with unexplained cytopenias, especially when the cytopenia is associated with teardrop-shaped red cells or nucleated red cells on peripheral blood smear, findings indicative of a marrow infiltrative process.⁷ The tissue diagnosis is established by obtaining a biopsy of the primary tumor or 1 of the metastatic sites. In localized disease, bronchoscopy (with endobronchial ultrasound, if necessary) with biopsy of the centrally located tumor and/or lymph node is required. Histologically, SCLC consists of monomorphic cells, a high nuclear-cytoplasmic ratio, and confluent necrosis. The tumor cells are positive for chromogranin, synaptophysin, and CD56 by immunohistochemistry, and very frequently are also positive for thyroid transcription factor 1.¹¹ Although serum tumor markers, including neuron-specific enolase and progastrin-releasing peptide, are frequently elevated in patients with SCLC, these markers are of limited value in clinical practice because they lack sensitivity and specificity.¹²

 

MANAGEMENT OF LIMITED-STAGE DISEASE

CASE CONTINUED

The patient undergoes FDG PET scan, which shows the presence of a hypermetabolic right hilar mass in addition to enlarged and hypermetabolic bilateral mediastinal lymph nodes. There are no other areas of FDG avidity. Brain MRI does not show any evidence of brain metastasis. Thus, the patient is confirmed to have limited-stage SCLC.

• What is the standard of care treatment for limited-stage SCLC?

SCLC is exquisitely sensitive to both chemotherapy and radiation, especially at the time of initial presentation. The standard of care treatment of limited-stage SCLC is 4 cycles of platinum-based chemotherapy in combination with thoracic radiation started within the first 2 cycles of chemotherapy (Figure 1).

This regimen yields an overall response rate between 75% and 90% and a complete response rate of 50%, with a median OS of 18 to 24 months.¹³ The cure rate with this approach is approximately 25%, with the remaining 75% of patients experiencing disease relapse within the first 5 years after completion of treatment. Attempts to improve the cure rate in the limited-stage setting by combining standard of care therapies with novel agents have been unsuccessful.

CHOICE OF CHEMOTHERAPY

Etoposide and cisplatin is the most commonly used initial combination chemotherapy regimen in limited-stage SCLC.¹⁴ This combination has largely replaced anthracycline-based regimens given its favorable efficacy and toxicity profile.^15–17 Several small randomized trials have shown comparable efficacy of carboplatin and etoposide in extensive-stage SCLC.^18–20 A meta-analysis of 4 randomized trials comparing cisplatin-based versus carboplatin-based regimens in 663 patients with SCLC (32% had limited-stage disease and 68% had extensive-stage disease) showed no statistically significant difference in response rate, progression-free survival (PFS), or OS between the 2 regimens.²¹ Therefore, in clinical practice carboplatin is frequently used instead of cisplatin in patients with extensive-stage disease. In patients with limited-stage disease, cisplatin is still the drug of choice. However, the toxicity profile of the 2 regimens is different. Cisplatin-based regimens are more commonly associated with neuropathy, nephrotoxicity, and chemotherapy-induced nausea/vomiting,¹⁸ while carboplatin-based regimens are more myelosuppressive.²² In addition, the combination of thoracic radiation with either of these regimens is associated with a higher risk of esophagitis, pneumonitis, and myelosuppression.²³ The use of myeloid growth factors is not recommended in patients undergoing concurrent chemoradiation.²⁴ Of note, intravenous etoposide is always preferred over oral etoposide, especially in the curative setting given the unreliable absorption and bioavailability of oral formulations.

THORACIC RADIOTHERAPY

Adding thoracic radiotherapy to platinum-etoposide chemotherapy improves local control and OS. Two meta-analyses of 13 trials including more than 2000 patients have shown a 25% to 30% decrease in local failure and a 5% to 7% increase in 2-year OS with chemoradiation compared to chemotherapy alone in limited-stage SCLC.^25,26 Early (within the first 2 cycles) concurrent thoracic radiation is superior to delayed and/or sequential radiation in terms of local control and OS.^23,27,28 The dose and fractionation of thoracic radiation in limited-stage SCLC has remained a controversial issue. The Eastern Cooperative Oncology Group/Radiation Therapy Oncology Group randomized trial compared 45 Gy of radiotherapy delivered twice daily over a period of 3 weeks to 45 Gy once daily over 5 weeks concurrently with chemotherapy. The twice daily regimen led to a 10% improvement in 5-year OS (26% versus 16%), but a higher incidence of grade 3 and 4 adverse events.¹³ Despite the survival advantage demonstrated by hyperfractionated radiotherapy, the results need to be interpreted with caution because the radiation doses are not biologically equivalent. In addition, the difficult logistics of patients receiving radiation twice a day has limited the routine implementation of this strategy. Subsequently, another randomized phase 3 trial (CONVERT) compared 45 Gy radiotherapy twice daily with 66 Gy radiotherapy once daily in limited-stage SCLC.²⁹ This trial did not show any difference in OS. The patients in the twice daily arm had a higher incidence of grade 4 neutropenia. Considering the results of these trials, both strategies—45 Gy fractionated twice daily or 60 Gy fractionated once daily, delivered concurrently with chemotherapy—are acceptable in the setting of limited-stage SCLC. However, quite often a hyperfractionated regimen is not feasible for patients and many radiation oncology centers. Hopefully, the ongoing CALGB 30610 study will clarify the optimal radiation schedule for limited-stage disease.

 

PROPHYLACTIC CRANIAL IRRADIATION

Approximately 75% of patients with limited-stage disease experience disease recurrence, and brain is the site of recurrence in approximately half of these patients.³⁰ Prophylactic cranial irradiation (PCI) consisting of 25 Gy radiotherapy delivered in 10 fractions has been shown to be effective in decreasing the incidence of cerebral metastases.^30–32 Although individual small studies have not shown a survival benefit of PCI because of small sample size and limited power, a meta-analysis of these studies has shown a 25% decrease in the 3-year incidence of brain metastasis and 5.4% increase in 3-year OS.³⁰ Most patients included in these studies had limited-stage disease. Therefore, PCI is the standard of care for patients with limited-stage disease who attain a partial or complete response to chemoradiation.

ROLE OF SURGERY

Surgical resection may be an acceptable choice in a very limited subset of patients with peripherally located small (< 5 cm) tumors where mediastinal lymph nodes have been confirmed to be uninvolved with complete mediastinal staging.^33,34 Most of the data in this setting are derived from retrospective studies.^35,36 A 5-year OS between 40% and 60% has been reported with this strategy in patients with clinical stage I disease. In general, when surgery is considered, lobectomy with mediastinal lymph node dissection followed by chemotherapy (if there is no nodal involvement) or chemoradiation (if nodal involvement) is recommended.^37,38 Wedge or segmental resections are not considered to be optimal surgical options.

MANAGEMENT OF EXTENSIVE-STAGE DISEASE

CASE CONTINUED

The patient receives 4 cycles of cisplatin and etoposide along with 70 Gy radiotherapy concurrently with the first 2 cycles of chemotherapy. His post-treatment CT scans show a partial response. He undergoes PCI 6 weeks after completion of treatment. At routine follow-up 18 months later, he is doing generally well except for mildly decreased appetite and an unintentional weight loss of 5 lb. CT scans demonstrate multiple hypodense liver lesions ranging from 7 mm to 2 cm in size and a 2-cm left adrenal gland lesion highly concerning for metastasis. FDG PET scan confirms that the adrenal and liver lesions are hypermetabolic. In addition, the PET scan shows multiple FDG-avid bone lesions throughout the spine. Brain MRI is negative for brain metastasis.

• What is the standard of care for treatment of extensive-stage disease?

Chemotherapy is the mainstay of treatment for extensive-stage SCLC; the goals of treatment are prolongation of survival, prevention or alleviation of cancer-related symptoms, and improvement in quality of life. The combination of etoposide with a platinum agent (carboplatin or cisplatin) is the preferred first-line treatment option. Carboplatin is more commonly used in clinical practice in this setting because of its comparable efficacy and better tolerability compared to cisplatin (Figure 2).²¹ A Japanese phase 3 trial comparing cisplatin plus irinotecan with cisplatin plus etoposide in the first-line setting in extensive-stage SCLC showed improvement in median and 2-year OS with the cisplatin/irinotecan regimen; however, 2 subsequent phase 3 trials conducted in the United States comparing these 2 regimens did not show any difference in OS. In addition, the cisplatin/irinotecan regimen was more toxic than the etoposide-based regimen.^39,40 Therefore, 4 to 6 cycles of platinum/etoposide remains the standard of care first-line treatment for extensive-stage SCLC in the United States. The combination yields a 60% to 70% response rate, but the majority of patients invariably experience disease progression, with a median OS of 9 to 11 months.⁴¹ Maintenance chemotherapy beyond the initial 4 to 6 cycles does not improve survival and is associated with higher cumulative toxicity.⁴²

 

Multiple attempts at improving first-line chemotherapy in extensive-stage disease have failed to show any meaningful difference in OS. For example, the addition of ifosfamide, palifosfamide, cyclophosphamide, taxane, or anthracycline to platinum doublet failed to show improvement in OS and led to more toxicity.^43–46 Additionally, the use of alternating or cyclic chemotherapies in an attempt to curb drug resistance has also failed to show survival benefit.^47–49 The addition of the antiangiogenic agent bevacizumab to standard platinum-based doublet has not prolonged OS in SCLC and has led to an unacceptably higher rate of tracheoesophageal fistula when used in conjunction with chemoradiation in limited-stage disease.^50–55 Finally, the immune checkpoint inhibitor ipilimumab in combination with platinum plus etoposide failed to improve PFS or OS compared to platinum plus etoposide alone in a recent phase 3 trial, and maintenance pembrolizumab after completion of platinum-based chemotherapy did not improve PFS.^56,57

More recently, a phase 2 study of pembrolizumab in extensive-stage SCLC (KEYNOTE 158) reported an overall response rate of 35.7%, median PFS of 2.1 months, and median OS of 14.6 months in patients who tested positive for programmed death ligand-1 (PD-L1) expression (which was defined as a PD-L1 Combined Positive Score ≥ 1).⁵⁸ The median duration of response has not been reached in this study, indicating that pembrolizumab may be a promising approach in patients with extensive-stage SCLC, especially for those with PD-L1–positive tumors.

Patients with extensive-stage disease who have brain metastasis at the time of diagnosis can be treated with systemic chemotherapy first if the brain metastases are asymptomatic and there is significant extracranial disease burden. In that case, whole brain radiotherapy should be given after completion of systemic therapy.

SECOND-LINE CHEMOTHERAPY

Despite being exquisitely chemosensitive, SCLC is associated with a very poor prognosis largely because of invariable disease progression following first-line therapy and lack of effective second-line treatment options that can lead to appreciable disease control. The choice of second-line treatment is predominantly determined by the time of disease relapse after first-line platinum-based therapy. If this interval is 6 months or longer, re-treatment utilizing the same platinum doublet is appropriate. However, if the interval is 6 months or less, second-line systemic therapy options should be explored. Unfortunately, the response rate tends to be less than 10% with most of the second-line therapies in platinum-resistant disease (defined as disease progression within 3 months of receiving platinum-based therapy). If disease progression occurs between 3 and 6 months after completion of platinum-based therapy, the response rate with second-line chemotherapy is in the range of 25%.^59,60

A number of second-line chemotherapy options have been explored in small studies, including topotecan, irinotecan, paclitaxel, docetaxel, temozolomide, vinorelbine, oral etoposide, gemcitabine, bendamustine, and CAV (cyclophosphamide, doxorubicin, vincristine) (Table 2).^61–77 Of these, topotecan (oral or intravenous) is the only second-line agent approved by the US Food and Drug Administration (FDA) for extensive-stage SCLC; this approval is based on a phase 3 trial showing a survival benefit with topotecan compared to best supportive care. The duration of second-line chemotherapy is not defined and is largely driven by duration of response and occurrence of side effects. Given the lack of effective treatment options, patients with platinum-resistant disease should be considered for clinical trials whenever possible. In patients with disease progression after second-line therapy, a third-line treatment can be considered if their Eastern Cooperative Oncology Group (ECOG) performance status is 0 to 2, but it is unclear whether further treatments would improve survival.

 

 

IMMUNOTHERAPY

The role of immune checkpoint inhibitors in the treatment of SCLC is evolving, and currently there are no FDA-approved immunotherapy agents for treating SCLC. A recently conducted phase 1/2 trial (CheckMate 032) studied the anti-programmed death(PD)-1 antibody nivolumab with or without the anti-cytotoxic T-lymphocyte–associated antigen (CTLA) -4 antibody ipilimumab in patients with relapsed SCLC. The authors reported response rates of 10% with nivolumab 3 mg/kg and 21% with nivolumab 1 mg/kg plus ipilimumab 3 mg/kg.^78,79 The 2-year OS was 26% with the combination and 14% with single-agent nivolumab. Only 18% of patients had PD-L1 expression of ≥ 1%, and the response rate did not correlate with PD-L1 status. The rate of grade 3 or 4 adverse events was approximately 20%, and only 10% of patients discontinued treatment because of toxicity. Based on these data, nivolumab plus ipilimumab is now included in the National Comprehensive Cancer Network guidelines as an option for patients with SCLC who experience disease relapse within 6 months of receiving platinum-based therapy;⁷ however, it is questionable whether routine use of this combination is justified based on currently available data. The evidence for the combination of nivolumab and ipilimumab remains limited. The efficacy and toxicity data from both randomized and nonrandomized cohorts were presented together, making it hard to interpret the results.

Another phase 1b study (KEYNOTE-028) evaluated the anti-PD-1 antibody pembrolizumab (10 mg/kg intravenously every 2 weeks) in patients with relapsed SCLC who had received 1 or more prior lines of therapy and had PD-L1 expression of ≥ 1%. This study showed a response rate of 33%, with a median duration of response of 19 months and 1-year OS of 38%.⁸⁰ Although only 28% of screened patients had PD-L1 expression of ≥ 1%, these results indicated that at least a subset of SCLC patients are able to achieve durable responses with immune checkpoint inhibition. A number of clinical trials utilizing immune checkpoint inhibitors in various combinations and settings are currently underway.

ROLE OF PROPHYLACTIC CRANIAL IRRADIATION

The role of PCI in extensive-stage SCLC is not clearly defined. A randomized phase 3 trial conducted by the European Organization for Research and Treatment of Cancer (EORTC) comparing PCI with no PCI in patients with extensive-stage SCLC who had a partial or complete response to initial platinum-based chemotherapy showed a decrease in the incidence of symptomatic brain metastasis and improvement in 1-year OS with PCI.⁸¹ However, this trial did not require mandatory brain imaging prior to PCI, and thus it is unclear if some patients in the PCI group had asymptomatic brain metastasis prior to enrollment and therefore received therapeutic benefit from brain radiation. Additionally, the dose and fractionation of PCI was not standardized across patient groups.

A more recent phase 3 study conducted in Japan that compared PCI (25 Gy in 10 fractions) with no PCI reported no difference in survival between the 2 groups.⁸² As opposed to the EORTC study, the Japanese study did require baseline brain imaging to confirm the absence of brain metastasis prior to enrollment. In addition, the control patients underwent periodic brain MRI to allow early detection of brain metastasis. Given the emergence of the new data, the impact of PCI on survival in patients with extensive-stage SCLC is unproven, and PCI likely has a role in a highly selected small group of patients with extensive-stage SCLC. PCI is not recommended for patients with poor performance status (ECOG performance score of 3 or 4) or underlying neurocognitive disorders.^34,83

The NMDA-receptor antagonist memantine can be used in patients undergoing PCI to delay the occurrence of cognitive dysfunction.⁶¹ Memantine 20 mg daily delayed time to cognitive decline and reduced the rate of decline in memory, executive function, and processing speed compared to placebo in patients receiving whole brain radiotherapy.⁸⁴

 

 

ROLE OF RADIOTHERAPY

A subset of patients with extensive-stage SCLC may benefit from consolidative thoracic radiotherapy after completion of platinum-based chemotherapy. A randomized trial that enrolled patients who achieved complete or near complete response after 3 cycles of cisplatin plus etoposide compared thoracic radiotherapy in combination with continued chemotherapy versus chemotherapy alone.⁸⁵ The median OS was longer with the addition of thoracic radiotherapy compared to chemotherapy alone. Another phase 3 trial did not show improvement in 1-year OS with consolidative thoracic radiotherapy, but 2-year OS and 6-month PFS were longer.⁸⁶ In general, consolidative thoracic radiotherapy benefits patients who have residual thoracic disease and low-bulk extrathoracic disease that has responded to systemic therapy.⁸⁷ In addition, patients who initially presented with bulky symptomatic thoracic disease should also be considered for consolidative radiation.

Similar to other solid tumors, radiotherapy should be utilized for palliative purposes in patients with painful bone metastasis, spinal cord compression, or brain metastasis. Surgery is generally not recommended for spinal cord compression given the short life expectancy of patients with extensive-stage disease. Whole brain radiotherapy is preferred over stereotactic radiosurgery because micrometastasis is frequently present even in the setting of 1 or 2 radiographically evident brain metastasis.

NOVEL THERAPIES

The very complex genetic landscape of SCLC accounts for its resistance to conventional therapy and high recurrence rate; however, at the same time this complexity can form the basis for effective targeted therapy for the disease. One of the major factors hindering the development of targeted therapies in SCLC is limited availability of tissue due to small tissue samples and the frequent presence of significant necrosis in the samples. In recent years, several different therapeutic strategies and targeted agents have been investigated for their potential role in SCLC. Several of them, including EGFR tyrosine kinase inhibitors (TKIs), BCR-ABL TKIs, mTOR inhibitors, and VEGF inhibitors, have not been shown to provide a survival advantage in this disease. Several others, including PARP inhibitors, cellular developmental pathway inhibitors, and antibody-drug conjugates, are being tested. A phase 1 study of veliparib combined with cisplatin and etoposide in patients with previously untreated extensive-stage SCLC demonstrated a complete response in 14.3%, a partial response in 57.1%, and stable disease in 28.6% of patients with an acceptable safety profile.⁸⁸ So far, none of these agents are approved for use in SCLC, and the majority are in early- phase clinical trials.⁸⁹

One of the emerging targets in the treatment of SCLC is delta-like protein 3 (DLL3). DLL3 is expressed on more than 80% of SCLC tumor cells and cancer stem cells. Rovalpituzumab tesirine is an antibody-drug conjugate consisting of humanized anti-DLL3 monoclonal antibody linked to SC-DR002, a DNA-crosslinking agent. A phase 1 trial of rovalpituzumab in patients with relapsed SCLC after 1 or 2 prior lines of therapy reported a response rate of 31% in patients with DLL3 expression of ≥ 50%. The median duration of response and median PFS were both 4.6 months.⁹⁰ Rovalpituzumab is currently in later phases of clinical trials and has a potential to serve as an option for patients with extensive-stage disease after disease progression on platinum-based therapy.

SUMMARY

Four to 6 cycles of carboplatin and etoposide remain the standard of care first-line treatment for patients with extensive stage SCLC. The only FDA-approved second-line treatment option is topotecan. Re-treatment with the original platinum doublet is a reasonable option for patients who have disease progression 6 months or longer after completion of platinum-based therapy. The immune checkpoint inhibitors pembrolizumab and combination nivolumab and ipilimumab have shown promising results in the second-line setting and beyond. The role of PCI has become more controversial in recent years, and periodic brain MRI in lieu of PCI is now an acceptable approach.

 

 

RESPONSE ASSESSMENT/SURVEILLANCE

For patients undergoing treatment for limited-stage SCLC, response assessment with contrast-enhanced CT of the chest/abdomen should be performed after completion of 4 cycles of chemotherapy and thoracic radiation.⁷ The surveillance guidelines consist of history, physical exam, and imaging every 3 months during the first 2 years, every 6 months during the third year, and annually thereafter. If PCI is not performed, brain MRI or contrast-enhanced CT scan should be performed every 3 or 4 months during the first 2 years of follow up. For extensive-stage disease, response assessment should be performed after every 2 cycles of therapy. After completion of therapy, history, physical exam, and imaging should be done every 2 months during the first year, every 3 or 4 months during years 2 and 3, every 6 months during years 4 and 5, and annually thereafter. Routine use of PET scan for surveillance is not recommended. Any new pulmonary nodule should prompt evaluation for a second primary lung malignancy. Finally, smoking cessation counseling is an integral part of management of any patient with SCLC and should be included with every clinic visit.

CONCLUSION

SCLC is a heterogeneous and genetically complex disease with a very high mortality rate. The current standard of care includes concurrent chemoradiation with cisplatin and etoposide for limited-stage SCLC and the combination of platinum and etoposide for extensive SCLC. A number of novel treatment approaches, including immune checkpoint inhibitors and antibody-drug conjugates, have had promising results in early clinical trials. Given the limited treatment options and large unmet need for new treatment options, enrollment in clinical trials is strongly recommended for patients with SCLC.

INTRODUCTION

According to the Surveillance, Epidemiology and End Results database, in 2017 there were 28,000 new cases of gastric cancer, accounting for 1.8% of all malignancies in the United States, and an estimated 10,960 gastric cancer–related deaths.¹ Worldwide, gastric cancer is the fifth most common malignancy and the third most common cause of death from any cancer.² The incidence of gastric cancer varies significantly by geographic region, with countries in Eastern Asia (China, Japan), Eastern Europe, and Central and South America accounting for 50% of all new cases.³ Although the incidence of gastric cancer has declined in recent years, this decrease has not been observed consistently across all nations.² In particular, the incidence of gastric cancers arising from the cardia has been increasing, which is perhaps due to a higher prevalence of obesity in Western societies.⁴

In this article, we review key aspects of management of metastatic gastric cancer, including selection of first- and second-line therapy, and discuss targeted agents and upcoming clinical trials.

EPIDEMIOLOGY AND RISK FACTORS

Chronic infection with Helicobacter pylori, a gram-negative bacterium, is a strong etiological factor for the development of gastric cancer, contributing to up to 70% of cases.² The pathogen can colonize the gastric mucosa, leading to chronic inflammation. Although most patients remain asymptomatic, 1% to 3% develop gastric cancer and another 0.1% develop mucosa-associated lymphoid tissue lymphoma.⁵ H. pylori infection is more commonly associated with cancer of the gastric body than with cancer of the gastroesophageal junction (GEJ). The increased burden of gastric cancer in countries in Eastern Asia, Latin America, and Eastern Europe has been correlated to the prevalence of chronic H. pylori infection in these areas.

Carcinogenesis secondary to H. pylori infection may occur via several mechanisms. First, H. pylori can release virulence factors, such as cytotoxin-associated gene A, vacuolating cytotoxin, and outer membrane proteins, into the cytosol of host cells, leading to changes in patterns of cell proliferation and apoptosis.⁶ These virulence factors can modulate the host immune system, attenuating it to promote dysplasia. In addition, continued recognition of these factors by the immune system leads to a persistent inflammatory response, with the release of cytokines such as interleukin (IL) -1β, IL-6, and IL-8. This leads to chronic mucosal damage, further promoting dysplasia with eventual transformation into adenocarcinoma.⁷ In Japan and Korea, where screening for H. pylori infection is routinely performed, there have been improvements in overall survival (OS) rates for gastric cancer, with 5-year OS rates of 70%.⁸ The International Agency for Research on Cancer recommends further research into population-based screening and treatment programs for patients with chronic H. pylori infection. However, despite this recommendation, optimal screening strategies are not clearly defined.⁹

Other risk factors for the development of gastric cancer include chronic gastroesophageal reflux disease; smoking; alcohol use; exposure to radiation; diets high in fats, salt, and smoked items and low in fruits and vegetables; obesity; and exposure to chemotherapeutic agents such as procarbazine.¹⁰ Another pathogen suspected, but not proven, to be associated with increased risk for gastric cancer is the Epstein-Barr virus, a human herpesvirus found in 80% of all gastric carcinomas with lymphoid features.¹¹ In addition, whether the use of medications such as statins and nonsteroidal anti-inflammatory drugs confers a decreased risk of gastric cancers remains unclear.¹⁰

EVALUATION

CASE PRESENTATION

A 55-year-old Caucasian man with a history of type 2 diabetes mellitus presents to the gastrointestinal (GI) clinic with a 6-month history of dysphagia. The dysphagia is worsened with ingestion of solids, particularly towards the end of the day. He states that the food often gets “stuck in the middle of the chest.” The patient denies any nausea or emesis but notes that he has a poor appetite. He reports having worsening mid-epigastric abdominal pain that is non-radiating, dull in character, and 6/10 in intensity. He also reports a 10-lb weight loss over the past 2 months. He has no previous history of reflux, chest pain, dyspnea, or cough. Review of systems is otherwise benign. Physical exam is within normal limits.

• Which tests should be conducted when gastric cancer is suspected?

 

Persistent epigastric abdominal pain and weight loss are the most common early symptoms of gastric cancer. Nausea, early satiety, dysphagia, and occult GI bleeding can be other presenting signs. Patients presenting with alarm symptoms of nausea, emesis, early satiety, abdominal pain, or weight loss should be fully evaluated with upper GI endoscopy. Early diagnosis of gastric cancer is essential in obtaining a curative resection. However, at least 40% of patients present with de novo metastatic disease at the time of initial diagnosis.¹² Gastric cancer spreads by direct extension through the gastric wall, with the liver, peritoneum, and regional lymph nodes being the most common sites of metastatic deposits.¹³ Classically, Virchow’s node, the left supraclavicular lymph node, is involved with metastatic gastric cancer. Involvement of the left axillary lymph node (Irish node) or a periumbilical nodule (Sister Mary Joseph node) may also be observed. Other, less commonly noted sites of metastatic disease include the ovaries, central nervous system, bone, lung, and soft tissues.¹³

Upper GI endoscopy is the best method for determining tumor location and extent and obtaining a specimen for a definitive tissue diagnosis.¹⁴ It is essential to accurately identify the location of the tumor in the stomach and relative to the GEJ. The American Joint Committee on Cancer classification defines tumors involving the GEJ with an epicenter no more than 2 cm into the proximal stomach as esophageal cancers.¹⁵ Tumors of the GEJ with their epicenter more than 2 cm into the proximal stomach are defined as gastric cancers. If metastatic disease is suspected, computed tomography (CT) scan of the chest, abdomen, and pelvis with oral and intravenous contrast can be obtained to determine the extent of disease spread. In the absence of any metastatic disease, endoscopic ultrasound (EUS) should be conducted to determine the depth of tumor invasion (T staging) and lymph node status. In the era of targeted therapy, patients with metastatic disease should undergo testing for human epidermal growth factor-2 (HER-2) expression, microsatellite instability (MSI), and programmed death ligand 1 (PD-L1) expression. Patients should be staged according to the TNM staging system.

FIRST-LINE TREATMENT OPTIONS

CASE CONTINUED

The patient undergoes esophagoduodenoscopy (EGD) and is found to have a gastric cardia mass extending into the distal esophagus. EUS also demonstrates multiple abdominal and mediastinal lymph nodes. No gastric outlet obstruction is found. Biopsy shows poorly differentiated invasive adenocarcinoma. Warthin–Starry stain is negative for H. pylori organism. The tumor cells are positive for cytokeratin (CK7), CK19, and mucin-1 gene (MUC1); focally positive for CK20; and negative for MUC2. HER2 testing results are reported as immunohistochemistry (IHC) 3+, consistent with strongly positive HER2 protein expression. Further IHC testing for mismatch repair (MMR) proteins shows intact nuclear expression of MLH1, MSH2, MSH6, and PMS2 protein, consistent with a low probability of MSI-high tumor. The tumor is found to be PD-L1 positive. Imaging reveals abnormal mass-like nodular thickening of the gastric wall, with an infiltrative opacity within the pancreatico-duodenal groove, suspicious for tumor infiltration. Multiple metastatic deposits are noted in the liver, peritoneum, and bilateral lungs. There is extensive gastrohepatic ligament and periportal lymphadenopathy and mild enlargement of the pulmonary hilar lymph nodes. These findings are consistent with stage 4 (T4bN3aM1) gastric cancer. Given these findings, staging laparoscopy is deferred.

• What are the first-line treatment options for patients with metastatic gastric cancer?

Patients with metastatic gastric cancer have a poor prognosis, and management is stratified based on performance status (Figure). In patients with good performance status, systemic chemotherapy is the mainstay of treatment. The goal of therapy is not curative, but rather treatment focuses on palliation of symptoms arising from tumor spread. Given this treatment goal, there has been considerable interest in clarifying the utility of chemotherapy as opposed to best supportive care. In a recent Cochrane review of 64 randomized control trials involving 11,698 patients, chemotherapy was found to improve OS by 6.7 months as compared to best supportive care (hazard ratio [HR] 0.3 [95% confidence interval {CI} 0.24 to 0.55]).¹⁶ Five classes of cytotoxic chemotherapeutic agents have demonstrated activity in gastric cancer. These include fluoropyrimidine (either infusional fluorouracil or capecitabine), platinum agents (cisplatin or oxaliplatin), taxanes (docetaxel or paclitaxel), anthracyclines (epirubicin), and irinotecan.¹³ Treatment options are further divided based on whether the patient has HER2-overexpressing or non-expressing malignancy.

 

 

HER2-NEGATIVE DISEASE

For patients with HER2-negative disease, National Comprehensive Cancer Network (NCCN) guidelines recommend using 2-drug combination regimens rather than 3 drugs, given concern for increased toxicity with 3-drug regimens.¹⁷ For patients with a performance status of 0 to 1, utilization of a 3-drug regimen is a reasonable alternative. The combination of a fluoropyrimidine with a platinum agent is considered the standard of care, with regimens such as fluorouracil, leucovorin, and oxaliplatin (FOLFOX) being commonly used.

Epirubicin-containing regimens have also been extensively studied in advanced gastric cancer. In a study of 274 previously untreated patients with GEJ cancers, the combination of epirubicin, cisplatin, and fluorouracil (ECF) was compared to fluorouracil, doxorubicin, and methotrexate (FAMTX). There was an OS benefit favoring ECF (8.9 months versus 5.7 months) at 1 year (95% CI 27% to 45%, P = 0.0009). The ECF regimen was associated with an increased risk of nausea, emesis, and alopecia, while more hematologic toxicity and infections were noted with the FAMTX regimen.¹⁸ In addition, in a phase 3 trial, Van Cutsem and colleagues examined the role of docetaxel in combination with cisplatin and fluorouracil (DCF) compared to cisplatin and fluorouracil alone. Addition of docetaxel led to improved OS and time to progression (9.2 months versus 8.6 months for cisplatin and fluorouracil alone, P = 0.02) but with an increased risk of grade 3 and 4 toxicities (69% versus 59%). These adverse events included neutropenia (82% versus 57% of cisplatin and fluorouracil patients), diarrhea (19% versus 8%), stomatitis (21% versus 27%), and fatigue (19% versus 14%).¹⁹

The landmark phase 3 REAL-2 study compared 4 chemotherapy regimens in patients with untreated advanced esophagogastric cancer. This study was conducted to determine if the efficacy of cisplatin and oxaliplatin, a third-generation platinum agent, is equivalent to that of fluorouracil and capecitabine, an oral fluoropyrimidine. In this trial, a 2 × 2 design was used to compare 4 regimens: ECF versus epirubicin, cisplatin, and capecitabine (ECX) versus epirubicin, oxaliplatin, and fluorouracil (EOF) versus epirubicin, oxaliplatin, and capecitabine (EOX). The study found EOX to be noninferior to ECF, with a trend towards improved OS compared to other combination regimens (11.2 months versus 9.9 months, HR 0.80 [95% CI 0.66 to 0.97], P = 0.02).²⁰ Thus, the study demonstrated that an oxaliplatin and capecitabine-based regimen could replace cisplatin and fluorouracil. Given that fluorouracil administration requires long continuous infusions, the oral-based capecitabine regimen is an attractive option for patients.

Several trials have demonstrated the equivalency of oxaliplatin with cisplatin in combination regimens for the treatment of advanced gastric cancer. Oxaliplatin has the benefit of an improved toxicity profile as compared to cisplatin, with the major dose-limiting toxicity being peripheral neuropathy

Given previous evidence that DCF (docetaxel, cisplatin, fluorouracil) is superior to cisplatin and fluorouracil alone, there was interest in determining if the addition of docetaxel to a backbone of fluorouracil, oxaliplatin, and leucovorin (FLO) could elicit a higher response rate. This concept was investigated in a phase 2 trial that assigned 54 patients with metastatic gastric or GEJ adenocarcinoma to receive biweekly infusions of oxaliplatin, leucovorin, fluorouracil, and docetaxel.²¹ Median time to response was 1.54 months, and the overall response rate was 57.7%. Median progression-free survival (PFS) was 5.2 months, and OS was 11.1 months. The most common grade 3 or 4 toxicities included neutropenia (48%), leukopenia (27.8%), diarrhea (14.8%), and fatigue (11.1%).

 

Irinotecan-based regimens have also been extensively studied in the first-line treatment of metastatic gastric cancer, particularly as an alternative to platinum-based therapy, but superiority has not been established. The combination of fluorouracil, leucovorin, and irinotecan (FOLFIRI) was compared to ECX in a phase 3 trial.²² The study enrolled 416 patients with locally advanced or metastatic gastric or GEJ cancer. At a median follow up of 31 months, the time to progression was longer in the FOLFIRI arm as compared to the ECX arm (5.1 months versus 4.2 months, P = 0.008), but there was no difference in OS (9.5 months versus 9.7 months, P = 0.95), median PFS (5.3 months versus 5.8 months, P = 0.96), or response rate (39.2% versus 37.8%). However, the FOLFIRI regimen had an improved toxicity profile, with a lower overall rate of grade 3 or 4 toxicity (69% versus 84%, P < 0.001). Given these findings, the FOLFIRI regimen is an acceptable alternative to platinum-based therapy in suitable patients.²²

HER2-POSITIVE DISEASE

The HER2 proto-oncogene, initially described in breast cancer, has been implicated in several malignancies, including gastric and esophageal cancer. Overexpression or amplification of HER2 can be found in up to 30% of gastric cancers.²³ For these patients, adding trastuzumab to a standard regimen of platinum and fluoropyrimidine is the standard of care. The prospective phase 3 Trastuzumab for Gastric Cancer (ToGA) trial randomly assigned 594 patients with HER2-positive gastric cancer to receive either cisplatin and fluorouracil or capecitabine and cisplatin with trastuzumab (n = 294) or without (n = 290) trastuzumab every 3 weeks for a total of 6 cycles, followed by maintenance trastuzumab until disease progression was noted.²⁴ HER2 positivity was defined as HER2 protein overexpression by IHC (cutoff of 3+) or gene amplification by fluorescence in situ hybridization (FISH); tumors with IHC 2+ patterns were followed with FISH studies to confirm positivity. The study found a higher incidence of HER2-positive tumors in patients with GEJ tumors compared to patients with distal gastric cancers (33% versus 20%).²⁴ In this trial, the addition of trastuzumab was associated with an improvement in OS: 13.5 months in the trastuzumab cohort versus 11.1 months in those receiving chemotherapy alone (HR 0.74, P = 0.0048). There was not a significant difference in toxicities between the 2 cohorts, with nausea, emesis, and neutropenia being the most common adverse events. Rates of overall grade 3 or 4 events were similar as well (68% in each cohort). Further exploratory analysis was also conducted according to HER2 status by dividing patients into a “high-expressor” group (n = 446), defined as patients with IHC 3+ tumors or IHC 2+ and FISH positivity, and a “low-expressor” group (n = 131), which included patients with IHC 0 or 1+ tumors. Analysis of patients in the 2 subgroups demonstrated an improved OS with the addition of trastuzumab for the high-expressor cohort, with a median OS of 16 months (HR 0.65 [95% CI 0.51 to 0.83]) compared to 11.8 months in those receiving only chemotherapy.

Dual HER2 blockade has been investigated in metastatic gastric cancer. The phase 3 randomized JACOB trial assigned 780 patients to receive either trastuzumab with a cisplatin/fluoropyrimidine regimen with or without the addition of pertuzumab; the primary end point was OS.²⁵ A non-statistically significant trend towards improvement in OS was found in the pertuzumab arm (17.5 months) as compared with the standard of care arm (14.2 months, HR 0.84, P = 0.0565). The pertuzumab/trastuzumab/chemotherapy cohort experienced a higher incidence of diarrhea (61.6% versus 35.1% in control arm). Cardiac toxicity was comparable in the 2 cohorts.

The Table provides a summary of relevant clinical trials in metastatic gastric cancer.

SECOND-LINE THERAPY

CASE CONTINUED

The patient receives capecitabine, oxaliplatin, and trastuzumab therapy for 6 cycles, followed by trastuzumab for another 3 cycles. While on therapy, he develops a painful right clavicular lesion. He undergoes magnetic resonance imaging of the right clavicle, which shows a lesion in the distal two-thirds of the right clavicle measuring 9.7 × 3.7 × 3.8 cm. The patient is started on palliative radiation to the clavicle. However, repeat CT imaging shows progressive liver metastases.

 

• What is the approach to second-line therapy for metastatic gastric cancer?

Improvements in our understanding of the molecular pathways that lead to tumorigenesis have contributed to the development of several targeted agents whose efficacy in gastric cancer is being investigated. The NCCN guidelines recommend that for all patients who progress on frontline therapy, second-line therapy consists of a combination of ramucirumab and paclitaxel. Other options include single-agent docetaxel, paclitaxel, irinotecan, or ramucirumab. Combination therapy using irinotecan with either docetaxel, fluorouracil, or cisplatin may also be used.

Ramucirumab, a human IgG1 monoclonal antibody that targets the vascular endothelial growth factor receptor 2 (VEGFR2), was initially approved in 2014 as monotherapy for patients who had previously progressed on first-line chemotherapy. Its approval was based on the results of the phase 3 randomized, double-blind placebo-controlled REGARD study.²⁶ The trial randomly assigned 355 patients with advanced gastric or GEJ adenocarcinoma and disease progression after first-line platinum-containing or fluoropyrimidine-containing chemotherapy to receive best supportive care plus either ramucirumab (n = 238) or placebo (n = 117). Monotherapy with ramucirumab significantly improved median OS compared with placebo (5.2 months versus. 3.8 months; HR 0.776 [95% CI 0.6 to 0.99], P = 0.047). There was also an improvement in PFS of 2.1 months in the ramucirumab cohort, as compared to 1.3 months in the placebo cohort (P < 0.0001). Patients in the ramucirumab arm experienced a higher incidence of hypertension (16% versus 8%), but all other adverse events occurred at comparable rates. Five deaths in the ramucirumab group were thought to be secondary to the study drug, as compared to 2 deaths in the placebo group.

In the subsequent phase 3 RAINBOW trial, the addition of ramucirumab to paclitaxel was investigated, with 330 patients assigned to the combination group and 335 to the paclitaxel-only group.²⁷ The trial again showed that combination therapy afforded patients a significant survival advantage compared to paclitaxel alone, with a median OS of 9.6 months versus 7.4 months for the monotherapy group (HR 0.807 [95% CI 0.678 to 0.962], P = 0.017). A PFS benefit of 4.4 months was observed in the combination therapy groups, as compared with 2.9 months in the monotherapy group (HR 0.635, P < 0.0001). The ramucirumab/paclitaxel group also had a higher overall response rate of 28% versus 16%. The combination cohort had an increased incidence of grade 3 or higher adverse hypertensive events (14% versus 2%) and neutropenia (41% versus 19%), while the incidence of grade 3 febrile neutropenic events was similar between the groups (3% versus 2%).

The addition of bevacizumab, another monoclonal antibody against VEGF, to standard chemotherapy regimens has been explored, but studies have failed to show a survival benefit with this agent in the first-line treatment of advanced gastric cancer. The phase 3 Avastin in Gastric Cancer (AVAGAST) trial was a multinational, randomized study where patients received either bevacizumab (n = 387) or placebo (n = 387) in addition to cisplatin and capecitabine.²⁸ The substitution of fluorouracil for capecitabine was permitted for patients who were unable to tolerate oral medications. Cisplatin was administered for a maximum of 6 cycles, while capecitabine and bevacizumab were administered until disease progression. The study failed to show an improvement in OS, with a median OS of 12.1 months noted in the bevacizumab cohort, as compared to 10.1 months in the placebo arm (HR 0.87 [95% CI 0.73 to 1.03], P = 0.1002). However, there was a modest improvement in median PFS (6.7 months versus 5.3 months; HR 0.80 [95% CI 0.68 to 0.93], P = 0.0037) and overall response rate (46% versus 37.4%, P = 0.0315). The most commonly reported grade 3 to 5 adverse events included neutropenia (35%), anemia (10%), and loss of appetite (8%). Interestingly, in a follow-up report, higher serum levels of VEGF-A were thought to correlate with an enhanced response to bevacizumab.²⁹ However, the routine use of biomarker analysis in selecting patients for treatment with bevacizumab in metastatic gastric cancer remains to be further clarified.

 

Use of other agents with anti-HER2 activity in the second-line treatment of patients who have experienced progression while on trastuzumab remains unclear. In the recent T-ACT trial, patients with disease refractory to frontline therapy with combination trastuzumab and fluoropyrimidine/platinum agents were randomly assigned to receive either weekly paclitaxel (n = 45) or weekly paclitaxel plus trastuzumab (n = 44).³⁰ Patients in the combination cohort received an initial dose of trastuzumab 8 mg/kg followed by 6 mg/kg every 3 weeks until progression. The study did not find a difference in either PFS (3.19 months versus 3.68 months; HR 0.91 [95% CI 0.67 to 1.22], P = 0.33) or OS (9.95 months versus 10.2 months; HR 1.23 [95% CI 0.75 to 1.99], P = 0.20). The study thus failed to show a benefit to continuing trastuzumab after progression in the first-line setting.

Lapatinib in combination with paclitaxel has been compared to paclitaxel alone for the treatment of advanced HER2-positive gastric cancer in an Asian population in the phase 3 TyTAN trial.³¹ With a primary end point of OS, the study randomly assigned 129 patients to receive paclitaxel alone and 132 patients to receive paclitaxel with lapatinib. There was a nonsignificant trend towards improvement in OS in the combination group (11 months) as compared to the paclitaxel-only group(8.9 months, P = 0.1044), with no significant difference in median PFS (5.4 months versus 4.4 months). However, it is important to note that only 15 patients in this trial had previously been exposed to trastuzumab. Another trial, the phase 3 GATSBY study, examined the efficacy of trastuzumab emtansine in the second-line setting compared to taxanes alone and failed to show any improvement in PFS or OS.³² Given these results, no alternative anti-HER2 therapy has been proven to be efficacious for patients who are trastuzumab refractory. Therefore, including anti-HER2 therapy in the second-line treatment of HER2-positive gastric cancer is not recommended.

IMMUNOTHERAPY AND OTHER TARGETED THERAPIES

Several other targeted therapies have been studied in advanced gastric cancer, without any demonstrable survival benefit. The PI3K/AKT/mTOR pathway is known to be involved in regulation of cell growth and angiogenesis, and the mTOR inhibitor everolimus is widely used to treat other malignancies, including breast cancer. The use of everolimus in the second-line setting was studied in the phase 3 GRANITE-1 trial, where it was compared to best supportive care and failed to provide any survival benefit.³³ Cetuximab, a recombinant human and mouse chimeric monoclonal antibody, and panitumumab, a recombinant human antibody against the epidermal growth factor receptor (EGFR), have also been examined in gastric and GEJ cancer patients. However, the large phase 3 EXPAND and REAL-3 trials did not show a survival benefit when these agents were added to standard chemotherapy.^34,35

Overexpression of MET, a proto-oncogene and tyrosine kinase receptor, has also been implicated in gastric cancer progression. The ligand for MET is the hepatocyte growth factor (HGF), and aberrant signaling of this pathway has been shown to correlate with an aggressive gastric cancer phenotype and poorer OS by promoting tumor growth and angiogenesis. However, no MET inhibitors thus far have been found to be clinically effective. RILOMET-1 and RILOMET-2 were phase 3 trials examining the efficacy of rilotumumab, a humanized anti-HGF antibody, in combination with chemotherapy (ECX and cisplatin with capecitabine, respectively) for the frontline treatment of MET-positive GEJ and gastric cancers. Both studies were discontinued due to a higher treatment-related mortality in patients receiving rilotumumab, with a higher incidence of adverse events due to disease progression being noted.³⁶ Similarly, onartuzumab, a monovalent monoclonal antibody against the MET receptor, was investigated in the phase 3 METGastric trial in combination with modified FOLFOX6 as first-line therapy for HER2-negative, MET-positive metastatic GEJ and gastric cancers. The study did not demonstrate any significant improvements in OS or PFS.³⁷

There has been significant interest in incorporating immunotherapy in the treatment of early and metastatic gastric cancer. Pembrolizumab is the first programmed death receptor (PD-1) inhibitor to be approved for treatment of patients with PD-L1−positive advanced gastric cancer who had previously received 2 or more lines of chemotherapy. Although earlier studies of pembrolizumab in lung cancer utilized the tumor proportion score (TPS) to determine PD-L1 positivity, this was not found to be applicable to gastric cancer. Instead, the combined positive score (CPS) is used in gastric cancer. The CPS evaluates the number of tumor cells and immune cells (macrophages and lymphocytes) that stain positive for PD-L1 relative to all viable tumor cells. Comparatively, the TPS only examines the percentage of viable tumor cells that show complete or partial positive staining for PD-L1. A CPS score of 1 or greater identifies patients who would be suitable candidates for pembrolizumab.

 

The approval of pembrolizumab was based on the positive findings from the recent KEYNOTE-059 trial.³⁸ The study included 259 patients who had previously received either fluoropyrimidine, cisplatin, or anti-HER2 therapy, with 148 patients (55%) of these patients having PD-L1−positive tumors. The PD-L1 status was determined using a pharmDx Kit, which is now approved by the US Food and Drug Administration to select patients who could benefit from pembrolizumab treatment. CPS was calculated as the number of PD-L1−staining cells divided by the total number of evaluated cells. The study included patients with microsatellite stable (MSI-S), undetermined, or deficient MMR status. The overall response rate to pembrolizumab across all patients was 11.6%, median PFS was 2 months, and the 12-month OS rate was 23.4%. In the subset of patients with MSI-H tumors, the overall response rate was 57.1%, with a complete response rate of 14.3%; in those with MSI-S tumors, the overall response rate was 9% and the complete response rate was 2.4%. Among patients with PD-L1–positive tumors, the overall response rate was 15.5% (95% CI 10.1% to 22.4). Common adverse events included fatigue, hypothyroidism, nausea, diarrhea, and arthralgia.³⁸

CASE CONCLUSION

This patient with metastatic gastric cancer receives second-line chemotherapy with ramucirumab and paclitaxel. Follow-up imaging shows persistent liver metastases and new lung metastasis. Because the tumor is PD-L1–positive, the patient receives 4 cycles of pembrolizumab, with no significant change noted in disease burden. He notes a significant decline in functional status with increased weight loss, nausea, emesis, and fatigue. The patient opts to forego any further therapy and instead chooses to pursue supportive care only.

SUMMARY

Gastric cancer is the third most common cause of cancer death worldwide. Common risk factors for developing gastric cancer include H. pylori infection, smoking, alcohol abuse, radiation exposure, high-fat diet, and obesity. Patients presenting with alarm symptoms of nausea, emesis, early satiety, abdominal pain, or weight loss should be fully evaluated with upper GI endoscopy. If there is suspicion for metastatic disease, CT evaluation of the chest, abdomen, and pelvis with oral and intravenous contrast should be obtained. Treatment of patients with metastatic gastric cancer is guided by their performance status at presentation. For patients with good performance status, a combination of platinum and fluoropyrimidine therapy, such as FOLFOX, can be considered. Doublet chemotherapy regimens are preferred over triplet chemotherapy regimens given their better tolerability. For patients with HER2-positive disease, the addition of trastuzumab to the platinum and fluoropyrimidine backbone is the standard of care in the first line.

Several targeted agents have been studied in patients progressing on initial therapy, with ramucirumab and paclitaxel being considered the regimen of choice in the second line. No anti-HER2 therapy has been approved for patients who are refractory to trastuzumab. Pembrolizumab is approved for use in patients who are PD-L1–positive and have previously progressed on at least 2 lines of chemotherapy. Pembrolizumab is also approved for the treatment of patients with unresectable or metastatic, MSI-H or MMR-deficient gastric cancers that have progressed after prior treatment and who have no satisfactory alternative treatment options.

INTRODUCTION

According to the Surveillance, Epidemiology and End Results database, in 2017 there were 28,000 new cases of gastric cancer, accounting for 1.8% of all malignancies in the United States, and an estimated 10,960 gastric cancer–related deaths.¹ Worldwide, gastric cancer is the fifth most common malignancy and the third most common cause of death from any cancer.² The incidence of gastric cancer varies significantly by geographic region, with countries in Eastern Asia (China, Japan), Eastern Europe, and Central and South America accounting for 50% of all new cases.³ Although the incidence of gastric cancer has declined in recent years, this decrease has not been observed consistently across all nations.² In particular, the incidence of gastric cancers arising from the cardia has been increasing, which is perhaps due to a higher prevalence of obesity in Western societies.⁴

In this article, we review key aspects of management of metastatic gastric cancer, including selection of first- and second-line therapy, and discuss targeted agents and upcoming clinical trials.

EPIDEMIOLOGY AND RISK FACTORS

Chronic infection with Helicobacter pylori, a gram-negative bacterium, is a strong etiological factor for the development of gastric cancer, contributing to up to 70% of cases.² The pathogen can colonize the gastric mucosa, leading to chronic inflammation. Although most patients remain asymptomatic, 1% to 3% develop gastric cancer and another 0.1% develop mucosa-associated lymphoid tissue lymphoma.⁵ H. pylori infection is more commonly associated with cancer of the gastric body than with cancer of the gastroesophageal junction (GEJ). The increased burden of gastric cancer in countries in Eastern Asia, Latin America, and Eastern Europe has been correlated to the prevalence of chronic H. pylori infection in these areas.

Carcinogenesis secondary to H. pylori infection may occur via several mechanisms. First, H. pylori can release virulence factors, such as cytotoxin-associated gene A, vacuolating cytotoxin, and outer membrane proteins, into the cytosol of host cells, leading to changes in patterns of cell proliferation and apoptosis.⁶ These virulence factors can modulate the host immune system, attenuating it to promote dysplasia. In addition, continued recognition of these factors by the immune system leads to a persistent inflammatory response, with the release of cytokines such as interleukin (IL) -1β, IL-6, and IL-8. This leads to chronic mucosal damage, further promoting dysplasia with eventual transformation into adenocarcinoma.⁷ In Japan and Korea, where screening for H. pylori infection is routinely performed, there have been improvements in overall survival (OS) rates for gastric cancer, with 5-year OS rates of 70%.⁸ The International Agency for Research on Cancer recommends further research into population-based screening and treatment programs for patients with chronic H. pylori infection. However, despite this recommendation, optimal screening strategies are not clearly defined.⁹

Other risk factors for the development of gastric cancer include chronic gastroesophageal reflux disease; smoking; alcohol use; exposure to radiation; diets high in fats, salt, and smoked items and low in fruits and vegetables; obesity; and exposure to chemotherapeutic agents such as procarbazine.¹⁰ Another pathogen suspected, but not proven, to be associated with increased risk for gastric cancer is the Epstein-Barr virus, a human herpesvirus found in 80% of all gastric carcinomas with lymphoid features.¹¹ In addition, whether the use of medications such as statins and nonsteroidal anti-inflammatory drugs confers a decreased risk of gastric cancers remains unclear.¹⁰

EVALUATION

CASE PRESENTATION

A 55-year-old Caucasian man with a history of type 2 diabetes mellitus presents to the gastrointestinal (GI) clinic with a 6-month history of dysphagia. The dysphagia is worsened with ingestion of solids, particularly towards the end of the day. He states that the food often gets “stuck in the middle of the chest.” The patient denies any nausea or emesis but notes that he has a poor appetite. He reports having worsening mid-epigastric abdominal pain that is non-radiating, dull in character, and 6/10 in intensity. He also reports a 10-lb weight loss over the past 2 months. He has no previous history of reflux, chest pain, dyspnea, or cough. Review of systems is otherwise benign. Physical exam is within normal limits.

• Which tests should be conducted when gastric cancer is suspected?

 

Persistent epigastric abdominal pain and weight loss are the most common early symptoms of gastric cancer. Nausea, early satiety, dysphagia, and occult GI bleeding can be other presenting signs. Patients presenting with alarm symptoms of nausea, emesis, early satiety, abdominal pain, or weight loss should be fully evaluated with upper GI endoscopy. Early diagnosis of gastric cancer is essential in obtaining a curative resection. However, at least 40% of patients present with de novo metastatic disease at the time of initial diagnosis.¹² Gastric cancer spreads by direct extension through the gastric wall, with the liver, peritoneum, and regional lymph nodes being the most common sites of metastatic deposits.¹³ Classically, Virchow’s node, the left supraclavicular lymph node, is involved with metastatic gastric cancer. Involvement of the left axillary lymph node (Irish node) or a periumbilical nodule (Sister Mary Joseph node) may also be observed. Other, less commonly noted sites of metastatic disease include the ovaries, central nervous system, bone, lung, and soft tissues.¹³

Upper GI endoscopy is the best method for determining tumor location and extent and obtaining a specimen for a definitive tissue diagnosis.¹⁴ It is essential to accurately identify the location of the tumor in the stomach and relative to the GEJ. The American Joint Committee on Cancer classification defines tumors involving the GEJ with an epicenter no more than 2 cm into the proximal stomach as esophageal cancers.¹⁵ Tumors of the GEJ with their epicenter more than 2 cm into the proximal stomach are defined as gastric cancers. If metastatic disease is suspected, computed tomography (CT) scan of the chest, abdomen, and pelvis with oral and intravenous contrast can be obtained to determine the extent of disease spread. In the absence of any metastatic disease, endoscopic ultrasound (EUS) should be conducted to determine the depth of tumor invasion (T staging) and lymph node status. In the era of targeted therapy, patients with metastatic disease should undergo testing for human epidermal growth factor-2 (HER-2) expression, microsatellite instability (MSI), and programmed death ligand 1 (PD-L1) expression. Patients should be staged according to the TNM staging system.

FIRST-LINE TREATMENT OPTIONS

CASE CONTINUED

The patient undergoes esophagoduodenoscopy (EGD) and is found to have a gastric cardia mass extending into the distal esophagus. EUS also demonstrates multiple abdominal and mediastinal lymph nodes. No gastric outlet obstruction is found. Biopsy shows poorly differentiated invasive adenocarcinoma. Warthin–Starry stain is negative for H. pylori organism. The tumor cells are positive for cytokeratin (CK7), CK19, and mucin-1 gene (MUC1); focally positive for CK20; and negative for MUC2. HER2 testing results are reported as immunohistochemistry (IHC) 3+, consistent with strongly positive HER2 protein expression. Further IHC testing for mismatch repair (MMR) proteins shows intact nuclear expression of MLH1, MSH2, MSH6, and PMS2 protein, consistent with a low probability of MSI-high tumor. The tumor is found to be PD-L1 positive. Imaging reveals abnormal mass-like nodular thickening of the gastric wall, with an infiltrative opacity within the pancreatico-duodenal groove, suspicious for tumor infiltration. Multiple metastatic deposits are noted in the liver, peritoneum, and bilateral lungs. There is extensive gastrohepatic ligament and periportal lymphadenopathy and mild enlargement of the pulmonary hilar lymph nodes. These findings are consistent with stage 4 (T4bN3aM1) gastric cancer. Given these findings, staging laparoscopy is deferred.

• What are the first-line treatment options for patients with metastatic gastric cancer?

Patients with metastatic gastric cancer have a poor prognosis, and management is stratified based on performance status (Figure). In patients with good performance status, systemic chemotherapy is the mainstay of treatment. The goal of therapy is not curative, but rather treatment focuses on palliation of symptoms arising from tumor spread. Given this treatment goal, there has been considerable interest in clarifying the utility of chemotherapy as opposed to best supportive care. In a recent Cochrane review of 64 randomized control trials involving 11,698 patients, chemotherapy was found to improve OS by 6.7 months as compared to best supportive care (hazard ratio [HR] 0.3 [95% confidence interval {CI} 0.24 to 0.55]).¹⁶ Five classes of cytotoxic chemotherapeutic agents have demonstrated activity in gastric cancer. These include fluoropyrimidine (either infusional fluorouracil or capecitabine), platinum agents (cisplatin or oxaliplatin), taxanes (docetaxel or paclitaxel), anthracyclines (epirubicin), and irinotecan.¹³ Treatment options are further divided based on whether the patient has HER2-overexpressing or non-expressing malignancy.

 

 

HER2-NEGATIVE DISEASE

For patients with HER2-negative disease, National Comprehensive Cancer Network (NCCN) guidelines recommend using 2-drug combination regimens rather than 3 drugs, given concern for increased toxicity with 3-drug regimens.¹⁷ For patients with a performance status of 0 to 1, utilization of a 3-drug regimen is a reasonable alternative. The combination of a fluoropyrimidine with a platinum agent is considered the standard of care, with regimens such as fluorouracil, leucovorin, and oxaliplatin (FOLFOX) being commonly used.

Epirubicin-containing regimens have also been extensively studied in advanced gastric cancer. In a study of 274 previously untreated patients with GEJ cancers, the combination of epirubicin, cisplatin, and fluorouracil (ECF) was compared to fluorouracil, doxorubicin, and methotrexate (FAMTX). There was an OS benefit favoring ECF (8.9 months versus 5.7 months) at 1 year (95% CI 27% to 45%, P = 0.0009). The ECF regimen was associated with an increased risk of nausea, emesis, and alopecia, while more hematologic toxicity and infections were noted with the FAMTX regimen.¹⁸ In addition, in a phase 3 trial, Van Cutsem and colleagues examined the role of docetaxel in combination with cisplatin and fluorouracil (DCF) compared to cisplatin and fluorouracil alone. Addition of docetaxel led to improved OS and time to progression (9.2 months versus 8.6 months for cisplatin and fluorouracil alone, P = 0.02) but with an increased risk of grade 3 and 4 toxicities (69% versus 59%). These adverse events included neutropenia (82% versus 57% of cisplatin and fluorouracil patients), diarrhea (19% versus 8%), stomatitis (21% versus 27%), and fatigue (19% versus 14%).¹⁹

The landmark phase 3 REAL-2 study compared 4 chemotherapy regimens in patients with untreated advanced esophagogastric cancer. This study was conducted to determine if the efficacy of cisplatin and oxaliplatin, a third-generation platinum agent, is equivalent to that of fluorouracil and capecitabine, an oral fluoropyrimidine. In this trial, a 2 × 2 design was used to compare 4 regimens: ECF versus epirubicin, cisplatin, and capecitabine (ECX) versus epirubicin, oxaliplatin, and fluorouracil (EOF) versus epirubicin, oxaliplatin, and capecitabine (EOX). The study found EOX to be noninferior to ECF, with a trend towards improved OS compared to other combination regimens (11.2 months versus 9.9 months, HR 0.80 [95% CI 0.66 to 0.97], P = 0.02).²⁰ Thus, the study demonstrated that an oxaliplatin and capecitabine-based regimen could replace cisplatin and fluorouracil. Given that fluorouracil administration requires long continuous infusions, the oral-based capecitabine regimen is an attractive option for patients.

Several trials have demonstrated the equivalency of oxaliplatin with cisplatin in combination regimens for the treatment of advanced gastric cancer. Oxaliplatin has the benefit of an improved toxicity profile as compared to cisplatin, with the major dose-limiting toxicity being peripheral neuropathy

Given previous evidence that DCF (docetaxel, cisplatin, fluorouracil) is superior to cisplatin and fluorouracil alone, there was interest in determining if the addition of docetaxel to a backbone of fluorouracil, oxaliplatin, and leucovorin (FLO) could elicit a higher response rate. This concept was investigated in a phase 2 trial that assigned 54 patients with metastatic gastric or GEJ adenocarcinoma to receive biweekly infusions of oxaliplatin, leucovorin, fluorouracil, and docetaxel.²¹ Median time to response was 1.54 months, and the overall response rate was 57.7%. Median progression-free survival (PFS) was 5.2 months, and OS was 11.1 months. The most common grade 3 or 4 toxicities included neutropenia (48%), leukopenia (27.8%), diarrhea (14.8%), and fatigue (11.1%).

 

Irinotecan-based regimens have also been extensively studied in the first-line treatment of metastatic gastric cancer, particularly as an alternative to platinum-based therapy, but superiority has not been established. The combination of fluorouracil, leucovorin, and irinotecan (FOLFIRI) was compared to ECX in a phase 3 trial.²² The study enrolled 416 patients with locally advanced or metastatic gastric or GEJ cancer. At a median follow up of 31 months, the time to progression was longer in the FOLFIRI arm as compared to the ECX arm (5.1 months versus 4.2 months, P = 0.008), but there was no difference in OS (9.5 months versus 9.7 months, P = 0.95), median PFS (5.3 months versus 5.8 months, P = 0.96), or response rate (39.2% versus 37.8%). However, the FOLFIRI regimen had an improved toxicity profile, with a lower overall rate of grade 3 or 4 toxicity (69% versus 84%, P < 0.001). Given these findings, the FOLFIRI regimen is an acceptable alternative to platinum-based therapy in suitable patients.²²

HER2-POSITIVE DISEASE

The HER2 proto-oncogene, initially described in breast cancer, has been implicated in several malignancies, including gastric and esophageal cancer. Overexpression or amplification of HER2 can be found in up to 30% of gastric cancers.²³ For these patients, adding trastuzumab to a standard regimen of platinum and fluoropyrimidine is the standard of care. The prospective phase 3 Trastuzumab for Gastric Cancer (ToGA) trial randomly assigned 594 patients with HER2-positive gastric cancer to receive either cisplatin and fluorouracil or capecitabine and cisplatin with trastuzumab (n = 294) or without (n = 290) trastuzumab every 3 weeks for a total of 6 cycles, followed by maintenance trastuzumab until disease progression was noted.²⁴ HER2 positivity was defined as HER2 protein overexpression by IHC (cutoff of 3+) or gene amplification by fluorescence in situ hybridization (FISH); tumors with IHC 2+ patterns were followed with FISH studies to confirm positivity. The study found a higher incidence of HER2-positive tumors in patients with GEJ tumors compared to patients with distal gastric cancers (33% versus 20%).²⁴ In this trial, the addition of trastuzumab was associated with an improvement in OS: 13.5 months in the trastuzumab cohort versus 11.1 months in those receiving chemotherapy alone (HR 0.74, P = 0.0048). There was not a significant difference in toxicities between the 2 cohorts, with nausea, emesis, and neutropenia being the most common adverse events. Rates of overall grade 3 or 4 events were similar as well (68% in each cohort). Further exploratory analysis was also conducted according to HER2 status by dividing patients into a “high-expressor” group (n = 446), defined as patients with IHC 3+ tumors or IHC 2+ and FISH positivity, and a “low-expressor” group (n = 131), which included patients with IHC 0 or 1+ tumors. Analysis of patients in the 2 subgroups demonstrated an improved OS with the addition of trastuzumab for the high-expressor cohort, with a median OS of 16 months (HR 0.65 [95% CI 0.51 to 0.83]) compared to 11.8 months in those receiving only chemotherapy.

Dual HER2 blockade has been investigated in metastatic gastric cancer. The phase 3 randomized JACOB trial assigned 780 patients to receive either trastuzumab with a cisplatin/fluoropyrimidine regimen with or without the addition of pertuzumab; the primary end point was OS.²⁵ A non-statistically significant trend towards improvement in OS was found in the pertuzumab arm (17.5 months) as compared with the standard of care arm (14.2 months, HR 0.84, P = 0.0565). The pertuzumab/trastuzumab/chemotherapy cohort experienced a higher incidence of diarrhea (61.6% versus 35.1% in control arm). Cardiac toxicity was comparable in the 2 cohorts.

The Table provides a summary of relevant clinical trials in metastatic gastric cancer.

SECOND-LINE THERAPY

CASE CONTINUED

The patient receives capecitabine, oxaliplatin, and trastuzumab therapy for 6 cycles, followed by trastuzumab for another 3 cycles. While on therapy, he develops a painful right clavicular lesion. He undergoes magnetic resonance imaging of the right clavicle, which shows a lesion in the distal two-thirds of the right clavicle measuring 9.7 × 3.7 × 3.8 cm. The patient is started on palliative radiation to the clavicle. However, repeat CT imaging shows progressive liver metastases.

 

• What is the approach to second-line therapy for metastatic gastric cancer?

Improvements in our understanding of the molecular pathways that lead to tumorigenesis have contributed to the development of several targeted agents whose efficacy in gastric cancer is being investigated. The NCCN guidelines recommend that for all patients who progress on frontline therapy, second-line therapy consists of a combination of ramucirumab and paclitaxel. Other options include single-agent docetaxel, paclitaxel, irinotecan, or ramucirumab. Combination therapy using irinotecan with either docetaxel, fluorouracil, or cisplatin may also be used.

Ramucirumab, a human IgG1 monoclonal antibody that targets the vascular endothelial growth factor receptor 2 (VEGFR2), was initially approved in 2014 as monotherapy for patients who had previously progressed on first-line chemotherapy. Its approval was based on the results of the phase 3 randomized, double-blind placebo-controlled REGARD study.²⁶ The trial randomly assigned 355 patients with advanced gastric or GEJ adenocarcinoma and disease progression after first-line platinum-containing or fluoropyrimidine-containing chemotherapy to receive best supportive care plus either ramucirumab (n = 238) or placebo (n = 117). Monotherapy with ramucirumab significantly improved median OS compared with placebo (5.2 months versus. 3.8 months; HR 0.776 [95% CI 0.6 to 0.99], P = 0.047). There was also an improvement in PFS of 2.1 months in the ramucirumab cohort, as compared to 1.3 months in the placebo cohort (P < 0.0001). Patients in the ramucirumab arm experienced a higher incidence of hypertension (16% versus 8%), but all other adverse events occurred at comparable rates. Five deaths in the ramucirumab group were thought to be secondary to the study drug, as compared to 2 deaths in the placebo group.

In the subsequent phase 3 RAINBOW trial, the addition of ramucirumab to paclitaxel was investigated, with 330 patients assigned to the combination group and 335 to the paclitaxel-only group.²⁷ The trial again showed that combination therapy afforded patients a significant survival advantage compared to paclitaxel alone, with a median OS of 9.6 months versus 7.4 months for the monotherapy group (HR 0.807 [95% CI 0.678 to 0.962], P = 0.017). A PFS benefit of 4.4 months was observed in the combination therapy groups, as compared with 2.9 months in the monotherapy group (HR 0.635, P < 0.0001). The ramucirumab/paclitaxel group also had a higher overall response rate of 28% versus 16%. The combination cohort had an increased incidence of grade 3 or higher adverse hypertensive events (14% versus 2%) and neutropenia (41% versus 19%), while the incidence of grade 3 febrile neutropenic events was similar between the groups (3% versus 2%).

The addition of bevacizumab, another monoclonal antibody against VEGF, to standard chemotherapy regimens has been explored, but studies have failed to show a survival benefit with this agent in the first-line treatment of advanced gastric cancer. The phase 3 Avastin in Gastric Cancer (AVAGAST) trial was a multinational, randomized study where patients received either bevacizumab (n = 387) or placebo (n = 387) in addition to cisplatin and capecitabine.²⁸ The substitution of fluorouracil for capecitabine was permitted for patients who were unable to tolerate oral medications. Cisplatin was administered for a maximum of 6 cycles, while capecitabine and bevacizumab were administered until disease progression. The study failed to show an improvement in OS, with a median OS of 12.1 months noted in the bevacizumab cohort, as compared to 10.1 months in the placebo arm (HR 0.87 [95% CI 0.73 to 1.03], P = 0.1002). However, there was a modest improvement in median PFS (6.7 months versus 5.3 months; HR 0.80 [95% CI 0.68 to 0.93], P = 0.0037) and overall response rate (46% versus 37.4%, P = 0.0315). The most commonly reported grade 3 to 5 adverse events included neutropenia (35%), anemia (10%), and loss of appetite (8%). Interestingly, in a follow-up report, higher serum levels of VEGF-A were thought to correlate with an enhanced response to bevacizumab.²⁹ However, the routine use of biomarker analysis in selecting patients for treatment with bevacizumab in metastatic gastric cancer remains to be further clarified.

 

Use of other agents with anti-HER2 activity in the second-line treatment of patients who have experienced progression while on trastuzumab remains unclear. In the recent T-ACT trial, patients with disease refractory to frontline therapy with combination trastuzumab and fluoropyrimidine/platinum agents were randomly assigned to receive either weekly paclitaxel (n = 45) or weekly paclitaxel plus trastuzumab (n = 44).³⁰ Patients in the combination cohort received an initial dose of trastuzumab 8 mg/kg followed by 6 mg/kg every 3 weeks until progression. The study did not find a difference in either PFS (3.19 months versus 3.68 months; HR 0.91 [95% CI 0.67 to 1.22], P = 0.33) or OS (9.95 months versus 10.2 months; HR 1.23 [95% CI 0.75 to 1.99], P = 0.20). The study thus failed to show a benefit to continuing trastuzumab after progression in the first-line setting.

Lapatinib in combination with paclitaxel has been compared to paclitaxel alone for the treatment of advanced HER2-positive gastric cancer in an Asian population in the phase 3 TyTAN trial.³¹ With a primary end point of OS, the study randomly assigned 129 patients to receive paclitaxel alone and 132 patients to receive paclitaxel with lapatinib. There was a nonsignificant trend towards improvement in OS in the combination group (11 months) as compared to the paclitaxel-only group(8.9 months, P = 0.1044), with no significant difference in median PFS (5.4 months versus 4.4 months). However, it is important to note that only 15 patients in this trial had previously been exposed to trastuzumab. Another trial, the phase 3 GATSBY study, examined the efficacy of trastuzumab emtansine in the second-line setting compared to taxanes alone and failed to show any improvement in PFS or OS.³² Given these results, no alternative anti-HER2 therapy has been proven to be efficacious for patients who are trastuzumab refractory. Therefore, including anti-HER2 therapy in the second-line treatment of HER2-positive gastric cancer is not recommended.

IMMUNOTHERAPY AND OTHER TARGETED THERAPIES

Several other targeted therapies have been studied in advanced gastric cancer, without any demonstrable survival benefit. The PI3K/AKT/mTOR pathway is known to be involved in regulation of cell growth and angiogenesis, and the mTOR inhibitor everolimus is widely used to treat other malignancies, including breast cancer. The use of everolimus in the second-line setting was studied in the phase 3 GRANITE-1 trial, where it was compared to best supportive care and failed to provide any survival benefit.³³ Cetuximab, a recombinant human and mouse chimeric monoclonal antibody, and panitumumab, a recombinant human antibody against the epidermal growth factor receptor (EGFR), have also been examined in gastric and GEJ cancer patients. However, the large phase 3 EXPAND and REAL-3 trials did not show a survival benefit when these agents were added to standard chemotherapy.^34,35

Overexpression of MET, a proto-oncogene and tyrosine kinase receptor, has also been implicated in gastric cancer progression. The ligand for MET is the hepatocyte growth factor (HGF), and aberrant signaling of this pathway has been shown to correlate with an aggressive gastric cancer phenotype and poorer OS by promoting tumor growth and angiogenesis. However, no MET inhibitors thus far have been found to be clinically effective. RILOMET-1 and RILOMET-2 were phase 3 trials examining the efficacy of rilotumumab, a humanized anti-HGF antibody, in combination with chemotherapy (ECX and cisplatin with capecitabine, respectively) for the frontline treatment of MET-positive GEJ and gastric cancers. Both studies were discontinued due to a higher treatment-related mortality in patients receiving rilotumumab, with a higher incidence of adverse events due to disease progression being noted.³⁶ Similarly, onartuzumab, a monovalent monoclonal antibody against the MET receptor, was investigated in the phase 3 METGastric trial in combination with modified FOLFOX6 as first-line therapy for HER2-negative, MET-positive metastatic GEJ and gastric cancers. The study did not demonstrate any significant improvements in OS or PFS.³⁷

There has been significant interest in incorporating immunotherapy in the treatment of early and metastatic gastric cancer. Pembrolizumab is the first programmed death receptor (PD-1) inhibitor to be approved for treatment of patients with PD-L1−positive advanced gastric cancer who had previously received 2 or more lines of chemotherapy. Although earlier studies of pembrolizumab in lung cancer utilized the tumor proportion score (TPS) to determine PD-L1 positivity, this was not found to be applicable to gastric cancer. Instead, the combined positive score (CPS) is used in gastric cancer. The CPS evaluates the number of tumor cells and immune cells (macrophages and lymphocytes) that stain positive for PD-L1 relative to all viable tumor cells. Comparatively, the TPS only examines the percentage of viable tumor cells that show complete or partial positive staining for PD-L1. A CPS score of 1 or greater identifies patients who would be suitable candidates for pembrolizumab.

 

The approval of pembrolizumab was based on the positive findings from the recent KEYNOTE-059 trial.³⁸ The study included 259 patients who had previously received either fluoropyrimidine, cisplatin, or anti-HER2 therapy, with 148 patients (55%) of these patients having PD-L1−positive tumors. The PD-L1 status was determined using a pharmDx Kit, which is now approved by the US Food and Drug Administration to select patients who could benefit from pembrolizumab treatment. CPS was calculated as the number of PD-L1−staining cells divided by the total number of evaluated cells. The study included patients with microsatellite stable (MSI-S), undetermined, or deficient MMR status. The overall response rate to pembrolizumab across all patients was 11.6%, median PFS was 2 months, and the 12-month OS rate was 23.4%. In the subset of patients with MSI-H tumors, the overall response rate was 57.1%, with a complete response rate of 14.3%; in those with MSI-S tumors, the overall response rate was 9% and the complete response rate was 2.4%. Among patients with PD-L1–positive tumors, the overall response rate was 15.5% (95% CI 10.1% to 22.4). Common adverse events included fatigue, hypothyroidism, nausea, diarrhea, and arthralgia.³⁸

CASE CONCLUSION

This patient with metastatic gastric cancer receives second-line chemotherapy with ramucirumab and paclitaxel. Follow-up imaging shows persistent liver metastases and new lung metastasis. Because the tumor is PD-L1–positive, the patient receives 4 cycles of pembrolizumab, with no significant change noted in disease burden. He notes a significant decline in functional status with increased weight loss, nausea, emesis, and fatigue. The patient opts to forego any further therapy and instead chooses to pursue supportive care only.

SUMMARY

Gastric cancer is the third most common cause of cancer death worldwide. Common risk factors for developing gastric cancer include H. pylori infection, smoking, alcohol abuse, radiation exposure, high-fat diet, and obesity. Patients presenting with alarm symptoms of nausea, emesis, early satiety, abdominal pain, or weight loss should be fully evaluated with upper GI endoscopy. If there is suspicion for metastatic disease, CT evaluation of the chest, abdomen, and pelvis with oral and intravenous contrast should be obtained. Treatment of patients with metastatic gastric cancer is guided by their performance status at presentation. For patients with good performance status, a combination of platinum and fluoropyrimidine therapy, such as FOLFOX, can be considered. Doublet chemotherapy regimens are preferred over triplet chemotherapy regimens given their better tolerability. For patients with HER2-positive disease, the addition of trastuzumab to the platinum and fluoropyrimidine backbone is the standard of care in the first line.

Several targeted agents have been studied in patients progressing on initial therapy, with ramucirumab and paclitaxel being considered the regimen of choice in the second line. No anti-HER2 therapy has been approved for patients who are refractory to trastuzumab. Pembrolizumab is approved for use in patients who are PD-L1–positive and have previously progressed on at least 2 lines of chemotherapy. Pembrolizumab is also approved for the treatment of patients with unresectable or metastatic, MSI-H or MMR-deficient gastric cancers that have progressed after prior treatment and who have no satisfactory alternative treatment options.

INTRODUCTION

According to the Surveillance, Epidemiology and End Results database, in 2017 there were 28,000 new cases of gastric cancer, accounting for 1.8% of all malignancies in the United States, and an estimated 10,960 gastric cancer–related deaths.¹ Worldwide, gastric cancer is the fifth most common malignancy and the third most common cause of death from any cancer.² The incidence of gastric cancer varies significantly by geographic region, with countries in Eastern Asia (China, Japan), Eastern Europe, and Central and South America accounting for 50% of all new cases.³ Although the incidence of gastric cancer has declined in recent years, this decrease has not been observed consistently across all nations.² In particular, the incidence of gastric cancers arising from the cardia has been increasing, which is perhaps due to a higher prevalence of obesity in Western societies.⁴

In this article, we review key aspects of management of metastatic gastric cancer, including selection of first- and second-line therapy, and discuss targeted agents and upcoming clinical trials.

EPIDEMIOLOGY AND RISK FACTORS

Chronic infection with Helicobacter pylori, a gram-negative bacterium, is a strong etiological factor for the development of gastric cancer, contributing to up to 70% of cases.² The pathogen can colonize the gastric mucosa, leading to chronic inflammation. Although most patients remain asymptomatic, 1% to 3% develop gastric cancer and another 0.1% develop mucosa-associated lymphoid tissue lymphoma.⁵ H. pylori infection is more commonly associated with cancer of the gastric body than with cancer of the gastroesophageal junction (GEJ). The increased burden of gastric cancer in countries in Eastern Asia, Latin America, and Eastern Europe has been correlated to the prevalence of chronic H. pylori infection in these areas.

Carcinogenesis secondary to H. pylori infection may occur via several mechanisms. First, H. pylori can release virulence factors, such as cytotoxin-associated gene A, vacuolating cytotoxin, and outer membrane proteins, into the cytosol of host cells, leading to changes in patterns of cell proliferation and apoptosis.⁶ These virulence factors can modulate the host immune system, attenuating it to promote dysplasia. In addition, continued recognition of these factors by the immune system leads to a persistent inflammatory response, with the release of cytokines such as interleukin (IL) -1β, IL-6, and IL-8. This leads to chronic mucosal damage, further promoting dysplasia with eventual transformation into adenocarcinoma.⁷ In Japan and Korea, where screening for H. pylori infection is routinely performed, there have been improvements in overall survival (OS) rates for gastric cancer, with 5-year OS rates of 70%.⁸ The International Agency for Research on Cancer recommends further research into population-based screening and treatment programs for patients with chronic H. pylori infection. However, despite this recommendation, optimal screening strategies are not clearly defined.⁹

Other risk factors for the development of gastric cancer include chronic gastroesophageal reflux disease; smoking; alcohol use; exposure to radiation; diets high in fats, salt, and smoked items and low in fruits and vegetables; obesity; and exposure to chemotherapeutic agents such as procarbazine.¹⁰ Another pathogen suspected, but not proven, to be associated with increased risk for gastric cancer is the Epstein-Barr virus, a human herpesvirus found in 80% of all gastric carcinomas with lymphoid features.¹¹ In addition, whether the use of medications such as statins and nonsteroidal anti-inflammatory drugs confers a decreased risk of gastric cancers remains unclear.¹⁰

EVALUATION

CASE PRESENTATION

A 55-year-old Caucasian man with a history of type 2 diabetes mellitus presents to the gastrointestinal (GI) clinic with a 6-month history of dysphagia. The dysphagia is worsened with ingestion of solids, particularly towards the end of the day. He states that the food often gets “stuck in the middle of the chest.” The patient denies any nausea or emesis but notes that he has a poor appetite. He reports having worsening mid-epigastric abdominal pain that is non-radiating, dull in character, and 6/10 in intensity. He also reports a 10-lb weight loss over the past 2 months. He has no previous history of reflux, chest pain, dyspnea, or cough. Review of systems is otherwise benign. Physical exam is within normal limits.

• Which tests should be conducted when gastric cancer is suspected?

 

Persistent epigastric abdominal pain and weight loss are the most common early symptoms of gastric cancer. Nausea, early satiety, dysphagia, and occult GI bleeding can be other presenting signs. Patients presenting with alarm symptoms of nausea, emesis, early satiety, abdominal pain, or weight loss should be fully evaluated with upper GI endoscopy. Early diagnosis of gastric cancer is essential in obtaining a curative resection. However, at least 40% of patients present with de novo metastatic disease at the time of initial diagnosis.¹² Gastric cancer spreads by direct extension through the gastric wall, with the liver, peritoneum, and regional lymph nodes being the most common sites of metastatic deposits.¹³ Classically, Virchow’s node, the left supraclavicular lymph node, is involved with metastatic gastric cancer. Involvement of the left axillary lymph node (Irish node) or a periumbilical nodule (Sister Mary Joseph node) may also be observed. Other, less commonly noted sites of metastatic disease include the ovaries, central nervous system, bone, lung, and soft tissues.¹³

Upper GI endoscopy is the best method for determining tumor location and extent and obtaining a specimen for a definitive tissue diagnosis.¹⁴ It is essential to accurately identify the location of the tumor in the stomach and relative to the GEJ. The American Joint Committee on Cancer classification defines tumors involving the GEJ with an epicenter no more than 2 cm into the proximal stomach as esophageal cancers.¹⁵ Tumors of the GEJ with their epicenter more than 2 cm into the proximal stomach are defined as gastric cancers. If metastatic disease is suspected, computed tomography (CT) scan of the chest, abdomen, and pelvis with oral and intravenous contrast can be obtained to determine the extent of disease spread. In the absence of any metastatic disease, endoscopic ultrasound (EUS) should be conducted to determine the depth of tumor invasion (T staging) and lymph node status. In the era of targeted therapy, patients with metastatic disease should undergo testing for human epidermal growth factor-2 (HER-2) expression, microsatellite instability (MSI), and programmed death ligand 1 (PD-L1) expression. Patients should be staged according to the TNM staging system.

FIRST-LINE TREATMENT OPTIONS

CASE CONTINUED

The patient undergoes esophagoduodenoscopy (EGD) and is found to have a gastric cardia mass extending into the distal esophagus. EUS also demonstrates multiple abdominal and mediastinal lymph nodes. No gastric outlet obstruction is found. Biopsy shows poorly differentiated invasive adenocarcinoma. Warthin–Starry stain is negative for H. pylori organism. The tumor cells are positive for cytokeratin (CK7), CK19, and mucin-1 gene (MUC1); focally positive for CK20; and negative for MUC2. HER2 testing results are reported as immunohistochemistry (IHC) 3+, consistent with strongly positive HER2 protein expression. Further IHC testing for mismatch repair (MMR) proteins shows intact nuclear expression of MLH1, MSH2, MSH6, and PMS2 protein, consistent with a low probability of MSI-high tumor. The tumor is found to be PD-L1 positive. Imaging reveals abnormal mass-like nodular thickening of the gastric wall, with an infiltrative opacity within the pancreatico-duodenal groove, suspicious for tumor infiltration. Multiple metastatic deposits are noted in the liver, peritoneum, and bilateral lungs. There is extensive gastrohepatic ligament and periportal lymphadenopathy and mild enlargement of the pulmonary hilar lymph nodes. These findings are consistent with stage 4 (T4bN3aM1) gastric cancer. Given these findings, staging laparoscopy is deferred.

• What are the first-line treatment options for patients with metastatic gastric cancer?

Patients with metastatic gastric cancer have a poor prognosis, and management is stratified based on performance status (Figure). In patients with good performance status, systemic chemotherapy is the mainstay of treatment. The goal of therapy is not curative, but rather treatment focuses on palliation of symptoms arising from tumor spread. Given this treatment goal, there has been considerable interest in clarifying the utility of chemotherapy as opposed to best supportive care. In a recent Cochrane review of 64 randomized control trials involving 11,698 patients, chemotherapy was found to improve OS by 6.7 months as compared to best supportive care (hazard ratio [HR] 0.3 [95% confidence interval {CI} 0.24 to 0.55]).¹⁶ Five classes of cytotoxic chemotherapeutic agents have demonstrated activity in gastric cancer. These include fluoropyrimidine (either infusional fluorouracil or capecitabine), platinum agents (cisplatin or oxaliplatin), taxanes (docetaxel or paclitaxel), anthracyclines (epirubicin), and irinotecan.¹³ Treatment options are further divided based on whether the patient has HER2-overexpressing or non-expressing malignancy.

 

 

HER2-NEGATIVE DISEASE

For patients with HER2-negative disease, National Comprehensive Cancer Network (NCCN) guidelines recommend using 2-drug combination regimens rather than 3 drugs, given concern for increased toxicity with 3-drug regimens.¹⁷ For patients with a performance status of 0 to 1, utilization of a 3-drug regimen is a reasonable alternative. The combination of a fluoropyrimidine with a platinum agent is considered the standard of care, with regimens such as fluorouracil, leucovorin, and oxaliplatin (FOLFOX) being commonly used.

Epirubicin-containing regimens have also been extensively studied in advanced gastric cancer. In a study of 274 previously untreated patients with GEJ cancers, the combination of epirubicin, cisplatin, and fluorouracil (ECF) was compared to fluorouracil, doxorubicin, and methotrexate (FAMTX). There was an OS benefit favoring ECF (8.9 months versus 5.7 months) at 1 year (95% CI 27% to 45%, P = 0.0009). The ECF regimen was associated with an increased risk of nausea, emesis, and alopecia, while more hematologic toxicity and infections were noted with the FAMTX regimen.¹⁸ In addition, in a phase 3 trial, Van Cutsem and colleagues examined the role of docetaxel in combination with cisplatin and fluorouracil (DCF) compared to cisplatin and fluorouracil alone. Addition of docetaxel led to improved OS and time to progression (9.2 months versus 8.6 months for cisplatin and fluorouracil alone, P = 0.02) but with an increased risk of grade 3 and 4 toxicities (69% versus 59%). These adverse events included neutropenia (82% versus 57% of cisplatin and fluorouracil patients), diarrhea (19% versus 8%), stomatitis (21% versus 27%), and fatigue (19% versus 14%).¹⁹

The landmark phase 3 REAL-2 study compared 4 chemotherapy regimens in patients with untreated advanced esophagogastric cancer. This study was conducted to determine if the efficacy of cisplatin and oxaliplatin, a third-generation platinum agent, is equivalent to that of fluorouracil and capecitabine, an oral fluoropyrimidine. In this trial, a 2 × 2 design was used to compare 4 regimens: ECF versus epirubicin, cisplatin, and capecitabine (ECX) versus epirubicin, oxaliplatin, and fluorouracil (EOF) versus epirubicin, oxaliplatin, and capecitabine (EOX). The study found EOX to be noninferior to ECF, with a trend towards improved OS compared to other combination regimens (11.2 months versus 9.9 months, HR 0.80 [95% CI 0.66 to 0.97], P = 0.02).²⁰ Thus, the study demonstrated that an oxaliplatin and capecitabine-based regimen could replace cisplatin and fluorouracil. Given that fluorouracil administration requires long continuous infusions, the oral-based capecitabine regimen is an attractive option for patients.

Several trials have demonstrated the equivalency of oxaliplatin with cisplatin in combination regimens for the treatment of advanced gastric cancer. Oxaliplatin has the benefit of an improved toxicity profile as compared to cisplatin, with the major dose-limiting toxicity being peripheral neuropathy

Given previous evidence that DCF (docetaxel, cisplatin, fluorouracil) is superior to cisplatin and fluorouracil alone, there was interest in determining if the addition of docetaxel to a backbone of fluorouracil, oxaliplatin, and leucovorin (FLO) could elicit a higher response rate. This concept was investigated in a phase 2 trial that assigned 54 patients with metastatic gastric or GEJ adenocarcinoma to receive biweekly infusions of oxaliplatin, leucovorin, fluorouracil, and docetaxel.²¹ Median time to response was 1.54 months, and the overall response rate was 57.7%. Median progression-free survival (PFS) was 5.2 months, and OS was 11.1 months. The most common grade 3 or 4 toxicities included neutropenia (48%), leukopenia (27.8%), diarrhea (14.8%), and fatigue (11.1%).

 

Irinotecan-based regimens have also been extensively studied in the first-line treatment of metastatic gastric cancer, particularly as an alternative to platinum-based therapy, but superiority has not been established. The combination of fluorouracil, leucovorin, and irinotecan (FOLFIRI) was compared to ECX in a phase 3 trial.²² The study enrolled 416 patients with locally advanced or metastatic gastric or GEJ cancer. At a median follow up of 31 months, the time to progression was longer in the FOLFIRI arm as compared to the ECX arm (5.1 months versus 4.2 months, P = 0.008), but there was no difference in OS (9.5 months versus 9.7 months, P = 0.95), median PFS (5.3 months versus 5.8 months, P = 0.96), or response rate (39.2% versus 37.8%). However, the FOLFIRI regimen had an improved toxicity profile, with a lower overall rate of grade 3 or 4 toxicity (69% versus 84%, P < 0.001). Given these findings, the FOLFIRI regimen is an acceptable alternative to platinum-based therapy in suitable patients.²²

HER2-POSITIVE DISEASE

The HER2 proto-oncogene, initially described in breast cancer, has been implicated in several malignancies, including gastric and esophageal cancer. Overexpression or amplification of HER2 can be found in up to 30% of gastric cancers.²³ For these patients, adding trastuzumab to a standard regimen of platinum and fluoropyrimidine is the standard of care. The prospective phase 3 Trastuzumab for Gastric Cancer (ToGA) trial randomly assigned 594 patients with HER2-positive gastric cancer to receive either cisplatin and fluorouracil or capecitabine and cisplatin with trastuzumab (n = 294) or without (n = 290) trastuzumab every 3 weeks for a total of 6 cycles, followed by maintenance trastuzumab until disease progression was noted.²⁴ HER2 positivity was defined as HER2 protein overexpression by IHC (cutoff of 3+) or gene amplification by fluorescence in situ hybridization (FISH); tumors with IHC 2+ patterns were followed with FISH studies to confirm positivity. The study found a higher incidence of HER2-positive tumors in patients with GEJ tumors compared to patients with distal gastric cancers (33% versus 20%).²⁴ In this trial, the addition of trastuzumab was associated with an improvement in OS: 13.5 months in the trastuzumab cohort versus 11.1 months in those receiving chemotherapy alone (HR 0.74, P = 0.0048). There was not a significant difference in toxicities between the 2 cohorts, with nausea, emesis, and neutropenia being the most common adverse events. Rates of overall grade 3 or 4 events were similar as well (68% in each cohort). Further exploratory analysis was also conducted according to HER2 status by dividing patients into a “high-expressor” group (n = 446), defined as patients with IHC 3+ tumors or IHC 2+ and FISH positivity, and a “low-expressor” group (n = 131), which included patients with IHC 0 or 1+ tumors. Analysis of patients in the 2 subgroups demonstrated an improved OS with the addition of trastuzumab for the high-expressor cohort, with a median OS of 16 months (HR 0.65 [95% CI 0.51 to 0.83]) compared to 11.8 months in those receiving only chemotherapy.

Dual HER2 blockade has been investigated in metastatic gastric cancer. The phase 3 randomized JACOB trial assigned 780 patients to receive either trastuzumab with a cisplatin/fluoropyrimidine regimen with or without the addition of pertuzumab; the primary end point was OS.²⁵ A non-statistically significant trend towards improvement in OS was found in the pertuzumab arm (17.5 months) as compared with the standard of care arm (14.2 months, HR 0.84, P = 0.0565). The pertuzumab/trastuzumab/chemotherapy cohort experienced a higher incidence of diarrhea (61.6% versus 35.1% in control arm). Cardiac toxicity was comparable in the 2 cohorts.

The Table provides a summary of relevant clinical trials in metastatic gastric cancer.

SECOND-LINE THERAPY

CASE CONTINUED

The patient receives capecitabine, oxaliplatin, and trastuzumab therapy for 6 cycles, followed by trastuzumab for another 3 cycles. While on therapy, he develops a painful right clavicular lesion. He undergoes magnetic resonance imaging of the right clavicle, which shows a lesion in the distal two-thirds of the right clavicle measuring 9.7 × 3.7 × 3.8 cm. The patient is started on palliative radiation to the clavicle. However, repeat CT imaging shows progressive liver metastases.

 

• What is the approach to second-line therapy for metastatic gastric cancer?

Improvements in our understanding of the molecular pathways that lead to tumorigenesis have contributed to the development of several targeted agents whose efficacy in gastric cancer is being investigated. The NCCN guidelines recommend that for all patients who progress on frontline therapy, second-line therapy consists of a combination of ramucirumab and paclitaxel. Other options include single-agent docetaxel, paclitaxel, irinotecan, or ramucirumab. Combination therapy using irinotecan with either docetaxel, fluorouracil, or cisplatin may also be used.

Ramucirumab, a human IgG1 monoclonal antibody that targets the vascular endothelial growth factor receptor 2 (VEGFR2), was initially approved in 2014 as monotherapy for patients who had previously progressed on first-line chemotherapy. Its approval was based on the results of the phase 3 randomized, double-blind placebo-controlled REGARD study.²⁶ The trial randomly assigned 355 patients with advanced gastric or GEJ adenocarcinoma and disease progression after first-line platinum-containing or fluoropyrimidine-containing chemotherapy to receive best supportive care plus either ramucirumab (n = 238) or placebo (n = 117). Monotherapy with ramucirumab significantly improved median OS compared with placebo (5.2 months versus. 3.8 months; HR 0.776 [95% CI 0.6 to 0.99], P = 0.047). There was also an improvement in PFS of 2.1 months in the ramucirumab cohort, as compared to 1.3 months in the placebo cohort (P < 0.0001). Patients in the ramucirumab arm experienced a higher incidence of hypertension (16% versus 8%), but all other adverse events occurred at comparable rates. Five deaths in the ramucirumab group were thought to be secondary to the study drug, as compared to 2 deaths in the placebo group.

In the subsequent phase 3 RAINBOW trial, the addition of ramucirumab to paclitaxel was investigated, with 330 patients assigned to the combination group and 335 to the paclitaxel-only group.²⁷ The trial again showed that combination therapy afforded patients a significant survival advantage compared to paclitaxel alone, with a median OS of 9.6 months versus 7.4 months for the monotherapy group (HR 0.807 [95% CI 0.678 to 0.962], P = 0.017). A PFS benefit of 4.4 months was observed in the combination therapy groups, as compared with 2.9 months in the monotherapy group (HR 0.635, P < 0.0001). The ramucirumab/paclitaxel group also had a higher overall response rate of 28% versus 16%. The combination cohort had an increased incidence of grade 3 or higher adverse hypertensive events (14% versus 2%) and neutropenia (41% versus 19%), while the incidence of grade 3 febrile neutropenic events was similar between the groups (3% versus 2%).

The addition of bevacizumab, another monoclonal antibody against VEGF, to standard chemotherapy regimens has been explored, but studies have failed to show a survival benefit with this agent in the first-line treatment of advanced gastric cancer. The phase 3 Avastin in Gastric Cancer (AVAGAST) trial was a multinational, randomized study where patients received either bevacizumab (n = 387) or placebo (n = 387) in addition to cisplatin and capecitabine.²⁸ The substitution of fluorouracil for capecitabine was permitted for patients who were unable to tolerate oral medications. Cisplatin was administered for a maximum of 6 cycles, while capecitabine and bevacizumab were administered until disease progression. The study failed to show an improvement in OS, with a median OS of 12.1 months noted in the bevacizumab cohort, as compared to 10.1 months in the placebo arm (HR 0.87 [95% CI 0.73 to 1.03], P = 0.1002). However, there was a modest improvement in median PFS (6.7 months versus 5.3 months; HR 0.80 [95% CI 0.68 to 0.93], P = 0.0037) and overall response rate (46% versus 37.4%, P = 0.0315). The most commonly reported grade 3 to 5 adverse events included neutropenia (35%), anemia (10%), and loss of appetite (8%). Interestingly, in a follow-up report, higher serum levels of VEGF-A were thought to correlate with an enhanced response to bevacizumab.²⁹ However, the routine use of biomarker analysis in selecting patients for treatment with bevacizumab in metastatic gastric cancer remains to be further clarified.

 

Use of other agents with anti-HER2 activity in the second-line treatment of patients who have experienced progression while on trastuzumab remains unclear. In the recent T-ACT trial, patients with disease refractory to frontline therapy with combination trastuzumab and fluoropyrimidine/platinum agents were randomly assigned to receive either weekly paclitaxel (n = 45) or weekly paclitaxel plus trastuzumab (n = 44).³⁰ Patients in the combination cohort received an initial dose of trastuzumab 8 mg/kg followed by 6 mg/kg every 3 weeks until progression. The study did not find a difference in either PFS (3.19 months versus 3.68 months; HR 0.91 [95% CI 0.67 to 1.22], P = 0.33) or OS (9.95 months versus 10.2 months; HR 1.23 [95% CI 0.75 to 1.99], P = 0.20). The study thus failed to show a benefit to continuing trastuzumab after progression in the first-line setting.

Lapatinib in combination with paclitaxel has been compared to paclitaxel alone for the treatment of advanced HER2-positive gastric cancer in an Asian population in the phase 3 TyTAN trial.³¹ With a primary end point of OS, the study randomly assigned 129 patients to receive paclitaxel alone and 132 patients to receive paclitaxel with lapatinib. There was a nonsignificant trend towards improvement in OS in the combination group (11 months) as compared to the paclitaxel-only group(8.9 months, P = 0.1044), with no significant difference in median PFS (5.4 months versus 4.4 months). However, it is important to note that only 15 patients in this trial had previously been exposed to trastuzumab. Another trial, the phase 3 GATSBY study, examined the efficacy of trastuzumab emtansine in the second-line setting compared to taxanes alone and failed to show any improvement in PFS or OS.³² Given these results, no alternative anti-HER2 therapy has been proven to be efficacious for patients who are trastuzumab refractory. Therefore, including anti-HER2 therapy in the second-line treatment of HER2-positive gastric cancer is not recommended.

IMMUNOTHERAPY AND OTHER TARGETED THERAPIES

Several other targeted therapies have been studied in advanced gastric cancer, without any demonstrable survival benefit. The PI3K/AKT/mTOR pathway is known to be involved in regulation of cell growth and angiogenesis, and the mTOR inhibitor everolimus is widely used to treat other malignancies, including breast cancer. The use of everolimus in the second-line setting was studied in the phase 3 GRANITE-1 trial, where it was compared to best supportive care and failed to provide any survival benefit.³³ Cetuximab, a recombinant human and mouse chimeric monoclonal antibody, and panitumumab, a recombinant human antibody against the epidermal growth factor receptor (EGFR), have also been examined in gastric and GEJ cancer patients. However, the large phase 3 EXPAND and REAL-3 trials did not show a survival benefit when these agents were added to standard chemotherapy.^34,35

Overexpression of MET, a proto-oncogene and tyrosine kinase receptor, has also been implicated in gastric cancer progression. The ligand for MET is the hepatocyte growth factor (HGF), and aberrant signaling of this pathway has been shown to correlate with an aggressive gastric cancer phenotype and poorer OS by promoting tumor growth and angiogenesis. However, no MET inhibitors thus far have been found to be clinically effective. RILOMET-1 and RILOMET-2 were phase 3 trials examining the efficacy of rilotumumab, a humanized anti-HGF antibody, in combination with chemotherapy (ECX and cisplatin with capecitabine, respectively) for the frontline treatment of MET-positive GEJ and gastric cancers. Both studies were discontinued due to a higher treatment-related mortality in patients receiving rilotumumab, with a higher incidence of adverse events due to disease progression being noted.³⁶ Similarly, onartuzumab, a monovalent monoclonal antibody against the MET receptor, was investigated in the phase 3 METGastric trial in combination with modified FOLFOX6 as first-line therapy for HER2-negative, MET-positive metastatic GEJ and gastric cancers. The study did not demonstrate any significant improvements in OS or PFS.³⁷

There has been significant interest in incorporating immunotherapy in the treatment of early and metastatic gastric cancer. Pembrolizumab is the first programmed death receptor (PD-1) inhibitor to be approved for treatment of patients with PD-L1−positive advanced gastric cancer who had previously received 2 or more lines of chemotherapy. Although earlier studies of pembrolizumab in lung cancer utilized the tumor proportion score (TPS) to determine PD-L1 positivity, this was not found to be applicable to gastric cancer. Instead, the combined positive score (CPS) is used in gastric cancer. The CPS evaluates the number of tumor cells and immune cells (macrophages and lymphocytes) that stain positive for PD-L1 relative to all viable tumor cells. Comparatively, the TPS only examines the percentage of viable tumor cells that show complete or partial positive staining for PD-L1. A CPS score of 1 or greater identifies patients who would be suitable candidates for pembrolizumab.

 

The approval of pembrolizumab was based on the positive findings from the recent KEYNOTE-059 trial.³⁸ The study included 259 patients who had previously received either fluoropyrimidine, cisplatin, or anti-HER2 therapy, with 148 patients (55%) of these patients having PD-L1−positive tumors. The PD-L1 status was determined using a pharmDx Kit, which is now approved by the US Food and Drug Administration to select patients who could benefit from pembrolizumab treatment. CPS was calculated as the number of PD-L1−staining cells divided by the total number of evaluated cells. The study included patients with microsatellite stable (MSI-S), undetermined, or deficient MMR status. The overall response rate to pembrolizumab across all patients was 11.6%, median PFS was 2 months, and the 12-month OS rate was 23.4%. In the subset of patients with MSI-H tumors, the overall response rate was 57.1%, with a complete response rate of 14.3%; in those with MSI-S tumors, the overall response rate was 9% and the complete response rate was 2.4%. Among patients with PD-L1–positive tumors, the overall response rate was 15.5% (95% CI 10.1% to 22.4). Common adverse events included fatigue, hypothyroidism, nausea, diarrhea, and arthralgia.³⁸

CASE CONCLUSION

This patient with metastatic gastric cancer receives second-line chemotherapy with ramucirumab and paclitaxel. Follow-up imaging shows persistent liver metastases and new lung metastasis. Because the tumor is PD-L1–positive, the patient receives 4 cycles of pembrolizumab, with no significant change noted in disease burden. He notes a significant decline in functional status with increased weight loss, nausea, emesis, and fatigue. The patient opts to forego any further therapy and instead chooses to pursue supportive care only.

SUMMARY

Gastric cancer is the third most common cause of cancer death worldwide. Common risk factors for developing gastric cancer include H. pylori infection, smoking, alcohol abuse, radiation exposure, high-fat diet, and obesity. Patients presenting with alarm symptoms of nausea, emesis, early satiety, abdominal pain, or weight loss should be fully evaluated with upper GI endoscopy. If there is suspicion for metastatic disease, CT evaluation of the chest, abdomen, and pelvis with oral and intravenous contrast should be obtained. Treatment of patients with metastatic gastric cancer is guided by their performance status at presentation. For patients with good performance status, a combination of platinum and fluoropyrimidine therapy, such as FOLFOX, can be considered. Doublet chemotherapy regimens are preferred over triplet chemotherapy regimens given their better tolerability. For patients with HER2-positive disease, the addition of trastuzumab to the platinum and fluoropyrimidine backbone is the standard of care in the first line.

Several targeted agents have been studied in patients progressing on initial therapy, with ramucirumab and paclitaxel being considered the regimen of choice in the second line. No anti-HER2 therapy has been approved for patients who are refractory to trastuzumab. Pembrolizumab is approved for use in patients who are PD-L1–positive and have previously progressed on at least 2 lines of chemotherapy. Pembrolizumab is also approved for the treatment of patients with unresectable or metastatic, MSI-H or MMR-deficient gastric cancers that have progressed after prior treatment and who have no satisfactory alternative treatment options.

INTRODUCTION

Transfusion therapy is an essential part of hematology practice, allowing for curative therapy of diseases such as leukemia, aplastic anemia, and aggressive lymphomas. Nonetheless, transfusions are associated with significant risks, including transmission of infections and transfusion-related reactions. Controversy remains about key issues in transfusion therapy, such as triggers for red cell transfusions. This article reviews the available blood products (Table 1) and indications for transfusion along with the associated risks, and also discusses specific clinical situations, such as massive transfusion.

BLOOD PRODUCTS

WHOLE BLOOD

Whole blood is the product of 1 unit of donated blood plus anticoagulant/preservative, and by definition contains 1 unit of plasma and red cells. Whole blood can be stored for 5 weeks. Although it was the standard product in the past, whole blood is rarely used since 1 unit of donated blood can now be fractionated into 1 unit of red blood cells (RBC), 1 unit of platelets, and 1 unit of fresh frozen plasma (FFP). Thus, the use of whole blood for just a single transfusion represents a waste of resources. There are 2 exceptions. One is autologous blood donations, which are whole blood units. Second, whole blood is increasingly being used in massive transfusions for trauma patients, with the rationale being that all essential blood components are being transfused at once.¹

PACKED RED CELLS

The remaining red cell mass after most of the plasma is removed is called the “packed” red cell unit (hematocrit = 70%–80%), and so red cells are often called “packed” red cells, or PRBC. A preservative is added to improve the flow of blood and to provide “nutrients” for the red cells, and this reduces the hematocrit to approximately 60%. The volume of a red cell unit is approximately 340 mL. In the average adult, 1 unit of RBC raises the hematocrit by 3%. The indications for transfusion of red cells are to increase red cell mass, and thus oxygen delivery, in patients who are compromised by their anemia.

Several randomized trials have helped define the indications for red cell transfusions and justify lower hematocrit thresholds for initiating transfusion. The TRICC (Transfusion Requirements in Critical Care Investigators for the Canadian Critical Care Trials Group) trial showed that in critical care patients (30-day mortality, 18.7%–23.3%), a conservative transfusion strategy of waiting until the hematocrit was below 21% had the same outcomes as transfusing at a threshold of 24%.² The TRACS (Transfusion Requirements After Cardiac Surgery) trial showed that a hematocrit target of 24% had the same benefit as a target of 30% in patients who had undergone cardiac bypass surgery.³ For patients with acute myocardial infarction, the outcomes were worse with aggressive transfusion at a hematocrit of 30% compared to 24%.⁴ In patients with upper gastrointestinal bleeding, a hemoglobin transfusion trigger of 7 g/dL was associated with a lower mortality than a trigger of 9 g/dL (5% versus 9%).⁵ Finally, the FOCUS (Transfusion Trigger Trial for Functional Outcomes in Cardiovascular Patients Undergoing Surgical Hip Fracture Repair) trial showed that in older patients (average age 80 years) who had undergone hip fracture surgery, transfusions based on symptoms and not a fixed trigger of 30% had the same outcomes but considerable savings in blood products.⁶ Based on these trials, decisions regarding when to transfuse patients should be based on symptoms and not “numbers.” Young patients, especially those with reversible anemias, can tolerate low blood counts and should not be transfused based on an arbitrary number.

PLATELETS

Several types of platelet products exist. One unit of platelet concentrate is derived from 1 unit of donor blood. Plateletpheresis from volunteer donors is also used to harvest platelets, with the resulting product referred to as plateletpheresis platelets. One unit of single-donor (pheresis) platelets is equivalent to 6 platelet concentrates. Finally, HLA-matched platelets are single-donor pheresis units that are obtained from an HLA-matched donor. This product should be ordered only if there is evidence of HLA antibodies (see Platelet Alloimmunization section).

The dose of platelets for the average patient is 6 units of platelet concentrate or 1 pheresis unit. In theory, 1 unit of platelet concentrate can raise the count by 5 to 7 × 10³/µL, but often this response is blunted by concurrent illness or bleeding. In patients who appear to have a poor response, the platelet count can be checked 15 minutes after platelet infusion. No rise or a minimal rise (< 2 × 10³/µL) in the platelet count is suggestive of platelet refractoriness, while a good 15-minute response but poor 24-hour count is more suggestive of consumption—fever, sepsis, drug, or splenomegaly—and not refractoriness.

The indication for platelet transfusion depends on the clinical situation. For patients with immune thrombocytopenia, platelets should not be transfused unless there is life-threatening bleeding. For stable patients with marrow aplasia from chemotherapy, a cut-off of a morning platelet count of less than 10 × 10³/µL has been shown to be as safe as higher levels for prophylactic transfusions.⁷ For patients with active bleeding, the platelet count should be kept above 50 × 10³/µL. Patients with acquired or inherited platelet dysfunction may benefit from transfusion no matter the platelet count.

 

 

Platelet Alloimmunization

Patients exposed to transfused white cells with different HLA antigens can develop antibodies to these antigens.⁸ Anti-HLA antibodies are common in patients who previously have received transfused blood that is not leukodepleted and in patients who have been pregnant. Since platelets carry class I HLA antigens, they will be rapidly destroyed by anti-HLA antibodies when transfused into these patients. In patients transfused for aplastic anemia or myelodysplasia, as many as 90% will become HLA-immunized. The incidence is lower in patients receiving chemotherapy but still can be as high as 60% to 90%.^9,10 Patients who have developed anti-HLA antibodies can respond to transfused platelets matched for HLA antigens. Unfortunately, some patients will either be a rare HLA type or be so heavily immunized that they will not respond to any platelet transfusion.

The significance of alloimmunization centers on 2 concepts: recognition and avoidance. Patients with HLA antibodies will fail to have an increment of their platelet counts with transfusions. Accordingly, patients who do not experience an increase in their count 15 minutes after the transfusion may have HLA antibodies. One can test for the presence of anti-HLA antibodies, although some patients instead have specific antiplatelet antibodies that will not respond to HLA-matched platelets. In patients who have been pregnant or previously transfused and are scheduled to undergo transplant or aggressive chemotherapy, it is wise to test for anti-HLA antibodies in order to plan their transfusion needs. The evidence suggests that transfused white cells are responsible for initiating the anti-HLA response. Trials have shown that giving leukodepleted blood products may reduce the incidence of alloimmunization, so patients who are not HLA-alloimmunized should receive only leukodepleted products.¹¹A difficult problem is bleeding in patients who are refractory to platelet transfusion.¹² Patients who test positive for the presence of anti-HLA antibodies can receive transfusions of HLA-matched platelets.¹³ Unfortunately, matched platelet transfusions are not effective in 20% to 70% of these patients. Also, since some loci are difficult to match, effective products may be unavailable. Finally, as many as 25% of patients have antiplatelet antibodies in which HLA-matched products will be ineffective. Platelet cross-matching can be performed to find compatible units for these patients, but this may not always be successful. In the patient who is totally refractory to platelet transfusion, consider drugs as an etiology of antiplatelet antibodies (especially vancomycin).¹⁴ Use of antifibrinolytic agents such as epsilon-aminocaproic acid or tranexamic acid may decrease the incidence of minor bleeding, but these are ineffective for major bleeding. “Platelet drips”—infusing either a platelet concentrate per hour or 1 plateletpheresis unit every 6 hours—may be given as a continuous infusion, but there is no evidence that this is helpful.¹⁵

FRESH FROZEN PLASMA

FFP is made from 1 unit of donated whole blood, with an average volume of 225 mL per unit. One unit of FFP can increase coagulation factor levels by 5% and fibrinogen by 10 mg/dL in the average stable patient. FFP can take 20 to 30 minutes to thaw before use, so in situations where FFP is needed quickly, the blood bank must be informed to “keep ahead” some units. Units of FFP that have been thawed but not used can be stored refrigerated for 5 days to prevent wasting blood products.

The indications for FFP are limited to several situations. These include a documented coagulation defect that can be corrected by a reasonable amount of FFP, such as factor V deficiency and factor XI deficiency, disseminated intravascular coagulation (DIC), reversal of warfarin, and massive transfusions. FFP is also used for the therapy of thrombotic thrombocytopenic purpura.

There is little justification for FFP transfusion in many of the clinical settings in which it is commonly used. For example, FFP is given for minor elevations of the INR in patients with liver disease, despite literature showing not only that the INR rise is not reflective of coagulation defects, but also that patients with liver disease may even be thrombophilic.^16,17 Reviews of FFP use found limited evidence-based indications for its use.^18,19 Also, several studies have shown that transfusion of FFP is not effective at reversing minor elevations of the INR (1.3–1.8).²⁰ In a meta-analysis, FFP was associated with increased risk for lung injury and a trend toward increased mortality.¹⁸

 

 

CRYOPRECIPITATE

Cryoprecipitate is produced from 1 unit of FFP that is thawed at 4°C. The precipitate is resuspended with 10 mL of saline or FFP and refrozen for storage. One unit contains at least 150 mg of fibrinogen and 80 units of factor VIII, along with von Willebrand factor. Thawing time for cryoprecipitate is approximately 20 minutes.

Cryoprecipitate is used to raise the fibrinogen level in patients with DIC or massive transfusion with hemodilution. It is third-line therapy in the treatment of type 1 von Willebrand disease and is second-line therapy in the treatment of patients with other types of von Willebrand disease. Currently, von Willebrand factor concentrates are the preferred replacement product for von Willebrand disease. Cryoprecipitate can be used as a source of factor VIII for hemophiliacs, but the preferred product for these patients is the super pure factor VIII concentrates or recombinant products. Cryoprecipitate can also be used to shorten the bleeding time of uremic patients, but its effectiveness for this is controversial.

GRANULOCYTES

Granulocytes are harvested by leukopheresis of normal donors, with a target yield of 1010 granulocytes from each donor. To reach this target, the donors are often “stimulated” with neutrophil growth factors. The harvesting procedure can take 3 hours and is associated with some risks to the donor (eg, citrate toxicity). The current indications for granulocytes are very limited since the advent of neutrophil growth factors and improved antimicrobials.²¹ They can be useful in the neutropenic patient with a documented bacterial infection in whom the white blood cell count is not expected to recover in the near future. Given the difficulty of keeping the count up, these transfusions have been mainly used in treating small children.

SPECIAL BLOOD PRODUCTS

IRRADIATED BLOOD PRODUCTS

Irradiation of blood is performed for only one reason: to prevent transfusion-related graft-versus-host disease (TGVHD) (Table 2).²² The irradiation can be performed at the blood center or in the transfusion service of larger hospitals. The units are not radioactive and can be transfused safely to other patients. There is increased leakage of potassium in irradiated units of blood, so the units need to be transfused within 14 days; in patients potentially sensitive to potassium (eg, neonates), the units must be transfused within 24 hours. Patients undergoing stem cell transplant, those receiving either interuterine transfusions or products from relatives, any patient with Hodgkin disease or receiving purine analogs or alemtuzumab, and patients with severe congenital immune deficiencies should receive irradiated blood. Most would also advocate that patients with hematologic malignances receiving chemotherapy receive irradiated products, but this is more controversial.

LEUKPDEPLETED BLOOD

Contamination of blood products by white blood cells is increasingly being recognized as a possible cause of adverse effects in transfused patients, including febrile transfusion reactions, inducing HLA alloimmunization, immunosuppression, disease transmission, and TGVHD. Reducing white cells can reduce the incidence of all of these complications except TGVHD. Currently, white cells are removed by infusion through filters that trap the cells. This can be done either at the bedside, in the blood bank, or at the donor center. The majority of red cells provided by blood centers in many areas of the country are already leukoreduced, eliminating the need for labor-intensive filtration at the transfusion center or bedside. Platelets collected by plateletpheresis methods can also be made leukocyte-poor. The current indications for leukodepleted productions are:

1. Prevention of febrile transfusion reactions in patients with previous documented reactions
2. Prevention of HLA alloimmunization (ineffective if patient has received 1 or more blood products not leukodepleted or is already HLA immunized)
3. Prevention of cytomegalovirus (CMV) infection

 

 

CMV-NEGATIVE BLOOD

CMV can be transmitted through any cellular blood product—red cells and platelets. For patients who are CMV-negative and receiving transplants, especially stem cell transplants, a new CMV infection can be devastating.²¹ For years only blood from CMV-negative donors was used to transfuse CMV-negative patients. This policy is effective in preventing CMV infection, but because 50% of the population is positive for CMV antibodies, it may potentially lead to shortages of products that could be transfused to the patient. Currently, leukoreduced blood products are used since leukofiltration of the blood is just as effective as transfusion of CMV-negative blood in preventing infections and allows greater use of all blood products.²³

COMPLICATIONS OF TRANSFUSIONS

HEMOLYTIC TRANSFUSION REACTION

There are 2 forms of hemolytic reactions—immediate and delayed.²⁴ The immediate reaction is associated with fevers, hypotension, back pain, and oliguria. In severe cases, DIC and renal failure may occur. The immediate reaction is due to transfusion of blood that reacts with the recipient’s preformed high-titer blood antigen antibodies, most often to ABO. This is fatal 2% of the time and occurs almost always as a result of errors in correct identification of the patient. Reactions are due to recipient antibodies attacking donated RBCs, resulting in release of hemoglobin and red cell membrane–antigen complexes. These complexes are believed to lead to the hypotension, fevers, chills, and renal damage associated with the hemolytic reaction. Treatment consists of immediately stopping the transfusion, notifying the blood bank, vigorous intravenous hydration to keep the urine output over 100 mL/hr, and supportive therapy.

The delayed reaction can range in severity from an abrupt drop in the hematocrit to normal response to transfusion but the patient developing a positive Coombs’ test. The delayed response is due to an anamnestic response to blood-group antigens. When the patient is exposed to the same antigen, there is a rise in antibody titer leading to the reaction. Some alloantibodies can lead to a brisk reaction, most often anti-Kidd. The frequency with which delayed transfusion reactions occur is underestimated because mild reactions often do not get worked up or even discovered.

ALLERGIC REACTIONS

Allergic reactions are common (1%–3% of transfusions) and occur in patients having antibodies to proteins in donor blood, which can lead to hives and itching with transfusions. Most of the time these allergic reactions are mild and can be treated with antihistamines. Prophylaxis with antihistamines is not indicated for future transfusions unless the reactions are frequent. Rarely these reactions can be associated with shock and hypotension. Patients who are immunoglobulin (Ig) A–deficient can develop anaphylactic reactions to IgA-containing blood products. Patients with severe allergic reactions need to have their IgA measured and, if deficient, receive only washed units or plasma from IgA-deficient donors to prevent future severe reactions.

FEBRILE REACTIONS

The most common transfusion reaction is a febrile reaction that occurs after the transfusion starts and that sometimes can be complicated by chills. This reaction often occurs due to the presence of leukocyte debris and cytokines in the donated blood. Therapy is supportive and involves stopping the transfusion and administering acetaminophen, but since hemolytic transfusion reactions can present with fever all patients need to be thoroughly evaluated. The incidence of reactions can be decreased by using leukodepleted blood and plateletpheresis platelets. Most patients do not benefit from receiving prophylactic acetaminophen for future transfusion unless they have multiple reactions.

 

 

TRANSFUSION-RELATED ACUTE LUNG INJURY

Once thought a rare complication, transfusion-related acute lung injury (TRALI) is increasingly being recognized, with an incidence of approximately 1:5000 patients; it is now the most frequent cause of transfusion-related death.^24,25 TRALI is noncardiac pulmonary edema and typically manifests clinically with hypoxemia, fever, bilateral infiltrates, and hypotension 2 to 6 hours after blood is given. Ventilatory support is often required. Recovery is usually rapid (24–48 hours) and complete. The etiology is complex. In many cases, transfused anti-HLA antibodies react with the recipient’s white cells leading to pulmonary damage. Another theory is that transfusion of preformed cytokines leads to pulmonary damage. Because plasma products from multiparous women are most often associated with anti-HLA antibodies, the restricted use of blood products from women has decreased the incidence of TRALI over the past few years.²⁶

TRANSFUSION-ASSOCIATED CIRCULATORY OVERLOAD

Increasingly it being recognized that volume overload resulting from transfusions can lead to significant morbidity.²⁷ Patients with heart or renal disease or patients who already have compromised fluid status are at risk for transfusion-associated circulatory overload (TACO). Another risk factor is transfusion of multiple blood products. Patients with TACO develop dyspnea within 6 hours of transfusion, but do not have fever or rash with the dyspnea. The diagnosis is made by demonstrating circulatory overload (eg, high venous pressure, B-type natriuretic peptide). Treatment is aggressive diuresis. Strategies to prevent TACO include judicious use of blood products, especially in patients at risk for TACO, and the use of prophylactic diuretics, especially with red cell or plasma transfusions.²⁸

TRANSFUSION-RELATED GRAFT-VERSUS-HOST DISEASE

TGVHD is a rare reaction, but one that is most often fatal.²⁹ TGVHD occurs when donor lymphocytes attack the blood recipient’s organs—skin, liver, intestines, and marrow. This is very rare in the normal blood recipient unless the donor and recipient share some HLA haplotypes.³⁰ In immunosuppressed patients, TGVHD can occur with lesser degrees of HLA similarity, with cases reported in blood recipients who are mainly patients with Hodgkin disease or acute leukemia undergoing chemotherapy, and in patients receiving purine analogs. TGVHD had not been reported in AIDS patients despite profound immunosuppression, perhaps because the milieu of the patient does not allow lymphocyte expansion. Symptoms of TGVHD are an erythematous rash that may progress to epidermal toxic necrolysis, liver dysfunction, diarrhea, and pancytopenia. TGVHD is prevented by irradiating blood products given to at-risk patients with 2500 to 3500 rads. Directed blood donation from all blood relatives should also be irradiated. TGVHD cannot be prevented by leukopoor blood because the minute amount of lymphocytes that are not filtered still can lead to these complications.

POST-TRANSFUSION PURPURA

Patients with post-transfusion purpura (PTP) develop severe thrombocytopenia (< 10 × 10³/µL) with often severe bleeding 1 to 2 weeks after receiving any type of blood product.³¹ Patients who develop PTP most often lack platelet antigen PLA1 or other platelet antigens. For unknown reasons, exposure to the antigens from the transfusion leads to rapid destruction of the patient’s own platelets. The diagnostic clue is thrombocytopenia in a patient, typically female, who has received a red cell or platelet blood product in the past 7 to 10 days. Treatment consists of intravenous immunoglobulin³² and plasmapheresis to remove the offending antibody. If patients with a history of PTP require further transfusions, only PLA1-negative platelets should be given.

IRON OVERLOAD

Every transfusion of red cells delivers approximately 250 mg of iron to the recipient. Since there is no natural way of ridding the body of iron, heavily transfused patients are at risk of iron overload. This is most often seen in children heavily transfused for thalassemia. Starting in the second decade of life, these individuals will develop endocrinopathies due to iron overload, liver problems, and often fatal cardiomyopathies. Studies have shown that chelation of iron with deferoxamine can be effective in preventing this fatal complication.³³ Oral iron chelators such as deferasirox and deferiprone are also effective. The risk of iron overload in heavily transfused patients with myelodysplasia or other transfusion-dependent anemias is unclear, and uncertainty exists about the need for chelation.³⁴

 

Young patients who face years of transfusions should be started on iron chelation to avoid iron overload. For older patients with transfusion-dependent anemia, iron chelation therapy should be considered if their life expectancy is long (years to decades) or special studies such as T2-weighted cardiac magnetic resonance imaging showing iron overloading.³⁵

INFECTIOUS COMPLICATIONS

Concern over transmission of HIV infection via blood products in the late 1980s led to both a reduction in blood product use and a greater awareness of infectious complications of transfusion and their prevention. However, no blood product can ever be assumed to be safe for 2 reasons. One is that blood products can transmit infections during a “window period”—the time before a contaminated product can be detected by testing. The second is that blood is not screened for all potential infections (eg, babesiosis or new infections such as West Nile virus at the start of the outbreak). Risk of infection is reduced in 2 ways: deferral of potential infectious donors and blood product testing.

As part of the donation process, potential blood donors are asked a series of questions to see if they have risk factors for infections (eg, recent travel to malarious areas, recent tattoos), and if they answer positive are deferred from donating blood. Blood products are then tested for infectious agents by a combination of methods including detection of viral antigen, antibody response to infections, and more recently polymerase chain reaction (PCR).³⁶ Current screening includes syphilis testing; testing for antibodies to HIV, HTLV (human T-lymphotropic virus), hepatitis C virus, hepatitis B core antigen (HBcAg), hepatitis B surface antigen, and PCR for HIV, hepatitis B virus, HCV, and West Nile virus. Some centers also test for Trypanosoma cruzi, the cause of Chagas disease.

In the past, the numerically most common transfusion-related disease was hepatitis, first B and then C.^37,38 The first step in eliminating these infections was to stop paying donors for blood products. With the introduction of effective testing for hepatitis B and then C, the incidence of transfusion-related hepatitis has plummeted.³⁶ For example, with the introduction of a diagnostic test for hepatitis C, the estimated risk has fallen from 5% to less than 1 per million. Currently, the risk of transmission of hepatitis B and C, HIV, and HTLV is less than 1 in a million.³⁸

Despite this testing, blood transfusions can transmit a variety of infections, including malaria and babesiosis.³⁹ Any new blood-borne infection introduced into the population can get into the blood supply as well. For example, at the start of the West Nile virus epidemic, there was a cluster of transfusion-transmitted cases that resulted in severe and sometimes fatal illness in immunosuppressed patients, but this issue has been addressed with the development of a PCR assay for screening blood.⁴⁰ The rate of transfusion-related babesiosis has been increasing and screening for the causative parasite is being considered.

 

 

MASSIVE TRANSFUSIONS

Acutely bleeding patients can require large amounts of transfusion products. Early data showed high mortality rates with transfusion of more than 20 units of blood,⁴¹ but with modern blood banking techniques and improved laboratory testing, this rate has decreased dramatically, with survival rates of 43% to 70% in patients transfused with more than 50 units of blood.⁴²

The basic approach to massive transfusions is to first transfuse the patient to maintain hemodynamic stability while specific blood tests are being obtained, and then to use the results of these early tests to guide the rest of the resuscitation. An important component is the ability to rapidly deliver standard packages of red cells, usually 6 to 10 units at a time, to the bleeding patient. To avoid delay while the patient’s blood is being typed, the first products delivered are blood group O Rh-positive units. Given the shortage of Rh-negative blood, this should be reserved for only empiric therapy of women of child-bearing age. Once the blood type is known, the patient can be switched over to type-specific blood.

In the past decade, there has been a shift toward increasing the amount of plasma given to patients receiving massive transfusions. This shift has occurred for 2 reasons. One is that modeling of coagulation changes in massive bleeding suggests the need for larger amounts of plasma to correct defects than have previously been recommended.⁴³ The other reason is based on analysis of resuscitation protocols used in military and civilian trauma centers showing that giving red cells and plasma units in a 1:1 ratio appears to be associated with improved outcomes in massive transfusion. Several studies have extended this concept to platelets, again suggesting improved survival with 1 unit of random donor platelets given 1:1 with red cells and plasma units. The PROPPR (Prospective Observational Multicenter Major Trauma Transfusion) study compared a 1:1:1 to 1:1:2 ratio in patients with severe trauma and major bleeding and found less exsanguination and faster achievement of hemostasis in the first 24 hours.⁴⁴ This has led to the widespread adoption of the 1:1 ratio by most trauma centers, and by default to other massive transfusion situations despite the lack of clinical trial data.⁴⁵

One barrier to increased use is that plasma is kept frozen and requires 20 minutes to thaw. Many institutions are now keeping inventories of thawed plasma available for immediate use, ranging from 2 to 4 units of group AB plasma to keeping their entire inventory as liquid plasma.⁴⁶ Plasma that is thawed but not used can be relabeled as “thawed plasma” and kept for up to 5 days. Also, many centers now use group A plasma for massive transfusions as this rarely leads to transfusion reactions and is much more available.⁴⁷ Research is currently under way on lyophilized plasma, which can be stored at room temperature and can be rapidly reconstituted for emergency use.

The standard approach for laboratory testing is obtaining 5 tests: hematocrit, platelet count, INR/prothrombin time, activated partial thromboplastin time (aPTT), and fibrinogen.⁴⁸ Product selection is guided by these tests, and they are repeated at regular intervals during the massive transfusion. A typical protocol is shown in Table 3. It is important as part of any protocol to have a flow chart that records laboratory results and products given that any member of the team can easily view.

 

The transfusion threshold for a low hematocrit depends on the stability of the patient. If the hematocrit is below 30% and the patient is bleeding or hemodynamically unstable, one should transfuse packed red cells. Stable patients can tolerate lower hematocrits, and an aggressive transfusion policy may even be detrimental.^2,49 If the patient is bleeding, has florid DIC, or has received platelet aggregation inhibitors, then keeping the platelet count above 50 × 10³/µL is reasonable. There are data regarding massive transfusion showing that keeping the platelet count above 50 × 10³/µL resulted in less microvascular bleeding.⁵⁰ The conventional dose of platelets is 6 to 8 platelet concentrates or 1 plateletpheresis unit.

While in the past fibrinogen targets of 50 to 100 mg/dL were recommended, recent data indicate that a target of 150 mg/dL or higher may be more appropriate.^51–53 Severe fibrinolysis may occur in certain clinical situations such as brain injuries, hepatic trauma, or ischemic limb reperfusion, and the use of large amounts of cryoprecipitate can be anticipated. In patients with an INR greater than 2 and an abnormal aPTT, one can give 2 to 4 units of FFP. For an aPTT greater than 1.5 times normal, 2 to 4 units of plasma should be given. Elevation of the aPTT above 1.8 times normal control is associated with microvascular bleeding in trauma patients.⁵⁴ Patients with marked abnormalities (eg, anaPTT more than 2 times normal) may require aggressive therapy with at least 15 to 30 mL/kg (4–8 units for an average adult) of plasma.⁵⁵

Recently there has been increasing interest in the use of thromboelastography (TEG) in massive transfusion.⁵⁶ This is a point-of-care assay performed on fresh whole blood that can assess multiple facets of hemostasis, including coagulation, platelet function, and fibrinolysis.^57,58 TEG is performed by placing a 0.35-mL sample of whole blood into an oscillating container with a sensor pin that measures the force of thrombus formation. TEG measures 5 parameters:

• r time: time from starting TEG until clot formation
• K time: time needed for tracing to go from 2 mm to 20 mm
• alpha angle: slope of tracing between r and K time
• MA: greatest amplitude of TEG tracing
• Whole blood lysis index: amplitude of tracing 60 minutes after MA.

Several centers have incorporated TEG into resuscitation protocols that include standardized strategies for responding to abnormalities. Data suggest that use of TEG may decrease the use of blood products, especially in cardiac surgery, but this has not been prospectively studied in massive transfusions.^56,59

COMPLICATIONS OF MASSIVE TRANSFUSIONS

Electrolyte abnormalities are unusual even in patients who receive massive transfusions.⁶⁰ Platelet concentrates and plasma contain citrate that can chelate calcium. However, the citrate is rapidly metabolized, and it is rare to see clinically significant hypocalcemia. Although empiric calcium replacement is often recommended, one study suggests that this is associated with a worse outcome and should not be done.⁶¹ If hypocalcemia is a clinical concern, then levels should be drawn to guide therapy. Stored blood is acidic, with a pH of 6.5 to 6.9. However, acidosis attributed solely to transfused blood is rare and most often is a reflection of the patient’s stability. Empirical bicarbonate replacement has been associated with severe alkalosis and is not recom mended.^62,63 Although potassium leaks out of stored red cells, even older units of blood contain only 8 mEq/L of potassium, so hyperkalemia is usually not a concern.

 

PATIENTS WITH AUTOIMMUNE HEMOLYTIC ANEMIA

Patients with autoimmune hemolytic anemia can be difficult to transfuse,⁶⁴ because the autoantibody can interfere with several aspects of the transfusion services evaluation. In some patients the autoantibody can be so strong that the patient’s blood type cannot be determined. In most patients, the final step of the cross-match—mixing the donor blood with recipient plasma—will show noncompatibility due to the autoantibodies reacting with any red cells.

The first step when transfusing a patient with autoimmune hemolytic anemia is to draw several tubes of blood for the transfusion service before any potential transfusions. This allows the transfusion service to remove the autoantibodies so they can screen for underlying alloantibodies. Second, if the patient requires immediate transfusion, then type-specific or O-negative blood should be given. If the patient has not been recently (months) transfused, the incidence of a severe transfusion reaction is low. The first unit should be infused slowly with close observation of the patient. For patients who have been multiply transfused, the use of an “in-vivo” cross-match may be helpful. This is where the patient is slowly transfused 10 to 15 mL of blood over 15 minutes. The the plasma and urine are then assessed for signs of hemolysis and, if negative, the remaining product is given.

REFUSAL OF BLOOD PRODUCTS

The initial step in managing patients who refuse blood products is to find out why they are refusing them. Many patients have an exaggerated fear of HIV and other infectious agents, so discussing the very low risk for infection transmission can often resolve the situation. The most common reason for refusal of blood products is religious belief. Jehovah’s Witness patients will refuse blood products due to their interpretation of the Bible.⁶⁵ All members will refuse red cells, plasma, and platelets, while decisions about “derived” blood products—products made by manipulation of the original donated units—are a matter of conscience. These include cryoprecipitate, intravenous gammaglobulin, and albumin.

In an elective situation, the first step is to discuss with the patient those products that are a matter of conscience and clearly document this. The patient’s blood count and iron stores should be assessed to identify any correctible causes of anemia or low iron stores before surgery. The use of erythropoietin to correct blood counts before surgery is controversial, as this may increase thrombosis risk and is contraindicated in patients with curable tumors.

For patients with acute blood loss, use of intravenous iron combined with high-dose erythropoietin is the most common approach to raise the blood count.⁶⁵ A recommended erythropoietin dose is 300 units/kg 3 times a week, dropping to 100 units/kg 3 times weekly until the goal hematocrit is reached. Another often overlooked step is to consolidate and minimize laboratory testing. The most important step is to be respectful of the patient and their beliefs. Many larger cities have liaisons that can help with interactions between Jehovah’s Witness patients and the health care system.

 

 

NON-TRANSFUSION THERAPIES FOR ACUTE BLEEDING

DESMOPRESSIN

Desmopressin (DDAVP) is a synthetic analog of antidiuretic hormone that raises the levels of both factor VIII and von Willebrand protein severalfold.⁶⁶ Desmopressin is effective in supporting hemostasis in patients with a wide variety of congenital and acquired bleeding disorders. However, desmopressin does not reduce blood loss before routine surgery in a healthy patient and should not be used for this purpose.⁶⁷

TRANEXAMIC ACID

Tranexamic acid is an antifibrinolytic agent that blocks the binding of plasmin to fibrin.⁶⁸ This agent was first shown to be useful in disorders that involve excessive fibrinolysis^69–73 or as adjunctive therapy for oral or dental procedures in patients with a bleeding diathesis. In patients with severe thrombocytopenia, the use of antifibrinolytic agents may reduce bleeding. Increasing data shows that tranexamic acid can prevent blood loss in a variety of surgeries including heart bypass, liver transplantation, and orthopedic surgery.⁷⁴ Patients across these settings have decreased blood loss and need for transfusion with no increased risk of thrombosis. The CRASH-2 study showed that the use of tranexamic acid significantly reduced mortality in trauma patients.⁷⁵ The WOMEN trial demonstrated that 1 g of tranexamic acid given to women with blood loss of more than 500 mL after vaginal delivery or 1000 mL after cesarean section has a risk reduction of death of 0.81 with no increased risk of thrombosis.⁷⁶ Given this abundant data, it is clear tranexamic acid needs to be part of any massive transfusion protocol.⁷⁷

RECOMBINANT FACTOR VIIa

Recombinant factor VIIa (rVIIa) was originally developed as a “bypass” agent to support hemostasis in hemophiliacs.⁷⁸ However, the use of rVIIa for a wide array of bleeding disorders, including patients with factor VII and XI deficiency and Glanzmann thrombasthenia, has been reported.⁷⁹ Increasingly, rVIIa is being used as a “universal hemostatic agent” for patients with uncontrolled bleeding from any mechanism.⁸⁰ Multiple case reports have described the use of rVIIa for bleeding in cardiac surgery patients, obstetrical bleeding, reversal of anticoagulation, and trauma.⁸¹ Unfortunately, little formal trial data exists to put these anecdotes into perspective, and formal review of clinical trial results has shown no benefit.^82,83 However, when used in older patients, especially those with vascular risk factors, the risk of arterial thrombosis appears to increase.⁸⁴ In the trials for intracranial hemorrhage, the thrombosis rate was 5% to 9%, and rates up to 10% for arterial events were seen in older patients in a review of all trials.^85–87 Given the lack of data but the evidence of risk, rVIIa use should be restricted to patients with documented bleeding disorders that have been shown to benefit by its use.

INTRODUCTION

Transfusion therapy is an essential part of hematology practice, allowing for curative therapy of diseases such as leukemia, aplastic anemia, and aggressive lymphomas. Nonetheless, transfusions are associated with significant risks, including transmission of infections and transfusion-related reactions. Controversy remains about key issues in transfusion therapy, such as triggers for red cell transfusions. This article reviews the available blood products (Table 1) and indications for transfusion along with the associated risks, and also discusses specific clinical situations, such as massive transfusion.

BLOOD PRODUCTS

WHOLE BLOOD

Whole blood is the product of 1 unit of donated blood plus anticoagulant/preservative, and by definition contains 1 unit of plasma and red cells. Whole blood can be stored for 5 weeks. Although it was the standard product in the past, whole blood is rarely used since 1 unit of donated blood can now be fractionated into 1 unit of red blood cells (RBC), 1 unit of platelets, and 1 unit of fresh frozen plasma (FFP). Thus, the use of whole blood for just a single transfusion represents a waste of resources. There are 2 exceptions. One is autologous blood donations, which are whole blood units. Second, whole blood is increasingly being used in massive transfusions for trauma patients, with the rationale being that all essential blood components are being transfused at once.¹

PACKED RED CELLS

The remaining red cell mass after most of the plasma is removed is called the “packed” red cell unit (hematocrit = 70%–80%), and so red cells are often called “packed” red cells, or PRBC. A preservative is added to improve the flow of blood and to provide “nutrients” for the red cells, and this reduces the hematocrit to approximately 60%. The volume of a red cell unit is approximately 340 mL. In the average adult, 1 unit of RBC raises the hematocrit by 3%. The indications for transfusion of red cells are to increase red cell mass, and thus oxygen delivery, in patients who are compromised by their anemia.

Several randomized trials have helped define the indications for red cell transfusions and justify lower hematocrit thresholds for initiating transfusion. The TRICC (Transfusion Requirements in Critical Care Investigators for the Canadian Critical Care Trials Group) trial showed that in critical care patients (30-day mortality, 18.7%–23.3%), a conservative transfusion strategy of waiting until the hematocrit was below 21% had the same outcomes as transfusing at a threshold of 24%.² The TRACS (Transfusion Requirements After Cardiac Surgery) trial showed that a hematocrit target of 24% had the same benefit as a target of 30% in patients who had undergone cardiac bypass surgery.³ For patients with acute myocardial infarction, the outcomes were worse with aggressive transfusion at a hematocrit of 30% compared to 24%.⁴ In patients with upper gastrointestinal bleeding, a hemoglobin transfusion trigger of 7 g/dL was associated with a lower mortality than a trigger of 9 g/dL (5% versus 9%).⁵ Finally, the FOCUS (Transfusion Trigger Trial for Functional Outcomes in Cardiovascular Patients Undergoing Surgical Hip Fracture Repair) trial showed that in older patients (average age 80 years) who had undergone hip fracture surgery, transfusions based on symptoms and not a fixed trigger of 30% had the same outcomes but considerable savings in blood products.⁶ Based on these trials, decisions regarding when to transfuse patients should be based on symptoms and not “numbers.” Young patients, especially those with reversible anemias, can tolerate low blood counts and should not be transfused based on an arbitrary number.

PLATELETS

Several types of platelet products exist. One unit of platelet concentrate is derived from 1 unit of donor blood. Plateletpheresis from volunteer donors is also used to harvest platelets, with the resulting product referred to as plateletpheresis platelets. One unit of single-donor (pheresis) platelets is equivalent to 6 platelet concentrates. Finally, HLA-matched platelets are single-donor pheresis units that are obtained from an HLA-matched donor. This product should be ordered only if there is evidence of HLA antibodies (see Platelet Alloimmunization section).

The dose of platelets for the average patient is 6 units of platelet concentrate or 1 pheresis unit. In theory, 1 unit of platelet concentrate can raise the count by 5 to 7 × 10³/µL, but often this response is blunted by concurrent illness or bleeding. In patients who appear to have a poor response, the platelet count can be checked 15 minutes after platelet infusion. No rise or a minimal rise (< 2 × 10³/µL) in the platelet count is suggestive of platelet refractoriness, while a good 15-minute response but poor 24-hour count is more suggestive of consumption—fever, sepsis, drug, or splenomegaly—and not refractoriness.

The indication for platelet transfusion depends on the clinical situation. For patients with immune thrombocytopenia, platelets should not be transfused unless there is life-threatening bleeding. For stable patients with marrow aplasia from chemotherapy, a cut-off of a morning platelet count of less than 10 × 10³/µL has been shown to be as safe as higher levels for prophylactic transfusions.⁷ For patients with active bleeding, the platelet count should be kept above 50 × 10³/µL. Patients with acquired or inherited platelet dysfunction may benefit from transfusion no matter the platelet count.

 

 

Platelet Alloimmunization

Patients exposed to transfused white cells with different HLA antigens can develop antibodies to these antigens.⁸ Anti-HLA antibodies are common in patients who previously have received transfused blood that is not leukodepleted and in patients who have been pregnant. Since platelets carry class I HLA antigens, they will be rapidly destroyed by anti-HLA antibodies when transfused into these patients. In patients transfused for aplastic anemia or myelodysplasia, as many as 90% will become HLA-immunized. The incidence is lower in patients receiving chemotherapy but still can be as high as 60% to 90%.^9,10 Patients who have developed anti-HLA antibodies can respond to transfused platelets matched for HLA antigens. Unfortunately, some patients will either be a rare HLA type or be so heavily immunized that they will not respond to any platelet transfusion.

The significance of alloimmunization centers on 2 concepts: recognition and avoidance. Patients with HLA antibodies will fail to have an increment of their platelet counts with transfusions. Accordingly, patients who do not experience an increase in their count 15 minutes after the transfusion may have HLA antibodies. One can test for the presence of anti-HLA antibodies, although some patients instead have specific antiplatelet antibodies that will not respond to HLA-matched platelets. In patients who have been pregnant or previously transfused and are scheduled to undergo transplant or aggressive chemotherapy, it is wise to test for anti-HLA antibodies in order to plan their transfusion needs. The evidence suggests that transfused white cells are responsible for initiating the anti-HLA response. Trials have shown that giving leukodepleted blood products may reduce the incidence of alloimmunization, so patients who are not HLA-alloimmunized should receive only leukodepleted products.¹¹A difficult problem is bleeding in patients who are refractory to platelet transfusion.¹² Patients who test positive for the presence of anti-HLA antibodies can receive transfusions of HLA-matched platelets.¹³ Unfortunately, matched platelet transfusions are not effective in 20% to 70% of these patients. Also, since some loci are difficult to match, effective products may be unavailable. Finally, as many as 25% of patients have antiplatelet antibodies in which HLA-matched products will be ineffective. Platelet cross-matching can be performed to find compatible units for these patients, but this may not always be successful. In the patient who is totally refractory to platelet transfusion, consider drugs as an etiology of antiplatelet antibodies (especially vancomycin).¹⁴ Use of antifibrinolytic agents such as epsilon-aminocaproic acid or tranexamic acid may decrease the incidence of minor bleeding, but these are ineffective for major bleeding. “Platelet drips”—infusing either a platelet concentrate per hour or 1 plateletpheresis unit every 6 hours—may be given as a continuous infusion, but there is no evidence that this is helpful.¹⁵

FRESH FROZEN PLASMA

FFP is made from 1 unit of donated whole blood, with an average volume of 225 mL per unit. One unit of FFP can increase coagulation factor levels by 5% and fibrinogen by 10 mg/dL in the average stable patient. FFP can take 20 to 30 minutes to thaw before use, so in situations where FFP is needed quickly, the blood bank must be informed to “keep ahead” some units. Units of FFP that have been thawed but not used can be stored refrigerated for 5 days to prevent wasting blood products.

The indications for FFP are limited to several situations. These include a documented coagulation defect that can be corrected by a reasonable amount of FFP, such as factor V deficiency and factor XI deficiency, disseminated intravascular coagulation (DIC), reversal of warfarin, and massive transfusions. FFP is also used for the therapy of thrombotic thrombocytopenic purpura.

There is little justification for FFP transfusion in many of the clinical settings in which it is commonly used. For example, FFP is given for minor elevations of the INR in patients with liver disease, despite literature showing not only that the INR rise is not reflective of coagulation defects, but also that patients with liver disease may even be thrombophilic.^16,17 Reviews of FFP use found limited evidence-based indications for its use.^18,19 Also, several studies have shown that transfusion of FFP is not effective at reversing minor elevations of the INR (1.3–1.8).²⁰ In a meta-analysis, FFP was associated with increased risk for lung injury and a trend toward increased mortality.¹⁸

 

 

CRYOPRECIPITATE

Cryoprecipitate is produced from 1 unit of FFP that is thawed at 4°C. The precipitate is resuspended with 10 mL of saline or FFP and refrozen for storage. One unit contains at least 150 mg of fibrinogen and 80 units of factor VIII, along with von Willebrand factor. Thawing time for cryoprecipitate is approximately 20 minutes.

Cryoprecipitate is used to raise the fibrinogen level in patients with DIC or massive transfusion with hemodilution. It is third-line therapy in the treatment of type 1 von Willebrand disease and is second-line therapy in the treatment of patients with other types of von Willebrand disease. Currently, von Willebrand factor concentrates are the preferred replacement product for von Willebrand disease. Cryoprecipitate can be used as a source of factor VIII for hemophiliacs, but the preferred product for these patients is the super pure factor VIII concentrates or recombinant products. Cryoprecipitate can also be used to shorten the bleeding time of uremic patients, but its effectiveness for this is controversial.

GRANULOCYTES

Granulocytes are harvested by leukopheresis of normal donors, with a target yield of 1010 granulocytes from each donor. To reach this target, the donors are often “stimulated” with neutrophil growth factors. The harvesting procedure can take 3 hours and is associated with some risks to the donor (eg, citrate toxicity). The current indications for granulocytes are very limited since the advent of neutrophil growth factors and improved antimicrobials.²¹ They can be useful in the neutropenic patient with a documented bacterial infection in whom the white blood cell count is not expected to recover in the near future. Given the difficulty of keeping the count up, these transfusions have been mainly used in treating small children.

SPECIAL BLOOD PRODUCTS

IRRADIATED BLOOD PRODUCTS

Irradiation of blood is performed for only one reason: to prevent transfusion-related graft-versus-host disease (TGVHD) (Table 2).²² The irradiation can be performed at the blood center or in the transfusion service of larger hospitals. The units are not radioactive and can be transfused safely to other patients. There is increased leakage of potassium in irradiated units of blood, so the units need to be transfused within 14 days; in patients potentially sensitive to potassium (eg, neonates), the units must be transfused within 24 hours. Patients undergoing stem cell transplant, those receiving either interuterine transfusions or products from relatives, any patient with Hodgkin disease or receiving purine analogs or alemtuzumab, and patients with severe congenital immune deficiencies should receive irradiated blood. Most would also advocate that patients with hematologic malignances receiving chemotherapy receive irradiated products, but this is more controversial.

LEUKPDEPLETED BLOOD

Contamination of blood products by white blood cells is increasingly being recognized as a possible cause of adverse effects in transfused patients, including febrile transfusion reactions, inducing HLA alloimmunization, immunosuppression, disease transmission, and TGVHD. Reducing white cells can reduce the incidence of all of these complications except TGVHD. Currently, white cells are removed by infusion through filters that trap the cells. This can be done either at the bedside, in the blood bank, or at the donor center. The majority of red cells provided by blood centers in many areas of the country are already leukoreduced, eliminating the need for labor-intensive filtration at the transfusion center or bedside. Platelets collected by plateletpheresis methods can also be made leukocyte-poor. The current indications for leukodepleted productions are:

1. Prevention of febrile transfusion reactions in patients with previous documented reactions
2. Prevention of HLA alloimmunization (ineffective if patient has received 1 or more blood products not leukodepleted or is already HLA immunized)
3. Prevention of cytomegalovirus (CMV) infection

 

 

CMV-NEGATIVE BLOOD

CMV can be transmitted through any cellular blood product—red cells and platelets. For patients who are CMV-negative and receiving transplants, especially stem cell transplants, a new CMV infection can be devastating.²¹ For years only blood from CMV-negative donors was used to transfuse CMV-negative patients. This policy is effective in preventing CMV infection, but because 50% of the population is positive for CMV antibodies, it may potentially lead to shortages of products that could be transfused to the patient. Currently, leukoreduced blood products are used since leukofiltration of the blood is just as effective as transfusion of CMV-negative blood in preventing infections and allows greater use of all blood products.²³

COMPLICATIONS OF TRANSFUSIONS

HEMOLYTIC TRANSFUSION REACTION

There are 2 forms of hemolytic reactions—immediate and delayed.²⁴ The immediate reaction is associated with fevers, hypotension, back pain, and oliguria. In severe cases, DIC and renal failure may occur. The immediate reaction is due to transfusion of blood that reacts with the recipient’s preformed high-titer blood antigen antibodies, most often to ABO. This is fatal 2% of the time and occurs almost always as a result of errors in correct identification of the patient. Reactions are due to recipient antibodies attacking donated RBCs, resulting in release of hemoglobin and red cell membrane–antigen complexes. These complexes are believed to lead to the hypotension, fevers, chills, and renal damage associated with the hemolytic reaction. Treatment consists of immediately stopping the transfusion, notifying the blood bank, vigorous intravenous hydration to keep the urine output over 100 mL/hr, and supportive therapy.

The delayed reaction can range in severity from an abrupt drop in the hematocrit to normal response to transfusion but the patient developing a positive Coombs’ test. The delayed response is due to an anamnestic response to blood-group antigens. When the patient is exposed to the same antigen, there is a rise in antibody titer leading to the reaction. Some alloantibodies can lead to a brisk reaction, most often anti-Kidd. The frequency with which delayed transfusion reactions occur is underestimated because mild reactions often do not get worked up or even discovered.

ALLERGIC REACTIONS

Allergic reactions are common (1%–3% of transfusions) and occur in patients having antibodies to proteins in donor blood, which can lead to hives and itching with transfusions. Most of the time these allergic reactions are mild and can be treated with antihistamines. Prophylaxis with antihistamines is not indicated for future transfusions unless the reactions are frequent. Rarely these reactions can be associated with shock and hypotension. Patients who are immunoglobulin (Ig) A–deficient can develop anaphylactic reactions to IgA-containing blood products. Patients with severe allergic reactions need to have their IgA measured and, if deficient, receive only washed units or plasma from IgA-deficient donors to prevent future severe reactions.

FEBRILE REACTIONS

The most common transfusion reaction is a febrile reaction that occurs after the transfusion starts and that sometimes can be complicated by chills. This reaction often occurs due to the presence of leukocyte debris and cytokines in the donated blood. Therapy is supportive and involves stopping the transfusion and administering acetaminophen, but since hemolytic transfusion reactions can present with fever all patients need to be thoroughly evaluated. The incidence of reactions can be decreased by using leukodepleted blood and plateletpheresis platelets. Most patients do not benefit from receiving prophylactic acetaminophen for future transfusion unless they have multiple reactions.

 

 

TRANSFUSION-RELATED ACUTE LUNG INJURY

Once thought a rare complication, transfusion-related acute lung injury (TRALI) is increasingly being recognized, with an incidence of approximately 1:5000 patients; it is now the most frequent cause of transfusion-related death.^24,25 TRALI is noncardiac pulmonary edema and typically manifests clinically with hypoxemia, fever, bilateral infiltrates, and hypotension 2 to 6 hours after blood is given. Ventilatory support is often required. Recovery is usually rapid (24–48 hours) and complete. The etiology is complex. In many cases, transfused anti-HLA antibodies react with the recipient’s white cells leading to pulmonary damage. Another theory is that transfusion of preformed cytokines leads to pulmonary damage. Because plasma products from multiparous women are most often associated with anti-HLA antibodies, the restricted use of blood products from women has decreased the incidence of TRALI over the past few years.²⁶

TRANSFUSION-ASSOCIATED CIRCULATORY OVERLOAD

Increasingly it being recognized that volume overload resulting from transfusions can lead to significant morbidity.²⁷ Patients with heart or renal disease or patients who already have compromised fluid status are at risk for transfusion-associated circulatory overload (TACO). Another risk factor is transfusion of multiple blood products. Patients with TACO develop dyspnea within 6 hours of transfusion, but do not have fever or rash with the dyspnea. The diagnosis is made by demonstrating circulatory overload (eg, high venous pressure, B-type natriuretic peptide). Treatment is aggressive diuresis. Strategies to prevent TACO include judicious use of blood products, especially in patients at risk for TACO, and the use of prophylactic diuretics, especially with red cell or plasma transfusions.²⁸

TRANSFUSION-RELATED GRAFT-VERSUS-HOST DISEASE

TGVHD is a rare reaction, but one that is most often fatal.²⁹ TGVHD occurs when donor lymphocytes attack the blood recipient’s organs—skin, liver, intestines, and marrow. This is very rare in the normal blood recipient unless the donor and recipient share some HLA haplotypes.³⁰ In immunosuppressed patients, TGVHD can occur with lesser degrees of HLA similarity, with cases reported in blood recipients who are mainly patients with Hodgkin disease or acute leukemia undergoing chemotherapy, and in patients receiving purine analogs. TGVHD had not been reported in AIDS patients despite profound immunosuppression, perhaps because the milieu of the patient does not allow lymphocyte expansion. Symptoms of TGVHD are an erythematous rash that may progress to epidermal toxic necrolysis, liver dysfunction, diarrhea, and pancytopenia. TGVHD is prevented by irradiating blood products given to at-risk patients with 2500 to 3500 rads. Directed blood donation from all blood relatives should also be irradiated. TGVHD cannot be prevented by leukopoor blood because the minute amount of lymphocytes that are not filtered still can lead to these complications.

POST-TRANSFUSION PURPURA

Patients with post-transfusion purpura (PTP) develop severe thrombocytopenia (< 10 × 10³/µL) with often severe bleeding 1 to 2 weeks after receiving any type of blood product.³¹ Patients who develop PTP most often lack platelet antigen PLA1 or other platelet antigens. For unknown reasons, exposure to the antigens from the transfusion leads to rapid destruction of the patient’s own platelets. The diagnostic clue is thrombocytopenia in a patient, typically female, who has received a red cell or platelet blood product in the past 7 to 10 days. Treatment consists of intravenous immunoglobulin³² and plasmapheresis to remove the offending antibody. If patients with a history of PTP require further transfusions, only PLA1-negative platelets should be given.

IRON OVERLOAD

Every transfusion of red cells delivers approximately 250 mg of iron to the recipient. Since there is no natural way of ridding the body of iron, heavily transfused patients are at risk of iron overload. This is most often seen in children heavily transfused for thalassemia. Starting in the second decade of life, these individuals will develop endocrinopathies due to iron overload, liver problems, and often fatal cardiomyopathies. Studies have shown that chelation of iron with deferoxamine can be effective in preventing this fatal complication.³³ Oral iron chelators such as deferasirox and deferiprone are also effective. The risk of iron overload in heavily transfused patients with myelodysplasia or other transfusion-dependent anemias is unclear, and uncertainty exists about the need for chelation.³⁴

 

Young patients who face years of transfusions should be started on iron chelation to avoid iron overload. For older patients with transfusion-dependent anemia, iron chelation therapy should be considered if their life expectancy is long (years to decades) or special studies such as T2-weighted cardiac magnetic resonance imaging showing iron overloading.³⁵

INFECTIOUS COMPLICATIONS

Concern over transmission of HIV infection via blood products in the late 1980s led to both a reduction in blood product use and a greater awareness of infectious complications of transfusion and their prevention. However, no blood product can ever be assumed to be safe for 2 reasons. One is that blood products can transmit infections during a “window period”—the time before a contaminated product can be detected by testing. The second is that blood is not screened for all potential infections (eg, babesiosis or new infections such as West Nile virus at the start of the outbreak). Risk of infection is reduced in 2 ways: deferral of potential infectious donors and blood product testing.

As part of the donation process, potential blood donors are asked a series of questions to see if they have risk factors for infections (eg, recent travel to malarious areas, recent tattoos), and if they answer positive are deferred from donating blood. Blood products are then tested for infectious agents by a combination of methods including detection of viral antigen, antibody response to infections, and more recently polymerase chain reaction (PCR).³⁶ Current screening includes syphilis testing; testing for antibodies to HIV, HTLV (human T-lymphotropic virus), hepatitis C virus, hepatitis B core antigen (HBcAg), hepatitis B surface antigen, and PCR for HIV, hepatitis B virus, HCV, and West Nile virus. Some centers also test for Trypanosoma cruzi, the cause of Chagas disease.

In the past, the numerically most common transfusion-related disease was hepatitis, first B and then C.^37,38 The first step in eliminating these infections was to stop paying donors for blood products. With the introduction of effective testing for hepatitis B and then C, the incidence of transfusion-related hepatitis has plummeted.³⁶ For example, with the introduction of a diagnostic test for hepatitis C, the estimated risk has fallen from 5% to less than 1 per million. Currently, the risk of transmission of hepatitis B and C, HIV, and HTLV is less than 1 in a million.³⁸

Despite this testing, blood transfusions can transmit a variety of infections, including malaria and babesiosis.³⁹ Any new blood-borne infection introduced into the population can get into the blood supply as well. For example, at the start of the West Nile virus epidemic, there was a cluster of transfusion-transmitted cases that resulted in severe and sometimes fatal illness in immunosuppressed patients, but this issue has been addressed with the development of a PCR assay for screening blood.⁴⁰ The rate of transfusion-related babesiosis has been increasing and screening for the causative parasite is being considered.

 

 

MASSIVE TRANSFUSIONS

Acutely bleeding patients can require large amounts of transfusion products. Early data showed high mortality rates with transfusion of more than 20 units of blood,⁴¹ but with modern blood banking techniques and improved laboratory testing, this rate has decreased dramatically, with survival rates of 43% to 70% in patients transfused with more than 50 units of blood.⁴²

The basic approach to massive transfusions is to first transfuse the patient to maintain hemodynamic stability while specific blood tests are being obtained, and then to use the results of these early tests to guide the rest of the resuscitation. An important component is the ability to rapidly deliver standard packages of red cells, usually 6 to 10 units at a time, to the bleeding patient. To avoid delay while the patient’s blood is being typed, the first products delivered are blood group O Rh-positive units. Given the shortage of Rh-negative blood, this should be reserved for only empiric therapy of women of child-bearing age. Once the blood type is known, the patient can be switched over to type-specific blood.

In the past decade, there has been a shift toward increasing the amount of plasma given to patients receiving massive transfusions. This shift has occurred for 2 reasons. One is that modeling of coagulation changes in massive bleeding suggests the need for larger amounts of plasma to correct defects than have previously been recommended.⁴³ The other reason is based on analysis of resuscitation protocols used in military and civilian trauma centers showing that giving red cells and plasma units in a 1:1 ratio appears to be associated with improved outcomes in massive transfusion. Several studies have extended this concept to platelets, again suggesting improved survival with 1 unit of random donor platelets given 1:1 with red cells and plasma units. The PROPPR (Prospective Observational Multicenter Major Trauma Transfusion) study compared a 1:1:1 to 1:1:2 ratio in patients with severe trauma and major bleeding and found less exsanguination and faster achievement of hemostasis in the first 24 hours.⁴⁴ This has led to the widespread adoption of the 1:1 ratio by most trauma centers, and by default to other massive transfusion situations despite the lack of clinical trial data.⁴⁵

One barrier to increased use is that plasma is kept frozen and requires 20 minutes to thaw. Many institutions are now keeping inventories of thawed plasma available for immediate use, ranging from 2 to 4 units of group AB plasma to keeping their entire inventory as liquid plasma.⁴⁶ Plasma that is thawed but not used can be relabeled as “thawed plasma” and kept for up to 5 days. Also, many centers now use group A plasma for massive transfusions as this rarely leads to transfusion reactions and is much more available.⁴⁷ Research is currently under way on lyophilized plasma, which can be stored at room temperature and can be rapidly reconstituted for emergency use.

The standard approach for laboratory testing is obtaining 5 tests: hematocrit, platelet count, INR/prothrombin time, activated partial thromboplastin time (aPTT), and fibrinogen.⁴⁸ Product selection is guided by these tests, and they are repeated at regular intervals during the massive transfusion. A typical protocol is shown in Table 3. It is important as part of any protocol to have a flow chart that records laboratory results and products given that any member of the team can easily view.

 

The transfusion threshold for a low hematocrit depends on the stability of the patient. If the hematocrit is below 30% and the patient is bleeding or hemodynamically unstable, one should transfuse packed red cells. Stable patients can tolerate lower hematocrits, and an aggressive transfusion policy may even be detrimental.^2,49 If the patient is bleeding, has florid DIC, or has received platelet aggregation inhibitors, then keeping the platelet count above 50 × 10³/µL is reasonable. There are data regarding massive transfusion showing that keeping the platelet count above 50 × 10³/µL resulted in less microvascular bleeding.⁵⁰ The conventional dose of platelets is 6 to 8 platelet concentrates or 1 plateletpheresis unit.

While in the past fibrinogen targets of 50 to 100 mg/dL were recommended, recent data indicate that a target of 150 mg/dL or higher may be more appropriate.^51–53 Severe fibrinolysis may occur in certain clinical situations such as brain injuries, hepatic trauma, or ischemic limb reperfusion, and the use of large amounts of cryoprecipitate can be anticipated. In patients with an INR greater than 2 and an abnormal aPTT, one can give 2 to 4 units of FFP. For an aPTT greater than 1.5 times normal, 2 to 4 units of plasma should be given. Elevation of the aPTT above 1.8 times normal control is associated with microvascular bleeding in trauma patients.⁵⁴ Patients with marked abnormalities (eg, anaPTT more than 2 times normal) may require aggressive therapy with at least 15 to 30 mL/kg (4–8 units for an average adult) of plasma.⁵⁵

Recently there has been increasing interest in the use of thromboelastography (TEG) in massive transfusion.⁵⁶ This is a point-of-care assay performed on fresh whole blood that can assess multiple facets of hemostasis, including coagulation, platelet function, and fibrinolysis.^57,58 TEG is performed by placing a 0.35-mL sample of whole blood into an oscillating container with a sensor pin that measures the force of thrombus formation. TEG measures 5 parameters:

• r time: time from starting TEG until clot formation
• K time: time needed for tracing to go from 2 mm to 20 mm
• alpha angle: slope of tracing between r and K time
• MA: greatest amplitude of TEG tracing
• Whole blood lysis index: amplitude of tracing 60 minutes after MA.

Several centers have incorporated TEG into resuscitation protocols that include standardized strategies for responding to abnormalities. Data suggest that use of TEG may decrease the use of blood products, especially in cardiac surgery, but this has not been prospectively studied in massive transfusions.^56,59

COMPLICATIONS OF MASSIVE TRANSFUSIONS

Electrolyte abnormalities are unusual even in patients who receive massive transfusions.⁶⁰ Platelet concentrates and plasma contain citrate that can chelate calcium. However, the citrate is rapidly metabolized, and it is rare to see clinically significant hypocalcemia. Although empiric calcium replacement is often recommended, one study suggests that this is associated with a worse outcome and should not be done.⁶¹ If hypocalcemia is a clinical concern, then levels should be drawn to guide therapy. Stored blood is acidic, with a pH of 6.5 to 6.9. However, acidosis attributed solely to transfused blood is rare and most often is a reflection of the patient’s stability. Empirical bicarbonate replacement has been associated with severe alkalosis and is not recom mended.^62,63 Although potassium leaks out of stored red cells, even older units of blood contain only 8 mEq/L of potassium, so hyperkalemia is usually not a concern.

 

PATIENTS WITH AUTOIMMUNE HEMOLYTIC ANEMIA

Patients with autoimmune hemolytic anemia can be difficult to transfuse,⁶⁴ because the autoantibody can interfere with several aspects of the transfusion services evaluation. In some patients the autoantibody can be so strong that the patient’s blood type cannot be determined. In most patients, the final step of the cross-match—mixing the donor blood with recipient plasma—will show noncompatibility due to the autoantibodies reacting with any red cells.

The first step when transfusing a patient with autoimmune hemolytic anemia is to draw several tubes of blood for the transfusion service before any potential transfusions. This allows the transfusion service to remove the autoantibodies so they can screen for underlying alloantibodies. Second, if the patient requires immediate transfusion, then type-specific or O-negative blood should be given. If the patient has not been recently (months) transfused, the incidence of a severe transfusion reaction is low. The first unit should be infused slowly with close observation of the patient. For patients who have been multiply transfused, the use of an “in-vivo” cross-match may be helpful. This is where the patient is slowly transfused 10 to 15 mL of blood over 15 minutes. The the plasma and urine are then assessed for signs of hemolysis and, if negative, the remaining product is given.

REFUSAL OF BLOOD PRODUCTS

The initial step in managing patients who refuse blood products is to find out why they are refusing them. Many patients have an exaggerated fear of HIV and other infectious agents, so discussing the very low risk for infection transmission can often resolve the situation. The most common reason for refusal of blood products is religious belief. Jehovah’s Witness patients will refuse blood products due to their interpretation of the Bible.⁶⁵ All members will refuse red cells, plasma, and platelets, while decisions about “derived” blood products—products made by manipulation of the original donated units—are a matter of conscience. These include cryoprecipitate, intravenous gammaglobulin, and albumin.

In an elective situation, the first step is to discuss with the patient those products that are a matter of conscience and clearly document this. The patient’s blood count and iron stores should be assessed to identify any correctible causes of anemia or low iron stores before surgery. The use of erythropoietin to correct blood counts before surgery is controversial, as this may increase thrombosis risk and is contraindicated in patients with curable tumors.

For patients with acute blood loss, use of intravenous iron combined with high-dose erythropoietin is the most common approach to raise the blood count.⁶⁵ A recommended erythropoietin dose is 300 units/kg 3 times a week, dropping to 100 units/kg 3 times weekly until the goal hematocrit is reached. Another often overlooked step is to consolidate and minimize laboratory testing. The most important step is to be respectful of the patient and their beliefs. Many larger cities have liaisons that can help with interactions between Jehovah’s Witness patients and the health care system.

 

 

NON-TRANSFUSION THERAPIES FOR ACUTE BLEEDING

DESMOPRESSIN

Desmopressin (DDAVP) is a synthetic analog of antidiuretic hormone that raises the levels of both factor VIII and von Willebrand protein severalfold.⁶⁶ Desmopressin is effective in supporting hemostasis in patients with a wide variety of congenital and acquired bleeding disorders. However, desmopressin does not reduce blood loss before routine surgery in a healthy patient and should not be used for this purpose.⁶⁷

TRANEXAMIC ACID

Tranexamic acid is an antifibrinolytic agent that blocks the binding of plasmin to fibrin.⁶⁸ This agent was first shown to be useful in disorders that involve excessive fibrinolysis^69–73 or as adjunctive therapy for oral or dental procedures in patients with a bleeding diathesis. In patients with severe thrombocytopenia, the use of antifibrinolytic agents may reduce bleeding. Increasing data shows that tranexamic acid can prevent blood loss in a variety of surgeries including heart bypass, liver transplantation, and orthopedic surgery.⁷⁴ Patients across these settings have decreased blood loss and need for transfusion with no increased risk of thrombosis. The CRASH-2 study showed that the use of tranexamic acid significantly reduced mortality in trauma patients.⁷⁵ The WOMEN trial demonstrated that 1 g of tranexamic acid given to women with blood loss of more than 500 mL after vaginal delivery or 1000 mL after cesarean section has a risk reduction of death of 0.81 with no increased risk of thrombosis.⁷⁶ Given this abundant data, it is clear tranexamic acid needs to be part of any massive transfusion protocol.⁷⁷

RECOMBINANT FACTOR VIIa

Recombinant factor VIIa (rVIIa) was originally developed as a “bypass” agent to support hemostasis in hemophiliacs.⁷⁸ However, the use of rVIIa for a wide array of bleeding disorders, including patients with factor VII and XI deficiency and Glanzmann thrombasthenia, has been reported.⁷⁹ Increasingly, rVIIa is being used as a “universal hemostatic agent” for patients with uncontrolled bleeding from any mechanism.⁸⁰ Multiple case reports have described the use of rVIIa for bleeding in cardiac surgery patients, obstetrical bleeding, reversal of anticoagulation, and trauma.⁸¹ Unfortunately, little formal trial data exists to put these anecdotes into perspective, and formal review of clinical trial results has shown no benefit.^82,83 However, when used in older patients, especially those with vascular risk factors, the risk of arterial thrombosis appears to increase.⁸⁴ In the trials for intracranial hemorrhage, the thrombosis rate was 5% to 9%, and rates up to 10% for arterial events were seen in older patients in a review of all trials.^85–87 Given the lack of data but the evidence of risk, rVIIa use should be restricted to patients with documented bleeding disorders that have been shown to benefit by its use.

INTRODUCTION

Transfusion therapy is an essential part of hematology practice, allowing for curative therapy of diseases such as leukemia, aplastic anemia, and aggressive lymphomas. Nonetheless, transfusions are associated with significant risks, including transmission of infections and transfusion-related reactions. Controversy remains about key issues in transfusion therapy, such as triggers for red cell transfusions. This article reviews the available blood products (Table 1) and indications for transfusion along with the associated risks, and also discusses specific clinical situations, such as massive transfusion.

BLOOD PRODUCTS

WHOLE BLOOD

Whole blood is the product of 1 unit of donated blood plus anticoagulant/preservative, and by definition contains 1 unit of plasma and red cells. Whole blood can be stored for 5 weeks. Although it was the standard product in the past, whole blood is rarely used since 1 unit of donated blood can now be fractionated into 1 unit of red blood cells (RBC), 1 unit of platelets, and 1 unit of fresh frozen plasma (FFP). Thus, the use of whole blood for just a single transfusion represents a waste of resources. There are 2 exceptions. One is autologous blood donations, which are whole blood units. Second, whole blood is increasingly being used in massive transfusions for trauma patients, with the rationale being that all essential blood components are being transfused at once.¹

PACKED RED CELLS

The remaining red cell mass after most of the plasma is removed is called the “packed” red cell unit (hematocrit = 70%–80%), and so red cells are often called “packed” red cells, or PRBC. A preservative is added to improve the flow of blood and to provide “nutrients” for the red cells, and this reduces the hematocrit to approximately 60%. The volume of a red cell unit is approximately 340 mL. In the average adult, 1 unit of RBC raises the hematocrit by 3%. The indications for transfusion of red cells are to increase red cell mass, and thus oxygen delivery, in patients who are compromised by their anemia.

Several randomized trials have helped define the indications for red cell transfusions and justify lower hematocrit thresholds for initiating transfusion. The TRICC (Transfusion Requirements in Critical Care Investigators for the Canadian Critical Care Trials Group) trial showed that in critical care patients (30-day mortality, 18.7%–23.3%), a conservative transfusion strategy of waiting until the hematocrit was below 21% had the same outcomes as transfusing at a threshold of 24%.² The TRACS (Transfusion Requirements After Cardiac Surgery) trial showed that a hematocrit target of 24% had the same benefit as a target of 30% in patients who had undergone cardiac bypass surgery.³ For patients with acute myocardial infarction, the outcomes were worse with aggressive transfusion at a hematocrit of 30% compared to 24%.⁴ In patients with upper gastrointestinal bleeding, a hemoglobin transfusion trigger of 7 g/dL was associated with a lower mortality than a trigger of 9 g/dL (5% versus 9%).⁵ Finally, the FOCUS (Transfusion Trigger Trial for Functional Outcomes in Cardiovascular Patients Undergoing Surgical Hip Fracture Repair) trial showed that in older patients (average age 80 years) who had undergone hip fracture surgery, transfusions based on symptoms and not a fixed trigger of 30% had the same outcomes but considerable savings in blood products.⁶ Based on these trials, decisions regarding when to transfuse patients should be based on symptoms and not “numbers.” Young patients, especially those with reversible anemias, can tolerate low blood counts and should not be transfused based on an arbitrary number.

PLATELETS

Several types of platelet products exist. One unit of platelet concentrate is derived from 1 unit of donor blood. Plateletpheresis from volunteer donors is also used to harvest platelets, with the resulting product referred to as plateletpheresis platelets. One unit of single-donor (pheresis) platelets is equivalent to 6 platelet concentrates. Finally, HLA-matched platelets are single-donor pheresis units that are obtained from an HLA-matched donor. This product should be ordered only if there is evidence of HLA antibodies (see Platelet Alloimmunization section).

The dose of platelets for the average patient is 6 units of platelet concentrate or 1 pheresis unit. In theory, 1 unit of platelet concentrate can raise the count by 5 to 7 × 10³/µL, but often this response is blunted by concurrent illness or bleeding. In patients who appear to have a poor response, the platelet count can be checked 15 minutes after platelet infusion. No rise or a minimal rise (< 2 × 10³/µL) in the platelet count is suggestive of platelet refractoriness, while a good 15-minute response but poor 24-hour count is more suggestive of consumption—fever, sepsis, drug, or splenomegaly—and not refractoriness.

The indication for platelet transfusion depends on the clinical situation. For patients with immune thrombocytopenia, platelets should not be transfused unless there is life-threatening bleeding. For stable patients with marrow aplasia from chemotherapy, a cut-off of a morning platelet count of less than 10 × 10³/µL has been shown to be as safe as higher levels for prophylactic transfusions.⁷ For patients with active bleeding, the platelet count should be kept above 50 × 10³/µL. Patients with acquired or inherited platelet dysfunction may benefit from transfusion no matter the platelet count.

 

 

Platelet Alloimmunization

Patients exposed to transfused white cells with different HLA antigens can develop antibodies to these antigens.⁸ Anti-HLA antibodies are common in patients who previously have received transfused blood that is not leukodepleted and in patients who have been pregnant. Since platelets carry class I HLA antigens, they will be rapidly destroyed by anti-HLA antibodies when transfused into these patients. In patients transfused for aplastic anemia or myelodysplasia, as many as 90% will become HLA-immunized. The incidence is lower in patients receiving chemotherapy but still can be as high as 60% to 90%.^9,10 Patients who have developed anti-HLA antibodies can respond to transfused platelets matched for HLA antigens. Unfortunately, some patients will either be a rare HLA type or be so heavily immunized that they will not respond to any platelet transfusion.

The significance of alloimmunization centers on 2 concepts: recognition and avoidance. Patients with HLA antibodies will fail to have an increment of their platelet counts with transfusions. Accordingly, patients who do not experience an increase in their count 15 minutes after the transfusion may have HLA antibodies. One can test for the presence of anti-HLA antibodies, although some patients instead have specific antiplatelet antibodies that will not respond to HLA-matched platelets. In patients who have been pregnant or previously transfused and are scheduled to undergo transplant or aggressive chemotherapy, it is wise to test for anti-HLA antibodies in order to plan their transfusion needs. The evidence suggests that transfused white cells are responsible for initiating the anti-HLA response. Trials have shown that giving leukodepleted blood products may reduce the incidence of alloimmunization, so patients who are not HLA-alloimmunized should receive only leukodepleted products.¹¹A difficult problem is bleeding in patients who are refractory to platelet transfusion.¹² Patients who test positive for the presence of anti-HLA antibodies can receive transfusions of HLA-matched platelets.¹³ Unfortunately, matched platelet transfusions are not effective in 20% to 70% of these patients. Also, since some loci are difficult to match, effective products may be unavailable. Finally, as many as 25% of patients have antiplatelet antibodies in which HLA-matched products will be ineffective. Platelet cross-matching can be performed to find compatible units for these patients, but this may not always be successful. In the patient who is totally refractory to platelet transfusion, consider drugs as an etiology of antiplatelet antibodies (especially vancomycin).¹⁴ Use of antifibrinolytic agents such as epsilon-aminocaproic acid or tranexamic acid may decrease the incidence of minor bleeding, but these are ineffective for major bleeding. “Platelet drips”—infusing either a platelet concentrate per hour or 1 plateletpheresis unit every 6 hours—may be given as a continuous infusion, but there is no evidence that this is helpful.¹⁵

FRESH FROZEN PLASMA

FFP is made from 1 unit of donated whole blood, with an average volume of 225 mL per unit. One unit of FFP can increase coagulation factor levels by 5% and fibrinogen by 10 mg/dL in the average stable patient. FFP can take 20 to 30 minutes to thaw before use, so in situations where FFP is needed quickly, the blood bank must be informed to “keep ahead” some units. Units of FFP that have been thawed but not used can be stored refrigerated for 5 days to prevent wasting blood products.

The indications for FFP are limited to several situations. These include a documented coagulation defect that can be corrected by a reasonable amount of FFP, such as factor V deficiency and factor XI deficiency, disseminated intravascular coagulation (DIC), reversal of warfarin, and massive transfusions. FFP is also used for the therapy of thrombotic thrombocytopenic purpura.

There is little justification for FFP transfusion in many of the clinical settings in which it is commonly used. For example, FFP is given for minor elevations of the INR in patients with liver disease, despite literature showing not only that the INR rise is not reflective of coagulation defects, but also that patients with liver disease may even be thrombophilic.^16,17 Reviews of FFP use found limited evidence-based indications for its use.^18,19 Also, several studies have shown that transfusion of FFP is not effective at reversing minor elevations of the INR (1.3–1.8).²⁰ In a meta-analysis, FFP was associated with increased risk for lung injury and a trend toward increased mortality.¹⁸

 

 

CRYOPRECIPITATE

Cryoprecipitate is produced from 1 unit of FFP that is thawed at 4°C. The precipitate is resuspended with 10 mL of saline or FFP and refrozen for storage. One unit contains at least 150 mg of fibrinogen and 80 units of factor VIII, along with von Willebrand factor. Thawing time for cryoprecipitate is approximately 20 minutes.

Cryoprecipitate is used to raise the fibrinogen level in patients with DIC or massive transfusion with hemodilution. It is third-line therapy in the treatment of type 1 von Willebrand disease and is second-line therapy in the treatment of patients with other types of von Willebrand disease. Currently, von Willebrand factor concentrates are the preferred replacement product for von Willebrand disease. Cryoprecipitate can be used as a source of factor VIII for hemophiliacs, but the preferred product for these patients is the super pure factor VIII concentrates or recombinant products. Cryoprecipitate can also be used to shorten the bleeding time of uremic patients, but its effectiveness for this is controversial.

GRANULOCYTES

Granulocytes are harvested by leukopheresis of normal donors, with a target yield of 1010 granulocytes from each donor. To reach this target, the donors are often “stimulated” with neutrophil growth factors. The harvesting procedure can take 3 hours and is associated with some risks to the donor (eg, citrate toxicity). The current indications for granulocytes are very limited since the advent of neutrophil growth factors and improved antimicrobials.²¹ They can be useful in the neutropenic patient with a documented bacterial infection in whom the white blood cell count is not expected to recover in the near future. Given the difficulty of keeping the count up, these transfusions have been mainly used in treating small children.

SPECIAL BLOOD PRODUCTS

IRRADIATED BLOOD PRODUCTS

Irradiation of blood is performed for only one reason: to prevent transfusion-related graft-versus-host disease (TGVHD) (Table 2).²² The irradiation can be performed at the blood center or in the transfusion service of larger hospitals. The units are not radioactive and can be transfused safely to other patients. There is increased leakage of potassium in irradiated units of blood, so the units need to be transfused within 14 days; in patients potentially sensitive to potassium (eg, neonates), the units must be transfused within 24 hours. Patients undergoing stem cell transplant, those receiving either interuterine transfusions or products from relatives, any patient with Hodgkin disease or receiving purine analogs or alemtuzumab, and patients with severe congenital immune deficiencies should receive irradiated blood. Most would also advocate that patients with hematologic malignances receiving chemotherapy receive irradiated products, but this is more controversial.

LEUKPDEPLETED BLOOD

Contamination of blood products by white blood cells is increasingly being recognized as a possible cause of adverse effects in transfused patients, including febrile transfusion reactions, inducing HLA alloimmunization, immunosuppression, disease transmission, and TGVHD. Reducing white cells can reduce the incidence of all of these complications except TGVHD. Currently, white cells are removed by infusion through filters that trap the cells. This can be done either at the bedside, in the blood bank, or at the donor center. The majority of red cells provided by blood centers in many areas of the country are already leukoreduced, eliminating the need for labor-intensive filtration at the transfusion center or bedside. Platelets collected by plateletpheresis methods can also be made leukocyte-poor. The current indications for leukodepleted productions are:

1. Prevention of febrile transfusion reactions in patients with previous documented reactions
2. Prevention of HLA alloimmunization (ineffective if patient has received 1 or more blood products not leukodepleted or is already HLA immunized)
3. Prevention of cytomegalovirus (CMV) infection

 

 

CMV-NEGATIVE BLOOD

CMV can be transmitted through any cellular blood product—red cells and platelets. For patients who are CMV-negative and receiving transplants, especially stem cell transplants, a new CMV infection can be devastating.²¹ For years only blood from CMV-negative donors was used to transfuse CMV-negative patients. This policy is effective in preventing CMV infection, but because 50% of the population is positive for CMV antibodies, it may potentially lead to shortages of products that could be transfused to the patient. Currently, leukoreduced blood products are used since leukofiltration of the blood is just as effective as transfusion of CMV-negative blood in preventing infections and allows greater use of all blood products.²³

COMPLICATIONS OF TRANSFUSIONS

HEMOLYTIC TRANSFUSION REACTION

There are 2 forms of hemolytic reactions—immediate and delayed.²⁴ The immediate reaction is associated with fevers, hypotension, back pain, and oliguria. In severe cases, DIC and renal failure may occur. The immediate reaction is due to transfusion of blood that reacts with the recipient’s preformed high-titer blood antigen antibodies, most often to ABO. This is fatal 2% of the time and occurs almost always as a result of errors in correct identification of the patient. Reactions are due to recipient antibodies attacking donated RBCs, resulting in release of hemoglobin and red cell membrane–antigen complexes. These complexes are believed to lead to the hypotension, fevers, chills, and renal damage associated with the hemolytic reaction. Treatment consists of immediately stopping the transfusion, notifying the blood bank, vigorous intravenous hydration to keep the urine output over 100 mL/hr, and supportive therapy.

The delayed reaction can range in severity from an abrupt drop in the hematocrit to normal response to transfusion but the patient developing a positive Coombs’ test. The delayed response is due to an anamnestic response to blood-group antigens. When the patient is exposed to the same antigen, there is a rise in antibody titer leading to the reaction. Some alloantibodies can lead to a brisk reaction, most often anti-Kidd. The frequency with which delayed transfusion reactions occur is underestimated because mild reactions often do not get worked up or even discovered.

ALLERGIC REACTIONS

Allergic reactions are common (1%–3% of transfusions) and occur in patients having antibodies to proteins in donor blood, which can lead to hives and itching with transfusions. Most of the time these allergic reactions are mild and can be treated with antihistamines. Prophylaxis with antihistamines is not indicated for future transfusions unless the reactions are frequent. Rarely these reactions can be associated with shock and hypotension. Patients who are immunoglobulin (Ig) A–deficient can develop anaphylactic reactions to IgA-containing blood products. Patients with severe allergic reactions need to have their IgA measured and, if deficient, receive only washed units or plasma from IgA-deficient donors to prevent future severe reactions.

FEBRILE REACTIONS

The most common transfusion reaction is a febrile reaction that occurs after the transfusion starts and that sometimes can be complicated by chills. This reaction often occurs due to the presence of leukocyte debris and cytokines in the donated blood. Therapy is supportive and involves stopping the transfusion and administering acetaminophen, but since hemolytic transfusion reactions can present with fever all patients need to be thoroughly evaluated. The incidence of reactions can be decreased by using leukodepleted blood and plateletpheresis platelets. Most patients do not benefit from receiving prophylactic acetaminophen for future transfusion unless they have multiple reactions.

 

 

TRANSFUSION-RELATED ACUTE LUNG INJURY

Once thought a rare complication, transfusion-related acute lung injury (TRALI) is increasingly being recognized, with an incidence of approximately 1:5000 patients; it is now the most frequent cause of transfusion-related death.^24,25 TRALI is noncardiac pulmonary edema and typically manifests clinically with hypoxemia, fever, bilateral infiltrates, and hypotension 2 to 6 hours after blood is given. Ventilatory support is often required. Recovery is usually rapid (24–48 hours) and complete. The etiology is complex. In many cases, transfused anti-HLA antibodies react with the recipient’s white cells leading to pulmonary damage. Another theory is that transfusion of preformed cytokines leads to pulmonary damage. Because plasma products from multiparous women are most often associated with anti-HLA antibodies, the restricted use of blood products from women has decreased the incidence of TRALI over the past few years.²⁶

TRANSFUSION-ASSOCIATED CIRCULATORY OVERLOAD

Increasingly it being recognized that volume overload resulting from transfusions can lead to significant morbidity.²⁷ Patients with heart or renal disease or patients who already have compromised fluid status are at risk for transfusion-associated circulatory overload (TACO). Another risk factor is transfusion of multiple blood products. Patients with TACO develop dyspnea within 6 hours of transfusion, but do not have fever or rash with the dyspnea. The diagnosis is made by demonstrating circulatory overload (eg, high venous pressure, B-type natriuretic peptide). Treatment is aggressive diuresis. Strategies to prevent TACO include judicious use of blood products, especially in patients at risk for TACO, and the use of prophylactic diuretics, especially with red cell or plasma transfusions.²⁸

TRANSFUSION-RELATED GRAFT-VERSUS-HOST DISEASE

TGVHD is a rare reaction, but one that is most often fatal.²⁹ TGVHD occurs when donor lymphocytes attack the blood recipient’s organs—skin, liver, intestines, and marrow. This is very rare in the normal blood recipient unless the donor and recipient share some HLA haplotypes.³⁰ In immunosuppressed patients, TGVHD can occur with lesser degrees of HLA similarity, with cases reported in blood recipients who are mainly patients with Hodgkin disease or acute leukemia undergoing chemotherapy, and in patients receiving purine analogs. TGVHD had not been reported in AIDS patients despite profound immunosuppression, perhaps because the milieu of the patient does not allow lymphocyte expansion. Symptoms of TGVHD are an erythematous rash that may progress to epidermal toxic necrolysis, liver dysfunction, diarrhea, and pancytopenia. TGVHD is prevented by irradiating blood products given to at-risk patients with 2500 to 3500 rads. Directed blood donation from all blood relatives should also be irradiated. TGVHD cannot be prevented by leukopoor blood because the minute amount of lymphocytes that are not filtered still can lead to these complications.

POST-TRANSFUSION PURPURA

Patients with post-transfusion purpura (PTP) develop severe thrombocytopenia (< 10 × 10³/µL) with often severe bleeding 1 to 2 weeks after receiving any type of blood product.³¹ Patients who develop PTP most often lack platelet antigen PLA1 or other platelet antigens. For unknown reasons, exposure to the antigens from the transfusion leads to rapid destruction of the patient’s own platelets. The diagnostic clue is thrombocytopenia in a patient, typically female, who has received a red cell or platelet blood product in the past 7 to 10 days. Treatment consists of intravenous immunoglobulin³² and plasmapheresis to remove the offending antibody. If patients with a history of PTP require further transfusions, only PLA1-negative platelets should be given.

IRON OVERLOAD

Every transfusion of red cells delivers approximately 250 mg of iron to the recipient. Since there is no natural way of ridding the body of iron, heavily transfused patients are at risk of iron overload. This is most often seen in children heavily transfused for thalassemia. Starting in the second decade of life, these individuals will develop endocrinopathies due to iron overload, liver problems, and often fatal cardiomyopathies. Studies have shown that chelation of iron with deferoxamine can be effective in preventing this fatal complication.³³ Oral iron chelators such as deferasirox and deferiprone are also effective. The risk of iron overload in heavily transfused patients with myelodysplasia or other transfusion-dependent anemias is unclear, and uncertainty exists about the need for chelation.³⁴

 

Young patients who face years of transfusions should be started on iron chelation to avoid iron overload. For older patients with transfusion-dependent anemia, iron chelation therapy should be considered if their life expectancy is long (years to decades) or special studies such as T2-weighted cardiac magnetic resonance imaging showing iron overloading.³⁵

INFECTIOUS COMPLICATIONS

Concern over transmission of HIV infection via blood products in the late 1980s led to both a reduction in blood product use and a greater awareness of infectious complications of transfusion and their prevention. However, no blood product can ever be assumed to be safe for 2 reasons. One is that blood products can transmit infections during a “window period”—the time before a contaminated product can be detected by testing. The second is that blood is not screened for all potential infections (eg, babesiosis or new infections such as West Nile virus at the start of the outbreak). Risk of infection is reduced in 2 ways: deferral of potential infectious donors and blood product testing.

As part of the donation process, potential blood donors are asked a series of questions to see if they have risk factors for infections (eg, recent travel to malarious areas, recent tattoos), and if they answer positive are deferred from donating blood. Blood products are then tested for infectious agents by a combination of methods including detection of viral antigen, antibody response to infections, and more recently polymerase chain reaction (PCR).³⁶ Current screening includes syphilis testing; testing for antibodies to HIV, HTLV (human T-lymphotropic virus), hepatitis C virus, hepatitis B core antigen (HBcAg), hepatitis B surface antigen, and PCR for HIV, hepatitis B virus, HCV, and West Nile virus. Some centers also test for Trypanosoma cruzi, the cause of Chagas disease.

In the past, the numerically most common transfusion-related disease was hepatitis, first B and then C.^37,38 The first step in eliminating these infections was to stop paying donors for blood products. With the introduction of effective testing for hepatitis B and then C, the incidence of transfusion-related hepatitis has plummeted.³⁶ For example, with the introduction of a diagnostic test for hepatitis C, the estimated risk has fallen from 5% to less than 1 per million. Currently, the risk of transmission of hepatitis B and C, HIV, and HTLV is less than 1 in a million.³⁸

Despite this testing, blood transfusions can transmit a variety of infections, including malaria and babesiosis.³⁹ Any new blood-borne infection introduced into the population can get into the blood supply as well. For example, at the start of the West Nile virus epidemic, there was a cluster of transfusion-transmitted cases that resulted in severe and sometimes fatal illness in immunosuppressed patients, but this issue has been addressed with the development of a PCR assay for screening blood.⁴⁰ The rate of transfusion-related babesiosis has been increasing and screening for the causative parasite is being considered.

 

 

MASSIVE TRANSFUSIONS

Acutely bleeding patients can require large amounts of transfusion products. Early data showed high mortality rates with transfusion of more than 20 units of blood,⁴¹ but with modern blood banking techniques and improved laboratory testing, this rate has decreased dramatically, with survival rates of 43% to 70% in patients transfused with more than 50 units of blood.⁴²

The basic approach to massive transfusions is to first transfuse the patient to maintain hemodynamic stability while specific blood tests are being obtained, and then to use the results of these early tests to guide the rest of the resuscitation. An important component is the ability to rapidly deliver standard packages of red cells, usually 6 to 10 units at a time, to the bleeding patient. To avoid delay while the patient’s blood is being typed, the first products delivered are blood group O Rh-positive units. Given the shortage of Rh-negative blood, this should be reserved for only empiric therapy of women of child-bearing age. Once the blood type is known, the patient can be switched over to type-specific blood.

In the past decade, there has been a shift toward increasing the amount of plasma given to patients receiving massive transfusions. This shift has occurred for 2 reasons. One is that modeling of coagulation changes in massive bleeding suggests the need for larger amounts of plasma to correct defects than have previously been recommended.⁴³ The other reason is based on analysis of resuscitation protocols used in military and civilian trauma centers showing that giving red cells and plasma units in a 1:1 ratio appears to be associated with improved outcomes in massive transfusion. Several studies have extended this concept to platelets, again suggesting improved survival with 1 unit of random donor platelets given 1:1 with red cells and plasma units. The PROPPR (Prospective Observational Multicenter Major Trauma Transfusion) study compared a 1:1:1 to 1:1:2 ratio in patients with severe trauma and major bleeding and found less exsanguination and faster achievement of hemostasis in the first 24 hours.⁴⁴ This has led to the widespread adoption of the 1:1 ratio by most trauma centers, and by default to other massive transfusion situations despite the lack of clinical trial data.⁴⁵

One barrier to increased use is that plasma is kept frozen and requires 20 minutes to thaw. Many institutions are now keeping inventories of thawed plasma available for immediate use, ranging from 2 to 4 units of group AB plasma to keeping their entire inventory as liquid plasma.⁴⁶ Plasma that is thawed but not used can be relabeled as “thawed plasma” and kept for up to 5 days. Also, many centers now use group A plasma for massive transfusions as this rarely leads to transfusion reactions and is much more available.⁴⁷ Research is currently under way on lyophilized plasma, which can be stored at room temperature and can be rapidly reconstituted for emergency use.

The standard approach for laboratory testing is obtaining 5 tests: hematocrit, platelet count, INR/prothrombin time, activated partial thromboplastin time (aPTT), and fibrinogen.⁴⁸ Product selection is guided by these tests, and they are repeated at regular intervals during the massive transfusion. A typical protocol is shown in Table 3. It is important as part of any protocol to have a flow chart that records laboratory results and products given that any member of the team can easily view.

 

The transfusion threshold for a low hematocrit depends on the stability of the patient. If the hematocrit is below 30% and the patient is bleeding or hemodynamically unstable, one should transfuse packed red cells. Stable patients can tolerate lower hematocrits, and an aggressive transfusion policy may even be detrimental.^2,49 If the patient is bleeding, has florid DIC, or has received platelet aggregation inhibitors, then keeping the platelet count above 50 × 10³/µL is reasonable. There are data regarding massive transfusion showing that keeping the platelet count above 50 × 10³/µL resulted in less microvascular bleeding.⁵⁰ The conventional dose of platelets is 6 to 8 platelet concentrates or 1 plateletpheresis unit.

While in the past fibrinogen targets of 50 to 100 mg/dL were recommended, recent data indicate that a target of 150 mg/dL or higher may be more appropriate.^51–53 Severe fibrinolysis may occur in certain clinical situations such as brain injuries, hepatic trauma, or ischemic limb reperfusion, and the use of large amounts of cryoprecipitate can be anticipated. In patients with an INR greater than 2 and an abnormal aPTT, one can give 2 to 4 units of FFP. For an aPTT greater than 1.5 times normal, 2 to 4 units of plasma should be given. Elevation of the aPTT above 1.8 times normal control is associated with microvascular bleeding in trauma patients.⁵⁴ Patients with marked abnormalities (eg, anaPTT more than 2 times normal) may require aggressive therapy with at least 15 to 30 mL/kg (4–8 units for an average adult) of plasma.⁵⁵

Recently there has been increasing interest in the use of thromboelastography (TEG) in massive transfusion.⁵⁶ This is a point-of-care assay performed on fresh whole blood that can assess multiple facets of hemostasis, including coagulation, platelet function, and fibrinolysis.^57,58 TEG is performed by placing a 0.35-mL sample of whole blood into an oscillating container with a sensor pin that measures the force of thrombus formation. TEG measures 5 parameters:

• r time: time from starting TEG until clot formation
• K time: time needed for tracing to go from 2 mm to 20 mm
• alpha angle: slope of tracing between r and K time
• MA: greatest amplitude of TEG tracing
• Whole blood lysis index: amplitude of tracing 60 minutes after MA.

Several centers have incorporated TEG into resuscitation protocols that include standardized strategies for responding to abnormalities. Data suggest that use of TEG may decrease the use of blood products, especially in cardiac surgery, but this has not been prospectively studied in massive transfusions.^56,59

COMPLICATIONS OF MASSIVE TRANSFUSIONS

Electrolyte abnormalities are unusual even in patients who receive massive transfusions.⁶⁰ Platelet concentrates and plasma contain citrate that can chelate calcium. However, the citrate is rapidly metabolized, and it is rare to see clinically significant hypocalcemia. Although empiric calcium replacement is often recommended, one study suggests that this is associated with a worse outcome and should not be done.⁶¹ If hypocalcemia is a clinical concern, then levels should be drawn to guide therapy. Stored blood is acidic, with a pH of 6.5 to 6.9. However, acidosis attributed solely to transfused blood is rare and most often is a reflection of the patient’s stability. Empirical bicarbonate replacement has been associated with severe alkalosis and is not recom mended.^62,63 Although potassium leaks out of stored red cells, even older units of blood contain only 8 mEq/L of potassium, so hyperkalemia is usually not a concern.

 

PATIENTS WITH AUTOIMMUNE HEMOLYTIC ANEMIA

Patients with autoimmune hemolytic anemia can be difficult to transfuse,⁶⁴ because the autoantibody can interfere with several aspects of the transfusion services evaluation. In some patients the autoantibody can be so strong that the patient’s blood type cannot be determined. In most patients, the final step of the cross-match—mixing the donor blood with recipient plasma—will show noncompatibility due to the autoantibodies reacting with any red cells.

The first step when transfusing a patient with autoimmune hemolytic anemia is to draw several tubes of blood for the transfusion service before any potential transfusions. This allows the transfusion service to remove the autoantibodies so they can screen for underlying alloantibodies. Second, if the patient requires immediate transfusion, then type-specific or O-negative blood should be given. If the patient has not been recently (months) transfused, the incidence of a severe transfusion reaction is low. The first unit should be infused slowly with close observation of the patient. For patients who have been multiply transfused, the use of an “in-vivo” cross-match may be helpful. This is where the patient is slowly transfused 10 to 15 mL of blood over 15 minutes. The the plasma and urine are then assessed for signs of hemolysis and, if negative, the remaining product is given.

REFUSAL OF BLOOD PRODUCTS

The initial step in managing patients who refuse blood products is to find out why they are refusing them. Many patients have an exaggerated fear of HIV and other infectious agents, so discussing the very low risk for infection transmission can often resolve the situation. The most common reason for refusal of blood products is religious belief. Jehovah’s Witness patients will refuse blood products due to their interpretation of the Bible.⁶⁵ All members will refuse red cells, plasma, and platelets, while decisions about “derived” blood products—products made by manipulation of the original donated units—are a matter of conscience. These include cryoprecipitate, intravenous gammaglobulin, and albumin.

In an elective situation, the first step is to discuss with the patient those products that are a matter of conscience and clearly document this. The patient’s blood count and iron stores should be assessed to identify any correctible causes of anemia or low iron stores before surgery. The use of erythropoietin to correct blood counts before surgery is controversial, as this may increase thrombosis risk and is contraindicated in patients with curable tumors.

For patients with acute blood loss, use of intravenous iron combined with high-dose erythropoietin is the most common approach to raise the blood count.⁶⁵ A recommended erythropoietin dose is 300 units/kg 3 times a week, dropping to 100 units/kg 3 times weekly until the goal hematocrit is reached. Another often overlooked step is to consolidate and minimize laboratory testing. The most important step is to be respectful of the patient and their beliefs. Many larger cities have liaisons that can help with interactions between Jehovah’s Witness patients and the health care system.

 

 

NON-TRANSFUSION THERAPIES FOR ACUTE BLEEDING

DESMOPRESSIN

Desmopressin (DDAVP) is a synthetic analog of antidiuretic hormone that raises the levels of both factor VIII and von Willebrand protein severalfold.⁶⁶ Desmopressin is effective in supporting hemostasis in patients with a wide variety of congenital and acquired bleeding disorders. However, desmopressin does not reduce blood loss before routine surgery in a healthy patient and should not be used for this purpose.⁶⁷

TRANEXAMIC ACID

Tranexamic acid is an antifibrinolytic agent that blocks the binding of plasmin to fibrin.⁶⁸ This agent was first shown to be useful in disorders that involve excessive fibrinolysis^69–73 or as adjunctive therapy for oral or dental procedures in patients with a bleeding diathesis. In patients with severe thrombocytopenia, the use of antifibrinolytic agents may reduce bleeding. Increasing data shows that tranexamic acid can prevent blood loss in a variety of surgeries including heart bypass, liver transplantation, and orthopedic surgery.⁷⁴ Patients across these settings have decreased blood loss and need for transfusion with no increased risk of thrombosis. The CRASH-2 study showed that the use of tranexamic acid significantly reduced mortality in trauma patients.⁷⁵ The WOMEN trial demonstrated that 1 g of tranexamic acid given to women with blood loss of more than 500 mL after vaginal delivery or 1000 mL after cesarean section has a risk reduction of death of 0.81 with no increased risk of thrombosis.⁷⁶ Given this abundant data, it is clear tranexamic acid needs to be part of any massive transfusion protocol.⁷⁷

RECOMBINANT FACTOR VIIa

Recombinant factor VIIa (rVIIa) was originally developed as a “bypass” agent to support hemostasis in hemophiliacs.⁷⁸ However, the use of rVIIa for a wide array of bleeding disorders, including patients with factor VII and XI deficiency and Glanzmann thrombasthenia, has been reported.⁷⁹ Increasingly, rVIIa is being used as a “universal hemostatic agent” for patients with uncontrolled bleeding from any mechanism.⁸⁰ Multiple case reports have described the use of rVIIa for bleeding in cardiac surgery patients, obstetrical bleeding, reversal of anticoagulation, and trauma.⁸¹ Unfortunately, little formal trial data exists to put these anecdotes into perspective, and formal review of clinical trial results has shown no benefit.^82,83 However, when used in older patients, especially those with vascular risk factors, the risk of arterial thrombosis appears to increase.⁸⁴ In the trials for intracranial hemorrhage, the thrombosis rate was 5% to 9%, and rates up to 10% for arterial events were seen in older patients in a review of all trials.^85–87 Given the lack of data but the evidence of risk, rVIIa use should be restricted to patients with documented bleeding disorders that have been shown to benefit by its use.

Twenty minutes after Ronald Reagan finished his inaugural address, the Islamic Republic of Iran announced the release of 52 American hostages that Iran had held for the last year of Jimmy Carter’s presidency. The timing led to the “October Surprise conspiracy theory,” in which some inferred a deliberate plan to influence an American election. We now refer to any political event timed to an election as an “October Surprise.” We are awaiting some type of surprise prior to this November’s elections. Events unfolding this fall will have generational effects on American politics, our health care delivery models, our financial security, individual rights and the democratic infrastructure of our country. This is not an election to sit out.

The proposed rule the Centers for Medicare & Medicaid published in the summer has generated a massive public response. The major issues (as discussed in last month’s GI & Hepatology News issue, including the editorial) include a dramatic change in documentation requirements and payment for evaluation and management (E/M) codes for both new and returning patients. While the reduction in documentation is laudable, the reduction in reimbursement for complex visits is not. At Michigan Medicine (2.2 million outpatient visits per year), reimbursements would go down by $3.5 million annually for our E/M visits. Most responses to the proposed rule requested a year’s delay and intensive analysis of work involved prior to reducing payments (see further comments at gastro.org, the AGA website).

This month we cover a new anti-obesity drug that shows cardiovascular safety. This is a welcome potential addition to our therapies since another story updates us on the relentless rise in obesity in this country. We cover the world’s alcohol use this month. On a financial note, the anticipated savings from biosimilars may be less than we hoped if data from Medicare Part D can be generalized. We also cover a diagnostic update about eosinophilic esophagitis.

I hope you enjoy this issue of GI & Hepatology News and all the AGA publications that provide you with up-to-date science, clinical information, and news about gastroenterology in general. Remember to vote. On the wall across from my desk as I sit as a leader in ambulatory care at Michigan Medicine, there is a large sign that grounds me. It reads, “Does this help the patient?”

John I. Allen, MD, MBA, AGAF
Editor in Chief

Twenty minutes after Ronald Reagan finished his inaugural address, the Islamic Republic of Iran announced the release of 52 American hostages that Iran had held for the last year of Jimmy Carter’s presidency. The timing led to the “October Surprise conspiracy theory,” in which some inferred a deliberate plan to influence an American election. We now refer to any political event timed to an election as an “October Surprise.” We are awaiting some type of surprise prior to this November’s elections. Events unfolding this fall will have generational effects on American politics, our health care delivery models, our financial security, individual rights and the democratic infrastructure of our country. This is not an election to sit out.

The proposed rule the Centers for Medicare & Medicaid published in the summer has generated a massive public response. The major issues (as discussed in last month’s GI & Hepatology News issue, including the editorial) include a dramatic change in documentation requirements and payment for evaluation and management (E/M) codes for both new and returning patients. While the reduction in documentation is laudable, the reduction in reimbursement for complex visits is not. At Michigan Medicine (2.2 million outpatient visits per year), reimbursements would go down by $3.5 million annually for our E/M visits. Most responses to the proposed rule requested a year’s delay and intensive analysis of work involved prior to reducing payments (see further comments at gastro.org, the AGA website).

This month we cover a new anti-obesity drug that shows cardiovascular safety. This is a welcome potential addition to our therapies since another story updates us on the relentless rise in obesity in this country. We cover the world’s alcohol use this month. On a financial note, the anticipated savings from biosimilars may be less than we hoped if data from Medicare Part D can be generalized. We also cover a diagnostic update about eosinophilic esophagitis.

I hope you enjoy this issue of GI & Hepatology News and all the AGA publications that provide you with up-to-date science, clinical information, and news about gastroenterology in general. Remember to vote. On the wall across from my desk as I sit as a leader in ambulatory care at Michigan Medicine, there is a large sign that grounds me. It reads, “Does this help the patient?”

John I. Allen, MD, MBA, AGAF
Editor in Chief

Twenty minutes after Ronald Reagan finished his inaugural address, the Islamic Republic of Iran announced the release of 52 American hostages that Iran had held for the last year of Jimmy Carter’s presidency. The timing led to the “October Surprise conspiracy theory,” in which some inferred a deliberate plan to influence an American election. We now refer to any political event timed to an election as an “October Surprise.” We are awaiting some type of surprise prior to this November’s elections. Events unfolding this fall will have generational effects on American politics, our health care delivery models, our financial security, individual rights and the democratic infrastructure of our country. This is not an election to sit out.

The proposed rule the Centers for Medicare & Medicaid published in the summer has generated a massive public response. The major issues (as discussed in last month’s GI & Hepatology News issue, including the editorial) include a dramatic change in documentation requirements and payment for evaluation and management (E/M) codes for both new and returning patients. While the reduction in documentation is laudable, the reduction in reimbursement for complex visits is not. At Michigan Medicine (2.2 million outpatient visits per year), reimbursements would go down by $3.5 million annually for our E/M visits. Most responses to the proposed rule requested a year’s delay and intensive analysis of work involved prior to reducing payments (see further comments at gastro.org, the AGA website).

This month we cover a new anti-obesity drug that shows cardiovascular safety. This is a welcome potential addition to our therapies since another story updates us on the relentless rise in obesity in this country. We cover the world’s alcohol use this month. On a financial note, the anticipated savings from biosimilars may be less than we hoped if data from Medicare Part D can be generalized. We also cover a diagnostic update about eosinophilic esophagitis.

I hope you enjoy this issue of GI & Hepatology News and all the AGA publications that provide you with up-to-date science, clinical information, and news about gastroenterology in general. Remember to vote. On the wall across from my desk as I sit as a leader in ambulatory care at Michigan Medicine, there is a large sign that grounds me. It reads, “Does this help the patient?”

John I. Allen, MD, MBA, AGAF
Editor in Chief