Obstetrics

There are approximately 4,000,000 births annually in the United States, and about 32% of them occur by cesarean delivery. Compared with vaginal birth, cesarean delivery is associated with an increased risk of endometritis (defined as fever plus uterine or abdominal tenderness). Although surgical complications cannot be eliminated entirely, surgeons are deeply dedicated to the continuous improvement of surgical practice in order to reduce the risk of complications.

With cesarean delivery, many surgical practices have been adopted universally to reduce postoperative complications, including administration of intravenous (IV) antibiotics before skin incision to minimize postoperative infection and the use of postoperative mechanical or pharmacologic interventions to help prevent venous thromboembolism and pulmonary embolism. Preoperative vaginal cleansing with povidone-iodine may reduce the risk of postoperative endometritis, but the practice is not currently common in the United States.

Should you adopt a policy of preoperative vaginal cleansing prior to cesarean delivery? The data suggest perhaps you should.

Data-driven support for povidone-iodine precesareanThree large randomized trials published within the past 10 years concluded that preoperative vaginal cleansing with povidone-iodine reduced the risk of postcesarean endometritis in women who also received prophylactic IV antibiotics (TABLE).¹⁻³ Vaginal cleansing did not reduce the rate of postpartum fever or wound infection in these studies.
 

Clinical factors that increased the risk of postpartum endometritis independent of vaginal cleansing included:

Authors of two recent, large nonrandomized studies also have reported that vaginal cleansing reduced the risk of postcesarean endometritis.^4,5 By contrast, investigators from one large trial from 2001 did not observe a decrease in endometritis with vaginal cleansing.⁶

Cochrane review of precesarean vaginal cleansingAuthors of a Cochrane review, in which they synthesized 7 studies involving 2,635 women, reported that vaginal cleansing with povidone-iodine immediately before cesarean delivery was associated with a reduced risk of postcesarean endometritis: 8.3% vs 4.3% in the control and vaginal cleansing groups, respectively, (risk ratio [RR], 0.45; 95% confidence interval [CI], 0.25−0.81).⁷

The positive effect of vaginal cleansing was particularly noteworthy in the subgroup of women with ruptured membranes (3 trials involving 272 women). The rates of endometritis in the control versus vaginal cleansing groups were 17.9% and 4.3%, respectively (RR, 0.24; 95% CI, 0.10−0.55).

Women who went into labor prior to cesarean delivery (523 women from 3 trials) also benefitted from vaginal cleansing, with endometritis rates of 13.0% and 7.4% in the control and vaginal cleansing groups, respectively (RR, 0.56; 95% CI, 0.34−0.95).

In this review, again, vaginal cleansing did not significantly reduce the rate of postoperative fever or wound infection.

Is vaginal cleansing prior to cesarean delivery best practice?In the United States, precesarean vaginal cleansing is not a common practice. To close the gap between current practice and what is potentially a best practice, two approaches to using vaginal cleansing could be instituted in delivery units.

Approach #1: A liberal clinical protocol. In this scenario, all women (who are not allergic to iodine or povidone-iodine) undergoing cesarean delivery should undergo vaginal cleansing. The World Health Organization conditionally recommends vaginal cleansing for all women undergoing a cesarean delivery.⁸

Approach #2: A focused clinical protocol. For this protocol, only women (again, who are not allergic to iodine or povidone-iodine) who have ruptured membranes or are in labor upon advanced cervical examination should receive vaginal cleansing.

The advantage of a liberal protocol is that vaginal preparation becomes embedded within the standard practice of cesarean delivery and, hence, is seldom overlooked. The upside of the focused protocol is that only those women most likely to benefit receive the intervention.

Tell me what you thinkWill you consider using vaginal cleansing in your practice? Please let me know your views on vaginal cleansing for cesarean delivery, as well as your clinical pearls on cesarean delivery surgery, at obgmanagement.com. In addition, weigh in on the Quick Poll posted to OBG Management’s homepage. Send your letter to the editor to [email protected].

There are approximately 4,000,000 births annually in the United States, and about 32% of them occur by cesarean delivery. Compared with vaginal birth, cesarean delivery is associated with an increased risk of endometritis (defined as fever plus uterine or abdominal tenderness). Although surgical complications cannot be eliminated entirely, surgeons are deeply dedicated to the continuous improvement of surgical practice in order to reduce the risk of complications.

With cesarean delivery, many surgical practices have been adopted universally to reduce postoperative complications, including administration of intravenous (IV) antibiotics before skin incision to minimize postoperative infection and the use of postoperative mechanical or pharmacologic interventions to help prevent venous thromboembolism and pulmonary embolism. Preoperative vaginal cleansing with povidone-iodine may reduce the risk of postoperative endometritis, but the practice is not currently common in the United States.

Should you adopt a policy of preoperative vaginal cleansing prior to cesarean delivery? The data suggest perhaps you should.

Data-driven support for povidone-iodine precesareanThree large randomized trials published within the past 10 years concluded that preoperative vaginal cleansing with povidone-iodine reduced the risk of postcesarean endometritis in women who also received prophylactic IV antibiotics (TABLE).¹⁻³ Vaginal cleansing did not reduce the rate of postpartum fever or wound infection in these studies.
 

Clinical factors that increased the risk of postpartum endometritis independent of vaginal cleansing included:

Authors of two recent, large nonrandomized studies also have reported that vaginal cleansing reduced the risk of postcesarean endometritis.^4,5 By contrast, investigators from one large trial from 2001 did not observe a decrease in endometritis with vaginal cleansing.⁶

Cochrane review of precesarean vaginal cleansingAuthors of a Cochrane review, in which they synthesized 7 studies involving 2,635 women, reported that vaginal cleansing with povidone-iodine immediately before cesarean delivery was associated with a reduced risk of postcesarean endometritis: 8.3% vs 4.3% in the control and vaginal cleansing groups, respectively, (risk ratio [RR], 0.45; 95% confidence interval [CI], 0.25−0.81).⁷

The positive effect of vaginal cleansing was particularly noteworthy in the subgroup of women with ruptured membranes (3 trials involving 272 women). The rates of endometritis in the control versus vaginal cleansing groups were 17.9% and 4.3%, respectively (RR, 0.24; 95% CI, 0.10−0.55).

Women who went into labor prior to cesarean delivery (523 women from 3 trials) also benefitted from vaginal cleansing, with endometritis rates of 13.0% and 7.4% in the control and vaginal cleansing groups, respectively (RR, 0.56; 95% CI, 0.34−0.95).

In this review, again, vaginal cleansing did not significantly reduce the rate of postoperative fever or wound infection.

Is vaginal cleansing prior to cesarean delivery best practice?In the United States, precesarean vaginal cleansing is not a common practice. To close the gap between current practice and what is potentially a best practice, two approaches to using vaginal cleansing could be instituted in delivery units.

Approach #1: A liberal clinical protocol. In this scenario, all women (who are not allergic to iodine or povidone-iodine) undergoing cesarean delivery should undergo vaginal cleansing. The World Health Organization conditionally recommends vaginal cleansing for all women undergoing a cesarean delivery.⁸

Approach #2: A focused clinical protocol. For this protocol, only women (again, who are not allergic to iodine or povidone-iodine) who have ruptured membranes or are in labor upon advanced cervical examination should receive vaginal cleansing.

The advantage of a liberal protocol is that vaginal preparation becomes embedded within the standard practice of cesarean delivery and, hence, is seldom overlooked. The upside of the focused protocol is that only those women most likely to benefit receive the intervention.

Tell me what you thinkWill you consider using vaginal cleansing in your practice? Please let me know your views on vaginal cleansing for cesarean delivery, as well as your clinical pearls on cesarean delivery surgery, at obgmanagement.com. In addition, weigh in on the Quick Poll posted to OBG Management’s homepage. Send your letter to the editor to [email protected].

There are approximately 4,000,000 births annually in the United States, and about 32% of them occur by cesarean delivery. Compared with vaginal birth, cesarean delivery is associated with an increased risk of endometritis (defined as fever plus uterine or abdominal tenderness). Although surgical complications cannot be eliminated entirely, surgeons are deeply dedicated to the continuous improvement of surgical practice in order to reduce the risk of complications.

With cesarean delivery, many surgical practices have been adopted universally to reduce postoperative complications, including administration of intravenous (IV) antibiotics before skin incision to minimize postoperative infection and the use of postoperative mechanical or pharmacologic interventions to help prevent venous thromboembolism and pulmonary embolism. Preoperative vaginal cleansing with povidone-iodine may reduce the risk of postoperative endometritis, but the practice is not currently common in the United States.

Should you adopt a policy of preoperative vaginal cleansing prior to cesarean delivery? The data suggest perhaps you should.

Data-driven support for povidone-iodine precesareanThree large randomized trials published within the past 10 years concluded that preoperative vaginal cleansing with povidone-iodine reduced the risk of postcesarean endometritis in women who also received prophylactic IV antibiotics (TABLE).¹⁻³ Vaginal cleansing did not reduce the rate of postpartum fever or wound infection in these studies.
 

Clinical factors that increased the risk of postpartum endometritis independent of vaginal cleansing included:

Authors of two recent, large nonrandomized studies also have reported that vaginal cleansing reduced the risk of postcesarean endometritis.^4,5 By contrast, investigators from one large trial from 2001 did not observe a decrease in endometritis with vaginal cleansing.⁶

Cochrane review of precesarean vaginal cleansingAuthors of a Cochrane review, in which they synthesized 7 studies involving 2,635 women, reported that vaginal cleansing with povidone-iodine immediately before cesarean delivery was associated with a reduced risk of postcesarean endometritis: 8.3% vs 4.3% in the control and vaginal cleansing groups, respectively, (risk ratio [RR], 0.45; 95% confidence interval [CI], 0.25−0.81).⁷

The positive effect of vaginal cleansing was particularly noteworthy in the subgroup of women with ruptured membranes (3 trials involving 272 women). The rates of endometritis in the control versus vaginal cleansing groups were 17.9% and 4.3%, respectively (RR, 0.24; 95% CI, 0.10−0.55).

Women who went into labor prior to cesarean delivery (523 women from 3 trials) also benefitted from vaginal cleansing, with endometritis rates of 13.0% and 7.4% in the control and vaginal cleansing groups, respectively (RR, 0.56; 95% CI, 0.34−0.95).

In this review, again, vaginal cleansing did not significantly reduce the rate of postoperative fever or wound infection.

Is vaginal cleansing prior to cesarean delivery best practice?In the United States, precesarean vaginal cleansing is not a common practice. To close the gap between current practice and what is potentially a best practice, two approaches to using vaginal cleansing could be instituted in delivery units.

Approach #1: A liberal clinical protocol. In this scenario, all women (who are not allergic to iodine or povidone-iodine) undergoing cesarean delivery should undergo vaginal cleansing. The World Health Organization conditionally recommends vaginal cleansing for all women undergoing a cesarean delivery.⁸

Approach #2: A focused clinical protocol. For this protocol, only women (again, who are not allergic to iodine or povidone-iodine) who have ruptured membranes or are in labor upon advanced cervical examination should receive vaginal cleansing.

The advantage of a liberal protocol is that vaginal preparation becomes embedded within the standard practice of cesarean delivery and, hence, is seldom overlooked. The upside of the focused protocol is that only those women most likely to benefit receive the intervention.

Tell me what you thinkWill you consider using vaginal cleansing in your practice? Please let me know your views on vaginal cleansing for cesarean delivery, as well as your clinical pearls on cesarean delivery surgery, at obgmanagement.com. In addition, weigh in on the Quick Poll posted to OBG Management’s homepage. Send your letter to the editor to [email protected].

Preterm birth (PTB) remains a major cause of perinatal morbidity and mortality, and so its prediction and prevention are 2 of the most important issues in obstetrics. Cervical length (CL) measured by ultrasound has been shown to be the best predictor; several interventions (vaginal progesterone and cerclage) have been shown to be effective at reducing PTB if a short CL is identified. In fact, both the American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine (SMFM) recommend CL being measured every 2 weeks from 16 to 23 weeks in singletons with prior spontaneous PTB (sPTB), with cerclage placed for CL less than 25 mm. Moreover, both ACOG and SMFM recommend that “universal CL screening” (CL measured in singletons without a prior sPTB) be considered as a single measurement at about 18 to 23 weeks.

Details of the study
Rhoades and colleagues present data on CL screening done by transabdominal ultrasound (TAU), as an alternative to transvaginal ultrasound (TVU). This study confirms early data:

There are several other reasons why TVU is considered the gold standard for CL screening, and instead TAU CL should be avoided as possible. All randomized controlled trials that showed benefit from interventions (vaginal progesterone, cerclage, pessary) aimed at decreasing PTB in women with short CL used TVU CL screening and never TAU CL screening. In addition, TAU CL is less accurate than TVU CL screening. On TAU, fetal parts can obscure the cervix, obesity makes it hard to visualize CL, the distance between probe and cervix is longer, manual pressure can mask CL shortening, and bladder filling can elongate CL.⁷ Cost-effectiveness studies show that TVU CL screening is more effective, and less costly, compared with TAU CL screening, even in singletons without a prior sPTB.⁸

Societies such as ACOG and SMFM all have recommended TVU CL for prediction and prevention of PTB, over TAU CL.^9,10 Importantly, a TVU CL should be done by sonographers educated and trained formally, through such programs as those made available by SMFM.¹¹

What this evidence means for practice
If CL assessment is done, TVU should be preferred, as it is the gold standard, and not TAU.
>>Vincenzo Berghella, MD

Share your thoughts on this article! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

Preterm birth (PTB) remains a major cause of perinatal morbidity and mortality, and so its prediction and prevention are 2 of the most important issues in obstetrics. Cervical length (CL) measured by ultrasound has been shown to be the best predictor; several interventions (vaginal progesterone and cerclage) have been shown to be effective at reducing PTB if a short CL is identified. In fact, both the American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine (SMFM) recommend CL being measured every 2 weeks from 16 to 23 weeks in singletons with prior spontaneous PTB (sPTB), with cerclage placed for CL less than 25 mm. Moreover, both ACOG and SMFM recommend that “universal CL screening” (CL measured in singletons without a prior sPTB) be considered as a single measurement at about 18 to 23 weeks.

Details of the study
Rhoades and colleagues present data on CL screening done by transabdominal ultrasound (TAU), as an alternative to transvaginal ultrasound (TVU). This study confirms early data:

There are several other reasons why TVU is considered the gold standard for CL screening, and instead TAU CL should be avoided as possible. All randomized controlled trials that showed benefit from interventions (vaginal progesterone, cerclage, pessary) aimed at decreasing PTB in women with short CL used TVU CL screening and never TAU CL screening. In addition, TAU CL is less accurate than TVU CL screening. On TAU, fetal parts can obscure the cervix, obesity makes it hard to visualize CL, the distance between probe and cervix is longer, manual pressure can mask CL shortening, and bladder filling can elongate CL.⁷ Cost-effectiveness studies show that TVU CL screening is more effective, and less costly, compared with TAU CL screening, even in singletons without a prior sPTB.⁸

Societies such as ACOG and SMFM all have recommended TVU CL for prediction and prevention of PTB, over TAU CL.^9,10 Importantly, a TVU CL should be done by sonographers educated and trained formally, through such programs as those made available by SMFM.¹¹

What this evidence means for practice
If CL assessment is done, TVU should be preferred, as it is the gold standard, and not TAU.
>>Vincenzo Berghella, MD

Share your thoughts on this article! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

Preterm birth (PTB) remains a major cause of perinatal morbidity and mortality, and so its prediction and prevention are 2 of the most important issues in obstetrics. Cervical length (CL) measured by ultrasound has been shown to be the best predictor; several interventions (vaginal progesterone and cerclage) have been shown to be effective at reducing PTB if a short CL is identified. In fact, both the American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine (SMFM) recommend CL being measured every 2 weeks from 16 to 23 weeks in singletons with prior spontaneous PTB (sPTB), with cerclage placed for CL less than 25 mm. Moreover, both ACOG and SMFM recommend that “universal CL screening” (CL measured in singletons without a prior sPTB) be considered as a single measurement at about 18 to 23 weeks.

Details of the study
Rhoades and colleagues present data on CL screening done by transabdominal ultrasound (TAU), as an alternative to transvaginal ultrasound (TVU). This study confirms early data:

There are several other reasons why TVU is considered the gold standard for CL screening, and instead TAU CL should be avoided as possible. All randomized controlled trials that showed benefit from interventions (vaginal progesterone, cerclage, pessary) aimed at decreasing PTB in women with short CL used TVU CL screening and never TAU CL screening. In addition, TAU CL is less accurate than TVU CL screening. On TAU, fetal parts can obscure the cervix, obesity makes it hard to visualize CL, the distance between probe and cervix is longer, manual pressure can mask CL shortening, and bladder filling can elongate CL.⁷ Cost-effectiveness studies show that TVU CL screening is more effective, and less costly, compared with TAU CL screening, even in singletons without a prior sPTB.⁸

Societies such as ACOG and SMFM all have recommended TVU CL for prediction and prevention of PTB, over TAU CL.^9,10 Importantly, a TVU CL should be done by sonographers educated and trained formally, through such programs as those made available by SMFM.¹¹

What this evidence means for practice
If CL assessment is done, TVU should be preferred, as it is the gold standard, and not TAU.
>>Vincenzo Berghella, MD

Share your thoughts on this article! Send your Letter to the Editor to [email protected]. Please include your name and the city and state in which you practice.

Chagas disease is a vector-borne parasitic disease, endemic to the Americas, that remains as little recognized by U.S. patients and practitioners as the obscure winged insects that transmit it.

Transmission occurs when triatomine bugs, commonly called “kissing bugs,” pierce the skin to feed and leave behind parasite-infected feces that can enter the bloodstream; pregnant women can also transmit Chagas to their newborns.

About a third of patients infected with Trypanosoma cruzi, the protozoan parasite that causes Chagas, will develop cardiac abnormalities such as cardiomyopathy, arrhythmias, and heart failure – often decades after becoming infected. In the United States, where blood banks began screening for Chagas in 2007, patients without symptoms are likely to learn they are positive only after donating blood.

Conventional wisdom has long maintained that Chagas is limited to Central and South America. But immigration from Chagas-endemic countries, such as El Salvador, Mexico, and Bolivia, means more people are living with the disease in the United States.

“One percent of the Latin American immigrant population we screen [in Los Angeles] has Chagas,” said Dr. Sheba K. Meymandi, cardiologist and director of Center of Excellence for Chagas Disease at Olive View–UCLA Medical Center in Los Angeles, who also works with the city’s health department to detect Chagas. “That’s huge.”

Meanwhile, blood banks are discovering more cases among people without ties to Latin America, and species of kissing bugs native to the southern United States are increasingly recognized as a non-negligible source of Chagas transmission. Of 39 Chagas cases reported to Texas health authorities in 2013 and 2014, 12 were thought to be locally acquired.

Dr. Heather Yun, an infectious disease specialist at the San Antonio Military Medical Center, said risk factors for local transmission are not well established, but “we think people who are living in poverty in substandard housing, people who spend a lot of time outdoors, especially at night, and people involved with direct blood contact with wild game in Southern parts of the United States” may be at higher risk.

A U.S. disease

Evidence is amassing quickly that Chagas is a U.S. disease. But U.S. clinicians still lag in their knowledge of it, say physicians treating Chagas cases. “In medical school we get a 2-hour lecture on it, and it’s always been presented as an exotic disease and one you don’t treat,” Dr. Meymandi said.

The persistent perception of Chagas as a foreign disease means clinicians are inclined to dismiss positive results from a blood screening, particularly from someone who is not from Latin America. Yet cardiologists, ID practitioners, obstetricians, and primary care physicians all need to be aware that cases do occur in the United States and are potentially treatable.

Dr. Laila Woc-Colburn, an infectious disease and tropical medicine specialist at Baylor College of Medicine in Houston, said many people with Chagas never make it to an infectious disease specialist or cardiologist for a work-up. “When you test positive on serology [after a blood donation], you get a letter recommending you consult your physician. Most will go to their primary care doctors, who might say ‘this isn’t a disease in the United States.’ In Houston, that is often the case.”

Dr. Meymandi, who has treated hundreds of patients with Chagas with and without cardiac involvement, said any physician with a potential Chagas case must act. “If you get someone that’s positive, it’s your duty as a physician to confirm the positivity with CDC,” she said.

Dr. Yun concurred. “The most important message is, do something,” she said. “Don’t just assume it’s a false positive.”

Diagnosis is not simple and requires testing beyond the initial ELISA assay used in blood-bank screening. Confirmatory tests must be carried out in coordination with the Centers for Disease Control and Prevention. Also, with no agents approved by the Food and Drug Administration to treat Chagas, treatment is available only through the CDC’s investigational drugs protocol. Both drugs used in Chagas, benznidazole and nifurtimox, come with serious adverse effects that must be closely monitored.

“It’s time consuming, filling out the forms, getting the consent, tracking and sending back lab results to CDC in order to get drugs – it’s not like you can just write a prescription,” Dr. Meymandi said. But, “if you don’t know how to treat the patient or don’t have time, find someone like me,” she noted, adding that she is available to counsel any physician daunted by a potential Chagas case.

Treatment options

No formal clinical algorithm exists for Chagas, but Dr. Meymandi, Dr. Yun, and Dr. Woc-Colburn all pointed to a 2007 JAMA article, which describes diagnosis and treatment protocols, as an important reference for clinicians to start with. It’s “the best approximation of a clinical guideline we have,” Dr. Yun said (JAMA. 2007;298[18]:2171-81. doi:10.1001/jama.298.18.2171).

Dr. Meymandi, who has treated more Chagas patients than has any other U.S. clinician, said that treatment has changed somewhat since the JAMA article was published. In 2007, she said, nifurtimox was the main drug available through CDC, while benznidazole, which is somewhat better tolerated and has shorter treatment duration, has since become the first-line agent.

“We’ve lowered the dose of benznidazole, maxing out at 400 mg/day to decrease the toxicity,” she said. Also, treatment is now being extended to some patients aged 60 years and older.

The decision to treat or not treat, clinicians say, depends on the patient’s age, disease progression, comorbidities and potential serious drug interactions, and willingness to tolerate side effects that, with nifurtimox especially, can include skin sloughing, rash, and psychological and neurologic symptoms including depression and peripheral neuropathy.

“If you don’t have side effects, you’re not taking the drugs,” Dr. Meymandi said. Dr. Woc-Colburn noted that polypharmacy was a major consideration when treating older adults for Chagas. “If I have a patient who has diabetes, obesity, [and] end-stage renal disease, it’s not going to be ideal to give [benznidazole].”

Recent, highly anticipated results from BENEFIT, a large randomized trial (n = 2,854) showed that benznidazole reduced parasite load but was not helpful in halting cardiac damage at 5 years’ follow-up in patients with established Chagas cardiomyopathy (N Engl J Med. 2015 Oct;373:1295-306. doi:10.1056/NEJMoa1507574).

Dr. Meymandi, whose earlier research established that Chagas cardiomyopathy carries significantly higher morbidity and mortality than does non–Chagas cardiomyopathy (Circulation. 2012;126:A18171), said that the BENEFIT results underscore the need for physicians to be bullish in their approach to treating Chagas soon after diagnosis.

“It doesn’t matter if they’re symptomatic or asymptomatic. You can’t wait till they progress to treat. If you wait for the progression of disease you’ve lost the battle. You can’t wait and follow conservatively until you see the complications, because once those complications have started the parasitic load is too high for you to have an impact,” she said.

Dr. Yun said that given the toxicity of current treatment, she hoped to see more studies show clearer evidence of clinical benefit, “either reductions in mortality or reductions in end organ disease.” Most studies “have focused on clearance of parasite, which is important, but it’s not as important decreasing the risk of death or cardiomyopathy or heart failure.”

Rick Tarleton, Ph.D., a biologist the University of Georgia, in Athens, who has worked on Chagas for more than 30 years, said that because Chagas pathology is directly tied to parasite load – and not, as people have suggested in the past, an autoimmune reaction resulting from parasite exposure – drug treatment may prove to be worthwhile even in patients with significant cardiac involvement.

“You get rid of the parasite, you get rid of the progression of the disease,” Dr. Tarleton said. Even the findings from the BENEFIT trial, he said, did not lead him to conclude that treatment in people with established cardiac disease was futile.

“If you’re treating people who are already chronically infected and showing symptoms, the question is not have you reversed the damage, it’s have you stopped accumulating damage,” he noted. “And a 5-year follow-up is probably not long enough to know whether you’ve stopped accumulating.”

“We have drugs, they’re not great, they do have side effects, they don’t always work,” Dr. Tarleton said. “But they’re better than nothing. And they ought to be more widely used.”

Dr. Meymandi said that current supplies of benznidazole at CDC are low and that a dozen patients at her clinic are awaiting treatment. Meanwhile, access may soon be complicated further by the announcement, this month, that KaloBios Pharmaceuticals had bought the rights to seek FDA approval of benznidazole and market it in the United States.

The same company’s CEO came under fire in recent months for acquiring rights to an inexpensive drug to treat toxoplasmosis in AIDS patients, then announcing a price increase from $13.50 to $750 a pill.

“Everyone’s really concerned,” Dr. Meymandi said, “because Chagas is a disease of the poor.”

Chagas disease is a vector-borne parasitic disease, endemic to the Americas, that remains as little recognized by U.S. patients and practitioners as the obscure winged insects that transmit it.

Transmission occurs when triatomine bugs, commonly called “kissing bugs,” pierce the skin to feed and leave behind parasite-infected feces that can enter the bloodstream; pregnant women can also transmit Chagas to their newborns.

About a third of patients infected with Trypanosoma cruzi, the protozoan parasite that causes Chagas, will develop cardiac abnormalities such as cardiomyopathy, arrhythmias, and heart failure – often decades after becoming infected. In the United States, where blood banks began screening for Chagas in 2007, patients without symptoms are likely to learn they are positive only after donating blood.

Conventional wisdom has long maintained that Chagas is limited to Central and South America. But immigration from Chagas-endemic countries, such as El Salvador, Mexico, and Bolivia, means more people are living with the disease in the United States.

“One percent of the Latin American immigrant population we screen [in Los Angeles] has Chagas,” said Dr. Sheba K. Meymandi, cardiologist and director of Center of Excellence for Chagas Disease at Olive View–UCLA Medical Center in Los Angeles, who also works with the city’s health department to detect Chagas. “That’s huge.”

Meanwhile, blood banks are discovering more cases among people without ties to Latin America, and species of kissing bugs native to the southern United States are increasingly recognized as a non-negligible source of Chagas transmission. Of 39 Chagas cases reported to Texas health authorities in 2013 and 2014, 12 were thought to be locally acquired.

Dr. Heather Yun, an infectious disease specialist at the San Antonio Military Medical Center, said risk factors for local transmission are not well established, but “we think people who are living in poverty in substandard housing, people who spend a lot of time outdoors, especially at night, and people involved with direct blood contact with wild game in Southern parts of the United States” may be at higher risk.

A U.S. disease

Evidence is amassing quickly that Chagas is a U.S. disease. But U.S. clinicians still lag in their knowledge of it, say physicians treating Chagas cases. “In medical school we get a 2-hour lecture on it, and it’s always been presented as an exotic disease and one you don’t treat,” Dr. Meymandi said.

The persistent perception of Chagas as a foreign disease means clinicians are inclined to dismiss positive results from a blood screening, particularly from someone who is not from Latin America. Yet cardiologists, ID practitioners, obstetricians, and primary care physicians all need to be aware that cases do occur in the United States and are potentially treatable.

Dr. Laila Woc-Colburn, an infectious disease and tropical medicine specialist at Baylor College of Medicine in Houston, said many people with Chagas never make it to an infectious disease specialist or cardiologist for a work-up. “When you test positive on serology [after a blood donation], you get a letter recommending you consult your physician. Most will go to their primary care doctors, who might say ‘this isn’t a disease in the United States.’ In Houston, that is often the case.”

Dr. Meymandi, who has treated hundreds of patients with Chagas with and without cardiac involvement, said any physician with a potential Chagas case must act. “If you get someone that’s positive, it’s your duty as a physician to confirm the positivity with CDC,” she said.

Dr. Yun concurred. “The most important message is, do something,” she said. “Don’t just assume it’s a false positive.”

Diagnosis is not simple and requires testing beyond the initial ELISA assay used in blood-bank screening. Confirmatory tests must be carried out in coordination with the Centers for Disease Control and Prevention. Also, with no agents approved by the Food and Drug Administration to treat Chagas, treatment is available only through the CDC’s investigational drugs protocol. Both drugs used in Chagas, benznidazole and nifurtimox, come with serious adverse effects that must be closely monitored.

“It’s time consuming, filling out the forms, getting the consent, tracking and sending back lab results to CDC in order to get drugs – it’s not like you can just write a prescription,” Dr. Meymandi said. But, “if you don’t know how to treat the patient or don’t have time, find someone like me,” she noted, adding that she is available to counsel any physician daunted by a potential Chagas case.

Treatment options

No formal clinical algorithm exists for Chagas, but Dr. Meymandi, Dr. Yun, and Dr. Woc-Colburn all pointed to a 2007 JAMA article, which describes diagnosis and treatment protocols, as an important reference for clinicians to start with. It’s “the best approximation of a clinical guideline we have,” Dr. Yun said (JAMA. 2007;298[18]:2171-81. doi:10.1001/jama.298.18.2171).

Dr. Meymandi, who has treated more Chagas patients than has any other U.S. clinician, said that treatment has changed somewhat since the JAMA article was published. In 2007, she said, nifurtimox was the main drug available through CDC, while benznidazole, which is somewhat better tolerated and has shorter treatment duration, has since become the first-line agent.

“We’ve lowered the dose of benznidazole, maxing out at 400 mg/day to decrease the toxicity,” she said. Also, treatment is now being extended to some patients aged 60 years and older.

The decision to treat or not treat, clinicians say, depends on the patient’s age, disease progression, comorbidities and potential serious drug interactions, and willingness to tolerate side effects that, with nifurtimox especially, can include skin sloughing, rash, and psychological and neurologic symptoms including depression and peripheral neuropathy.

“If you don’t have side effects, you’re not taking the drugs,” Dr. Meymandi said. Dr. Woc-Colburn noted that polypharmacy was a major consideration when treating older adults for Chagas. “If I have a patient who has diabetes, obesity, [and] end-stage renal disease, it’s not going to be ideal to give [benznidazole].”

Recent, highly anticipated results from BENEFIT, a large randomized trial (n = 2,854) showed that benznidazole reduced parasite load but was not helpful in halting cardiac damage at 5 years’ follow-up in patients with established Chagas cardiomyopathy (N Engl J Med. 2015 Oct;373:1295-306. doi:10.1056/NEJMoa1507574).

Dr. Meymandi, whose earlier research established that Chagas cardiomyopathy carries significantly higher morbidity and mortality than does non–Chagas cardiomyopathy (Circulation. 2012;126:A18171), said that the BENEFIT results underscore the need for physicians to be bullish in their approach to treating Chagas soon after diagnosis.

“It doesn’t matter if they’re symptomatic or asymptomatic. You can’t wait till they progress to treat. If you wait for the progression of disease you’ve lost the battle. You can’t wait and follow conservatively until you see the complications, because once those complications have started the parasitic load is too high for you to have an impact,” she said.

Dr. Yun said that given the toxicity of current treatment, she hoped to see more studies show clearer evidence of clinical benefit, “either reductions in mortality or reductions in end organ disease.” Most studies “have focused on clearance of parasite, which is important, but it’s not as important decreasing the risk of death or cardiomyopathy or heart failure.”

Rick Tarleton, Ph.D., a biologist the University of Georgia, in Athens, who has worked on Chagas for more than 30 years, said that because Chagas pathology is directly tied to parasite load – and not, as people have suggested in the past, an autoimmune reaction resulting from parasite exposure – drug treatment may prove to be worthwhile even in patients with significant cardiac involvement.

“You get rid of the parasite, you get rid of the progression of the disease,” Dr. Tarleton said. Even the findings from the BENEFIT trial, he said, did not lead him to conclude that treatment in people with established cardiac disease was futile.

“If you’re treating people who are already chronically infected and showing symptoms, the question is not have you reversed the damage, it’s have you stopped accumulating damage,” he noted. “And a 5-year follow-up is probably not long enough to know whether you’ve stopped accumulating.”

“We have drugs, they’re not great, they do have side effects, they don’t always work,” Dr. Tarleton said. “But they’re better than nothing. And they ought to be more widely used.”

Dr. Meymandi said that current supplies of benznidazole at CDC are low and that a dozen patients at her clinic are awaiting treatment. Meanwhile, access may soon be complicated further by the announcement, this month, that KaloBios Pharmaceuticals had bought the rights to seek FDA approval of benznidazole and market it in the United States.

The same company’s CEO came under fire in recent months for acquiring rights to an inexpensive drug to treat toxoplasmosis in AIDS patients, then announcing a price increase from $13.50 to $750 a pill.

“Everyone’s really concerned,” Dr. Meymandi said, “because Chagas is a disease of the poor.”

Chagas disease is a vector-borne parasitic disease, endemic to the Americas, that remains as little recognized by U.S. patients and practitioners as the obscure winged insects that transmit it.

Transmission occurs when triatomine bugs, commonly called “kissing bugs,” pierce the skin to feed and leave behind parasite-infected feces that can enter the bloodstream; pregnant women can also transmit Chagas to their newborns.

About a third of patients infected with Trypanosoma cruzi, the protozoan parasite that causes Chagas, will develop cardiac abnormalities such as cardiomyopathy, arrhythmias, and heart failure – often decades after becoming infected. In the United States, where blood banks began screening for Chagas in 2007, patients without symptoms are likely to learn they are positive only after donating blood.

Conventional wisdom has long maintained that Chagas is limited to Central and South America. But immigration from Chagas-endemic countries, such as El Salvador, Mexico, and Bolivia, means more people are living with the disease in the United States.

“One percent of the Latin American immigrant population we screen [in Los Angeles] has Chagas,” said Dr. Sheba K. Meymandi, cardiologist and director of Center of Excellence for Chagas Disease at Olive View–UCLA Medical Center in Los Angeles, who also works with the city’s health department to detect Chagas. “That’s huge.”

Meanwhile, blood banks are discovering more cases among people without ties to Latin America, and species of kissing bugs native to the southern United States are increasingly recognized as a non-negligible source of Chagas transmission. Of 39 Chagas cases reported to Texas health authorities in 2013 and 2014, 12 were thought to be locally acquired.

Dr. Heather Yun, an infectious disease specialist at the San Antonio Military Medical Center, said risk factors for local transmission are not well established, but “we think people who are living in poverty in substandard housing, people who spend a lot of time outdoors, especially at night, and people involved with direct blood contact with wild game in Southern parts of the United States” may be at higher risk.

A U.S. disease

Evidence is amassing quickly that Chagas is a U.S. disease. But U.S. clinicians still lag in their knowledge of it, say physicians treating Chagas cases. “In medical school we get a 2-hour lecture on it, and it’s always been presented as an exotic disease and one you don’t treat,” Dr. Meymandi said.

The persistent perception of Chagas as a foreign disease means clinicians are inclined to dismiss positive results from a blood screening, particularly from someone who is not from Latin America. Yet cardiologists, ID practitioners, obstetricians, and primary care physicians all need to be aware that cases do occur in the United States and are potentially treatable.

Dr. Laila Woc-Colburn, an infectious disease and tropical medicine specialist at Baylor College of Medicine in Houston, said many people with Chagas never make it to an infectious disease specialist or cardiologist for a work-up. “When you test positive on serology [after a blood donation], you get a letter recommending you consult your physician. Most will go to their primary care doctors, who might say ‘this isn’t a disease in the United States.’ In Houston, that is often the case.”

Dr. Meymandi, who has treated hundreds of patients with Chagas with and without cardiac involvement, said any physician with a potential Chagas case must act. “If you get someone that’s positive, it’s your duty as a physician to confirm the positivity with CDC,” she said.

Dr. Yun concurred. “The most important message is, do something,” she said. “Don’t just assume it’s a false positive.”

Diagnosis is not simple and requires testing beyond the initial ELISA assay used in blood-bank screening. Confirmatory tests must be carried out in coordination with the Centers for Disease Control and Prevention. Also, with no agents approved by the Food and Drug Administration to treat Chagas, treatment is available only through the CDC’s investigational drugs protocol. Both drugs used in Chagas, benznidazole and nifurtimox, come with serious adverse effects that must be closely monitored.

“It’s time consuming, filling out the forms, getting the consent, tracking and sending back lab results to CDC in order to get drugs – it’s not like you can just write a prescription,” Dr. Meymandi said. But, “if you don’t know how to treat the patient or don’t have time, find someone like me,” she noted, adding that she is available to counsel any physician daunted by a potential Chagas case.

Treatment options

No formal clinical algorithm exists for Chagas, but Dr. Meymandi, Dr. Yun, and Dr. Woc-Colburn all pointed to a 2007 JAMA article, which describes diagnosis and treatment protocols, as an important reference for clinicians to start with. It’s “the best approximation of a clinical guideline we have,” Dr. Yun said (JAMA. 2007;298[18]:2171-81. doi:10.1001/jama.298.18.2171).

Dr. Meymandi, who has treated more Chagas patients than has any other U.S. clinician, said that treatment has changed somewhat since the JAMA article was published. In 2007, she said, nifurtimox was the main drug available through CDC, while benznidazole, which is somewhat better tolerated and has shorter treatment duration, has since become the first-line agent.

“We’ve lowered the dose of benznidazole, maxing out at 400 mg/day to decrease the toxicity,” she said. Also, treatment is now being extended to some patients aged 60 years and older.

The decision to treat or not treat, clinicians say, depends on the patient’s age, disease progression, comorbidities and potential serious drug interactions, and willingness to tolerate side effects that, with nifurtimox especially, can include skin sloughing, rash, and psychological and neurologic symptoms including depression and peripheral neuropathy.

“If you don’t have side effects, you’re not taking the drugs,” Dr. Meymandi said. Dr. Woc-Colburn noted that polypharmacy was a major consideration when treating older adults for Chagas. “If I have a patient who has diabetes, obesity, [and] end-stage renal disease, it’s not going to be ideal to give [benznidazole].”

Recent, highly anticipated results from BENEFIT, a large randomized trial (n = 2,854) showed that benznidazole reduced parasite load but was not helpful in halting cardiac damage at 5 years’ follow-up in patients with established Chagas cardiomyopathy (N Engl J Med. 2015 Oct;373:1295-306. doi:10.1056/NEJMoa1507574).

Dr. Meymandi, whose earlier research established that Chagas cardiomyopathy carries significantly higher morbidity and mortality than does non–Chagas cardiomyopathy (Circulation. 2012;126:A18171), said that the BENEFIT results underscore the need for physicians to be bullish in their approach to treating Chagas soon after diagnosis.

“It doesn’t matter if they’re symptomatic or asymptomatic. You can’t wait till they progress to treat. If you wait for the progression of disease you’ve lost the battle. You can’t wait and follow conservatively until you see the complications, because once those complications have started the parasitic load is too high for you to have an impact,” she said.

Dr. Yun said that given the toxicity of current treatment, she hoped to see more studies show clearer evidence of clinical benefit, “either reductions in mortality or reductions in end organ disease.” Most studies “have focused on clearance of parasite, which is important, but it’s not as important decreasing the risk of death or cardiomyopathy or heart failure.”

Rick Tarleton, Ph.D., a biologist the University of Georgia, in Athens, who has worked on Chagas for more than 30 years, said that because Chagas pathology is directly tied to parasite load – and not, as people have suggested in the past, an autoimmune reaction resulting from parasite exposure – drug treatment may prove to be worthwhile even in patients with significant cardiac involvement.

“You get rid of the parasite, you get rid of the progression of the disease,” Dr. Tarleton said. Even the findings from the BENEFIT trial, he said, did not lead him to conclude that treatment in people with established cardiac disease was futile.

“If you’re treating people who are already chronically infected and showing symptoms, the question is not have you reversed the damage, it’s have you stopped accumulating damage,” he noted. “And a 5-year follow-up is probably not long enough to know whether you’ve stopped accumulating.”

“We have drugs, they’re not great, they do have side effects, they don’t always work,” Dr. Tarleton said. “But they’re better than nothing. And they ought to be more widely used.”

Dr. Meymandi said that current supplies of benznidazole at CDC are low and that a dozen patients at her clinic are awaiting treatment. Meanwhile, access may soon be complicated further by the announcement, this month, that KaloBios Pharmaceuticals had bought the rights to seek FDA approval of benznidazole and market it in the United States.

The same company’s CEO came under fire in recent months for acquiring rights to an inexpensive drug to treat toxoplasmosis in AIDS patients, then announcing a price increase from $13.50 to $750 a pill.

“Everyone’s really concerned,” Dr. Meymandi said, “because Chagas is a disease of the poor.”

Women who become pregnant while in remission from Hodgkin lymphoma were not at increased risk for cancer relapse, according to an analysis of data from Swedish health care registries combined with medical records.

Of 449 women who were diagnosed with Hodgkin lymphoma between 1992 and 2009, 144 (32%) became pregnant during follow-up, which started 6 months after diagnosis, when the disease was assumed to be in remission. Only one of these women experienced a pregnancy-associated relapse, which was defined as a relapse occurring during pregnancy or within 5 years of delivery. Of the women who did not become pregnant, 46 had a relapse.

©Jupiterimages/thinkstockphotos.com

The effect of pregnancy on relapse has been a concern of patients and clinicians, but “our findings suggest that the risk of pregnancy-associated relapse does not need to be taken into account in family planning for women whose Hodgkin lymphoma is in remission,” said Caroline E. Weibull of Karolinska Institutet in Stockholm, and her associates.

The researchers used the nationwide “Swedish Cancer Register” to identify all cases of Hodgkin lymphoma (reporting is mandatory) and merged this data with clinical information from other registries and medical records.

The pregnancy rates were similar among women who had limited- and advanced-stage disease and among women with and without B symptoms at diagnosis – a finding that negates consideration of a so-called “healthy mother effect” in protecting against relapse, they wrote (J Clin Onc. 2015 Dec. 14 [doi:10.1200/JCO.2015.63.3446]).

The researchers also found that the absolute risk for relapse was highest in the first 2-3 years after diagnosis, which suggests that women should be advised, “if possible, to wait 2 years after cessation of treatment before becoming pregnant.” Additionally, the relapse rate more than doubled in women aged 30 years or older at diagnosis, compared with women aged 18-24 years at diagnosis – a finding consistent with previous research, they noted.

Women in the study were aged 18-40 at diagnosis. Follow-up ended on the date of relapse, the date of death, or at the end of 2010, whichever came first.

Women who become pregnant while in remission from Hodgkin lymphoma were not at increased risk for cancer relapse, according to an analysis of data from Swedish health care registries combined with medical records.

Of 449 women who were diagnosed with Hodgkin lymphoma between 1992 and 2009, 144 (32%) became pregnant during follow-up, which started 6 months after diagnosis, when the disease was assumed to be in remission. Only one of these women experienced a pregnancy-associated relapse, which was defined as a relapse occurring during pregnancy or within 5 years of delivery. Of the women who did not become pregnant, 46 had a relapse.

©Jupiterimages/thinkstockphotos.com

The effect of pregnancy on relapse has been a concern of patients and clinicians, but “our findings suggest that the risk of pregnancy-associated relapse does not need to be taken into account in family planning for women whose Hodgkin lymphoma is in remission,” said Caroline E. Weibull of Karolinska Institutet in Stockholm, and her associates.

The researchers used the nationwide “Swedish Cancer Register” to identify all cases of Hodgkin lymphoma (reporting is mandatory) and merged this data with clinical information from other registries and medical records.

The pregnancy rates were similar among women who had limited- and advanced-stage disease and among women with and without B symptoms at diagnosis – a finding that negates consideration of a so-called “healthy mother effect” in protecting against relapse, they wrote (J Clin Onc. 2015 Dec. 14 [doi:10.1200/JCO.2015.63.3446]).

The researchers also found that the absolute risk for relapse was highest in the first 2-3 years after diagnosis, which suggests that women should be advised, “if possible, to wait 2 years after cessation of treatment before becoming pregnant.” Additionally, the relapse rate more than doubled in women aged 30 years or older at diagnosis, compared with women aged 18-24 years at diagnosis – a finding consistent with previous research, they noted.

Women in the study were aged 18-40 at diagnosis. Follow-up ended on the date of relapse, the date of death, or at the end of 2010, whichever came first.

Women who become pregnant while in remission from Hodgkin lymphoma were not at increased risk for cancer relapse, according to an analysis of data from Swedish health care registries combined with medical records.

Of 449 women who were diagnosed with Hodgkin lymphoma between 1992 and 2009, 144 (32%) became pregnant during follow-up, which started 6 months after diagnosis, when the disease was assumed to be in remission. Only one of these women experienced a pregnancy-associated relapse, which was defined as a relapse occurring during pregnancy or within 5 years of delivery. Of the women who did not become pregnant, 46 had a relapse.

©Jupiterimages/thinkstockphotos.com

The effect of pregnancy on relapse has been a concern of patients and clinicians, but “our findings suggest that the risk of pregnancy-associated relapse does not need to be taken into account in family planning for women whose Hodgkin lymphoma is in remission,” said Caroline E. Weibull of Karolinska Institutet in Stockholm, and her associates.

The researchers used the nationwide “Swedish Cancer Register” to identify all cases of Hodgkin lymphoma (reporting is mandatory) and merged this data with clinical information from other registries and medical records.

The pregnancy rates were similar among women who had limited- and advanced-stage disease and among women with and without B symptoms at diagnosis – a finding that negates consideration of a so-called “healthy mother effect” in protecting against relapse, they wrote (J Clin Onc. 2015 Dec. 14 [doi:10.1200/JCO.2015.63.3446]).

The researchers also found that the absolute risk for relapse was highest in the first 2-3 years after diagnosis, which suggests that women should be advised, “if possible, to wait 2 years after cessation of treatment before becoming pregnant.” Additionally, the relapse rate more than doubled in women aged 30 years or older at diagnosis, compared with women aged 18-24 years at diagnosis – a finding consistent with previous research, they noted.

Women in the study were aged 18-40 at diagnosis. Follow-up ended on the date of relapse, the date of death, or at the end of 2010, whichever came first.

A number of drugs are available for ophthalmic use. This review focuses on single drug products and, although combination drug products are not discussed, these formulations typically include the drugs reviewed here.

Since it appears that ophthalmic medications are commonly used for a wide range of conditions and ages, one would expect to see numerous reports of their use in the eyes of pregnant or breastfeeding patients. Unfortunately, the opposite is true. The majority of the drugs have no human pregnancy or lactation data. When there are human data, it invariably involves the systemic use of the drug for other indications, rather than its ophthalmic use. Moreover, the animal reproduction data are usually not relevant because they involve systemic drugs (e.g., IV or oral). Consequently, determining the level of risk an ophthalmic drug presents to an embryo and/or fetus is primarily based on time and dose, the two cardinal principles of teratology.

Avoiding exposure during organogenesis – the period when a drug can cause developmental toxicity (altered growth, structural anomalies, functional and/or behavioral deficits, or death) – is usually best, but may not be possible in some cases, including glaucoma, eye infections, and eye surgery. Fortunately, the systemic concentrations of drugs applied topically to the eye are typically thought to be low, even though the levels of most drugs have not been studied. Thus, the risk to the embryo and/or fetus in most cases can be considered low and the drug classified as compatible in pregnancy and breastfeeding.

If a topical eye drug must be used during pregnancy or lactation, teach the patient how to decrease the amount of drug reaching the systemic circulation. This involves placing pressure over the tear duct in the corner of the eye for 1 minute or more, then removing any excess solution with absorbent tissue.

In the sections below, drugs are shown by indication or by pharmacologic class. The term “human eye data” refers to the use of the drug in pregnancy and/or lactation.

Glaucoma

If you are caring for a pregnant patient who is being treated for glaucoma, two recent reviews may be helpful: Surv Ophthalmol. 2011 Jul-Aug;56(4):324-35 and Curr Opin Ophthalmol. 2014 Mar;25(2):93-7.

Sympathomimetics (alpha-adrenergic agonists) include apraclonidine (Iopidine), which has no human eye data, and brimonidine (Alphagan P), which has one case report in pregnancy and breastfeeding showing no fetal or nursing infant harm.

Four of the five beta-adrenergic blockers have no human eye data: betaxolol (Betoptic), carteolol (Ocupress), levobunolol (Betagan), and metipranolol, but there are two case reports for timolol (Betimol, Istalol, Timoptic, Timoptic-XE) showing no fetal harm in one newborn and growth restriction in the other.

Among the miotics, there are limited human eye data for pilocarpine (Isopto Carpine) and no fetal harm was observed. For echothiophate iodide (Phospholine Iodide), there is one case report of a normal full-term infant whose mother was treated with the agent up to 32 weeks’ gestation and then with pilocarpine for 8 weeks (Arch Ophthalmol. 1968 Mar;79[3]:283-5).

Fuse/Thinkstock.com

Carbonic anhydrase inhibitors include brinzolamide (Azopt), which has no human eye data, and dorzolamide (Trusopt), which has growth restriction in one case treated with fixed combination of dorzolamide and timolol. There are no human eye data for unoprostone isopropyl (Rescula), a synthetic docosanoid.

Of the four prostaglandin analogs, three have no human eye data, bimatoprost (Lumigan), tafluprost (Zioptan), and travoprost (Travatan Z). There were 11 pregnancies exposed to latanoprost (Xalatan). The outcomes of these cases were one lost to follow-up, one miscarriage, and nine infants without congenital anomalies (Am J Ophthalmol. 2004 Aug;138[2]:305-6).

Mitomycin (Mitosol) is an antimetabolite that is given topically to the surgical site of glaucoma filtration surgery. No reports of its use in pregnant humans have been located. According to the manufacturer, clinically significant systemic concentrations are not expected.

Antiseptics

Povidone-iodine (Betadine) is indicated for prepping of the periocular region. There does not appear to be any risk to the embryo-fetus or nursing infant from this indication.

Antihistamines

The four ophthalmic agents in this class are alcaftadine (Lastacaft), azelastine (Optivar), emedastine (Emadine), and epinastine (Elestat). There are no human eye data for these agents but, like antihistamines given systemically, they are probably compatible in pregnancy and lactation.

Antihistamine-mast cell stabilizers

There are no human eye data for bepotastine (Bepreve), ketotifen (Alaway), and olopatadine (Pataday, Patanol). Peak plasma concentrations of bepotastine were 5.1-7.3 ng/mL for 1-2 hours after instillation and were less than 2 ng/mL at 24 hours. It appears that these drugs can also be classified as compatible in pregnancy and lactation.

Anti-infectives

Ten anti-infectives are available for topical treatment of eye infections (systemic concentrations if known): besifloxacin (Besivance) (0.43 ng/mL), ciprofloxacin (less than 2.5 ng/mL), gentamicin, gatifloxacin (Zymaxid) (less than 5 ng/mL), levofloxacin (Iquix) (10.9 ng/mL), moxifloxacin (Vigamox) (2.7 ng/mL), ofloxacin (1.9 ng/mL), sulfacetamide (Isopto Cetamide), and tobramycin. The tenth agent, natamycin (Natacyn), is an antifungal. According to the manufacturer, systemic absorption is not expected.

None of these agents have human eye data, but they are usually considered compatible in pregnancy and breastfeeding when systemic formulations are used, so they should be compatible when used in the eye.

Antivirals

Ganciclovir (Zirgan) has no human eye data but, according to the manufacturer, the daily ophthalmic dose is about 0.04% and 0.1% of the oral and IV doses, respectively. Thus, minimal systemic exposure is expected. Trifluridine (Viroptic) also has no eye human data. As reported in the product information, detectable blood concentrations of the drug were not found in healthy normal subjects indicating that systemic absorption was negligible.

Corticosteroids

Among the nine corticosteroid products, six are suspensions or ointments, one is an injection, and two are implants. In most nonpregnant patients receiving the dexamethasone (Ozurdex) intravitreal implant, plasma dexamethasone concentrations were undetectable (less than 50 pg/mL) but, in some, ranged from 52 pg/mL to 102 pg/mL. There is only one case report describing the use of topical dexamethasone suspension (Maxidex) in pregnancy. In that case, dexamethasone and clindamycin were given in the first and second trimesters and the woman eventually gave birth to a normal full term infant (Int Ophthalmol. 1998-1999;22[2]:85-8).

For the fluocinolone (Retisert) ocular implant, systemic absorption of detectable amounts of the drug have not been observed. In a second report, a patient who had type 1 diabetes and was 6 months pregnant received a 2-mg intravitreal injection of triamcinolone (Triesence) in both eyes. No adverse effects in the mother or fetus were noted (Clin Ophthalmol. 2011;5:439-41).

There are no human eye data for five topical corticosteroids: difluprednate (Durezol), fluorometholone (Flarex, Fluor-OP, FML), loteprednol (Alrex, Lotemax), prednisolone (Econopred), and rimexolone (Vexol). Only two of the five drugs had information about systemic absorption. For difluprednate, blood levels were below the quantification limit (50 ng/mL). Extremely low levels were detected after the use of rimexolone with a mean serum concentration of 130 pg/mL (range less than 80-470 pg/mL).

Cycloplegics-mydriatics

This class includes four anticholinergic agents: atropine, cyclopentolate (Cyclogyl), homatropine, and tropicamide (Mydriacyl). Systemic absorption has not been studied for these drugs and there are no reports of their use in pregnancy or lactation.

Cystine-depleting agents

There are no reports describing the use of cysteamine (Cystaran) in human pregnancy or lactation. Since the drug is given as one drop in each eye every waking hour, transfer to the systemic circulation should be expected, but the amount, if it occurs, has not been reported.

Immunologics

There are no reports on the use of cyclosporine (Restasis) eye drops in human pregnancy or during breastfeeding. However, data for the systemic use of the drug during these conditions has shown it to be low risk. After long term ophthalmic use, blood concentrations of the drug were below the quantitation limit of 0.1 ng/mL.

Local anesthetics

The three drugs in this class are lidocaine, proparacaine (Alcaine), and tetracaine (Altacaine, Tetravisc). There is no information regarding the use of these agents in pregnancy or lactation. Since they are used for brief periods, the risk to an embryo or nursing infant appears to be nil.

Mast cell stabilizers

There are three drugs in this class that can be used topically in the eye: cromolyn, lodoxamide (Alomide), and nedocromil (Alocril). There are no human eye data for these agents. Systemic concentrations of lodoxamide were below the detection limit of 2.5 ng/mL.

Miotics

The two miotics are acetylcholine (Miochol E) and carbachol (Carbastat, Miostat). Both are used immediately before eye surgery. The amount, if any, in the systemic circulation is unknown. No reports of human eye use in pregnancy or lactation have been found.

NSAIDs

The five drugs in this class are bromfenac (Xibrom), diclofenac (Voltaren), flurbiprofen (Ocufen), ketorolac, and nepafenac (Nevanac). Although it has not been studied, the estimated plasma level for bromfenac is less than 50 ng/mL. For diclofenac, the plasma concentration was below the limit of quantification (10 ng/mL). In a study conducted with ketorolac, only 5 of 26 subjects had detectable plasma concentrations (10.7 to 22.5 ng/mL). These levels were about 2% of the peak plasma levels after oral dosing. For nepafenac, the peak plasma concentrations of the parent drug and active metabolite were 0.3 and 0.4 ng/mL, respectively. These data suggest that the risk to the embryo-fetus or a nursing infant are nil.

Photodynamic therapy

Verteporfin (Visudyne) is given intravenously. Three reports have described its use in three pregnancies. In all cases, the infants were healthy at birth and had normal growth (Acta Ophthalmol Scand. 2004 Oct;82[5]:623-4, Eye [Lond]. 2009 Jun;23[6]:1479, and Aust N Z J Obstet Gynaecol. 2009 Apr;49[2]:236-7).

Proteolytic enzymes

No human eye data were found for ocriplasmin (Jetrea), an agent given as an intravitreal injection. Although it was not studied, detectable levels of the drug in the systemic circulation are not expected, according to the manufacturer.

Selective vascular endothelial growth factor antagonists

There are three agents in this class: aflibercept (Eylea), pegaptanib (Macugen), and ranibizumab (Lucentis). There are no human eye data for these drugs. All are given as intravitreal injection. The systemic concentrations of the three drugs were a mean 0.02 mcg/mL; 80 ng/mL (after a dose of 10 times the recommended dose); and an estimated minimum 0.22 ng/mL, respectively. These data suggest that the risk to the embryo-fetus or a nursing infant is nil.

Sympathomimetics (decongestants)

There are four ophthalmic agents in this class and none have human eye data. Naphazoline (Naphcon, Vasocon) is the only one of the four that requires a prescription. No reports describing the systemic absorption, if any, have been found. The other three agents are available over the counter. They are oxymetazoline (Visine L.R.), phenylephrine, and tetrahydrozoline (Opti-Clear). Although the amount reaching the systemic circulation was not provided by the manufacturers for these three agents, a caution was placed on phenylephrine stating that systemic absorption, although rare, might cause alpha-adrenergic effects, such as a rise in blood pressure and reflex atropine-sensitive bradycardia.

The risk to the embryo-fetus or nursing infant from ophthalmic drugs appears to be low, with the possible exception of phenylephrine. Nevertheless, if any of these agents are used in pregnancy or during breastfeeding, careful assessment of the embryo-fetus and nursing infant should be conducted. Moreover, research on the potential effects of ophthalmic drugs on the embryo-fetus and nursing infant are desperately needed.

Mr. Briggs is clinical professor of pharmacy at the University of California, San Francisco; and adjunct professor of pharmacy at the University of Southern California, Los Angeles, and Washington State University, Spokane. He is coauthor of “Drugs in Pregnancy and Lactation,” and coeditor of “Diseases, Complications, and Drug Therapy in Obstetrics.” He has no relevant financial disclosures.

A number of drugs are available for ophthalmic use. This review focuses on single drug products and, although combination drug products are not discussed, these formulations typically include the drugs reviewed here.

Since it appears that ophthalmic medications are commonly used for a wide range of conditions and ages, one would expect to see numerous reports of their use in the eyes of pregnant or breastfeeding patients. Unfortunately, the opposite is true. The majority of the drugs have no human pregnancy or lactation data. When there are human data, it invariably involves the systemic use of the drug for other indications, rather than its ophthalmic use. Moreover, the animal reproduction data are usually not relevant because they involve systemic drugs (e.g., IV or oral). Consequently, determining the level of risk an ophthalmic drug presents to an embryo and/or fetus is primarily based on time and dose, the two cardinal principles of teratology.

Avoiding exposure during organogenesis – the period when a drug can cause developmental toxicity (altered growth, structural anomalies, functional and/or behavioral deficits, or death) – is usually best, but may not be possible in some cases, including glaucoma, eye infections, and eye surgery. Fortunately, the systemic concentrations of drugs applied topically to the eye are typically thought to be low, even though the levels of most drugs have not been studied. Thus, the risk to the embryo and/or fetus in most cases can be considered low and the drug classified as compatible in pregnancy and breastfeeding.

If a topical eye drug must be used during pregnancy or lactation, teach the patient how to decrease the amount of drug reaching the systemic circulation. This involves placing pressure over the tear duct in the corner of the eye for 1 minute or more, then removing any excess solution with absorbent tissue.

In the sections below, drugs are shown by indication or by pharmacologic class. The term “human eye data” refers to the use of the drug in pregnancy and/or lactation.

Glaucoma

If you are caring for a pregnant patient who is being treated for glaucoma, two recent reviews may be helpful: Surv Ophthalmol. 2011 Jul-Aug;56(4):324-35 and Curr Opin Ophthalmol. 2014 Mar;25(2):93-7.

Sympathomimetics (alpha-adrenergic agonists) include apraclonidine (Iopidine), which has no human eye data, and brimonidine (Alphagan P), which has one case report in pregnancy and breastfeeding showing no fetal or nursing infant harm.

Four of the five beta-adrenergic blockers have no human eye data: betaxolol (Betoptic), carteolol (Ocupress), levobunolol (Betagan), and metipranolol, but there are two case reports for timolol (Betimol, Istalol, Timoptic, Timoptic-XE) showing no fetal harm in one newborn and growth restriction in the other.

Among the miotics, there are limited human eye data for pilocarpine (Isopto Carpine) and no fetal harm was observed. For echothiophate iodide (Phospholine Iodide), there is one case report of a normal full-term infant whose mother was treated with the agent up to 32 weeks’ gestation and then with pilocarpine for 8 weeks (Arch Ophthalmol. 1968 Mar;79[3]:283-5).

Fuse/Thinkstock.com

Carbonic anhydrase inhibitors include brinzolamide (Azopt), which has no human eye data, and dorzolamide (Trusopt), which has growth restriction in one case treated with fixed combination of dorzolamide and timolol. There are no human eye data for unoprostone isopropyl (Rescula), a synthetic docosanoid.

Of the four prostaglandin analogs, three have no human eye data, bimatoprost (Lumigan), tafluprost (Zioptan), and travoprost (Travatan Z). There were 11 pregnancies exposed to latanoprost (Xalatan). The outcomes of these cases were one lost to follow-up, one miscarriage, and nine infants without congenital anomalies (Am J Ophthalmol. 2004 Aug;138[2]:305-6).

Mitomycin (Mitosol) is an antimetabolite that is given topically to the surgical site of glaucoma filtration surgery. No reports of its use in pregnant humans have been located. According to the manufacturer, clinically significant systemic concentrations are not expected.

Antiseptics

Povidone-iodine (Betadine) is indicated for prepping of the periocular region. There does not appear to be any risk to the embryo-fetus or nursing infant from this indication.

Antihistamines

The four ophthalmic agents in this class are alcaftadine (Lastacaft), azelastine (Optivar), emedastine (Emadine), and epinastine (Elestat). There are no human eye data for these agents but, like antihistamines given systemically, they are probably compatible in pregnancy and lactation.

Antihistamine-mast cell stabilizers

There are no human eye data for bepotastine (Bepreve), ketotifen (Alaway), and olopatadine (Pataday, Patanol). Peak plasma concentrations of bepotastine were 5.1-7.3 ng/mL for 1-2 hours after instillation and were less than 2 ng/mL at 24 hours. It appears that these drugs can also be classified as compatible in pregnancy and lactation.

Anti-infectives

Ten anti-infectives are available for topical treatment of eye infections (systemic concentrations if known): besifloxacin (Besivance) (0.43 ng/mL), ciprofloxacin (less than 2.5 ng/mL), gentamicin, gatifloxacin (Zymaxid) (less than 5 ng/mL), levofloxacin (Iquix) (10.9 ng/mL), moxifloxacin (Vigamox) (2.7 ng/mL), ofloxacin (1.9 ng/mL), sulfacetamide (Isopto Cetamide), and tobramycin. The tenth agent, natamycin (Natacyn), is an antifungal. According to the manufacturer, systemic absorption is not expected.

None of these agents have human eye data, but they are usually considered compatible in pregnancy and breastfeeding when systemic formulations are used, so they should be compatible when used in the eye.

Antivirals

Ganciclovir (Zirgan) has no human eye data but, according to the manufacturer, the daily ophthalmic dose is about 0.04% and 0.1% of the oral and IV doses, respectively. Thus, minimal systemic exposure is expected. Trifluridine (Viroptic) also has no eye human data. As reported in the product information, detectable blood concentrations of the drug were not found in healthy normal subjects indicating that systemic absorption was negligible.

Corticosteroids

Among the nine corticosteroid products, six are suspensions or ointments, one is an injection, and two are implants. In most nonpregnant patients receiving the dexamethasone (Ozurdex) intravitreal implant, plasma dexamethasone concentrations were undetectable (less than 50 pg/mL) but, in some, ranged from 52 pg/mL to 102 pg/mL. There is only one case report describing the use of topical dexamethasone suspension (Maxidex) in pregnancy. In that case, dexamethasone and clindamycin were given in the first and second trimesters and the woman eventually gave birth to a normal full term infant (Int Ophthalmol. 1998-1999;22[2]:85-8).

For the fluocinolone (Retisert) ocular implant, systemic absorption of detectable amounts of the drug have not been observed. In a second report, a patient who had type 1 diabetes and was 6 months pregnant received a 2-mg intravitreal injection of triamcinolone (Triesence) in both eyes. No adverse effects in the mother or fetus were noted (Clin Ophthalmol. 2011;5:439-41).

There are no human eye data for five topical corticosteroids: difluprednate (Durezol), fluorometholone (Flarex, Fluor-OP, FML), loteprednol (Alrex, Lotemax), prednisolone (Econopred), and rimexolone (Vexol). Only two of the five drugs had information about systemic absorption. For difluprednate, blood levels were below the quantification limit (50 ng/mL). Extremely low levels were detected after the use of rimexolone with a mean serum concentration of 130 pg/mL (range less than 80-470 pg/mL).

Cycloplegics-mydriatics

This class includes four anticholinergic agents: atropine, cyclopentolate (Cyclogyl), homatropine, and tropicamide (Mydriacyl). Systemic absorption has not been studied for these drugs and there are no reports of their use in pregnancy or lactation.

Cystine-depleting agents

There are no reports describing the use of cysteamine (Cystaran) in human pregnancy or lactation. Since the drug is given as one drop in each eye every waking hour, transfer to the systemic circulation should be expected, but the amount, if it occurs, has not been reported.

Immunologics

There are no reports on the use of cyclosporine (Restasis) eye drops in human pregnancy or during breastfeeding. However, data for the systemic use of the drug during these conditions has shown it to be low risk. After long term ophthalmic use, blood concentrations of the drug were below the quantitation limit of 0.1 ng/mL.

Local anesthetics

The three drugs in this class are lidocaine, proparacaine (Alcaine), and tetracaine (Altacaine, Tetravisc). There is no information regarding the use of these agents in pregnancy or lactation. Since they are used for brief periods, the risk to an embryo or nursing infant appears to be nil.

Mast cell stabilizers

There are three drugs in this class that can be used topically in the eye: cromolyn, lodoxamide (Alomide), and nedocromil (Alocril). There are no human eye data for these agents. Systemic concentrations of lodoxamide were below the detection limit of 2.5 ng/mL.

Miotics

The two miotics are acetylcholine (Miochol E) and carbachol (Carbastat, Miostat). Both are used immediately before eye surgery. The amount, if any, in the systemic circulation is unknown. No reports of human eye use in pregnancy or lactation have been found.

NSAIDs

The five drugs in this class are bromfenac (Xibrom), diclofenac (Voltaren), flurbiprofen (Ocufen), ketorolac, and nepafenac (Nevanac). Although it has not been studied, the estimated plasma level for bromfenac is less than 50 ng/mL. For diclofenac, the plasma concentration was below the limit of quantification (10 ng/mL). In a study conducted with ketorolac, only 5 of 26 subjects had detectable plasma concentrations (10.7 to 22.5 ng/mL). These levels were about 2% of the peak plasma levels after oral dosing. For nepafenac, the peak plasma concentrations of the parent drug and active metabolite were 0.3 and 0.4 ng/mL, respectively. These data suggest that the risk to the embryo-fetus or a nursing infant are nil.

Photodynamic therapy

Verteporfin (Visudyne) is given intravenously. Three reports have described its use in three pregnancies. In all cases, the infants were healthy at birth and had normal growth (Acta Ophthalmol Scand. 2004 Oct;82[5]:623-4, Eye [Lond]. 2009 Jun;23[6]:1479, and Aust N Z J Obstet Gynaecol. 2009 Apr;49[2]:236-7).

Proteolytic enzymes

No human eye data were found for ocriplasmin (Jetrea), an agent given as an intravitreal injection. Although it was not studied, detectable levels of the drug in the systemic circulation are not expected, according to the manufacturer.

Selective vascular endothelial growth factor antagonists

There are three agents in this class: aflibercept (Eylea), pegaptanib (Macugen), and ranibizumab (Lucentis). There are no human eye data for these drugs. All are given as intravitreal injection. The systemic concentrations of the three drugs were a mean 0.02 mcg/mL; 80 ng/mL (after a dose of 10 times the recommended dose); and an estimated minimum 0.22 ng/mL, respectively. These data suggest that the risk to the embryo-fetus or a nursing infant is nil.

Sympathomimetics (decongestants)

There are four ophthalmic agents in this class and none have human eye data. Naphazoline (Naphcon, Vasocon) is the only one of the four that requires a prescription. No reports describing the systemic absorption, if any, have been found. The other three agents are available over the counter. They are oxymetazoline (Visine L.R.), phenylephrine, and tetrahydrozoline (Opti-Clear). Although the amount reaching the systemic circulation was not provided by the manufacturers for these three agents, a caution was placed on phenylephrine stating that systemic absorption, although rare, might cause alpha-adrenergic effects, such as a rise in blood pressure and reflex atropine-sensitive bradycardia.

The risk to the embryo-fetus or nursing infant from ophthalmic drugs appears to be low, with the possible exception of phenylephrine. Nevertheless, if any of these agents are used in pregnancy or during breastfeeding, careful assessment of the embryo-fetus and nursing infant should be conducted. Moreover, research on the potential effects of ophthalmic drugs on the embryo-fetus and nursing infant are desperately needed.

Mr. Briggs is clinical professor of pharmacy at the University of California, San Francisco; and adjunct professor of pharmacy at the University of Southern California, Los Angeles, and Washington State University, Spokane. He is coauthor of “Drugs in Pregnancy and Lactation,” and coeditor of “Diseases, Complications, and Drug Therapy in Obstetrics.” He has no relevant financial disclosures.

A number of drugs are available for ophthalmic use. This review focuses on single drug products and, although combination drug products are not discussed, these formulations typically include the drugs reviewed here.

Since it appears that ophthalmic medications are commonly used for a wide range of conditions and ages, one would expect to see numerous reports of their use in the eyes of pregnant or breastfeeding patients. Unfortunately, the opposite is true. The majority of the drugs have no human pregnancy or lactation data. When there are human data, it invariably involves the systemic use of the drug for other indications, rather than its ophthalmic use. Moreover, the animal reproduction data are usually not relevant because they involve systemic drugs (e.g., IV or oral). Consequently, determining the level of risk an ophthalmic drug presents to an embryo and/or fetus is primarily based on time and dose, the two cardinal principles of teratology.

Avoiding exposure during organogenesis – the period when a drug can cause developmental toxicity (altered growth, structural anomalies, functional and/or behavioral deficits, or death) – is usually best, but may not be possible in some cases, including glaucoma, eye infections, and eye surgery. Fortunately, the systemic concentrations of drugs applied topically to the eye are typically thought to be low, even though the levels of most drugs have not been studied. Thus, the risk to the embryo and/or fetus in most cases can be considered low and the drug classified as compatible in pregnancy and breastfeeding.

If a topical eye drug must be used during pregnancy or lactation, teach the patient how to decrease the amount of drug reaching the systemic circulation. This involves placing pressure over the tear duct in the corner of the eye for 1 minute or more, then removing any excess solution with absorbent tissue.

In the sections below, drugs are shown by indication or by pharmacologic class. The term “human eye data” refers to the use of the drug in pregnancy and/or lactation.

Glaucoma

If you are caring for a pregnant patient who is being treated for glaucoma, two recent reviews may be helpful: Surv Ophthalmol. 2011 Jul-Aug;56(4):324-35 and Curr Opin Ophthalmol. 2014 Mar;25(2):93-7.

Sympathomimetics (alpha-adrenergic agonists) include apraclonidine (Iopidine), which has no human eye data, and brimonidine (Alphagan P), which has one case report in pregnancy and breastfeeding showing no fetal or nursing infant harm.

Four of the five beta-adrenergic blockers have no human eye data: betaxolol (Betoptic), carteolol (Ocupress), levobunolol (Betagan), and metipranolol, but there are two case reports for timolol (Betimol, Istalol, Timoptic, Timoptic-XE) showing no fetal harm in one newborn and growth restriction in the other.

Among the miotics, there are limited human eye data for pilocarpine (Isopto Carpine) and no fetal harm was observed. For echothiophate iodide (Phospholine Iodide), there is one case report of a normal full-term infant whose mother was treated with the agent up to 32 weeks’ gestation and then with pilocarpine for 8 weeks (Arch Ophthalmol. 1968 Mar;79[3]:283-5).

Fuse/Thinkstock.com

Carbonic anhydrase inhibitors include brinzolamide (Azopt), which has no human eye data, and dorzolamide (Trusopt), which has growth restriction in one case treated with fixed combination of dorzolamide and timolol. There are no human eye data for unoprostone isopropyl (Rescula), a synthetic docosanoid.

Of the four prostaglandin analogs, three have no human eye data, bimatoprost (Lumigan), tafluprost (Zioptan), and travoprost (Travatan Z). There were 11 pregnancies exposed to latanoprost (Xalatan). The outcomes of these cases were one lost to follow-up, one miscarriage, and nine infants without congenital anomalies (Am J Ophthalmol. 2004 Aug;138[2]:305-6).

Mitomycin (Mitosol) is an antimetabolite that is given topically to the surgical site of glaucoma filtration surgery. No reports of its use in pregnant humans have been located. According to the manufacturer, clinically significant systemic concentrations are not expected.

Antiseptics

Povidone-iodine (Betadine) is indicated for prepping of the periocular region. There does not appear to be any risk to the embryo-fetus or nursing infant from this indication.

Antihistamines

The four ophthalmic agents in this class are alcaftadine (Lastacaft), azelastine (Optivar), emedastine (Emadine), and epinastine (Elestat). There are no human eye data for these agents but, like antihistamines given systemically, they are probably compatible in pregnancy and lactation.

Antihistamine-mast cell stabilizers

There are no human eye data for bepotastine (Bepreve), ketotifen (Alaway), and olopatadine (Pataday, Patanol). Peak plasma concentrations of bepotastine were 5.1-7.3 ng/mL for 1-2 hours after instillation and were less than 2 ng/mL at 24 hours. It appears that these drugs can also be classified as compatible in pregnancy and lactation.

Anti-infectives

Ten anti-infectives are available for topical treatment of eye infections (systemic concentrations if known): besifloxacin (Besivance) (0.43 ng/mL), ciprofloxacin (less than 2.5 ng/mL), gentamicin, gatifloxacin (Zymaxid) (less than 5 ng/mL), levofloxacin (Iquix) (10.9 ng/mL), moxifloxacin (Vigamox) (2.7 ng/mL), ofloxacin (1.9 ng/mL), sulfacetamide (Isopto Cetamide), and tobramycin. The tenth agent, natamycin (Natacyn), is an antifungal. According to the manufacturer, systemic absorption is not expected.

None of these agents have human eye data, but they are usually considered compatible in pregnancy and breastfeeding when systemic formulations are used, so they should be compatible when used in the eye.

Antivirals

Ganciclovir (Zirgan) has no human eye data but, according to the manufacturer, the daily ophthalmic dose is about 0.04% and 0.1% of the oral and IV doses, respectively. Thus, minimal systemic exposure is expected. Trifluridine (Viroptic) also has no eye human data. As reported in the product information, detectable blood concentrations of the drug were not found in healthy normal subjects indicating that systemic absorption was negligible.

Corticosteroids

Among the nine corticosteroid products, six are suspensions or ointments, one is an injection, and two are implants. In most nonpregnant patients receiving the dexamethasone (Ozurdex) intravitreal implant, plasma dexamethasone concentrations were undetectable (less than 50 pg/mL) but, in some, ranged from 52 pg/mL to 102 pg/mL. There is only one case report describing the use of topical dexamethasone suspension (Maxidex) in pregnancy. In that case, dexamethasone and clindamycin were given in the first and second trimesters and the woman eventually gave birth to a normal full term infant (Int Ophthalmol. 1998-1999;22[2]:85-8).

For the fluocinolone (Retisert) ocular implant, systemic absorption of detectable amounts of the drug have not been observed. In a second report, a patient who had type 1 diabetes and was 6 months pregnant received a 2-mg intravitreal injection of triamcinolone (Triesence) in both eyes. No adverse effects in the mother or fetus were noted (Clin Ophthalmol. 2011;5:439-41).

There are no human eye data for five topical corticosteroids: difluprednate (Durezol), fluorometholone (Flarex, Fluor-OP, FML), loteprednol (Alrex, Lotemax), prednisolone (Econopred), and rimexolone (Vexol). Only two of the five drugs had information about systemic absorption. For difluprednate, blood levels were below the quantification limit (50 ng/mL). Extremely low levels were detected after the use of rimexolone with a mean serum concentration of 130 pg/mL (range less than 80-470 pg/mL).

Cycloplegics-mydriatics

This class includes four anticholinergic agents: atropine, cyclopentolate (Cyclogyl), homatropine, and tropicamide (Mydriacyl). Systemic absorption has not been studied for these drugs and there are no reports of their use in pregnancy or lactation.

Cystine-depleting agents

There are no reports describing the use of cysteamine (Cystaran) in human pregnancy or lactation. Since the drug is given as one drop in each eye every waking hour, transfer to the systemic circulation should be expected, but the amount, if it occurs, has not been reported.

Immunologics

There are no reports on the use of cyclosporine (Restasis) eye drops in human pregnancy or during breastfeeding. However, data for the systemic use of the drug during these conditions has shown it to be low risk. After long term ophthalmic use, blood concentrations of the drug were below the quantitation limit of 0.1 ng/mL.

Local anesthetics

The three drugs in this class are lidocaine, proparacaine (Alcaine), and tetracaine (Altacaine, Tetravisc). There is no information regarding the use of these agents in pregnancy or lactation. Since they are used for brief periods, the risk to an embryo or nursing infant appears to be nil.

Mast cell stabilizers

There are three drugs in this class that can be used topically in the eye: cromolyn, lodoxamide (Alomide), and nedocromil (Alocril). There are no human eye data for these agents. Systemic concentrations of lodoxamide were below the detection limit of 2.5 ng/mL.

Miotics

The two miotics are acetylcholine (Miochol E) and carbachol (Carbastat, Miostat). Both are used immediately before eye surgery. The amount, if any, in the systemic circulation is unknown. No reports of human eye use in pregnancy or lactation have been found.

NSAIDs

The five drugs in this class are bromfenac (Xibrom), diclofenac (Voltaren), flurbiprofen (Ocufen), ketorolac, and nepafenac (Nevanac). Although it has not been studied, the estimated plasma level for bromfenac is less than 50 ng/mL. For diclofenac, the plasma concentration was below the limit of quantification (10 ng/mL). In a study conducted with ketorolac, only 5 of 26 subjects had detectable plasma concentrations (10.7 to 22.5 ng/mL). These levels were about 2% of the peak plasma levels after oral dosing. For nepafenac, the peak plasma concentrations of the parent drug and active metabolite were 0.3 and 0.4 ng/mL, respectively. These data suggest that the risk to the embryo-fetus or a nursing infant are nil.

Photodynamic therapy

Verteporfin (Visudyne) is given intravenously. Three reports have described its use in three pregnancies. In all cases, the infants were healthy at birth and had normal growth (Acta Ophthalmol Scand. 2004 Oct;82[5]:623-4, Eye [Lond]. 2009 Jun;23[6]:1479, and Aust N Z J Obstet Gynaecol. 2009 Apr;49[2]:236-7).

Proteolytic enzymes

No human eye data were found for ocriplasmin (Jetrea), an agent given as an intravitreal injection. Although it was not studied, detectable levels of the drug in the systemic circulation are not expected, according to the manufacturer.

Selective vascular endothelial growth factor antagonists

There are three agents in this class: aflibercept (Eylea), pegaptanib (Macugen), and ranibizumab (Lucentis). There are no human eye data for these drugs. All are given as intravitreal injection. The systemic concentrations of the three drugs were a mean 0.02 mcg/mL; 80 ng/mL (after a dose of 10 times the recommended dose); and an estimated minimum 0.22 ng/mL, respectively. These data suggest that the risk to the embryo-fetus or a nursing infant is nil.

Sympathomimetics (decongestants)

There are four ophthalmic agents in this class and none have human eye data. Naphazoline (Naphcon, Vasocon) is the only one of the four that requires a prescription. No reports describing the systemic absorption, if any, have been found. The other three agents are available over the counter. They are oxymetazoline (Visine L.R.), phenylephrine, and tetrahydrozoline (Opti-Clear). Although the amount reaching the systemic circulation was not provided by the manufacturers for these three agents, a caution was placed on phenylephrine stating that systemic absorption, although rare, might cause alpha-adrenergic effects, such as a rise in blood pressure and reflex atropine-sensitive bradycardia.

The risk to the embryo-fetus or nursing infant from ophthalmic drugs appears to be low, with the possible exception of phenylephrine. Nevertheless, if any of these agents are used in pregnancy or during breastfeeding, careful assessment of the embryo-fetus and nursing infant should be conducted. Moreover, research on the potential effects of ophthalmic drugs on the embryo-fetus and nursing infant are desperately needed.

Mr. Briggs is clinical professor of pharmacy at the University of California, San Francisco; and adjunct professor of pharmacy at the University of Southern California, Los Angeles, and Washington State University, Spokane. He is coauthor of “Drugs in Pregnancy and Lactation,” and coeditor of “Diseases, Complications, and Drug Therapy in Obstetrics.” He has no relevant financial disclosures.