Cleveland Clinic Journal of Medicine

Randomized trials that included more than 20,000 medical patients have shown that anticoagulant therapy is safe and effective in preventing venous thromboembolism (VTE), ie, deep vein thrombosis and pulmonary embolism.

However, these trials were done in hospitalized patients, who typically had an acute medical illness and who, if eligible, received a short (7- to 10-day) course of anticoagulant prophylaxis.

Little attention has been given to VTE prophylaxis in residents of long-term care facilities. These patients have risk profiles similar to those of hospitalized medical patients. Some of them may have been transferred from an acute care hospital. In addition, most are elderly, and many have reduced mobility and are at risk for illnesses such as stroke and cardiorespiratory insufficiency, which increase the risk of VTE.

VTE in residents of long-term care facilities is a growing concern. By some estimates, by the year 2030 more than 20% of the US population (70.2 million people) will be over 65 years of age.¹ Of those who reached age 65 in 1990, an estimated 43% will enter a nursing home at least once before they die—32% for 3 months, 24% for at least a year, and 9% for at least 5 years.²

Against this background, the objectives of this review are to consider:

• The scope of the problem of VTE in long-term care residents
• Why VTE prophylaxis is often overlooked in medical patients
• Evidence—or lack of evidence—for the safety and efficacy of VTE prophylaxis in long-term care residents and other medical patients
• Available options for VTE prophylaxis
• Which long-term care residents should or should not be considered for prophylaxis.

THE TRUE SCOPE OF THE PROBLEM IS UNKNOWN

The incidence of acute VTE among nursing home residents is reported to be 1.3 events per 100 person-years.³ About 8% of cases of pulmonary embolism and 10% of cases of deep venous thrombosis in the elderly are in nursing home residents.⁴

However, only 20% of patients with VTE have typical symptoms such as leg pain and swelling or acute dyspnea and chest pain, while 80% have no symptoms.⁵

Furthermore, deep venous thrombosis is more likely to be clinically silent in patients whose mobility is impaired, such as nursing home residents, as the symptoms arising from obstruction of venous flow are more pronounced with walking.

Pulmonary embolism is also underdiagnosed in this group. An autopsy study of 234 nursing home residents found undiagnosed pulmonary embolism to be the cause of death in 8%, and 40% of cases of pulmonary embolism were not suspected before the patient died.⁶ Yet pulmonary embolism has a higher case-fatality rate in the elderly than in younger patients, particularly when elderly patients have comorbidities.⁷

A reason why the diagnosis is so often missed is that pulmonary embolism can present atypically in the elderly, with syncope being more common and tachycardia being less common than in younger patients.⁸

Since so many cases of VTE are clinically silent and most long-term care residents who die do not undergo autopsy, the true scope of VTE as a clinical problem in these patients is unknown. Consequently, the best way to diagnose, prevent, and treat VTE is also unclear.

WHY IS VTE PREVENTION SO OFTEN OVERLOOKED IN MEDICAL PATIENTS?

In general, nonsurgical patients receive suboptimal thromboprophylaxis. National and international chart audits and cross-sectional studies show that only 16% to 33% of hospitalized medical patients at risk for VTE receive appropriate anticoagulant prophylaxis.⁹ Though no audits in long-term care facilities have been published, the rate of appropriate prophylaxis is likely comparable to or possibly less than that in medical patients in the hospital. In contrast, in surgical patients the rate is much higher—up to 90%.^10,11

Why is VTE prophylaxis so underused in medical patients?

One reason is that we do not really know the baseline risk of VTE in medical patients, particularly in those with chronic illness who require long-term care.¹² This is relevant because, in the absence of data about patients’ baseline risk, anticoagulant prophylaxis should be ordered selectively, as it poses known risks of bleeding. The risk is greater in elderly people with comorbidities, as are the associated costs.

In addition, relatively few studies have assessed thromboprophylaxis in medical patients, especially in residents of long-term care facilities.

Another reason is that we lack practice guidelines for patients who need long-term care. The well-accepted guidelines from the American College of Chest Physicians (ACCP) cite advanced age and immobility as risk factors for VTE and strongly recommend prophylaxis in acutely ill medical patients who have limited mobility and an additional risk factor such as infection or cancer.¹³ Though elderly residents of long-term care facilities may share some of these risk factors, the ACCP guidelines make no specific recommendations for this group.

The attitudes of health care professionals may also pose a barrier. Lloyd et al (unpublished data, 2009) surveyed 1,601 health care professionals in Ontario, Canada, in 2007, to assess potential barriers to anticoagulant prophylaxis in hospitalized medical patients. Respondents cited concerns about the risk of bleeding from anticoagulants, lack of clear indications and contraindications for anticoagulant prophylaxis, and lack of time to consider VTE prophylaxis in every patient. (They did not, however, cite disagreement with guidelines or patient discomfort from subcutaneous anticoagulant injections as barriers.) It is reasonable to assume that these attitudes may also pose a problem in long-term care residents.

Finally, no randomized trials have evaluated the efficacy and safety of anticoagulant drugs or mechanical methods of prophylaxis in long-term care residents. Studies have shown that a short course (7–10 days) of an anticoagulant drug effectively prevents VTE in acutely ill patients, but the efficacy of an extended course in patients with chronic illness who require long-term care is not clear. Therefore, recommendations about thromboprophylaxis in long-term care residents should be made with the caveat that they are based on indirect evidence from other patient groups. This is a considerable limitation.

OPTIONS FOR THROMBOPROPHYLAXIS IN LONG-TERM CARE RESIDENTS

Options for thromboprophylaxis fall into two broad categories: anticoagulant drugs and mechanical devices.

Anticoagulant prophylactic drugs

These agents have been assessed in randomized trials in surgical or acutely ill medical patients, although fondaparinux and tinzaparin are not approved for use in medical patients. Furthermore, none of them has been evaluated in residents of long-term care facilities.

The choice of anticoagulant for prophylaxis is determined largely by clinical factors.

Low-molecular-weight heparins are popular both in and out of the hospital because they have predictable pharmacokinetic properties, they come in convenient prefilled syringes, and they can be given once daily. However, some of them may bioaccumulate in patients with impaired renal function, as they are cleared primarily by the kidney.

Unfractionated heparin is likely to be safer in patients with severe renal insufficiency (creatinine clearance < 30 mL/min), as it is cleared via nonrenal mechanisms.

However, a recent single-arm trial of dalteparin 5,000 IU once daily in critically ill patients with severe renal insufficiency found no evidence of an excessive anticoagulant effect or of drug bioaccumulation.¹⁵ Dalteparin may thus be an alternative to unfractionated heparin in medical patients with impaired renal function.

Fondaparinux, a newer anticoagulant, is also given once daily. It is the anticoagulant of choice in patients who have had heparin-induced thrombocytopenia because it is not derived from heparin and likely does not cross-react with heparin-induced thrombocytopenia antibodies.^16,17

Limited data on benefit of prophylactic anticoagulant drugs

As mentioned, the trials that confirmed the efficacy and safety of anticoagulant prophylaxis were in surgical patients and hospitalized medical patients, not elderly long-term care residents. The poor evidence for anticoagulant prophylaxis in these patients may be strengthened if extended-duration, out-of-hospital prophylaxis were shown to be effective in medical patients. Long-term care residents could more reasonably be compared with medical patients discharged home with a chronic or resolving illness than with those who are hospitalized.

There is some evidence, although with caveats, that extended anticoagulant prophylaxis, started after an acute illness has resolved, confers a benefit. A recent randomized trial compared extended-duration and short-duration prophylaxis (5 weeks vs 10 days) with enoxaparin 40 mg once daily in 4,726 medical patients with impaired mobility.¹⁸ The risk of any VTE event was 44% lower with extended-duration prophylaxis (2.8% vs 4.9%; P = .001) and the risk of symptomatic VTE was 73% lower (0.3% vs 1.1%; P = .004), and this benefit persisted 2 months after treatment was stopped (3.0% vs 5.2%; P = .0015). However, extended treatment conferred a fourfold higher risk of major bleeding (0.6% vs 0.15%; P = .019).

These findings should also be considered in terms of absolute benefit and harm. Treating 1,000 patients for 5 weeks instead of 10 days would prevent eight episodes of symptomatic VTE (absolute risk reduction = 0.8%, number needed to treat = 125) at the cost of four to five episodes of major bleeding (absolute risk increase = 0.45%, number needed to harm = 222). This is a modest net therapeutic benefit.

The therapeutic benefit would be greater if we consider all episodes of VTE, both symptomatic and asymptomatic. Treating 1,000 patients for 5 weeks would prevent 20 episodes of symptomatic or asymptomatic VTE (absolute risk reduction = 2.1%, number needed to treat = 48). However, the clinical importance of asymptomatic VTE is questionable.

Given these considerations, if extended-duration anticoagulant prophylaxis is considered, it should be for patients at highest risk to optimize both its net therapeutic benefits and its cost-effectiveness.

Mechanical prophylaxis

Mechanical thromboprophylactic devices—graduated or elastic compression stockings and intermittent pneumatic compression devices—are effective when used by themselves in surgical patients.¹³ However, in a randomized controlled trial in patients with ischemic stroke, the rate of VTE was 10.0% with graduated compression stockings in addition to “usual care VTE prophylaxis” vs 10.5% with usual care alone, and patients in the stocking group had a fourfold higher risk of developing skin breaks, ulcers, blisters, or necrosis (5% vs 1%; odds ratio 4.18; 95% CI 2.4–7.3).¹⁹ Furthermore, improperly fitted stockings, especially those that are thigh-length, can be uncomfortable to wear and difficult to apply.

Overall, the role of mechanical thromboprophylaxis in long-term care facilities is not clear. If it is considered, there should be a compelling reason to use it—for example, for patients at high risk in whom anticoagulants are contraindicated because of ongoing bleeding or a higher risk of bleeding (eg, recent gastrointestinal bleeding, hemorrhagic stroke, coagulopathy, or thrombocytopenia). Furthermore, if stockings are used, they should be properly fitted and routinely monitored for adverse effects, since elderly patients are likely to be most susceptible to skin breakdown.

WHICH LONG-TERM CARE RESIDENTS SHOULD RECEIVE VTE PROPHYLAXIS?

Old age and immobility are not the only risk factors

The current ACCP guidelines suggest considering thromboprophylaxis for hospitalized medical patients over age 75 who cannot walk without assistance.¹³ However, we lack evidence to suggest a similar strategy in long-term care residents.

The ACCP guidelines are based on data on risk. Nearly 25% of elderly patients with confirmed pulmonary embolism had been immobile prior to their diagnosis.⁸ In addition, prolonged bed rest (> 14 days) has been reported to be the strongest independent risk factor for symptomatic deep venous thrombosis, increasing the risk more than fivefold.²⁰ Advanced age is also considered a risk factor for VTE, as risk starts to increase at age 40 and doubles each decade of life thereafter.¹⁸

No study has assessed the impact of these factors on the risk of VTE in long-term care residents. Since most of such patients are elderly and have impaired mobility, we believe a more selective approach should be used in assigning VTE risk status, one that does not use advanced age and immobility as the only criteria for starting thromboprophylaxis.

Residents of long-term care facilities may be immobile because of underlying illness or disability, such as cognitive impairment, sensory impairment (eg, poor access to corrective lenses and hearing aids), or poor access to assist devices (eg, walkers, canes). In addition, iatrogenic factors that decrease mobility such as indwelling bladder catheters and physical restraints are also common in such patients.

Efforts to improve mobility should be encouraged. However, we recommend that thromboprophylaxis be considered only in patients who have both impaired mobility and an intercurrent acute medical illness such as an acute infection or acute inflammatory disease.¹³

A related issue is the difference between long-term care residents with a chronic but stable disease and those with acute disease. Patients with acute exacerbations of congestive heart failure or chronic obstructive lung disease may be considered for thromboprophylaxis, as they become more comparable to acutely ill medical patients in whom clinical trials have shown the effectiveness of anticoagulant prophylaxis. On the other hand, patients with these diseases who remain stable may not need prophylaxis.

This approach avoids giving long-term anticoagulant prophylaxis to patients who have irreversible diseases and limits the use of these drugs and devices to higher-risk periods.

Consider thromboprophylaxis if…

• An acute exacerbation of congestive heart failure or chronic obstructive pulmonary disease
• Acute infection (eg, urosepsis, pneumonia, cellulitis, infectious diarrhea)
• An acute exacerbation of an inflammatory disease (eg, rheumatoid arthritis)
• Active cancer (eg, patient receiving radiation therapy or chemotherapy)
• Immobility and prior VTE.

Do not routinely consider prophylaxis if…

We also believe patients should not be routinely considered for anticoagulant VTE prophylaxis if they have:

• Chronic but stable cardiorespiratory disease
• Chronic but stable infectious or inflammatory disease
• Terminal cancer with very limited life expectancy
• Any contraindication to anticoagulants (eg, active bleeding, recent bleeding, coagulopathy, thrombocytopenia).

ANTICOAGULANT PROPHYLAXIS POSES RISKS IN LONG-TERM CARE RESIDENTS

Bleeding is the principal risk

The incidence of clinically important bleeding associated with anticoagulant prophylaxis is 0.2% to 5.6%, and the risk of fatal bleeding is 0.02% to 0.5%.^21–24

As no randomized trial has examined anticoagulant prophylaxis in elderly long-term care residents, their bleeding risk with this therapy is unclear. However, older patients are likely to be at higher risk than younger patients because they have more comorbidities, take more drugs that could interact with heparin and potentiate bleeding, and have fragile skin, predisposing to injury from subcutaneous injections.

Also, renal function tends to decline with age. In a retrospective study of 854 outpatients over age 65, 29% had moderate renal insufficiency (creatinine clearance 30–50 mL/min), and 6% had severe renal insufficiency (creatinine clearance < 30 mL/min).²⁵ Recent evidence suggests that some low-molecular-weight heparins (dalteparin and tinzaparin) do not bioaccumulate in patients with impaired renal function. However, enoxaparin and fondaparinux should be used with caution in patients with moderate to severe renal impairment.

Though much attention has recently been paid to increasing anticoagulant doses if the patient is obese, residents of long-term care facilities are more likely to be underweight. Dose adjustment should be considered when a low-molecular-weight heparin or fondaparinux is given to patients weighing less than 50 kg.

The other major risk of anticoagulant prophylaxis is heparin-induced thrombocytopenia, an infrequent but life-threatening complication caused by the formation of antibodies to the heparin-derived anticoagulant and a platelet surface antigen. It is associated with moderate thrombocytopenia and an incidence of venous or arterial thrombosis that is over 50%.²⁶

No study has assessed the incidence of heparin-induced thrombocytopenia in long-term care residents. A meta-analysis reported that the risk with anticoagulant prophylaxis was 1.6% with unfractionated heparin (95% confidence interval [CI] 1.2%–2.1%) and 0.6% with low-molecular-weight heparin (95% CI 0.4%–0.9%), and that this risk increased with the duration of prophylaxis.²⁷ If anticoagulant prophylaxis were given to all long-term care residents for extended durations (eg, for the duration of reduced mobility), the incidence and prevalence of heparin-induced thrombocytopenia would likely become a major concern.

Whenever anticoagulant prophylaxis is considered, the risks of both thrombosis and bleeding should be considered. Patients who are receiving anticoagulant prophylaxis should also be monitored for bleeding and heparin-induced thrombocytopenia. This is particularly true in long-term care residents, in whom the risks and benefits of anticoagulant prophylaxis are extrapolated from data from other populations.

MORE RESEARCH IS NEEDED

To date, we lack audits of thromboprophylaxis, clinical practice guidelines, and clear indications and contraindications for anticoagulant prophylaxis in long-term care residents. In the absence of such data, extrapolating the efficacy and safety of thromboprophylaxis from hospitalized patients to long-term care residents is difficult.

Clearly, additional research is needed to identify which long-term care residents would benefit most from thromboprophylaxis. In the meantime, a selective approach to identifying patients who should be considered for thromboprophylaxis should be adopted.

Randomized trials that included more than 20,000 medical patients have shown that anticoagulant therapy is safe and effective in preventing venous thromboembolism (VTE), ie, deep vein thrombosis and pulmonary embolism.

However, these trials were done in hospitalized patients, who typically had an acute medical illness and who, if eligible, received a short (7- to 10-day) course of anticoagulant prophylaxis.

Little attention has been given to VTE prophylaxis in residents of long-term care facilities. These patients have risk profiles similar to those of hospitalized medical patients. Some of them may have been transferred from an acute care hospital. In addition, most are elderly, and many have reduced mobility and are at risk for illnesses such as stroke and cardiorespiratory insufficiency, which increase the risk of VTE.

VTE in residents of long-term care facilities is a growing concern. By some estimates, by the year 2030 more than 20% of the US population (70.2 million people) will be over 65 years of age.¹ Of those who reached age 65 in 1990, an estimated 43% will enter a nursing home at least once before they die—32% for 3 months, 24% for at least a year, and 9% for at least 5 years.²

Against this background, the objectives of this review are to consider:

• The scope of the problem of VTE in long-term care residents
• Why VTE prophylaxis is often overlooked in medical patients
• Evidence—or lack of evidence—for the safety and efficacy of VTE prophylaxis in long-term care residents and other medical patients
• Available options for VTE prophylaxis
• Which long-term care residents should or should not be considered for prophylaxis.

THE TRUE SCOPE OF THE PROBLEM IS UNKNOWN

The incidence of acute VTE among nursing home residents is reported to be 1.3 events per 100 person-years.³ About 8% of cases of pulmonary embolism and 10% of cases of deep venous thrombosis in the elderly are in nursing home residents.⁴

However, only 20% of patients with VTE have typical symptoms such as leg pain and swelling or acute dyspnea and chest pain, while 80% have no symptoms.⁵

Furthermore, deep venous thrombosis is more likely to be clinically silent in patients whose mobility is impaired, such as nursing home residents, as the symptoms arising from obstruction of venous flow are more pronounced with walking.

Pulmonary embolism is also underdiagnosed in this group. An autopsy study of 234 nursing home residents found undiagnosed pulmonary embolism to be the cause of death in 8%, and 40% of cases of pulmonary embolism were not suspected before the patient died.⁶ Yet pulmonary embolism has a higher case-fatality rate in the elderly than in younger patients, particularly when elderly patients have comorbidities.⁷

A reason why the diagnosis is so often missed is that pulmonary embolism can present atypically in the elderly, with syncope being more common and tachycardia being less common than in younger patients.⁸

Since so many cases of VTE are clinically silent and most long-term care residents who die do not undergo autopsy, the true scope of VTE as a clinical problem in these patients is unknown. Consequently, the best way to diagnose, prevent, and treat VTE is also unclear.

WHY IS VTE PREVENTION SO OFTEN OVERLOOKED IN MEDICAL PATIENTS?

In general, nonsurgical patients receive suboptimal thromboprophylaxis. National and international chart audits and cross-sectional studies show that only 16% to 33% of hospitalized medical patients at risk for VTE receive appropriate anticoagulant prophylaxis.⁹ Though no audits in long-term care facilities have been published, the rate of appropriate prophylaxis is likely comparable to or possibly less than that in medical patients in the hospital. In contrast, in surgical patients the rate is much higher—up to 90%.^10,11

Why is VTE prophylaxis so underused in medical patients?

One reason is that we do not really know the baseline risk of VTE in medical patients, particularly in those with chronic illness who require long-term care.¹² This is relevant because, in the absence of data about patients’ baseline risk, anticoagulant prophylaxis should be ordered selectively, as it poses known risks of bleeding. The risk is greater in elderly people with comorbidities, as are the associated costs.

In addition, relatively few studies have assessed thromboprophylaxis in medical patients, especially in residents of long-term care facilities.

Another reason is that we lack practice guidelines for patients who need long-term care. The well-accepted guidelines from the American College of Chest Physicians (ACCP) cite advanced age and immobility as risk factors for VTE and strongly recommend prophylaxis in acutely ill medical patients who have limited mobility and an additional risk factor such as infection or cancer.¹³ Though elderly residents of long-term care facilities may share some of these risk factors, the ACCP guidelines make no specific recommendations for this group.

The attitudes of health care professionals may also pose a barrier. Lloyd et al (unpublished data, 2009) surveyed 1,601 health care professionals in Ontario, Canada, in 2007, to assess potential barriers to anticoagulant prophylaxis in hospitalized medical patients. Respondents cited concerns about the risk of bleeding from anticoagulants, lack of clear indications and contraindications for anticoagulant prophylaxis, and lack of time to consider VTE prophylaxis in every patient. (They did not, however, cite disagreement with guidelines or patient discomfort from subcutaneous anticoagulant injections as barriers.) It is reasonable to assume that these attitudes may also pose a problem in long-term care residents.

Finally, no randomized trials have evaluated the efficacy and safety of anticoagulant drugs or mechanical methods of prophylaxis in long-term care residents. Studies have shown that a short course (7–10 days) of an anticoagulant drug effectively prevents VTE in acutely ill patients, but the efficacy of an extended course in patients with chronic illness who require long-term care is not clear. Therefore, recommendations about thromboprophylaxis in long-term care residents should be made with the caveat that they are based on indirect evidence from other patient groups. This is a considerable limitation.

OPTIONS FOR THROMBOPROPHYLAXIS IN LONG-TERM CARE RESIDENTS

Options for thromboprophylaxis fall into two broad categories: anticoagulant drugs and mechanical devices.

Anticoagulant prophylactic drugs

These agents have been assessed in randomized trials in surgical or acutely ill medical patients, although fondaparinux and tinzaparin are not approved for use in medical patients. Furthermore, none of them has been evaluated in residents of long-term care facilities.

The choice of anticoagulant for prophylaxis is determined largely by clinical factors.

Low-molecular-weight heparins are popular both in and out of the hospital because they have predictable pharmacokinetic properties, they come in convenient prefilled syringes, and they can be given once daily. However, some of them may bioaccumulate in patients with impaired renal function, as they are cleared primarily by the kidney.

Unfractionated heparin is likely to be safer in patients with severe renal insufficiency (creatinine clearance < 30 mL/min), as it is cleared via nonrenal mechanisms.

However, a recent single-arm trial of dalteparin 5,000 IU once daily in critically ill patients with severe renal insufficiency found no evidence of an excessive anticoagulant effect or of drug bioaccumulation.¹⁵ Dalteparin may thus be an alternative to unfractionated heparin in medical patients with impaired renal function.

Fondaparinux, a newer anticoagulant, is also given once daily. It is the anticoagulant of choice in patients who have had heparin-induced thrombocytopenia because it is not derived from heparin and likely does not cross-react with heparin-induced thrombocytopenia antibodies.^16,17

Limited data on benefit of prophylactic anticoagulant drugs

As mentioned, the trials that confirmed the efficacy and safety of anticoagulant prophylaxis were in surgical patients and hospitalized medical patients, not elderly long-term care residents. The poor evidence for anticoagulant prophylaxis in these patients may be strengthened if extended-duration, out-of-hospital prophylaxis were shown to be effective in medical patients. Long-term care residents could more reasonably be compared with medical patients discharged home with a chronic or resolving illness than with those who are hospitalized.

There is some evidence, although with caveats, that extended anticoagulant prophylaxis, started after an acute illness has resolved, confers a benefit. A recent randomized trial compared extended-duration and short-duration prophylaxis (5 weeks vs 10 days) with enoxaparin 40 mg once daily in 4,726 medical patients with impaired mobility.¹⁸ The risk of any VTE event was 44% lower with extended-duration prophylaxis (2.8% vs 4.9%; P = .001) and the risk of symptomatic VTE was 73% lower (0.3% vs 1.1%; P = .004), and this benefit persisted 2 months after treatment was stopped (3.0% vs 5.2%; P = .0015). However, extended treatment conferred a fourfold higher risk of major bleeding (0.6% vs 0.15%; P = .019).

These findings should also be considered in terms of absolute benefit and harm. Treating 1,000 patients for 5 weeks instead of 10 days would prevent eight episodes of symptomatic VTE (absolute risk reduction = 0.8%, number needed to treat = 125) at the cost of four to five episodes of major bleeding (absolute risk increase = 0.45%, number needed to harm = 222). This is a modest net therapeutic benefit.

The therapeutic benefit would be greater if we consider all episodes of VTE, both symptomatic and asymptomatic. Treating 1,000 patients for 5 weeks would prevent 20 episodes of symptomatic or asymptomatic VTE (absolute risk reduction = 2.1%, number needed to treat = 48). However, the clinical importance of asymptomatic VTE is questionable.

Given these considerations, if extended-duration anticoagulant prophylaxis is considered, it should be for patients at highest risk to optimize both its net therapeutic benefits and its cost-effectiveness.

Mechanical prophylaxis

Mechanical thromboprophylactic devices—graduated or elastic compression stockings and intermittent pneumatic compression devices—are effective when used by themselves in surgical patients.¹³ However, in a randomized controlled trial in patients with ischemic stroke, the rate of VTE was 10.0% with graduated compression stockings in addition to “usual care VTE prophylaxis” vs 10.5% with usual care alone, and patients in the stocking group had a fourfold higher risk of developing skin breaks, ulcers, blisters, or necrosis (5% vs 1%; odds ratio 4.18; 95% CI 2.4–7.3).¹⁹ Furthermore, improperly fitted stockings, especially those that are thigh-length, can be uncomfortable to wear and difficult to apply.

Overall, the role of mechanical thromboprophylaxis in long-term care facilities is not clear. If it is considered, there should be a compelling reason to use it—for example, for patients at high risk in whom anticoagulants are contraindicated because of ongoing bleeding or a higher risk of bleeding (eg, recent gastrointestinal bleeding, hemorrhagic stroke, coagulopathy, or thrombocytopenia). Furthermore, if stockings are used, they should be properly fitted and routinely monitored for adverse effects, since elderly patients are likely to be most susceptible to skin breakdown.

WHICH LONG-TERM CARE RESIDENTS SHOULD RECEIVE VTE PROPHYLAXIS?

Old age and immobility are not the only risk factors

The current ACCP guidelines suggest considering thromboprophylaxis for hospitalized medical patients over age 75 who cannot walk without assistance.¹³ However, we lack evidence to suggest a similar strategy in long-term care residents.

The ACCP guidelines are based on data on risk. Nearly 25% of elderly patients with confirmed pulmonary embolism had been immobile prior to their diagnosis.⁸ In addition, prolonged bed rest (> 14 days) has been reported to be the strongest independent risk factor for symptomatic deep venous thrombosis, increasing the risk more than fivefold.²⁰ Advanced age is also considered a risk factor for VTE, as risk starts to increase at age 40 and doubles each decade of life thereafter.¹⁸

No study has assessed the impact of these factors on the risk of VTE in long-term care residents. Since most of such patients are elderly and have impaired mobility, we believe a more selective approach should be used in assigning VTE risk status, one that does not use advanced age and immobility as the only criteria for starting thromboprophylaxis.

Residents of long-term care facilities may be immobile because of underlying illness or disability, such as cognitive impairment, sensory impairment (eg, poor access to corrective lenses and hearing aids), or poor access to assist devices (eg, walkers, canes). In addition, iatrogenic factors that decrease mobility such as indwelling bladder catheters and physical restraints are also common in such patients.

Efforts to improve mobility should be encouraged. However, we recommend that thromboprophylaxis be considered only in patients who have both impaired mobility and an intercurrent acute medical illness such as an acute infection or acute inflammatory disease.¹³

A related issue is the difference between long-term care residents with a chronic but stable disease and those with acute disease. Patients with acute exacerbations of congestive heart failure or chronic obstructive lung disease may be considered for thromboprophylaxis, as they become more comparable to acutely ill medical patients in whom clinical trials have shown the effectiveness of anticoagulant prophylaxis. On the other hand, patients with these diseases who remain stable may not need prophylaxis.

This approach avoids giving long-term anticoagulant prophylaxis to patients who have irreversible diseases and limits the use of these drugs and devices to higher-risk periods.

Consider thromboprophylaxis if…

• An acute exacerbation of congestive heart failure or chronic obstructive pulmonary disease
• Acute infection (eg, urosepsis, pneumonia, cellulitis, infectious diarrhea)
• An acute exacerbation of an inflammatory disease (eg, rheumatoid arthritis)
• Active cancer (eg, patient receiving radiation therapy or chemotherapy)
• Immobility and prior VTE.

Do not routinely consider prophylaxis if…

We also believe patients should not be routinely considered for anticoagulant VTE prophylaxis if they have:

• Chronic but stable cardiorespiratory disease
• Chronic but stable infectious or inflammatory disease
• Terminal cancer with very limited life expectancy
• Any contraindication to anticoagulants (eg, active bleeding, recent bleeding, coagulopathy, thrombocytopenia).

ANTICOAGULANT PROPHYLAXIS POSES RISKS IN LONG-TERM CARE RESIDENTS

Bleeding is the principal risk

The incidence of clinically important bleeding associated with anticoagulant prophylaxis is 0.2% to 5.6%, and the risk of fatal bleeding is 0.02% to 0.5%.^21–24

As no randomized trial has examined anticoagulant prophylaxis in elderly long-term care residents, their bleeding risk with this therapy is unclear. However, older patients are likely to be at higher risk than younger patients because they have more comorbidities, take more drugs that could interact with heparin and potentiate bleeding, and have fragile skin, predisposing to injury from subcutaneous injections.

Also, renal function tends to decline with age. In a retrospective study of 854 outpatients over age 65, 29% had moderate renal insufficiency (creatinine clearance 30–50 mL/min), and 6% had severe renal insufficiency (creatinine clearance < 30 mL/min).²⁵ Recent evidence suggests that some low-molecular-weight heparins (dalteparin and tinzaparin) do not bioaccumulate in patients with impaired renal function. However, enoxaparin and fondaparinux should be used with caution in patients with moderate to severe renal impairment.

Though much attention has recently been paid to increasing anticoagulant doses if the patient is obese, residents of long-term care facilities are more likely to be underweight. Dose adjustment should be considered when a low-molecular-weight heparin or fondaparinux is given to patients weighing less than 50 kg.

The other major risk of anticoagulant prophylaxis is heparin-induced thrombocytopenia, an infrequent but life-threatening complication caused by the formation of antibodies to the heparin-derived anticoagulant and a platelet surface antigen. It is associated with moderate thrombocytopenia and an incidence of venous or arterial thrombosis that is over 50%.²⁶

No study has assessed the incidence of heparin-induced thrombocytopenia in long-term care residents. A meta-analysis reported that the risk with anticoagulant prophylaxis was 1.6% with unfractionated heparin (95% confidence interval [CI] 1.2%–2.1%) and 0.6% with low-molecular-weight heparin (95% CI 0.4%–0.9%), and that this risk increased with the duration of prophylaxis.²⁷ If anticoagulant prophylaxis were given to all long-term care residents for extended durations (eg, for the duration of reduced mobility), the incidence and prevalence of heparin-induced thrombocytopenia would likely become a major concern.

Whenever anticoagulant prophylaxis is considered, the risks of both thrombosis and bleeding should be considered. Patients who are receiving anticoagulant prophylaxis should also be monitored for bleeding and heparin-induced thrombocytopenia. This is particularly true in long-term care residents, in whom the risks and benefits of anticoagulant prophylaxis are extrapolated from data from other populations.

MORE RESEARCH IS NEEDED

To date, we lack audits of thromboprophylaxis, clinical practice guidelines, and clear indications and contraindications for anticoagulant prophylaxis in long-term care residents. In the absence of such data, extrapolating the efficacy and safety of thromboprophylaxis from hospitalized patients to long-term care residents is difficult.

Clearly, additional research is needed to identify which long-term care residents would benefit most from thromboprophylaxis. In the meantime, a selective approach to identifying patients who should be considered for thromboprophylaxis should be adopted.

Randomized trials that included more than 20,000 medical patients have shown that anticoagulant therapy is safe and effective in preventing venous thromboembolism (VTE), ie, deep vein thrombosis and pulmonary embolism.

However, these trials were done in hospitalized patients, who typically had an acute medical illness and who, if eligible, received a short (7- to 10-day) course of anticoagulant prophylaxis.

Little attention has been given to VTE prophylaxis in residents of long-term care facilities. These patients have risk profiles similar to those of hospitalized medical patients. Some of them may have been transferred from an acute care hospital. In addition, most are elderly, and many have reduced mobility and are at risk for illnesses such as stroke and cardiorespiratory insufficiency, which increase the risk of VTE.

VTE in residents of long-term care facilities is a growing concern. By some estimates, by the year 2030 more than 20% of the US population (70.2 million people) will be over 65 years of age.¹ Of those who reached age 65 in 1990, an estimated 43% will enter a nursing home at least once before they die—32% for 3 months, 24% for at least a year, and 9% for at least 5 years.²

Against this background, the objectives of this review are to consider:

• The scope of the problem of VTE in long-term care residents
• Why VTE prophylaxis is often overlooked in medical patients
• Evidence—or lack of evidence—for the safety and efficacy of VTE prophylaxis in long-term care residents and other medical patients
• Available options for VTE prophylaxis
• Which long-term care residents should or should not be considered for prophylaxis.

THE TRUE SCOPE OF THE PROBLEM IS UNKNOWN

The incidence of acute VTE among nursing home residents is reported to be 1.3 events per 100 person-years.³ About 8% of cases of pulmonary embolism and 10% of cases of deep venous thrombosis in the elderly are in nursing home residents.⁴

However, only 20% of patients with VTE have typical symptoms such as leg pain and swelling or acute dyspnea and chest pain, while 80% have no symptoms.⁵

Furthermore, deep venous thrombosis is more likely to be clinically silent in patients whose mobility is impaired, such as nursing home residents, as the symptoms arising from obstruction of venous flow are more pronounced with walking.

Pulmonary embolism is also underdiagnosed in this group. An autopsy study of 234 nursing home residents found undiagnosed pulmonary embolism to be the cause of death in 8%, and 40% of cases of pulmonary embolism were not suspected before the patient died.⁶ Yet pulmonary embolism has a higher case-fatality rate in the elderly than in younger patients, particularly when elderly patients have comorbidities.⁷

A reason why the diagnosis is so often missed is that pulmonary embolism can present atypically in the elderly, with syncope being more common and tachycardia being less common than in younger patients.⁸

Since so many cases of VTE are clinically silent and most long-term care residents who die do not undergo autopsy, the true scope of VTE as a clinical problem in these patients is unknown. Consequently, the best way to diagnose, prevent, and treat VTE is also unclear.

WHY IS VTE PREVENTION SO OFTEN OVERLOOKED IN MEDICAL PATIENTS?

In general, nonsurgical patients receive suboptimal thromboprophylaxis. National and international chart audits and cross-sectional studies show that only 16% to 33% of hospitalized medical patients at risk for VTE receive appropriate anticoagulant prophylaxis.⁹ Though no audits in long-term care facilities have been published, the rate of appropriate prophylaxis is likely comparable to or possibly less than that in medical patients in the hospital. In contrast, in surgical patients the rate is much higher—up to 90%.^10,11

Why is VTE prophylaxis so underused in medical patients?

One reason is that we do not really know the baseline risk of VTE in medical patients, particularly in those with chronic illness who require long-term care.¹² This is relevant because, in the absence of data about patients’ baseline risk, anticoagulant prophylaxis should be ordered selectively, as it poses known risks of bleeding. The risk is greater in elderly people with comorbidities, as are the associated costs.

In addition, relatively few studies have assessed thromboprophylaxis in medical patients, especially in residents of long-term care facilities.

Another reason is that we lack practice guidelines for patients who need long-term care. The well-accepted guidelines from the American College of Chest Physicians (ACCP) cite advanced age and immobility as risk factors for VTE and strongly recommend prophylaxis in acutely ill medical patients who have limited mobility and an additional risk factor such as infection or cancer.¹³ Though elderly residents of long-term care facilities may share some of these risk factors, the ACCP guidelines make no specific recommendations for this group.

The attitudes of health care professionals may also pose a barrier. Lloyd et al (unpublished data, 2009) surveyed 1,601 health care professionals in Ontario, Canada, in 2007, to assess potential barriers to anticoagulant prophylaxis in hospitalized medical patients. Respondents cited concerns about the risk of bleeding from anticoagulants, lack of clear indications and contraindications for anticoagulant prophylaxis, and lack of time to consider VTE prophylaxis in every patient. (They did not, however, cite disagreement with guidelines or patient discomfort from subcutaneous anticoagulant injections as barriers.) It is reasonable to assume that these attitudes may also pose a problem in long-term care residents.

Finally, no randomized trials have evaluated the efficacy and safety of anticoagulant drugs or mechanical methods of prophylaxis in long-term care residents. Studies have shown that a short course (7–10 days) of an anticoagulant drug effectively prevents VTE in acutely ill patients, but the efficacy of an extended course in patients with chronic illness who require long-term care is not clear. Therefore, recommendations about thromboprophylaxis in long-term care residents should be made with the caveat that they are based on indirect evidence from other patient groups. This is a considerable limitation.

OPTIONS FOR THROMBOPROPHYLAXIS IN LONG-TERM CARE RESIDENTS

Options for thromboprophylaxis fall into two broad categories: anticoagulant drugs and mechanical devices.

Anticoagulant prophylactic drugs

These agents have been assessed in randomized trials in surgical or acutely ill medical patients, although fondaparinux and tinzaparin are not approved for use in medical patients. Furthermore, none of them has been evaluated in residents of long-term care facilities.

The choice of anticoagulant for prophylaxis is determined largely by clinical factors.

Low-molecular-weight heparins are popular both in and out of the hospital because they have predictable pharmacokinetic properties, they come in convenient prefilled syringes, and they can be given once daily. However, some of them may bioaccumulate in patients with impaired renal function, as they are cleared primarily by the kidney.

Unfractionated heparin is likely to be safer in patients with severe renal insufficiency (creatinine clearance < 30 mL/min), as it is cleared via nonrenal mechanisms.

However, a recent single-arm trial of dalteparin 5,000 IU once daily in critically ill patients with severe renal insufficiency found no evidence of an excessive anticoagulant effect or of drug bioaccumulation.¹⁵ Dalteparin may thus be an alternative to unfractionated heparin in medical patients with impaired renal function.

Fondaparinux, a newer anticoagulant, is also given once daily. It is the anticoagulant of choice in patients who have had heparin-induced thrombocytopenia because it is not derived from heparin and likely does not cross-react with heparin-induced thrombocytopenia antibodies.^16,17

Limited data on benefit of prophylactic anticoagulant drugs

As mentioned, the trials that confirmed the efficacy and safety of anticoagulant prophylaxis were in surgical patients and hospitalized medical patients, not elderly long-term care residents. The poor evidence for anticoagulant prophylaxis in these patients may be strengthened if extended-duration, out-of-hospital prophylaxis were shown to be effective in medical patients. Long-term care residents could more reasonably be compared with medical patients discharged home with a chronic or resolving illness than with those who are hospitalized.

There is some evidence, although with caveats, that extended anticoagulant prophylaxis, started after an acute illness has resolved, confers a benefit. A recent randomized trial compared extended-duration and short-duration prophylaxis (5 weeks vs 10 days) with enoxaparin 40 mg once daily in 4,726 medical patients with impaired mobility.¹⁸ The risk of any VTE event was 44% lower with extended-duration prophylaxis (2.8% vs 4.9%; P = .001) and the risk of symptomatic VTE was 73% lower (0.3% vs 1.1%; P = .004), and this benefit persisted 2 months after treatment was stopped (3.0% vs 5.2%; P = .0015). However, extended treatment conferred a fourfold higher risk of major bleeding (0.6% vs 0.15%; P = .019).

These findings should also be considered in terms of absolute benefit and harm. Treating 1,000 patients for 5 weeks instead of 10 days would prevent eight episodes of symptomatic VTE (absolute risk reduction = 0.8%, number needed to treat = 125) at the cost of four to five episodes of major bleeding (absolute risk increase = 0.45%, number needed to harm = 222). This is a modest net therapeutic benefit.

The therapeutic benefit would be greater if we consider all episodes of VTE, both symptomatic and asymptomatic. Treating 1,000 patients for 5 weeks would prevent 20 episodes of symptomatic or asymptomatic VTE (absolute risk reduction = 2.1%, number needed to treat = 48). However, the clinical importance of asymptomatic VTE is questionable.

Given these considerations, if extended-duration anticoagulant prophylaxis is considered, it should be for patients at highest risk to optimize both its net therapeutic benefits and its cost-effectiveness.

Mechanical prophylaxis

Mechanical thromboprophylactic devices—graduated or elastic compression stockings and intermittent pneumatic compression devices—are effective when used by themselves in surgical patients.¹³ However, in a randomized controlled trial in patients with ischemic stroke, the rate of VTE was 10.0% with graduated compression stockings in addition to “usual care VTE prophylaxis” vs 10.5% with usual care alone, and patients in the stocking group had a fourfold higher risk of developing skin breaks, ulcers, blisters, or necrosis (5% vs 1%; odds ratio 4.18; 95% CI 2.4–7.3).¹⁹ Furthermore, improperly fitted stockings, especially those that are thigh-length, can be uncomfortable to wear and difficult to apply.

Overall, the role of mechanical thromboprophylaxis in long-term care facilities is not clear. If it is considered, there should be a compelling reason to use it—for example, for patients at high risk in whom anticoagulants are contraindicated because of ongoing bleeding or a higher risk of bleeding (eg, recent gastrointestinal bleeding, hemorrhagic stroke, coagulopathy, or thrombocytopenia). Furthermore, if stockings are used, they should be properly fitted and routinely monitored for adverse effects, since elderly patients are likely to be most susceptible to skin breakdown.

WHICH LONG-TERM CARE RESIDENTS SHOULD RECEIVE VTE PROPHYLAXIS?

Old age and immobility are not the only risk factors

The current ACCP guidelines suggest considering thromboprophylaxis for hospitalized medical patients over age 75 who cannot walk without assistance.¹³ However, we lack evidence to suggest a similar strategy in long-term care residents.

The ACCP guidelines are based on data on risk. Nearly 25% of elderly patients with confirmed pulmonary embolism had been immobile prior to their diagnosis.⁸ In addition, prolonged bed rest (> 14 days) has been reported to be the strongest independent risk factor for symptomatic deep venous thrombosis, increasing the risk more than fivefold.²⁰ Advanced age is also considered a risk factor for VTE, as risk starts to increase at age 40 and doubles each decade of life thereafter.¹⁸

No study has assessed the impact of these factors on the risk of VTE in long-term care residents. Since most of such patients are elderly and have impaired mobility, we believe a more selective approach should be used in assigning VTE risk status, one that does not use advanced age and immobility as the only criteria for starting thromboprophylaxis.

Residents of long-term care facilities may be immobile because of underlying illness or disability, such as cognitive impairment, sensory impairment (eg, poor access to corrective lenses and hearing aids), or poor access to assist devices (eg, walkers, canes). In addition, iatrogenic factors that decrease mobility such as indwelling bladder catheters and physical restraints are also common in such patients.

Efforts to improve mobility should be encouraged. However, we recommend that thromboprophylaxis be considered only in patients who have both impaired mobility and an intercurrent acute medical illness such as an acute infection or acute inflammatory disease.¹³

A related issue is the difference between long-term care residents with a chronic but stable disease and those with acute disease. Patients with acute exacerbations of congestive heart failure or chronic obstructive lung disease may be considered for thromboprophylaxis, as they become more comparable to acutely ill medical patients in whom clinical trials have shown the effectiveness of anticoagulant prophylaxis. On the other hand, patients with these diseases who remain stable may not need prophylaxis.

This approach avoids giving long-term anticoagulant prophylaxis to patients who have irreversible diseases and limits the use of these drugs and devices to higher-risk periods.

Consider thromboprophylaxis if…

• An acute exacerbation of congestive heart failure or chronic obstructive pulmonary disease
• Acute infection (eg, urosepsis, pneumonia, cellulitis, infectious diarrhea)
• An acute exacerbation of an inflammatory disease (eg, rheumatoid arthritis)
• Active cancer (eg, patient receiving radiation therapy or chemotherapy)
• Immobility and prior VTE.

Do not routinely consider prophylaxis if…

We also believe patients should not be routinely considered for anticoagulant VTE prophylaxis if they have:

• Chronic but stable cardiorespiratory disease
• Chronic but stable infectious or inflammatory disease
• Terminal cancer with very limited life expectancy
• Any contraindication to anticoagulants (eg, active bleeding, recent bleeding, coagulopathy, thrombocytopenia).

ANTICOAGULANT PROPHYLAXIS POSES RISKS IN LONG-TERM CARE RESIDENTS

Bleeding is the principal risk

The incidence of clinically important bleeding associated with anticoagulant prophylaxis is 0.2% to 5.6%, and the risk of fatal bleeding is 0.02% to 0.5%.^21–24

As no randomized trial has examined anticoagulant prophylaxis in elderly long-term care residents, their bleeding risk with this therapy is unclear. However, older patients are likely to be at higher risk than younger patients because they have more comorbidities, take more drugs that could interact with heparin and potentiate bleeding, and have fragile skin, predisposing to injury from subcutaneous injections.

Also, renal function tends to decline with age. In a retrospective study of 854 outpatients over age 65, 29% had moderate renal insufficiency (creatinine clearance 30–50 mL/min), and 6% had severe renal insufficiency (creatinine clearance < 30 mL/min).²⁵ Recent evidence suggests that some low-molecular-weight heparins (dalteparin and tinzaparin) do not bioaccumulate in patients with impaired renal function. However, enoxaparin and fondaparinux should be used with caution in patients with moderate to severe renal impairment.

Though much attention has recently been paid to increasing anticoagulant doses if the patient is obese, residents of long-term care facilities are more likely to be underweight. Dose adjustment should be considered when a low-molecular-weight heparin or fondaparinux is given to patients weighing less than 50 kg.

The other major risk of anticoagulant prophylaxis is heparin-induced thrombocytopenia, an infrequent but life-threatening complication caused by the formation of antibodies to the heparin-derived anticoagulant and a platelet surface antigen. It is associated with moderate thrombocytopenia and an incidence of venous or arterial thrombosis that is over 50%.²⁶

No study has assessed the incidence of heparin-induced thrombocytopenia in long-term care residents. A meta-analysis reported that the risk with anticoagulant prophylaxis was 1.6% with unfractionated heparin (95% confidence interval [CI] 1.2%–2.1%) and 0.6% with low-molecular-weight heparin (95% CI 0.4%–0.9%), and that this risk increased with the duration of prophylaxis.²⁷ If anticoagulant prophylaxis were given to all long-term care residents for extended durations (eg, for the duration of reduced mobility), the incidence and prevalence of heparin-induced thrombocytopenia would likely become a major concern.

Whenever anticoagulant prophylaxis is considered, the risks of both thrombosis and bleeding should be considered. Patients who are receiving anticoagulant prophylaxis should also be monitored for bleeding and heparin-induced thrombocytopenia. This is particularly true in long-term care residents, in whom the risks and benefits of anticoagulant prophylaxis are extrapolated from data from other populations.

MORE RESEARCH IS NEEDED

To date, we lack audits of thromboprophylaxis, clinical practice guidelines, and clear indications and contraindications for anticoagulant prophylaxis in long-term care residents. In the absence of such data, extrapolating the efficacy and safety of thromboprophylaxis from hospitalized patients to long-term care residents is difficult.

Clearly, additional research is needed to identify which long-term care residents would benefit most from thromboprophylaxis. In the meantime, a selective approach to identifying patients who should be considered for thromboprophylaxis should be adopted.

A 40-year-old man who works as a roofer began, 1 week ago, to experience episodes of generalized weakness, perioral numbness, and diffuse paresthesias. In the past he has had recurring nosebleeds but no history of other medical conditions.

His recent “spells” come on abruptly and spontaneously, without warning, and last about 15 minutes. He never loses consciousness, but he reports a feeling of derealization or an out-of-body experience—he can hear the people around him talking during the spells, but he feels that everything is far away. He has been having about three episodes per day. They typically occur after mild exertion or heavy lifting, and each episode resolves with complete rest. He has had no nausea, vomiting, loss of bowel or bladder control, fever, chills, or traumatic brain injury.

The patient first reported to the emergency department of a local hospital for evaluation. There, he underwent computed tomography (CT) of the head without contrast, which showed nothing abnormal. However, he had an episode while in the emergency department, which prompted his physician to admit him to the hospital.

In the hospital, he underwent an extensive medical evaluation. CT angiography revealed no evidence of vasculitis or occlusive disease. Results of electroencephalography during these spells were normal. Results of magnetic resonance imaging of the cervical and lumbar spine were also normal.

A neurologist was consulted. Concerned that the spells were due to paradoxical emboli coming through a patent foramen ovale, the neurologist recommended transthoracic echocardiography with agitated saline. This study showed a normal ejection fraction and a right-to-left shunt through a left pulmonary arteriovenous malformation (AVM). Unfortunately, the shunt fraction could not be estimated because the patient had another episode during the procedure, and so the procedure was cut short. CT of the chest confirmed a large AVM in the upper lobe of the left lung (Figure 1).

The patient is transferred

The patient’s physician requested that he be transferred to Mayo Clinic for further evaluation.

When he arrived, we performed a complete physical examination, in which we noted scattered erythematous maculopapular telangiectases in the lower lips and significant digital clubbing (Figure 2). He could not recall any family members having rheumatologic or cardiovascular diseases, but he recalled that his father has oral telangiectases and recurrent epistaxis.

His examination was interrupted by yet another spell, during which his oxygen saturation fell to 85%. We immediately started giving him oxygen by nasal cannula, which raised his oxygen saturation to 96%, and the spell promptly ended.

After his physical examination was completed and his records from the other hospital were reviewed, a diagnosis was made. No further diagnostic studies were pursued.

WHICH IS THE MOST LIKELY DIAGNOSIS?

1. Based on the information available, which of the following is the most likely diagnosis?

• Generalized tonic-clonic seizures
• Osler-Weber-Rendu disease
• Subarachnoid hemorrhage
• Conversion disorder
• Atrial septal defect

Generalized tonic-clonic seizures begin with abrupt loss of consciousness, followed by stiffening of the body and extremities. This is the tonic phase, which may last for 1 minute. The clonic phase follows, characterized by abnormal jerking and teeth-clenching (raising the concern that the patient will bite his or her tongue). The clonic phase lasts 1 to 2 minutes. After a seizure, confusion and headache are common. On electroencephalography, epileptiform abnormalities are documented in about 23% of patients with a first documented seizure.¹

Our patient’s history of remaining fully conscious and of having normal electroencephalographic findings during his spells does not suggest generalized tonic-clonic seizures.

Osler-Weber-Rendu disease is also known as hereditary hemorrhagic telangiectasia (HHT). Its pathophysiology is complex, and it is believed to be related to mutations in an endothelial protein² that lead to abnormal vascular structures. The estimated prevalence in European studies is 1 in 5,000; in Japanese studies it is 1 in 8,000.^3–4

The diagnosis of HHT is based on four clinical criteria:

• Spontaneous and recurrent epistaxis
• Multiple mucocutaneous telangiectases
• Pulmonary, cerebral, or gastrointestinal AVMs
• A first-degree relative with the disease.

The presence of three or four of these criteria establishes a “definite” diagnosis, while fewer than two makes it “unlikely.”⁵ Since the spectrum of this disease is wide, varying from mild epistaxis to iron-deficiency anemia, its diagnosis is often missed.⁶

Our patient meets at least three of the criteria—recurrent epistaxis, oral telangiectases, and a CT-documented pulmonary AVM. His father has a history of oral telangiectases and epistaxis but was never formally diagnosed with HHT. The patient presented with spells of weakness and paresthesias from worsening hypoxemia due to an enlarged pulmonary AVM. Thus, based on these features, HHT is the most likely diagnosis.

Subarachnoid hemorrhage is commonly from a ruptured cerebral aneurysm. Common symptoms include sudden, severe headaches with focal neurologic deficits, a stiff neck, brief loss of consciousness, nausea, and vomiting.⁷

Our patient’s CT scan showed no intracranial bleeding, and CT angiography showed no evidence of aneurysm. Thus, he has neither clinical nor radiographic features of subarachnoid hemorrhage.

Conversion disorder is typically associated with psychological stressors.⁸ It is characterized by the sudden onset of neurologic deficits such as blindness, paralysis, and numbness that cannot be explained by a general medical condition.

Our patient has a known pulmonary AVM with clinical and laboratory findings of hypoxemia that explain his spells. Therefore, the diagnosis of conversion disorder cannot be made.

A right-to-left intracardiac shunt can be present in patients with patent foramen ovale, atrial septal defects with shunt reversal, Eisenmenger syndrome, or tetralogy of Fallot (even in adults). It can present with hypoxemia and neurologic weakness.

Our patient’s echocardiogram ruled out these conditions.

MANIFESTATIONS OF HEREDITARY HEMORRHAGIC TELANGIECTASIA

2. Which is the most common clinical manifestation of HHT?

• Epistaxis
• Mucocutaneous telangiectases
• Hematochezia
• Dyspnea

Epistaxis is the most common presentation, occurring in more than 90% of patients.⁹ Many patients experience only mild occasional nosebleeds that are not frequent or severe enough to cause anemia or to lead to medical treatment or consultation. Others, however, have heavy, frequent bleeding that requires invasive interventions.¹⁰

Mucocutaneous telangiectases are the second most common clinical manifestation, documented in about 75% of patients. They are cosmetically unpleasant but rarely bleed. They occur most commonly on the face, lips, tongue, and fingertips, and they increase in size and number with age.¹¹

Gastrointestinal bleeding, sometimes manifesting as hematochezia, occurs in one-third of people with HHT. It most commonly presents with iron-deficiency anemia in patients over age 40.¹²

Dyspnea. Pulmonary AVMs occur in 30% to 50% of affected people, but interestingly, most patients with pulmonary AVMs have no respiratory symptoms, including dyspnea.

In pulmonary AVMs, abnormal vessels replace normal capillary beds, creating a capillary-free communication between the pulmonary and systemic circulations. This abnormal connection prevents blood from the pulmonary arterial system from being oxygenated, resulting in hypoxemia and secondary polycythemia, as in our patient. One-third of patients have evidence of right-to-left shunting, such as the clubbing in our patient.^9,13

Other, less common complications of HHT include seizures or hemorrhage from cerebral AVMs and stroke and brain abscesses from paradoxical embolization due to the loss of the capillary filter in the pulmonary vascular bed. Hepatic involvement may result in portal hypertension and hepatic encephalopathy.¹⁴

Back to our patient

As mentioned above, during one of the patient’s spells of paresthesia and weakness, we noted his oxygen saturation by oximetry was 85%. At that time, his arterial Po₂ was also low at 50 mm Hg (normal 70–100). With oxygen supplementation, his spell completely resolved and his Po₂ improved to 80 mm Hg. Though the shunt fraction of his pulmonary AVM was never measured, it was likely less than 30% of the cardiac output, as his hypoxemia improved with oxygen supplementation alone.¹⁵ When he was taken off oxygen supplementation, his spells recurred, but with oxygen support he remained clinically stable.

MANAGEMENT

3. Which is the next logical step in our patient’s management?

• Consult a surgeon for lobectomy
• Consult an interventional radiologist for embolization therapy
• Transfer to the intensive care unit for elective intubation
• Observe with close follow-up

Untreated pulmonary AVMs enlarge at an estimated rate of 0.3 mm/year. The estimated death rate is up to 15.8% per year, with most deaths resulting from stroke, cerebral abscess, hemoptysis, and hemothorax.^16–18 Common indications for treatment are progressively enlarging lesions, symptomatic hypoxemia, and paradoxical embolization.¹⁹ Pulmonary AVMs in which the feeding artery is 3 mm or more in diameter require treatment.

Embolization therapy, in which the AVM is occluded angiographically, is considered a first-line treatment for pulmonary AVM, with a procedural success rate (defined as involution of the AVM) of 97%.²⁰ Embolization therapy allows patients to avoid major surgery, with its potential complications, and it has a shorter recovery time.

Surgical procedures such as excision, vascular ligation, or lobectomy can be considered if the lesion cannot be treated by embolization or if the patient has an anaphylactic allergy to contrast dyes.

This patient had no clinical signs of impending respiratory failure requiring elective intubation.

Since he was experiencing symptoms, there is no role for observation in this case.

Back to our patient

An interventional radiologist was consulted, and the patient underwent bilateral pulmonary artery angiography with successful coil embolization of his large left-upper-lobe AVM. He was weaned off oxygen and had no further spells of generalized weakness and paresthesias.

Given his father’s history of recurrent epistaxis and oral telangiectases, the patient asks about the risk of his children acquiring this disease.

GENETICS OF HEREDITARY HEMORRHAGIC TELANGIECTASIA

4. Which of the following is the inheritance pattern for HHT?

• Autosomal dominant
• Autosomal recessive
• Maternal inheritance
• X-linked recessive

The inheritance pattern is autosomal dominant with variable expression and penetrance. At least four different mutations have been identified in genes on chromosomes 9 and 12 that result in abnormal vascular malformations.^21–24 The other modes of inheritance have not been described in HHT.

RECOMMENDATIONS FOR OUR PATIENT

5. Which of the following is not recommended for our patient?

• Consideration of genetic testing
• Consideration of screening of first-degree relatives
• Dental prophylaxis
• Scuba diving

Genetic testing. The molecular diagnosis of HHT is primarily based on sequencing of the entire coding regions of the ENG and ALK1 genes on chromosomes 9 and 12, respectively. The interpretation of these results is quite complex. The clinical genetics laboratories in North America that currently offer molecular diagnostic testing for HHT recommend that testing be coordinated and ordered through a center that specializes in this disease or by a genetics professional. Testing of the index case is performed to confirm the clinical diagnosis and also to determine if genetic testing will be possible in at-risk relatives. Further genetic testing should be pursued in at-risk family members only if the index case has a positive result.²⁵

Screening of relatives. Given that HHT is an autosomal dominant disease, the current practice is to offer molecular genetic screening early in life for first-degree relatives.^25,26 The external signs such as telangiectases and nosebleeds may not manifest until the second or third decade of life. However, AVMs in the brain, spinal cord, lungs, and liver are usually congenital and may present suddenly and with serious complications, even in childhood.

Dental prophylaxis. People with HHT and a pulmonary AVM are at risk of bacteremia and consequent brain abscesses after dental procedures. Antibiotic prophylaxis is therefore highly recommended.²⁷

One sport to avoid. There have been several case reports of paradoxical air emboli occurring in patients with HHT complicated by a pulmonary AVM. Hsu et al²⁸ reported a 31-year-old man with an undiagnosed large pulmonary AVM and HHT who became comatose with diffuse bilateral hemispheric brain swelling on head CT after scuba diving, due to air embolism.

The HHT Foundation International recommends that people with this disease avoid scuba diving (the only sport to be avoided) owing to the risk of air emboli from small lung AVMs. It also recommends that patients alert health care providers about their risk of air embolism whenever intravenous access is being established.

Back to our patient

The patient met with a geneticist, and blood was collected for genetic testing before he was sent home. Additionally, the need to screen his first-degree relatives was thoroughly discussed. Four days after discharge he returned to work, and his spells have not recurred. He has a follow-up appointment scheduled with a pulmonologist specializing in this disease for the results of genetic testing and for continued management.

TAKE-HOME POINTS

• The diagnosis of HHT is based on the following four clinical criteria: spontaneous or recurrent epistaxis, multiple mucocutaneous telangiectases, visceral involvement (eg, cerebral, pulmonary, or gastrointestinal AVM), and a first-degree relative with this disease.
• The diagnosis may be confirmed with genetic testing.
• The diagnosis may be underreported, given the wide spectrum of disease presentation, from inconsequential epistaxis to massive gastrointestinal bleeding.
• HHT is autosomal dominant, and therefore all first-degree relatives should be screened.

A 40-year-old man who works as a roofer began, 1 week ago, to experience episodes of generalized weakness, perioral numbness, and diffuse paresthesias. In the past he has had recurring nosebleeds but no history of other medical conditions.

His recent “spells” come on abruptly and spontaneously, without warning, and last about 15 minutes. He never loses consciousness, but he reports a feeling of derealization or an out-of-body experience—he can hear the people around him talking during the spells, but he feels that everything is far away. He has been having about three episodes per day. They typically occur after mild exertion or heavy lifting, and each episode resolves with complete rest. He has had no nausea, vomiting, loss of bowel or bladder control, fever, chills, or traumatic brain injury.

The patient first reported to the emergency department of a local hospital for evaluation. There, he underwent computed tomography (CT) of the head without contrast, which showed nothing abnormal. However, he had an episode while in the emergency department, which prompted his physician to admit him to the hospital.

In the hospital, he underwent an extensive medical evaluation. CT angiography revealed no evidence of vasculitis or occlusive disease. Results of electroencephalography during these spells were normal. Results of magnetic resonance imaging of the cervical and lumbar spine were also normal.

A neurologist was consulted. Concerned that the spells were due to paradoxical emboli coming through a patent foramen ovale, the neurologist recommended transthoracic echocardiography with agitated saline. This study showed a normal ejection fraction and a right-to-left shunt through a left pulmonary arteriovenous malformation (AVM). Unfortunately, the shunt fraction could not be estimated because the patient had another episode during the procedure, and so the procedure was cut short. CT of the chest confirmed a large AVM in the upper lobe of the left lung (Figure 1).

The patient is transferred

The patient’s physician requested that he be transferred to Mayo Clinic for further evaluation.

When he arrived, we performed a complete physical examination, in which we noted scattered erythematous maculopapular telangiectases in the lower lips and significant digital clubbing (Figure 2). He could not recall any family members having rheumatologic or cardiovascular diseases, but he recalled that his father has oral telangiectases and recurrent epistaxis.

His examination was interrupted by yet another spell, during which his oxygen saturation fell to 85%. We immediately started giving him oxygen by nasal cannula, which raised his oxygen saturation to 96%, and the spell promptly ended.

After his physical examination was completed and his records from the other hospital were reviewed, a diagnosis was made. No further diagnostic studies were pursued.

WHICH IS THE MOST LIKELY DIAGNOSIS?

1. Based on the information available, which of the following is the most likely diagnosis?

• Generalized tonic-clonic seizures
• Osler-Weber-Rendu disease
• Subarachnoid hemorrhage
• Conversion disorder
• Atrial septal defect

Generalized tonic-clonic seizures begin with abrupt loss of consciousness, followed by stiffening of the body and extremities. This is the tonic phase, which may last for 1 minute. The clonic phase follows, characterized by abnormal jerking and teeth-clenching (raising the concern that the patient will bite his or her tongue). The clonic phase lasts 1 to 2 minutes. After a seizure, confusion and headache are common. On electroencephalography, epileptiform abnormalities are documented in about 23% of patients with a first documented seizure.¹

Our patient’s history of remaining fully conscious and of having normal electroencephalographic findings during his spells does not suggest generalized tonic-clonic seizures.

Osler-Weber-Rendu disease is also known as hereditary hemorrhagic telangiectasia (HHT). Its pathophysiology is complex, and it is believed to be related to mutations in an endothelial protein² that lead to abnormal vascular structures. The estimated prevalence in European studies is 1 in 5,000; in Japanese studies it is 1 in 8,000.^3–4

The diagnosis of HHT is based on four clinical criteria:

• Spontaneous and recurrent epistaxis
• Multiple mucocutaneous telangiectases
• Pulmonary, cerebral, or gastrointestinal AVMs
• A first-degree relative with the disease.

The presence of three or four of these criteria establishes a “definite” diagnosis, while fewer than two makes it “unlikely.”⁵ Since the spectrum of this disease is wide, varying from mild epistaxis to iron-deficiency anemia, its diagnosis is often missed.⁶

Our patient meets at least three of the criteria—recurrent epistaxis, oral telangiectases, and a CT-documented pulmonary AVM. His father has a history of oral telangiectases and epistaxis but was never formally diagnosed with HHT. The patient presented with spells of weakness and paresthesias from worsening hypoxemia due to an enlarged pulmonary AVM. Thus, based on these features, HHT is the most likely diagnosis.

Subarachnoid hemorrhage is commonly from a ruptured cerebral aneurysm. Common symptoms include sudden, severe headaches with focal neurologic deficits, a stiff neck, brief loss of consciousness, nausea, and vomiting.⁷

Our patient’s CT scan showed no intracranial bleeding, and CT angiography showed no evidence of aneurysm. Thus, he has neither clinical nor radiographic features of subarachnoid hemorrhage.

Conversion disorder is typically associated with psychological stressors.⁸ It is characterized by the sudden onset of neurologic deficits such as blindness, paralysis, and numbness that cannot be explained by a general medical condition.

Our patient has a known pulmonary AVM with clinical and laboratory findings of hypoxemia that explain his spells. Therefore, the diagnosis of conversion disorder cannot be made.

A right-to-left intracardiac shunt can be present in patients with patent foramen ovale, atrial septal defects with shunt reversal, Eisenmenger syndrome, or tetralogy of Fallot (even in adults). It can present with hypoxemia and neurologic weakness.

Our patient’s echocardiogram ruled out these conditions.

MANIFESTATIONS OF HEREDITARY HEMORRHAGIC TELANGIECTASIA

2. Which is the most common clinical manifestation of HHT?

• Epistaxis
• Mucocutaneous telangiectases
• Hematochezia
• Dyspnea

Epistaxis is the most common presentation, occurring in more than 90% of patients.⁹ Many patients experience only mild occasional nosebleeds that are not frequent or severe enough to cause anemia or to lead to medical treatment or consultation. Others, however, have heavy, frequent bleeding that requires invasive interventions.¹⁰

Mucocutaneous telangiectases are the second most common clinical manifestation, documented in about 75% of patients. They are cosmetically unpleasant but rarely bleed. They occur most commonly on the face, lips, tongue, and fingertips, and they increase in size and number with age.¹¹

Gastrointestinal bleeding, sometimes manifesting as hematochezia, occurs in one-third of people with HHT. It most commonly presents with iron-deficiency anemia in patients over age 40.¹²

Dyspnea. Pulmonary AVMs occur in 30% to 50% of affected people, but interestingly, most patients with pulmonary AVMs have no respiratory symptoms, including dyspnea.

In pulmonary AVMs, abnormal vessels replace normal capillary beds, creating a capillary-free communication between the pulmonary and systemic circulations. This abnormal connection prevents blood from the pulmonary arterial system from being oxygenated, resulting in hypoxemia and secondary polycythemia, as in our patient. One-third of patients have evidence of right-to-left shunting, such as the clubbing in our patient.^9,13

Other, less common complications of HHT include seizures or hemorrhage from cerebral AVMs and stroke and brain abscesses from paradoxical embolization due to the loss of the capillary filter in the pulmonary vascular bed. Hepatic involvement may result in portal hypertension and hepatic encephalopathy.¹⁴

Back to our patient

As mentioned above, during one of the patient’s spells of paresthesia and weakness, we noted his oxygen saturation by oximetry was 85%. At that time, his arterial Po₂ was also low at 50 mm Hg (normal 70–100). With oxygen supplementation, his spell completely resolved and his Po₂ improved to 80 mm Hg. Though the shunt fraction of his pulmonary AVM was never measured, it was likely less than 30% of the cardiac output, as his hypoxemia improved with oxygen supplementation alone.¹⁵ When he was taken off oxygen supplementation, his spells recurred, but with oxygen support he remained clinically stable.

MANAGEMENT

3. Which is the next logical step in our patient’s management?

• Consult a surgeon for lobectomy
• Consult an interventional radiologist for embolization therapy
• Transfer to the intensive care unit for elective intubation
• Observe with close follow-up

Untreated pulmonary AVMs enlarge at an estimated rate of 0.3 mm/year. The estimated death rate is up to 15.8% per year, with most deaths resulting from stroke, cerebral abscess, hemoptysis, and hemothorax.^16–18 Common indications for treatment are progressively enlarging lesions, symptomatic hypoxemia, and paradoxical embolization.¹⁹ Pulmonary AVMs in which the feeding artery is 3 mm or more in diameter require treatment.

Embolization therapy, in which the AVM is occluded angiographically, is considered a first-line treatment for pulmonary AVM, with a procedural success rate (defined as involution of the AVM) of 97%.²⁰ Embolization therapy allows patients to avoid major surgery, with its potential complications, and it has a shorter recovery time.

Surgical procedures such as excision, vascular ligation, or lobectomy can be considered if the lesion cannot be treated by embolization or if the patient has an anaphylactic allergy to contrast dyes.

This patient had no clinical signs of impending respiratory failure requiring elective intubation.

Since he was experiencing symptoms, there is no role for observation in this case.

Back to our patient

An interventional radiologist was consulted, and the patient underwent bilateral pulmonary artery angiography with successful coil embolization of his large left-upper-lobe AVM. He was weaned off oxygen and had no further spells of generalized weakness and paresthesias.

Given his father’s history of recurrent epistaxis and oral telangiectases, the patient asks about the risk of his children acquiring this disease.

GENETICS OF HEREDITARY HEMORRHAGIC TELANGIECTASIA

4. Which of the following is the inheritance pattern for HHT?

• Autosomal dominant
• Autosomal recessive
• Maternal inheritance
• X-linked recessive

The inheritance pattern is autosomal dominant with variable expression and penetrance. At least four different mutations have been identified in genes on chromosomes 9 and 12 that result in abnormal vascular malformations.^21–24 The other modes of inheritance have not been described in HHT.

RECOMMENDATIONS FOR OUR PATIENT

5. Which of the following is not recommended for our patient?

• Consideration of genetic testing
• Consideration of screening of first-degree relatives
• Dental prophylaxis
• Scuba diving

Genetic testing. The molecular diagnosis of HHT is primarily based on sequencing of the entire coding regions of the ENG and ALK1 genes on chromosomes 9 and 12, respectively. The interpretation of these results is quite complex. The clinical genetics laboratories in North America that currently offer molecular diagnostic testing for HHT recommend that testing be coordinated and ordered through a center that specializes in this disease or by a genetics professional. Testing of the index case is performed to confirm the clinical diagnosis and also to determine if genetic testing will be possible in at-risk relatives. Further genetic testing should be pursued in at-risk family members only if the index case has a positive result.²⁵

Screening of relatives. Given that HHT is an autosomal dominant disease, the current practice is to offer molecular genetic screening early in life for first-degree relatives.^25,26 The external signs such as telangiectases and nosebleeds may not manifest until the second or third decade of life. However, AVMs in the brain, spinal cord, lungs, and liver are usually congenital and may present suddenly and with serious complications, even in childhood.

Dental prophylaxis. People with HHT and a pulmonary AVM are at risk of bacteremia and consequent brain abscesses after dental procedures. Antibiotic prophylaxis is therefore highly recommended.²⁷

One sport to avoid. There have been several case reports of paradoxical air emboli occurring in patients with HHT complicated by a pulmonary AVM. Hsu et al²⁸ reported a 31-year-old man with an undiagnosed large pulmonary AVM and HHT who became comatose with diffuse bilateral hemispheric brain swelling on head CT after scuba diving, due to air embolism.

The HHT Foundation International recommends that people with this disease avoid scuba diving (the only sport to be avoided) owing to the risk of air emboli from small lung AVMs. It also recommends that patients alert health care providers about their risk of air embolism whenever intravenous access is being established.

Back to our patient

The patient met with a geneticist, and blood was collected for genetic testing before he was sent home. Additionally, the need to screen his first-degree relatives was thoroughly discussed. Four days after discharge he returned to work, and his spells have not recurred. He has a follow-up appointment scheduled with a pulmonologist specializing in this disease for the results of genetic testing and for continued management.

TAKE-HOME POINTS

• The diagnosis of HHT is based on the following four clinical criteria: spontaneous or recurrent epistaxis, multiple mucocutaneous telangiectases, visceral involvement (eg, cerebral, pulmonary, or gastrointestinal AVM), and a first-degree relative with this disease.
• The diagnosis may be confirmed with genetic testing.
• The diagnosis may be underreported, given the wide spectrum of disease presentation, from inconsequential epistaxis to massive gastrointestinal bleeding.
• HHT is autosomal dominant, and therefore all first-degree relatives should be screened.

A 40-year-old man who works as a roofer began, 1 week ago, to experience episodes of generalized weakness, perioral numbness, and diffuse paresthesias. In the past he has had recurring nosebleeds but no history of other medical conditions.

His recent “spells” come on abruptly and spontaneously, without warning, and last about 15 minutes. He never loses consciousness, but he reports a feeling of derealization or an out-of-body experience—he can hear the people around him talking during the spells, but he feels that everything is far away. He has been having about three episodes per day. They typically occur after mild exertion or heavy lifting, and each episode resolves with complete rest. He has had no nausea, vomiting, loss of bowel or bladder control, fever, chills, or traumatic brain injury.

The patient first reported to the emergency department of a local hospital for evaluation. There, he underwent computed tomography (CT) of the head without contrast, which showed nothing abnormal. However, he had an episode while in the emergency department, which prompted his physician to admit him to the hospital.

In the hospital, he underwent an extensive medical evaluation. CT angiography revealed no evidence of vasculitis or occlusive disease. Results of electroencephalography during these spells were normal. Results of magnetic resonance imaging of the cervical and lumbar spine were also normal.

A neurologist was consulted. Concerned that the spells were due to paradoxical emboli coming through a patent foramen ovale, the neurologist recommended transthoracic echocardiography with agitated saline. This study showed a normal ejection fraction and a right-to-left shunt through a left pulmonary arteriovenous malformation (AVM). Unfortunately, the shunt fraction could not be estimated because the patient had another episode during the procedure, and so the procedure was cut short. CT of the chest confirmed a large AVM in the upper lobe of the left lung (Figure 1).

The patient is transferred

The patient’s physician requested that he be transferred to Mayo Clinic for further evaluation.

When he arrived, we performed a complete physical examination, in which we noted scattered erythematous maculopapular telangiectases in the lower lips and significant digital clubbing (Figure 2). He could not recall any family members having rheumatologic or cardiovascular diseases, but he recalled that his father has oral telangiectases and recurrent epistaxis.

His examination was interrupted by yet another spell, during which his oxygen saturation fell to 85%. We immediately started giving him oxygen by nasal cannula, which raised his oxygen saturation to 96%, and the spell promptly ended.

After his physical examination was completed and his records from the other hospital were reviewed, a diagnosis was made. No further diagnostic studies were pursued.

WHICH IS THE MOST LIKELY DIAGNOSIS?

1. Based on the information available, which of the following is the most likely diagnosis?

• Generalized tonic-clonic seizures
• Osler-Weber-Rendu disease
• Subarachnoid hemorrhage
• Conversion disorder
• Atrial septal defect

Generalized tonic-clonic seizures begin with abrupt loss of consciousness, followed by stiffening of the body and extremities. This is the tonic phase, which may last for 1 minute. The clonic phase follows, characterized by abnormal jerking and teeth-clenching (raising the concern that the patient will bite his or her tongue). The clonic phase lasts 1 to 2 minutes. After a seizure, confusion and headache are common. On electroencephalography, epileptiform abnormalities are documented in about 23% of patients with a first documented seizure.¹

Our patient’s history of remaining fully conscious and of having normal electroencephalographic findings during his spells does not suggest generalized tonic-clonic seizures.

Osler-Weber-Rendu disease is also known as hereditary hemorrhagic telangiectasia (HHT). Its pathophysiology is complex, and it is believed to be related to mutations in an endothelial protein² that lead to abnormal vascular structures. The estimated prevalence in European studies is 1 in 5,000; in Japanese studies it is 1 in 8,000.^3–4

The diagnosis of HHT is based on four clinical criteria:

• Spontaneous and recurrent epistaxis
• Multiple mucocutaneous telangiectases
• Pulmonary, cerebral, or gastrointestinal AVMs
• A first-degree relative with the disease.

The presence of three or four of these criteria establishes a “definite” diagnosis, while fewer than two makes it “unlikely.”⁵ Since the spectrum of this disease is wide, varying from mild epistaxis to iron-deficiency anemia, its diagnosis is often missed.⁶

Our patient meets at least three of the criteria—recurrent epistaxis, oral telangiectases, and a CT-documented pulmonary AVM. His father has a history of oral telangiectases and epistaxis but was never formally diagnosed with HHT. The patient presented with spells of weakness and paresthesias from worsening hypoxemia due to an enlarged pulmonary AVM. Thus, based on these features, HHT is the most likely diagnosis.

Subarachnoid hemorrhage is commonly from a ruptured cerebral aneurysm. Common symptoms include sudden, severe headaches with focal neurologic deficits, a stiff neck, brief loss of consciousness, nausea, and vomiting.⁷

Our patient’s CT scan showed no intracranial bleeding, and CT angiography showed no evidence of aneurysm. Thus, he has neither clinical nor radiographic features of subarachnoid hemorrhage.

Conversion disorder is typically associated with psychological stressors.⁸ It is characterized by the sudden onset of neurologic deficits such as blindness, paralysis, and numbness that cannot be explained by a general medical condition.

Our patient has a known pulmonary AVM with clinical and laboratory findings of hypoxemia that explain his spells. Therefore, the diagnosis of conversion disorder cannot be made.

A right-to-left intracardiac shunt can be present in patients with patent foramen ovale, atrial septal defects with shunt reversal, Eisenmenger syndrome, or tetralogy of Fallot (even in adults). It can present with hypoxemia and neurologic weakness.

Our patient’s echocardiogram ruled out these conditions.

MANIFESTATIONS OF HEREDITARY HEMORRHAGIC TELANGIECTASIA

2. Which is the most common clinical manifestation of HHT?

• Epistaxis
• Mucocutaneous telangiectases
• Hematochezia
• Dyspnea

Epistaxis is the most common presentation, occurring in more than 90% of patients.⁹ Many patients experience only mild occasional nosebleeds that are not frequent or severe enough to cause anemia or to lead to medical treatment or consultation. Others, however, have heavy, frequent bleeding that requires invasive interventions.¹⁰

Mucocutaneous telangiectases are the second most common clinical manifestation, documented in about 75% of patients. They are cosmetically unpleasant but rarely bleed. They occur most commonly on the face, lips, tongue, and fingertips, and they increase in size and number with age.¹¹

Gastrointestinal bleeding, sometimes manifesting as hematochezia, occurs in one-third of people with HHT. It most commonly presents with iron-deficiency anemia in patients over age 40.¹²

Dyspnea. Pulmonary AVMs occur in 30% to 50% of affected people, but interestingly, most patients with pulmonary AVMs have no respiratory symptoms, including dyspnea.

In pulmonary AVMs, abnormal vessels replace normal capillary beds, creating a capillary-free communication between the pulmonary and systemic circulations. This abnormal connection prevents blood from the pulmonary arterial system from being oxygenated, resulting in hypoxemia and secondary polycythemia, as in our patient. One-third of patients have evidence of right-to-left shunting, such as the clubbing in our patient.^9,13

Other, less common complications of HHT include seizures or hemorrhage from cerebral AVMs and stroke and brain abscesses from paradoxical embolization due to the loss of the capillary filter in the pulmonary vascular bed. Hepatic involvement may result in portal hypertension and hepatic encephalopathy.¹⁴

Back to our patient

As mentioned above, during one of the patient’s spells of paresthesia and weakness, we noted his oxygen saturation by oximetry was 85%. At that time, his arterial Po₂ was also low at 50 mm Hg (normal 70–100). With oxygen supplementation, his spell completely resolved and his Po₂ improved to 80 mm Hg. Though the shunt fraction of his pulmonary AVM was never measured, it was likely less than 30% of the cardiac output, as his hypoxemia improved with oxygen supplementation alone.¹⁵ When he was taken off oxygen supplementation, his spells recurred, but with oxygen support he remained clinically stable.

MANAGEMENT

3. Which is the next logical step in our patient’s management?

• Consult a surgeon for lobectomy
• Consult an interventional radiologist for embolization therapy
• Transfer to the intensive care unit for elective intubation
• Observe with close follow-up

Untreated pulmonary AVMs enlarge at an estimated rate of 0.3 mm/year. The estimated death rate is up to 15.8% per year, with most deaths resulting from stroke, cerebral abscess, hemoptysis, and hemothorax.^16–18 Common indications for treatment are progressively enlarging lesions, symptomatic hypoxemia, and paradoxical embolization.¹⁹ Pulmonary AVMs in which the feeding artery is 3 mm or more in diameter require treatment.

Embolization therapy, in which the AVM is occluded angiographically, is considered a first-line treatment for pulmonary AVM, with a procedural success rate (defined as involution of the AVM) of 97%.²⁰ Embolization therapy allows patients to avoid major surgery, with its potential complications, and it has a shorter recovery time.

Surgical procedures such as excision, vascular ligation, or lobectomy can be considered if the lesion cannot be treated by embolization or if the patient has an anaphylactic allergy to contrast dyes.

This patient had no clinical signs of impending respiratory failure requiring elective intubation.

Since he was experiencing symptoms, there is no role for observation in this case.

Back to our patient

An interventional radiologist was consulted, and the patient underwent bilateral pulmonary artery angiography with successful coil embolization of his large left-upper-lobe AVM. He was weaned off oxygen and had no further spells of generalized weakness and paresthesias.

Given his father’s history of recurrent epistaxis and oral telangiectases, the patient asks about the risk of his children acquiring this disease.

GENETICS OF HEREDITARY HEMORRHAGIC TELANGIECTASIA

4. Which of the following is the inheritance pattern for HHT?

• Autosomal dominant
• Autosomal recessive
• Maternal inheritance
• X-linked recessive

The inheritance pattern is autosomal dominant with variable expression and penetrance. At least four different mutations have been identified in genes on chromosomes 9 and 12 that result in abnormal vascular malformations.^21–24 The other modes of inheritance have not been described in HHT.

RECOMMENDATIONS FOR OUR PATIENT

5. Which of the following is not recommended for our patient?

• Consideration of genetic testing
• Consideration of screening of first-degree relatives
• Dental prophylaxis
• Scuba diving

Genetic testing. The molecular diagnosis of HHT is primarily based on sequencing of the entire coding regions of the ENG and ALK1 genes on chromosomes 9 and 12, respectively. The interpretation of these results is quite complex. The clinical genetics laboratories in North America that currently offer molecular diagnostic testing for HHT recommend that testing be coordinated and ordered through a center that specializes in this disease or by a genetics professional. Testing of the index case is performed to confirm the clinical diagnosis and also to determine if genetic testing will be possible in at-risk relatives. Further genetic testing should be pursued in at-risk family members only if the index case has a positive result.²⁵

Screening of relatives. Given that HHT is an autosomal dominant disease, the current practice is to offer molecular genetic screening early in life for first-degree relatives.^25,26 The external signs such as telangiectases and nosebleeds may not manifest until the second or third decade of life. However, AVMs in the brain, spinal cord, lungs, and liver are usually congenital and may present suddenly and with serious complications, even in childhood.

Dental prophylaxis. People with HHT and a pulmonary AVM are at risk of bacteremia and consequent brain abscesses after dental procedures. Antibiotic prophylaxis is therefore highly recommended.²⁷

One sport to avoid. There have been several case reports of paradoxical air emboli occurring in patients with HHT complicated by a pulmonary AVM. Hsu et al²⁸ reported a 31-year-old man with an undiagnosed large pulmonary AVM and HHT who became comatose with diffuse bilateral hemispheric brain swelling on head CT after scuba diving, due to air embolism.

The HHT Foundation International recommends that people with this disease avoid scuba diving (the only sport to be avoided) owing to the risk of air emboli from small lung AVMs. It also recommends that patients alert health care providers about their risk of air embolism whenever intravenous access is being established.

Back to our patient

The patient met with a geneticist, and blood was collected for genetic testing before he was sent home. Additionally, the need to screen his first-degree relatives was thoroughly discussed. Four days after discharge he returned to work, and his spells have not recurred. He has a follow-up appointment scheduled with a pulmonologist specializing in this disease for the results of genetic testing and for continued management.

TAKE-HOME POINTS

• The diagnosis of HHT is based on the following four clinical criteria: spontaneous or recurrent epistaxis, multiple mucocutaneous telangiectases, visceral involvement (eg, cerebral, pulmonary, or gastrointestinal AVM), and a first-degree relative with this disease.
• The diagnosis may be confirmed with genetic testing.
• The diagnosis may be underreported, given the wide spectrum of disease presentation, from inconsequential epistaxis to massive gastrointestinal bleeding.
• HHT is autosomal dominant, and therefore all first-degree relatives should be screened.

Many clinicians are concerned about a possible interaction between the proton pump inhibitor omeprazole (Prilosec) and the antiplatelet drug clopidogrel (Plavix), which is often given to patients as part of dual antiplatelet therapy after an acute coronary syndrome or a percutaneous coronary intervention. Indeed, the US Food and Drug Administration (FDA) has warned that omeprazole reduces the antiplatelet effect of clopidogrel.

Although we should not take such warnings lightly, we also should not be alarmed. The data on which the FDA warning was based came mostly from laboratory assays of platelet function. Preliminary results from a randomized, controlled clinical trial with hard end points show that, for the time being, we should not change the way we manage patients.

PROTON PUMP INHIBITORS DECREASE GASTROINTESTINAL BLEEDING

Dual antiplatelet therapy with aspirin and clopidogrel decreases the risk of major adverse cardiac events after an acute coronary syndrome or a percutaneous coronary intervention compared with aspirin alone.¹ However, it also increases the risk of gastrointestinal bleeding. A recent analysis determined that dual antiplatelet therapy was the most significant risk factor associated with serious or fatal gastrointestinal bleeding in high-risk survivors of myocardial infarction.²

Given the risks of significant morbidity and death in patients on dual antiplatelet therapy who develop gastrointestinal bleeding, an expert consensus panel recommended the use of proton pump inhibitors in patients on dual antiplatelet therapy who have risk factors for gastrointestinal bleeding.³ Accordingly, these drugs are commonly used for gastrointestinal protection in patients requiring dual antiplatelet therapy.

A POSSIBLE CYP450 INTERACTION

Clopidogrel is metabolized from a prodrug to its active metabolite by the cytochrome P450 (CYP450) system. Proton pump inhibitors also are metabolized by the CYP450 system.⁴ Proton pump inhibitors are thought to diminish the activity of clopidogrel via inhibition of the CYP2C19 isoenzyme. However, the clinical significance of this inhibition is not clear. Different drugs of this class inhibit the CYP450 system to varying degrees.

The potential interaction between proton pump inhibitors and clopidogrel is worrisome for many physicians, since adverse cardiovascular outcomes are more common in patients in whom the antiplatelet response to clopidogrel is impaired.¹ This interaction led to the publication of numerous articles, and prompted the FDA to carefully analyze the potential clinical implications.

In several randomized trials, omeprazole diminished the response to clopidogrel (measured via platelet function assays).^5,6 It is unclear if this is a class effect, as proton pump inhibitors other than omeprazole have not consistently been shown to have this effect.^6,7 Observational studies of the effect of co-administration of a proton pump inhibitor and clopidogrel on cardiovascular outcomes following acute coronary syndromes have had conflicting findings.^8–11

THE FDA ISSUES AN ADVISORY

Given the reports of an impaired platelet response to clopidogrel with omeprazole, the FDA asked the manufacturer for data on this potential interaction. The data showed diminished platelet inhibition when clopidogrel was co-administered with omeprazole or when the two were taken 12 hours apart.

On November 17, 2009, the FDA issued a patient advisory and updated the patient safety information on the package insert for clopidogrel about this drug interaction.¹² Specifically, the FDA warns that omeprazole reduces the antiplatelet effect of clopidogrel by about 50%. The FDA warning sparked debate in the medical community, as the decision was based in part on ex vivo data.

POST HOC ANALYSES FROM RANDOMIZED CONTROLLED TRIALS

Several post hoc analyses of large randomized clinical trials have studied the potential interaction between proton pump inhibitors and clopidogrel.

In the Clopidogrel for the Reduction of Events During Observation (CREDO) trial, clopidogrel reduced the incidence of death, myocardial infarction, or stroke to a similar extent regardless of baseline use of a proton pump inhibitor.¹³

In patients undergoing percutaneous coronary intervention, the Prasugrel in Comparison to Clopidogrel for Inhibition of Platelet Activation and Aggregation—Thrombolysis in Myocardial Infarction 44 (PRINCIPLE-TIMI 44) trial found that those taking a proton pump inhibitor had significantly less platelet inhibition with clopidogrel compared with those not on one.¹⁴ However, patients taking prasugrel (Effient) and a proton pump inhibitor only had a slight trend towards diminished platelet inhibition.¹⁴

The Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel—Thrombolysis in Myocardial Infarction 38 (TRITON-TIMI 38) found that proton pump inhibitors did not influence the long-term outcome of cardiovascular death, myocardial infarction, or stroke for patients on clopidogrel or prasugrel after an acute coronary syndrome.¹⁴ A subanalysis did not reveal any differences between omeprazole or other drugs of this class as to an effect on the primary outcome.

Though informative, the results of these post hoc analyses need to be validated with data from randomized clinical trials.

‘COGENT’ TRIAL HALTED EARLY, BUT PRELIMINARY RESULTS AVAILABLE

The Clopidogrel and the Optimization of Gastrointestinal Events (COGENT) trial was the first randomized clinical study of the effect of the interaction between clopidogrel and omeprazole on cardiovascular and gastrointestinal outcomes.¹⁵ In a double-blind fashion, patients with acute coronary syndromes or undergoing percutaneous coronary interventions were randomized to receive a fixed-dose combination pill containing either clopidogrel and delayed-release omeprazole or clopidogrel alone. All patients also received aspirin.

Unfortunately, the trial was stopped early because the sponsor declared bankruptcy. However, preliminary results revealed no significant difference in cardiovascular outcomes for patients on clopidogrel and omeprazole compared with clopidogrel alone.¹⁵ Furthermore, adverse gastrointestinal events were significantly fewer in patients on clopidogrel and omeprazole.

Thus, omeprazole appears to be safe and may offer gastrointestinal protection to patients on dual antiplatelet therapy, though we need to await publication of the full results.

‘SPICE ’ TRIAL TO EVALUATE POSSIBLE MECHANISMS OF INTERACTION

The Evaluation of the Influence of Statins and Proton Pump Inhibitors on Clopidogrel Antiplatelet Effects (SPICE) trial is a mechanistic study that will evaluate platelet function and genetic polymorphisms in patients on clopidogrel and aspirin after a percutaneous coronary intervention. They will be randomized to statin therapy plus different proton pump inhibitors.¹⁶ Prior concerns about an ex vivo interaction between clopidogrel and certain statins were not validated by clinical data.¹⁷

OUR RECOMMENDATIONS

Based on the current evidence, patients on aspirin and clopidogrel who have an indication for a proton pump inhibitor or who are at risk of gastrointestinal bleeding can continue or start taking a proton pump inhibitor, including omeprazole.

Switching to another proton pump inhibitor is not currently supported by any randomized clinical trial, nor is changing to a histamine H₂-receptor antagonist. The effect of proton pump inhibitors other than omeprazole on clopidogrel is unclear, and it is not known if the interaction with clopidogrel is a class effect or specific to certain drugs of this class.¹⁸ On the other hand, we still have no compelling evidence of any major clinical interaction between alternative proton pump inhibitors and clopidogrel.¹⁸

Also, separating the dosing times of clopidogrel and omeprazole by 12 hours is not supported by any randomized clinical trial, and runs contrary to at least some ex vivo data.

It is important that all physicians assess the need for a proton pump inhibitor in their patients, as overuse of these drugs has been documented in certain settings.¹⁹

Clopidogrel and omeprazole share a common metabolic link via CYP2C19. Omeprazole, along with some other proton pump inhibitors, interacts with clopidogrel at the level of the CYP450 system. Platelet function studies show that platelet inhibition by clopidogrel is impaired, though the astute clinician should be aware of the wide variability associated with platelet function assays and clopidogrel.^1,20 However, what may appear to be an interaction at the enzymatic level does not necessarily translate into worse clinical outcomes. Additionally, reliance on nonrandomized studies rather than on randomized clinical trials can be misleading.