Risk Factors for Venous Thromboembolism in Patients with Human Immunodeficiency Virus Infection

Katie L. Kiser, Pharm.D.; Melissa E. Badowski, Pharm.D.


Pharmacotherapy. 2010;30(12):1292-1302. 

In This Article

Host Risk Factors

Many established factors increase the risk for VTE in the general population; however, the risk is further increased in the presence of HIV. Several factors—sex, age, protein S deficiency, protein C deficiency, and the presence of antiphospholipid antibodies—are thought to be associated with VTE in patients with HIV.


The incidence of VTE in the general population does not appear to differ significantly by sex, based on epidemiologic studies. Women do have several specific risk factors including use of oral contraceptives, hormone replacement therapy, and pregnancy that increase their risk. However, the balance of risk in men seems to stem from an increase in risk from non–male-specific risk factors such as trauma.[22] For patients with HIV, evidence supports a nonsignificant increased risk in the incidence of VTE in men. In one study, an almost 2-fold increased risk for VTE was found in males compared with females, although the difference was not significant (annual incidence 0.72% vs 0.36%, relative risk 0.5, 95% confidence interval [CI] 0.1–2.2%).[18] Another study conducted in over 42,000 patients with HIV also found a nonsignificant increased risk in incidence of VTE in males versus females (odds ratio 1 vs 0.8, 95% CI 0.5–1.3).[12] Based on the available data, sex of the patient should not be included as a risk factor for VTE risk assessment in patients with HIV.


In the general population, advancing age (> 40 yrs) is a known risk factor for thrombosis.[22] However, disparate evidence exists regarding the association between age and venous thrombosis in patients with HIV. One study reported the median age at the time of venous thrombosis in patients with HIV to be 43 years, which is nearly 20 years younger than the median age for patients without HIV.[11,23] Similarly, another study established that patients with HIV who were younger than 50 years had a significantly higher rate of VTE/year compared with age-matched, healthy controls (3.31% vs 0.53%, p<0.0001).[6] Another group, however, reported that HIV-infected patients aged 45 years or older were at a significantly higher risk for thrombosis compared with those aged 14–44 years (adjusted odds ratio [OR] 1.9, 95% CI 1.4–2.7).[12] Results of this study correlate well with what is understood about venous thrombosis risk in the general population—that is, that risk increases with increasing age. However, the potentially large cohort of patients who were adolescents could significantly impact this result. The other limited baseline population characteristics reported in these studies were similar.

The evidence about age as a risk factor for VTE in patients with HIV may also be conflicting due to an overwhelming number of other risk factors that may or may not be present in this population. It is reasonable to conclude that advancing age remains a standard risk factor for VTE; however, evidence supports that there may also be an increased risk for thrombosis in patients with HIV at a younger age compared with the general population.

Protein S Deficiency

In the coagulation cascade (Figure 1), protein S serves as a cofactor, along with activated protein C, in the inactivation of factors V and VIII. Protein S can be found in free and bound form. Only the free form is available as a cofactor to activated protein C. Protein S deficiency is defined when only 25–50% of protein S activity is remaining.[24] The frequency of free protein S deficiency is shown in Table 1.[25–27] The increased frequency in patients with HIV is thought to correlate to a 6-fold increased risk for VTE compared with the general population.[25–27] In the HIV population, 27–76% had protein S deficiency (both free and bound), with 12% of those patients having a VTE.[10,18,25–33]

Figure 1.

Coagulation cascade.

Type III protein S deficiency is the most common presentation in patients with HIV.[26,34] This type of deficiency is characterized by a normal total protein S level with a decrease in both free protein S and functional protein S activity. There are many proposed mechanisms behind the functional protein S deficiency in patients with HIV. Examples include reduced synthesis by the liver, abnormalities in endothelial cell function, irregular division between free and bound protein S, abnormal activation of the coagulation cascade, and free protein S interaction with other immune complexes in the plasma.[15,25] A mechanism that has specific supportive data in patients with HIV is the abnormal changes in endothelial cell function.[29] This occurs when the free functional protein S binds to these abnormal endothelial cells, thus resulting in decreased free protein S available in the plasma.[35]

There is conflicting evidence as to whether the decrease in free protein S may be secondary to other mechanisms including an increased binding of free protein S to C4b binding protein (C4bBP), severity of HIV infection, or immuno-suppression represented by CD4+ cell count. Normally, 60% of protein S is bound to C4bBP, and the remaining 40% is free in the plasma. During inflammation, the amount of C4bBP can increase and thus bind more protein S causing the percentage of free protein S to decrease.[27] Most studies in patients with HIV have not found a significant increase in C4bBP;[26,28,31,33] however, one study did report an inverse relationship between C4bBP and free protein S levels in patients with HIV.[32] Table 2 provides a summary of patient characteristics from studies related to protein S deficiency in patients with HIV. Free protein S decreased significantly and progressively with decreasing CD4+ cell count, and C4bBP inversely increased with decreasing CD4+ cell count. The authors concluded that the C4bBP level could be acutely elevated in most immunocompromised patients, thus resulting in increased binding of free protein S.[32] The evidence here, again, is complicated by the many risk factors present in patients with HIV or AIDS. With regard to immunosuppression and CD4+ cell count, one study confirmed that free protein S activity was significantly decreased (37.6% vs 69.8%, p=0.0001) in patients with AIDS versus patients with HIV.[25] Similarly, another study confirmed that free protein S deficiency was significantly higher in patients with AIDS versus those with HIV (100% vs 53%, p<0.001).[34] When specifically looking at CD4+ cell count, another group discovered a trend of decreasing free protein S activity of 71.9%, 50.9%, and 54.6% for decreasing CD4+ cell counts of greater than 400, 200–399, and less than 200 cells/mm3, respectively.[32] The free protein S level was significantly lower for both CD4+ cell less than 200 and from 200-399 cells/mm3 as compared with CD4+ cell greater than 400 cell/mm3. These results differ from those found in three other studies that did not suggest a linear correlation between decreases in protein S and CD4+ cell count or protein S deficiency and CD4+ cell count.[26,27,30] One of the studies was conducted in an HIV population, most of whom had AIDS,[27] whereas less than half of the participants in the other studies were classified as having AIDS.[26,30] Most studies do support a correlation between decreased free protein S activity and decreased CD4+ cell count.[25,28,32,36]

Taking the conflicting evidence into account, the results of these studies do not support that the protein S deficiency observed in patients with HIV or AIDS is secondary to an acute elevation of C4bBP level. The evidence does support a correlation between immunosuppression and decreased CD4+ cell count with protein S deficiency.

Protein C Deficiency

In the clotting cascade, protein C becomes activated and serves to inactivate factors V and VIII. Protein C deficiency can present in two forms: a decreased amount of available protein C and a normal amount of protein C but a decrease in the functionality. The frequency of protein C deficiency is shown in Table 1 and can be as high as 25% during an acute opportunistic infection.[27] The frequency of VTE in the setting of protein C deficiency in patients with HIV is not clear; however, one study documented 28 patients with AIDS who had 34 venous thromboembolic events over a 42-month span of time.[10] These patients had no other noted risk factors for VTE, and 2 (7%) of these 28 patients were found to have protein C deficiency.[10] Table 2 provides more background information on studies related to protein C deficiency in patients with HIV.

Similar to the mechanism behind protein S deficiency, protein C deficiency in the setting of HIV disease is thought to be secondary to immunosuppression. One study established a direct correlation between decreased CD4+ cell count and decreased protein C activity.[32] With a normal reference range of 60–140% for protein C activity, patients with a CD4+ cell count of less than 200 cells/mm3 had a protein C activity of 81.8%, those with a CD4+ cell count of 200–399 cells/mm3 had 102.7%, and those with a CD4+ cell greater than 400 cells/mm3 had 110.3%. Those with a CD4+ cell count of less than 200 cells/mm3 had protein C activity significantly less than those with CD4+ cell counts greater than 400 cells/mm3 (p<0.005). This evidence supports the theory that functional protein C declines with immunosuppression, thus increasing the risk of VTE.

Antiphospholipid Antibodies

There are two main antibodies, anticardiolipin and lupus anticoagulant, that are associated with the autoimmune disease antiphospholipid syndrome. The frequency of antiphospholipid syndrome in the general population is 2–4%, and the syndrome is clearly linked to increased risk of venous and arterial thrombosis. However, the relationship between the presence of antibodies and thrombotic risk has not been fully elucidated. Anticardiolipin antibodies have been commonly reported in HIV-positive patients, with a prevalence of 7.7–94%.[27,28,37–45] There are two different types of anticardiolipin antibodies, immunoglobulin (Ig) G and IgM. The literature reports the IgG anticardiolipin antibody to be positive in 41–94% of patients with HIV[38,39,40,42,46] compared with the IgM anticardiolipin antibody, which is reported to be positive in only 7% of patients with HIV.[38,46] As stated above, the clinical relevance of the elevation of anticardiolipin antibodies is controversial. An association between the presence of antiphospholipid antibodies and infection is well documented and thought to be transient and not associated with antiphospholipid syndrome; however, there have also been reports of HIV-inducing pathogenic antibodies.[47] There are both reports of no association [38,39,42,48] and a positive association between having positive anticardiolipin antibodies and thrombosis in patients with HIV.[49–53]

Contrasting with anticardiolipin antibodies, which are often found positive in patients with HIV, lupus anticoagulant is much more variable in its manifestation. The frequency of positive lupus anticoagulant in patients with HIV ranges from 0–72%.[10,27,34,36,54,55] One study evaluated 61 patients with HIV and positive for antiphospholipid antibodies, who could be with or without opportunistic infections and antiretroviral therapy, and compared them with 55 patients without HIV and positive for antiphospholipid antibodies during an acute infectious disease like syphilis, and with 45 patients with well-documented antiphospholipid syndrome.[55] The study established that lupus anticoagulant was present significantly more often in those with HIV and antiphospholipid syndrome compared with those who had syphilis, respectively (72%, 81%, 0%, p<0.0001). They concluded that the lupus anticoagulant activity was secondary to a hyperreactive immune system, was not pathogenic, and did not increase the risk of thrombosis. Although lupus anticoagulant is often associated with an increased risk of thrombosis, no data, to our knowledge, support a direct association between lupus anticoagulant positivity in patients with HIV or AIDS and thrombosis.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: