icon star paper   Articles  
Back grey_arrow_rt.gif
Male Sexual Dysfunction Associated With Antiretroviral Therapy
  Amy Colson and researchers from Boston reported in the Journal of AIDS (Vol 30, No 1, Mai 1, 2002) on male sexual dysfunction and antiretroviral therapy.
To determine whether treatment with protease inhibitors (PIs) is associated with male sexual dysfunction, we conducted a retrospective, cohort study of 254 adult male PI recipients who received care from the staff-model division of a large managed care organization in New England between 1993 and 1998. After a median of 5.0 years of observation, 80 incident cases of sexual dysfunction were observed. Relative to unexposed individuals, the rate of sexual dysfunction adjusted for confounding was most elevated with use of ritonavir (hazard ratio [HR], 2.83; 95% confidence interval [CI], 1.34-5.97; p = .006) followed by indinavir (HR, 1.69; 95% CI, 0.84-3.37; p = .14), nelfinavir (HR, 1.53; 95% CI, 0.66-3.54; p = .32) and saquinavir (HR, 1.25; 95% CI, 0.53-2.96; p = .60). We conclude that PIs, especially ritonavir, appear to increase the risk of sexual dysfunction.
Editorial note: below the author's discuss factors for sexual dysfunction besides HIV therapy, with particular note regarding depression. But other factors examined are alcoholism, diabetes, and hypertension. Depression may reflect a number of factors including a feeling of being stigmatized, lipodystrophy, worry, stage of HIV diseaseand other psychiatric-related factors.
Protease inhibitor (PI)-based combination antiretroviral therapy has been associated with metabolic and endocrinologic adverse effects that were not apparent during premarketing clinical trials, such as hyperlipidemia, insulin resistance, and body fat redistribution (1). More recently, preliminary reports have suggested an increased incidence of sexual dysfunction among men treated with PIs (2-5). To investigate the potential association between PI therapy and sexual dysfunction, we conducted a retrospective cohort study of male PI recipients.
Study Setting
Harvard Vanguard Medical Associates (HVMA), a multispecialty group practice affiliated with Harvard Pilgrim Heath Care, a large New England health maintenance organization, was the source of the study population. All outpatient encounters at HVMA are recorded in a computerized medical record system, which includes vital signs, providers' notes in their entirety, and laboratory test results. All diagnoses, prescribed treatments and laboratory tests are also coded, which facilitates case and exposure identification (6). Records of prescriptions filled at HVMA pharmacies are captured in a computerized pharmacy record, which is fully linkable to the computerized medical record.
We identified a total of 254 men who met the age, membership, and pharmacy-use eligibility criteria for this study. Three patients were subsequently excluded from the cohort because they had documented sexual dysfunction before January 1, 1993. the median age at the start of observation was 37 years (range, 21-68 years). The median observation time per patient was 5.0 years (3.3-5.8 years). Overall, the cohort had advanced HIV disease; 82% of study participants had a CD4+ lymphocyte count <200 cells and 82% had a serum HIV-1 RNA > 10,000 copies/mL at the start of protease inhibitor therapy. Comorbid conditions associated with sexual dysfunction were relatively common. Thirty-five percent of study subjects had a coded diagnosis of depression at some point during the observation period, 14% hypertension, 12% alcoholism, and 7% diabetes mellitus. Indinavir was the most frequently prescribed PI (62% of cohort members) followed by nelfinavir (46%), ritonavir (46%), and saquinavir (41%). Ritonavir and saquinavir were usually prescribed together as was the case in 79 patients (31% of the cohort). Among the nucleoside analogues, lamivudine (3TC), stavudine (d4T), and zidovudine (ZDV) were prescribed to at least three quarters of the study population. Nonnucleoside reverse transcriptase inhibitors were the least frequently used drug class during the study period (27% of cohort).
A total of 80 patients (32%) had at least one presentation with sexual dysfunction confirmed by chart review during the study period. As shown in Table 2, the unadjusted rate of sexual dysfunction was elevated during any protease inhibitor therapy (hazard ratio [HR], 2.08; 95% confidence interval [CI], 1.23-3.50; p = .006). When the four noninvestigational PIs were examined individually, ritonavir was significantly associated with sexual dysfunction (HR, 2.65; 95% CI, 1.28-5.50; p = .009) whereas indinavir (HR, 1.81; 95% CI, 0.96-3.42; p = .07), nelfinavir (HR, 1.52; 95% CI, 0.69-3.33; p = .30), and saquinavir (HR, 1.19; 95% CI, 0.50-2.82; p = .30) had HRs >1 but were not statistically significant (p >.05).
As a class, nucleoside reverse transcriptase inhibitors also exhibited a possible, unadjusted association with sexual dysfunction (HR, 1.66; 95% CI, 0.98-2.83; p = .06). However, only 3TC was statistically significant (HR, 2.04; 95% CI, 1.05-3.95; p = .03) when the individual nucleoside reverse transcriptase inhibitors were examined . No crude association with sexual dysfunction was found for nonnucleoside reverse transcriptase inhibitors (HR, 1.17; 95% CI, 0.45-3.06; p = .75). The rate of sexual dysfunction was also increased among participants who had a diagnosis of depression (HR, 2.27; 95% CI, 1.46-3.52; p = .0003). Laboratory markers of HIV disease progression, including CD4 count and HIV-1 RNA level, were not significantly associated with sexual dysfunction in our cohort.
After adjustment for age, CD4 count, and diagnoses of depression, alcoholism, hypertension, and diabetes, PI treatment remained significantly associated with an initial presentation of sexual dysfunction (HR, 2.22; 95% CI, 1.24-3.95; p = .007) (Table 3). In contrast, treatment with nucleoside reverse transcriptase inhibitors as a class was no longer associated with sexual dysfunction after adjusting for confounding by other factors including PIs (HR, 1.22; 95% CI, 0.69-2.18; p = .49).
When the individual PIs and nucleoside reverse-transcriptase inhibitors were examined in a fully adjusted model, HRs for ritonavir, indinavir, nelfinavir, and saquinavir remained largely unchanged, as seen in Table 3. Likewise, the association of sexual dysfunction with 3TC was only slightly attenuated (HR, 1.92; 95% CI, 1.00-3.68; p = .05) after adjusting for PI use and the other possible confounding factors. Among the nonantiretroviral variables, depression remained a significant risk factor for documented sexual dysfunction after adjusting for confounding. No interactions between concurrent exposure to ritonavir and saquinavir (p = .78) nor to 3TC and ZDV (p = .47) were observed.
Among those participants whose data included hormone level measurements, no significant differences in serum total testosterone, free testosterone, luteinizing hormone, prolactin, or TSH were observed between those with and without documented sexual dysfunction. However, the number of patients with data on available hormone levels was too small to permit a full analysis of the possible role of hormonal alterations in the development of antiretroviral-therapy-associated sexual dysfunction. In addition, the prevalence of exposure to prescription androgens and other anabolic steroids among all PI recipients during our observation period was only 10%, which suggests the relationship between antiretroviral therapy and sexual dysfunction was not meaningfully confounded by the use of exogenous anabolic steroids. This observation is further supported by preliminary analyses that indicate a similar prevalence of exposure to prescription androgens and other anabolic steroids among men who ultimately presented with sexual dysfunction and those who did not (results not shown).
HIV-1 RNA load itself was also not associated with sexual dysfunction after adjusting for potential confounders.
Several studies conducted before the current era of potent combination antiretroviral therapy have suggested that sexual dysfunction is common among men with HIV infection, especially individuals with advanced-stage disease. It is probable that a number of pathophysiologic disturbances contribute to sexual dysfunction in this population including neurologic, endocrinologic, psychiatric, and infectious factors (10). In particular, impaired production of testosterone, a hormone essential to normal male sexual function, has been associated with HIV infection and the prevalence of this deficiency appears to increase with advancing disease stage (11-18).
PI-based combination regimens came into widespread use for the treatment of HIV infection in early 1996 and resulted in impressive declines in HIV-related morbidity and mortality (19-22). Because studies conducted before the PI era suggested that the incidence of sexual dysfunction increased with stage of HIV disease, a decline in sexual dysfunction among patients treated with effective therapy was expected. However, preliminary reports have suggested the opposite (2-5). Martinez et al. (2) reported new-onset erectile dysfunction with or without loss of libido during treatment with a PI in combination with nucleoside-analogue reverse-transcriptase inhibitors in 14 young (mean age, 37 years), eugonadal men with well-controlled HIV infection (viral load undetectable in 11 of 14) (2). A subsequently reported survey of 115 male PI recipients found that 22% experienced erectile dysfunction despite well controlled HIV disease (median CD4 count, 0.493 109 cells/L, HIV-1 RNA <80 copies/mL in 56% of patients), a median age of 37 years, and normal serum total testosterone levels in all 8 patients for whom testosterone measurements were available (5). In addition, an anonymous survey of 1141 patients with HIV infection found that PI-experienced men and women reported significantly higher rates of decreased sexual interest compared with PI-inexperienced men and women (48% vs. 32%; p < .05). The rate of impotence was also higher among the male PI recipients than nonrecipients (44% vs. 27%; p < .01). (4).
Results from the present investigation support an association between the PIs, especially ritonavir, and sexual dysfunction among men with HIV-1 infection, which is not explained by known confounders or case-detection bias related to the marketing of sildenafil after March 1998. Nevertheless, our study is limited by its retrospective design, which required us to obtain relevant exposure and outcome data from preexisting clinical records rather than by using a specifically designed questionnaire, which would have been possible with a prospective study. This may have resulted in underdetection of cases of sexual dysfunction. Moreover, it is possible that decreased fear about infecting a potential sexual partner or an improved sense of well-being resulting from effective combination antiretroviral therapy may have caused patients to focus more on issues of sexual function. This could have led to increased presentations with sexual dysfunction during treatment with combination antiretroviral therapy and an apparent association between PIs and sexual dysfunction. Against this interpretation, however, are the lack of associations between sexual dysfunction and higher CD4 count or lower HIV-1 viral load (indicators of treatment response) in this study population and the lack of a significant attenuation of the effect estimates after adjustment for CD4 count.
The possible association between 3TC and sexual dysfunction was unexpected and cannot be fully attributed to confounding by concomitant prescribing of PIs found in the automated pharmacy records because adjustment for concurrent use of PIs only minimally reduced the observed association.
In summary, the results of this study support previous observations that PIs may interfere with normal sexual function in men. Furthermore, the risk of sexual dysfunction may vary with different PIs. Prospective investigation is needed to clarify the relationship between antiretroviral therapy and sexual dysfunction in men as well as to examine the impact of antiretroviral therapy on sexual function in women.
  icon paper stack View Older Articles   Back to Top   www.natap.org