Association between sexual disturbances and sexual hormones with specific antiretroviral drugs

Sexual disturbances have been related to protease inhibitor therapy, but the effect of specific protease or non-nucleoside reverse transcriptase inhibitors is largely unknown. We analysed the rate of sexual dysfunction and the sexual hormonal profile in patients undergoing antiretroviral therapy. All antiretroviral drugs were associated with different degrees of sexual dysfunction; the highest rates with indinavir and the lowest with nevirapine. Although these drugs were associated with increases in testosterone and 17β-oestradiol, sexual disturbances were not related to alterations in the sexual hormonal pattern.

Sexual disturbances and hypogonadism were relatively common among HIV-infected patients with advanced disease in the pre-highly active antiretroviral therapy (HAART) era [^1-6]. More recently, sexual dysfunction has been described in patients receiving protease inhibitors (PI) [^7,8]. However, the different antiretroviral agents that could be involved, either PI or other drugs, have not been studied separately. We have investigated the possible association of several PI and non-nucleoside reverse transcriptase inhibitors (NNRTI) with sexual disturbances in a numerous group of HIV-infected patients in stable clinical condition.

Of a large cohort of patients who were evaluated for sexual disturbances, which included erectile dysfunction, decreased libido or impaired ejaculation, we selected those patients who were being treated with two nucleoside reverse transcriptase inhibitors plus either one PI or one NNRTI. The study group consisted of 135 patients, who did not have any acute, severe illness, and who underwent a total of 231 evaluations that included specific questioning about symptoms of sexual dysfunction and determinations of sexual hormones [testosterone, 17β-oestradiol, luteinizing hormone (LH) and follicle-stimulating hormone (FSH)]. Five different antiretroviral drugs were considered, three PI (indinavir, nelfinavir, and saquinavir with or without low-dose ritonavir enhancement), and two NNRTI (nevirapine and efavirenz). The number of evaluations for each of the five drugs studied were: indinavir 34 evaluations, nelfinavir 67, saquinavir (with or without ritonavir) 26, nevirapine 55, and efavirenz 49. The control group consisted of 46 HIV-infected patients who did not receive any antiretroviral treatment and who underwent a total of 53 evaluations. The chi-square test and the Mann-Whitney U test were used for the statistical comparison of proportions and means, respectively. The level of significance was established at P < 0.05 for a two-tailed test.

The mean age of the study group was 37.8 years [95% confidence intervals (CI) 36.9-38.7], the mean period on HAART was 24.6 months (95% CI 23.1-26), the mean CD4 cell count was 479.6 cells × 10⁶/l (95% CI 451.7-507.4), and 80.4% had undetectable (less than 50 copies/ml) viral loads. For the control group these values were: mean age 35.2 years (95% CI 33.8-36.5), mean CD4 cell counts 387.3 cells × 10⁶/l (95% CI 296.4-478.1), undetectable viral load 6.1%.

The rates of sexual disturbances in patients under treatment were 32.4% for indinavir, 28.6% for efavirenz, 26.9% for saquinavir, 22.4% for nelfinavir and 12.7% for nevirapine, whereas the rate for untreated patients was 3.8%. There were no significant differences among the individual drugs except for the comparison of nevirapine with indinavir (P = 0.025) and nevirapine with efavirenz (P = 0.045). Regarding the comparisons with the control group, there were significant differences among untreated patients and patients treated with indinavir (P = 0.0003), nelfinavir (P = 0.004), saquinavir (P = 0.002), and efavirenz (P = 0.0006). Of all drugs under study, only nevirapine was not associated with significant differences with respect to untreated patients (P = 0.09).

Fig. 1 shows the values of testosterone and 17β-oestradiol for each drug and for untreated controls as well as the statistical comparisons among them. As can be seen, the highest values of testosterone were observed in patients receiving saquinavir, and the highest 17β-oestradiol values were seen in patients treated with nevirapine. The lowest values of both hormones corresponded to untreated patients. The levels of sexual hormones were similar in patients complaining of sexual disturbances and in those who did not report such symptoms (mean testosterone 27.9 versus 27.6 nmol/l, P = 0.99; 17β-oestradiol 180.2 versus 185.5 pmol/l, P = 0.84; 17β-oestradiol:testosterone ratio 6.92 versus 7.23, P = 0.95; FSH 4.3 versus 5.3 U/l, P = 0.06; LH 5.6 versus 6.1 U/l, P = 0.2, respectively).

Fig. 1

A recent multicentre study involving men and women found decreased libido in 40% and decreased sexual potency in 34% of the patients under treatment with PI [⁹]. In contrast, another study with a lower number of patients [¹⁰] failed to find differences in the sexual performance of patients treated with PI in comparison with patients treated with other drugs. Both studies found a relationship between sexual dysfunction and indinavir-containing regimens. However, neither reported the frequency of sexual dysfunction for individual PI, analysed the effect of NNRTI or evaluated the sexual hormonal profile of the patients.

Another study [¹¹] found that 9% of their patients had erectile dysfunction, that 70% of the patients with erectile dysfunction were receiving indinavir, and that this drug was predictive of this disorder in a multivariate analysis. However, indinavir was the most commonly prescribed drug in that study and, apparently, indinavir was combined with other PI, and perhaps NNRTI, in an unknown number of patients. Although it was not specified in that report, it seems that 14% of the patients receiving indinavir, either alone or in combination with other drugs, developed erectile dysfunction. We found that indinavir was the drug most commonly associated with diverse sexual disturbances, including erectile dysfunction, involving one-third of the patients treated with this antiretroviral agent. In contrast, the drug associated with fewer sexual disturbances in these patients was nevirapine.

Sexual disturbances do not seem to be related to hormonal alterations. We observed similar values of testosterone, 17β-oestradiol, 17β-oestradiol:testosterone ratio, LH and FSH in patients with and without these symptoms. In addition, serum testosterone was increased, rather than decreased, in patients treated with antiretroviral drugs compared with untreated patients, perhaps as a result of metabolic interactions, whereas the opposite was found with regard to the existence of symptoms. Finally, nevirapine, the drug less commonly associated with sexual disturbances, was also the drug associated with the highest values of oestrogens and the oestrogen:testosterone ratio.

There are many factors related to sexual dysfunction in the general population that also apply to HIV-infected individuals [¹²]. In addition, neuropathic and neuropsychological impairment have been found in a small number of HIV-infected patients with erectile dysfunction [¹¹]. Although combination antiretroviral therapy has resulted in marked improvements in patients' health status, these drugs have been associated with many adverse effects. Sexual disturbances seem to constitute another untoward effect of these drugs not mediated through sexual hormonal alterations.

Julio Collazos

Jose Mayo

Eduardo Martínez

Sofía Ibarra