Objectives We sought to determine the prevalence of phosphodiesterase type 5 inhibitor (PDE-5) mediated drug-drug interactions (DDIs) in males with HIV infection receiving antiretroviral therapy (ART) and identify factors associated with PDE-5-mediated DDIs. Methods Male US Military HIV Natural History Study participants diagnosed with erectile dysfunction (ED) and having a PDE-5 inhibitor and potentially-interacting ART co-dispensed within 30 days were included. DDIs were defined according to criteria found in published guidelines and drug information resources. The primary outcome of interest was overall PDE-5 inhibitor-mediated DDI prevalence and episode duration. A secondary logistic regression analysis was performed on those with and without DDIs to identify factors associated with initial DDI episode. Results A total of 235 male participants with ED met inclusion criteria. The majority were White (50.6%) or African American (40.4%). Median age at medication co-dispensing (45 years), duration of HIV infection (14 years), and duration of ED (1 year) did not differ between the two groups (p>0.05 for all). PDE-5 inhibitors included sildenafil (n = 124), vardenafil (n = 99), and tadalafil (n = 14). ART regimens included RTV-boosted protease inhibitors (PIs) atazanavir (n = 83) or darunavir (n = 34), and COBI-boosted elvitegravir (n = 43). Potential DDIs occurred in 181 (77.0%) participants, of whom 122 (67.4%) had multiple DDI episodes. The median DDI duration was 8 (IQR 1–12) months. In multivariate analyses, non-statistically significant higher odds of DDIs were observed with RTV-boosted PIs or PI-based ART (OR 2.13, 95% CI 0.85–5.37) and in those with a diagnosis of major depressive disorder (OR 1.74, 95% CI 0.83–3.64). Conclusions PDE-5-mediated DDIs were observed in the majority of males with HIV infection on RTV- or COBI-boosted ART in our cohort. This study highlights the importance of assessing for DDIs among individuals on ART, especially those on boosted regimens.

HIV infection is associated with increased risk of erectile dysfunction (ED); however, factors associated with ED remain unclear. We evaluated the prevalence of ED among men living with HIV and factors associated with ED diagnosis in the US Military HIV Natural History Study (NHS). MethodsA retrospective cohort study evaluated participants in the NHS, a cohort of HIV-positive active duty members and beneficiaries with HIV infection. Men with a diagnosis of ED after HIV diagnosis were included. Cohort controls without ED diagnosis were matched 2:1 by age at HIV diagnosis and duration of follow-up. Multivariate logistic regression models were used to identify factors associated with ED. ResultsA total of 543 of 5682 male participants (9.6% prevalence) had a diagnosis of ED, of whom 488 were included in the analysis. The median (interquartile range, IQR) age at ED diagnosis was 43 (37.0-49.0) years and the time from HIV diagnosis to antiretroviral therapy (ART) start was longer for cases (5.0 years, IQR: 2.0-9.0) than for controls (3.0 years, 1.0-6.0; P < 0.01). Cases had higher proportions of multiple comorbid conditions, including depression (33.4% vs. 21.7%), tobacco use (19.7% vs. 9.0%) and sleep apnoea (14.8% vs. 4.2%) compared with controls (P < 0.01 for all). Logistic regression showed increased odds of ED for delayed ART initiation > 4 years [odds ratio (OR) = 2.05, 95% confidence interval (CI): 1.56-2.71], protease inhibitor use ≥ 1 year (OR = 1.81, 95% CI: 1.38-2.38) and sleep apnoea (OR = 2.60, 95% CI: 1.68-4.01). ConclusionsErectile dysfunction was common in men with HIV and associated factors included both HIVrelated and traditional factors.

BackgroundHIV-infected men have increased incidence of erectile dysfunction (ED) compared with men without HIV infection. Risk factors for ED among HIV-positive individuals have not been widely described.MethodsA retrospective cohort study was completed evaluating participants in the US Military HIV Natural History Study, a cohort of HIV-infected active duty members and beneficiaries. Men with a diagnosis of ED after HIV diagnosis were included (n = 488). Cohort controls (n = 976) without ED diagnosis were matched 2:1 by age at HIV diagnosis. Multivariate logistic regression model was used to identify risk factors for ED.ResultsAt HIV diagnosis, the median CD4 count was similar for cases (523 cells/μL, IQR 396–675) and controls (508 cells/μL, IQR 366–673; P = 0.310) and the overall median age was 32 years. At ED diagnosis, cases had a median age of 43 years (IQR 37.0–49.0) and 445 (92.3%) were on antiretroviral therapy (ART). The median time from HIV diagnosis to ART start was longer for cases (5.0 years, IQR 2.0–9.0) compared with controls (3.0 years, IQR 1.0–6.0; P < 0.001). Cases had a higher proportion of the following diagnoses compared with controls (P < 0.001 for all): depression (33.4% vs. 21.7%), hypertension (37.9% vs. 20.4%), hyperlipidemia (54.3% vs. 32.4%), tobacco use (31.1% vs. 23.1%), sleep apnea (14.8% vs. 4.2%) and diabetes/cardiovascular disease (CVD)(10.5% vs. 4.7%). Multivariate logistic regression model is reported below (table).Logistic Regression Model to Predict ED Characteristic Odds Ratio 95% CI P-value Sleep apnea2.621.69–4.05<0.001Time from HIV diagnosis to ART start > mean2.071.58–2.71<0.001African-American race1.760.90–3.420.096Diabetes/cardiovascular disease1.611.01–2.580.048Tobacco use1.420.99–2.040.057Hypertension1.361.02–1.820.034Hyperlipidemia1.260.96–1.640.092Depression1.240.94–1.630.130CD4 count <200 cells/μL at HIV diagnosis0.630.32–1.250.184Prior protease inhibitor use0.430.31–0.60<0.001ConclusionDelay in ART initiation as well as risk factors for and presence of CVD were associated with ED in HIV-infected persons. Mitigating risk factors and optimizing comorbidities is important to improve sexual health and reduce ED in HIV-infected persons.Disclosures All authors: No reported disclosures.