Benefits and risks of switching from protease inhibitors to nevirapine with stable background therapy in patients with low or undetectable viral load: a multicentre study

Abstract

Combination antiretroviral therapy with HIV-1 protease inhibitors (PI) has changed the natural history of HIV infection, but PI have also been associated with a number of metabolic side-effects, including glucose intolerance, hyperlipidaemia and lipodystrophy [1,2]. Furthermore, currently available PI require multiple daily doses and pills, with consequent patient discomfort and problems of adherence. Induction–maintenance strategies with a reduction of antiretroviral agents fail to maintain viral suppression [3–5]. On the contrary, a simplification strategy with a switch from PI to non-nucleoside reverse transcriptase inhibitors (NNRTI) has been successfully attempted in some case series [6,7]. NNRTI have potent antiviral activity, a more favourable dosing schedule and seemingly fewer metabolic side-effects than PI [8,9]. To investigate further the safety and effectiveness of treatment simplification by switching from PI to nevirapine in patients with low or undetectable viral load, we retrospectively analysed records from two centres. NNRTI-naive patients were selected who had switched from PI to nevirapine, without changing nucleoside reverse transcriptase inhibitors, with viral loads of less than 500 copies/ml. A total of 53 HIV-infected patients fulfilled these criteria, and started nevirapine 200 mg a day for 2 weeks and 200 mg twice a day subsequently. Oral prednisone 25–50 mg a day for the first 2 weeks was given as prophylaxis for hypersensitivity reactions in 27 patients (51%). The median age at switch was 39 years, 40% of patients were women; transmission categories were heterosexual (40%), homosexual (34%) and intravenous drug users (26%). Patients had a median experience of five antiretroviral agents and 28.5 months’ exposure. The median PI experience was 18 months. Patients were taking the following PI before switching: indinavir (36), ritonavir (seven), ritonavir plus saquinavir (hard gel) (four), nelfinavir (three), and saquinavir (hard gel) (three); furthermore, 10 patients were taking zidovudine, 43 stavudine, five didanosine and 48 lamivudine. The median CD4 cell count was 429 cells/μl and HIV RNA was less than 50 copies/ml in 89% and 50–490 copies/ml in the remaining patients. Viral load was less than 500 copies/ml after a median of 18 weeks (range 4–58). Side-effects before PI discontinuation were: lipodystrophy (28 patients: peripheral fat wasting plus abnormal fat accumulation in 13, peripheral fat wasting alone in 13, abnormal fat accumulation alone in two), gastroenteric side effects (25), hyperlipidaemia (38), hyperbilirubinaemia (five), diabetes (three), nephrolithiasis (five), paraesthesias (two), hair loss (one). In nine patients the reason for the switch was their requirement for treatment simplification. Six patients (11%) presented adverse events that required nevirapine discontinuation, with five (9%) definitive discontinuations: four skin rashes, including one Stevens–Johnson's syndrome, one hepatitis and one severe methadone withdrawal. The use of prednisone was not associated with a reduction of skin rash and of discontinuation. All the remaining patients reported a better tolerability of nevirapine over PI. Gastroenteric side-effects, hyperbilirubinaemia, nephrolithiasis, paraesthesias and hair loss resolved in all patients after PI discontinuation. Three months after the switch, mean plasma levels of triglycerides were significantly reduced (see Table 1); fasting total cholesterol was reduced in a subgroup with higher baseline levels (> 280 mg/dl), whereas those with normal baseline values showed a significant increase. Fasting glucose levels were significantly reduced in patients with baseline levels of over 90 mg/dl. After a median follow up of 8 months (19 patient-years), 70% of patients with lipodystrophy reported a perceived improvement of their body shape, substantiated by an improved waist : hip ratio.Table 1: Baseline and follow-up metabolic, virological and immunological parameters in patients switching from protease inhibitors to nevirapine. The mean CD4 cell counts significantly increased 3, 6 and 9 months after the switch to nevirapine (see Table 1). The presence of lipodystrophy at treatment switch was significantly associated with a shorter time to virological rebound [viral load > 50 copies/ml during follow-up; log rank P = 0.01; odds ratio (OR) = 8.6, 95% confidence intervals (CI) 1.1–69.1]. Having experienced five or more drugs before the switch to nevirapine showed a trend towards an association with earlier viral rebound (P = 0.07; OR = 3.6, 95% CI 0.8–17.5), whereas adverse events to nevirapine and having a viral load between 50 and 500 copies/ml at switch were not associated with virological failure. Our findings confirm that a switch from PI- to nevirapine-containing combinations in patients with side-effects or perceived difficulties in undergoing PI therapy results in improved tolerability of the treatment in most cases. Some of the metabolic disorders show improvement but not resolution, at least in the medium term, whereas CD4 cell counts continue to increase over time. Some patients show increased viral replication, but the lack of controls does not allow a comparison with patients continuing PI therapy. Until more findings from prospective, controlled studies become available [10], individuals with a previous exposure to a large number of antiretroviral agents should undergo simplification with nevirapine with caution, and the potential benefits must be carefully weighed against the risk of adverse reactions. The finding of an association between baseline lipodystrophy and subsequent viral rebound prompts an investigation of the effects of lipodystrophy on adherence and pharmacokinetics. Andrea De Lucaa Francesco Baldinia Antonella Cingolania Simona Di Giambenedettoa Mauro Zaccarellib Valerio Tozzib Adriana Ammassaria Rita Murria Andrea Antinorib