Virological outcome and management of persistent low-level viraemia in HIV-1-infected patients: 11 years of the Swiss HIV Cohort Study.

Background: Management of persistent low-level viraemia (pLLV) in patients on combined antiretroviral therapy (cART) with previously undetectable HIV viral loads (VLs) is challenging. We examined virological outcome and management among patients enrolled in the Swiss HIV Cohort Study (SHCS). Methods: In this retrospective study (2000-2011), pLLV was defined as a VL of 21-400 copies/ml on >= three consecutive plasma samples with >= 8 weeks between first and last analyses, in patients undetectable for >= 24 weeks on cART. Control patients had >= three consecutive undetectable VLs over >= 32 weeks. Virological failure (VF), analysed in the pLLV patient group, was defined as a VL>400 copies/ml. Results: Among 9,972 patients, 179 had pLLV and 5,389 were controls. Compared to controls, pLLV patients were more often on unboosted protease inhibitor (PI)- based (adjusted odds ratio [aOR; 95% CI] 3.2 [1.8, 5.9]) and nucleoside/nucleotide reverse transcriptase inhibitor NRTI)-only combinations (aOR 2.1 [1.1, 4.2]) than on non-nucleoside reverse transcriptase inhibitor and boosted PI-based regimens. At 48 weeks, 102/155 pLLV patients (66%) still had pLLV, 19/155 (12%) developed VF and 34/155 (22%) had undetectable VLs. Predictors of VF were previous VF (aOR 35 [3.8, 315]), unboosted PI-based (aOR 12.8 [1.7, 96]) or NRTI-only combinations (aOR 115 [6.8, 1,952]), and VLs>200 during pLLV (aOR 3.7 [1.1, 12]). No VF occurred in patients with persistent very LLV (21-49 copies/ml; n=26). At 48 weeks, 29/39 patients (74%) who changed cART had undetectable VLs, compared with 19/74 (26%) without change (P<0.001). Conclusions: Among patients with pLLV, VF was predicted by previous VF, cART regimen and VL >= 200. Most patients who changed cART had undetectable VLs 48 weeks later. These findings support cART modification for pLLV>200 copies/ml.