Evaluation of the Efficacy and Safety of Switching to Tenofovir, Emtricitabine, and Rilpivirine in Treatment-Experienced Patients

To the Editors: Rilpivirine (RPV), a recently licensed nonnucleoside reverse transcriptase inhibitor, has shown noninferior efficacy and a favorable safety profile compared with efavirenz (EFV) in treatment-naive HIV-1–infected adults.1,2 In the SPIRIT study, noninferior efficacy has also been observed among patients who switched to an RPV-containing versus protease inhibitor (PI)-containing antiretroviral (ARV) regimen.3 Switching from tenofovir (TDF)/emtricitabine (FTC)/EFV to TDF/FTC/RPV was also shown to be a safe and efficacious option for 49 virologically suppressed HIV-infected patients.4 To date, there has been no confirmation of the results from this trial in real-world settings. Moreover, changes in residual viremia at ultralow detection levels, which could potentially predict virologic failure,5 have never been evaluated after switching to an RPV-containing ARV regimen. The objective of our study was then (1) to assess efficacy and safety of switching to TDF/FTC/RPV among patients whose HIV was already suppressed (below 20 copies/mL) and (2) to determine whether rates of ultrasensitive HIV detection shed light on the efficacy of this treatment strategy. We conducted a prospective study at Saint-Antoine Hospital (Paris, France). From September 2012 to May 2013, all patients with plasma HIV-RNA <20 copies per milliliter switching to TDF/FTC/RPV were consecutively included. Clinical and biological (immunological, virological, liver function, lipids) data were collected at baseline, month-3 (M3), and month-6 (M6). In a subset of patients, ultrasensitive viral loads (US-VL) were measured in plasma by an adapted Cobas/TaqmanHIV-1ultrasensitive assay (quantification limit <1 copy/mL) before and after switching ARV regimens. Virologic failure was defined as 2 consecutive HIV-RNA >20 copies per milliliter. Sequences of protease and reverse transcriptase genes on RNA or DNA at baseline and at the time of virologic failure were determined using ANRS consensus-technique primer sequences, as described at http://www.hivfrenchresistance.org. Statistical analyses were performed with STATA12.1 (College Station, TX), and significance was determined using a P value of <0.05. Signed informed consent to use biomedical data for medical research was obtained from all patients. One-hundred fifty-five patients were included. As indicated in Table 1, RPV was mainly given to patients previously treated with a PI (48.4%), whereas 63.2% switched therapy to simplify their current regimen and 6% (n = 9) had previously documented virologic failure. Of the 118 patients with genotypic information before baseline, resistance mutations had been observed to the following ARVs in 15 different patients: TDF/FTC/RPV (K65R, Q151M, Y188L, G190A, n = 1), FTC/TDF (K70E, M184I, n = 1; K65R, M184V, n = 1; K65R, Q151M, n = 1), FTC and possibly TDF (M41L, E44D, D67N, M184V, L210W, T215Y, n = 1), possibly TDF (M41L, D67N, T69A/D/N, L210W, T215N/S/Y, n = 1), FTC (M41L, A62V, M184I/V, L210W, T215Y, K219R, n = 1; M41L, D67N, K70R, M184V, T215E, K219Q, n = 1; M184I, n = 1), and EFV and nevirapine (NVP) (n = 6, isolated mutation K103N or K103N associated with M184I/V, G190A). All other baseline characteristics are detailed in Table 1.TABLE 1: Clinical, Immunological, and Viral Characteristics of 155 Patients With Viral Load <20 Copies per Milliliter at Baseline and During Follow-upMedian follow-up up was 6.5 months (range: 1.0–12.0). Median CD4 cell count remained stable [baseline: 634/mm3; M3: 590/mm3 (N = 129); M6: 631/mm3 (N = 121)]. No significant changes in mean low-density lipoprotein cholesterol, high-density lipoprotein cholesterol, total cholesterol, and triglycerides were observed. Of patients with available virologic samples during follow-up, 124/134 (92.5%) and 119/125 (95.2%) maintained virologic suppression (HIV-RNA <20 copies/mL) at M3 and M6, respectively. These results were similar for patients previously treated with a PI-containing versus NNRTI-containing ARV regimen (% with HIV-RNA <20 copies/mL: 93.6% and 92.4% at M3, 96.6% and 92.4% at M6, respectively). Among patients with US-VL before and after switch, 64/82 (78.1%) patients at baseline and 61/82 (74.4%) during follow-up had a US-VL <1 copy per milliliter. Both of these proportions were not significantly different when comparing patients actively followed in our department (1903/2601, 73.2%; Pearson's χ2, P = 0.3 and 0.4, respectively). When examining changes after switching to TDF/FTC/RPV, 51/64 (79.7%) patients with US-VL <1 copy per milliliter at baseline maintained US-VL <1 copy per milliliter, whereas 10/18 (55.6%) patients with US-VL between 1 and 20 copies per milliliter at baseline decreased to US-VL <1 copy per milliliter. TDF/FTC/RPV was discontinued in 11 patients (7.1%) after median treatment duration of 4 months (range: 1–6). Three patients previously treated with PI-containing ARV regimen exhibited confirmed virologic failure (1.9%), all of whom had a US-VL <1 copy per milliliter at baseline. Patient 1 had a VL of 64 copies per milliliter at M3 and, before switch, had harbored a resistant HIV strain with the D67N, L74I, V75L, M184I, V90I, and K103N mutations (conferring resistance to TDF, FTC, and NVP/EFV). Patient 2 had a VL of 118 copies per milliliter at M3. The G190A mutation associated with resistance to NVP/EFV had been found from a previous genotypic test at baseline, whereas no new resistance mutations were detected at the time of virologic failure. Patient 3 had a VL of 2518 copies per milliliter at M3. A previous genotypic test found the D67N, K70R, M184V, T215I, K219Q, and K103N mutations conferring resistance to thymidine analogs, FTC, and NVP/EFV at baseline, whereas the patient developed a V179I mutation during the follow-up. A proton-pump inhibitor was inadvertently given to this patient a few days after initiating TDF/FTC/RPV. The remaining 8 patients discontinued TDF/FTC/RPV because of adverse events. The reasons were as follows: elevated transaminases (n = 2), difficulty sleeping (n = 2), gastrointestinal disorder (n = 1), initiating proton-pump inhibitor (n = 1) or anti-HCV therapy (n = 1), and arthralgia (n = 1). In the 2 patients with abnormalities in liver function, 5 and 10-fold than normal increases in transaminases occurred in the first month of treatment and returned to normal levels in the 2 months after interruption of TDF/FTC/RPV. No other causes of cytolysis were found (acute viral hepatitis, syphilis hepatitis, or toxic-related hepatitis). Our study confirms the results of Mills et al and of SPIRIT study among treatment-experienced HIV-infected patients with controlled HIV-RNA replication,3,4 demonstrating largely adequate efficacy and safety. Virologic suppression was maintained for a large majority of patients and only a few discontinued treatment because of adverse events. Very few patients had virologic failure while undergoing TDF/FTC/RPV. Interestingly, both K103N and M184V mutations had been detected years before in 2 of the patients failing this regimen. Isolated K103N has not been shown to be associated with RPV-resistance (http://www.hivfrenchresistance.org; update September 2013, version 23). Furthermore, a recently published study described 3 patients with isolated K103N who maintained virologic success after switching from boosted PI regimen to TDF/FTC/RPV.6 Nevertheless, an additional mutation conferring NRTI resistance could increase the impact of the K103N mutation. The history of previous treatment failure(s) and resistance genotyping should be taken into account when considering switching ARV therapy to avoid “functional” monotherapy. This point has been previously highlighted in another treatment-switch study, particularly for drugs with low genetic barrier to resistance.7 To the best of our knowledge, this study is the first that has evaluated US-VL after switching to RPV. The percentage of patients with undetectable US-VL pre- and postswitch was the same as observed in our infectious disease clinic, whereas at the same time the proportion of patients changing from undetectable to detectable (15.9%) and vice versa (12.2%) were very similar. Taken together, these findings suggest that therapeutic switch does not affect US-VL replication more than what would be expected. Further analysis with more patients and longer periods would be necessary to assess the long-term efficacy of TDF/FTC/RPV on US-VL. Although the clinical impact of US-VL remains debatable, its importance lies with the potential of virologic failure among patients with residual viremia.5,8