Original article Efficacy and safety of a switch to rilpivirine-based regimens in treatment-experienced HIV-1-infected patients: a cohort study

Transcription

1 Antiviral Therapy 2016; 21: (doi: /IMP3010) Original article Efficacy and safety of a switch to rilpivirine-based regimens in treatment-experienced HIV-1-infected patients: a cohort study Sandrine Gazaignes 1, Matthieu Resche-Rigon 2,3, Caroline Gatey 1, Chloe Yang 4, Blandine Denis 1, Julien Fonsart 5, Kristell Desseaux 2,3, Michel Guionie 6, Willy Rozenbaum 1, Constance Delaugerre 7,8, Jean Michel Molina 1,7 * 1 Department of Infectious Diseases, Saint-Louis Hospital, Assistance Publique Hôpitaux de Paris, University of Paris Diderot, Paris, France 2 Biostatistics and Computer Medicine, Saint-Louis Hospital, Assistance Publique Hôpitaux de Paris, University of Paris Diderot, Paris, France 3 ECSTRA Team INSERM, UMR 1153, Paris, France 4 Department of Medicine, University of Toronto, Toronto, ON, Canada 5 Department of Biochemistry, Saint-Louis Hospital, Assistance Publique Hôpitaux de Paris, Paris, France 6 COREVIH Ile de France Est, Paris, France 7 INSERM UMR 941, Paris, France 8 Department of Virology, Saint-Louis Hospital, Assistance Publique Hôpitaux de Paris and University of Paris Diderot, Paris, France *Corresponding author jean-michel.molina@aphp.fr Background: Rilpivirine (RPV) is a second-generation once-daily non-nucleoside reverse transcriptase inhibitor (NNRTI) which has shown non-inferior antiviral activity to efavirenz in treatment-naive patients. Data in treatmentexperienced patients are more limited. We wished to assess the efficacy and safety of a switch to RPV-based regimens in well-suppressed treatment-experienced patients. Methods: Between September 2012 and June 2013, all antiretroviral therapy (ART)-experienced HIV-1-infected patients with a plasma HIV RNA level <50 copies/ml, and switching to an RPV-based regimen, were analysed in this retrospective observational monocentric cohort study. The primary end point was the proportion of patients with virological success defined as a plasma HIV RNA level <50 copies/ml at 12 months using the FDA snapshot algorithm. Results: A total of 281 participants were studied and 97% received a combination of RPV/tenofovir disoproxil fumarate/emtricitabine. At month 12, the rate of virological success was 59% and increased to 72% using available data beyond month 12. Sixteen (6%) patients experienced virological failure, which was associated with the presence of the M184V/I resistance mutation in prior genotypes (P=0.02) and the use of a non-nnrti as third agent before the switch (P=0.03). RPV-based regimens were overall well tolerated and only 23 (8%) patients discontinued ART because of adverse events, mostly neuropsychiatric adverse events. Switching to RPV was associated with significant but modest improvement of the lipid profile. Conclusions: In patients fully suppressed on ART, a switch to an RPV-based regimen should only be considered in the absence of prior virological failure or resistance mutations to nucleoside reverse transcriptase inhibitors and NNRTIs to avoid virological failures. Introduction Rilpivirine (RPV) is second-generation non-nucleoside reverse transcriptase inhibitor (NNRTI) that has recently been approved for the treatment of HIV-1 infection [1]. RPV is taken once daily with meals as a single tablet in combination with other antiretroviral drugs or as a fixeddose combination with emtricitabine (FTC) and tenofovir disoproxil fumarate (TDF). RPV approval was granted on the basis of pooled 96-week efficacy and safety analysis of two large placebo-controlled Phase III clinical trials, ECHO and THRIVE, which demonstrated the non-inferiority and better tolerability of RPV compared with efavirenz (EFV) for first-line treatment in patients with a baseline viral load below 100,000 copies/ml [2 4]. These results were confirmed in a third open-label study with RPV/FTC/TDF fixed-dose combination, the STAR trial [5] International Medical Press (print) (online) 329

2 S Gazaignes et al. Given that RPV is well tolerated and convenient to take, it is also an attractive option for virologically suppressed treatment-experienced patients willing to switch their current treatment regimen to improve tolerability or convenience. Data with RPV in treatmentexperienced patients have been thus far limited and a single randomized trial has shown that the combination of RPV/FTC/TDF was able to maintain virological suppression in patients switching from a ritonavir-boosted protease inhibitor-based (PI/r) antiretroviral therapy (ART) [6 13]. The purpose of our study was to assess the safety and efficacy of switching to an RPV-based regimen in treatment-experienced patients who are virologically suppressed in real life conditions. Methods Study design and population We performed a single-centre, observational, retrospective cohort study enrolling all ART-experienced adult HIV-1-infected patients with plasma HIV-1 RNA level <50 copies/ml switching to an RPV-based regimen during the time period from 1 September 2012 (date of RPV approval in France) to 18 June 2013, and followed at the Saint-Louis Hospital in Paris, France. There were no pre-defined criteria for switching to an RPV-based regimen and this decision was left to each physician s discretion. We were however willing to prospectively collect data for each patient switching to RPV in order to assess the safety and efficacy of this switch in a reallife setting. Clinical and biological data were collected prospectively at the time of switch (baseline [BL]), and up to month 12 (M12) after the switch using the Nadis electronic database [14]. Data collected included reasons for the switch, patient demographics, route of HIV infection, duration of HIV infection, CD4 + T-cell counts, duration of undetectable plasma HIV-1 RNA levels before the switch, results of prior genotypic resistance tests when available with detection of resistance associated mutations (RAMs) to NNRTIs and nucleoside reverse transcriptase inhibitors (NRTIs), treatment history (prior virological failure [VF] on an NNRTI + NRTIs, exposure to prior regimens with NNRTIs or NRTIs), time from ART initiation to switch, hepatitis B or C coinfection, antiretroviral drugs used before the switch and ART used in combination with RPV after the switch. Laboratory parameters included full blood cell counts, CD4 + T-cell count, plasma HIV-1 RNA, alanine aminotransferase (ALAT), aspartate aminotransferase (ASAT), total cholesterol, high-density lipoprotein (HDL) and low-density lipoprotein (LDL) cholesterol, triglycerides, and creatinine levels, and estimated glomerular filtration rate (egfr) using the Cockroft Gault formula. These BL parameters were either collected at the time of the switch or at the closest visit before the switch. All participants provided written informed consent prior to inclusion in the Nadis Database which was approved by the Paris Saint-Louis ethics committee (CPP Paris Saint-Louis) and the CNIL (Commission Nationale Informatique et Libertés). Treatment-naive patients, those who had received RPV prior to the study period (in the setting of clinical trials), and patients with a plasma HIV-1 RNA level >50 copies/ml at BL were excluded from this study. Efficacy assessments The primary end point was the proportion of patients with a plasma HIV-1 RNA level <50 copies/ml at M12 using the FDA snapshot analysis. VF was defined as a confirmed HIV-1 RNA level >50 copies/ml or an HIV-1 RNA level >50 copies/ml followed by RPV discontinuation. Secondary efficacy end points included the proportion of patients with a plasma HIV-1 RNA level <50 copies/ml over time, the change in CD4 + T-cell counts from BL, and the emergence of RAMs in patients with VF. In case of VF, a genotypic test was performed on the confirmatory sample to detect emerging resistance mutations and measure RPV plasma concentrations to assess drug adherence. Historical genotypic resistance tests were collected at screening and analysed using the 2013 ANRS resistance group algorithm. RPV concentration was assessed on plasma samples after protein precipitation by liquid chromatography tandem mass spectrometry on ThermoFisher Scientific U3000RS coupled with a TSQ Quantum Ultra (Thermo Scientific, Villebon-sur-Yvette, France). Lower limit of quantification was 4 ng/ml with a lower limit of detection of 1 ng/ml and linearity from 4 to 400 ng/ml. Mean trough RPV plasma concentrations in the ECHO and THRIVE study ranged from 50 to 80 ng/ml [15]. Safety assessments The primary safety end point was the proportion of patients who discontinued RPV for safety reasons. Safety was investigated by examining adverse events (AEs) and laboratory abnormalities up to M12. The severity of AEs was evaluated using the 2008 ANRS grading scale for AEs [16]. Statistical analysis The primary analysis was an intent-to-treat (ITT) analysis including all patients who met the inclusion criteria, using the FDA snapshot algorithm, at M12. Patients were classified in three categories: virological success when plasma HIV-1 RNA level was <50 copies/ml at M12, virological failure when patients experienced two consecutive plasma HIV-1 RNA levels 50 copies/ml or International Medical Press

3 Switch to rilpivirine in HIV-infected patients one viral load above 50 copies/ml and discontinuation of RPV, and no data when virological data were missing because of loss to follow-up, discontinuation of RPV for reasons other than VF, or missing data in the window period. The window period around M12 ranged from month 10 to month 14. Data outside the window period were excluded. Sensitivity analyses were performed using available data obtained after M12 in patients who remained on their RPV-based regimen to assess the rate of virological success. Data are presented as medians with IQR for continuous variables and frequencies with percentages for qualitative variables. Changes in CD4 + T-cell counts, plasma creatinine, total HDL and LDL cholesterol from BL to M12 were assessed using paired Wilcoxon tests. Comparisons between patients with VF and virological success were performed to assess BL risk factors associated with VF including demographics, nadir CD4 + T-cell count, duration of HIV-1 suppression in plasma, peak of plasma HIV-1 RNA levels, prior use of NNRTIs, type of third agents used prior to the switch (NNRTIs: EFV or nevirapine) versus non-nnrtibased regimens (including PIs, raltegravir and NRTIs alone), prior VF with an NNRTI-based regimen, prior NNRTI-associated RAMs and history of the M184V/I mutation. Categorical variables were compared using Fisher exact tests while quantitative variables were compared using Wilcoxon signed rank tests. All tests were two-sided at the 0.05 significance level. Analyses were performed using R statistical package (version 3.1.0; the R Foundation, Vienna, Austria). Results Study participants Among the 3,491 HIV-1-infected patients followed at Saint-Louis Hospital during the study period, 385 patients received RPV and 281 treatment-experienced patients with a plasma HIV-1 RNA level <50 copies/ml who switched to an RPV-based regimen were analysed in this study (Figure 1). Subjects were mostly male, Caucasian and men who have sex with men (MSM), with a median age of 47 years (Table 1). Patients had been on ART for a median of 7 years and had a median BL CD4 + T-cell count of 630 cells/mm 3. Median duration of plasma HIV-1 RNA level <50 copies/ml before the switch was 38 months. The vast majority of patients, 273 (97%), received before the switch a combination of two nucleoside/ nucleotide reverse transcriptase inhibitors with a third agent. Among them, 80% received FTC/TDF, 15% lamivudine (3TC)/abacavir and 5% other combinations. The third agent was an NNRTI in 39% (EFV in 32%), PIs in 51%, integrase inhibitors in 7% (all received raltegravir) and another NRTI in 3%. All patients but eight (3%) switched to a fixed-dose combination of RPV/FTC/TDF. Main reason for switch was treatment simplification (n=177, 63%) and poor tolerability of their current regimen (n=93, 33%), mostly because of neuropsychiatric symptoms (n=65, 23%, all receiving EFV) and also gastrointestinal symptoms (4%). Efficacy At M12, using the FDA snapshot algorithm, 167 (59%) patients maintained full virological suppression (plasma HIV-1 RNA level <50 copies/ml) and met virological success, and 16 (6%) experienced VF (Figure 2). Among the 98 (35%) patients who had no data in the window period, 23 (8%) discontinued RPV for AEs, 6 discontinued RPV for other reasons (4 due to patient decision and 2 to avoid drug interaction), 2 patients died of suicide, 35 (12%) patients had no plasma viral load data in the window period and 32 (11%) were lost to follow-up. A sensitivity analysis performed with plasma viral load data available beyond M12 in 35 patients who Figure 1. Study flowchart 3,491 Patients followed at Saint-Louis Hospital 385 Patients treated with RPV 281 Patients included Exclusion of patients treated in trials: 9 Exclusion of treatment-naive patients: 46 Exclusion of patients with VL>50 copies/ml: 42 Exclusion of patients without RPV: 7 32 (11%) Patients lost to follow-up 38 (14%) Patients discontinued RPV 7 due to VF 23 due to AEs 6 due to other reasons 2 patients died 211 (75%) Patients remained on RPV at M12 AE, adverse event; M12, month 12; RPV, rilpivirine; VF, virological failure; VL, viral load. Antiviral Therapy

4 S Gazaignes et al. Table 1. Baseline characteristics of the 281 patients Characteristic n or median % or IQR Male gender Age, years Ethnicity Sub-Saharan Africa Europe North Africa Mode of HIV infection Homosexual Heterosexual Intravenous drug users, transfusion, others 13 5 Duration of HIV infection, years CD4 + T-cell count at baseline, cells/mm Duration of undetectable HIV RNA before switch, months Time between first ART and switch, years CDC category C Hepatitis coinfection HBV (HBsAg-positive) 17 6 HCV (Ab anti-hcv-positive) ART regimen before the switch Combination of 2 N(t)RTIs FTC/TDF TC/ABC TC/AZT 14 5 Other combination of 2 NRTIs 2 <1 Third agent EFV NVP 15 5 ETR 5 2 PI + RTV ATV + RTV DRV + RTV LPV + RTV 15 5 ATV unboosted 21 8 INSTI 18 7 NRTIs 8 3 Reasons for switch Simplification of treatment regimen Neuropsychiatric AEs Gastrointestinal AEs 12 4 Dyslipidaemia 4 1 Lipodystrophy 3 1 Other AEs 9 3 To avoid drug interaction 5 2 Other reasons 6 2 Ab, antibodies; ABC, abacavir; AEs, adverse events; ART, antiretroviral therapy; ATV, atazanavir; AZT, zidovudine; DRV, darunavir; EFV, efavirenz; ETR, etravirine; FTC, emtricitabine; HBsAg, hepatitis B surface antigen; INSTI, integrase inhibitor; LPV, lopinavir; N(t)RTI, nucleoside/nucleotide reverse transcriptase inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; NVP, nevirapine; PI, protease inhibitor; RTV, ritonavir; TDF, tenofovir disoproxil fumarate; 3TC, lamivudine. remained under the same RPV-based regimen yielded a similar rate of VF (6%, no additional VF) but the rate of virological suppression increased to 72% (202 patients) since all 35 patients had a plasma HIV-1 RNA level Figure 2. Primary efficacy end point at M12 using the FDA snapshot analysis Percentage M12 Post-M12 Virological success Virological failure No data M12 >M12 A sensitivity analysis using available HIV-1 viral load data beyond month (M) 12 is also presented. Values below x-axis are number of patients. <50 copies/ml at some time beyond M12 (Figure 2). In addition, using a threshold of 200 copies/ml instead of 50 copies/ml to define virological success, the rate of virological suppression would increase further to 74% and the rate of VF decrease to 3.5% (data not shown). Median CD4 + T-cell count remained unchanged from BL to M12 (+5 cells/mm 3, IQR ; P=0.45). Virological failure and resistance Sixteen (6%) patients experienced VF at M12 (Table 2). Among 10 patients with genotypic resistance tests available before the switch, 4 had previous NRTI RAMs and 1 had previous NNRTI RAMs which were unknown at the time of switch in 2 patients. 4 of these 16 patients had a history of previous ART failure (2 with NNRTI-based regimen) and the pattern of resistance mutations suggested that at least 2 other patients (patients 3 and 6) also had experienced previous failure. VF was detected as early as one month after the switch in four patients, and at month 3 in three other patients. RPV plasma levels were measured at the time of VF in 12/16 patients and were consistent with good treatment adherence in 10/12, with only 2 patients with no drug detected. Among the 16 patients with VF, resistance genotype at the time of VF was performed in all but 1 patient, but was not amplified in 6. Among the remaining 9 patients, new emerging NRTI or NNRTI RAMs were identified in 2 and 5 patients, respectively (Table 2). When assessing BL risk factors associated with VF, previous M184V/I mutation and use of non-nnrtis as third agent were significantly associated with VF International Medical Press

5 Switch to rilpivirine in HIV-infected patients Table 2. Characteristics of patients with VF RPV level VL at VF, Time at VF, Previous genotypic test Genotypic test at VF Pt Previous VF Previous ART Duration a copies/ml of VF ng/ml NRTI RAM NNRTI RAM NRTI RAM NNRTI RAM 1 With NRTI 3TC/ABC; ATV 5.8 years 428 M6 122 M41L; M184V; M184V E138K T215Y 2 With, NRTI, FTC/TDF; RAL 2.3 years 346,054 M1 85 M41L; D67N; M41L; D67N; K103N; V106I; NNRTI, PI T69D; K70R; T69D; K70R; H221Y; M230L L74I; M184V; L74I; M184V; T215F; K219Q T215F; K219Q 3 No DRV+RTV; RAL 3.4 years 37,105 M3 303 M41L; D67N; K103N; V108I M41L; D67N; K103N; V179I; K70R; T215F; K70R; T215F; Y181C K219Q; M184V K219Q; M184V 4 No 3TC/AZT; ATV 3.2 years 54 M1 No data M41L Not amplified Not amplified 5 No FTC/TDF; ATV+RTV 3 years 40,985 M3 34 No data No data M184V K103N; E138A/K; V179I; P225H 6 No FTC/TDF; ATV+RTV 3 months 33,908 M1 139 No data No data D67N; K70R; K101E; Y181I M184V; T215F; K219Q 7 With NNRTI NVP; ATV+RTV 8.5 years 1,328 M No data No data Not amplified Not amplified 8 With PI FTC/TDF; ATV+RTV 4.2 years 51 M12 No data No mutation No mutation Not amplified Not amplified 9 No FTC/TDF; ATV 8 months 52 M6 No data No data No data Not amplified Not amplified 10 No 3TC/ABC; RAL 1.7 years 97 M6 62 No mutation No mutation Not amplified Not amplified 11 No AZT/3TC/ABC 12.8 years 90 M12 No data No data No data Not amplified Not amplified 12 No FTC/TDF/EFV 1 month 411,786 M1 <1 No mutation No mutation No data No data 13 No FTC/TDF; LPV+RTV 4.3 years 455 M3 160 No mutation No mutation No mutation No mutation 14 No FTC/TDF/EFV 1 month 224 M No mutation No mutation No mutation No mutation 15 No FTC/TDF; LPV+RTV 6 years 16,036 M12 <1 No data No data No mutation No mutation 16 No FTC/TDF; RAL 2.3 years 88 M12 88 No mutation No mutation No mutation No mutation a Duration of undetectable plasma HIV RNA before switch (years/months). ABC, abacavir; AEs, adverse events; ART, antiretroviral therapy; ATV, atazanavir; AZT, zidovudine; DRV, darunavir; EFV, efavirenz; ETR, etravirine; FTC, emtricitabine; INSTI, integrase inhibitor; LPV, lopinavir; M, month; NNRTI, non-nucleoside reverse transcriptase inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; NVP, nevirapine; PI, protease inhibitor; Pt, patient; RAL, raltegravir; RAM, resistance-associated mutation; RPV, rilpivirine; RTV, ritonavir; TDF, tenofovir disoproxil fumarate; VF, virological failure; 3TC, lamivudine. whereas nadir CD4 + T-cell count, and duration of undetectable plasma VL before the switch or peak viral load were not (Table 3). There was also a trend toward more VF in patients who have experienced previous VF while receiving NNRTI-based regimens. Safety A total of 50 (18%) patients reported at least 1 clinical AE, 32 (11%) were neuropsychiatric AEs, 16 (6%) were gastrointestinal AEs and 9 (3%) patients experienced rash. Twenty-three (8%) participants discontinued RPV due to one or more AEs, including 16 (6%) patients with neuropsychiatric disorders (sleep disturbances, vertigo and dizziness), 7 (2%) with gastrointestinal AEs, 5 (<2%) with rash, 3 with hepatitis and 1 with renal failure. Two patients, both young MSM aged 28 and 34 years, died during the study because of suicide (Additional file 1). The first patient had presented with anxiety and depression since the diagnosis of HIV-infection in He was started on EFV/TDF/FTC in September 2012 and was switched to RPV/TDF/FTC in April 2013 because of persistent depressive mood, nightmares and insomnia. His psychiatrist introduced paroxetine at the same time. In May 2013, following a separation from his boyfriend, his depression increased, the dose of paroxetine was also increased and alprazolam and zopiclone were added. He committed suicide 6 months later. The second patient was diagnosed with HIV-infection in He was started on EFV/TDF/FTC in 2010 and was addicted to amphetamines. He presented multiple episodes of delirious puffs following amphetamine intake and started treatment with methotrimeprazine and olanzapine in 2012 while hospitalized in psychiatry. He was switched to RPV/TDF/FTC in January 2013 because of severe depression. In March 2013, following a problem with his partner, he presented a new episode of delirium with amphetamines and expressed suicidal ideas. He committed suicide one month later. From BL to M12, a small but significant increase in median ASAT and ALAT level was observed in 161 evaluable patients: +4 UI/l (IQR=-1 9) and +5 UI/l (IQR -3 5) respectively (P<0.001). A total of 39 and 58 patients Antiviral Therapy

6 S Gazaignes et al. Table 3. Baseline risk factors associated with VF VFs (n=16), Virological n/median success (n=202), Parameters (%/IQR) n/median (%/IQR) P-value Age, years 47 (42 48) 47 (40 53) 0.90 Male gender 10 (63) 156 (77) 0.22 Mode of HIV infection 0.15 Heterosexual 9 (56) 65 (32) MSM 6 (38) 119 (59) Others 1 (6) 18 (9) Highest HIV VL, log 4.9 ( ) 5 ( ) 0.59 copies/ml CD4 + T-cell count nadir, 210 ( ) 260 ( ) 0.37 cells/mm 3 Undetectable plasma VL 38 (17 560) 37 (17 78) 0.67 before the switch, months Previous VF with NNRTI 2 (12.5) 4 (2) 0.06 regimen Previous RAMs to NNRTIs a 1/10 (10) 3/109 (3) 0.30 Previous M184V/I a 3/10 (30) 5/109 (5) 0.02 NNRTIs as third agent 2 (13) 82 (41) 0.03 before switch vs others (PIs, RAL, NRTIs) a Genotypic data were available for only 10/16 patient with virological failure (VF) and 109/202 patient with virological success. MSM, men who have sex with men; NNRTI, non-nucleoside reverse transcriptase inhibitor; NRTI, nucleoside reverse transcriptase inhibitor; PI, protease inhibitor; RAL, raltegravir; RAM, resistance-associated mutation; VL, viral load. experienced elevations in ASAT and ALAT levels of at least 1 grade, respectively, but only 3 patients developed grade 3 ALAT/ASAT elevation leading to treatment discontinuation. Median change from BL in serum creatinine was +6 µmol/l (IQR= ; P<0.001) in 164 evaluable patients. Twenty-three patients developed a grade 1 elevation of plasma creatinine and 1 patient discontinued RPV because of renal failure. This patient switched from zidovudine/3tc and boosted fosamprenavir to RPV/TDF/FTC and had to stop his regimen because of an increase in plasma creatinine to 495 µmol/l at month 2. Plasma creatinine level returned to BL 6 months later. The serum lipid profile slightly improved from BL to M12 following the switch to RPV-based regimens. Significant changes in median total cholesterol (TC; -0.5 mmol/l, IQR ; P<0.001), LDL cholesterol (-0.26 mmol/l, IQR ; P<0.001), HDL cholesterol (-0.06 mmol/l, IQR ; P<0.008), triglycerides (-0.12 mmol/l, IQR ; P=0.011) and TC/HDL cholesterol ratio (-0.28, IQR ; P=0.017) were seen in 89 evaluable patients. No significant change from BL to M12 was seen in median fasting plasma glucose levels in 81 assessable patients (0.03 mmol/l, IQR ; P=0.29). Discussion This study aimed to assess the efficacy and safety of a switch to RPV-based regimens in a real-life setting among 281 virologically suppressed treatment-experienced patients in a single centre in Paris, France. The rate of virological success at 12 months in this retrospective observational cohort study, defined as the proportion of patients who maintained a suppressed plasma RNA level <50 copies/ml, was only 59%, which is lower than previously reported in clinical trials and small cohort studies, despite a moderate proportion of patients lost to follow-up (11%) [6 13]. However in this study a few patients who were likely to fail because of prior NNRTI or NRTIs RAMs, or because of prior VF to NNRTI or NRTI-based regimens were inappropriately switched to RPV-based regimens. Also, this study was conducted in real-life conditions and therefore a substantial number of patients had missing data within the M12 window period used for the FDA-snapshot analysis. Indeed 35 (12%) patients who were still under follow-up and remained under the same RPVbased regimen had no viral load data available during the M12 window period, but using a sensitivity analysis including data beyond M12 for these 35 patients who remained all virologically suppressed, the rate of virological success increased to 72% (Figure 2). This relatively lower rate of treatment success in this open-label cohort is explained by higher rates than expected of both VFs and treatment discontinuations because of AEs. Sixteen (6%) patients experienced VF during the study period although VF led to treatment discontinuation in only 7 (2.5%) patients. Indeed, 6 patients had low viral loads (<100 copies/ml) and two others were probably not compliant according to the lack of RPV detection in plasma (Table 2). Longer follow-up will provide a better assessment of the risk of VF with this regimen. Interestingly, among patients with VF, 4 had previous NRTI RAMs, 1 of whom had additional NNRTI-based mutations, and 4 have reported previous VF with NRTI, NNRTI or PI-based regimens. Among the 9 patients with VF who had a resistance genotype available at the time of VF, new emerging NRTI or NNRTI RAMs were identified in 2 and 5 patients, respectively. Also, we assessed BL risk factors associated with VF and identified the presence of the M184V mutation in prior genotypes and the use of a non-nnrti third agent before the switch with an increased risk of VF. There was also a trend toward a higher risk of VF among patients with a history of previous failure with NNRTI-based regimens. Our study lacked power, however, to assess the risk of VF in patients with NNRTI RAMs, since only four patients had known NNRTI RAMs at the time of the switch. These data are consistent with previous International Medical Press

7 Switch to rilpivirine in HIV-infected patients reports suggesting the efficacy of a switch from either EFV or nevirapine to RPV [9,11,12]. They also reinforce the need to restrict the switch to an RPV-based regimen to patients not only fully suppressed under their current regimen, but also to those with a previous genotype available, without any RPV/TDF/FTC-associated RAMs, or without any history of VF, especially when they are switched from non NNRTI-based regimens, such as PIs or raltegravir [6]. It was definitely unwise in our study to have switched to RPV-based regimens patients with prior RAMs to NNRTI and NRTIs, as well as patients with previous VFs to NNRTI and NRTIs containing regimens. These data are unfortunately not always easy to collect or retrieve in patient charts from patients followed for years and sometimes at other institutions. In these cases, when treatment history is not reliable, and a resistance genotype not available, it is safe to avoid a switch to RPV except maybe if the patient is currently suppressed under an NNRTI-based regimen as previously reported with EFV and nevirapine [9,11,12]. In some cases, performing a resistance genotype of HIV proviral DNA in peripheral blood lymphocytes could be helpful, but this test is not perfect as it may underestimate the risk of resistance as compared to plasma HIV RNA [17]. Rilpivirine is an attractive NNRTI because of its good safety profile, which has been proven consistently better than EFV in randomized trials and also because of its convenience when coformulated in a single pill with TDF and FTC [2 5]. Indeed, simplification was the main reason for the switch for 63% of the patients and 97% in this study received this single pill combination. In this study, up to 8% of patients discontinued RPV during follow-up for AEs, a rate that is somewhat higher than previously reported. Patients are probably less carefully switched to RPV in such a cohort study than in a randomized trial, but this rate of treatment discontinuation probably better reflects the safety in real-life settings where patients have multiple treatment options should they experience tolerability issues. On the other hand, since the study is open-label it is difficult to ascertain that reported AEs were due to RPV. Indeed, in this study, 16 patients (6%) discontinued RPV for neuropsychiatric AEs, mostly sleep disturbances. Such sleeping disorders, although less frequent than with EFV, have been already reported in randomized trials with RPV [2 5]. However, neuropsychiatric disorders were also one of the main reasons for the switch in 65 patients (23%) all of whom were receiving EFV before the switch. It is therefore difficult to know whether these symptoms were drug-related or due to associated psychiatric comorbidities which are quite prevalent in HIVinfected patients. Similarly, the two patients who died both had a long history of psychiatric disorders, and it would be difficult to know whether the switch to RPV was responsible for this outcome, as recently suggested with EFV [18]. Longer follow-up of this cohort will also be important to better address this issue. Other AEs, although rare, were more likely to be related to the use of RPV or the combination of RPV plus TDF/FTC: gastrointestinal AEs, renal failure (one patient), hepatitis (3 patients) and rashes (5 patients). As previously reported there was a modest improvement in the lipid profile following the switch to RPV/ TDF/FTC in our study where half of patients switched from a boosted PI regimen. The clinical relevance of this slightly improved lipid profile is yet unclear. Another limitation of this study was that it was mainly conducted in male Caucasian MSM and its results cannot therefore be easily extrapolated to other populations. In conclusion, our study reinforced the need to carefully select virologically suppressed patients who are considered for a switch to an RPV-based regimen, here mainly a switch to RPV/TDF/FTC. Patients without a reliable treatment history, those without a resistance genotypic test available, with prior RAMs to NNRTIs and NRTIs, and those with previous VF should avoid such a switch due to a higher risk of VF. In addition, although the tolerability of RPV overall is good, some patients may still complain of neuropsychiatric AEs which might lead to treatment discontinuation. Acknowledgements This study has been presented in part at the 53rd ICAAC Meeting, September 2013, Denver, CO, USA, and at HIV Drug Therapy, October 2014, Glasgow, UK. This study was sponsored, in part, by a grant from Gilead sciences who had no role in data collection or data analysis. We would like to thank David Simo and Hervé Akpe (both from Hopital St Louis, Paris) who participated in data collection for this study. Disclosure statement J-MM has received grants from Gilead Sciences and Merck, and has participated in advisory boards for Gilead Sciences, Merck, Janssen, Bristol Myers Squibb, ViiV and Tobira. CD has received grants from Merck and has participated in advisory boards for Gilead Sciences, Merck, Bristol Myers Squibb. All other authors declare no competing interests. Additional file Additional file 1: A table of adverse events reported during the study period can be found at com/uploads/documents/3705_gazaignes_addfile1.pdf Antiviral Therapy

8 S Gazaignes et al. References 1. Schafer JJ, Short WR. Rilpivirine, a novel non-nucleoside reverse transcriptase inhibitor for the management of HIV-1 infection: a systematic review. Antivir Ther 2012; 17: Cohen CJ, Andrade-Villanueva J, Clotet B, et al. Rilpivirine versus efavirenz with two background nucleoside or nucleotide reverse transcriptase inhibitors in treatmentnaive adults infected with HIV-1 (THRIVE): a Phase 3, randomised, non-inferiority trial. Lancet 2011; 378: Molina JM, Cahn P, Grinsztejn B, et al. Rilpivirine versus efavirenz with tenofovir and emtricitabine in treatmentnaive adults infected with HIV-1 (ECHO): a Phase 3 randomised double-blind active-controlled trial. Lancet 2011; 378: Cohen CJ, Molina JM, Cahn P, et al. Efficacy and safety of rilpivirine (TMC 278) at 48-weeks in treatment-naive, HIV-1 patients: Pooled results from the Phase 3 doubleblind randomized ECHO and THRIVE trials. J Acquir Immune Defic Syndr 2012; 60: Cohen C, Wohl D, Arribas J, et al. Week 48 results from a randomized clinical trial of rilpivirine/emtricitabine/ tenofovir disoproxil fumarate in treatment naive HIVinfected adults. AIDS 2014; 28: Palella FJ, Fisher M, Tebas P, et al. Simplification to rilpivirine/emtricitabine/tenofovir disoproxil fumarate from ritonavir-boosted protease inhibitor antiretroviral therapy in a randomized trial of HIV-1 RNA-suppressed participants. AIDS 2014; 28: Amiel C, Schneider V, Guessant S, et al. Initiation of rilpivirine, tenofovir and emtricitabine (RPV/TDF/ FTC) regimen in 363 patients with virological vigilance assessment in real life. J Antimicrob Chemother 2014; 69: Gantner P, Reinhart S, Partisani M, et al. Switching to emtricitabine, tenofovir and rilpivirine as single tablet regimen in virologically suppressed HIV-1-infected patients: a cohort study. HIV Med 2015; 16: Mills AM, Cohen C, Dejesus E, et al. Efficacy and safety 48 weeks after switching from efavirenz to rilpivirine using emtricitabine/tenofovir disoproxil fumarate-based single tablet regimens. HIV Clin Trials 2013; 14: Rokx C, Verbon A, Rijnders BJA. Successful switch to rilpivirine/tenofovir/emtricitabine in HIV-1 infected patients with an isolated K103N mutation acquired during prior nonnucleoside reverse transcriptase inhibitor therapy. HIV Med 2014; 15: Allavena C, Dailly E, Reliquet V, et al. Switching from tenofovir/emtricitabine and nevirapine to a tenofovir/ emtricitabine/rilpivirine single-tablet regimen in virologically suppressed, HIV-1-infected subjects. J Antimicrob Chemother 2014; 69: Rokx C, Blonk M, Verbon A, et al. The efficacy, pharmacokinetics, and safety of a nevirapine to rilpivirine switch in virologically suppressed HIV-1-infected patients. J Acquir Immune Defic Syndr 2015; 68: Gianotti N, Poli A, Nozza S, et al. Efficacy and safety in clinical practice of a rilpivirine, tenofovir and emtricitabine single-tablet regimen in virologically suppressed HIVpositive patients on stable antiretroviral therapy. J Int AIDS Soc 2015; 18: Pugliese P, Cuzin L, Cabié A, et al. A large French prospective cohort of HIV-infected patients: the Nadis Cohort. HIV Med 2009; 10: Crauwels HM, van Schaick E, van heeswijk RPG, et al. Effect of intrinsic and extrinsic factors on the pharmacokinetics of TMC 278 in antiretroviral-naive HIV-i-infected patients in ECHO and THRIVE. 10th International Congress on Drug Therapy in HIV Infection November 2010, Glasgow, UK. Poster P ANRS. ANRS scale to grade the severity of adverse events in adults. (Accessed 17 March 2016.) Available from www. anrs.fr/content/download/2242/12805/file/anrs-gradeei- V1-En-2008.pdf 17. Delaugerre C, Braun J, Charreau I, et al. Comparison of resistance mutation patterns in historical plasma HIV RNA genotypes with those in current proviral HIV DNA genotypes among extensively treated patients with suppressed replication. HIV Med 2012; 13: Mollan KR, Smurzynski M, Eron JJ, et al. Association between efavirenz as initial therapy for HIV-1 infection and increased risk for suicidal ideation or attempted or completed suicide: an analysis of trial data. Ann Intern Med 2014; 161:1 10. Accepted 4 November 2015; published online 13 November International Medical Press