Abstract
Aims
The purpose of this study was to assess the antiviral activity of the rilpivirine/emtricitabine/tenofovir disoproxil fumarate combination and to describe the pharmacokinetics of rilpivirine and its association with resistance in clinical routine.
Methods
A retrospective multicentre cohort study was performed in both naive and pretreated HIV patients receiving the once‐daily rilpivirine/emtricitabine/tenofovir disoproxil fumarate regimen. Immuno‐virologic and resistance data, and rilpivirine plasma trough concentrations were collected over the follow‐up. Statistical analyses were performed to evaluate the relationship between rilpivirine pharmacokinetics and virological response. Receiver operating characteristic (ROC) curve analysis was performed to determine the best target rilpivirine trough concentration.
Results
Overall, 379 patients were included. After a median follow‐up of 28 months, 26% of patients discontinued mainly due to toxicity and the virological success rate was 65.7%. Virological failure occurred in 5% of patients. A significant proportion of patients with HIV‐RNA > 40 copies/mL displayed rilpivirine plasma trough concentrations below the currently used 50 ng/mL efficacy threshold at both M6 (28%) and M12 (31%), in agreement with a significant lower median rilpivirine plasma trough concentration compared with patients virologically suppressed. Half of the patients with virologic failure who acquired rilpivirine resistance mutations had at least one suboptimal rilpivirine trough concentration. The optimal target for rilpivirine trough concentration was 70 ng/mL (sensitivity 75.4%; specificity 61.5%).
Conclusions
This study shows the impact of rilpivirine plasma trough concentration on both virological response and the emergence of rilpivirine mutations. Moreover, our results suggest that a higher target of rilpivirine trough concentration could be proposed in clinical practice.
Keywords: antiretroviral therapy, HIV, pharmacodynamics, pharmacokinetics, rilpivirine
What is known about this subject
Rilpivirine is widely prescribed for the treatment of HIV‐1 infection both in induction or maintenance strategies.
The trough plasma concentration of rilpivirine is highly correlated with virological response.
An important inter‐individual rilpivirine PK variability is described in clinical trials as well as in clinical routine that could affect both efficacy and safety of the treatment.
What this study adds
A significant proportion of patients with HIV‐RNA > 40 copies/mL displayed rilpivirine plasma Ctrough below the currently used 50 ng/mL efficacy threshold.
A significant impact of rilpivirine plasma Ctrough on virological response and a relationship with the emergence of rilpivirine‐associated mutations.
A higher efficacy target for rilpivirine plasma Ctrough should be considered to obtain virological response.
1. INTRODUCTION
The combination of rilpivirine (RPV), a second‐generation non‐nucleoside reverse transcriptase inhibitor (NNRTI) and two nucleos(t)ide reverse transcriptase inhibitors (NRTI), emtricitabine (FTC) and tenofovir disoproxil fumarate (TDF) is commonly used for the treatment of HIV‐1 infection both in naive and treatment‐experienced patients. This combination is available in a co‐formulated single tablet regimen (STR) (RPV/FTC/TDF 25/200/245 mg) taken once daily (QD), which improves treatment adherence and maximizes the success of antiretroviral (ARV) treatment. 1 , 2 The phase III studies proved that RPV/FTC/TDF had an antiviral efficacy similar to efavirenz (EFV)‐based regimens in naive patients (NP) with a plasma HIV‐1 viral load (VL) at baseline ≤100,000 copies/mL. 3 , 4 , 5 The efficacy was also demonstrated in patients virologically suppressed switching from either EFV/FTC/TDF or a protease inhibitor‐based regimen to RPV/FTC/TDF. 6 , 7 Moreover, RPV/FTC/TDF provides major advantages in terms of tolerance compared to other STRs. Indeed, data from phase III studies showed that side effects, such as neurological symptoms with a grade ≥ 2, were significantly less frequent with RPV‐based regimens than with EFV (16% vs 31%). Moreover, discontinuations due to adverse effects occurred at a lower rate with RPV. 3 , 4 RPV displays a different pattern of resistance mutations than first‐generation NNRTIs due to a greater flexibility and a specific interaction with the reverse transcriptase binding site. 8 It is making RPV an appropriate alternative treatment for patients failing first NNRTI regimens. Nonetheless, RPV is not recommended for patients harbouring a baseline VL > 100,000 copies/mL, related to higher rates of virological failure (VF) and resistance selection. 4 , 9 Overall, the most frequently selected mutations impacting NNRTI and NRTI were E138K (72%) and M184I (69%), respectively. 10
The pharmacokinetics (PK) properties of RPV, such as both the food‐ and pH‐dependent intestinal absorption and the intense hepatic metabolism through the cytochrome P‐450 3A4 (CYP3A4), enhance the variability of RPV exposure, and might compromise the antiviral efficacy. Population pharmacokinetic (Pop‐PK) models developed in cohorts of unselected HIV‐1‐infected adult patients showed a moderate interpatient variability of RPV PK parameters but a lower apparent volume of distribution associated with a shorter terminal elimination half‐life (t 1/2) than reported in the Summary of Product Characteristics. 11 , 12 This suggests that, in routine clinical practice, patients may be at greater risk of suboptimal exposure than expected, which may increase the risk of developing HIV‐resistant variants as demonstrated in vitro and in vivo with other ARV treatments. 13 , 14 In phase III studies, a strong relationship between RPV trough plasma concentration (Ctrough) and efficacy was established whatever the level of baseline VL (≤100,000 copies/mL or >100,000 copies/mL). Patients with an RPV Ctrough > 50 ng/mL have an 80% probability of achieving virological success. 15 Therefore, an efficacy threshold of 50 ng/mL is currently recommended in clinical practice as a minimal target RPV Ctrough to increase the probability of virological response in clinical practice. 16
However, studies combining PK and efficacy of RPV are lacking in real‐life contexts. 17 , 18 , 19 , 20 , 21 , 22 Hence, the present study assessed the antiviral activity of the RPV/FTC/TDF regimen and described the association of RPV exposure with virologic response and resistance in a cohort of NP and treatment‐experienced patients (TEP) over 3 years of follow‐up.
2. PATIENTS AND METHODS
2.1. Patients
Ambulatory patients who had initiated RPV/FTC/TDF regimen between November 2012 and November 2015 in the University hospital of Marseille and Bichat‐Claude Bernard (Paris) were included retrospectively in this observational study. The date when RPV/FTC/TDF therapy was started was the baseline for each patient. Patients were followed up for a period of 3 years after RPV/FTC/TDF initiation. This study was carried out in compliance with the International guidelines for human research protection (Declaration of Helsinki and ICH‐GCP). 23 , 24 Patients provided written informed consent for the use of their medical records on Nadis® (Fedialis, Marly‐Le‐Roi, France; electronic medical record for patients infected by HIV, HBV and HCV, approved by the French Commission Nationale Informatique et Liberté; registration number: 2001/762876/nadiscnil.doc).
2.2. Clinical and biological assessments
Baseline data of the patients, including demographic characteristics, therapeutic status, prior ARV therapy, reasons for switch, CD4 and VL, were collected from Nadis®. Data were described by median and range or frequency (%). Clinical follow‐up data, including VL and CD4, were collected at routine follow‐up visits at 6 months (M6), M12, M24 and M36 ± 2 months and were reported as median and interquartile range (IQR). The proportion of treatment interruptions and their reasons were investigated. Virological response was defined as the proportion of individuals who achieved viral suppression (plasma VL ≤ 40 copies/mL). The response rates were calculated using the intention‐to‐treat principle, whereby all missing data were treated as failures. The proportion of virological failures (VF), defined as a confirmed plasma VL > 40 copies/mL, the resistance profile associated to VF, the number and amplitude of viral blips (transient VL > 40 copies/mL preceded and followed by a VL ≤ 40 copies/mL) in patients virologically suppressed, and the CD4 counts were also collected. NNRTI and NRTI‐resistance mutations were obtained from the last genotype available prior to starting RPV/FTC/TDF treatment and from the cumulative genotype. In the case of VF, NNRTI and NRTI‐resistance mutations were also collected if genotype was available. Resistance mutations were defined and interpreted according to the ANRS (Agence Nationale de Recherche sur le Sida et les hépatites virales) drug resistance algorithm (http://hivfrenchresistance.org, version 27).
2.3. RPV pharmacokinetics
Blood samples were collected at steady state, at various registered post‐dose times ranging from 1 to 31 hours and over the course of the RPV/FTC/TDF treatment. Ctrough was defined as the concentration at 24 hours ± 6 hours after the last drug intake. RPV plasma concentrations were determined by a sensitive and validated reverse‐phase ultra‐performance liquid chromatography coupled to a tandem mass spectrometry method, as previously described. 25 The assay was validated over a calibration range of 10 to 5,000 ng/mL with a lower limit of quantification of 10 ng/mL. Both accuracy and precision were less than 15%.
2.4. Statistical analysis
Intra‐individual variability was evaluated by averaging the coefficients of variation (CV) (defined as the ratio of the standard deviation to the mean) of all the available RPV Ctrough from each individual patient throughout the follow‐up period. Inter‐individual variability was calculated using the CV for the mean of available RPV Ctrough from each subject.
Comparison of RPV Ctrough between patients with VL ≤ 40 and > 40 copies/mL was analysed by the non‐parametric Mann–Whitney U‐test. The percentage of patients with RPV Ctrough below the threshold of 50 ng/mL was compared between both populations by Fisher exact test. Analyses were performed at M6, M12, M24 and M36.
Optimal target for the RPV Ctrough was determined using receiver operating characteristic (ROC) curve analysis. Area under the curve (AUC), sensitivity and specificity values were used for the evaluation of the optimal cut‐off.
Statistical analyses were performed using R software v 3.5.3 (https://cran.r-project.org/). A P‐value of <0.05 was considered statistically significant.
3. RESULTS
3.1. Demographic and pharmacodynamic data at baseline
A total of 65 NP and 314 TEP were included in the study of which 270 TEP (86%) had an undetectable baseline VL. The population consisted of 71% male patients. The median age at inclusion was 44 years (range: 19–82 years). Baseline characteristics of NP and TEP are shown in Table 1. Sex ratio and median age were similar between NP and TEP. The median number of ARV regimens received by TEP before RPV/FTC/TDF therapy was four and boosted PI was predominant in the previous regimens. Simplification was the main reason for switching, followed by adverse effects.
TABLE 1.
Patients characteristics at baseline
| Parameters median (range) or n (%) | Naive population (NP) (n = 65) | Treatment‐experienced population (TEP) (n = 314) |
|---|---|---|
| Male gender | 53 (81.5%) | 215 (68.5%) |
| Age, years | 37 (19–70) | 44 (19–82) |
| CD4, cells/mm 3 | 480 (24–1609) | 602 (20–2351) |
| Nadir CD4, cells/mm 3 | 373 (24–1609) | 247 (1–1275) |
| Plasma HIV RNA, copies/mL | 15,318 (40–3141 000) | 40 (40–285 022) |
| % ≤ 40 copies/mL | 3 (4.6%) | 270 (85.9%) |
| Co‐infection | ||
| HBV | 3 (4.6%) | 31 (9.9%) |
| HCV | 3 (4.6%) | 32 (10.2%) |
| HBV and HCV | – | 4 (1.3%) |
| Number of prior ARV regimens | – | 4 (2–26) |
| Time since first antiretroviral medication, years | – | 6 (0.08–25.6) |
| Last ART treatment | ||
| 2 NRTIs + 1 PI/r | – | 176 (56.1%) |
| 2 NRTIs + 1 NNRTI | – | 103 (32.8%) |
| 2 NRTIs + II | – | 22 (7.0%) |
| Other | – | 13 (4.1%) |
| Reasons for switching | ||
| Simplification | – | 166 (52.8%) |
| Adverse effects | – | 106 (33.8%) |
| Virologic failure | – | 10 (3.2%) |
| Other | – | 25 (8.0%) |
| Not available | – | 7 (2.2%) |
| Available genotype | 54 (83.1%) | 229 (72.9%) |
| HIV‐RNA not amplifiable | 7 (10.8%) | 20 (6.4%) |
| Not available | 4 (6.2%) | 65 (20.7%) |
| WT genotype | 43 (79.6%) | 147 (64.2%) |
| Presence of resistance mutations | 11 (20.4%) | 82 (35.8%) |
| NNRTI mutations only | 7 (13%) | 35 (15.3%) |
| NRTI mutations only | 2 (3.7%) | 27 (11.8%) |
| NNRTI + NRTI mutations | 2 (3.7%) | 20 (8.7%) |
PI/r: boosted protease inhibitor; NNRTI: non‐nucleoside reverse transcriptase inhibitor; NRTI: nucleoside/nucleotide reverse transcriptase inhibitor; II: integrase inhibitor; HBV: Hepatitis B virus; HCV: Hepatitis C virus; WT: wild type.
Among the 283 patients (74.7%) with an available genotype, 93 (32.9%) presented resistance mutations at baseline (Table 1). Among them, a total of 64 patients (68.8%), 9 NP and 55 TEP, harboured NNRTI resistance mutations: 45 patients with n = 1 mutation, 13 (n = 2), 5 (n = 3) and 1 (n = 4). Among them, 16 patients (14 TEP and 2 NP) presented specific mutations conferring resistance to RPV. Moreover, based on the cumulative genotype, five more TEP showed mutations conferring resistance to RPV. The NNRTI mutations were: A98S (n = 12), K103N (n = 9), V179I (n = 9), V90I (n = 6), V179VI (n = 5), A98G (n = 3), K101R (n = 3), G190A (n = 3), K103H/N/S/T (n = 2), V106I (n = 2), H221H (n = 2), P236M (n = 2), A98AE (n = 1), A98AG (n = 1), K101Q (n = 1), K101KR (n = 1), K101KC (n = 1), K103R (n = 1), K103KN (n = 1), V106VI (n = 1), V108I (n = 1), V108VI (n = 1), V179I/L/M/T (n = 1) and M230P (n = 1). The mutations conferring resistance to RPV were: E138A (n = 6), V179L (n = 5), M230MI (n = 3), E138K (n = 2), K101E (n = 1), V179D (n = 1), Y181YC (n = 1), Y188L (n = 1), Y181C (n = 1) and K103N+ L00I (n = 1). The K103N mutation was present in 13 patients, always in association with other mutations except in one patient.
Among the patients with resistance mutations at baseline, 51 patients (54.8%) harboured NRTI resistance mutations. Most of them (n = 27) harboured only one NRTI mutation and other patients had between two and seven NRTI mutations: 2 (n = 8), 3 (n = 3), 4 (n = 8), 5 (n = 2), 6 (n = 1) and 7 (n = 2). The NRTI mutations identified were: M184V (n = 15), M41L (n = 13), D67N (n = 13), V118I (n = 8), T215Y (n = 8), K70R (n = 6), K219Q (n = 5), T69N (n = 4), M184V/I (n = 4), L210W (n = 4), T215F (n = 3), K70KR (n = 2), K70KT (n = 2), L74I (n = 2), T215N (n = 2), T215S (n = 2), K19E (n = 2), E44ED (n = 1), T69N/S (n = 1), T69S (n = 1), M41ML (n = 1), M184MV (n = 1), M184MIV (n = 1), L210G (n = 1), T215C (n = 1), T215L (n = 1), T215TS (n = 1), T215CDG (n = 1), T215TFIS (n = 1) and T215K (n = 1).
3.2. Clinical follow‐up
The median follow‐up duration was 28 months. Thirteen patients were lost to follow‐up at M6 (n = 4), M12 (n = 2), M24 (n = 6), M36 (n = 1), and 15 patients were transferred to another facility (4 at M6, 3 at M12, 3 at M24 and 5 at M36). Of the 351 patients remaining (55 NP and 296 TEP), 91 (26%) including 20 NP (36.4%) and 71 TEP (23.9%), discontinued the treatment at M6 (n = 14), M12 (n = 16), M24 (n = 37) and M36 (n = 24). The most frequent cause of discontinuation in both populations was toxicity (n = 43, 47%). Among adverse event (AE)‐related interruptions, neuropsychiatric and renal side effects occurred in 32.6% (n = 14) and 20.9% (n = 9) of patients, respectively. The other AE‐related interruptions were gastro‐intestinal (n = 2), osteoarticular (n = 3), dyslipidaemia (n = 3), hepatic (n = 2), cutaneous (n = 1) and unknown (n = 9). The other reasons leading to treatment interruption were VF (20.9%), simplification (13.2%), pregnancy (10.9%), patient decision (2.2%), unknown (4.4%) and one TEP died from bronchial cancer (not related to the treatment).
3.3. Virological outcomes
Virologic success was achieved in 88.6%, 84.2%, 75.2% and 65.7% of patients at M6, M12, M24 and M36, respectively (Figure 1). Among TEP with plasma VL detectable at baseline (n = 44), the proportion of patients in virologic success was 84.1%, 61.4%, 52.3% and 38.6% at M6, M12, M24 and M36, respectively. Regarding initially virologically‐suppressed TEP patients (n = 270), the proportions were 92.6%, 88.9%, 81.1%, 74.8%, respectively.
FIGURE 1.
Virologic suppression at 6 months (M6), 12 months (M12), 24 months (M24) and 36 months (M36) for pretreated patients (TEP) and naive patients (NP)
Percentage of patients with viral load (VL) ≤ 40 copies/mL at M6, M12, M24 and M36 for general population (blue), TEP (orange) and NP (grey). N represent the number of patients reaching VL ≤ 40 copies/mL
Seven TEP (2.2%) and four NP (6.2%) had isolated viral blip during the treatment. One TEP presented another viral blip 6 months after the first blip episode. The median viral blip amplitude was 222 copies/mL (range 57–1,087 copies/mL). All these patients maintained virologic suppression until the end of the follow‐up period. The CD4 count increased from a median of 577 cells/mm3 (IQR: 411–770) at baseline to 591 cells/mm3 (IQR: 450–800), 610 cells/mm3 (IQR: 479–840), 651 cells/mm3 (IQR: 492–826), 681 cells/mm3 (IQR: 508–840) at M6, M12, M24 and M36, respectively.
Nineteen patients (5%) discontinued the regimen because of VF, in similar proportions between NP (6.1%) and TEP (4.8%), respectively at M6 (4 TEP), M12 (2 NP; 5 TEP), M24 (1 NP; 6 TEP) and M36 (1 NP). Among the 15 TEP who failed the regimen, 6 (40%) were virologically suppressed at baseline.
3.4. Resistance analysis
Among the 19 patients with VF, HIV genotyping was performed and available for 17 patients. The details of the resistance profile and pharmacological data of these patients at baseline and at the time of failure are summarized in Table 2. Eight (47%) patients had acquired RPV mutations at failure, namely E138K (n = 3), M230L (n = 2), E138Q (n = 1), E138KQ (n = 1), Y181C (n = 1), Y181I (n = 1), Y188YFL (n = 1) and M230I (n = 1). Among these eight patients, a suboptimal RPV plasma Ctrough was observed once for one patient and twice for three patients during the follow‐up. Among the 21 patients with pre‐existing RPV‐associated resistance mutations, 12 were virologically suppressed at M36, one was lost to follow‐up, seven discontinued the treatment and only one experienced a virologic failure at M5, while displaying RPV Ctrough above 50 ng/mL. However, this patient had the combination K103N + L100I at baseline, which conferred a complete resistance to RPV. Only three out of the 13 patients harbouring the K103N mutation at baseline failed the regimen, including the patient with the K103N + L100I mutations.
TABLE 2.
Resistance profile and pharmacological data of the 19 patients with virologic failure
| Previous cART before the switch | VL at baseline copies/mL | VL at failure copies/mL | Time of VF month | RT mutations at baseline | RT mutations at failure | RPV Ctrough ng/mL | ||
|---|---|---|---|---|---|---|---|---|
| 1 | NP | – | 6971 | 169 800 | 10 | None | K103KN, E138KQ, Y188YFL | 23 (M3); 23 (M10) |
| 2 | NP | – | 15 804 | 40 | 10 | A98S | A98S | 97 (M2); 69 (M3); 106 (M8) |
| 3 | NP | – | 286 890 | 212 | 19 | K103R, V179I | K65T, K103R, V179I, Y181C, M230L | 260 (M14) |
| 4 | NP | – | NA | 498 | 27 | None | None | 84 (M7) |
| 5 | TEP |
Atazanavir/r TDF/FTC |
285 022 | 236 | 3 | None | None | 198 (M1); 68 (M3) |
| 6 | TEP |
Lopinavir/r Lamivudine/abacavir |
153 340 | 145 600 | 9 | D67N, T69DN, K70R, V118I, M184V, T215F, K219Q |
D67N, T69D, K70R, V118I, V179I, Y181I M184V, T215F, K219Q |
62 (M1); 82 (M5); 56 (M9) |
| 7 | TEP |
Darunavir/r TDF/FTC |
40 | 521 | 3 | None | None | 155 (M1) |
| 8 | TEP | Darunavir/r TDF/FTC | 81 | 66 | 5 | NA | T69DN | 104 (M5) |
| 9 | TEP | Raltegravir TDF/FTC | 40 | 49 | 17 | None | None | 52 a (M6) |
| 10 | TEP | Darunavir/r TDF/FTC | 58 | 94 | 16 | None | None | 70 (M9) |
| 11 | TEP | Fosamprenavir/r TDF/FTC | 8465 | 803 | 12 | None | V106A, E138K, M184V, L210W, T215D | 12 (M7); 10 (M10) |
| 12 | TEP | Efavirenz lamivudine/abacavir | 15 952 | 495 | 22 | None | K65R, E138Q, M184V M230L | 28 (M1); 47 (M3); 67 (M12) |
| 13 | TEP | Atazanavir/r lamivudine/abacavir | 40 | 61 | 13 | M41L, A98S, V179I, M184V, T215K | M41L, A98S, V179I, E138K M184V, T215K | 94 (M4); 119 (M9) |
| 14 | TEP | Raltegravir Saquinavir/r | 5185 | 16 657 | 5 | A98G K103N L100I | NA | 118 (M1); 181 (M4) |
| 15 | TEP | Efavirenz TDF/FTC | 40 | 62 | 12 | None | M41I, M184I G190R, M230I | 248 (M5) |
| 16 | TEP | NA | 40 | 50 | 9 | NA | Amplification negative | 74 (M6) |
| 17 | TEP | Darunavir/r TDF/FTC | 50 | 51 | 15 | NA | None | 103 (M9) |
| 18 | TEP | Efavirenz TDF/FTC | 338 | 375 | 22 | M41L, V90I, K103N, V108I, V118I, M184V, T215F | M41L, V90I, K103N, V108I, V118I, E138K, V179I, M184V, T215F | 199 (M1); 232 (M2); 119 (M3); 84 (M4); 29 (M7) |
| 19 | TEP | Atazanavir/r TDF/FTC | 40 | 186 | 9 | D67N, K70R, A98AE, K101R, K103N, M184V, K219Q | D67N, K70R, A98G, K103N, V108I, M184V, K219Q | 34 (M1); 175 (M10); 168 (M13) |
NP: naive patients; TEP: treatment‐experienced patients; TDF: tenofovir disoproxil fumarate; FTC: emtricitabine; NA: not available; cART: combination antiretroviral therapy; RPV: rilpivirine; VL: plasma HIV‐1 viral load; VF: virological failure; RT: reverse transcriptase; r: ritonavir. Mutations indicated in bold are specific mutations conferring resistance to rilpivirine; Ctrough values indicated in bold are suboptimal (< 50 ng/mL);
RPV concentration was collected 12 hours after drug intake.
3.5. RPV pharmacokinetics
A total of 779 RPV plasma concentrations were measured in the population and on average two samples per patient were available. The median values of RPV concentrations were 99 ng/mL (IQR: 68–143 ng/mL) and 101 (IQR: 68–155 ng/mL) in NP and TEP, respectively. Among these samples, 662 RPV Ctrough were available in 343 patients, respectively in 55 NP and 288 TEP. The analysis of RPV Ctrough values is presented in Table 3. Over the follow‐up period, suboptimal RPV Ctrough were observed for 64 patients (18.7%), which was similar in NP (18.2%) and TEP (18.8%). The intra‐ and inter‐subject variabilities were 27% and 54%, respectively and similar between NP and TEP.
TABLE 3.
RPV trough concentrations
| Parameters median (IQR) or n (%) | General population (n = 343) | Naive population (NP) (n = 55) | Treatment‐experienced population (TEP) (n = 288) |
|---|---|---|---|
| RPV C trough , ng/mL | 96 (66–148) | 98 (64–141) | 96 (66–149) |
| Inter‐subject variability, % | 54% | 61% | 52% |
| Intra‐subject variability, % | 27% | 29% | 26% |
| Number of RPV C trough | 662 | 112 | 550 |
| < 50 ng/mL | 89 (13.4%) | 21 (18.8%) | 68 (12.4%) |
| Number of patients with a | 64 (18.7%) | 10 (18.2%) | 54 (18.8%) |
| RPV C trough < 50 ng/mL | |||
| Only one | 48 (75%) | 4 (40%) | 44 (81.5%) |
| ≥ 2 | 16 (25%) | 6 (60%) | 10 (18.5%) |
Ctrough: trough plasma concentration; IQR: interquartile range; RPV: rilpivirine.
At M6, the median (IQR; n) RPV Ctrough was significantly lower in patients with VL > 40 copies/mL compared with patients virologically suppressed: 72 ng/mL (39–102; n = 18) vs 97 ng/mL (68–143; n = 265) (P = 0.03). Similar results were found at M12: 68 ng/mL (49–103; n = 13) vs 103 ng/mL (71–153; n = 286) (P = 0.006) (Figure 2). The proportion of patients with RPV Ctrough below 50 ng/mL was significantly higher for patients with detectable VL compared with those virologically‐suppressed at M6 (28% vs 11%; P = 0.04), as well as at M12 (31% vs 11%; P = 0.05). The analysis could not be performed at M24 and M36 due to a too small number of patients with detectable VL (n = 8 and n = 2, respectively). Figure 3 shows that the optimal RPV Ctrough cut‐off value at M12, which gives the best balance between sensitivity or specificity for efficacy, was 70 ng/mL from ROC curve analysis (AUC: 0.72, sensitivity = 0.75, specificity = 0.62).
FIGURE 2.
Boxplot of rilpivirine trough plasma concentration (RPV Ctrough) in patients with VL ≤ 40 and > 40 copies/mL at M6 and M12. Comparison of RPV Ctrough values between patients with VL ≤ 40 (green) and > 40 copies/mL (purple). The boundaries of the box indicate the 25th and 75th percentiles, respectively. The line inside the box represents the median and the whiskers correspond to the 5th and 95th percentiles. Black dots represent outliers
FIGURE 3.
ROC curve analysis to identify the optimal RPV Ctrough cut‐off value for achieving virologic response.The true positive rate (sensitivity) is plotted as a function of the true negative rate (specificity) for different RPV Ctrough cut‐off values. The diagonal line represents a sensitivity and specificity of 1. The best balance between sensitivity and specificity was achieved at the cut‐off value of 70 ng/mL (sensitivity = 75%; specificity = 62%)
4. DISCUSSION
In the present study, we present the RPV Ctrough values and antiviral efficacy results of a cohort of both NP and TEP treated with the RPV/FTC/TDF regimen and followed up for 3 years. At 48 weeks (M12), undetectable plasma VL was achieved in more than 80% of patients (85% of TEP and 80% of NP). This result was in agreement with findings in previous cohort studies. 21 , 22 Moreover, our result was also close to those reported in the ECHO/THRIVE and SPIRIT trials with 84% and 89% of virological success at 48 weeks for NP and TEP, respectively. 3 , 4 , 6 Virologic success decreases at week 96 (M24) and week 144 (M36), more significantly in NP, which can be attributed to a higher proportion of patients lost to follow‐up and treatment discontinuations in this group. Indeed, in our study, 24% of patients had stopped the treatment, more frequently in NP (30.8% vs 22.6%). These results were higher than in other published observational studies reporting discontinuation rates of 13–16% in TEP but with shorter follow‐up of 12–16 months. 17 , 18 In contrast, Sculier et al. reported 25% of discontinuation after a median time of 18.4 months. 22 Finally, our result was lower than the 34% showed by Bernaud et al., where the median time of follow‐up was similar (36 months). 26 At M12, the proportion of treatment discontinuation was also higher (8.5%) than reported in clinical trials (3% in ECHO/THRIVE 2 , 3 and 2.4% in SPIRIT 5 ), in line with other cohort studies. 21 , 22 , 26 The major reason for treatment discontinuation was toxicity, which was in agreement with previous observational cohort studies. Neuropsychiatric effects were the main reason such as in the study by Gianotti et al. 18 Gastrointestinal disorders are frequently reported as the first cause of treatment interruption. However, in our study, only two cases were observed. Our results confirm that in clinical practice, side effects lead more frequently to ARV treatment switch.
The rate of VFs observed over 3 years in our study was in accordance with those previously reported in cohort studies ranging from 1.6% to 5.9% at M6 to M12. 20 , 21 , 26 Among patients carrying RPV‐associated resistance mutations at baseline, only one failed the treatment. This patient should not have received RPV as he presented the K103N + L100I mutations at baseline that confer complete resistance to RPV. Among patients with VF, two had FTC resistance‐associated mutations and one harboured mutations associated with an intermediate resistance to TDF, which may have affected treatment efficacy.
The median observed RPV Ctrough of 96 ng/mL was slightly higher than the value of 74 ng/mL reported at week 48 in the ECHO/THRIVE PK substudy. 27 A higher inter‐subject variability was also observed, as expected with real‐life data, which may be explained by a strong variability in RPV absorption. The intra‐subject variability of 27% was lower than the inter‐subject variability reported in the ECHO/THRIVE trials. 28 A suboptimal RPV Ctrough was found in 19% of patients, which was similar in NP and TEP. This result was lower than the 29% predicted by the Pop‐PK model performed in the same context of routine clinical practice. 12 Suboptimal exposure may be the consequence of several factors such as drug–drug interactions (DDI), PK variabilities or adherence difficulties. Thus, in clinical practice, the use of therapeutic drug monitoring (TDM) for ARV is valuable to rapidly detect patients with suboptimal exposure.
A significant relationship between RPV Ctrough and virological response has been demonstrated at both M6 and M12. Indeed, we highlighted that the median RPV Ctrough value was significantly lower in patients with detectable VL compared with patients virologically suppressed (−26% at M6 and −34% at M12). Moreover, almost one third of patients with RPV Ctrough below the 50 ng/mL threshold had detectable VL at M6, which was significantly higher compared with those virologically suppressed. A similar significant trend was observed at M12. These results are consistent with those from the first RPV concentration–response model that we recently published, showing that RPV Ctrough impacts both the proportion of undetectable patients and the time to obtain virological success. 29
The optimal RPV Ctrough cut‐off value of 70 ng/mL determined using ROC curve analysis was higher than the currently used 50 ng/mL value, in line with our recently published concentration–response model. 29 Indeed, based on simulations, we had shown that an RPV Ctrough of 100 ng/mL would be necessary in induction treatment to reach more than 80% of virological success in both NP and non‐virologically‐suppressed TEP. 29 In the current study, we present analysis of the data from the entire population, including the TEP virologically suppressed at baseline, which could explain that the RPV Ctrough cut‐off found is lower than the previous identified target of 100 ng/mL. From a pharmacological and virological point of view, it is consistent to consider a higher Ctrough value in induction compared with a maintenance treatment as the intrinsic efficacy required to obtain the undetectability is higher. Therefore, we could consider two thresholds for RPV Ctrough according to the baseline virologic status of the patients. However, our target for RPV Ctrough must be interpreted with caution because our study included a small number of patients with detectable VL at baseline.
Overall, 47% of the VF (i.e. eight out of 17 with genotyping data) have developed resistance to RPV. Among these patients, one out of two had at least one suboptimal observed RPV Ctrough during the monitoring. Reasons for suboptimal RPV plasma exposure observed in half of the VF who developed resistance to RPV may be a poor or variable treatment adherence, food intake failure or unknown DDI. Moreover, two other patients had unexplained high RPV Ctrough that could also suggest an erratic treatment adherence. Furthermore, the nine other patients who failed without emergence of mutation to RPV always displayed RPV Ctrough ≥ 50 ng/mL during the follow‐up, except one for whom RPV Ctrough was suboptimal only once during the follow‐up. However, if we consider the 70 ng/mL target for RPV Ctrough, 50% of the patients who failed the regimen, and five out of eight among those who developed resistances to RPV, presented at least one suboptimal RPV Ctrough throughout their follow‐up. These results strengthen the hypothesis of the relationship between suboptimal RPV exposure and drug resistance development.
In conclusion, our study, carried out over a long period of follow‐up, highlighted the impact of suboptimal RPV Ctrough on both virologic response and the emergence of RPV mutations. Moreover, we observed that a significant proportion of patients displayed an RPV Ctrough below the target cut‐off value of 50 ng/mL during the follow‐up. Consequently, these results strengthen the use of TDM in real‐life contexts to rapidly detect suboptimal exposure and to discuss dose adjustment strategies to reduce the risk of emergence of resistance mutations leading to virologic failure. Moreover, these results, along with our recently published concentration–response model, suggest that a higher target of RPV Ctrough might be recommended in clinical practice, particularly in induction strategies.
COMPETING INTERESTS
There are no competing interest to declare.
CONTRIBUTORS
N.N.: PhD student; conceptualization, data acquisition, statistical analysis, interpretation; main writer of the manuscript. M.L.: data acquisition, writing review and editing. N.B.: methodology of the analysis, writing review and editing. F.G.: conceptualization, methodology of analysis, writing review and editing. D.D., C.T.: virologists, data validation, writing review and editing. C.D., S.B., S.M., Y.Y.: recruitment of the patients, writing review and editing. G.P.: pharmacologist, supervision of the analysis, writing review and editing. B.L.: pharmacologist, supervision of the analysis, writing review and editing. C.S.: principal investigator, pharmacologist, conceptualization, methodology, supervision of the analysis, data validation, writing and editing.
Néant N, Lê MP, Bouazza N, et al. Usefulness of therapeutic drug monitoring of rilpivirine and its relationship with virologic response and resistance in a cohort of naive and pretreated HIV‐infected patients. Br J Clin Pharmacol. 2020;86:2404–2413. 10.1111/bcp.14344
Principal investigator, Caroline Solas, Pharm D, PhD.
DATA AVAILABILITY STATEMENT
Research data are not shared.
REFERENCES
- 1. European Medicines Agency . (2011). Summary product of characteristics EVIPLERA. http://www.ema.europa.eu/docs/en_GB/document_library/EPAR_‐_Product_Information/human/002312/WC500118802.pdf.
- 2. Clay PG, Nag S, Graham CM, et al. Meta‐analysis of studies comparing single and multi‐tablet fixed dose combination HIV treatment regimens. Medicine (Baltimore). 2015;94(42):e1677. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Molina JM, Cahn P, Grinsztejn B, et al. Rilpivirine versus efavirenz with tenofovir and emtricitabine in treatment‐naive adults infected with HIV‐1 (ECHO): a phase 3 randomised double‐blind active‐controlled trial. Lancet. 2011;378(9787):238‐246. [DOI] [PubMed] [Google Scholar]
- 4. Cohen CJ, Andrade‐Villanueva J, Clotet B, et al. Rilpivirine versus efavirenz with two background nucleoside or nucleotide reverse transcriptase inhibitors in treatment‐naive adults infected with HIV‐1 (THRIVE): a phase 3, randomised, non‐inferiority trial. Lancet. 2011;378(9787):229‐237. [DOI] [PubMed] [Google Scholar]
- 5. Van Lunzen J, Antinori A, Cohen CJ, et al. Rilpivirine vs. efavirenz‐based single‐tablet regimens in treatment‐naive adults: week 96 efficacy and safety from a randomized phase 3b study. Aids. 2016;30(2):251‐259. [DOI] [PubMed] [Google Scholar]
- 6. 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(5):216‐223. [DOI] [PubMed] [Google Scholar]
- 7. Palella FJ Jr, 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(3):335‐344. [DOI] [PubMed] [Google Scholar]
- 8. Das K, Bauman J, Clark AJ, et al. High‐resolution structures of HIV‐1 reverse transcriptase/TMC278 complexes: strategic flexibility explains potency against resistance mutations. Proc Natl Acad Sci U S A. 2008;105(5):1466‐1471. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Porter DP, Kulkarni R, Fralich T, Miller M, White K. 96‐week resistance analyses of the STaR study: rilpivirine/emtricitabine/tenofovir DF versus efavirenz/emtricitabine/tenofovir DF in antiretroviral‐naive, HIV‐1‐infected subjects. HIV Clin Trials. 2015;16(1):30‐38. [DOI] [PubMed] [Google Scholar]
- 10. Cohen CJ, Molina JM, Cahn P, et al. Efficacy and safety of rilpivirine (TMC278) versus efavirenz at 48 weeks in treatment‐naive HIV‐1‐infected patients: pooled results from the phase 3 double‐blind randomized ECHO and THRIVE trials. J Acquir Immune Defic Syndr. 2012;60(1):33‐42. [DOI] [PubMed] [Google Scholar]
- 11. Néant N, Gattacceca F, Lê MP, et al. Population pharmacokinetics of rilpivirine in HIV‐1‐infected patients treated with the single‐tablet regimen rilpivirine/tenofovir/emtricitabine. Eur J Clin Pharmacol. 2018;74(4):473‐481. [DOI] [PubMed] [Google Scholar]
- 12. Aouri M, Barcelo C, Guidi M, et al. Population pharmacokinetics and pharmacogenetics analysis of rilpivirine in HIV‐1 infected individuals. Antimicrob Agents Chemother. 2016;61(1). 10.1128/AAC.00899-16 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Molla A, Korneyeva M, Gao Q, et al. Ordered accumulation of mutations in HIV protease confers resistance to ritonavir. Nat Med. 1996;2(7):760‐766. [DOI] [PubMed] [Google Scholar]
- 14. Fabbiani M, Bracciale L, Ragazzoni E, et al. Relationship between antiretroviral plasma concentration and emergence of HIV‐1 resistance mutations at treatment failure. Infection. 2011;39(6):563‐569. [DOI] [PubMed] [Google Scholar]
- 15. Brochot A, De La Rosa G, Vis P, et al. Generalised additive modelling of virologic response to the NNRTIs rilpivirine (rilpivirine, TMC278) and efavirenz (EFV) in treatment‐naive HIV‐infected patients: pooled data from ECHO and THRIVE. In: Abstracts of the Thirteenth European AIDS, Belgrade, Serbia, 2011. European AIDS Clinical Society, Brussels, Belgium Abstract PS12/7.
- 16. Rapport Morlat . Prise en charge médicale des personnes vivant avec le VIH recommandations du groupe d'experts rapport 2018. Annexe Pharmacologique 2018. https://cns.sante.fr/wp-content/uploads/2018/03/experts-vih_pharmacologie.pdf
- 17. Bagella P, De Socio GV, Ricci E, et al. Durability, safety, and efficacy of rilpivirine in clinical practice: results from the SCOLTA project. Infect Drug Resist. 2018;11:615‐623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. 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 HIV‐positive patients on stable antiretroviral therapy. J Int AIDS Soc. 2015;18(1):20037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Gagliardini R, Bandera A, Zaccarelli M, et al. 3‐year efficacy and durability of simplification to single tablet regimens: a comparison between co‐formulated efavirenz/emtricitabine/tenofovir and rilpivirine/emtricitabine/tenofovir. Antivir Ther. 2018;23(2):139‐148. [DOI] [PubMed] [Google Scholar]
- 20. Pinnetti C, Di Giambenedetto S, Maggiolo F. Simplification to co‐formulated rilpivirine/emtricitabine/tenofovir in virologically suppressed patients: data from a multicenter cohort. J Int AIDS Soc. 2014;17(4Suppl 3):19812. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Cazanave C, Reigadas S, Mazubert C, et al. Switch to rilpivirine/emtricitabine/tenofovir single‐tablet regimen of human immunodeficiency virus‐1 RNA‐suppressed patients, Agence Nationale de Recherches Sur le SIDA et les Hépatites Virales CO3 Aquitaine cohort, 2012–2014. Open Forum Infect Dis. 2015;2(1):ofv018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Sculier D, Gayet‐Ageron A, Battegay M. Rilpivirine use in the Swiss HIV cohort study: a prospective cohort study. BMC Infect Dis. 2017;17:476. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. World Medical Association . (2013). World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. Last amended by the 64th WMA general assembly, Brazil. https://www.wma.net/wp-content/uploads/2016/11/DoH-Oct2013-JAMA.pdf [DOI] [PubMed]
- 24. International conference on harmonization . (1996) ICH harmonised tripartite guideline: guideline for good clinical practice E6 (R1). https://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E6/E6_R1_Guideline.pdf
- 25. Quaranta S, Woloch C, Paccou A, Giocanti M, Solas C, Lacarelle B. Validation of an electrospray ionization LC‐MS/MS method for quantitative analysis of raltegravir, etravirine, and 9 other antiretroviral agents in human plasma samples. Ther Drug Monit. 2009;31(6):695‐702. [DOI] [PubMed] [Google Scholar]
- 26. Bernaud C, Khatchatourian L, Rodallec A, et al. Optimizing the virological success of tenofovir DF/FTC/rilpivirine in HIV‐infected naive and virologically suppressed patients through strict clinical and virological selection. Infect Dis (Lond). 2016;48(10):754‐759. [DOI] [PubMed] [Google Scholar]
- 27. Crauwels HM, Van Schaick E, Van Heeswijk RPG, Vanveggel S, Boven K, Vis P. Effect of intrinsic and extrinsic factors on the pharmacokinetics of TMC278 in antiretroviral‐naive, HIV‐1‐ infected patients in ECHO and THRIVE. J Int AIDS Soc. 2010;13(Suppl 4):P186. [Google Scholar]
- 28. Food and Drug Administration . 2010. Center for drug evaluation and research (CDER) application number: 202022orig1s000. http://www.accessdata.fda.gov/drugsatfda_docs/nda/2011/202022Orig1s000MedR.pdf
- 29. Néant N, Solas C, Bouazza N, et al. Concentration–response model of rilpivirine in a cohort of HIV‐1 infected naive and pre‐treated patients. J Antimicrob Chemother. 2019;74(7):1992‐2002. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Research data are not shared.