Five-year follow-up of patients enrolled in the NEAT 001/ANRS 143 randomized clinical trial: NEAT 001/ANRS 143 LONG TERM study Free

Abstract

Introduction

The NEAT 001/ANRS 143 study, a Phase III, open-label, non-inferiority, randomized, multicentre clinical trial, demonstrated the non-inferiority of the NRTI-sparing regimen of ritonavir-boosted darunavir + raltegravir to the standard triple drug regimen of ritonavir-boosted darunavir + tenofovir disoproxil fumarate/emtricitabine for first-line combination ART in HIV adults naive to antiretrovirals.¹ In prespecified subgroup analyses, this finding was confirmed for subjects with CD4+ counts >200 cells/mm³, whereas in subjects with ≤200 cells/mm³, the ritonavir-boosted darunavir + raltegravir arm was inferior to the standard triple regimen. A non-significant difference toward more failures in the NRTI-sparing treatment group was also observed in subjects with baseline HIV-1 RNA of ≥100 000 copies/mL.

International HIV treatment guidelines have recommended for the past two decades that individuals should initiate ART with three antiretroviral drugs. Since 2017, the preference has been a combination of two NRTIs and an integrase strand transfer inhibitor, either bictegravir, dolutegravir or raltegravir.^2,3 It is only recently that a dual regimen, combining dolutegravir and lamivudine, has demonstrated an efficacy similar to that of an integrase strand transfer inhibitor-based triple regimen, and this combination was ranked in December 2019 as one of the recommended initial ART regimens.^4,5 However, the association of dolutegravir with potential teratogenic effects (neural tube defects), neuropsychiatric symptoms and weight gain have raised concern.^6–8 There are also recent reports on increased risk of weight gain with tenofovir alafenamide.⁸ Based on NEAT 001 results, guidelines indicate that when abacavir, tenofovir alafenamide and tenofovir disoproxil fumarate cannot be used or are not optimal, a regimen of ritonavir-boosted darunavir once daily plus raltegravir twice daily could be considered in subjects with HIV RNA <100 000 copies/mL and CD4+ cell count >200 cells/mm³.^2,3 However, few long-term data are available in subjects having initiated ART with an NRTI-sparing regimen. We assessed the 5 year outcome of subjects enrolled in the NEAT 001/ANRS 143 randomized clinical trial (comparing ritonavir-boosted darunavir + raltegravir versus darunavir + tenofovir disoproxil fumarate/emtricitabine).

Patients and methods

The NEAT 001/ANRS 143 trial was a Phase III, open-label, randomized trial performed in 78 sites in 15 European countries. The full study design and procedures are described elsewhere (ClinicalTrials.gov: NCT01066962).¹ Eight hundred and five antiretroviral-naive adults were randomized between August 2010 and September 2011 in a 1:1 ratio to receive either two pills of darunavir 400 mg and ritonavir 100 mg once daily combined with one pill of raltegravir 400 mg twice daily (a five pills twice-daily regimen) or one pill of tenofovir disoproxil fumarate/emtricitabine 245/200 mg fixed-dose combination once daily (a four pills once-daily regimen). All participants remained in the study until the last patient enrolled had completed the Week 96 visit (W96; October 2013). To obtain long-term outcomes of individuals enrolled in NEAT 001, we conducted a retrospective non-interventional audit, in 2015–17, to collect information on clinical progression (AIDS and non-AIDS serious events), occurrence of significant (grade 4, death) adverse events, CD4 cell counts, HIV RNA, change in ART and reasons for changes, up to 5 years post-enrolment. Data were obtained from subjects’ charts and/or electronic medical records through an anonymized electronic case report form. We also assessed, over the full treatment period (NEAT 001/ANRS 143 and post-trial audit), mean weight (kg) change and mean creatinine (μmol/L) increase according to exposure for >50% of follow-up to an integrase strand transfer inhibitor versus never exposed to an integrase strand transfer inhibitor and exposure for >50% of follow-up to tenofovir disoproxil fumarate versus never exposed to tenofovir disoproxil fumarate, respectively. Data were analysed without imputation. χ² (or Fisher’s exact) or Student’s t-tests were used for comparisons. This long-term audit was conducted in 39 sites in eight countries, from the initial NEAT 001/ANRS 143 sites/countries.

Ethics

The study was approved by The French National Agency for Medicines and Health Products (ANSM) (no. A100-154, 17 February 2010). The Ethics Committee of each participating centre approved the study protocol. Because of the retrospective nature of data collection, only subjects’ non-opposition was collected.

Results

Among the 39 sites taking part in the NEAT 001/ANRS 143 LONG TERM study (NLT), last NEAT 001 visit (W96) was conducted in 485/546 randomized subjects. Of these 485 individuals, 430 were enrolled in NLT (201 of the 231 initially randomized to raltegravir, and 229 of the 254 initially randomized to tenofovir disoproxil fumarate/emtricitabine), with a median follow-up of 44.5 months post-W96 (IQR: 41.7–47.3) (Table 1). The baseline characteristics of the 430 enrolled patients were similar overall to those of the global population (Table S1, available as Supplementary data at JAC Online). The proportion of AIDS, non-AIDS events, virological rebound and serious adverse events occurring during NLT follow-up did not differ between groups (Table 1). Considering all follow-up, since NEAT 001, rates of ART discontinuation for virological failure were 6.8% and 5.9% in the raltegravir and tenofovir disoproxil fumarate groups, respectively (P = 0.90), and rates of ART discontinuation for adverse event were 10.2% and 15.8%, respectively (P = 0.15); discontinuations for virological failure were more frequent in subjects with baseline CD4 <200 cells/mm³ (11.9% versus 5.3%; P = 0.077). During NLT, the rates of discontinuation for virological failure and of adverse events were low: 2.05% and 4.2%, respectively, between W96 and Year 4 of follow-up, and 0% and 2.6%, respectively, between Year 4 and 5 of follow-up (Table 1); between W96 and Year 4, 3.28% of subjects with baseline CD4 <200 cells/mm³ and 1.83% with baseline CD4  >200 cells/mm³ had discontinuations for virological failure (P = 0.37), whereas between Year 4 and Year 5, one of the 380 subjects discontinued for virological failure (Table 1). At the last available data, >50% of participants in both groups were receiving boosted darunavir, with no difference between the initial randomized groups (54.2% versus 49.3%, P = 0.39) (Table 1), 12.4% of participants were on raltegravir, significantly more frequently in the initial raltegravir group (24.9% versus 1.5%, P < 0.001) (Table 1); the proportions of total participants on boosted darunavir monotherapy or dual therapy of boosted darunavir plus lamivudine were 8.5% versus 7.9% in the raltegravir versus the tenofovir disoproxil fumarate/emtricitabine group (P = 0.97). Mean weight gain between NEAT 001 baseline and last follow-up was 3.33 ± 8.19 kg versus 4.08 ± 6.49 kg in subjects exposed for >50% of follow-up to an integrase strand transfer inhibitor versus those never exposed to an integrase strand transfer inhibitor (P = 0.48). Mean increase in serum creatinine between baseline and last available data was 9.33 ± 13.61 μmol/L versus 5.94 ± 14.05 μmol/L in subjects exposed for >50% of follow-up to tenofovir disoproxil fumarate versus never exposed to tenofovir disoproxil fumarate (P = 0.067). Mean changes in estimated glomerular filtration rate (Cockroft–Gault formula) during follow-up were not significantly different between subjects exposed for >50% of follow-up to tenofovir disoproxil fumarate or never exposed to tenofovir disoproxil fumarate (−7.45 ± 22.07 mL/min versus −11.64 ± 23.85 mL/min, respectively; P = 0.25).

Discussion

Previous studies of NRTI-sparing regimens, in which an NNRTI was combined with a ritonavir-boosted PI, as alternative first-line ART, showed no compelling evidence of benefit over standard regimens.^9,10 NEAT 001/ANRS 143 was the first study to demonstrate that an NRTI-sparing regimen consisting of a dual combination of a ritonavir-boosted PI and an integrase strand transfer inhibitor was well tolerated and exhibited non-inferior efficacy to a standard triple combination of two NRTIs + a ritonavir-boosted PI, although there were more failures in those with CD4  < 200 cells/mm³ at enrolment.¹ However, this clinical trial was conducted over 96 weeks of treatment, and long-term data are needed to clarify whether initiating ART with such an NRTI-sparing strategy is associated with any long-term benefit or hazard. Raltegravir remains an interesting option in some cases, such as second-generation integrase strand transfer inhibitor neuropsychiatric side effects, weight gain or drug–drug interactions. This retrospective audit of long-term outcomes was performed a median of nearly 6 years after treatment initiation with ritonavir-boosted darunavir + raltegravir. Most subjects experienced change(s) in their initial regimen, which mainly reflect the availability of new antiretrovirals and/or more convenient regimens over time, especially more diverse single-tablet regimens and recent use of dual therapies. Indeed, at last follow-up, three-quarters of subjects enrolled in NEAT 001/ANRS 143 were no longer receiving their initial treatment, with a higher but not significant difference for those randomized to twice-daily raltegravir + once-daily ritonavir-boosted darunavir. Interestingly, although boosted PIs are no longer recommended as preferred first-line ART in major guidelines and are used less and less in Europe for maintenance therapy, mainly to avoid ritonavir- or cobicistat-induced drug–drug interactions,¹¹ nearly half of our study population was still receiving boosted darunavir at last follow-up, with similar proportions between subjects who initiated ART with the latter in combination with raltegravir or with tenofovir disoproxil fumarate/emtricitabine. Studies have demonstrated the high virological efficacy of ritonavir-boosted darunavir + lamivudine dual therapy, as both an initial and a maintenance regimen, with the absence of emergence of PI resistance mutations, as well as a good safety profile.^12,13 Such relatively high persistence on ritonavir-boosted darunavir after 6 years of treatment initiation may also be related to good tolerability and the wish to spare other classes in subjects doing well on such an antiretroviral regimen, with the possibility to simplify either to a single-tablet regimen (darunavir/cobicistat/tenofovir alafenamide/emtricitabine) or to boosted darunavir dual or monotherapy.^14,15 In contrast, few subjects were receiving raltegravir at the end of the 6 year period of follow-up, almost exclusively in the subgroup of individuals initially randomized to ritonavir-boosted darunavir + raltegravir. This low rate of persistence on (or new use of) raltegravir after 6 years on ART did not seem to be related to an excess of virological failure or treatment-limiting tolerability, and is probably explained by the request to move away from twice-daily to daily therapy (raltegravir 600 mg two pills once daily not being available before 2017, at the end of our study),¹⁶ absence of a fixed drug regimen and availability of much easier, one-pill once-daily, regimens, including within the integrase strand transfer inhibitor class. Notwithstanding this, a quarter of subjects initiated on raltegravir were still receiving this antiretroviral at 6 year follow-up, which may reflect the tolerability advantages and the benefit of a low risk of drug–drug interactions, counterbalancing the need for twice-daily intake.¹⁷ We have previously shown, in NEAT 001/ANRS 143, that, after 96 weeks on therapy, subjects on ritonavir-boosted darunavir + raltegravir had a similar degree of satisfaction with the extent of unwanted side effects, and of desire to continue their treatment unchanged, as those on ritonavir-boosted darunavir + tenofovir disoproxil fumarate/emtricitabine.¹⁸ The proportion of discontinuations for reasons of tolerability was 13.2%, with no difference between the two initial treatment groups. This rate was much higher than in Phase III randomized clinical trials of ritonavir-boosted darunavir + tenofovir disoproxil fumarate/emtricitabine (4.7% in the ARTEMIS study)¹⁹ or raltegravir + tenofovir disoproxil fumarate/emtricitabine (5.0% at 5 years in the STARTMRK study)²⁰ which could reflect the real-life situation of post-trial follow-up, where patients and physicians are more prone to modify an antiretroviral regimen as a result of minor adverse events, owing to the availability of many viable alternative regimens.²¹ The rate of serious adverse events and absence of virological suppression (HIV-1 RNA >50 copies/mL) remained low throughout post-trial follow-up, reflecting the good safety profile and high success rate with most recent antiretroviral regimens.²²

In conclusion, after a median of 5.6 years, individuals initiating ART with ritonavir-boosted darunavir + raltegravir or ritonavir-boosted darunavir + tenofovir disoproxil fumarate/emtricitabine experienced few serious clinical events. A quarter of subjects were still receiving their initial regimen, most discontinuations being for reasons unrelated to adverse events or virological failure.

This long-term follow-up study has limitations: retrospective data collection, loss of randomization with risk of bias, absence of enrolment of all initial participants although information was obtained in 88.7%, and discontinuations for reasons other than treatment-related events. The latter point reflects real-life situations, where patients having finished the trial period might benefit from treatment modifications for simplification, comfort, prevention of drug interactions, and so on.

In conclusion, our report describes the long-term follow-up of patients previously included in a large clinical trial of first-line ART with no major differences in outcomes in individuals initiated on a ritonavir-boosted darunavir-containing regimen with either two NRTIs or raltegravir.

Acknowledgements

We thank all the patients participating in the study, Thomas Jovelin (NADIS^® Data Manager, COREVIH Pays de la Loire, France), Maxime Lebeaupin (Project Leader, DRCI, CHU Nantes, France) and Tanguy Roman (Data Manager, DRCI, CHU Nantes, France).

Members of the NEAT 001/ANRS 143 LONG TERM STUDY GROUP

Study team: François Raffi, Maelle Ningre, Thomas Jovelin, Tanguy Roman, Maxime Lebeaupin, Sabrina Banon. Belgium: Stéphane de Wit, MD, Coca Nescoi, MD (CHU St Pierre, Brussels). France: Yann Quertainmont, MD, Martine Molle, CRA (Bicêtre Hospital, Paris), Yazdan, Yazdanpanah, MD, Zélie Julia, CRA (Bichat Hospital, Paris), Lionel Piroth, MD, Sandrine Gohier, CRA (Dijon), Thomas Guimard, MD, Armelle Pegeot, CRA (La Roche sur Yon), David Zucman, MD, Dominique Bornarel, CRA (Foch), Jacques Reynes, MD, Christine Tramoni, CRA (Montpellier), O. Robineau, MD, M-Christine Marien, CRA (Tourcoing), L. Weiss, MD, Corinne Jung, CRA (Pompidou Hospital, Paris), Jean-Daniel Lelièvre, MD, Chrystel Chesnel, CRA (Henri Mondor), F. Raffi, MD, Ernesto Paredes, CRA (Nantes), Philippe Morlat, MD, Sabrina Caldato, CRA (Bordeaux), Didier Neau, MD, Séverine Le Puil, CRA (Bordeaux), André Cabié, MD, Christophe Cerland, CRA (Fort de France), Christine Katlama, MD, Ludovic Lenclume, CRA (Pitié Salpêtrière Hospital, Paris), Jean-Michel Molina, MD, Astrid Olier, CRA (Saint Louis Hospital, Paris), Pierre Tattevin, MD, Anne Meheut, CRA (Rennes), Pierre Delobel, MD, Sandra Lagarrigue, CRA (Toulouse), Didier Neau, MD (Bordeaux), Pierre-Marie Girard, MD, Manuela Le Cam, CRA (Saint Antoine Hospital, Paris). Germany: Heiko Jessen, MD, Anna Müller-Claud, CRA (Gemeinschaftspraxis Jessen-Jessen-Stein), Christoph Stephan, MD, Claudia Wengenroth, CRA (Wolfgang Goethe University, Frankfurt), Christian Hoffmann, MD, Susanne Heesch, CRA (ICH study centre, Hamburg) Stefan Esser, MD, Pia Schenk-Westkamp, CRA (Essen), Albrecht Stoehr MD, Meike Brenner (Asklepios-Klinik St. Georg, Hamburg), Jurgen Rockstroh, MD, Brigitta Becker, CRA (Rheinische Friedrich-Wilhelms-Universität Bonn). Ireland: Paddy Mallon, MD, Alan Macken (Mater Misericordiae, Dublin). Italy: Andrea Antinori, MD, Maria Maddalena Plazzi, CRA (Istituto Naziona e per le Malattie ‘Lazzaro Spallanzani’, Roma). Poland: Andrzej Horban, MD, Elzbieta Bakowska, CRA (Warsaw). Sweden: Anders Thalme, MD, Sofia Sandberg, CRA (Stockholm), Magnus Gisslen, MD, Lissie Johansson, CRA (Gothenburg), Carl-Johan Treutiger, MD, Elisabet Storgard, CRA (Stockholm). UK: Ian Williams, MD, Andrea Cartier, CRA (Mortimer Market Centre, London), Margaret Johnson, MD, Nargis Hemat, CRA (Royal Free Hospital, London), Mark Gompels, MD, Louise Jennings, CRA (Southmead Hospital, Bristol), Alan Winston, MD, Ajerico Del Rosario, CRA (Saint Mary’s Hospital, London), Anton Pozniak, MD, Jacob Lowman, CRA (Saint Stephen’s Centre Chelsea and Westminster Hospital, London), Julie Fox, MD, Alice Sharp, CRA (Saint Thomas’ Hospital, London), Clifford Leen, MD, Sheila Morris, CRA (Western General Hospital, Edinburgh).

Funding

The study was funded by the NEAT-ID Foundation through a grant from Janssen Laboratories (TMC114IFD4034). Nantes University Hospital was the sponsor (RC15_0003).

Transparency declarations

F.R. received research funding or honoraria from or consulted for Gilead Sciences, Janssen, Merck and ViiV Healthcare. A.P. has served on advisory boards and presented at educational symposia of Gilead Sciences, Janssen, Merck and ViiV Healthcare, and his hospital has received research funding from Gilead Sciences, Janssen, Merck and ViiV Healthcare. J-M.M. received a grant from Gilead and advisory board fees from Gilead, ViiV Healthcare, Merck and Sanofi. A.A. has served as a paid consultant to Gilead Sciences, Janssen-Cilag, Merck and ViiV Healthcare, and received research funding through National Institute for Infectious Diseases Lazzaro Spallanzani IRCCS from Gilead Sciences, Janssen-Cilag and ViiV Healthcare. S. De Wit received grants from Gilead Sciences, Janssen, MSD and ViiV Healthcare, all paid to his institution. All other authors: none to declare.

Supplementary data

Table S1 is available as Supplementary data at JAC Online.

References

Author notes