Maintenance darunavir/ritonavir monotherapy to prevent perinatal HIV transmission, ANRS-MIE 168 MONOGEST study

Abstract

Objectives

Because NRTIs can have fetal toxicities, we evaluated a perinatal NRTI-sparing strategy to prevent perinatal HIV transmission. Our primary objective was to determine the proportion maintaining a viral load (VL) of <50 copies/mL up to delivery on darunavir/ritonavir monotherapy, without requiring treatment intensification.

Methods

In a one-arm, multicentre Phase 2 clinical trial, eligible patients in the first trimester of pregnancy on ART with plasma VL < 50 copies/mL received maintenance monotherapy with darunavir/ritonavir, 600/100 mg twice daily. VL was monitored monthly. ART was intensified in the case of VL > 50 copies/mL. Neonates received nevirapine prophylaxis for 14 days.

Results

Of 89 patients switching to darunavir/ritonavir monotherapy, 4 miscarried before 22 weeks’ gestation, 2 changed treatment for elevated liver enzymes without virological failure, and 83 were evaluable for the main outcome. Six had virological failure confirmed on a repeat sample (median VL = 193 copies/mL; range 78–644), including two before switching to monotherapy. In these six cases, ART was intensified with tenofovir disoproxil fumarate/emtricitabine. The success rate was 75/83, 90.4% (95% CI, 81.9%–95.7%) considering two patients with VL missing at delivery as failures, and 77/83, 92.8% (95% CI, 84.9%–97.3%) when considering them as successes since both had undetectable VL on darunavir/ritonavir throughout pregnancy. In ITT, the last available VL before delivery was <50 copies/mL in all of the patients. There was no case of perinatal HIV transmission.

Conclusions

Darunavir/ritonavir maintenance monotherapy required intensification in nearly 10% of cases. This limits its widespread use, thus other regimens should be evaluated in order to limit exposure to antiretrovirals, particularly NRTIs, during pregnancy.

Introduction

Viral suppression throughout pregnancy with ART eliminates perinatal transmission (PT) of HIV, regardless of the number or type of antiretroviral drugs.¹ Current guidelines recommend triple therapy with two NRTIs for pregnant women living with HIV, associated with a boosted PI (PI/ritonavir), an NNRTI or an integrase strand transfer inhibitor (INSTI).^2–4 In France, the first-line association was two NRTIs and a PI/ritonavir since the early 2000s.⁵

However, NRTI-related toxicities are reported in fetuses and neonates,⁶ including mitochondrial toxicities following exposure to zidovudine and lamivudine in clinical^7,8,9,10 and experimental studies.^11,12 Nuclear DNA alterations are reported in cord blood cells with zidovudine and tenofovir.^13,14 Long-term follow-up of children exposed to zidovudine in utero has shown conflicting evidence of subclinical myocardial dysfunction.^15,16,17 Exposure to didanosine, an NRTI which is no longer used, was associated with some childhood cancers.¹⁸ There are conflicting reports of an association between tenofovir exposure and altered growth and bone mineral content.^19,20 Also, postnatal zidovudine as prophylaxis is related to haematological abnormalities ^21,22 and hyperlactataemia²³ in infants.

Paradoxically, despite known toxicities, NRTIs are still recommended in pregnancy and for neonatal prophylaxis. No NRTI-sparing regimen has yet been studied in pregnant women, except for a randomized controlled trial of lopinavir/ritonavir monotherapy.^24,25

The intrapartum component of PT prevention with zidovudine is not required when maternal plasma viral load (VL) is suppressed.²⁶ For postnatal prophylaxis, zidovudine remains standard care in most countries, but nevirapine prophylaxis is now an option in France,^5,27 based on reassuring neonatal data.^{28,29,30,28,31}

After in utero exposure to PIs, no fetal toxicity was described except for increased bilirubinaemia with atazanavir/ritonavir.³² Placental transfer of PIs is lower than for NRTIs.³³ There is now a long experience on the use of darunavir/ritonavir during pregnancy.^{34–38,39,40} Dual therapy with darunavir/ritonavir and lamivudine or emtricitabine was reported in a small retrospective study.⁴¹

Ritonavir-boosted darunavir monotherapy has been used in non-pregnant persons for its virological potency, tolerability and high genetic barrier to resistance NRTI-sparing therapy but has not been previously studied in pregnancy. With darunavir/ritonavir monotherapy, 80% to 90% of patients^42,43 maintained viral suppression at 48 to 96 weeks. In cases of virological failure, the risk of emergence of drug-resistance mutations was low and resumption of triple therapy was effective.

The purpose of this study was to evaluate, in women starting pregnancy with suppressed viraemia on combination ART, an NRTI-sparing maintenance strategy using darunavir/ritonavir monotherapy during pregnancy and nevirapine monotherapy as neonatal prophylaxis.

The primary objective was to estimate the proportion of women maintaining a VL of <50 copies/mL until delivery after switching to darunavir/ritonavir monotherapy early in pregnancy, without requiring treatment intensification for virological failure. Secondary objectives were to estimate the proportion of women with VL < 50 copies/mL at delivery in ITT, the proportions of miscarriage, of treatment changes for intolerance or other reasons, to identify factors associated with virological failures, and to determine darunavir and ritonavir plasma concentrations.

Methods

A one-arm, open-label, multicentre Phase II clinical trial was performed in 24 ANRS-MIE EPF centres in France.

Study population and intervention

Patients in the first trimester of pregnancy with HIV-1 and long-term antiretroviral success were informed about the study by their providers; those who consented were assessed for eligibility at the screening visit, and if eligible were offered enrolment with informed written consent.

Eligibility

Inclusion criteria were pregnancy, gestational age under 15 weeks, maternal age ≥18 years, documented HIV-1 infection, current treatment with at least two ARVs, viral suppression for at least 12 months, defined by VL < 50 copies/mL (a single value between 50 and 399 copies/mL was not an exclusion criterion if followed by two successive VL < 50 copies/mL at least 1 month before enrolment), CD4 cells ≥250 cells/mm³, healthcare coverage and informed written consent. The inclusion criteria for enrolment of the neonate in the paediatric component of the study were enrolment of the mother in the trial, and informed written consent by both parents or legal guardians.

Non-inclusion criteria were infection with HIV-2, age <18 years, HIV-1 plasma VL of (i) >50 copies/mL at screening; or (ii) >400 copies/mL or between 50 and 399 copies/mL on two consecutive measurements in the last 12 months, CD4 count <250 cells/mm³, history of CD4 count nadir <200 cells/mm³, history of treatment failure, intolerance to darunavir or ritonavir, HBV coinfection (HBsAg positive and/or detectable HBV DNA), history of resistance to any PI or to nevirapine, intended absence from the study centre, or person under a legal guardianship. Infants were excluded from the paediatric component of the study in the absence of consent by both parents or legal guardians.

Study regimen and follow-up

The maternal regimen was darunavir/ritonavir (600/100 mg orally twice daily). Women on darunavir/ritonavir-based ART combination therapy switched directly to darunavir/ritonavir monotherapy. Women who were taking other ART combinations replaced their third antiretroviral (PI/ritonavir, NNRTI or INSTI) with darunavir/ritonavir 600/100 mg twice daily, while maintaining the previous NRTI backbone to check for tolerance. If clinical tolerance of darunavir/ritonavir was satisfactory after 2 weeks, the NRTIs were stopped. In the case of intolerance, another ART was decided by the investigator, and follow-up was continued.

Plasma HIV VL was monitored monthly, according to usual pregnancy guidelines. In the case of VL ≥ 50 copies/mL, testing was repeated within 14 days. If the VL was <50 copies/mL, darunavir/ritonavir monotherapy was continued, whereas if it remained ≥50 copies/mL, treatment was intensified. The choice of the intensified regimen was left up to the clinicians.

At delivery, no intrapartum zidovudine was added if the most recent maternal VL was below 50 copies/mL, except for complications such as chorioamnionitis or intrapartum haemorrhage, according to French guidelines.⁵

After delivery, the choice of maternal ART was left up to the clinician and patient. As recommended in France, there was no breastfeeding.

Neonatal prophylaxis

Neonates received prophylaxis with nevirapine once daily for 14 days at a fixed dose: 15 mg for birthweights of >2.5 kg, 10 mg for between 2 and 2.5 kg and 2 mg/kg/day for <2 kg, according to French⁵ and WHO guidelines.⁴⁴ Preterm neonates (before 37 weeks of gestation, WG) received 4 weeks of zidovudine prophylaxis. In the case of maternal virological failure, usual guidelines were to be applied.⁵ Enrolment of neonates required additional consent from both parents. In the case of the co-parent not signing consent, particularly when the mother did not accept to inform the co-parent of her HIV status, infant prophylaxis was prescribed by their paediatrician following usual guidelines.

Maternal study visits were monthly with the obstetrician and HIV specialist as per usual care, with physical examination and biology, including plasma HIV-1-RNA, blood cell counts, liver enzymes, creatinine, urine albumin, CD4/CD8 counts at each trimester, first-trimester TORCH screening, and hepatitis C and hepatitis B, with HBV DNA testing in the case of positive anti-HBc antibodies, even when HBsAg was negative. Gestational diabetes was screened with a 75 g oral glucose challenge test at 24–28 WG. A six-point self-completed questionnaire for adherence was administered at each visit. Additional visits were performed in the case of VL increase or biological or clinical anomalies.

Total HIV-1 DNA was quantified on stored samples by ultrasensitive quantitative PCR (qPCR) in blood samples collected at enrolment and at delivery with the Generic DNA cell Biocentric assay.⁴⁵ Plasma HIV-1 RNA was quantified by ultrasensitive RT–PCR on large plasma volumes adapted on the Alinity m HIV-1 assay (Alinity m HIV-1) (quantification threshold: 6 copies/mL).

Infants’ follow-up was according to practice guidelines⁴⁶ until 6 months, with physical and biological tests including HIV-1-RNA and/or DNA PCR for diagnosis of HIV status.⁴⁷

Endpoints

The primary endpoint was the rate of maternal plasma VL < 50 copies/mL maintained until delivery after switching to darunavir/ritonavir monotherapy, without requiring treatment intensification for virological failure.

Secondary endpoints were the proportion of plasma VL < 50 copies/mL at delivery in an ITT analysis, treatment changes for intolerance, toxicity or other reasons, and factors associated with virological failure. Pregnancy outcomes were described. Any case of mother-to-child transmission was to be considered as a serious adverse event and analysed immediately.

Determination of darunavir and ritonavir plasma concentrations

Total darunavir and ritonavir and unbound (darunavir only) plasma concentrations were determined at Day 0, Week (W)4, W24–28, W32 and delivery, 12 h after the last drug intake (C_min) by UPLC coupled with tandem MS (Waters Acquity UPLC-TQD, Milford, MA, USA).^48,49

Plasma protein binding analysis involved an ultrafiltration assay (Centrifree^®; Millipore, Molsheim, France). Limits of quantification (LOQs) were 5 ng/mL for darunavir and 10 ng/mL for ritonavir. Total darunavir C_min values were interpreted according to the 10-fold in vitro protein binding-corrected median effective concentration required to induce 50% virological response (EC₅₀) of 550 ng/mL.⁴⁹

Statistical methods

The trial was designed as a one-stage Phase 2 trial, with an expected observed success rate of 95%, power set at 80% and a two-sided alpha of 5% for the exact test of the comparison of the observed proportion of success against the minimum bound of success set at 85%. The 95% CI of proportions was estimated using the exact Clopper–Pearson method for binomial proportion. We initially planned to recruit 120 evaluable patients, but based on the accrual rate, the protocol was amended to enrol a minimum of 80 evaluable patients.

The primary analysis concerned the proportion of virological success defined by a VL of <50 copies/mL at childbirth among women who remained on darunavir/ritonavir monotherapy, excluding miscarriages before 22 WG and/or treatment changes for intolerance. The protocol planned to consider values not measured within 7 days of delivery as failures in the primary analysis and to perform a case-by-case review of such cases if all VLs during the pregnancy were <50 copies/mL.

The proportion of virological success was also estimated in a secondary ITT analysis among all women who were enrolled to monotherapy, regardless of subsequent changes.

Initial characteristics potentially associated with virological failure were analysed by comparing percentages (categorical variables), means and medians (continuous variables) between women with failure and women whose darunavir/ritonavir monotherapy was maintained until delivery. Analyses were carried out with the software SAS^® V9 and STATA^® V11.

Ethics

All women were enrolled with their informed, written consent. The trial was registered (EudraCT # 2015-002542-31, Clinicaltrials.gov #NCT02738502) and was approved by the Committee for Protection of Persons (CPP Ile de France VII) on 22 January 2016 and the French Medicines Agency on 17 December 2015.

Results

We enrolled 91 women between 6 June 2016 and 4 June 2019 (Figure 1), at a median of 14 WG (IQR 12–15). Ten women who started their pregnancies with regimens not containing darunavir/ritonavir first switched to darunavir/ritonavir-based triple therapy for 2 weeks before stopping the NRTI component. Another 81 with darunavir/ritonavir-based regimens at screening switched directly to darunavir/ritonavir monotherapy. Two patients did not proceed to darunavir/ritonavir monotherapy, one for elevated liver enzymes at enrolment, and another because of a miscarriage. Overall, 89 women started darunavir/ritonavir monotherapy, among whom 4 had miscarriages before 22 WG.

Among 85 women who delivered at 22 WG or later, 2 had elevated liver enzymes and switched to tenofovir disoproxil fumarate/emtricitabine + raltegravir (400 mg twice daily), including one who withdrew her consent at 26 WG, so no subsequent data are available. Both had VL < 50 copies/mL at the time of switch. Thus, 84 patients were followed until delivery and 83 were evaluable for the primary endpoint.

Characteristics at enrolment are described in Table 1. Baseline characteristics of women who miscarried did not differ from the overall population and none stopped darunavir/ritonavir monotherapy for adverse effects or virological failure.

Pregnancy outcomes were unremarkable, including preterm birth, birthweight, (Table 2). In two cases, IV zidovudine was administered because of placental bleeding during labour.

Virological endpoints

Treatment was intensified for virological failure in 6/83 (median VL = 193 copies/mL; range 78–252, confirmed on a second sample with a median VL of 178 copies/mL, range 64–1703) (Table 3). Another two patients had VL missing at delivery; both had monthly undetectable VL on darunavir/ritonavir throughout their pregnancies, including 33 and 13 days before delivery. When considering them as failures, the success rate was 75/83, 90.4% (95% CI with the binomial exact method, 81.9%–95.7%), which was not significantly greater than the minimum success rate, which we had set at 85% (exact two-sided test, P = 0.22). When considering them as successes, the success rate was 77/83, 92.8% (95% CI with the binomial exact method, 84.9%–97.3%). This 92.8% success rate was greater, according to the exact two-sided test, than the minimum success rate as defined in the study protocol (P = 0.047).

Among women who maintained viral suppression through delivery, five had a single VL measurement ≥50 copies/mL during the pregnancy with subsequent values <50 copies/mL without ART change, thus considered as viral blips.

For three patients, VL was already ≥50 copies/mL on the day of starting monotherapy (Table 4). When checked within the following 2 weeks according to the protocol, one virological failure was confirmed and treatment was intensified (with a delay of 56 days) and two patients had VL < 50 copies/mL, one of whom had virological failure 140 days later requiring treatment intensification, while the other maintained virological success on monotherapy. Among the six patients with virological failure, the time from darunavir/ritonavir monotherapy to virological failure ranged from 0 to 139 days (Table 4).

Treatment intensification consisted of adding tenofovir disoproxil fumarate/emtricitabine to darunavir/ritonavir and was successful in all six patients. In ITT analysis, the last VL during the pregnancy was below 50 copies/mL for all (88/88) patients enrolled on darunavir/ritonavir monotherapy (excluding one who withdrew consent).

The 6 women who required ART intensification had a lower median baseline CD4 cell count (516 versus 725 cells/mm³, respectively; P = 0.027), and had a higher proportion of highly sensitive VL ≥ 6 copies/mL at enrolment (50% versus 11.8%, P = 0.038), compared with the 77 women who maintained virological suppression on monotherapy (Table 5). The PBMC-associated HIV-DNA concentration was not significantly higher in the failure group. Genotyping resistance testing was performed for the six cases of virological failure, showing no acquisition of resistance-associated mutations. There were one or more issues with treatment adherence in all six patients who had virological failure; however, adherence issues did not differ significantly from the patients who maintained virological success (data not shown).

Pharmacological results

Total and unbound darunavir C_min post darunavir/ritonavir (600/100 mg twice daily) regimen were stable throughout pregnancy (Table 6). They were adequate, with total darunavir C_min > 550 ng/mL in all women except two at W4 and one at W24–28. Unbound darunavir C_min values were >55 ng/mL in all women except one at W4, one at W24–28, one at W32 and one at delivery. Trough darunavir concentrations in the six failures, at enrolment, W4 and nearest to the failure, were not lower than in patients who maintained virological success.

Adverse events

In addition to the 2 cases where darunavir/ritonavir was stopped for intolerance, there were 10 cases of liver enzyme elevation, of which 3 were attributable to pregnancy-related conditions (1 case of pre-eclampsia with HELLP syndrome and 2 cases of cholestasis of pregnancy) and 7 that were not explained by hepatitis or pregnancy-related conditions. The rate of preterm delivery was 9.5% and the rate of birth defects was 3.6%. No cases of mother-to-child transmission were reported.

Discussion

Main findings

Treatment with darunavir/ritonavir monotherapy was successful in 90.4% when considering two missing VLs at delivery as failures, and 92.8% in the sensitivity analysis counting the two patients with missing VLs at delivery as successes since they had virological suppression throughout the pregnancy. The success rate was significantly higher than the minimal success rate set in the study protocol in the sensitivity analysis, but not in the primary analysis.

ITT analysis showed that no patient had plasma HIV-RNA above 50 copies/mL near delivery. This is reassuring regarding the risk of PT, providing monthly monitoring of maternal VL, as recommended for all pregnant women with HIV.⁵

The treatment regimen was chosen on the basis of data showing plasma concentrations of darunavir decreased in the second and third trimesters of pregnancy and target concentrations were best attained with twice-daily dosage.^38,50 Plasma concentrations of darunavir monotherapy were above the antiviral efficacy threshold in most patients from enrolment through to delivery, suggesting overall good adherence to this regimen, despite the twice-daily regimen. The virological failures did not seem clearly related to insufficient concentrations. Inter-individual plasma pharmacokinetics variability was as expected for boosted PIs, and low compared with the integrase inhibitor raltegravir.⁵¹ Protein binding of darunavir was stable throughout pregnancy despite small changes related to plasma protein concentrations.

The rate of virological failure in our study was lower than observed in the French perinatal cohort in the same period among women taking classical triple combination therapies, where over 10% had VLs above 50 copies/mL near delivery.^1,52 In pregnant women, as well as other populations, the main determinant of treatment failure is poor adherence. An important consideration when making any change to an ART regimen is whether this will improve or impair adherence. Among women switching ART in pregnancy for concern about fetal safety, previous reports showed contrasting findings.^53,54

The factors predictive of success were undetectable ultrasensitive HIV-1 RNA and high baseline CD4 cell count, as previously reported outside pregnancy.⁵⁵ In non-pregnant persons, other factors predictive of virological success of boosted-PI monotherapy were good compliance, low HIV-1 proviral DNA and prolonged control of viral replication.^55,56,57

Regarding safety, monotherapy avoids issues related to nucleoside analogue exposure, but not those related to darunavir/ritonavir. Liver enzyme elevation, which has been previously reported with boosted-PI therapy,⁵⁸ was observed in the absence of common pregnancy-related conditions (cholestasis of pregnancy or pre-eclampsia) in seven of our patients. While boosted-PI regimens have been associated with preterm delivery,^59,60 the incidence was 9%, which is unremarkable and below the rate of 14% in women receiving triple ART in the French EPF cohort.⁵⁹ One hypothesis could be that preterm delivery was associated with lopinavir, which was widely used in the previous studies,⁶¹ rather than a class effect of PIs or a general effect⁶² of therapy, or possibly with other time-related changes in management.

Strengths and weaknesses

This is the first study to date to evaluate a regimen of PT prevention without any exposure to NRTIs during pregnancy, delivery or the neonatal period. The other strengths are the prospective multicentre design and high standards of follow-up and monitoring.

An important limitation of the study was the high-resource setting and the highly selected population, which may not reflect real-world programmes, thus reducing the external validity in other settings,⁶³ especially in the case of breastfeeding.⁶⁴ Also, we had to limit the number of patients to enrol, leading to loss of power. We chose an open-label one-arm interventional design, since a randomized equivalence trial would require a much larger enrolment and was not feasible.

Perspectives

Current guidelines do not recommend PI monotherapy in non-pregnant individuals. The rationale for NRTI-free strategies is to reduce the risk of toxicities in the special context of pregnancy, where it may be preferable to limit exposure of the fetus to multiple medications. Such decisions should be based on consideration of the benefits and risks according to each patient’s individual virological history and personal preferences and priorities, in a framework of shared decision-making. The place for darunavir/ritonavir monotherapy would be in limited circumstances, such as pregnant HIV-1 or HIV-2 controllers having undetectable plasma HIV RNA without ART, or patients with a strong preference to not use NRTIs in pregnancy who can be expected from our findings to have an acceptably low risk of virological failure on darunavir/ritonavir monotherapy. This pilot study paves the way for the evaluation of other simplified regimens during pregnancy, which could offer optimal viral control while reducing maternal and fetal NRTI exposure.

Acknowledgements

Members of the Monogest study group

Research coordinators: Sandrine DELMAS, Catherine CAPITANT, Isabelle FOURNIER (Inserm SC10-US19), Juliette SAILLARD, Lucie MARCHAND (ANRS/MIE). Director of Inserm SC10: Laurence MEYER. Data Safety Monitoring Board: Elie AZRIA, Elisabeth ROUVEIX, Tessa GOETGHEBUER, Francis BARIN, Valériane LEROY. Scientific Committee: Cédric ARVIEUX, Stéphane BLANCHE, Marc DOMMERGUES, Pierre FRANGE, Corinne GUERIN, Lucie MARCHAND, Gilles PEYTAVIN, Hélène POLLARD, Jean-Christophe PLANTIER, Véronique AVETTAND FENOEL, Roland TUBIANA. ANRS/MIE Clinical Trial Safety Department: Alpha DIALLO, Christelle PAUL. Participating Centres: Hôpital Louis Mourier, Colombes (Laurent MANDELBROT, Jeanne SIBIUDE, Françoise MEIER, Corinne FLOCH), Hôpital Pitié-Salpêtrière, Paris (Roland TUBIANA, Marc DOMMERGUES, Christine BLANC), Hôpital Tenon, Paris (Ana CANESTRI, Lise SELLERET), Hôpital Saint André, CHU Bordeaux, Bordeaux (Philippe MORLAT, Mojgan HESSAMFAR, Sabrina CALDATO), Hôpital Bicêtre, Le Kremlin-Bicêtre (Delphine PERETTI, Marie HOULLIER, Corinne FOURCADE), Hôpital Antoine Béclère, Clamart (Véronique CHAMBRIN, Alexandra LETOURNEAU, Laure CLECH), Centre Hospitalier du Sud Francilien, Evry (Amélie CHABROL, Bérengère CANON, Michèle GRANIER), Hôpital Trousseau, Paris (Eida BUI, Catherine DOLLFUS), CHU de Toulouse, Toulouse (Caroline SIMON-TOULZA, Edith BRAZET, Julie ABBAL), Hôpital Bichat, Paris (Sophie MATHERON, Agnès BOURGEOIS MOINE), CHU Rennes (Cédric ARVIEUX), Hôpital Cochin, Paris (Emmanuelle PANNIER, Valérie MARCOU), Centre Hospitalier de Perpignan (Marie MEDUS, Ilona WAJSZCZAK, Germaine BACHELARD), Centre Hospitalier Victor Dupouy, Argenteuil (Philippe GENET, Nathalie TORDJEMAN, Dominique BRAULT), Hôpital Lariboisière, Paris (Philippe GENET, Agathe RAMI, Barbara MARAUX), Hôpital Jean-Verdier, Bondy (Julie BOTTERO, Amélie BENBARA, Eric LACHASSINE), Groupe Hospitalier L’Archet, Nice (Eric ROSENTHAL, Alissa NAQVI, André BONGAIN, Fabrice MONPOUX), Hôpital de la Croix Rousse—GH Nord, Lyon (Laurent COTTE, Cyril HUISSOUD, Jean-Marc LABAUNE), Hôpital Pellegrin, CHU de Bordeaux, Bordeaux (Michel DUPON, Denis ROUX, Christophe ELLEAU), Hôpital Necker Enfants Malades, Paris (Claudine DUVIVIER, Marine DRIESSEN, Pierre FRANGE, Véronique AVETTAND-FENOEL, Elise GARDIENNET), Hôpital Européen Georges-Pompidou, Paris (Marina KARMOCHKINE), Hôpital Foch, Suresnes (David ZUCMAN, Juliette LAPERRELLE), CHU de Nantes (Cécile BRUNET-CARTIER, Norbert WINER, Véronique RELIQUET). Research assistants: Marina MAMBERT, Axel LEVIER, Hayette LADJAL, Cléa HOUEL, Barbara LEBAS, Manal RAHMOUN, Véronique ELIETTE, Martine RESCH, Elise GARDIENNET. Statistician: Jérôme LE CHENADEC. Data manager: Elisa AREZES.

The preliminary results of the trial were presented at the Conference on Retroviruses and Opportunistic Infections (CROI), Virtual CROI 2021 (Abstract ID 570).

Funding

This work was supported by the Agence Nationale de Recherches sur le Sida et les Hépatites Virales/Maladies Infectieuses Emergentes for research on AIDS, viral hepatitis and Emerging infectious diseases (Study number ANRS-168).

Transparency declarations

P.F. received research grants (to his institution) from the French National Agency for AIDS Research (ANRS), consultancy fees from MSD France and honoraria and travel grants from ViiV Healthcare, Bristol-Myers Squibb, Janssen Cilag, Gilead Sciences and MSD France for participation in educational programmes and conferences. G.P. has received travel grants, consultancy fees, honoraria, or study grants from various pharmaceutical companies, including Gilead Sciences, Merck, Pfizer, Takeda, Theratechnologies and ViiV Healthcare. V.A.F. received research grants (to her institution) from the French National Agency for AIDS Research (ANRS), and honoraria and travel grants from ViiV Healthcare and Gilead Sciences for participation in educational programmes and conferences. The other authors report no conflicts of interest regarding this work.

Author contributions

Laurent Mandelbrot: conceptualization, investigation, writing, funding acquisition; Roland Tubiana: conceptualization, investigation, writing; Pierre Frange: investigation, writing; Gilles Peytavin: investigation, analysis, writing; Jerome Le Chenadec: data management, formal analysis; Ana Canestri: investigation; Philippe Morlat: investigation, editing; Cécile Brunet-Cartier: investigation; Jeanne Sibiude: investigation, writing, analysis; Delphine Peretti: investigation; Véronique Chambrin: investigation; Amélie Chabrol: investigation; Eida Bui: investigation; Caroline Simon-Toulza: investigation; Lucie Marchand: supervision; Christelle Paul: supervision; Sandrine Delmas: project administration; Véronique Avettand-Fenoel: investigation, analysis; Josiane Warszawski: methodology, validation, analysis.

References

Author notes