Background and objective(s) Clinical trials and cohorts are powerful epidemiological tools to study changes in health outcomes and biological markers over time. To ensure the validity of these studies, measurement tools must provide consistent and reproducible results over the follow-up period, allowing observed changes to be attributed to the outcomes of interest. Ideally, biomarker assessments would be conducted under standardized conditions—using the same devices, assays, and operators—within a short timeframe. However, meeting these conditions is challenging in long time and/or large-scale studies, due to inherent constraints like human and financial resources, assay availability, and logistical factors. Additionally, clinical studies may rely on Research Use Only (RUO) assays, which are not marketed for daily practice and are more likely to be impacted by variations in assay lots or updates in measurement methods (e.g. reagents, antibodies). In this study, we leveraged data from the MEMENTO cohort to develop a statistical framework addressing (i) the change in the assay lot between the different analysis time points (lot-to-lot validation, LTLV) and (ii) the change in the method of measure (bridging). Material and Methods The MEMENTO cohort enrolled 2323 individuals at risk of Alzheimer's disease (AD) across 26 French Memory Clinic centers. Blood samples were collected at enrolment (M0) and then on average at 24 (M24) and 48 months (M48) of follow-up and stored in a centralized biobank. AD blood biomarkers, including amyloid-β40 (Aβ40), Aβ42, 181-phosphorylated Tau (p-tau181), and Neurofilament light chain (NfL), were centrally measured at the Plateforme Analytique de Recherche en Santé (PARS) of the Bordeaux University Hospital using RUO Quanterix® immunoassay kits. M0 samples were analyzed in 2021, while M24 and M48 samples were measured in 2023 using updated versions of the immunoassay kits. To provide a bridging equation, a random subset of M0 samples was reanalyzed using the same kits as for M24/M48 but from different lots. To address changes in measurement methods, bridging formulas for M0 samples were derived using five statistical regression models: linear, quadratic, weighted Deming, Passing-Bablok, and a four-parameter logistic curve. Models were compared based on their R-squared (R²) and Root Mean Squared Error (RMSE), estimated via leave-one-out cross-validation. The LTLV was assessed using Passing-Bablok regression models. Results Fifty M0 samples were randomly selected to span the full range of biomarker values observed at M24 and M48. Conversion equations performed differently depending on the biomarker, with R² values ranging from 0.89 for NfL to 0.36 for Aβ42. Prediction errors were comparable across linear, weighted Deming, and Passing-Bablok regression models. For LTLV, the best performances were observed for NfL (R²=0.97) and Aβ42 (R²=0.91). However, for all four biomarkers, the regression slope was significantly greater than 1, indicating systematic differences between assay lots. Conclusion Our study highlights the critical challenges posed by assay lot variability and method changes when studying fluid biomarker trajectories in long term longitudinal studies. These findings demonstrate the feasibility of addressing these issues using statistical frameworks and underscore the importance of adopting appropriate methods to ensure reliable data interpretation in long-term clinical studies.

This article offers a well-referenced expert opinion on the need to strengthen inclusive and ethical data systems as a foundation for equitable public health leadership. Drawing on recent experiences from COVID-19, maternal health, and climate-related health risks, the authors argue for concrete actions that public health professionals and institutions can take to make inequities more visible and address them effectively.

BackgroundThe association between the pattern of cortical thickness (CT) and executive dysfunction (ED) in mild cognitive impairment (MCI) and subjective cognitive complaints (SCC) is still poorly understood. We aimed to investigate the association between CT and ED in a large French cohort (MEMENTO) of 2323 participants with MCI or SCC.MethodsAll participants with available CT and executive function data (verbal fluency and Trail Making Test [TMT]) were selected (n=1924). Linear regressions were performed to determine relationships between executive performance and the brain parenchymal fraction (BPF) and CT using FreeSurfer.ResultsThe global executive function score was related to the BPF (sß: 0.091, P<0.001) and CT in the right supramarginal (sß: 0.060, P=0.041) and right isthmus cingulate (sß: 0.062, P=0.011) regions. Literal verbal fluency was related to the BPF (sß: 0.125, P<0.001) and CT in the left parsorbitalis region (sß: 0.045, P=0.045). Semantic verbal fluency was related to the BPF (sß: 0.101, P<0.001) and CT in the right supramarginal region (sß: 0.061, P=0.042). The time difference between the TMT parts B and A was related to the BPF (sß: 0.048, P=0.045) and CT in the right precuneus (sß: 0.073, P=0.019) and right isthmus cingulate region (sß: 0.054, P=0.032).ConclusionsIn a large clinically based cohort of participants presenting with either MCI or SCC (a potential early stage of Alzheimer's disease [AD]), ED was related to the BPF and CT in the left pars orbitalis, right precuneus, right supramarginal, and right isthmus cingulate regions. This pattern of lesions adds knowledge to the conventional anatomy of ED and could contribute to the early diagnosis of AD.

Introduction: There are few data on the frequency of virological remission in African individuals after treatment with antiretroviral therapy (ART) in primary HIV infection (PHI). Methods: We studied participants (n = 82) from South Africa and Uganda in Short Pulse Antiretroviral Treatment at HIV-1 Seroconversion, the first trial of treatment interruption in African individuals with PHI randomized to deferred ART or 48 weeks of immediate ART. All were female and infected with non-B HIV subtypes, mainly C. We measured HIV DNA in CD4+ T cells, CD4+ cell count, plasma viral load (pVL), cell-associated HIV RNA and T-cell activation and exhaustion. We explored associations with clinical progression and time to pVL rebound after treatment interruption (n = 22). Data were compared with non-African Short Pulse Antiretroviral Treatment at HIV-1 Seroconversion participants. Results: Pretherapy pVL and integrated HIV DNA were lower in Africans compared with non-Africans (median 4.16 vs. 4.72 log10 copies/ml and 3.07 vs. 3.61 log10 copies/million CD4+ T cells, respectively; P < 0.001). Pre-ART HIV DNA in Africans was associated with clinical progression (P = 0.001, HR per log10 copies/million CD4+ T cells increase (95% CI) 5.38 (1.95–14.79)) and time to pVL rebound (P = 0.034, HR per log10 copies/ml increase 4.33 (1.12–16.84)). After treatment interruption, Africans experienced longer duration of viral remission than non-Africans (P < 0.001; HR 3.90 (1.75–8.71). Five of 22 African participants (22.7%) maintained VL less than 400 copies/ml over a median of 188 weeks following treatment interruption. Conclusion: We find evidence of greater probability of virological remission following treatment interruption among African participants, although we are unable to differentiate between sex, ethnicity and viral subtype. The finding warrants further investigation.

Les événements stressants de vie pourraient contribuer aux plaintes et déficits au stade de déficit cognitif léger (mild cognitive impairment), et à une progression plus rapide vers la démence de type Alzheimer (MA). L'objectif de ce travail est de préciser l'impact du stress chez des patients non-déments consultant un centre mémoire et d'étudier son substrat en TEP cérébrale au 18FDG. Cette étude s'appuie sur la cohorte MEMENTO, une cohorte nationale clinique de 2323 participants. Nous avons identifié à l'inclusion 512 sujets présentant une plainte cognitive subjective ou un déficit cognitif léger, pour lesquels les variables suivantes étaient disponibles : âge, sexe, niveau éducatif, statut tabagique, statut TEP amyloïde, statut APO-E, CDR, MMSE, scores NPI, TEP au 18FDG et échelle de stress. Cette dernière était cotée entre 0 et 10 (stress maximal) par les patients en fonction de la gêne ressentie. Des analyses statistiques ont été conduites pour établir le lien entre ces variables. Une analyse SPM12 de corrélation TEP au 18FDG voxel-à-voxel sur cerveau entier a également été réalisée selon le niveau de stress rapporté (en tenant compte des co-variables suivantes : âge, sexe, niveau éducatif, statut amyloïde, CDR, MMSE). Le score moyen à l'échelle de stress sur les 512 sujets de l'échantillon analytique était de 3,5 (écart-type) 2,6 (min 0–max 10). Ce score n'était pas modifié selon les statuts APO-E ɛ4, TEP amyloïde ou tabagique (ANOVA p > 0,25). Les personnes dont les scores de stress étaient les plus élevés étaient plus fréquemment des femmes, jeunes, moins diplômées, avec une plus grande sévérité aux scores CDR, MMSE et NPI (ANOVA/Pearson, p < 0,04). Après prise en compte des facteurs de confusion, la sévérité du stress exprimé était associée à un hypermétabolisme de la région amygdalienne droite et à un hypométabolisme temporal inférieur/latéral droit (p < 0,001, k > 180). Le stress associé aux événements de vie chez des patients non-déments consultant en centre mémoire est associé à une plus grande sévérité clinique, indépendante du statut APO-E ɛ4 et amyloïde, avec une dysfonction métabolique du lobe temporal droit en TEP au 18FDG. Des analyses complémentaires seront nécessaires pour préciser si ce profil est associé à une réorganisation différentielle des réseaux cérébraux associés à la MA avec un possible impact sur la conversion vers une démence.

was considered as normal. For patients and controls comfort, BT was systematically stopped after 20 min. Plasmatic vWF antigen was assayed by electroimmunodiffusion ac­ cording to Laurell. Plasmatic vWF activity was measured using fresh washed platelets. As the two assays gave strongly correlated results, only the vWF activity expressed as international units is reported. As expected, the vWF values in T-HIV+ were significantly higher than in controls (Table 1; p = 0.0004), on the contrary, the BT in T-HIV+ was significantly shorter than in T-HIV(p = 0.002). Whereas, an inverse relationship between bleeding time and platelet count was associated with broad statistical scatter in different publications (7), in our study, the linear regression analysis showed in T-HIV+ a strong inverse asso­ ciation between BT and platelet count (Fig. 1 A; p = 0.02). In Fig. IB, BT values are obviously lower for the highest plasmatic von Willebrand factor levels, and the linear regression analysis indi­ cates that the BT values are inversely associated with the plasmatic von Willebrand factor levels adjusted for platelet counts within the HIVinfected group (p = 0.01). It has been previously shown that BT is related to the platelet vWF (8), it appeared in this study, that in throm­ bocytopenic HIV-infected patients, BT is strongly inversely associated with the plasmatic vWF values. This moderating effect of high concen­ trations of plasmatic vWF on BT may take part in the slight bleeding tendency in such patients.

high-throughput sequencing to observe how SAMHD1 expression alters the mutational profile (frequency and spectra) of integrated proviruses.We will explore how mutation rates of HIV-2 can be manipulated through the use of nucleoside analogs and RNRI drugs to explore what effects these compounds have on the HIV-2 mutation profile.Using single-cycle infectivity assays as well as long-term spreading experiments, we will be able to correlate mutagenesis with viral evolution and infectivity data to explore how sensitive these two viruses are to changes in viral mutation.This work will serve to understand how HIV-2 operates at a lower mutation frequency than HIV-1, elucidate the relationship between mutagenesis and infectivity for the two viruses, and provide insights into the contrasting phenotypes observed between the viruses.

Background: We evaluate differences in timing of cART (combined antiretroviral treatment) initiation by geographical origin in male and female HIV-positive patients in the Collaboration of Observational HIV Epidemiological Research Europe, a large European Collaboration of HIV Cohorts.Methods: We included individuals recruited in Western Europe between January 1997 and March 2013, with known geographical origin and at least 1 CD4(+) cell countmeasurement while cART-naive. Timing of cART was assessed through modified time-to-eventmethods, in which a scale of CD4(+) cell counts was used instead of time, with cART being the outcome. We estimated the median CD4(+) cell count at cART initiation (estimated CD4(+) levels at which the probability of having started cART is 50%) using Kaplan-Meier and adjusted hazard ratios of cART initiation using Cox regression.Results: Of 151 674 individuals, 110 592 (72.9%) were men. Median (95% confidence interval) CD4(+) cell count falls far below 250 cells/ml in all groups and was lowest in sub-Saharan African [SSA: 161 (158-167)], Caribbean men [161 (150-174)] and in Asian women [Asian Continent and Oceania: 185 (165-197)]. Among men, the adjusted probability of cART initiation was lower in migrants compared with natives, but differences depended on initial CD4(+) cell count. For example, in the group with more than 500 CD4(+) at recruitment, they were 45% (36-53%), 30% (17-40%) and 25% (19-30%) lower for Caribbean, Eastern European and SSA men, respectively. In women, no meaningful differences were observed between natives and most migrant groups. However, SSA women had a 31% (24-38%) higher probability of cART initiation when recruited at a CD4(+) more than 500 cells/ml and 9% (4-14%) lower when recruited at CD4(+) less than 100 cells/ml.Conclusion: Most migrant men initiate cART at lower CD4(+) cell count than natives, whereas this does not hold for migrant women.

La randomisation d’un essai clinique en double insu doit être opérationnalisée en tenant compte des contraintes logistiques pour maintenir l’impossibilité de discerner l’allocation pour le prochain participant. Dans un essai vaccinal contre Ebola en Guinée, Liberia et Sierra Leone, l’accès permanent aux solutions techniques traditionnelles pour centraliser le processus d’allocation (Internet, répondeur vocal) ne peut être garanti. Deux procédures de randomisation ont été considérées, et l’impact sur une sous-étude a été évalué. PREVAC est un essai de phase II, multicentrique, randomisé, en double-insu, évaluant trois stratégies vaccinales versus placebo (cinq bras) selon un ratio 2:1:2:1:1. Quatre sites doivent inclure 4900 participants. La randomisation est stratifiée sur le site. Avant le début de l’essai, des rouleaux d’étiquettes de codes-barres détachables avec les n° d’identification des seringues (SID) correspondant aux bras sont préparés pour chaque site. Dans les deux procédures, la randomisation a lieu au moment de la vaccination du participant quand le SID est associé à l’identifiant du participant (PID) sur le cahier d’observation. Les vaccins sont reconstitués sur site le jour de la vaccination par un pharmacien sans insu et doivent être utilisée dans l’heure qui suit. Une première méthode, similaire à une méthode utilisée dans un précédent essai vaccinal contre Ebola au Liberia (PREVAIL I), implique la préparation quotidienne de plusieurs sacs contenant chacun sept seringues étiquetées avec les SID. Cinq rouleaux d’étiquettes de SID, un pour chaque bras, permettent aux pharmaciens de préparer sept seringues qui sont placées dans un sac opaque et tirées au sort lors de la vaccination. Une seconde méthode implique l’utilisation d’une application disponible sur un ordinateur en local qui permet aux pharmaciens d’identifier le bras alloué à chaque participant. Avec cette approche, un rouleau unique d’étiquettes de SID est préparé et lorsque qu’un participant se présente pour la vaccination, le pharmacien détache l’étiquette de SID, scanne le code-barres et l’application (indépendante d’Internet) indique le vaccin à préparer. La sous-étude doit inclure 196 participants sur un site. Sans stratification de la randomisation sur la participation à la sous-étude qui permettait d’assurer un ratio des bras identique entre l’étude principale et la sous-étude, nous avons procédé à des simulations afin d’estimer les risques de déséquilibres entre les bras pour les deux procédures en testant plusieurs hypothèses (consentement et prévalence du VIH). Les deux procédures de randomisation satisfont l’impossibilité de discerner à l’avance l’allocation et le maintien ensuite de l’insu. La première méthode permet une préparation anticipée des seringues et facilite le flux des participants mais comporte un risque de déséquilibre si un sac n’est pas complètement utilisé. La seconde méthode permet une flexibilité du nombre de participants à inclure chaque jour et respecte le délai de stérilité des vaccins, mais peut ralentir le flux des participants. La seconde méthode a été adoptée dans l’essai PREVAC. Dans la sous-étude, les simulations montrent un risque de déséquilibre quasiment nul pour les deux procédures. Cet exemple montre les procédures opératoires à mettre en œuvre pour respecter la randomisation dans un essai de taille importante et leur complexité en présence de contraintes d’utilisation des produits expérimentaux et d’études ancillaires.

OBJECTIVES:Single nucleotide polymorphisms in the cytochrome P450 (CYP) 2B6 gene have been associated with high interindividual variation in efavirenz pharmacokinetics. However, clinical data on the relationship of CYP2B6 polymorphisms with the occurrence of efavirenz-induced central nervous system (CNS) symptoms are limited.METHODS:We analysed four polymorphisms in the CYP2B6 (516 G>T), CYP3A5 (6986 A>G) and ATP-binding cassette, sub-family B, member 1 (ABCB1) (2677 G>T/A and 3435 C>T) genes in HIV-infected adults virologically suppressed on a protease inhibitor-based regimen who switched to a regimen containing emtricitabine, didanosine and efavirenz in the setting of the ANRS ALIZE trial. Kaplan-Meier methods and Cox regression analysis were used to investigate their association with efavirenz plasma levels and CNS events up to 48 months after switching.RESULTS:In total, 191 patients with a median age of 41 years, who were 87% male and 85% Caucasian, were enrolled in the study. Variant allelic frequencies were 0.49, 0.93, 0.59 and 0.63 for CYP2B6 516, CYP3A5 392, ABCB1 2677 and ABCB1 3435, respectively. The median efavirenz plasma concentration (MEPC) was 2.2 mg/L [interquartile range (IQR) 1.7-2.8 mg/L] and was significantly higher in patients with the deficient CYP2B6 516T. Overall, 242 CNS events were reported in 104 individuals (54%). No correlation was found between MEPC and CNS events. The occurrence of a first CNS event was lower in patients with the CYP2B6 516 G/G genotype vs. CYP2B6 516 T genotypes [50% (IQR: 40-60%) vs. 66% (IQR: 56-75%), respectively; P = 0.02]. In an adjusted Cox regression model, there was a tendency towards a higher risk of a first CNS event among carriers of the variant CYP2B6 516 T allele (relative risk 1.4 [95% CI, 0.99-2.1]; P?=?.06), compared with noncarriers.CONCLUSIONS:The deficient CYP2B6 516 T allele is associated with higher efavirenz plasma drug levels and more frequent CNS-related symptoms.

Background: Socioeconomic inequality challenges population-level implementation of health interventions. We investigated differences by educational level in clinical, virological, and immunological responses to combined antiretroviral treatment (cART) in HIV-positive men and women in Collaboration of Observational HIV Epidemiological Research in Europe, a European collaboration.Methods: Data were pooled from 15 cohorts in eight countries of patients initiating cART in 1996-2013 with data on educational level categorized in UNESCO/ISCED classifications. Kaplan-Meier curves, Cox and piecewise linear mixed models were used.Results: Of 24 069 HIV-positive patients, 9% had not completed primary education, 32% had completed primary, 44% secondary, and 15% tertiary education. Overall, 21% were women, who were overrepresented in lower educational strata. During 132 507 person-years of follow-up, 1081 individuals died; cumulative mortality decreased with higher educational level (P< 0.001). Over 122 765 person-years, new AIDS events or death occurred in 2598 individuals; differences by education were more marked than for death alone (P< 0.001). Virological response was achieved by 67% of patients without completed basic education, 85% with completed primary education, 82% with secondary, and 87% with tertiary (P< 0.001). Patients with higher education had higher CD4(+) cell count at cART initiation and at each time after cART but rate of CD4(+) cell count recovery did not differ. Differences in mortality and clinical responses were similar for men and women and were not entirely explained by delayed HIV diagnosis and late cART initiation.Conclusion: HIV-positive patients with lower educational level had worse responses to cART and survival in European countries with universal healthcare. To maximize the population impact of cART, Europe needs to decrease the socioeconomic divide. Copyright (C) 2017 Wolters Kluwer Health, Inc. All rights reserved.

Summary Objectives: To present the European landscape regarding the re-use of health administrative data for research. Methods: We present some collaborative projects and solutions that have been developed by Nordic countries, Italy, Spain, France, Germany, and the UK, to facilitate access to their health data for research purposes. Results: Research in public health is transitioning from siloed systems to more accessible and re-usable data resources. Following the example of the Nordic countries, several European countries aim at facilitating the re-use of their health administrative databases for research purposes. However, the ecosystem is still a complex patchwork, with different rules, policies, and processes for data provision. Conclusion: The challenges are such that with the abundance of health administrative data, only a European, overarching public health research infrastructure, is able to efficiently facilitate access to this data and accelerate research based on these highly valuable resources.

L'expertise par les pairs est actuellement le mode d'évaluation des projets de recherche soumis à des appels à projets institutionnels. L'objectif de ce travail est de décrire le processus d'expertise des projets de recherche institutionnels et de dégager des propositions de recommandations afin d'améliorer le processus d'expertise académique par les pairs dans le contexte français. Une recherche de la littérature a été réalisée en juillet 2015 en interrogeant deux bases de données (Pubmed et Web of science) à partir des mots clés : peer review ou grant dans le titre et le résumé sans restriction de date. D'autres sites d'agences de recherche comme l'Agence nationale ou le Conseil européen de la recherche ont été visités. Les principaux items explorés étaient : la gestion des conflits d'intérêt, le contenu d'une grille d'expertise, l'organisation générale de l'expertise d'un appel d'offre et la valorisation de l'expertise. Parmi les 1940 articles sélectionnés, 459 décrivaient le processus d'expertise de projets ou d'articles de recherche. Seuls 66 portant sur l'expertise des projets de recherche ont été analysés, parmi lesquels 71 % (47/66) ont été publiés entre 2010 et 2015 : 39 % (26/66) étaient des études observationnelles, 17 % (11/66) des retours d'expérience sur l'expertise des projets de recherche, 12 % (8/66) des opinions, 9 % (6/66) des études interventionnelles, 7 % (5/66) des discussions, 7 % (5/66) des commentaires, 5 % (3/66) des éditoriaux, 2 % (1/66) un rapport et 2 % (1/66) une méta-analyse. Parmi les 66 articles : 71 % (47/66) portaient sur l'organisation générale de l'expertise, 14 % (9/66) concernaient la valorisation de la recherche, 11 % (7/66) portaient sur les conflits d'intérêts et 4 % (3/66) concernaient la valorisation de l'expertise. Trente-huit propositions de recommandations concernant l'organisation générale de l'expertise ont été identifiées. Elles portaient sur la sélection et le recrutement des experts, l'anonymat, la formation des experts et le processus de l'expertise (réunions, notation, etc.). Parmi ces propositions, 8 concernaient la gestion des conflits d'intérêt (déclaration, indépendance, etc.), 27 la grille d'expertise et son contenu (originalité du projet, intérêt scientifique, méthodologie, faisabilité, budget, etc.) et 5 la valorisation de l'expertise (financière, intellectuelle). Cette revue nous permet de dégager des propositions de recommandations pouvant servir lors du processus d'expertise des projets de recherche institutionnels. Afin d'identifier les items les plus pertinents, une enquête Delphi auprès des différents acteurs de la recherche clinique sera menée. Cette enquête permettra d'établir une liste de recommandations consensuelles qui seront ultérieurement validées selon la méthodologie de consensus formalisé.

Linking health and administrative data for maternal, child and young adult health. 9. European Public Health Conference

V Bouteloup, C Sabin, A Mocroft, L Gras, N Pantazis, V Le Moing, A d’Arminio Monforte, M Mary-Krause, B Roca, JM Miro, M Battegay, N Brockmeyer, J Berenguer, P Morlat, N Obel, S De Wit, G F€atkenheuer, R Zangerle, J Ghosn, S P erez-Hoyos, M Campbell, M Prins, G Chêne, L Meyer, M Dorrucci, C Torti and R Thi ebaut The Standard Reference Distribution of CD4 Response to HAART Project Team for the Collaboration of Observational HIV Epidemiological Research Europe (COHERE) in EuroCoord* CIC 1401, CHU de Bordeaux, Bordeaux, France, INSERM U1219 – Centre Inserm Bordeaux Population Health, Universit e de Bordeaux, Bordeaux, France, ISPED, Centre INSERM U1219-Bordeaux Population Health, Universit e de Bordeaux, Bordeaux, France, Research Department of Infection & Population Health, UCL, London, UK, Stichting HIV Monitoring, Amsterdam, The Netherlands, Department of Hygiene, Epidemiology & Medical Statistics, Athens University Medical School, Athens, Greece, Montpellier University, Montpellier, France, Infectious Diseases Unit, Department of Health Sciences, San Paolo University Hospital, Milan, Italy, INSERM, Institut Pierre Louis d’ epid emiologie et de Sant e Publique (IPLESP UMRS 1136), UPMC Univ Paris 06, Sorbonne Universit es, F-75013, Paris, France, Hospital General of Castellon, Castell on, Spain, Infectious Diseases Service. Hospital Clinic – IDIBAPS, University of Barcelona, Barcelona, Spain, Division of Infectious Diseases and Hospital Epidemiology, Department of Clinical Research, University Hospital of Basel, Basel, Switzerland, Department of Dermatology, Venerology – Center for Sexual Health and Medicine, Ruhr-Universit€at Bochum, Bochum, Germany, Instituto de Investigaci on Sanitaria Gregorio Mara~ n on (IiSGM), Hospital General Universitario Gregorio Mara~ n on, Madrid, Spain, Service de M edecine Interne et Maladies Infectieuses, Hôpital Saint-Andr e, Bordeaux, France, Department of Infectious Diseases, Copenhagen University Hospital, Copenhagen, Denmark, Department of Infectious Diseases, St Pierre University Hospital, Universit e Libre de Bruxelles, Brussels, Belgium, Department of Internal Medicine, University of Cologne and German Centre for Infection Research (DZIF), Cologne, Germany, Medical University Innsbruck, Innsbruck, Austria, APHP, Unit e Fonctionnelle de Th erapeutique en Immuno-Infectiologie, Centre Hospitalier Universitaire Hôtel Dieu, Paris, France, Facult e de M edecine Site Necker, Sorbonne Paris Cit e, Universit e Paris Descartes, EA 7327, Paris, France, Vall d’Hebr on Institut de Recerca (VHIR), Barcelona, Spain, Universitat Aut onoma de Barcelona, Barcelona, Spain, CHIP, Department of Infectious Diseases, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark, Division of Infectious Diseases, Department of Internal Medicine, Center for Infection and Immunity Amsterdam (CINIMA), Academic Medical Center, Amsterdam, The Netherlands, Department of Infectious Diseases, Public Health Service, Amsterdam, The Netherlands, CHU de Bordeaux, Pole de Sante Publique, Service d’Information Medicale, F-33000, Bordeaux, France, INSERM, U1018, Epidemiology of HIV, Reproduction, Paediatrics, CESP; University Paris-Sud, Paris, France, Department of Public Health and Epidemiology, Bicêtre Hospital, AP-HP, Le Kremlin Bicêtre, Paris, France, Department of Infectious, Parasitic and Immune-mediated Diseases, Istituto Superiore di Sanit a, Rome, Italy and Unit of Infectious and Tropical Diseases, Department of Medical and Surgical Sciences, University “Magna Graecia”, Catanzaro, Italy

ObjectivesThe aim of the study was to determine the time to, and risk factors for, triple‐class virological failure (TCVF) across age groups for children and adolescents with perinatally acquired HIV infection and older adolescents and adults with heterosexually acquired HIV infection.MethodsWe analysed individual patient data from cohorts in the Collaboration of Observational HIV Epidemiological Research Europe (COHERE). A total of 5972 participants starting antiretroviral therapy (ART) from 1998, aged < 20 years at the start of ART for those with perinatal infection and 15–29 years for those with heterosexual infection, with ART containing at least two nucleoside reverse transcriptase inhibitors (NRTIs) and a nonnucleoside reverse transcriptase inhibitor (NNRTI) or a boosted protease inhibitor (bPI), were followed from ART initiation until the most recent viral load (VL) measurement. Virological failure of a drug was defined as VL > 500 HIV‐1 RNA copies/mL despite ≥ 4 months of use. TCVF was defined as cumulative failure of two NRTIs, an NNRTI and a bPI.ResultsThe median number of weeks between diagnosis and the start of ART was higher in participants with perinatal HIV infection compared with participants with heterosexually acquired HIV infection overall [17 (interquartile range (IQR) 4–111) vs. 8 (IQR 2–38) weeks, respectively], and highest in perinatally infected participants aged 10–14 years [49 (IQR 9–267) weeks]. The cumulative proportion with TCVF 5 years after starting ART was 9.6% [95% confidence interval (CI) 7.0−12.3%] in participants with perinatally acquired infection and 4.7% (95% CI 3.9−5.5%) in participants with heterosexually acquired infection, and highest in perinatally infected participants aged 10–14 years when starting ART (27.7%; 95% CI 13.2−42.1%). Across all participants, significant predictors of TCVF were those with perinatal HIV aged 10–14 years, African origin, pre‐ART AIDS, NNRTI‐based initial regimens, higher pre‐ART viral load and lower pre‐ART CD4.ConclusionsThe results suggest a beneficial effect of starting ART before adolescence, and starting young people on boosted PIs, to maximize treatment response during this transitional stage of development.

ObjectivesThe aim of this work was to provide a reference for the CD4 T‐cell count response in the early months after the initiation of combination antiretroviral therapy (cART) in HIV‐1‐infected patients.MethodsAll patients in the Collaboration of Observational HIV Epidemiological Research Europe (COHERE) cohort who were aged ≥ 18 years and started cART for the first time between 1 January 2005 and 1 January 2010 and who had at least one available measurement of CD4 count and a viral load ≤ 50 HIV‐1 RNA copies/mL at 6 months (± 3 months) after cART initiation were included in the study. Unadjusted and adjusted references curves and predictions were obtained using quantile regressions.ResultsA total of 28 992 patients were included in the study. The median CD4 T‐cell count at treatment initiation was 249 [interquartile range (IQR) 150, 336] cells/μL. The median observed CD4 counts at 6, 9 and 12 months were 382 (IQR 256, 515), 402 (IQR 274, 543) and 420 (IQR 293, 565) cells/μL. The two main factors explaining the variation of CD4 count at 6 months were AIDS stage and CD4 count at cART initiation. A CD4 count increase of ≥ 100 cells/mL is generally required in order that patients stay ‘on track’ (i.e. with a CD4 count at the same percentile as when they started), with slightly higher gains required for those starting with CD4 counts in the higher percentiles. Individual predictions adjusted for factors influencing CD4 count were more precise.ConclusionsReference curves aid the evaluation of the immune response early after antiretroviral therapy initiation that leads to viral control.