Search

Background: In relapsing-remitting multiple sclerosis (RRMS), early identification of suboptimal responders can prevent disability progression. Objective: We aimed to develop and validate a dynamic score to guide the early decision to switch from first- to second-line therapy. Methods: Using time-dependent propensity scores (PS) from a French cohort of 12,823 patients with RRMS, we constructed one training and two validation PS-matched cohorts to compare the switched patients to second-line treatment and the maintained patients. We used a frailty Cox model for predicting individual hazard ratios (iHRs). Results: From the validation PS-matched cohort of 348 independent patients with iHR ⩽ 0.69, we reported the 5-year relapse-free survival at 0.14 (95% confidence interval (CI) 0.09–0.22) for the waiting group and 0.40 (95% CI 0.32–0.51) for the switched group. From the validation PS-matched cohort of 518 independent patients with iHR > 0.69, these values were 0.37 (95% CI 0.30–0.46) and 0.44 (95% CI 0.37–0.52), respectively. Conclusions: By using the proposed dynamic score, we estimated that at least one-third of patients could benefit from an earlier switch to prevent relapse.

Rationale & Objective Deceased donor acute kidney injury (AKI) frequently leads to kidney discards, but its impact on long-term graft survival in kidney transplant recipients remains unclear. We investigated the association between deceased donor AKI assessed using back-estimation of baseline serum creatinine (Scr) and graft survival. Study Design Observational cohort study. Setting & Participants Adult patients represented within the French CRISTAL registry who received a single kidney allograft from brain-dead deceased donors between January 2006 and December 2017. Exposure A back-estimated Scr baseline value was derived for an assumed glomerular filtration rate at 75 mL/min/1.73 m2, using the MDRD Study equation. A refined classification system for donor AKI was implemented as follows: no AKI, undetermined AKI/chronic kidney disease (CKD), recovery from AKI, and ongoing AKI. Outcome Death-censored graft survival. Analytical Approach Multivariable Cox models using a robust variance estimator for paired kidneys from the same donor. Results We classified 26,786 recipients as follows: no AKI (n = 19,276); undetermined AKI/CKD (n = 1,745); recovery from AKI (n = 2,392); and ongoing AKI (n = 3,373). We observed 4,458 kidney graft losses during a median follow-up period of 5.7 years. Compared with no AKI, ongoing AKI was associated with an increased risk of graft failure (hazard ratio [HR], 1.24 [95% CI, 1.13-1.35]). The HRs for graft failure in the undetermined AKI/CKD and recovery from AKI groups (1.22 [95% CI, 1.07-1.38] and 1.18 [95% CI, 1.06-1.31], respectively) were similar to those observed in the ongoing AKI group. The adverse effect of deceased donor AKI was no longer evident when relying either on the admission or the lowest Scr throughout the procurement procedure as baseline Scr. Limitations No measurement of urine output in donors. Conclusions Deceased donor ongoing AKI, undetermined AKI/CKD, and recovery from AKI according to back-estimated baseline Scr are associated with decreased graft survival. The definition of baseline Scr as the first value measured on admission would have led to a misclassification bias and erroneous estimates.

Background Identifying biomarkers to predict kidney transplant failure and to define new therapeutic targets requires more comprehensive understanding of the immune response to chronic allogeneic stimulation. Methods We investigated the frequency and function of CD8+ T cell subsets-including effector memory (EM) and terminally differentiated EM (TEMRA) CD8+ T cells-in blood samples from 284 kidney transplant recipients recruited 1 year post-transplant and followed for a median of 8.3 years. We also analyzed CD8+ T cell reactivity to donor-specific PBMCs in 24 patients who had received living-donor kidney transplants. Results Increased frequency of circulating TEMRA CD8+ T cells at 1 year post-transplant associated with increased risk of graft failure during follow-up. This association remained after adjustment for a previously reported composite of eight clinical variables, the Kidney Transplant Failure Score. In contrast, increased frequency of EM CD8+ T cells associated with reduced risk of graft failure. A distinct TEMRA CD8+ T cell subpopulation was identified that was characterized by expression of FcγRIIIA (CD16) and by high levels of proinflammatory cytokine secretion and cytotoxic activity. Although donor-specific stimulation induced a similar rapid, early response in EM and TEMRA CD8+ T cells, CD16 engagement resulted in selective activation of TEMRA CD8+ T cells, which mediated antibody-dependent cytotoxicity. Conclusions At 1 year post-transplant, the composition of memory CD8+ T cell subsets in blood improved prediction of 8-year kidney transplant failure compared with a clinical-variables score alone. A subpopulation of TEMRA CD8+ T cells displays a novel dual mechanism of activation mediated by engagement of the T-cell receptor or of CD16. These findings suggest that TEMRA CD8+ T cells play a pivotal role in humoral and cellular rejection and reveal the potential value of memory CD8+ T cell monitoring for predicting risk of kidney transplant failure.

International audience; Background: The COVID-19 pandemic has resulted in worldwide kidney transplantation (KT) moratoriums. The impacts of these moratoriums on the life expectancy of KT candidates remain unclear.Methods: We simulated the evolution of several French candidate populations for KT using a multistate semi-Markovian approach and according to moratorium durations ranging from 0 to 24 mo. The transition rates were modeled from the 63 927 French patients who began dialysis or were registered on the waiting list for KT between 2011 and 2019.Results: Among the 8350 patients active on the waiting list at the time of the French KT moratorium decided on March 16, 2020, for 2.5 mo, we predicted 4.0 additional months (confidence interval [CI], 2.8-5.0) on the waiting list and 42 additional deaths (CI, -70 to 150) up to March 16, 2030, compared with the scenario without moratorium. In this population, we reported a significant impact for a 9-mo moratorium duration: 135 attributable deaths (CI, 31-257) up to March 16, 2030. Patients who became active on the list after March 2020 were less impacted; there was a significant impact for an 18-mo moratorium (175 additional deaths [CI, 21-359]) in the 10 862 prevalent end-stage renal disease patients on March 16, 2020 and for a 24-mo moratorium (189 additional deaths [CI, 10-367]) in the 16 355 incident end-stage renal disease patients after this date.Conclusion: The temporary moratorium of KT during a COVID-19 peak represents a sustainable decision to free up hospitals' resources if the moratorium does not exceed a prolonged period.

BackgroundMost studies comparing the efficacy of hypothermic machine perfusion (HMP) versus static cold storage (SCS) are based on short-term outcomes. We aimed to better evaluate the mid-term impact of HMP in patients receiving expanded criteria donor (ECD) kidneys.MethodsThe analyses were based on the French Données Informatisées et VAlidées en Transplantation (DIVAT) observational cohort. Patients aged ≥45 years transplanted for the first or second times from an ECD donor since 2010 were studied. Our study reported the graft and/or patient survivals and the incidence of acute rejection episode. The Cox models and the Kaplan–Meier estimators, weighted on the propensity score, were used to study the times-to-events.ResultsAmong the 2019 included patients, 1073 were in the SCS group versus 946 in the HMP group. The mean life expectancy with functioning graft was 5.7 years [95% confidence interval (CI) 5.4–6.1] for the HMP cohort followed-up for 8 years post-transplantation versus 6.0 years (95% CI 5.7–6.2) for the SCS group. These mid-term results were comparable in the patients receiving grafts from donors aged ≥70 years and in the transplantations with cold ischaemia time ≥18 h.ConclusionsOur study challenges the utility of using HMP to improve mid-term patient and graft survival. Nevertheless, the improvement of the short-term outcomes is indisputable. It is necessary to continue technological innovations to obtain long-term results.

In causal studies, the near-violation of the positivity may occur by chance, because of sample-to-sample fluctuation despite the theoretical veracity of the positivity assumption in the population. It may mostly happen when the exposure prevalence is low or when the sample size is small. We aimed to compare the robustness of g-computation (GC), inverse probability weighting (IPW), truncated IPW, targeted maximum likelihood estimation (TMLE), and truncated TMLE in this situation, using simulations and one real application. We also tested different extrapolation situations for the sub-group with a positivity violation. The results illustrated that the near-violation of the positivity impacted all methods. We demonstrated the robustness of GC and TMLE-based methods. Truncation helped in limiting the bias in near-violation situations, but at the cost of bias in normal conditions. The application illustrated the variability of the results between the methods and the importance of choosing the most appropriate one. In conclusion, compared to propensity score-based methods, methods based on outcome regression should be preferred when suspecting near-violation of the positivity assumption.

Background Although kidney transplantation (KT) is considered the best treatment for end-stage renal disease (ESRD), there are concerns about its benefit in the obese population because of the increased incidence of post-transplant adverse events. We compared patients who underwent KT versus patients awaiting KT on dialysis. Methods We estimated the life expectancy [restricted mean survival time (RMST)] for a 10-year follow-up by matching on time-dependent propensity scores. The primary outcome was time to death. Results In patients with a body mass index (BMI) ≥30 kg/m2 (n = 2155 patients per arm), the RMST was 8.23 years [95% confidence interval (CI) 8.05–8.40] in the KT group versus 8.00 years (95% CI 7.82–8.18) in the awaiting KT group, a difference of 2.71 months (95% CI −0.19–5.63). In patients with a BMI ≥35 kg/m2 (n = 212 patients per arm), we reported no significant difference [8.56 years (95% CI 7.96–9.08) versus 8.66 (95% CI 8.10–9.17)]. Hence we deduced that KT in patients with a BMI between 30 and 35 kg/m2 was beneficial in terms of life expectancy. Conclusion Regarding the organ shortage, KT may be questionable for those with a BMI ≥35 kg/m2. These results do not mean that a BMI ≥35 kg/m2 should be a barrier to KT, but it should be accounted for in allocation systems to better assign grafts and maximize the overall life expectancy of ESRD patients.

Introduction Les capacites discriminantes d’un score sont souvent rapportees en utilisant l’aire sous la courbe ROC (AUC) non-ajustee. En transplantation renale, le Kidney Donor Risk Index (KDRI), score de qualite des greffons, a ete construit sur ce modele. Description Nous avons etudie les capacites du KDRI pour discriminer le risque d’echec de greffe renale, en considerant les covariables du receveur et de la transplantation qui pourraient fausser cette interpretation. Methodes Plus de 4000 patients ont ete inclus. Le critere de jugement principal etait le delai entre la transplantation et l’echec de greffe (retour en dialyse, retransplantation, deces). Nous avons calcule les capacites discriminantes non-ajustees du KDRI en utilisant l’AUC de la courbe ROC dependante du temps. Ensuite, nous avons etudie les capacites discriminantes intrinseques du KDRI, estimees par l’AUC de la courbe ROC dependante du temps standardisee et ponderee par les caracteristiques du receveur et de la greffe. Resultats L’AUC non-ajustee variait de 65 % (IC95 % [62 %–68 %]) a 68 % (IC95 % [65 %–70 %]) pour des pronostics de 1 a 7 ans de greffe. L’AUC ajustee sur les variables d’ajustement variait de 55 % (IC95 % [51–60]) a 56 % (IC95 % [52–59]) pour des pronostics de 1 a 7 ans de greffe. Les variables d’ajustement retenues etaient âge, IMC, antecedents d’hypertension, de maladie cardiovasculaire ou de cancer du receveur, temps passe en dialyse, temps d’ischemie froide, incompatibilites HLA, immunisation pre-greffe ( Fig. 1 ). Conclusion Les capacites discriminantes du KDRI sur cette cohorte de receveurs francais sont faibles, et semblent moindres encore lorsque les caracteristiques du receveur et de la transplantation sont prises en consideration. Considerer les variables d’ajustement dans l’evaluation des performances discriminantes d’un marqueur semble donc important. L’etude de l’effet causal sur la survie patient–greffon de la qualite de la transplantation definie par le KDRI a egalement montre de faibles differences de survie, suggerant une possible extension du pool de greffons disponibles.

BACKGROUND: In kidney transplantation, dynamic prediction of patient and kidney graft survival (DynPG) may help to promote therapeutic alliance by delivering personalized evidence-based information about long-term graft survival for kidney transplant recipients. The objective of the current study is to externally validate the DynPG. METHODS: Based on 6 baseline variables, the DynPG can be updated with any new serum creatinine measure available during the follow-up. From an external validation sample of 1637 kidney recipients with a functioning graft at 1-year posttransplantation from 2 European transplantation centers, we assessed the prognostic performance of the DynPG. RESULTS: As one can expect from an external validation sample, differences in several recipient, donor, and transplantation characteristics compared with the learning sample were observed. Patients were mainly transplanted from deceased donors (91.6% versus 84.8%; P < 0.01), were less immunized against HLA class I (18.4% versus 32.7%; P < 0.01) and presented less comorbidities (62.2% for hypertension versus 82.7%, P < 0.01; 25.1% for cardiovascular disease versus 33.9%, P < 0.01). Despite these noteworthy differences, the area under the ROC curve varied from 0.70 (95% confidence interval [CI], 0.64-0.76) to 0.76 (95% CI, 0.64-0.88) for prediction times at 1 and 6 years posttransplantation respectively, and calibration plots revealed reasonably accurate predictions. CONCLUSIONS: We validated the prognostic capacities of the DynPG in terms of both discrimination and calibration. Our study showed the robustness of the DynPG for informing both the patient and the physician, and its transportability for a cohort presenting different features than the one used for the DynPG development. ispartof: TRANSPLANTATION vol:105 issue:2 pages:396-403 ispartof: location:United States status: published

Background No study to our knowledge has examined the use of observational data to emulate a clinical trial whereby patients at the time of kidney transplant proposal are randomly assigned to an awaiting transplantation or transplantation group. The main methodologic issue is definition of the baseline for dialyzed patients assigned to awaiting transplantation, resulting in the inability to use common propensity score-based approaches. We aimed to use time-dependent propensity score to better appraise the benefit of transplantation. Methods We studied 23,231 patients included in the French registry and on a transplant waiting list from 2005 to 2016. The main outcome was time to death. By matching on time-dependent propensity score, we obtained 10,646 pairs of recipients (transplantation group) versus comparable patients remaining on dialysis (awaiting transplantation group). Results The baseline exposure, that is, pseudo-randomization, was matching time. Median follow-up time was 3.5 years. At 10 years' follow-up, the restricted mean survival time was 8.8 years [95% confidence interval (CI) = 8.7, 8.9] in the transplantation group versus 8.2 years (95% CI = 8.1, 8.3) in the awaiting transplantation group. The corresponding life expectancy gain was 6.8 months (95% CI = 5.5, 8.2), and this corresponded to one prevented death at 10 years for 13 transplantations. Conclusions Our study has estimated the life expectancy gain due to kidney transplantation. It confirms transplantation as the best treatment for end-stage renal disease. Furthermore, we demonstrated that this simple method should also be considered for estimating marginal effects of time-dependent exposures.

International audience; Barbiturates are proposed as a second/third line treatment for intracranial hypertension in traumatic brain injury (TBI) patients, but the literature remains uncertain regarding their benefit/risk balance. We aimed to evaluate the impact of barbiturates therapy in TBI patients with early intracranial hypertension on the intensive care unit (ICU) survival, the occurrence of ventilator-associated pneumonia (VAP), and the patient’s functional status at three months. We used the French AtlanREA prospective cohort of trauma patients. Using a propensity score-based methodology (inverse probability of treatment weighting), we compared patients having received barbiturates within the first 24 hours of admission (barbiturates group) and those who did not (control group). We used cause-specific Cox models for ICU survival and risk of VAP, and logistic regression for the 3-month Glasgow Outcome Scale (GOS) evaluation. Among the 1396 patients with severe trauma, 383 had intracranial hypertension on admission and were analyzed. Among them, 96 (25.1%) received barbiturates. The early use of barbiturates was significantly associated with increased ICU mortality (HR = 1.85, 95%CI 1.03–3.33). However, barbiturates treatment was not significantly associated with VAP (HR = 1.02, 95%CI 0.75–1.41) or 3-month GOS (OR = 1.67, 95%CI 0.84–3.33). Regarding the absence of relevant clinical trials, our results suggest that each early prescription of barbiturates requires a careful assessment of the benefit/risk ratio.

Introduction Tandis que les approches d’apprentissage automatique (c.-a-d., « machine learning ») sont de plus en plus utilisees en prediction, leurs applications en inference causale sont plus recentes. Nous proposons une approche combinant la G-computation et l’apprentissage automatique pour estimer l’effet causal d’une exposition binaire sur un evenement binaire. Methodes Nous avons evalue et compare, via une etude de simulation, les performances de regressions logistiques penalisees, d’un reseau de neurones, d’une machine a vecteurs de support, d’une classification renforcee, d’arbres de regression et d’un « super learner ». Nous avons etudie six scenarios differents faisant varier les tailles d’echantillon et les relations entre les covariables, l’exposition binaire et l’evenement binaire. Resultats L’utilisation du super learner pour specifier le modele de nuisance de la G-computation surpassait les autres approches en termes de biais ou de variance lors de l’estimation de l’effet causal, particulierement pour les tailles d’echantillon moderees. L’utilisation de la machine a support de vecteurs permettait egalement d’obtenir des estimations causales performantes, bien que le biais moyen fut legerement plus eleve qu’avec le super learner. Conclusion L’utilisation d’approches d’apprentissage automatique peut etre pertinente en inference causale. Contrairement a une idee recue, ceci est egalement vrai pour des tailles d’echantillon constituees d’une centaine d’individus comme dans la majorite des etudes medicales. La G-computation avec le super learner est disponible dans le package R RISCA.

International audience; Purpose: The long-term effect of beta-interferon and glatiramer acetate on multiple sclerosis (MS) disability progression has resulted in controversial results, probably due to a lack of appropriate control of biases as raised in observational studies. In particular, the time of the therapeutic decision is difficult to define when the controls are not treated.Methods: This retrospective observational study was based on a series of patients from the MS expert center in Rennes, France. We used a time-dependent propensity score defined as the linear predictor of a Cox model estimating the hazard of being treated at each time from MS onset. The matching procedure resulted in two groups: patients matched as treated and as not yet treated. The restricted mean times (RMST) to reach a moderate level of disability or worsening of the disability were compared between the two groups in an intention-to-treat analysis.Results: Of the 2383 patients included in the study, 556 were matched as treated. The matching procedure provided a good balance of both the time-fixed and the time-dependent covariates. A slight difference was observed for the time to reach a moderate level of disability, in favor of the "not yet treated" group (difference in the RMST: -0.62 [-0.91; -0.33]) while no difference was found in terms of worsening of the disability (-0.03 [-0.24; 0.33]).Conclusion: This unexpected result is probably due to unmeasured confounders. However, this time-dependent PS warrants consideration in long-term effectiveness studies.

International audience

Objectives We aimed to illustrate that considering covariates can lead to meaningful interpretation of the discriminative capacities of a prognostic marker. For this, we evaluated the ability of the Kidney Donor Risk Index (KDRI) to discriminate kidney graft failure risk. Study design and setting From 4114 French patients, we estimated the adjusted area under the time-dependent ROC curve by standardizing the marker and weighting the observations. By weighting the contributions, we also studied the impact of KDRI-based transplantations on the patient and graft survival. Results The covariate-adjusted AUC varied from 55% (95% confidence interval [CI]: 51–60%) for a prognostic up to 1 year post-transplantation to 56% (95% CI: 52–59%) up to 7 years. The Restricted Mean Survival Time (RMST) was 6.44 years for high-quality graft recipients (95% CI: 6.30–6.56) and would have been 6.31 years (95% CI: 6.13–6.46) if they had medium-quality transplants. The RMST was 5.10 years for low-quality graft recipients (95% CI: 4.90–5.31) and would have been 5.52 years (95% CI: 5.17–5.83) if they had medium-quality transplants. Conclusion We demonstrated that the KDRI discriminative capacities were mainly explained by the recipient characteristics. We also showed that counterfactual estimations, often used in causal studies, are also interesting in predictive studies, especially regarding the new available methods.

Background. Whilst there are numbers of publications comparing the relationship between the body mass index (BMI) of kidney transplant recipients and graft/patient survival, no study was performed in French patients. Methods. In this study, cause-specific Cox models were used to study patients and graft survival and several other different times-to-event. Logistic regressions were performed to study surgical complications at 30 days post-transplantation as well as delayed graft function. Results. Among the 4691 included patients, 747 patients were considered obese with a BMI level higher than 30 kg/m². We observed a higher mortality of obese recipients (HR = 1.37, p = 0.0086) and higher risks of serious bacterial infections (HR = 1.24, p = 0.0006) and cardiac complications (HR = 1.45, p

In time-to-event settings, g-computation and doubly robust estimators are based on discrete-time data. However, many biological processes are evolving continuously over time. In this paper, we extend the g-computation and the doubly robust standardisation procedures to a continuous-time context. We compare their performance to the well-known inverse-probability-weighting (IPW) estimator for the estimation of the hazard ratio and restricted mean survival times difference, using a simulation study. Under a correct model specification, all methods are unbiased, but g-computation and the doubly robust standardisation are more efficient than inverse probability weighting. We also analyse two real-world datasets to illustrate the practical implementation of these approaches. We have updated the R package RISCA to facilitate the use of these methods and their dissemination., Accepted for publication in Statistical Methods in Medical Research, 16 pages, including 4 figures and 2 tables

Background One of the main challenges in multiple sclerosis (MS) is to predict disease progression based on patient characteristics and therapeutic strategies. We therefore performed a systematic review to critically appraise the composite tools available for this purpose. Methods We performed electronic database searches in MEDLINE, EMBASE, Web of Science and the Cochrane Library. We included studies in English or French that developed and/or validated a predictive model for MS patients. Two reviewers independently screened articles by title and abstract. Three teams of two reviewers assessed the full text of each relevant study. Results Database searches yielded 6,035 studies after deduplication. Among the 42 screened full texts, 15 articles satisfied the eligibility criteria. Of these, six articles examined the development of predictive tools, six articles aimed to validate existing tools and three articles proposed both development and validation. We identified numerous methodological pitfalls, especially the lack of adequate validations in terms of discrimination and calibration. Only two scoring systems were externally validated several times: the Rio and the modified Rio scores. Nevertheless, their accuracies were highly variable, ranging from 65% to 91%. Conclusions Overall, there is a lack of validated predictive tools in MS, and further external validation of the existing ones are required. Demonstration of the clinical usefulness is also needed prior to being transferred into clinical practice. Finally, our study illustrates that the MS literature needs to integrate good standards in developing and validating predictive models.

Controlling for confounding bias is crucial in causal inference. Distinct methods are currently employed to mitigate the effects of confounding bias. Each requires the introduction of a set of covariates, which remains difficult to choose, especially regarding the different methods. We conduct a simulation study to compare the relative performance results obtained by using four different sets of covariates (those causing the outcome, those causing the treatment allocation, those causing both the outcome and the treatment allocation, and all the covariates) and four methods: g-computation, inverse probability of treatment weighting, full matching and targeted maximum likelihood estimator. Our simulations are in the context of a binary treatment, a binary outcome and baseline confounders. The simulations suggest that considering all the covariates causing the outcome led to the lowest bias and variance, particularly for g-computation. The consideration of all the covariates did not decrease the bias but significantly reduced the power. We apply these methods to two real-world examples that have clinical relevance, thereby illustrating the real-world importance of using these methods. We propose an R package RISCA to encourage the use of g-computation in causal inference.

Introduction Une des complications les plus severes du diabete de type 1 est l’atteinte renale pouvant mener a l’insuffisance renale terminale. Bien que des progres majeurs ont ete faits pour ralentir la progression de l’insuffisance renale certains patients diabetiques de type 1 (D1) necessitent une prise en charge en dialyse. Ces patients semblent avoir plus de comorbidites et un moins bon pronostic que les patients non diabetiques (D0) ou que les patients diabetiques de type 2 (D2) au demarrage de la dialyse. Description Nous avons utilise le Registre d’Epidemiologie et d’Information en Nephrologie (REIN) qui collige les informations sur tous les patients au demarrage de la dialyse en France. Methodes Sur cette cohorte, nous avons compare la survie des differents groupes a l’aide de modeles a risques proportionnels de Cox. Nous avons ensuite explore differents chemins causaux en realisant des modeles plus ou moins ajustes et complete les analyses de sensibilites en remplacant les donnees manquantes par imputations multiples ( Fig. 1 ). Resultats Entre le 1 janvier 2010 et le 31 decembre 2016, 1596 patients D1 ont ete compare a 39540 patients D0 et 28170 patients D2, avec un suivi jusqu’au 31 decembre 2017. Les patients D1 avaient 26 % de plus de risque de deces que les patients non diabetiques (HR 1,26 IC 95 % [1,14-1,40] p Conclusion Cette cohorte quasi exhaustive des patients diabetiques de type 1 demarrant la dialyse montre le mauvais pronostic des patients D1 en termes de survie. Le sur-ajustement des modeles statistiques risque d’entrainer une sous estimation du risque en considerant des complications du diabetes comme des comorbidites independantes.

BACKGROUND: Identifying biomarkers to predict kidney transplant failure and to define new therapeutic targets requires more comprehensive understanding of the immune response to chronic allogeneic stimulation. METHODS: We investigated the frequency and function of CD8(+) T cell subsets—including effector memory (EM) and terminally differentiated EM (TEMRA) CD8(+) T cells—in blood samples from 284 kidney transplant recipients recruited 1 year post-transplant and followed for a median of 8.3 years. We also analyzed CD8(+) T cell reactivity to donor-specific PBMCs in 24 patients who had received living-donor kidney transplants. RESULTS: Increased frequency of circulating TEMRA CD8(+) T cells at 1 year post-transplant associated with increased risk of graft failure during follow-up. This association remained after adjustment for a previously reported composite of eight clinical variables, the Kidney Transplant Failure Score. In contrast, increased frequency of EM CD8(+) T cells associated with reduced risk of graft failure. A distinct TEMRA CD8(+) T cell subpopulation was identified that was characterized by expression of FcγRIIIA (CD16) and by high levels of proinflammatory cytokine secretion and cytotoxic activity. Although donor-specific stimulation induced a similar rapid, early response in EM and TEMRA CD8(+) T cells, CD16 engagement resulted in selective activation of TEMRA CD8(+) T cells, which mediated antibody-dependent cytotoxicity. CONCLUSIONS: At 1 year post-transplant, the composition of memory CD8(+) T cell subsets in blood improved prediction of 8-year kidney transplant failure compared with a clinical-variables score alone. A subpopulation of TEMRA CD8(+) T cells displays a novel dual mechanism of activation mediated by engagement of the T-cell receptor or of CD16. These findings suggest that TEMRA CD8(+) T cells play a pivotal role in humoral and cellular rejection and reveal the potential value of memory CD8(+) T cell monitoring for predicting risk of kidney transplant failure.

ObjectiveIn this study, we compared the effectiveness of teriflunomide (TRF) and dimethyl fumarate (DMF) on both clinical and MRI outcomes in patients followed prospectively in the Observatoire Français de la Sclérose en Plaques.MethodsA total of 1,770 patients with relapsing-remitting multiple sclerosis (RRMS) (713 on TRF and 1,057 on DMF) with an available baseline brain MRI were included in intention to treat. The 1- and 2-year postinitiation outcomes were relapses, increase of T2 lesions, increase in Expanded Disability Status Scale score, and reason for treatment discontinuation. Propensity scores (inverse probability weighting) and logistic regressions were estimated.ResultsThe confounder-adjusted proportions of patients were similar in TRF- compared to DMF-treated patients for relapses and disability progression after 1 and 2 years. However, the adjusted proportion of patients with at least one new T2 lesion after 2 years was lower in DMF compared to TRF (60.8% vs 72.2%, odds ratio [OR] 0.60, p < 0.001). Analyses of reasons for treatment withdrawal showed that lack of effectiveness was reported for 8.5% of DMF-treated patients vs 14.5% of TRF-treated patients (OR 0.54, p < 0.001), while adverse events accounted for 16% of TRF-treated patients and 21% of DMF-treated patients after 2 years (OR 1.39, p < 0.001).ConclusionsAfter 2 years of treatment, we found similar effectiveness of DMF and TRF in terms of clinical outcomes, but with better MRI-based outcomes for DMF-treated patients, resulting in a lower rate of treatment discontinuation due to lack of effectiveness.Classification of evidenceThis study provides Class III evidence that for patients with RRMS, TRF and DMF have similar clinical effectiveness after 2 years of treatment.

International audience; Surveillance biopsies after renal transplantation remain debatable. To drive the decision of such intervention, we propose a predictive score of abnormal histology at 1-year post-transplantation, named 1-year Renal Biopsy Index (1-RBI). We studied 466 kidney recipients from the DIVAT cohort alive with a functioning graft and a surveillance biopsy at 1-year post-transplantation. Patients displaying abnormal histology (49%) (borderline, acute rejection, interstitial fibrosis and tubular atrophy [IFTA] grade 2 or 3, glomerulonephritis) were compared to the normal or subnormal (IFTA grade 1) histology group. Obtained from a lasso penalized logistic regression , the 1-RBI was composed of recipient gender, serum creatinine at 3, 6, and 12 month post-transplantation and anticlass II immunization at transplantation (internal validation: AUC = 0.71, 95% CI [0.53-0.83]; external validation: AUC = 0.62, 95% CI [0.58-0.66]). While we could not determinate a threshold able to identify patients at high chance of normal or subnormal histology, we estimated and validated a discriminating threshold capable of identifying a subgroup of 15% of the patients with a risk of abnormal histology higher than 80%. The 1-RBI is computable online at www.divat.fr. The 1-RBI could be a useful tool to standardize 1-year biopsy proposal and may for instance help to indicate one in case of high risk of abnormal histology.

International audience; BACKGROUND: A significant number of studies have compared graft outcomes between patients with Pre-emptive Kidney Transplantation (PreKT) and patients on Dialysis before their Kidney Transplantation (DiaKT). These studies have suffered from the limitation that the DiaKT group is composed of all the dialysed patients, including those placed on a waiting list at the time of their first dialysis session. This seriously questions the comparability of these patients with those placed on the waiting list a long time before the need for renal replacement therapy. The aim of this study was to precisely evaluate the causal effect of PreKT from deceased donors.METHODS: Data were extracted from the multicentric French DIVAT (Données Informatisées et VAlidées en Transplantation) cohort. The DiaKT group was composed of patients placed on the waiting list with an initial intention of pre-emptive transplantation. Cause-specific Cox models with propensity scores (inverse probability weighting) were used to study the patient and graft outcomes.RESULTS: Among the 1138 included patients, 554 patients were in the PreKT group. The outcomes of the PreKT group were similar compared with the DiaKT group. In particular, the life expectancy with a functioning graft was 8.51 years [95% confidence interval (CI) 8.20-8.81] for the PreKT recipients versus 8.49 years (95% CI 8.15-8.84) for the DiaKT recipients.CONCLUSIONS: Our results challenge the utility of PreKTs from deceased donors, especially with regard to the consequential increase in the waiting list.

Calculation of the utility scores for the different patient health and care states. The corresponding text explains the calculation of the utility scores for the different patient health and care states. (PDF 109â kb)

International audience; OBJECTIVES: Structural valve deterioration (SVD) remains a major bioprosthesis-related complication, as recently described for the Mitroflow~valve (models LX and 12A) (LivaNova, London, United Kingdom). The real incidence of the SVD risk remains unclear, often due to methodologic pitfalls by systematically using the Kaplan-Meier estimator and/or the Cox model. In this report, we propose for the first time a precise statistical modeling of this issue. METHODS: Five hundred sixty-one patients who underwent aortic valve replacement with the aortic Mitroflow valve between 2002 and 2007 were included. We used an illness-death model for interval-censored data. Median follow-up was 6.6~years; 103 cases of SVD were diagnosed. RESULTS: The 4-year and 7-year SVD cumulative incidences after the first anniversary of surgery were 15.2% (95% confidence interval, 11.9-19.1) and 31.0% (95% confidence interval, 25.8-37.2), respectively. Female gender, dyslipidemia, chronic obstructive pulmonary disease, and severe patient-prosthesis mismatch were significant risk factors of SVD. The occurrence of SVD was associated with a 2-fold increase in the risk of death. CONCLUSIONS: Appropriate statistical models should be used to avoid underestimating the SVD complication associated with worse long-term survival.

International audience; BACKGROUND: In relapsing-remitting multiple sclerosis (RRMS), relapse severity and residual disability are difficult to predict. Nevertheless, this information is crucial both for guiding relapse treatment strategies and for informing patients. OBJECTIVE: We, therefore, developed and validated a clinical-based model for predicting the risk of residual disability at 6 months post-relapse in MS. METHODS: We used the data of 186 patients with RRMS collected during the COPOUSEP multicentre trial. The outcome was an increase of ≥ 1 EDSS point 6 months post-relapse treatment. We used logistic regression with LASSO penalization to construct the model, and bootstrap cross-validation to internally validate it. The model was externally validated with an independent retrospective French single-centre cohort of 175 patients. RESULTS: The predictive factors contained in the model were age > 40 years, shorter disease duration, EDSS increase ≥ 1.5 points at time of relapse, EDSS = 0 before relapse, proprioceptive ataxia, and absence of subjective sensory disorders. Discriminative accuracy was acceptable in both the internal (AUC 0.82, 95% CI [0.73, 0.91]) and external (AUC 0.71, 95% CI [0.62, 0.80]) validations. CONCLUSION: The predictive model we developed should prove useful for adapting therapeutic strategy of relapse and follow-up to individual patients.

Background Informing kidney transplant recipients of their prognosis and disease progression is of primary importance in a patient-centred vision of care. By participating in decisions from the outset, transplant recipients may be more adherent to complex medical regimens due to their enhanced understanding. Methods We proposed to include repeated measurements of serum creatinine (SCr), in addition to baseline characteristics, in order to obtain dynamic predictions of the graft failure risk that could be updated continuously during patient follow-up. Adult recipients from the French Données Informatisées et VAlidées en Transplantation (DIVAT) cohort transplanted for the first or second time from a heart-beating or living donor and alive with a functioning graft at 1 year post-transplantation were included. Results The model was composed of six baseline parameters, in addition to the SCr evolution. We validated the dynamic predictions by evaluating both discrimination and calibration accuracy. The area under the receiver operating characteristic curve varied from 0.72 to 0.76 for prediction times at 1 and 6 years post-transplantation, respectively, while calibration plots showed correct accuracy. We also provided an online application tool (https://shiny.idbc.fr/DynPG). Conclusion We have created a tool that, for the first time in kidney transplantation, predicts graft failure risk both at an individual patient level and dynamically. We believe that this tool would encourage willing patients into participative medicine.

International audience; PURPOSE: From the MINDACT trial, Cardoso et al. did not demonstrate a significant efficacy for adjuvant chemotherapy (CT) for women with early-stage breast cancer presenting high clinical and low genomic risks. Our objective was to assess the usefulness of the 70-gene signature in this population by using an alternative endpoint: the number of Quality-Adjusted Life-Years (QALYs), i.e., a synthetic measure of quantity and quality of life. METHODS: Based on the results of the MINDACT trial, we simulated a randomized clinical trial consisting of 1497 women with early-stage breast cancer presenting high clinical and low genomic risks. The individual preferences for the different health states and corresponding decrements were obtained from the literature. RESULTS: The gain in terms of 5-year disease-free survival was 2.8% (95% CI from -\,0.1 to 5.7%, from 90.4% for women without CT to 93.3% for women with CT). In contrast, due to the associated side effects, CT significantly reduced the number of QALYs by 62~days (95% CI from 55 to 70 days, from 4.13~years for women without CT to 3.96~years for women with CT). CONCLUSION: Our results support the conclusions published by Cardoso et al. by providing additional evidence that the 70-gene signature can be used to avoid overtreatment by CT for women with high clinical risk but low genomic risk.

International audience; In 2002, the United Network for Organ Sharing proposed increasing the pool of donor kidneys to include Expanded Criteria Donor (ECD). Outside the USA, the ECD definition remains the one used without questioning whether such a graft allocation criterion is valid worldwide. We performed a meta-analysis to quantify the differences between ECD and Standard Criteria Donor (SCD) transplants. We paid particular attention to select studies in which the methodology was appropriate and we took into consideration the geographical area. Thirty-two publications were included. Only five studies, all from the USA, reported confounder-adjusted hazard ratios comparing the survival outcomes between ECD and SCD kidney transplant recipients. These five studies confirmed that ECD recipients seemed to have poorer prognosis. From 29 studies reporting appropriate survival curves, we estimated the 5-year pooled nonadjusted survivals for ECD and SCD recipients. The relative differences between the two groups were lower in Europe than in North America, particularly for death-censored graft failure. It is of primary importance to propose appropriate studies for external validation of the ECD criteria in non-US kidney transplant recipients.

Grégoire Couvrat-Desvergnes, Apolline Salama, and Ludmilla Le Berre contributed equally to this work.; International audience; BACKGROUND: Rabbit-generated antithymocyte globulins (ATGs), which target human T cells, are widely used as immunosuppressive agents during treatment of kidney allograft recipients. However, ATGs can induce immune complex diseases, including serum sickness disease (SSD). Rabbit and human IgGs have various antigenic differences, including expression of the sialic acid Neu5Gc and α-1-3-Gal (Gal), which are not synthesized by human beings. Moreover, anti-Neu5Gc antibodies have been shown to preexist and be elicited by immunization in human subjects. This study aimed to assess the effect of SSD on long-term kidney allograft outcome and to compare the immunization status of grafted patients presenting with SSD following ATG induction treatment.METHODS: We analyzed data from a cohort of 889 first kidney graft recipients with ATG induction (86 with SSD [SSD(+)] and 803 without SSD [SSD(-)]) from the Données Informatisées et Validées en Transplantation data bank. Two subgroups of SSD(+) and SSD(-) patients that had received ATG induction treatment were then assessed for total anti-ATG, anti-Neu5Gc, and anti-Gal antibodies using ELISA assays on sera before and after transplantation.RESULTS: SSD was significantly associated with long-term graft loss (>10 years, P = 0.02). Moreover, SSD(+) patients exhibited significantly elevated titers of anti-ATG (P = 0.043) and anti-Neu5Gc (P = 0.007) IgGs in late post-graft samples compared with SSD(-) recipients.CONCLUSION: In conclusion, our data indicate that SSD is a major contributing factor of late graft loss following ATG induction and that anti-Neu5Gc antibodies increase over time in SSD(+) patients.FUNDING: This study was funded by Société d'Accélération du Transfert de Technologies Ouest Valorisation, the European FP7 "Translink" research program, the French National Agency of Research, Labex Transplantex, the Natural Science and Engineering Research Council of Canada, and the Canadian Foundation for Innovation.

Dans le contexte de maladies curables a longue duree d’evolution, tel qu’un cancer ou un accident vasculaire cerebral, les patients ont un risque de deces superieur a celui de la population generale au moment du diagnostic. Leur prise en charge peut reduire le risque de deces lie a la maladie pour retrouver un risque proche de celui de la population generale. Ces patients peuvent alors etre consideres comme gueris. Pour etudier le risque de deces lie ou independant a la pathologie, on peut utiliser des modeles de survie cause-specifique. Toutefois, la cause de survenue d’un deces n’est pas toujours connue ou attribuable. Les modeles de survie relative sont une alternative interessante puisqu’ils ne requierent pas l’identification des causes de deces. L’approche par modele de survie relative a risque additif suppose le risque instantane observe λO de deces au temps t defini comme la somme du risque instantane de deces en population generale λP pour un âge donne (Aget) au temps t et d’un exces de risque instantane λE de deces attribuable a la maladie au temps t : λO(t) = λP (Aget) + λE(t) (Esteve et al., 1990). Ces modeles ont ete proposes pour etudier la guerison des patients (Yu et al., 2013). Le delai de guerison est ainsi defini comme le temps a partir duquel le risque en exces λE(t) devient negligeable et tend durablement vers 0 (Boussari et al., 2018). Le risque observe rejoint alors le risque de deces en population generale. On peut etudier alternativement un risque observe defini relativement au risque attendu en population generale. Dans les modeles de survie relative a risque multiplicatif, le risque instantane observe λO de deces au temps t est proportionnel au risque en population generale λP pour un âge donne (Aget) : λO(t) = λP (Aget) * λR(t), avec λR(t) un risque instantane de deces relatif a la maladie (Andersen et al., 1985). Cette approche n’a, a notre connaissance, jamais ete utilisee pour estimer un delai de guerison. Nous proposons une definition originale du delai de guerison a l’aide des modeles de survie relative a risque multiplicatif, comme le delai a partir duquel le risque relatif λR(t) est durablement proche de 1 ou inferieur. Le risque instantane observe rejoint alors le risque de deces en population generale. L’objectif de ce travail est de comparer les approches de survie relative a risque additif et a risque multiplicatif dans leur capacite a estimer correctement un delai de guerison, a l’aide d’une etude de simulation. Nous simulons 1000 jeux de 1000 patients pour lesquels on observe le couple de variables aleatoires (Ti, δi) ou T est le minimum entre le temps de censure C (δ=0) et le temps de survenue T* de l’evenement attendu (δ=1), ici le deces. Les temps d’evenements sont simules selon la procedure de Beyersmann et al. (2009). On considere un scenario de guerison avec un risque en exces qui tend vers 0 apres 5 ans post-inclusion. Pour chaque jeu, le delai de guerison de chaque individu est estime en fonction de covariables explicatives selon les deux approches de survie relative. Pour les modeles de survie relative a risque additif, le risque en exces est par definition strictement positif. L’estimation d’un delai de guerison se trouve donc en limite de l’espace de definition. Illustree par notre etude de simulation, l’approche par survie relative a risque multiplicatif qui estime le delai de guerison sans etre en limite de definition, permet une plus grande precision dans l’estimation de ce delai.

Introduction Dans les etudes observationnelles, les methodes basees sur le score de propension connaissent une popularite croissante pour estimer l’effet causal populationnel d’une exposition, d’un traitement ou d’une autre sorte d’intervention. Parmi ces differentes methodes, l’« Inverse-Probability-Weighting (IPW) [1] est decrite comme etant la plus fiable. Egalement basee sur le modele causal a resultats potentiels, la « g-computation » (GC) [2] constitue une alternative prometteuse aux scores de propension. Cependant, aucune etude de simulation n’a compare ces deux methodes dans un contexte de survie en presence de censure a droite. Methodes Nous proposons une etude de simulation comparant la GC et l’IPW en presence de divers taux de censure a droite. Sachant que l’estimateur le plus courant dans ce cadre, le hazard ratio, pose question pour tirer des conclusions causales claires, nous avons choisi de nous interesser a un estimateur absolu : la difference de temps de survie moyen restreint a un horizon temporel (RMST) [3] . Puisqu’il n’y a pas de consensus dans la litterature sur quelle ensemble de covariables ajuster dans ce contexte, nous investiguons egalement deux ensembles : toutes les variables causant l’evenement a l’inclusion ou seulement le sous-vecteur causant en plus l’exposition a l’inclusion. Resultats Dans notre etude de simulation, nous rapportons que la GC et l’IPW estiment de facon non-biaisee la difference de RMST, mais la GC est associee a une plus forte puissance statistique particulierement avec un ajustement sur les variables causant l’evenement. Ces resultats ont ete illustres par deux applications sur des donnees de vie reelles tirees d’etudes recemment publiees dans le domaine de la sclerose en plaques et de la transplantation renale. Conclusion Nous illustrons ici la puissance superieure de la GC par rapport au score de propension pour estimer un effet causal marginal absolu en presence de confusion non-dependante du temps. Notre travail s’inscrit donc dans la continuite d’une litterature emergente qui remet en question l’utilisation accrue des scores de propension par les statisticiens appliques pour l’estimation des effets causaux marginaux. Nous proposons en sus un package R dans lesquels la GC a ete implementee pour encourager son utilisation en inference causale.

Introduction Aucune etude basee sur des donnees observationnelles en transplantation renale n’a reussi a imiter un essai clinique ou les patients seraient randomises entre la decision d’acceptation ou de refus d’une proposition de greffon renal. La premiere limite methodologique d’une telle etude est le caractere dependant du temps de l’exposition, rendant complexe la definition d’un temps d’origine pour les patients non exposes. La seconde limite est la violation de l’hypothese de positivite et ainsi de la comparabilite des groupes au moment de la survenue de l’exposition a cause de facteurs de confusion dependants du temps. Ces deux limites empechent l’utilisation de methodes basees sur les scores de propension classiques. L’une des possibilites est l’utilisation de scores de propension dependants du temps. Notre objectif est d’evaluer le benefice d’accepter la transplantation renale par rapport au fait de rester en dialyse en mimant un essai clinique par l’emploi d’une methode basee sur les scores de propension dependants du temps. Methode Au total, 23 231 patients inclus dans le registre francais REIN inscrits sur liste d’attente de transplantation entre 2005 et 2016 ont ete inclus. Le critere de jugement principal est le deces. Par un appariement sur un score de propension dependant du temps, nous avons obtenu 10 707 paires de patients comparables et differant uniquement sur le bras de traitement : transplantes (groupe « accepter ») ou restant en dialyse (groupe « refus »). Resultats Le temps d’origine etait le temps d’appariement, c’est-a-dire le pseudo temps de randomisation dans l’essai clinique imite. Le temps median de suivi est de 3,5 annees. A 10 ans de suivi, l’esperance de vie etait de 8,8 ans (IC95 % : 8,7 a 8,9) dans le groupe « accepter » contre 8,2 ans (IC95 % : 8,1 a 8,3) dans le groupe « refus ». Cela correspond a un gain d’esperance de vie de 6,8 mois (IC95 % : 5,5 a 8,2) soit une mort evitee pour 12 transplantations acceptees a 10 ans. Conclusion Notre etude a estime pour la premiere fois le gain d’esperance de vie apportee par l’acceptation d’une proposition de transplantation renale par rapport au refus. Cela confirme que la transplantation renale est le meilleur traitement disponible pour l’insuffisance renale chronique terminale. Par ailleurs, nous avons demontre que cette methode simple peut etre utilisee pour estimer l’effet marginal d’une exposition dependante du temps.

International audience; Multistate models with interval-censored data, such as the illness-death model, are still not used to any considerable extent in medical research regardless of the significant literature demonstrating their advantages compared to usual survival models. Possible explanations are their uncommon availability in classical statistical software or, when they are available, by the limitations related to multivariable modelling to take confounding into consideration. In this paper, we propose a strategy based on propensity scores that allows population causal effects to be estimated: the inverse probability weighting in the illness semi-Markov model with interval-censored data. Using simulated data, we validated the performances of the proposed approach. We also illustrated the usefulness of the method by an application aiming to evaluate the relationship between the inadequate size of an aortic bioprosthesis and its degeneration or/and patient death. We have updated the R package multistate to facilitate the future use of this method.