Estimating time-to-onset of adverse drug reactions from spontaneous reporting databases.

BACKGROUND: Analyzing time-to-onset of adverse drug reactions from treatment exposure contributes to meeting pharmacovigilance objectives, i.e. identification and prevention. Post-marketing data are available from reporting systems. Times-to-onset from such databases are right-truncated because some patients who were exposed to the drug and who will eventually develop the adverse drug reaction may do it after the time of analysis and thus are not included in the data. Acknowledgment of the developments adapted to right-truncated data is not widespread and these methods have never been used in pharmacovigilance. We assess the use of appropriate methods as well as the consequences of not taking right truncation into account (naive approach) on parametric maximum likelihood estimation of time-to-onset distribution. METHODS: Both approaches, naive or taking right truncation into account, were compared with a simulation study. We used twelve scenarios for the exponential distribution and twenty-four for the Weibull and log-logistic distributions. These scenarios are defined by a set of parameters: the parameters of the time-to-onset distribution, the probability of this distribution falling within an observable values interval and the sample size. An application to reported lymphoma after anti TNF- α treatment from the French pharmacovigilance is presented. RESULTS: The simulation study shows that the bias and the mean squared error might in some instances be unacceptably large when right truncation is not considered while the truncation-based estimator shows always better and often satisfactory performances and the gap may be large. For the real dataset, the estimated expected time-to-onset leads to a minimum difference of 58 weeks between both approaches, which is not negligible. This difference is obtained for the Weibull model, under which the estimated probability of this distribution falling within an observable values interval is not far from 1. CONCLUSIONS: It is necessary to take right truncation into account for estimating time-to-onset of adverse drug reactions from spontaneous reporting databases.

Estimation in a competing risks proportional hazards model under length-biased sampling with censoring.

What population does the sample represent? The answer to this question is of crucial importance when estimating a survivor function in duration studies. As is well-known, in a stationary population, survival data obtained from a cross-sectional sample taken from the population at time t(0) represents not the target density f (t) but its length-biased version proportional to t f (t), for t > 0. The problem of estimating survivor function from such length-biased samples becomes more complex, and interesting, in presence of competing risks and censoring. This paper lays out a sampling scheme related to a mixed Poisson process and develops nonparametric estimators of the survivor function of the target population assuming that the two independent competing risks have proportional hazards. Two cases are considered: with and without independent censoring before length biased sampling. In each case, the weak convergence of the process generated by the proposed estimator is proved. A well-known study of the duration in power for political leaders is used to illustrate our results. Finally, a simulation study is carried out in order to assess the finite sample behaviour of our estimators.

Assessing Long-term Treatment Benefits Using Complementary Statistical Approaches: An In Silico Analysis of the Phase III Keynote-045 and Checkmate-214 Immune Checkpoint Inhibitor Trials.

BACKGROUND: The Keynote-045 trial illustrates that the long-term benefit (LTB) of treatment does not always translate to improved progression-free survival (PFS). Milestone survival and flexible parametric survival model with cure (FPCM) have been proposed as complementary statistical approaches to more comprehensively evaluate LTBs of treatments. OBJECTIVE: The current study compares milestone survival and FPCM analyses to evaluate treatment effects of immune checkpoint inhibitor (ICI) phase III trials. DESIGN, SETTING, AND PARTICIPANTS: Individual patient data, from initial and follow-up analyses of Keynote-045 (urothelial cancer) and Checkmate-214 (advanced renal cell carcinoma), were reconstructed for PFS. OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS: Each trial was reanalyzed using the Cox proportional hazard regression and two complementary methods (milestone survival and FPCM) to estimate treatment impact on the LTB. RESULTS AND LIMITATIONS: For each trial, there was evidence of nonproportional hazards. For the long-term analysis of the Keynote-045 trial, FPCM identified a time-dependent effect on PFS, but the Cox model found no statistical difference in PFS (hazard ratio, 0.90; 95% confidence interval, 0.75-1.08). Milestone survival and FPCM identified improvements in the LTB fractions. This was consistent with the results from the reanalysis of Keynote-045, based on the shorter follow-up, although the LTB fraction was not retained. The increase in PFS in Checkmate-214 was identified by both Cox model and FPCM. Experimental treatment-dependent improvement in the LTB fraction was demonstrated using milestone survival and FPCM. The LTB fraction estimated with FPCM was consistent with the results from the reanalysis of the shorter follow-up period. CONCLUSIONS: Although ICIs show substantial shifts toward LTBs in terms of PFS, based on a conventional Kaplan-Meier or Cox model analysis, our approach provides an alternative assessment of benefit-risk ratios for new therapeutics and facilitates communicating risk to patients. Kidney patients treated with ICIs can be counseled that they are potentially cured, but future work will need to definitively validate this conclusion. PATIENT SUMMARY: Although immune checkpoint inhibitor treatments show substantial shifts toward long-term benefits in terms of progression-free survival, a more rigorous attempt to quantify this shift, rather than simply using a Kaplan-Meier estimate or comparing progression-free survival curves using the classic Cox model, is warranted. Our results suggest that advanced renal cell carcinoma patients who had not received a previous treatment are functionally cured by nivolumab and ipilimumab, which is not the case for second-line urothelial carcinoma.

Assessment of Treatment Effects and Long-term Benefits in Immune Checkpoint Inhibitor Trials Using the Flexible Parametric Cure Model: A Systematic Review.

Importance: Compared with standard cytotoxic therapies, randomized immune checkpoint inhibitor (ICI) phase 3 trials reveal delayed benefits in terms of patient survival and/or long-term response. Such outcomes generally violate the assumption of proportional hazards, and the classical Cox proportional hazards regression model is therefore unsuitable for these types of analyses. Objective: To evaluate the ability of the flexible parametric cure model (FPCM) to estimate treatment effects and long-term responder fractions (LRFs) independently of prespecified time points. Evidence Review: This systematic review used reconstructed individual patient data from ICI advanced or metastatic melanoma and lung cancer phase 3 trials extracted from the literature. Trials published between January 1, 2010, and October 1, 2019, with long-term follow-up periods (maximum follow-up, ≥36 months in first line and ≥30 months otherwise) were selected to identify LRFs. Individual patient data for progression-free survival were reconstructed from the published randomized ICI phase 3 trial results. The FPCM was applied to estimate treatment effects on the overall population and on the following components of the population: LRF and progression-free survival in non-long-term responders. Results obtained were compared with treatment effects estimated using the Cox proportional hazards regression model. Findings: In this systematic review, among the 23 comparisons studied using the FPCM, a statistically significant association between the time-to-event component and experimental treatment was observed in the main analyses and confirmed in the sensitivity analyses of 18 comparisons. Results were discordant for 4 comparisons that were not significant by the Cox proportional hazards regression model. The LRFs varied from 1.5% to 12.7% for the control arms and from 4.6% to 38.8% for the experimental arms. Differences in LRFs varied from 2% to 29% and were significantly increased in the experimental compared with the control arms, except for 4 comparisons. Conclusions and Relevance: This systematic review of reconstructed individual patient data found that the FPCM was a complementary approach that provided a comprehensive and pertinent evaluation of benefit and risk by assessing whether ICI treatment was associated with an increased probability of patients being long-term responders or with an improved progression-free survival in patients who were not long-term responders.

The path of pre-ribosomes through the nuclear pore complex revealed by electron tomography.

Determining the path of single ribonucleoprotein (RNP) particles through the 100 nm-wide nuclear pore complex (NPC) by fluorescence microscopy remains challenging due to resolution limitation and RNP labeling constraints. By using high-pressure freezing and electron tomography, here we captured snapshots of the translocation of native RNP particles through NPCs in yeast and analyzed their trajectory at nanometer-scale resolution. Morphological and functional analyses indicate that these particles mostly correspond to pre-ribosomes. They are detected in 5-6% of the NPCs, with no apparent bias for NPCs adjacent to the nucleolus. Their path closely follows the central axis of the NPC through the nuclear and inner rings, but diverges at the cytoplasmic ring, suggesting interactions with the cytoplasmic nucleoporins. By applying a probabilistic queueing model to our data, we estimated that the dwell time of pre-ribosomes in the yeast NPC is ~90 ms. These data reveal distinct steps of pre-ribosome translocation through the NPC.