Critical reappraisal of phase III trials with immune checkpoint inhibitors in non-proportional hazards settings

Group sequential methods based on supremum logrank statistics under proportional and nonproportional hazards

Despite the widespread use of Cox regression for modeling treatment effects in clinical trials, in immunotherapy oncology trials and other settings therapeutic benefits are not immediately realized thereby violating the proportional hazards assumption. Weighted logrank tests and the so-called Maxcombo test involving the combination of multiple logrank test statistics have been advocated to increase power for detecting effects in these and other settings where hazards are nonproportional. We describe a testing framework based on supremum logrank statistics created by successively analyzing and excluding early events, or obtained using a moving time window. We then describe how such tests can be conducted in a group sequential trial with interim analyses conducted for potential early stopping of benefit. The crossing boundaries for the interim test statistics are determined using an easy-to-implement Monte Carlo algorithm. Numerical studies illustrate the good frequency properties of the proposed group sequential methods.

Dynamic nature of BRAF or KRAS p.G12C mutations in second-line therapy for advanced colorectal cancer patients: do early and late effects exist?

INTRODUCTION

The mitogen-activated protein kinase (MAPK) signalling network aberrations in metastatic colorectal cancer (mCRC) generate intrinsic dynamic effects and temporal variations that are crucial but often overlooked in clinical trial populations. Here, we investigate the time-varying impact of MAPK pathway mutation genotype on each treatment line's contribution to the overall clinical course.

METHODS

The PROMETEO study focused on mCRC patients undergoing second-line treatment at 20 hospitals. We evaluated genotypes and employed flexible models to analyse the dynamic effect of each mutation.

RESULTS

We examined data derived from 1160 patients. The effects of KRAS G12C or G12V, and BRAF V600E are clearly time-varying, with unexpected consequences such as the deleterious effect of BRAF V600E vs other genotypes dissipating over time when subjects receive antiangiogenics, or KRAS G12V and G12C showing increasing aggressiveness over time. Thus, contrary to expectations, the 12-month survival rate from the second line for those who survived >6 months was 49.9% (95% CI, 32.7-67.3) for KRAS G12C and 59% (95% CI, 38.5-80.6) for BRAF V600E.

CONCLUSIONS

The dynamic perspective is essential for understanding the behaviour of tumours with specific genotypes, especially from the second line onward. This may be relevant in patient monitoring and treatment decision-making, particularly in cases with distinct mutations.

Randomized controlled trials and alternative study designs in surgical oncology

Surgery is central to the cure of most solid cancers and an integral part of modern multimodal cancer management for early and advanced stage cancers. Decisions made by surgeons and multidisciplinary team members are based on best available knowledge for the defined clinical situation at hand. While surgery is both an art and a science, good decision-making requires data that are robust, valid, representative and, applicable to most if not all patients with a specific cancer. Such data largely comes from clinical observations and registries, and more preferably from trials conducted with the specific purpose of arriving at new answers. As part of the ESSO core curriculum development an increased focus has been put on the need to enhance research literacy among surgical candidates. As an expansion of the curriculum catalogue list and to enhance the educational value, we here present a set of principles and emerging concepts which applies to surgical oncologist for reading, understanding, planning and contributing to future surgeon-led cancer trials.

New late‐emphasis and combination tests based on infimum and supremum logrank statistics with application in oncology trials

Immunotherapy cancer clinical trials routinely feature an initial period during which the treatment is given without evident therapeutic benefit, which may be followed by a period during which an effective therapy reduces the hazard for event occurrence. The nature of this treatment effect is incompatible with the proportional hazards assumption, which has prompted much work on the development of alternative effect measures of frameworks for testing. We consider tests based on individual and combination of early- and late-emphasis infimum and supremum logrank statistics, describe how they can be implemented, and evaluate their performance in simulation studies. Through this work and illustrative applications we conclude that this class of test statistics offers a new and powerful framework for assessing treatment effects in cancer clinical trials involving immunotherapies.

Multi-state survival models with treatment effects and biomarkers: Simulations for study design assessment

BACKGROUND

Comparative effectiveness studies of cancer therapeutics in observational data face confounding by patterns of clinical treatment over time. The validity of survival analysis in longitudinal health records depends on study design choices including index date definition and model specification for covariate adjustment.

METHODS

Overall survival in cancer is a multi-state transition process with mortality and treatment switching as competing risks. Parametric Weibull regression quantifies proportionality of hazards across lines of therapy in real-world cohorts of 12 solid tumor types. Study design assessments compare alternative analytic models in simulations with realistic disproportionality. The multi-state simulation framework is adaptable to alternative treatment effect profiles and exposure patterns.

RESULTS

Event-specific hazards of treatment-switching and death are not proportional across lines of therapy in 12 solid tumor types. Study designs that include all eligible lines of therapy per subject showed lower bias and variance than designs that select one line per subject. Confounding by line number was effectively mitigated across a range of simulation scenarios by Cox proportional hazards models with stratified baseline hazards and inverse probability of treatment weighting.

CONCLUSION

Quantitative study design assessment can inform the planning of observational research in clinical oncology by demonstrating the potential impact of model misspecification. Use of empirical parameter estimates in simulation designs adapts analytic recommendations to the clinical population of interest.

Medicine before and after David Cox

Herein we recount the legacy of Sir David Roxbee Cox (15 July 1924 - 18 January 2022) from the perspective of practicing clinicians. His-pioneering work in developing the logistic and Cox proportional hazard regression models revolutionized the analysis and interpretation of categorical and time-to-event survival outcomes in modern medicine. This legacy is an inspiration for all those who follow on Sir David Cox's path.

A Review of Overall Survival Extrapolations of Immune-Checkpoint Inhibitors Used in Health Technology Assessments by the French Health Authorities

OBJECTIVES

Extrapolation is often required to inform cost-effectiveness (CE) evaluations of immune-checkpoint inhibitors (ICIs) since survival data from pivotal clinical trials are seldom complete. The objectives of this study were to evaluate the accuracy of estimates of long-term overall survival (OS) predicted in French CE assessment reports of ICIs, and to identify models presenting the best fit to the observed long-term survival data.

METHODS

A systematic review of French assessment reports of ICIs in the metastatic setting since inception until May 2020 was performed. A targeted literature review was conducted to collect associated extended follow-up of randomized controlled trials (RCTs) used in the CE assessment reports. Difference between projected and observed OS was calculated. A range of standard parametric and spline-based models were applied to the extended follow-up data from the RCT to determine the best-fitting survival models.

RESULTS

Of the 121 CE assessment reports published, 11 reports met the inclusion criteria. OS was underestimated in 73 percent of the CE assessment reports. The mean relative difference between each source was -13 percent (median: -15 percent; IQR: -0.4 to 26 percent). Models providing the best fit were those that could reflect nonmonotonic hazards.

CONCLUSIONS

Based on the available data at the time of submission, longer-term survival of ICIs was not fully captured by the extrapolation models used in CE assessments. Standard and flexible parametric models which can capture nonmonotonic hazard functions provided the best fit to the extended follow-up data. However, these models may still have performed poorly if fitted to survival data available at the time of submission to the French National Authority for Health.

Learning in times of stress: Lessons from COVID-19 that will last throughout this century

Systems tend toward inertia until an external pressure pushes them toward change; thus, a situation of crisis such as the COVID-19 pandemic represents an opportunity for technological innovation. The prevailing need for treatments and vaccines has impelled innovation in the world of randomized clinical trials (RCT), resorting to ideas that had been floating around for a while. Is this merely a circumstantial phenomenon or are new methods here to stay?

Bayesian interpretation of immunotherapy trials with dynamic treatment effects

INTRODUCTION

The mechanism of action of immune checkpoints inhibitors hinders the writing of rational statistical analysis plans for phase III randomised clinical trials (RCTs) because of their unpredictable dynamic effects. The purpose is to illustrate the advantages of Bayesian reporting of treatment efficacy analysis in immunotherapy RCTs, in contrast to frequentist reporting.

METHOD

Fourteen RCTs (one with two pairwise comparisons) that failed to achieve their primary objective (overall survival, OS) were selected. These RCTs were reanalysed using Bayesian Cox models with dynamic covariate coefficients and time-invariant models.

RESULTS

The RCTs that met inclusion criteria were 7 lung cancer trials, various other tumours, with antiPD1, antiPDL1 or antiCTLA4 therapies. The minimum detectable effect (δ_S) was superior to the true benefit observed in all cases, in conditions of non-proportional hazards. Schoenfeld tests indicated the existence of PH assumption violations (p<0.05) in 6/15 cases. The Bayesian Cox models revealed a probability of benefit >79% in all the RCTs, with the therapeutic equivalence hypothesis unlikely. The OS curves diverged after a median of 9.1 months. Since the divergency, no non-proportionality was evinced in 13/15, while the Wald tests achieved p<0.05 in 12/15 datasets. In all cases, the Bayesian Cox models with dynamic coefficients detected fluctuations of the hazard ratio, and increased 2-year OS was the most likely hypothesis.

CONCLUSION

We recommend progressively implementing Bayesian and dynamic analyses in all RCTs of immunotherapy to interpret and assess the credibility of frequentist results.

Low Programmed Death-Ligand 1-Expressing Subgroup Outcomes of First-Line Immune Checkpoint Inhibitors in Gastric or Esophageal Adenocarcinoma

PURPOSE

The US Food and Drug Administration has granted regulatory approval for the use of nivolumab-an immune checkpoint inhibitor (ICI)-in the first-line treatment of advanced gastric or esophageal adenocarcinoma (GEAC), regardless of programmed death-ligand 1 (PD-L1) expression. However, the efficacy of ICIs in low PD-L1-expressing tumors remains unclear.

MATERIALS AND METHODS

This study aims to reconstruct unreported Kaplan-Meier (KM) plots of PD-L1 combined positive score (CPS) subgroups of randomized phase III trials comparing the addition of ICIs with conventional chemotherapy in the first-line treatment of GEAC. A graphical reconstructive algorithm was adopted to estimate time-to-event outcomes from reported overall survival and progression-free survival (OS and PFS) KM plots describing overall or subgroup cohorts. Using reconstructed time-to-event outcomes, KMSubtraction conducts bipartite matching of patients from the reported subgroup among the overall cohort. By excluding matched patients, KM plots and survival analyses of the unreported subgroups were retrieved.

RESULTS

CheckMate-649, KEYNOTE-062, and KEYNOTE-590 were included. Two PD-L1 subgroups were identified with data unreported in the primary manuscripts: PD-L1 CPS 1-4 from CheckMate-649 and PD-L1 CPS 1-9 from KEYNOTE-062. No significant differences in OS and PFS were demonstrated in ICI-chemotherapy combinations when compared with chemotherapy among CheckMate-649 PD-L1 CPS 1-4 (OS: hazard ratio [HR] = 0.950, 95% CI, 0.747 to 1.209, P = .678; PFS: HR = 0.958, 95% CI, 0.743 to 1.236, P = .743) and KEYNOTE-062 PD-L1 CPS 1-9 subgroups. In the KEYNOTE-062 PD-L1 CPS 1-9 subgroup, patients treated with pembrolizumab had an increased hazard of tumor progression (HR = 2.092, 95% CI, 1.661 to 2.635, P < .001).

CONCLUSION

Using KMSubtraction, data of PD-L1 subgroups previously unreported by primary manuscripts of pivotal clinical trials were retrieved. These data suggest the lack of benefit in the addition of ICI to chemotherapy in low PD-L1-expressing GEAC tumors.

Causal Considerations Can Inform the Interpretation of Surprising Associations in Medical Registries

An exploratory analysis of registry data from 2437 patients with advanced gastric cancer revealed a surprising association between astrological birth signs and overall survival (OS) with p = 0.01. After dichotomizing or changing the reference sign, p-values <0.05 were observed for several birth signs following adjustments for multiple comparisons. Bayesian models with moderately skeptical priors still pointed to these associations. A more plausible causal model, justified by contextual knowledge, revealed that these associations arose from the astrological sign association with seasonality. This case study illustrates how causal considerations can guide analyses through what would otherwise be a hopeless maze of statistical possibilities.

Impact of a non-constant baseline hazard on detection of time-dependent treatment effects: a simulation study

Background

Non-proportional hazards are common with time-to-event data but the majority of randomised clinical trials (RCTs) are designed and analysed using approaches which assume the treatment effect follows proportional hazards (PH). Recent advances in oncology treatments have identified two forms of non-PH of particular importance - a time lag until treatment becomes effective, and an early effect of treatment that ceases after a period of time. In sample size calculations for treatment effects on time-to-event outcomes where information is based on the number of events rather than the number of participants, there is crucial importance in correct specification of the baseline hazard rate amongst other considerations. Under PH, the shape of the baseline hazard has no effect on the resultant power and magnitude of treatment effects using standard analytical approaches. However, in a non-PH context the appropriateness of analytical approaches can depend on the shape of the underlying hazard.

Methods

A simulation study was undertaken to assess the impact of clinically plausible non-constant baseline hazard rates on the power, magnitude and coverage of commonly utilized regression-based measures of treatment effect and tests of survival curve difference for these two forms of non-PH used in RCTs with time-to-event outcomes.

Results

In the presence of even mild departures from PH, the power, average treatment effect size and coverage were adversely affected. Depending on the nature of the non-proportionality, non-constant event rates could further exacerbate or somewhat ameliorate the losses in power, treatment effect magnitude and coverage observed. No single summary measure of treatment effect was able to adequately describe the full extent of a potentially time-limited treatment benefit whilst maintaining power at nominal levels.

Conclusions

Our results show the increased importance of considering plausible potentially non-constant event rates when non-proportionality of treatment effects could be anticipated. In planning clinical trials with the potential for non-PH, even modest departures from an assumed constant baseline hazard could appreciably impact the power to detect treatment effects depending on the nature of the non-PH. Comprehensive analysis plans may be required to accommodate the description of time-dependent treatment effects.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12874-021-01372-0.

Survival outcomes of patients with advanced melanoma from 2013 to 2017: Results of a nationwide population-based registry

BACKGROUND

The treatment landscape has completely changed for advanced melanoma. We report survival outcomes and the differential impact of prognostic factors over time in daily clinical practice.

METHODS

From a Dutch nationwide population-based registry, patients with advanced melanoma diagnosed from 2013 to 2017 were analysed (n = 3616). Because the proportional hazards assumption was violated, a multivariable Cox model restricted to the first 6 months and a multivariable landmark Cox model from 6 to 48 months were used to assess overall survival (OS) of cases without missing values. The 2017 cohort was excluded from this analysis because of the short follow-up time.

RESULTS

Median OS of the 2013 and 2016 cohort was 11.7 months (95% confidence interval [CI]: 10.4-13.5) and 17.7 months (95% CI: 14.9-19.8), respectively. Compared with the 2013 cohort, the 2016 cohort had superior survival in the Cox model from 0 to 6 months (hazard ratio [HR] = 0.55 [95% CI: 0.43-0.72]) and in the Cox model from 6 to 48 months (HR = 0.68 [95% CI: 0.57-0.83]). Elevated lactate dehydrogenase levels, distant metastases in ≥3 organ sites, brain and liver metastasis and Eastern Cooperative Oncology Group performance score of ≥1 had stronger association with inferior survival from 0 to 6 months than from 6 to 48 months. BRAF-mutated melanoma had superior survival in the first 6 months (HR = 0.50 [95% CI: 0.42-0.59]).

CONCLUSION(S)

Prognosis for advanced melanoma in the Netherlands has improved from 2013 to 2016. Prognostic importance of most evaluated factors was higher in the first 6 months after diagnosis. BRAF-mutated melanoma was only associated with superior survival in the first 6 months.

Non-proportional hazards in immuno-oncology: Is an old perspective needed?

A fundamental concept in two-arm non-parametric survival analysis is the comparison of observed versus expected numbers of events on one of the treatment arms (the choice of which arm is arbitrary), where the expectation is taken assuming that the true survival curves in the two arms are identical. This concept is at the heart of the counting-process theory that provides a rigorous basis for methods such as the log-rank test. It is natural, therefore, to maintain this perspective when extending the log-rank test to deal with non-proportional hazards, for example, by considering a weighted sum of the "observed - expected" terms, where larger weights are given to time periods where the hazard ratio is expected to favor the experimental treatment. In doing so, however, one may stumble across some rather subtle issues, related to difficulties in the interpretation of hazard ratios, that may lead to strange conclusions. An alternative approach is to view non-parametric survival comparisons as permutation tests. With this perspective, one can easily improve on the efficiency of the log-rank test, while thoroughly controlling the false positive rate. In particular, for the field of immuno-oncology, where researchers often anticipate a delayed treatment effect, sample sizes could be substantially reduced without loss of power.