Time to deterioration in cancer randomized clinical trials for patient-reported outcomes data: a systematic review

[Special issue PRO] A demonstration of estimands and sensitivity analyses for time-to-deterioration of patient reported outcomes

In oncology trials, health-related quality of life (HRQoL), specifically patient-reported symptom burden and functional status, can support the interpretation of survival endpoints, such as progression-free survival. However, applying time-to-event endpoints to patient-reported outcomes (PRO) data is challenging. For example, in time-to-deterioration analyses clinical events such as disease progression are common in many settings and are often handled through censoring the patient at the time of occurrence; however, disease progression and HRQoL are often related leading to informative censoring. Special consideration to the definition of events and intercurrent events (ICEs) is necessary. In this work, we demonstrate time-to-deterioration of PRO estimands and sensitivity analyses to answer research questions using composite, hypothetical, and treatment policy strategies applied to a single endpoint of disease-related symptoms. Multiple imputation methods under both the missing-at-random and missing-not-at-random assumptions are used as sensitivity analyses of primary estimands. Hazard ratios ranged from 0.52 to 0.66 over all the estimands and sensitivity analyses modeling a robust treatment effect favoring the treatment in time to disease symptom deterioration or death. Differences in the estimands include how people who experience disease progression or discontinue the randomized treatment due to AEs are accounted for in the analysis. We use the estimand framework to define interpretable and principled approaches for different time-to-deterioration research questions and provide practical recommendations. Reporting the proportions of patient events and patient censoring by reason helps understand the mechanisms that drive the results, allowing for optimal interpretation.

[Special issue PRO] Considering endpoints for comparative tolerability of cancer treatments using patient report given the estimand framework

Regulatory agencies are advancing the use of systematic approaches to collect patient experience data, including patient-reported outcomes (PROs), in cancer clinical trials to inform regulatory decision-making. Due in part to clinician under-reporting of symptomatic adverse events, there is a growing recognition that evaluation of cancer treatment tolerability should include the patient experience, both in terms of the overall side effect impact and symptomatic adverse events. Methodologies around implementation, analysis, and interpretation of "patient" reported tolerability are under development, and current approaches are largely descriptive. There is robust guidance for use of PROs as efficacy endpoints to compare cancer treatments, but it is unclear to what extent this can be relied-upon to develop tolerability endpoints. An important consideration when developing endpoints to compare tolerability between treatments is the linkage of trial design, objectives, and statistical analysis. Despite interest in and frequent collection of PRO data in oncology trials, heterogeneity in analyses and unclear PRO objectives mean that design, objectives, and analysis may not be aligned, posing substantial challenges for the interpretation of results. The recent ICH E9 (R1) estimand framework represents an opportunity to help address these challenges. Efforts to apply the estimand framework in the context of PROs have primarily focused on efficacy outcomes. In this paper, we discuss considerations for comparing the patient-reported tolerability of different treatments in an oncology trial context.

Time to deterioration of patient-reported outcomes as a surrogate of overall survival: a meta-analysis

BACKGROUND

Overall survival is the optimal marker of treatment efficacy in randomized clinical trials (RCTs) but can take considerable time to mature. Progression-free survival (PFS) has served as an early surrogate of overall survival but is imperfect. Time to deterioration in quality of life (QOL) measures could be a surrogate for overall survival.

METHODS

Phase 3 RCTs in solid malignancies that reported overall survival, PFS, and time to deterioration in QOL or physical function published between January 1, 2010, and June 30, 2022, were evaluated. Weighted regression analysis was used to assess the relationship between PFS, time to deterioration in QOL, and time to deterioration in physical function with overall survival. The coefficient of determination (R2) was used to quantify surrogacy.

RESULTS

In total, 138 phase 3 RCTs were included. Of these, 47 trials evaluated immune checkpoint inhibitors and 91 investigated non-immune checkpoint inhibitor agents. Time to deterioration in QOL (137 RCTs) and time to deterioration in physical function (75 RCTs) performed similarly to PFS as surrogates for overall survival (R2 = 0.18 vs R2 = 0.19 and R2 = 0.10 vs R2 = 0.09, respectively). For immune checkpoint inhibitor studies, time to deterioration in physical function had a higher association with overall survival than with PFS (R2 = 0.38 vs R2 = 0.19), and PFS and time to deterioration in physical function did not correlate with each other (R2 = 0). When time to deterioration in physical function and PFS are used together, the coefficient of determination increased (R2 = 0.57).

CONCLUSIONS

Time to deterioration in physical function appears to be an overall survival surrogate measure of particular importance for immune checkpoint inhibitor treatment efficacy. The combination of time to deterioration in physical function with PFS may enable better prediction of overall survival treatment benefit in RCTs of immune checkpoint inhibitors than either PFS or time to deterioration in physical function alone.

The estimand framework had implications in time to patient-reported outcomes deterioration analyses in cancer clinical trials

OBJECTIVES

To apply the estimand framework in time to deterioration (TTD) analysis of patient-reported outcomes (PROs), and identify the appropriate statistical methods to deal with intercurrent event (IEs) such as death.

STUDY DESIGN AND SETTING

Data from phase II randomized trial were used. We estimated TTD using European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-C30 questionnaire with death as the IE, by applying Kaplan-Meier (K.M.) estimator and Cox proportional hazards (PH) model. The Fine-Gray approach was explored, accounting for death as a competing risk. The estimands targeted by the aforementioned methods were defined.

RESULTS

We analyzed the data of 64 patients with available questionnaires at baseline. The most notable differences in TTD estimates were observed for deterioration in physical functioning: the hazard ratios were 0.44 [95% CI 0.22-0.90] and 0.62 [95% CI 0.36-1.07] by either ignoring death (31 events) or considering it as deterioration (58 events), respectively (Cox-PH model). When considering death as a competing event (Fine-Gray model), the sub-HRs was 0.51 [95% CI 0.26-1.01].

CONCLUSION

Depending on the proportion and distribution of deaths occurring before deterioration between arms, the Fine-Gray competing risks model should be considered rather than KM estimator and Cox PH model to reflect the patient's experience of the disease and treatment burden.

Patient-reported outcomes of maintenance rucaparib in patients with recurrent ovarian carcinoma in ARIEL3, a phase III, randomized, placebo-controlled trial

PURPOSE

To compare NFOSI-18 Disease Related Symptoms - Physical (DRSP), Total score, and side effect bother between maintenance rucaparib (600 mg twice daily) vs. placebo in the phase III ARIEL3 trial.

METHODS

ARIEL3 (NCT01968213) included patients with ovarian carcinoma who responded to second-line or later platinum-based chemotherapy. The NFOSI-18 DRS-P and Total scales were secondary endpoints. The NFOSI-18 contains a side effect impact item (GP5): "I am bothered by side effects of treatment." We compared treatment arms on change from baseline of DRS-P and Total scores using mixed models with repeated measures (MRMM). Time to first and confirmed deterioration of NFOSI-18 DRS-P and Total scales were analyzed using Cox regression. We also calculated the proportion of patients reporting moderate to high side effect bother on GP5.

RESULTS

In the intention-to-treat (ITT) cohort, mean change from baseline favored the placebo. Compared to placebo, rucaparib was associated with higher risk of deterioration [e.g., 4-point deteriorator definition hazard ratio (HR): 1.85; 95% CI: 1.46, 2.36; median time to first deterioration on DRSP: 1.9 vs. 7.0 months]. Confirmed deterioration results resembled those for first deterioration. Proportions of patients reporting moderate/high side effect bother on GP5 fluctuated around 20% across treatment cycles. Results in BRCA mutant and homologous recombination deficient cohorts were generally similar to those from the ITT cohort.

CONCLUSION

This placebo-controlled study in the maintenance therapy setting provides a unique view of the impact of PARP inhibition on the patient-reported outcomes that are commonly used in ovarian cancer clinical trials. Information regarding the adverse side effect impact of PARP inhibitors should be weighed against their clinical benefit.

Patient-reported outcomes in metastatic renal cell carcinoma trials using combinations versus sunitinib as first-line treatment

Over the past 5 years, several new immunotherapy treatments have been tested for metastatic renal cell carcinoma (mRCC). Clinical trials assessing combinations of different immunotherapies, or of an immunotherapy with a tyrosine kinase inhibitor (TKI), have reported improved clinical outcomes compared with the standard of care - that is, treatments using TKIs alone. However, to understand the holistic impact of new treatments on patients, physicians must also consider effects on health-related quality of life (HRQoL). As patient-reported outcome measures (PROMs) on HRQoL are often treated as a secondary outcome in clinical trials, their collection and reporting are non-standardized and, therefore, difficult to compare and interpret. However, results from six clinical trials indicate that two immunotherapy treatments overwhelmingly outperform sunitinib in HRQoL measurements: nivolumab plus cabozantinib (CheckMate 9ER) and atezolizumab plus bevacizumab (IMmotion151). An additional two treatments generally outperform sunitinib: nivolumab plus ipilimumab (CheckMate 214) and lenvatinib plus pembrolizumab (CLEAR). Of three studies that reported no difference from sunitinib, two suffered design flaws that might have obscured HRQoL benefits (JAVELIN Renal 101 and KEYNOTE-426). To ensure future HRQoL data are of the highest quality and comparable across trials, future studies should adopt best practices for the design, analysis and reporting of PROMs.

Cardiovascular Reasons for Access to a Tertiary Oncological Emergency Service: The CARILLON Study

BACKGROUND

The prevalence of acute cardiovascular diseases (CVDs) in cancer patients is steadily increasing and represents a significant reason for admission to the emergency department (ED).

METHODS

We conducted a prospective observational study, enrolling consecutive patients with cancer presenting to a tertiary oncological ED and consequently admitted to the oncology ward. Two groups of patients were identified based on main symptoms that lead to ED presentation: symptoms potentially related to CVD vs. symptoms potentially not related to CVD. The aims of the study were to describe the prevalence of symptoms potentially related to CVD in this specific setting and to evaluate the prevalence of definite CV diagnoses at discharge. Secondary endpoints were new intercurrent in-hospital CV events occurrence, length of stay in the oncology ward, and mid-term mortality for all-cause.

RESULTS

A total of 469 patients (51.8% female, median age 68.0 [59.1-76.3]) were enrolled. One hundred and eighty-six out of 469 (39.7%) presented to the ED with symptoms potentially related to CVD. Baseline characteristics were substantially similar between the two study groups. A discharge diagnosis of CVD was confirmed in 24/186 (12.9%) patients presenting with symptoms potentially related to CVD and in no patients presenting without symptoms potentially related to CVD (p < 0.01). During a median follow-up of 3.4 (1.2-6.5) months, 204 (43.5%) patients died (incidence rate of 10.1 per 100 person/months). No differences were found between study groups in terms of all-cause mortality (hazard ratio [HR]: 0.85, 95% confidence interval [CI] 0.64-1.12), new in-hospital CV events (HR: 1.03, 95% CI 0.77-1.37), and length of stay (p = 0.57).

CONCLUSIONS

In a contemporary cohort of cancer patients presenting to a tertiary oncological ED and admitted to an oncology ward, symptoms potentially related to CVD were present in around 40% of patients, but only a minority were actually diagnosed with an acute CVD.

Improvement of postoperative quality of life in patients with esophageal squamous cell carcinoma: does tea consumption have a role?

Background

To investigate the effect of tea consumption on the improvement of postoperative quality of life in male patients with esophageal squamous cell carcinoma (ESCC).

Methods

The quality of life information of 290 male patients with ESCC was collected. The time to deterioration and the number of events in each area of quality of life was calculated by time-to-deterioration (TTD) model. The association between postoperative tea drinking and postoperative quality of life in male ESCC patients was investigated using the Cox proportional risk model.

Results

Postoperative tea-drinking patients experienced delayed TTD in multiple domains, including general health, physical, role, emotional, and cognitive function, fatigue, nausea and vomiting, dyspnea, loss of appetite, constipation, diarrhea, eating problems, difficulty swallowing, choking while swallowing saliva, dry mouth, taste difficulties, coughing, and speech problems. The multivariate Cox regression analysis showed that drinking tea after surgery improved quality of life, including physical function (HR = 0.722, 95% CI: 0.559-0.933), role function (HR = 0.740, 95% CI: 0.557-0.983), eating problems (HR = 0.718, 95% CI: 0.537-0.960), odynophagia (HR = 0.682, 95% CI: 0.492-0.945), trouble swallowing saliva (HR = 0.624, 95% CI: 0.444-0.877), coughing (HR = 0.627, 95% CI: 0.442-0.889) and speech problems (HR = 0.631, 95% CI: 0.441-0.903). Furthermore, the improvement was more significant in patients who drank tea before surgery and continued to drink tea after surgery.

Conclusions

Postoperative tea drinking had a positive effect on delay in clinical deterioration and improvements in multiple functions and symptoms associated with ESCC in men.

Health-Related Quality of Life Outcomes in Patients with Resected Epidermal Growth Factor Receptor-Mutated Non-Small Cell Lung Cancer Who Received Adjuvant Osimertinib in the Phase III ADAURA Trial

PURPOSE

In the phase III ADAURA trial, adjuvant treatment with osimertinib versus placebo, with/without prior adjuvant chemotherapy, resulted in a statistically significant and clinically meaningful disease-free survival benefit in completely resected stage IB-IIIA EGFR-mutated (EGFRm) non-small cell lung cancer (NSCLC). We report health-related quality of life (HRQoL) outcomes from ADAURA.

PATIENTS AND METHODS

Patients randomized 1:1 received oral osimertinib 80 mg or placebo for 3 years or until recurrence/discontinuation. HRQoL (secondary endpoint) was measured using the Short Form-36 (SF-36) health survey at baseline, 12, and 24 weeks, then every 24 weeks until recurrence or treatment completion/discontinuation. Exploratory analyses of SF-36 score changes from baseline until week 96 and time to deterioration (TTD) were performed in the overall population (stage IB-IIIA; N = 682). Clinically meaningful changes were defined using the SF-36 manual.

RESULTS

Baseline physical/mental component summary (PCS/MCS) scores were comparable between osimertinib and placebo (range, 46-47) and maintained to Week 96, with no clinically meaningful differences between arms; difference in adjusted least squares (LS) mean [95% confidence intervals (CI), -1.18 (-2.02 to -0.34) and -1.34 (-2.40 to -0.28), for PCS and MCS, respectively. There were no differences between arms for TTD of PCS and MCS; HR, 1.17 (95% CI, 0.82-1.67) and HR, 0.98 (95% CI, 0.70-1.39), respectively.

CONCLUSIONS

HRQoL was maintained with adjuvant osimertinib in patients with stage IB-IIIA EGFRm NSCLC, who were disease-free after complete resection, with no clinically meaningful differences versus placebo, further supporting adjuvant osimertinib as a new treatment in this setting. See related commentary by Patil and Bunn, p. 2204.

%TTD and %TUDD: New SAS macro programs to calculate the survival data of the time to deterioration for patient-reported outcomes data in oncology

BACKGROUND AND OBJECTIVE

Longitudinal analysis of patient-reported outcome (PRO) data remains challenging, as no standardization of statistical methods has been proposed, making comparison of PRO results between clinical trials difficult. In this context, the time to deterioration approach has recently been proposed and is regularly used as a modality of longitudinal PRO analysis in oncology.

METHODS

Two new SAS macro programs were developed, %TTD and %TUDD, which implement longitudinal analysis of PRO data according to the time to deterioration approach. These programs implement the recommended deterioration definitions. We described the programs with their different functionalities.

RESULTS

The %TTD macro calculates the time to first or transient deterioration, and the %TUDD macro calculates the time until definitive deterioration. These macros allow to obtain the survival variables from the time to deterioration approach. We illustrate our programs by presenting different applications on the randomized phase II AFUGEM GERCOR clinical trial.

CONCLUSION

The implementation of the deterioration definitions in SAS software allows the dissemination of this approach, in order to move toward the goal of standardization of longitudinal PRO analysis in oncology clinical trials.

Time to deterioration of patient-reported outcomes in non-small cell lung cancer: exploring different definitions

Purpose

The clinical relevance of different time-to-deterioration (TTD) definitions for patient-reported outcomes were explored.

Methods

TTD definitions differing by reference score and deterioration event were used to analyse data from the phase 3 FLAURA trial of first-line osimertinib versus erlotinib or gefitinib in patients with EGFR-mutated advanced non-small cell lung cancer. Pre-specified key symptoms were fatigue, appetite loss, cough, chest pain and dyspnoea, scored using the European Organisation for Research and Treatment of Cancer QLQ-C30 and QLQ-LC13 questionnaires (≥ 10-point difference = clinically relevant).

Results

No significant treatment differences in TTD (distributions) were observed using definitions based on transient or definitive deterioration alone. TTD definitions based on definitive, sustained deterioration, with death not included as an event, yielded a significant treatment difference for dyspnoea (hazard ratio [HR] 0.71; P = 0.034) when baseline was the reference, and for cough (HR 0.70; P = 0.009) and dyspnoea (HR 0.71; P = 0.004) when best previous score was the reference. With death included as an event, treatment differences were significant for dyspnoea (HR 0.70; P = 0.025) when baseline was the reference, and for cough (HR 0.70; P = 0.011), dyspnoea (HR 0.71; P = 0.003) and chest pain (HR 0.71; P = 0.038) when best previous score was the reference. Irrespective of definition, TTD for appetite loss and fatigue did not differ significantly between arms.

Conclusion

This exploratory work showed that different TTD definitions yield different magnitudes of treatment difference, highlighting the importance of pre-specifying TTD definitions upfront in clinical trials.

Clinical trial registration

ClinicalTrials.gov NCT02296125.

Supplementary Information

The online version contains supplementary material available at 10.1007/s11136-022-03088-0.

Effect of Regorafenib in Delaying Definitive Deterioration in Health-Related Quality of Life in Patients with Advanced Cancer of Three Different Tumor Types

Purpose

The efficacy and safety of regorafenib have been demonstrated in phase 3 trials for multiple tumor types, including metastatic colorectal cancer (mCRC) (CORRECT [NCT01103323]; CONCUR [NCT01584830]), advanced gastrointestinal stromal tumor (GIST) (GRID [NCT01271712]), and hepatocellular carcinoma (HCC) (RESORCE [NCT01774344]). The objective of this post hoc exploratory analysis was to explore the impact of regorafenib on delaying health-related quality of life (HRQOL) deterioration across these tumor types.

Patients and Methods

HRQOL data (assessed with EORTC QLQ-C30 and EQ-5D questionnaires) were pooled for all trials to determine time until definitive deterioration (TUDD), defined as the patient’s first minimal clinically important deterioration in HRQOL score from baseline that does not resolve, using stratified Kaplan–Meier estimators and Cox proportional hazards models adjusted for relevant trial, cancer type, and baseline covariates. Additional analyses based on cancer type were conducted by pooling mCRC trials (CORRECT and CONCUR) and pooling the two mCRC trials with the HCC trial (RESORCE).

Results

A total of 1699 patients with HRQOL data were pooled across the four trials. The results showed that regorafenib significantly delayed TUDD compared with placebo across all three tumor types. Median time to deterioration across the five scales ranged from 16.3 to 24.1 weeks for regorafenib and 8.6 to 12.1 weeks for placebo. The results from the individual studies, the pooled mCRC trials, and the pooled mCRC and HCC trials were similar to the overall pooled results.

Conclusion

A pooled analysis of four phase 3 trials demonstrated that regorafenib delayed a clinically relevant exploratory endpoint, defined as TUDD, compared with placebo across three different tumor types (mCRC, GIST, and HCC), which supports a novel benefit of the impact of regorafenib with respect to patients with these three types of cancers by allowing initial declines in HRQOL to resolve and patients the opportunity to continue treatment.

Time to health-related quality of life improvement analysis was developed to enhance evaluation of modern anticancer therapies

OBJECTIVES

Major advances have recently been made in the treatments of cancer, which now also have the potential to improve patients' health-related quality of life (HRQOL). We propose the time to HRQOL improvement (TTI) and the time to sustained HRQOL improvement (TTSI) as potentially important cancer outcomes to be used in longitudinal HRQOL analyses.

STUDY DESIGN AND SETTING

As proof of principle, we defined TTI and TTSI, using the Fine-Gray model to include competing risks in estimates, in a case study in real life of a cohort of newly diagnosed patients with cancer receiving a targeted therapy. HRQOL was evaluated before and during therapy with six assessments over a 24-month period, using the well-validated European Organization for Research and Treatment of Cancer Quality of Life Questionnaire-Core 30.

RESULTS

For each assessed HRQOL domain, we assessed TTI and TTSI and estimated the cumulative incidence of patients' clinically meaningful improvements, also accounting for the occurrence of competing events.

CONCLUSION

TTI and TTSI are potentially important outcomes in the era of modern anticancer therapies. The analysis of TTI and TTSI by competing risks approach will further add to the statistical methods that can be used to inform on the impact of cancer therapies on patients' HRQOL.

Evaluation of Cognitive Function in Trials Testing New-Generation Hormonal Therapy in Patients with Prostate Cancer: A Systematic Review

Simple Summary

In patients with prostate cancer, the use of new-generation hormonal therapy, added to androgen deprivation therapy, requires careful evaluation of cognitive function. The aim of this systematic review is to describe the evidence about cognitive function in randomized trials testing new-generation hormonal therapy (abiraterone, enzalutamide, apalutamide, darolutamide). For each trial, we assessed the availability of both investigator-assessed cognitive impairment and disorders and patient-reported evaluation of cognitive function. Out of 19 trials, the investigator-based evaluation of cognitive impairment was available in seven (36.8%), while patient-reported evaluation of cognitive function results was presented only in one trial (5.3%). This analysis shows that, despite cognitive deterioration could be relevant in patients with prostate cancer, clinical development of new-generation hormonal drugs has not included a systematic evaluation of cognitive function.

Abstract

In patients with prostate cancer, earlier use and longer duration of new-generation hormonal therapy (NGHT), added to androgen deprivation therapy, requires careful evaluation of cognitive function. The aim of this systematic review is to describe the evidence about cognitive function in all the randomized trials (RCTs) testing NGHT (abiraterone, enzalutamide, apalutamide, darolutamide). We assessed the availability of both investigator-assessed cognitive impairment and disorders and patient-reported evaluation of cognitive function. Nineteen RCTs (17,617 patients) were included. The investigator-based evaluation of cognitive impairment was available in seven RCTs (36.8%). In total, 19/19 RCTs (100%) included patient-reported outcomes (PROs) collection, but PRO tools adopted allowed evaluation of cognitive function in two RCTs (10.5%). Among them, PRO-based cognitive function results were presented only in one RCT (5.3%): in ENZAMET, mean changes from baseline were worse with enzalutamide than with placebo, but deterioration-free survival favored enzalutamide. Despite cognitive deterioration could be relevant, clinical development of NGHT has not included a systematic evaluation of cognitive function. Assessment by investigators is at risk of underreporting, and commonly used PROs do not allow proper cognitive function analysis. Furthermore, the methodology of analysis can jeopardize the interpretation of results. Although direct comparisons are scanty, there could be differences between different NGHTs.

Longitudinal analysis of health-related quality of life in cancer clinical trials: methods and interpretation of results

PURPOSE

Health-related quality of life (HRQoL) is assessed by self-administered questionnaires throughout the care process. Classically, two longitudinal statistical approaches were mainly used to study HRQoL: linear mixed models (LMM) or time-to-event models for time to deterioration/time until definitive deterioration (TTD/TUDD). Recently, an alternative strategy based on generalized linear mixed models for categorical data has also been proposed: the longitudinal partial credit model (LPCM). The objective of this article is to evaluate these methods and to propose recommendations to standardize longitudinal analysis of HRQoL data in cancer clinical trials.

METHODS

The three methods are first described and compared through statistical, methodological, and practical arguments, then applied on real HRQoL data from clinical cancer trials or published prospective databases. In total, seven French studies from a collaborating group were selected with longitudinal collection of QLQ-C30. Longitudinal analyses were performed with the three approaches using SAS, Stata and R software.

RESULTS

We observed concordant results between LMM and LPCM. However, discordant results were observed when we considered the TTD/TUDD approach compared to the two previous methods. According to methodological and practical arguments discussed, the approaches seem to provide additional information and complementary interpretations. LMM and LPCM are the most powerful methods on simulated data, while the TTD/TUDD approach gives more clinically understandable results. Finally, for single-item scales, LPCM is more appropriate.

CONCLUSION

These results pledge for the recommendation to use of both the LMM and TTD/TUDD longitudinal methods, except for single-item scales, establishing them as the consensual methods for publications reporting HRQoL.