Trial-level Surrogacy Research Articles

Assessing surrogacy using restricted mean survival time ratio for overall survival in non-small cell lung cancer immunotherapy studies.

Proportional hazards (PH) assumption is often violated in cancer immunotherapy studies. Restricted mean survival time (RMST) ratio is a valid metric to quantify the size of treatment effect when non-proportional hazard (NPH) is present. This study investigated the use of RMST ratio and hazard ratio (HR) in studying progression-free survival (PFS) as a surrogate endpoint for overall survival (OS) in non-small cell lung cancer immunotherapy trials. Trial level data were collected from 14 phase III trials published between 2012 and 2018. A weighted least-square regression (WLSR) was performed to evaluate the trial-level surrogacy. Surrogacy was evaluated via the association between RMST ratios for PFS and OS and between HRs for PFS and OS. Using data extracted from published articles, low to moderate correlation (0.49) between PFS and OS was observed for HR while low correlation (0.35) was observed for RMST ratio. When trials violating PH in PFS were included, more consistent correlations for both HR (0.43) and RMST ratio (0.44) were observed. In summary, the strength of PFS surrogacy for OS depends on whether HR or RMST ratio are chosen. RMST ratio and additional sensitivity analysis should be considered in addition to HR.

Abstract 10351: Validity and Relevance of Clinical Worsening as a Composite Endpoint in Pulmonary Arterial Hypertension Trials

Background: Clinical worsening (CW) is commonly used as an endpoint in pulmonary arterial hypertension (PAH) trials. The use of composite endpoints is based on the assumption that each component is interchangeable. Objectives: We aimed to assess the trial-level surrogacy of CW for predicting overall mortality (considered adequate when coefficient of determination R 2 trial >0.70) in PAH trials. We also assessed whether the various CW components were similar in terms of: 1) frequency of occurrence; 2) treatment-related relative risk (RR) reduction and; 3) importance to patients. Methods: We searched MEDLINE, Embase, and the Cochrane Library (01/1990-12/2020) for RCTs evaluating the effects of PAH treatment on CW. The R 2 trial between the RR for CW and mortality (including after censoring) was analysed by regression analysis. The frequency of occurrence of the CW components, as well as the RR reduction of the CW components and those considered of critical, major and mild-to-moderate importance determined through a web-based survey on 257 PAH patients were assessed. Results: From 4,778 references, we included 35 studies (7,744 pts). The effects of PAH-specific therapies on CW modestly correlated with mortality (R 2 trial 0.32). Most of the CW defining-events were PAH-related hospitalizations (30.5%) and symptomatic progressions (48.0%), whereas deaths (14.1%), treatment escalations (6.8%) and transplantations or atrial septostomies (0.6%) were unusual. The RR across components differed by >0.4 in 20/35 studies (57%). Furthermore, the composite endpoints combined components of both critical to moderate importance to patients in 31/35 studies (89%). The pooled treatment effects only became significant when endpoints of major (RR 0.67;95%CI 0.60-0.74;p<0.01) and moderate (RR 0.61;0.52-0.70;p<0.01) importance to patients were added to critical outcomes (RR 1.02;95%CI 0.78-1.32;p=0.91). Consistently, 8/12 (67%) trials with documented decreases in CW were no longer significant when outcomes of moderate importance to patients were excluded. Conclusion: CW is not a surrogate outcome for mortality in PAH trials. Moreover, components of CW largely vary in frequency, response to therapy and importance to patients and are thus not interchangeable.

Progression-free survival assessed per immune-related or conventional response criteria, which is the better surrogate endpoint for overall survival in trials of immune-checkpoint inhibitors in lung cancer: A systematic review and meta-analysis.

Progression‐free survival (PFS) has been used as a surrogate endpoint for overall survival (OS) in lung cancer trials. The pattern of response to immune‐checkpoint inhibitors (ICIs) differs from that to conventional chemotherapy, so immune‐related response evaluation criteria were proposed. This study aims at determining which PFS measure, PFS assessed per immune‐related response evaluation criteria (iPFS), or conventional criteria (cPFS), is the better surrogate endpoint for OS in trials of ICIs in lung cancer. We selected clinical trials in lung cancer that administered ICIs to at least one arm and reported both median OS and median PFS from PubMed, Embase, and The Cochrane Library. We compared the correlation between treatment effect (hazard ratio) on OS and cPFS or iPFS and the correlation between median OS and median cPFS or iPFS using weighted linear regression at trial level. We analyzed 78 ICI arms (13,438 patients) from 54 studies, including 66 arms with cPFS, seven arms with iPFS, and five arms with both kinds of PFS. We demonstrated an excellent correlation between treatment effect (hazard ratio) on OS and iPFS (R WLS 2 = 0.91), while the correlation was moderate for cPFS (R WLS 2 = 0.38). Similarly, the correlation between median OS and median iPFS was also strong (R WLS 2 ranging from 0.86 to 0.96) across different phases of trials and different types of lung cancer, ICI, and treatment modalities, while it was much weaker for median cPFS (R WLS 2 ranging from 0.28 to 0.88). In conclusion, iPFS provides better trial‐level surrogacy for OS than cPFS in trials of ICIs in lung cancer.

Reevaluating Disease-Free Survival as an Endpoint vs Overall Survival in Stage III Adjuvant Colon Cancer Trials.

Disease-free survival (DFS) with a 3-year median follow-up (3-year DFS) was validated as a surrogate for overall survival (OS) with a 5-year median follow-up (5-year OS) in adjuvant chemotherapy colon cancer (CC) trials. Recent data show further improvements in OS and survival after recurrence in patients who received adjuvant FOLFOX. Hence, reevaluation of the association between DFS and OS and determination of the optimal follow-up duration of OS to aid its utility in future adjuvant trials are needed. Individual patient data from 9 randomized studies conducted between 1998 and 2009 were included; 3 trials tested biologics. Trial-level surrogacy examining the correlation of treatment effect estimates of 3-year DFS with 5 to 6.5-year OS was evaluated using both linear regression (RWLS2) and Copula bivariate (RCopula2) models and reported with 95% confidence intervals (CIs). For R2, a value closer to 1 indicates a stronger correlation. Data from a total of 18 396 patients were analyzed (median age = 59 years; 54.0% male), with 54.1% having low-risk tumors (T1-3 and N1), 31.6% KRAS mutated, 12.3% BRAF mutated, and 12.4% microsatellite instability high or deficient mismatch repair tumors. Trial-level correlation between 3-year DFS and 5-year OS remained strong (RWLS2 = 0.82, 95% CI = 0.67 to 0.98; RCopula2 = 0.92, 95% CI = 0.83 to 1.00) and increased as the median follow-up of OS extended. Analyses limited to trials that tested biologics showed consistent results. Three-year DFS remains a validated surrogate endpoint for 5-year OS in adjuvant CC trials. The correlation was likely strengthened with 6 years of follow-up for OS.

Refining neoadjuvant therapy clinical trial design for muscle-invasive bladder cancer before cystectomy: a joint US Food and Drug Administration and Bladder Cancer Advocacy Network workshop.

The success of the use of novel therapies in the treatment of advanced urothelial carcinoma has contributed to growing interest in evaluating these therapies at earlier stages of the disease. However, trials evaluating these therapies in the neoadjuvant setting must have clearly defined study elements and appropriately selected end points to ensure the applicability of the trial and enable interpretation of the study results. To advance the development of rational trial design, a public workshop jointly sponsored by the US Food and Drug Administration and the Bladder Cancer Advocacy Network convened in August 2019. Clinicians, clinical trialists, radiologists, biostatisticians, patients, advocates and other stakeholders discussed key elements and end points when designing trials of neoadjuvant therapy for muscle-invasive bladder cancer (MIBC), identifying opportunities to refine eligibility, design and end points for neoadjuvant trials in MIBC. Although pathological complete response (pCR) is already being used as a co-primary end point, both individual-level and trial-level surrogacy for time-to-event end points, such as event-free survival or overall survival, remain incompletely characterized in MIBC. Additionally, use of pCR is limited by heterogeneity in pathological evaluation and the fact that the magnitude of pCR improvement that might translate into a meaningful clinical benefit remains unclear. Given existing knowledge gaps, capture of highly granular patient-related, tumour-related and treatment-related characteristics in the current generation of neoadjuvant MIBC trials will be critical to informing the design of future trials.

Radiographic progression-free survival as a surrogate endpoint of overall survival in men with metastatic castrate-resistant prostate cancer.

5057 Background: Radiographic progression-free survival (rPFS) is commonly used as a co-primary endpoint in randomized clinical trials in men with metastatic castrate-resistant prostate cancer (mCRPC). However, rPFS has not been established as a valid surrogate endpoint of overall survival (OS) in men with mCRPC. Here, we hypothesized that rPFS is a reliable surrogate for OS in mCRPC. We also explored whether PFS is a valid surrogate endpoint of OS at the aggregate trial level. Methods: We performed a systematic search of the literature encompassing the period January 2004-December 2020 using PubMed and clinical trials.gov to identify completed phase III trials in mCRPC post-docetaxel. Eligible trials had to be randomized phase III therapeutic trials that reported OS, PFS or rPFS. OS was measured from the date of random assignment to date of death from any cause or date of last follow-up. rPFS was defined as the time from random assignment to date of disease progression on CT and/or Tc bone scan per trial definition or death from any cause, whichever occurred first. PFS included PSA progression as a component of the composite endpoint. Trial level surrogacy was evaluated by fitting linear regression on the treatment effect of rPFS (or PFS) and OS (in other words, the weighted linear regression of the log(hazard ratio) of OS on the log(hazard ratio) of rPFS). It was pre-specified that rPFS would be considered a valid surrogate for OS if R² was 0·7 or higher. Results: We identified 33 in men with mCRPC post docetaxel approval. We assessed the association between PFS and OS in 29,456 patients from 30 trials. Overall, a moderate correlation was observed at the trial level between OS and PFS ( R2 = 0.46, 95 %CI = 0.20-0.68) in these trials. In 18 trials with 16,818 mCRPC patients where rPFS was considered as a key endpoint, a moderate correlation between the treatment effects on rPFS and OS was observed at the trial level ( R2= 0.65, 95% CI = 0.23-0.87). Conclusions: This meta-analysis demonstrates moderate correlation between treatment effects of rPFS and OS in patients with mCRPC. However, rPFS did not meet the pre-specified surrogacy threshold of 0.7. Clinical trial information: several.

A joint frailty-copula model for meta-analytic validation of failure time surrogate endpoints in clinical trials.

In a meta-analysis framework, the classical approach for the validation of time-to-event surrogate endpoint is based on a two-step analysis. This approach often raises estimation issues. Recently, we proposed a one-step validation approach based on a joint frailty model. This approach was quite time consuming, despite parallel computing, due to individual-level frailties used to take into account heterogeneity in the data at the individual level. We now propose an alternative one-step approach for evaluating surrogacy, using a joint frailty-copula model. The model includes two correlated random effects treatment-by-trial interaction and a shared random effect associated with the baseline risks. At the individual level, the joint survivor functions of time-to-event endpoints are linked using copula functions. We used splines for the baseline hazard functions. We estimated parameters and hazard function using a semiparametric penalized marginal likelihood method, considering various numerical integration methods. Both individual-level and trial-level surrogacy were evaluated using Kendall's tau and coefficient of determination. The performance of the estimators was evaluated using simulation studies. The model was applied to individual patient data meta-analyses in advanced ovarian cancer to assess progression-free survival as a surrogate for overall survival, as part of the evaluation of new therapy. The model showed good performance and was quite robust regarding the integration methods and data variation, regardless of the surrogacy evaluation criteria. Kendall's Tau was better estimated using the Clayton copula model compared to the joint frailty model. The proposed model reduces the convergence and model estimation issues encountered in the two-step approach.

Disease-free survival as a surrogate for overall survival in neoadjuvant trials of gastroesophageal adenocarcinoma: Pooled analysis of individual patient data from randomized controlled trials.

4533 Background: Disease-free survival (DFS) is an appealing surrogate endpoint for overall survival (OS) in trials on neoadjuvant or adjuvant cancer therapy, because it is available faster and with less follow-up effort. The aim of this study was to assess if DFS can be a valid surrogate endpoint for OS when comparing neoadjuvant treatment followed by surgery to surgery alone for gastroesophageal adenocarcinoma. Methods: Individual patient data (IPD) from eight randomized controlled trials (n = 1,126 patients) which compared neoadjuvant therapy followed by surgery with surgery alone for gastroesophageal adenocarcinoma were used for the analysis. Correlation between OS-time and DFS-time was calculated. Kaplan-Meier survival curves and corresponding hazard ratios (HRs) for treatment effects were separately determined for each trial. Subsequently, HRs were pooled in a meta-analysis using a random-effects model. An error-in-variables linear regression model was used to compare observed and predicted values. The minimum treatment effect on DFS necessary to predict a non-zero treatment effect on OS was estimated by calculating the surrogate threshold effect. Results: OS-time correlated strongly with DFS-time. HRs for OS and DFS were highly similar for all single trials. The meta-analysis yielded almost identical overall HRs for treatment effects on OS and DFS. The determination coefficient for the association between HRs for OS and DFS was 0.912 (95% confidence interval 0.75-1.0), indicating a strong trial-level surrogacy between OS and DFS. The surrogate threshold effect was calculated at 0.79, indicating that a future trial yielding a hazard ratio for the treatment effect on DFS < 0.79 could be expected with a 95% probability to yield a hazard ratio for the treatment effect on OS < 1. Conclusions: DFS and OS strongly correlate both after neoadjuvant therapy followed by surgery and after surgery alone for gastroesophageal adenocarcinoma. Likewise, the treatment effects on the two endpoints are very similar. Consequently, DFS can be regarded an appropriate surrogate endpoint for OS in trials on neoadjuvant therapy for gastroesophageal adenocarcinoma.

Disease-free survival as a surrogate for overall survival in neoadjuvant trials of gastroesophageal adenocarcinoma: Pooled analysis of individual patient data from randomised controlled trials

IntroductionDisease-free survival (DFS) is increasingly being used as surrogate end-point for overall survival (OS) in cancer trials. So far, there has been no validation of the surrogacy of DFS for OS for neoadjuvant treatment of gastroesophageal adenocarcinoma. MethodsThe study uses individual patient data (IPD) from eight randomised controlled trials (RCTs) (n = 1126 patients) comparing neoadjuvant therapy followed by surgery with surgery alone for gastroesophageal adenocarcinoma. Correlation between OS time and DFS time was calculated to evaluate individual-level surrogacy. For each trial, survival curves using the Kaplan-Meier method were plotted and hazard ratios (HRs) on the treatment effects were calculated for OS and DFS separately. Those HRs were pooled in a random-effects meta-analysis. Observed HRs were compared with predicted HRs for OS using results from an error-in-variables linear regression model accounting for the uncertainty about the estimated effect. The strength of the association was quantified by the coefficient of determination to assess trial-level surrogacy. The surrogate threshold effect was calculated to determine the minimum treatment effect on DFS necessary to predict a non-zero treatment effect on OS. ResultsA strong correlation between OS time and DFS time was observed, indicating a high individual-level surrogacy. For all RCTs, estimated HRs for OS and DFS were highly similar. In the meta-analysis, the overall HR for OS was virtually identical to that for DFS. The estimated coefficient of determination r2 for the association between HRs for OS and DFS was 0.912 (95% confidence interval: 0.75–1.0), indicating a very good fit of the regression model and thus a strong trial-level surrogacy between OS and DFS. The surrogate threshold effect based on the regression analysis was 0.79. DiscussionBased on strong correlations between DFS and OS, as well as a strong correlation of the treatment effects of the two end-points in the error-in-variable regression, DFS seems an appropriate surrogate marker for OS in randomised trials of neoadjuvant chemotherapy or chemoradiotherapy for gastroesophageal adenocarcinoma.

Re-evaluating disease-free survival (DFS) as an endpoint versus overall survival (OS) in adjuvant colon cancer (CC) trials with chemotherapy +/- biologics: An updated surrogacy analysis based on 18,886 patients (pts) from the Accent database.

3502 Background: DFS with 3 years median follow-up (3yDFS) was validated as a surrogate for OS with 5 years median follow-up (5yOS) in adjuvant chemotherapy CC trials prior. Recent data showed improved survival after recurrence and OS, over time, in pts who received adjuvant FOLFOX. Hence, re-evaluation of the association between DFS and OS, as well as the optimal follow-up of OS to aid its utility in future adjuvant trials is needed. Methods: Individual patient data from 8 randomized adjuvant studies conducted from 1998-2009 were included; 3 trials tested anti-VEGF or anti-EGFR agents. Trial-level surrogacy examining the correlation of treatment effect estimates (i.e. hazard ratios) of 3yDFS and 5y to 8yOS was evaluated using both linear regression (R2WLS) and Copula bivariate (R2Copula) models. For the R2, a value closer to 1 indicates a stronger correlation. Prespecified criteria for surrogacy required either R2WLS or R2Copula ≥ 0.80 and neither < 0.7, with lower-bound 95% Confidence Interval (CI) > 0.60. The rank correlation coefficient (ρ) quantified the individual-level surrogacy. Results: Total of 18,886 pts were analyzed, with median age 60, 54% male, 83% stage III, 59% > 12 nodes examined. Median follow-up for survival ranged from 5 to 10 years across trials. Trial level correlation between 3yDFS and OS remained strong (R2WLS ≥0.74; R2Copula ≥ 0.89) and increased as the median follow-up of OS extended longer (see table). Analyses limited to stage III pts and/or trials tested biologics showed consistent results. Conclusions: 3yDFS remains a validated surrogate endpoint for 5yOS in adjuvant trials in CC pts per prespecified criteria. The correlation was strengthened with more than 6 years of follow-up for OS. [Table: see text]

One-step validation method for surrogate endpoints using data from multiple randomized cancer clinical trials with failure-time endpoints.

A surrogate endpoint can be used instead of the most relevant clinical endpoint to assess the efficiency of a new treatment. Before being used, a surrogate endpoint must be validated based on appropriate methods. Numerous validation approaches have been proposed with the most popular used in a context of meta-analysis, based on a two-step analysis strategy. For two failure-time endpoints, two association measurements are usually used, Kendall's τ at the individual level and the adjusted coefficient of determination ( ) at the trial level. However, is not always available due to model estimation constraints. We propose a one-step validation approach based on a joint frailty model, including both individual-level and trial-level random effects. Parameters have been estimated using a semiparametric penalized marginal log-likelihood method, and various numerical integration approaches were considered. Both individual- and trial-level surrogacy were evaluated using a new definition of Kendall's τ and the coefficient of determination. Estimators' performances were evaluated using simulation studies and satisfactory results were found. The model was applied to individual patient data meta-analyses in gastric cancer to assess disease-free survival as a surrogate for overall survival, as part of the evaluation of adjuvant therapy.

A reflection on the causal interpretation of individual-level surrogacy

ABSTRACTAt the beginning of the 21st century, a new paradigm was introduced for the evaluation of surrogate endpoints based on meta-analysis. In this paradigm, the putative surrogate is assessed at two different levels, the so-called, trial and individual level. Trial level surrogacy is defined as the association between the expected causal treatment effects across different trials populations, whereas the individual level is defined as the association between the surrogate and true endpoints, after adjusting by trial and treatment. It has been argued that the individual level surrogacy does not have a causal interpretation and, consequently, it is a poor metric of surrogacy. In the present work, an alternative definition of individual level surrogacy is introduced based on individual causal treatment effects. In addition, using the maximum entropy principle, a direct link between the individual level surrogacy, as defined in the meta-analytic approach, and the newly proposed definition is established. This new perspective sets the individual level surrogacy in a more coherent framework with respect to the trial level and bridges the two main schools of thought in this domain, namely, the causal inference and meta-analytic schools.

Event-Free Survival As a Surrogate Endpoint for Overall Survival in Previously Untreated Acute Myeloid Leukemia: An Individual Patient-Level Analysis of Multiple Randomized Trials (Alliance A151614)

Background: Overall survival (OS) remains the definitive primary efficacy endpoint to evaluate previously untreated acute myeloid leukemia (AML) therapies, but it requires prolonged follow-up. An earlier endpoint assessed post-treatment would expedite clinical trial conduct and accelerate patient access to effective new therapies. Our objective was to formally evaluate event-free survival (EFS) as a surrogate endpoint for OS in untreated AML.Methods: Individual patient data were analyzed from 2,475 patients (pts) from 4 multicenter, randomized controlled phase III trials of active treatment in previously untreated AML using anthracycline and cytarabine induction chemotherapy as the concurrent control (CALGB 10201, n=506, enrollment period 2003-2006, age 60-88 years (y); CALGB 10603, n=717, enrollment period 2008-2015, age 18-60 y, FLT3-mutated pts only; SWOG 0106, n=595, enrollment period 2004-2009, age 18-60 y; ECOG-ACRIN 1900, n=657, enrollment period 2002-2008, age 17-60 y). Individual patient-level surrogacy examines the association between the individual patients' EFS and OS time after adjusting for treatment effect, and was assessed using the copula bivariate survival model (Kendall's tau). Trial-level surrogacy measures how precisely the treatment effect on OS can be predicted on the basis of observed treatment effect on EFS, and was evaluated using both linear regression (R2WLS) weighted by trial size and Copula bivariate (R2Copula) models. Pre-specified criteria for surrogacy required either R2WLS or R2Copula ≥0.80, neither below 0.7, with either lower bound 95% Confidence Interval (CI) >0.60. Sensitivity analyses were conducted using different EFS definitions (Table 1).Results: With a median follow-up of 50.2 months for the 896 patients still alive, the median OS and EFS across all four trials were 20.9 months (95% CI: 19.0-22.7) and 5.6 months (95% CI: 4.5-6.4), respectively. Trial-level surrogacy for EFS was strong (R2WLS=0.79; R2Copula=0.89), indicating a high correlation of treatment effect between EFS and OS. At the individual patient-level, however, EFS showed weak association with OS (tau= 0.52), compared to the strength of trial-level surrogacy. The discrepancy between patient-level EFS and OS was greatest among patients who did not achieve a CR, followed by those who achieved a CR but relapsed (Figure 1).Sensitivity analysis on alternative EFS definitions showed that the trial-level surrogacy was similar, but individual patient-level surrogacy varied across different EFS definitions (Table 1). This is consistent with what we previously reported (ASH 2016): EFS estimates differed considerably based on the definition of induction failure (IF) in a single arm setting, but this had minimal impact on the estimation of the treatment effect using EFS in randomized trials. In addition, when considering only relapse and death as events (definition 4), both individual patient- and trial-level correlations were high.Conclusions: Correlation between EFS and OS was impacted by patients not achieving CR during induction. Despite the lack of patient-level correlation, a strong correlation between hazard ratios for treatment effects was observed between EFS and OS on the trial level. Hence, it remains debatable whether EFS represents a clinical benefit in itself for patient with untreated AML considering the strong correlation in treatment effects. Further validation is needed due to the small number of trials included and the heterogeneity across trials.Acknowledgment: We gratefully acknowledge the important contributions of the late Dr. Stephen H. Petersdorf, SWOG S0106 Study Chair.Support: U10CA180821, U10CA180882, U10CA180794, U10CA180820, U10CA180888;Clinicaltrials.gov Identifiers: NCT00085124 (10201), NCT00651261 (10603), NCT01253070 (11001), NCT00085709 (SWOG S0106), and NCT00049517 (ECOG-ACRIN E1900) [Display omitted] DisclosuresUy:Curis: Consultancy; GlycoMimetics: Consultancy. Larson:Pfizer: Consultancy, Research Funding; BristolMyers Squibb: Consultancy, Research Funding; Novartis: Consultancy, Research Funding; Ariad/Takeda: Consultancy, Research Funding.

PCN59 - ASSOCIATION BETWEEN HAZARD RATIOS OF SURROGATE TIME-TO-EVENT ENDPOINTS AND OVERALL SURVIVAL IN ADVANCED/METASTATIC CANCERS

Objectives: Surrogate endpoints can support early access to novel therapies. In trial-level endpoint validation studies, the association between treatment effect e.g., the hazard ratio (HR) on both the surrogate and hard endpoint can be estimated. We aimed to review studies reporting an association between HRs of surrogate time-to-event endpoints and overall survival (OS) in advanced/metastatic cancers. Methods: A systematic review was conducted using Medline and Embase. We included full reports assessing the association between HRs of surrogate time-to-event endpoints and OS in advanced/metastatic cancer indications. The following information was extracted: study characteristics, association measure, use of weighted analyses, logarithmic transformation of HRs, use of multivariate analysis, evaluation of crossover impact, use of IQWiG framework, estimating surrogate threshold effect (STE), and reported results and/or regression equations. Results: Forty-five studies were included. Retrieved studies were conducted in 16 different cancer indications. Different methods were used to assess associations, including Spearman's/Pearson’s correlation coefficient and linear regression analysis. Weighted analyses, logarithmic transformation of HRs and multivariate analysis were implemented in 35, 26 and 10 studies, respectively. Few studies assessed the crossover impact on the association (8 studies) and implemented IQWiG framework and STE assessment (11 studies and 3 studies, respectively). Detailed results are extracted, summarized and will be presented. Conclusions: There is inconsistency in conducting/reporting of trial-level endpoint validation studies in advanced/metastatic cancers. Future studies would benefit from building a structured data analysis checklist. Also, trial-level surrogacy was not assessed in all advanced/metastatic cancers and the strength of association varied across indications. The generalizability of results from one indication to another is limited.

Progression-Free Survival as a Surrogate End Point for Overall Survival in First-Line Diffuse Large B-Cell Lymphoma: An Individual Patient-Level Analysis of Multiple Randomized Trials (SEAL).

Purpose Overall survival (OS) is the definitive and best-established primary efficacy end point to evaluate diffuse large B-cell lymphoma (DLBCL) therapies, but it requires prolonged follow-up. An earlier end point assessed post-treatment would expedite clinical trial conduct and accelerate patient access to effective new therapies. Our objective was to formally evaluate progression-free survival (PFS) and PFS at 24 months (PFS24) as surrogate end points for OS in first-line DLBCL. Patients and Methods Individual patient data were analyzed from 7,507 patients from 13 multicenter randomized controlled trials of active treatment in previously untreated DLBCL, published after 2002, with sufficient PFS data to predict treatment effects on OS. Trial-level surrogacy examining the correlation of treatment effect estimates of PFS/PFS24 and OS was evaluated using both linear regression ( R2WLS) and Copula bivariable ( R2Copula) models. Prespecified criteria for surrogacy required either R2WLS or R2Copula ≥ 0.80 and neither < 0.7, with lower-bound 95% CI > 0.60. Results Trial-level surrogacy for PFS was strong ( R2WLS = 0.83; R2Copula = 0.85) and met the predefined criteria for surrogacy. At the patient level, PFS strongly correlated with OS. The surrogate threshold effect had a hazard ratio of 0.89. Surrogacy was consistent across comparisons with or without rituximab and with rituximab maintenance trials. Trial-level surrogacy for PFS24 was relatively strong ( R2WLS = 0.77; R2Copula = 0.78) but did not meet prespecified criteria. At the patient level, PFS24 significantly correlated with OS. The surrogate threshold effect had an odds ratio of 1.51. Conclusion This large pooled analysis of individual patient data supports PFS as a surrogate end point for OS in future randomized controlled trials evaluating chemoimmunotherapy in DLBCL. Use of this end point may expedite therapeutic development with the intent of bringing novel therapies to this patient population years before OS results are mature.

Neoadjuvant rectal score as individual-level surrogate for disease-free survival in rectal cancer in the CAO/ARO/AIO-04 randomized phase III trial

BackgroundSurrogate end points in rectal cancer after preoperative chemoradiation are lacking as their statistical validation poses major challenges, including confirmation based on large phase III trials. We examined the prognostic role and individual-level surrogacy of neoadjuvant rectal (NAR) score that incorporates weighted cT, ypT and ypN categories for disease-free survival (DFS) in 1191 patients with rectal carcinoma treated within the CAO/ARO/AIO-04 phase III trial. Patients and methodsCox regression models adjusted for treatment arm, resection status, and NAR score were used in multivariable analysis. The four Prentice criteria (PC1–4) were used to assess individual-level surrogacy of NAR for DFS. ResultsAfter a median follow-up of 50months, the addition of oxaliplatin to fluorouracil-based chemoradiotherapy (CRT) significantly improved 3-year DFS [75.9% (95% confidence interval [CI] 72.30% to 79.50%) versus 71.3% (95% CI 67.60% to 74.90%); P=0.034; PC 1) and resulted in a shift toward lower NAR groups (P=0.034, PC 2) compared with fluorouracil-only CRT. The 3-year DFS was 91.7% (95% CI 88.2% to 95.2%), 81.8% (95% CI 78.4% to 85.1%), and 58.1% (95% CI 52.4% to 63.9%) for low, intermediate, and high NAR score, respectively (P<0.001; PC 3). NAR score remained an independent prognostic factor for DFS [low versus high NAR: hazard ratio (HR) 4.670; 95% CI 3.106–7.020; P<0.001; low versus intermediate NAR: HR 1.971; 95% CI 1.303–2.98; P=0.001] in multivariable analysis. Notwithstanding the inherent methodological difficulty in interpretation of PC 4 to establish surrogacy, the treatment effect on DFS was captured by NAR, supporting satisfaction of individual-level PC 4. ConclusionOur study validates the prognostic role and individual-level surrogacy of NAR score for DFS within a large randomized phase III trial. NAR score could help oncologists to speed up response-adapted therapeutic decision, and further large phase III trial data sets should aim to confirm trial-level surrogacy.

Surrogate and Intermediate Endpoints in Randomized Trials: What's the Goal?

Establishing trial-level surrogacy of an intermediate endpoint for predicting survival benefit in future trials is extremely challenging because of the extrapolations required, but there are other useful drug development and patient management applications of intermediate endpoints. Clin Cancer Res; 24(10); 2239-40. ©2018 AACRSee related article by Mushti et al., p. 2268.

Surrogate end-points for overall survival in 22 neoadjuvant trials of gastro-oesophageal cancers

BackgroundThe surrogacy between disease-free survival (DFS) and overall survival (OS) has been evaluated in patients with gastric cancer who received adjuvant chemotherapy. Similar analyses in patients who have undergone neoadjuvant chemoradiotherapy (CTRT) or chemotherapy (CT) for gastro-oesophageal (GE) cancer have not yet been performed. MethodsWe evaluated the correlation between DFS and pathologic complete response (pCR) with OS in patients with GE carcinomas enrolled in randomised studies. A weighted linear regression analysis was used to evaluate surrogacy. Correlations were evaluated by squared correlation R2. Surrogacy of DFS and pCR with OS was assessed through the correlation between end-points and OS (individual-level surrogacy), and between the treatment effects on the end-points (trial-level surrogacy). Correlations were weighted by the number of patients in the intent-to-treat population. ResultsTwenty-two trials were included for a total of 4749 patients that received CTRT or CT for gastric or oesophageal cancers. Treatment effect on surrogates did not correlate with treatment effect on OS (R2 = 0.27, and R2 = 0.17, for DFS and pCR) at the trial level. In subgroup analysis surrogacy of DFS did not vary according to histology or treatment modality but was good in gastric cancer and poor in oesophageal cancers. ConclusionsDFS and pCR did not correlate with OS, and should not be used as surrogate end-points in patients with GE cancer who have undergone neoadjuvant therapy. OS should still represent the primary end-point to compare treatments in future randomised clinical trials.

Pathologic complete response and disease-free survival are not surrogate endpoints for 5-year survival in rectal cancer: an analysis of 22 randomized trials.

We performed a literature-based analysis of randomized clinical trials to assess the pathologic complete response (pCR) (ypT0N0 after neoadjuvant therapy) and 3-year disease-free survival (DFS) as potential surrogate endpoints for 5-year overall survival (OS) in rectal cancer treated with neoadjuvant (chemo)radiotherapy (CT)RT. A systematic literature search of PubMed, EMBASE, the Web of Science, SCOPUS, CINAHL, and the Cochrane Library was performed. Treatment effects on 3-year DFS and 5-year OS were expressed as rates of patients alive (%), and those on pCR as differences in pCR rates (∆pCR%). A weighted regression analysis was performed at individual- and trial-level to test the association between treatment effects on surrogate (∆pCR% and ∆3yDFS) and the main clinical outcome (∆5yOS). Twenty-two trials involving 10,050 patients, were included in the analysis. The individual level surrogacy showed that the pCR% and 3-year DFS were poorly correlated with 5-year OS (R=0.52; 95% CI, 0.31-0.91; P=0.002; and R=0.60; 95% CI, 0.36-1; P=0.002). The trial-level surrogacy analysis confirmed that the two treatment effects on surrogates (∆pCR% and ∆3yDFS) are not strong surrogates for treatment effects on 5-year OS % (R=0.2; 95% CI, -0.29-0.78; P=0.5 and R=0.64; 95% CI, 0.29-1; P=0.06). These findings were confirmed in neoadjuvant CTRT studies but not in phase III trials were 3-year DFS could still represent a valid surrogate. This analysis does not support the use of pCR and 3-year DFS% as appropriate surrogate endpoints for 5-year OS% in patients with rectal cancer treated with neoadjuvant therapy.

Multitrial Evaluation of Progression-Free Survival as a Surrogate End Point for Overall Survival in First-Line Extensive-Stage Small-Cell Lung Cancer

We previously reported that progression-free survival (PFS) may be a candidate surrogate end point for overall survival (OS) in first-line extensive-stage small-cell lung cancer (ES-SCLC) using data from three randomized trials (Foster, Cancer 2011). In this validation study (N0424-Alliance), we assessed the patient-level and trial-level surrogacy of PFS using data from seven new first-line phase II/III ES-SCLC trials and across all 10 trials as well (seven new, three previous). Individual patient data were utilized across the seven new trials (2259 patients) and all 10 trials (2855 patients). Patient-level surrogacy (Kendall's τ) was assessed using the Clayton copula bivariate survival model. Trial-level surrogacy was assessed through association of the log hazard ratios on OS and PFS across trials, including weighted (by trial size) least squares regression (WLS R²) of Cox model effects and correlation of the copula effects (copula R²). The minimum effect on the surrogate (MES) needed to detect a nonzero treatment effect on OS was also calculated. The median OS and PFS across all 10 trials were 9.8 and 5.9 months, respectively. PFS showed strong surrogacy within the 7 new trials (copula R² = 0.90 [standard error = 0.27], WLS R² = 0.83 [95% confidence interval: 0.43, 0.95]; MES = 0.67, and Kendall's τ = 0.58) and across all 10 trials (copula R² = 0.81 [standard errors = 0.25], WLS R² = 0.77 [95% confidence interval: 0.47-0.91], MES = 0.70, and Kendall's τ = 0.57). PFS demonstrated strong surrogacy for OS in first-line ES-SCLC based on this external validation study of individual patient data. PFS is a good alternative end point to OS and should be considered when resource constraints (time or patient) might make it useful or desirable in place of OS. Additional analyses are needed to assess its appropriateness for targeted agents in this disease setting.