Improving the clinical meaning of surrogate endpoints: An empirical assessment of clinical progression in phase III oncology trials.

Disease progression in clinical trials is commonly defined by radiologic measures. However, clinical progression may be more meaningful to patients, may occur even when radiologic criteria for progression are not met, and often requires a change in therapy in clinical practice. The objective of this study was to determine the utilization of clinical progression criteria within progression-based trial endpoints among phase III trials testing systemic therapies for metastatic solid tumors. The primary manuscripts and protocols of phase III trials were reviewed for whether clinical events, such as refractory pain, tumor bleeding, or neurologic compromise, could constitute a progression event. Univariable logistic regression computed odds ratios (OR) and 95% CI for associations between trial-level covariates and clinical progression. A total of 216 trials enrolling 148,190 patients were included, with publication dates from 2006 through 2020. A major change in clinical status was included in the progression criteria of 13% of trials (n = 27), most commonly as a secondary endpoint (n = 22). Only 59% of trials (n = 16) reported distinct clinical progression outcomes that constituted the composite surrogate endpoint. Compared with other disease sites, genitourinary trials were more likely to include clinical progression definitions (16/33 [48%] vs. 11/183 [6%]; OR, 14.72; 95% CI, 5.99 to 37.84; p < .0001). While major tumor-related clinical events were seldom considered as disease progression events, increased attention to clinical progression may improve the meaningfulness and clinical applicability of surrogate endpoints for patients with metastatic solid tumors.

An information-theoretic approach for the assessment of a continuous outcome as a surrogate for a binary true endpoint based on causal inference: Application to vaccine evaluation.

Within the causal association paradigm, a method is proposed to assess the validity of a continuous outcome as a surrogate for a binary true endpoint. The methodology is based on a previously introduced information-theoretic definition of surrogacy and has two main steps. In the first step, a new model is proposed to describe the joint distribution of the potential outcomes associated with the putative surrogate and the true endpoint of interest. The identifiability issues inherent to this type of models are handled via sensitivity analysis. In the second step, a metric of surrogacy new to this setting, the so-called individual causal association is presented. The methodology is studied in detail using theoretical considerations, some simulations, and data from a randomized clinical trial evaluating an inactivated quadrivalent influenza vaccine. A user-friendly R package Surrogate is provided to carry out the evaluation exercise.

Percentage area reduction at week 4 as a prognostic indicator of complete healing in patients treated with standard of care: a post hoc analysis.

Early indicators of healing provide valuable information on the potential benefit of treatment. In patients with hard-to-heal (chronic) diabetic foot ulcers (DFUs), timely intervention is critical. Ulcers that fail to show measurable progress within four weeks of treatment are considered recalcitrant. These ulcers increase the risk of soft tissue infection, osteomyelitis and lower extremity amputation. A prognostic indicator or surrogate marker allows for rapid evaluation of treatment efficacy and safety. An inverse correlation between a percentage area reduction (PAR) of ≤50% at week 4 and complete healing by week 12 has been previously established; however, the data were derived from a standard of care (SoC) arm of clinical trials that are over a decade old. In this post hoc analysis, data from a large multicentre prospective randomised controlled trial were reviewed to assess PAR at week 4 as a prognostic indicator in patients treated with SoC. Overall, 65.4% (17/26) of patients with PAR >50% at week 4 achieved complete closure at week 12. The receiver operating characteristic (ROC) curve for area reduction by week 4 showed strong discrimination for predicting non-healing (area under the ROC curve: 0.92; p<0.001; positive predictive value: 70.6%; negative predictive value: 87.2%). These findings are consistent with previous studies and support the use of four-week PAR as a prognostic indicator.

A systematic review of potential biomarkers for bacterial burden and treatment efficacy assessment in TB platform-based clinical trials.

Adaptive platform trials can be more efficient than classic trials for developing new treatments. Moving from culture-based to simpler- or faster-to-measure biomarkers as efficacy surrogates may enhance this advantage. We performed a systematic review of treatment efficacy biomarkers in adults with tuberculosis. Platform trials can span different development phases. We grouped biomarkers as: α, bacterial load estimates used in phase 2a trials; ß, early and end-of treatment endpoints, phase 2b-c trials; γ, post-treatment or trial-level estimates, phase 2c-3 trials. We considered as analysis unit (biomarker entry) each combination of biomarker, predicted outcome, and their respective measurement times or intervals. Performance metrics included: sensitivity, specificity, area under the receiver-operator curve (AUC), and correlation measures, ,and classified as poor, promising, or good. Eighty-six studies included 22864 participants. From 1356 biomarker entries, 318 were reported with the performance metrics of interest, with 103 promising and 41 good predictors. Group results: α, mycobacterial RNA and Lipoarabinomannan in sputum, and host metabolites in urine; ß, mycobacterial RNA and host transcriptomic or cytokine signatures for early treatment response; γ, host transcriptomics for recurrence. A combination of biomarkers from different categories could help designing more efficient platform trials. Efforts to develop efficacy surrogates should be better coordinated.

Solutions for surrogacy validation with longitudinal outcomes for a gene therapy.

Valid surrogate endpoints S can be used as a substitute for a true outcome of interest T to measure treatment efficacy in a clinical trial. We propose a causal inference approach to validate a surrogate by incorporating longitudinal measurements of the true outcomes using a mixed modeling approach, and we define models and quantities for validation that may vary across the study period using principal surrogacy criteria. We consider a surrogate-dependent treatment efficacy curve that allows us to validate the surrogate at different time points. We extend these methods to accommodate a delayed-start treatment design where all patients eventually receive the treatment. Not all parameters are identified in the general setting. We apply a Bayesian approach for estimation and inference, utilizing more informative prior distributions for selected parameters. We consider the sensitivity of these prior assumptions as well as assumptions of independence among certain counterfactual quantities conditional on pretreatment covariates to improve identifiability. We examine the frequentist properties (bias of point and variance estimates, credible interval coverage) of a Bayesian imputation method. Our work is motivated by a clinical trial of a gene therapy where the functional outcomes are measured repeatedly throughout the trial.

On the individual surrogate paradox.

When the primary outcome is difficult to collect, a surrogate endpoint is typically used as a substitute. It is possible that for every individual, the treatment has a positive effect on the surrogate, and the surrogate has a positive effect on the primary outcome, but for some individuals, the treatment has a negative effect on the primary outcome. For example, a treatment may be substantially effective in preventing the stroke for everyone, and lowering the risk of stroke is universally beneficial for life expectancy; however, the treatment may still cause death for some individuals. We define such paradoxical phenomenon as the individual surrogate paradox. The individual surrogate paradox is proposed to capture the treatment effect heterogeneity, which is unable to be described by either the surrogate paradox based on average causal effect or that based on distributional causal effect. We investigate the existing surrogate criteria in terms of whether the individual surrogate paradox could manifest. We find that only the strong binary surrogate can avoid such paradox without additional assumptions. Utilizing the sharp bounds, we propose novel criteria to exclude the individual surrogate paradox. Our methods are illustrated in an application to determine the effect of the intensive glycemia on the risk of development or progression of diabetic retinopathy.

Use of copula to model within-study association in bivariate meta-analysis of binomial data at the aggregate level: A Bayesian approach and application to surrogate endpoint evaluation.

Bivariate meta-analysis provides a useful framework for combining information across related studies and has been utilized to combine evidence from clinical studies to evaluate treatment efficacy on two outcomes. It has also been used to investigate surrogacy patterns between treatment effects on the surrogate endpoint and the final outcome. Surrogate endpoints play an important role in drug development when they can be used to measure treatment effect early compared to the final outcome and to predict clinical benefit or harm. The standard bivariate meta-analytic approach models the observed treatment effects on the surrogate and the final outcome outcomes jointly, at both the within-study and between-studies levels, using a bivariate normal distribution. For binomial data, a normal approximation on log odds ratio scale can be used. However, this method may lead to biased results when the proportions of events are close to one or zero, affecting the validation of surrogate endpoints. In this article, we explore modeling the two outcomes on the original binomial scale. First, we present a method that uses independent binomial likelihoods to model the within-study variability avoiding to approximate the observed treatment effects. However, the method ignores the within-study association. To overcome this issue, we propose a method using a bivariate copula with binomial marginals, which allows the model to account for the within-study association. We applied the methods to an illustrative example in chronic myeloid leukemia to investigate the surrogate relationship between complete cytogenetic response and event-free-survival.

Clinical and research applications of multiplexed immunohistochemistry and in situ hybridization.

Over the past decade, invention and adoption of novel multiplexing technologies for tissues have made increasing impacts in basic and translational research and, to a lesser degree, clinical medicine. Platforms capable of highly multiplexed immunohistochemistry or in situ RNA measurements promise evaluation of protein or RNA targets at levels of plex and sensitivity logs above traditional methods - all with preservation of spatial context. These methods promise objective biomarker quantification, markedly increased sensitivity, and single-cell resolution. Increasingly, development of novel technologies is enabling multi-omic interrogations with spatial correlation of RNA and protein expression profiles in the same sample. Such sophisticated methods will provide unprecedented insights into tissue biology, biomarker science, and, ultimately, patient health. However, this sophistication comes at significant cost, requiring extensive time, practical knowledge, and resources to implement. This review will describe the technical features, advantages, and limitations of currently available multiplexed immunohistochemistry and spatial transcriptomic platforms. © 2021 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Development of a framework and decision tool for the evaluation of health technologies based on surrogate endpoint evidence.

In the drive toward faster patient access to treatments, health technology assessment (HTA) agencies and payers are increasingly faced with reliance on evidence based on surrogate endpoints, increasing decision uncertainty. Despite the development of a small number of evaluation frameworks, there remains no consensus on the detailed methodology for handling surrogate endpoints in HTA practice. This research overviews the methods and findings of four empirical studies undertaken as part of COMED (Pushing the Boundaries of Cost and Outcome Analysis of Medical Technologies) program work package 2 with the aim of analyzing international HTA practice of the handling and considerations around the use of surrogate endpoint evidence. We have synthesized the findings of these empirical studies, in context of wider contemporary body of methodological and policy-related literature on surrogate endpoints, to develop a web-based decision tool to support HTA agencies and payers when faced with surrogate endpoint evidence. Our decision tool is intended for use by HTA agencies and their decision-making committees together with the wider community of HTA stakeholders (including clinicians, patient groups, and healthcare manufacturers). Having developed this tool, we will monitor its use and we welcome feedback on its utility.

[Evaluation considerations for using surrogate endpoints in anticancer clinical trials].

The prolongation of patient's overall survival is the accepted as gold standard to prove clinical values of anti-cancer drugs. However, if overall survival is taken as the primary endpoint in clinical trials for cancer types with a relatively good prognosis in the process of new anti-cancer drug research and development, the time to market the drugs will be prolonged due to the long follow-up time. In addition, overall survival is often interfered by confounding factors such as follow-up treatment. Therefore, regulatory agencies have established an accelerated review model using surrogate endpoints for the approval of new anti-cancer drugs, but there are still some problems in the use of surrogate endpoints in cancer clinical trials. From the perspective of new drug review, the authors expounds the key points of confirming and rationally using surrogate endpoints in clinical trials of anti-cancer drugs, which will improve the level of clinical trials of new anti-cancer drugs and accelerate the development of anti-tumor drugs.

Are radiological endpoints surrogate outcomes of overall survival in hepatocellular carcinoma treated with transarterial chemoembolization?

BACKGROUND& AIMS: Time to progression (TTP) and progression-free survival (PFS) are commonly used as surrogate endpoints in oncology trials. We aimed to assess the surrogacy relationship of TTP and PFS with overall survival (OS) in studies of transarterial chemoembolization (TACE) for unresectable hepatocellular carcinoma (u-HCC) by innovative methods. METHODS: A search of databases for studies of TACE for u-HCC reporting both OS and TTP or PFS was performed. Individual patient data were extracted from TTP/PFS and OS Kaplan-Meier curves of TACE arms. Pooled median TTP and OS were obtained from random-effect model. The surrogate relationships of hazard ratios (HRs) and median TTP for OS were evaluated by the coefficient of determination R2 . RESULTS: We identified 13 studies comparing TACE vs systemic therapy or vs TACE plus systemic therapy and including 1932 TACE-treated patients. Pooled median OS was 11.2 months (95% confidence interval [95%CI] 7.9-17.8), and pooled median TTP was 5.4 months (95%CI 3.8-8.0). Heterogeneity among studies was highly significant for both outcomes. The correlation between HR TTP and HR OS was moderate (R2  = 0.65. 95%CI 0.08-0.81). R2 value was 0.04 (95%CI 0.00-0.35) between median TTP and median OS. CONCLUSION: In studies of TACE for u-HCC, the surrogate relationship of radiology-based endpoints with OS is moderate. Multiple endpoints including hepatic decompensation, macrovascular invasion and extrahepatic spread are needed for future trials comparing systemic therapies or combination of TACE with systemic therapies vs TACE alone.

A single-center retrospective analysis of outcome measures and consolidation strategies for relapsed and refractory primary CNS lymphoma.

BACKGROUND: Relapsed or refractory primary CNS lymphoma (rrPCNSL) is a rare and challenging malignancy for which better evidence is needed to guide management. METHODS: We present a retrospective cohort of 66 consecutive patients with rrPCNSL treated at the University of Washington between 2000 and 2020. Immunosuppressed and secondary CNS lymphoma patients were excluded. RESULTS: During a median follow-up of 40.5 months from initial diagnosis, median OS for relapsed disease was 14.1 (0.2-88.5) months and median PFS was 11.0 (0.2-73.9) months. At diagnosis (r2 = 0.85, p < 0.001), first relapse (r2 = 0.69, p < 0.001), multiple relapses (r2 = 0.97, p < 0.001) PFS was highly correlated with OS. In contrast, there was no correlation between the duration of subsequent progression-free intervals. No difference in PFS or OS was seen between CSF or intraocular relapse and parenchymal relapse. Patients reinduced with high-dose methotrexate-based (HD-MTX) regimens had an overall response rate (ORR) of 86.7%. Consolidation with autologous stem cell transplant (ASCT) was associated with longer PFS compared to either no consolidation (p = 0.01) and trended to longer PFS when compared to other consolidation strategies (p = 0.06). OS was similarly improved in patients consolidated with ASCT compared with no consolidation (p = 0.04), but not compared with other consolidation (p = 0.22). Although patients receiving ASCT were younger, KPS, sex, and number of recurrences were similar between consolidation groups. A multivariate analysis confirmed an independent effect of consolidation group on PFS (p = 0.01), but not OS. CONCLUSIONS: PFS may be a useful surrogate endpoint which predicts OS in PCNSL. Consolidation with ASCT was associated with improved PFS in rrPCNSL.

Use of the likelihood reduction factor in a path analysis framework to quantify surrogacy in clinical trials.

In clinical trials, surrogate endpoints are useful when the endpoint of interest is difficult to measure or requires a long follow-up time. Current methodology for validating surrogate endpoints encounters challenges in the presence of collinearity between the treatment and surrogate endpoint, which is often present in clinical trials. The proposed methods adapt current methodology in the structural framework of path analysis to quantify the validity of a surrogate endpoint. The path analysis framework provides an improved interpretation of treatment effect. Through derivation and simulation we show the proposed path likelihood reduction factor (LRF P ), is less biased and more robust than current methodology in cases of collinearity between the treatment and surrogate endpoint, with notable improvement when surrogacy is weak or moderate. LRF P can be expanded to evaluate multiple correlated surrogate endpoints, which as shown through simulation, is also less biased and more robust than current methodology in the case of collinearity between the treatment and surrogate endpoint.

Incorporating baseline covariates to validate surrogate endpoints with a constant biomarker under control arm.

A surrogate endpoint S in a clinical trial is an outcome that may be measured earlier or more easily than the true outcome of interest T. In this work, we extend causal inference approaches to validate such a surrogate using potential outcomes. The causal association paradigm assesses the relationship of the treatment effect on the surrogate with the treatment effect on the true endpoint. Using the principal surrogacy criteria, we utilize the joint conditional distribution of the potential outcomes T, given the potential outcomes S. In particular, our setting of interest allows us to assume the surrogate under the placebo, S(0) , is zero-valued, and we incorporate baseline covariates in the setting of normally distributed endpoints. We develop Bayesian methods to incorporate conditional independence and other modeling assumptions and explore their impact on the assessment of surrogacy. We demonstrate our approach via simulation and data that mimics an ongoing study of a muscular dystrophy gene therapy.

Optical coherence tomography in adult adrenoleukodystrophy: a cross-sectional and longitudinal study.

BACKGROUND: Adrenoleukodystrophy (ALD) encompasses different neurological phenotypes, ranging from the most severe cerebral forms (C-ALD) to the less severe adrenomyeloneuropathy (AMN). As visual system can be varyingly involved, we aimed at exploring whether optical coherence tomography (OCT) may detect retinal abnormalities and their longitudinal changes in adult ALD patients. METHODS: In this cross-sectional and longitudinal study, we measured the thicknesses of peripapillary retinal nerve fiber layer (pRNFL), macular ganglion cell complex (mGCC), and segmented inner and outer macula at baseline and their changes over time in 11 symptomatic adult ALD males and 10 age- and sex-matched healthy controls. Statistical analyses were performed for the patients as complete group, and splitting them into two subgroups, one (C-ALD) with and the other (AMN) without cerebral parieto-occipital white matter (WM) lesions. RESULTS: In the complete ALD group and in the C-ALD subgroup, the average pRNFL, mGCC, and inner macula were significantly thinner than in controls (p ≤ 0.01), whereas in the AMN subgroup, they were constantly, though non-significantly, thinner. Significant outer macula thinning was also observed (p < 0.01). In the complete ALD group, follow-up assessment (mean 26.8 months, range 8-48) showed mildly progressive thinning of inferior pRNFL, average mGCC, and inner macula. CONCLUSIONS: In adult ALD patients, OCT can reveal retinal abnormalities which are prominent in the more compromised patients, namely those with parieto-occipital WM lesions. The inferior pRNFL, average mGCC and inner macula thicknesses might be sensitive-to-change OCT parameters, but their utility and consistency for short-term longitudinal studies deserve further investigations.

Surrogate endpoint evaluation using data from one large global randomized controlled trial.

BACKGROUND: Robust identification of surrogate endpoints can help accelerate the development of pharmacotherapies for diseases traditionally evaluated using true endpoints associated with prolonged follow-up. The meta-analysis-based surrogate endpoint evaluation (SEE) integrates data from multiple, usually smaller, trials to statistically confirm a surrogate endpoint as a robust proxy for the true endpoint. To test the applicability of SEE when only a single, larger trial is available, we analysed the cardiovascular (CV) survival endpoint from the large multinational trial LEADER (9340 subjects) that confirmed the CV safety of a diabetes drug (liraglutide). We evaluated if using country as a trial unit adequately facilitated the meta-analysis and calculation of R2 by country group. METHODS: Data were grouped by country, ensuring at least 30 CV deaths (497 in total) in each of the nine resulting by-country groups. In a two-step SEE on the grouped dataset, we first fitted the group-specific Cox proportional hazard models; next, on the trial-level, we regressed the estimated hazard ratio (HR; liraglutide vs placebo) of the true endpoints (CV death: 497 events, or all-cause death: 828 events) on the HR of the surrogate endpoint (major CV adverse event [MACE]: 1302 events) and derived the group-specific R2 and its 95% confidence interval (CI). RESULTS: Group-level surrogacy of MACE was supported for CV death but not for all-cause death, with [Formula: see text] values of 0.85 [0.63;1.00]95% CI and 0.23 [0.00;0.67]95% CI, respectively. Sensitivity analyses using different grouping approaches (e.g. grouping by region) corroborated the robustness of the conclusions as well as the appropriateness of the data-grouping approaches. CONCLUSIONS: We derived a specific grouping approach to successfully apply SEE on data from a single trial. This may allow for the statistically robust identification and validation of surrogate endpoints based on the abundance of large monolithic outcome trials conducted as part of drug development programmes in, for example, diabetes.

Proportion of treatment effect mediated by surrogate endpoints.

One of the central aims in randomized clinical trials is to find well-validated surrogate endpoints to reduce the sample size and/or duration of trials. Clinical researchers and practitioners have proposed various surrogacy measures for assessing candidate surrogate endpoints. However, most existing surrogacy measures have the following shortcomings: (i) they often fall outside the range [0,1], (ii) they are imprecisely estimated, and (iii) they ignore the interaction associations between a treatment and candidate surrogate endpoints in the evaluation of the surrogacy level. To overcome these difficulties, we propose a new surrogacy measure, the proportion of treatment effect mediated by candidate surrogate endpoints (PMS), based on the decomposition of the treatment effect into direct, indirect, and interaction associations mediated by candidate surrogate endpoints. In addition, we validate the advantages of PMS through Monte Carlo simulations and the application to empirical data from ORIENT (the Olmesartan Reducing Incidence of Endstage Renal Disease in Diabetic Nephropathy Trial).

Change in Albuminuria and GFR as End Points for Clinical Trials in Early Stages of CKD: A Scientific Workshop Sponsored by the National Kidney Foundation in Collaboration With the US Food and Drug Administration and European Medicines Agency.

The US Food and Drug Administration (FDA) and European Medicines Agency (EMA) are currently willing to consider a 30% to 40% glomerular filtration rate (GFR) decline as a surrogate end point for kidney failure for clinical trials of kidney disease progression under appropriate conditions. However, these end points may not be practical for early stages of kidney disease. In March 2018, the National Kidney Foundation sponsored a scientific workshop in collaboration with the FDA and EMA to evaluate changes in albuminuria or GFR as candidate surrogate end points. Three parallel efforts were presented: meta-analyses of observational studies (cohorts), meta-analyses of clinical trials, and simulations of trial design. In cohorts, after accounting for measurement error, relationships between change in urinary albumin-creatinine ratio (UACR) or estimated GFR (eGFR) slope and the clinical outcome of kidney disease progression were strong and consistent. In trials, the posterior median R2 of treatment effects on the candidate surrogates with the clinical outcome was 0.47 (95% Bayesian credible interval [BCI], 0.02-0.96) for early change in UACR and 0.72 (95% BCI, 0.05-0.99) when restricted to baseline UACR>30mg/g, and 0.97 (95% BCI, 0.78-1.00) for total eGFR slope at 3 years and 0.96 (95% BCI, 0.63-1.00) for chronic eGFR slope (ie, the slope excluding the first 3 months from baseline, when there might be acute changes in eGFR). The magnitude of the relationships of changes in the candidate surrogates with risk for clinical outcome was consistent across cohorts and trials: a UACR reduction of 30% or eGFR slope reduction by 0.5 to 1.0mL/min/1.73m2 per year were associated with an HR of ∼0.7 for the clinical outcome in cohorts and trials. In simulations, using GFR slope as an end point substantially reduced the required sample size and duration of follow-up compared with the clinical end point when baseline eGFR was high, treatment effects were uniform, and there was no acute effect of the treatment. We conclude that both early change in albuminuria and GFR slope fulfill criteria for surrogacy for use as end points in clinical trials for chronic kidney disease progression under certain conditions, with stronger support for change in GFR than albuminuria. Implementation requires understanding conditions under which each surrogate is likely to perform well and restricting its use to those settings.

Clinical trials to assess adjuvant therapeutics for severe malaria.

Despite potent anti-malarial treatment, mortality rates associated with severe falciparum malaria remain high. To attempt to improve outcome, several trials have assessed a variety of potential adjunctive therapeutics, however none to date has been shown to be beneficial. This may be due, at least partly, to the therapeutics chosen and clinical trial design used. Here, we highlight three themes that could facilitate the choice and evaluation of putative adjuvant interventions for severe malaria, paving the way for their assessment in randomized controlled trials. Most clinical trials of adjunctive therapeutics to date have been underpowered due to the large number of participants required to reach mortality endpoints, rendering these study designs challenging and expensive to conduct. These limitations may be mitigated by the use of risk-stratification of participants and application of surrogate endpoints. Appropriate surrogate endpoints include direct measures of pathways causally involved in the pathobiology of severe and fatal malaria, including markers of host immune and endothelial activation and microcirculatory dysfunction. We propose using circulating markers of these pathways to identify high-risk participants that would be most likely to benefit from adjunctive therapy, and further by adopting these biomarkers as surrogate endpoints; moreover, choosing interventions that target deleterious host immune responses that directly contribute to microcirculatory dysfunction, multi-organ dysfunction and death; and, finally, prioritizing where possible, drugs that act on these pathways that are already approved by the FDA, or other regulators, for other indications, and are known to be safe in target populations, including children. An emerging understanding of the critical role of the host response in severe malaria pathogenesis may facilitate both clinical trial design and the search of effective adjunctive therapeutics.

Bayesian hierarchical meta-analytic methods for modeling surrogate relationships that vary across treatment classes using aggregate data.

Surrogate endpoints play an important role in drug development when they can be used to measure treatment effect early compared to the final clinical outcome and to predict clinical benefit or harm. Such endpoints are assessed for their predictive value of clinical benefit by investigating the surrogate relationship between treatment effects on the surrogate and final outcomes using meta-analytic methods. When surrogate relationships vary across treatment classes, such validation may fail due to limited data within each treatment class. In this paper, two alternative Bayesian meta-analytic methods are introduced which allow for borrowing of information from other treatment classes when exploring the surrogacy in a particular class. The first approach extends a standard model for the evaluation of surrogate endpoints to a hierarchical meta-analysis model assuming full exchangeability of surrogate relationships across all the treatment classes, thus facilitating borrowing of information across the classes. The second method is able to relax this assumption by allowing for partial exchangeability of surrogate relationships across treatment classes to avoid excessive borrowing of information from distinctly different classes. We carried out a simulation study to assess the proposed methods in nine data scenarios and compared them with subgroup analysis using the standard model within each treatment class. We also applied the methods to an illustrative example in colorectal cancer which led to obtaining the parameters describing the surrogate relationships with higher precision.