‘Limited Evidence’ for Surrogate Endpoints in Medical Oncology Trials

The validity of surrogate endpoints for overall survival in medical oncology trials is questioned

medwireNews: There is little correlation between overall survival (OS) and the surrogate endpoints commonly used in oncology clinical trials, suggests a systematic review of trial-level meta-analyses.

Moreover, the validation studies examined were found to focus on subsets of oncology trials, with few analyses including findings of unpublished trials, the researchers caution in JAMA Internal Medicine.

“Our findings call into question the widespread use of surrogate end points in oncology as a basis for treatment decisions”, say Vinay Prasad, from the National Cancer Institute in Bethesda, Maryland, USA, and co-authors.

The researchers identified 36 meta-analyses of medical oncology trials reporting on the relationship between OS and surrogate endpoints, such as response rate, progression-free survival, event-free survival, time to progression, pathological complete response and locoregional control.

Ten of the analyses included systematic reviews of published literature and 14 reported on published literature and meeting abstracts, while four of the meta-analyses described only a “convenience sample” where the authors were able to easily obtain the data.

Just five studies attempted to include data from unpublished trials and these included only 51.1% of 684 eligible studies identified in their meta-analyses.

“Unpublished trials (and those that do not report data that meta-analysts require) may have poorer correlations than those that are published, and such discrepancies may contribute to reluctance among sponsors and authors to submit those findings to journals and conferences and to a poorer likelihood of acceptance”, explain Vinay Prasad et al.

Using Institute of Quality and Efficiency in Health Care criteria, the researchers found just 23% of the 65 different trial-level analyses reported showed strong correlations with OS, while 25% of correlations were of medium strength and 52% were of low strength.

Of the high-strength correlations, six were identified in studies of the adjuvant setting – including three in colorectal cancer – six in the metastatic setting and three in trials of patients with locally advanced disease. Four of the strong correlations were found in studies where multiple other studies had previously found lower strength relationships.

Noting that the use of surrogate endpoints, and in particular progression-free survival, has become more common and is used for the regulatory approval of agents and their acceptance by insurance providers, the researchers write: “Our analysis confirms that the use of this end point occurs beyond the specific settings in which it has been validated.”

They conclude: “Simply because a surrogate is measurable does not make it predictive or meaningful, and our analysis builds on a prior study that concludes that the association between surrogates and survival in oncology is generally poor.”


Prasad V, Kim C, Burotto M, Vandross A. The strength of association between surrogate end points and survival in oncology. A systematic review of trial-level meta-analyses. JAMA Intern Med 2015; Advance online publication 22 June. doi:10.1001/jamainternmed.2015.2829

medwireNews (www.medwireNews.com ) is an independent medical news service provided by Springer Healthcare Limited. © Springer Healthcare Ltd; 2015