A comprehensive analysis of immunotherapy trials reveals a paucity of translational research endpoints in trial design


Session type:


Sophie Raby1,Dr Claire Connell2,Dr Tobias Janowitz2
1The Christie,2Addenbrookes Hospital



There has been a rapid growth in immunotherapy trials. Long term durable responses occur, but only in a subset of patients. As yet no accurate method of identifying patients most likely to benefit has been identified. Factors behind non-response are unclear but include 1) inherent tumour characteristics, 2) immunogenicity factors such as tumour burden and previous treatments and 3) clinical trial design. By performing a cross-sectional analysis of registered trials investigating agents thought to stimulate Tcells we aimed to detect trends in these. In particular we aimed to assess the extent to which trials sought to develop novel biomarkers of response.


A pubmed literature search was conducted to establish a list of known Tcell checkpoints, co-stimulatory receptors, ligands, and antibodies targeting these. These search terms were entered into clinicaltrials.gov. Study details were downloaded as datasets for review by two independent assessors.


We identified 350 trials of immunomodulatory antibodies targeting PD-1, CTLA4, PD-L1, PD-L2, LAG3, B7-H3, CD137, Ox40, CD27 and GITR. Longitudinal analysis by registration date shows a steady increase in these. As some cancer types are thought to be more immunogenic, we looked at cancer type. Unsurprisingly, melanoma trials represent the largest proportion, but there is a shift towards testing in cancers considered less immunogenic, with a significant increase in NSCLC trials. Only 39% of trials measured a dynamic immune endpoint as a specified outcome. T and Bcell number or function were the most commonly analysed. There was a significant increase in Pdl1 expression measurement.


This analysis provides comprehensive data on the rapid growth of immunotherapy trials and highlights that despite the multiplicity and variability of potential dynamic biomarkers available, there is a poor uptake. Whilst the future of immunotherapy is not in doubt, biomarkers are essential to understand the considerable lack of response and help guide further trial efforts.