Development of adapted RECIST criteria to assess response in lymphoma and their comparison to the International Workshop Criteria

RECIST (response evaluation criteria in solid tumours) uses a unidimensional approach to tumour measurement and has been widely adopted for assessing the response rate of new therapies in solid tumour clinical trials. For lymphoma, the IWC (International Workshop Criteria), based on bidimensional product assessment, is generally utilised. We adapted RECIST for use in lymphoma and compared responses with the IWC in three Phase II lymphoma trials (n = 115). Measures of agreement estimated the concordance between the adapted RECIST and the IWC response assessments. A Pearson's coefficient estimated the correlation between changes in uni- and bidimensional measurements in a subset of patients (n = 75). All measures of agreement were very high [kappa=0.86 (95% CI: 0.76-0.95), percent agreement 0.93 (95% CI: 0.87-0.97), positive agreement 0.90 (95% CI: 0.87-0.98), negative agreement 0.92 (95% CI: 0.89-0.98)]. Pearson's coefficient was 0.92 (95% CI: 0.87, 0.95). The lymphoma-adapted RECIST is simpler to apply than the IWC and yields near identical response rates. The adapted RECIST should be considered for inclusion into any new draft of the IWC.