Background and objectivesGlofitamab, tafasitamab, loncastuximab tesirine, polatuzumab, and selinexor have been proposed for the treatment of relapsed-refractory diffuse large B-cell lymphoma (DLBCL). We studied the pattern of overall survival (OS) for these five agents.MethodsWe reconstructed patient-level data from the Kaplan-Meier OS graphs published in five pivotal trials. For this purpose, we used an artificial intelligence technique (the Shiny method). Reconstructed survival curves were subjected to standard statistics to perform cross-trial indirect comparisons; medians and hazard ratios (HRs) with 95% confidence interval (CI) were estimated for each treatment.ResultsUsing glofitamab (a bispecific antibody) as a common comparator, our analysis of OS yielded the following results: a) tafasitamab plus lenalidomide, HR: 0.514 (95% CI: 0.341 to 0.776; P=0.0015); b) polatuzumab vedotin, HR: 0.822 (95% CI: 0.509 to 1.327); c) selinexor, HR: 1.170 (95% CI: 0.852 to 1.603); and d) loncastuximab tesirine, HR: 1.120 (95% CI: 0.868 to 1.659). Medians were estimated as follows: a) tafasitamab plus lenalidomide, 26.5 months (95% CI: 18.9 to NA); b) polatuzumab vedotin, 12.5 months (95% CI: 9.03 to NA); c) glofitamab, 11.7 months (95% CI: 7.96 to 18.0); d) loncastuximab tesirine, 10.2 months (95% CI: 6.97 to 11.6); and e) selinexor, 10.1 months (95% CI: 6.72 to 14.2).ConclusionsThese comparative results represent an original finding generated by the Shiny method. Although these comparisons are indirect, our analysis offers a useful synthesis of the outcomes of these treatments. According to these results, glofitamab, despite its improved mechanism of action, does not seem to confer any OS advantage compared with the other four treatments.