Improved Survival for Multiple Myeloma in Denmark Based on Autologous Stem Cell Transplantation and Novel Drug Therapy in Collaborative Trials: Analysis of Accrual, Prognostic Variables, Selection Bias, and Clinical Behavior on Survival in More Than 1200 Patients in Trials of the Nordic Myeloma Study Group

Background: An unexplained survival difference was observed in the Nordic Myeloma Study Group (NMSG) high-dose therapy trial 5/94 in Denmark compared with Sweden and Norway; however, this difference was eliminated in the subsequent NMSG trial 7/98. It was hypothesized that a detailed analysis of potential explanations would reveal important information for future designs of clinical trials for multiple myeloma (MM) patients in Denmark. Patients and Methods: The analysis is based on 3 consecutive clinical trials coordinated by NMSG from 1990 to 2000: NMSG 4/90 including 583 patients, NMSG 5/94 including 274 patients and NMSG 7/98 including 414 patients with newly diagnosed MM. Event-free and total survival rates were calculated according to the Kaplan-Meier method, and survival comparisons were made by the log-rank test. The Cox proportional hazards regression model was used to estimate the prognostic importance of selected variables. Results: The analysis revealed no differences in disease stages, prognostic variables, or inclusion bias at diagnosis between the 3 consecutive NMSG trials. However, the number of initial treatment failures was low, and post-relapse survival was superior in Swedish patients as compared to Danish patients. These differences were explained by a defensive clinical practice in Denmark during 1994-1997 for patients with poor risk refractory or relapsed disease. Conclusion: These initially observed differences were subsequently eliminated most likely as a consequence of international collaboration improving diagnosis, research infrastructure, clinical training, and education as planned within the European Myeloma Network (EMN).