Journal of Clinical Oncology, Vol 10, 200-209, Copyright © 1992 by American Society of Clinical Oncology
The optimal timing of autologous bone marrow transplantation in Hodgkin's disease patients after a chemotherapy relapse
CE Desch, MR Lasala, TJ Smith and BE Hillner
Massey Cancer Center, Medical College of Virginia, Virginia Commonwealth University, Richmond 23298-0037.
PURPOSE: The optimal sequence of salvage chemotherapy (SC) and autologous
bone marrow transplantation (ABMT) for Hodgkin's disease (HD) patients who
relapse after primary chemotherapy is unknown. We created a decision
analysis model to determine the optimal treatment strategy and the most
cost-effective approach. METHODS: The decision tree simulated a 25-year-old
HD patient who relapsed less than 12 months after mechlorethamine,
vincristine, procarbazine, and prednisone (MOPP) chemotherapy. Four
strategies used ABMT in some sequence with SC; the final strategy
considered SC alone. Clinical data were derived from 17 published reports
chosen by explicit criteria. Costs of care were estimated from the
published literature and institutional experience. RESULTS: The optimal
strategy was ABMT in second relapse, which was superior to the SC-only
option by 1.9 years at an incremental cost of $26,200 per each year of life
saved. When the probabilities of complete remission and disease-free
survival were reduced for SC, similar to the clinical expectation of SC
after a seven- or eight-drug regimen like MOPP/doxorubicin, bleomycin, and
vinblastine with or without dacarbazine (MOPP/ABV[D]), ABMT in first
relapse was the preferred strategy and provided 6 additional months.
However, when the data from favorable (or unfavorable) SC and ABMT reports
were compared head-to-head in this model, SC followed by ABMT in second
relapse was always optimal. CONCLUSIONS: All relapsed HD patients should
plan to use ABMT in some sequence with SC, if necessary. In most situations
the optimal strategy is ABMT in second relapse. This analysis will assist
clinicians in planning treatment for relapsed HD patients. It could be
refined if historical series were updated to report the incidence and
outcomes of SC relapse from seven- or eight-drug regimens.