Efficacy and cost-effectiveness of autologous bone marrow transplantation in metastatic breast cancer. Estimates using decision analysis while awaiting clinical trial results.

OBJECTIVE

To assess the efficacy and cost-effectiveness of standard chemotherapy and high-dose chemotherapy with autologous bone marrow transplantation (ABMT) in metastatic breast cancer.

DESIGN

Decision analysis model using a Markov process.

SETTING

Response and recurrence rates from the published literature for standard therapy and from case series of ABMT. Costs were based on local charges and on adjusted Medicare data.

PATIENTS

Hypothetical cohorts of women with metastatic breast cancer who had no bone marrow involvement and no comorbid illness.

INTERVENTION

The standard chemotherapy cohort received cyclophosphamide, doxorubicin, and fluorouracil. The ABMT cohort was treated with intense induction chemotherapy, then additional high-dose chemotherapy following a remission, with ABMT support.

MAIN OUTCOME MEASURES

Anticipated survival, incremental cost per year of life, and incremental cost per quality-adjusted year of life gained using a 5-year time horizon. Rigorous sensitivity analyses were done, including assessing a benefit "tail" of normal life expectancy for those free of disease after 5 years.

RESULTS

ABMT was the preferred approach under almost all assumptions, but the size of the benefit varied greatly. ABMT had a survival benefit of 6.0 months at 5 years at an incremental cost of $115,800 per year of life saved. If patients who were free of disease after 5 years had normal survival, the benefit was 18.1 months at an incremental cost of $28,600 per year. The benefit of ABMT was primarily dependent on whether the recurrence risk was constant or decreases after a finite period of time.

CONCLUSION

Using reasonable assumptions, ABMT provided a substantial benefit but at a cost that may be untenable. Decision analysis highlights the limitations in the currently available data and the assumptions made for the emotional question of using ABMT in metastatic breast cancer. The model supports the need for randomized clinical trials.