Cost-effectiveness of intraoperative radiation therapy for breast conservation.

Abstract

6081 Background: Results from the TARGIT trial at 4 years median follow-up for women >50 yrs with node negative tumors show equivalent outcomes in local recurrence, however, it has been argued that long-term follow-up is required before adoption. Modeling can inform this decision by evaluating the tradeoffs between costs, adverse outcomes, and quality of life. Methods: Using a Markov model, a cost-utility analysis compared lumpectomy with Intraoperative Radiation Therapy (IORT) versus standard External Beam Radiation Therapy (EBRT). Local recurrence rates (LRR) were based on results from the TARGIT trial. For the base case, LRR for EBRT was assumed to progress linearly to an expected 2.4% at 10 years. LRR for IORT was set to 5.5%, resulting in a more than 3-fold increase from the observed, published 4-year rate. Utilities were derived from the literature. Baseline utility for no disease after RT was set to .92 for IORT and EBRT. Medicare CPT codes were used for direct costs and non-medical costs were ascertained. Quality adjusted life years (QALY), costs, and the incremental cost-effectiveness ratio (ICER) were calculated over 10 years. Results: In the baseline analysis, IORT is less costly than EBRT, but less effective in terms of QALYs. The use of EBRT led to an increase in .011 QALYS (<5 days) but an incremental increase in $2840, resulting in an ICER of more than $250,000 per QALY. For the ICER to decrease to $81,000/QALY the observed LRR for IORT must increase 8.8 fold from the 4-year rate. Sensitivity analyses demonstrated the ICER to be affected most by the costs of radiation treatment, LRR, and utility measures for post-lumpectomy IORT and EBRT. If the utility for IORT were increased by .002 (out of 1.000), IORT becomes cost-saving, offering greater QALY’s for less cost. Conclusions: The ICER for EBRT is much higher than the accepted willingness-to-pay threshold of $75,000 on the conservative assumption that the utility for no disease after RT is equal for both IORT and EBRT, and that LRR for IORT would more than triple at 10 years. Only a minute improvement in utility for IORT—which is likely since it is a single dose given at time of surgery and avoids multiple visits for EBRT—makes the IORT treatment less expensive and more effective.