The Probability of Locoregional Control in Patients With Locoregional Recurrent Breast Cancer Treated With Postoperative Reirradiation and Hyperthermia (RADHY): A Continuous Thermal Dose-effect Relationship.

Abstract

Purpose: Mild hyperthermia (HT) (39-43 °C) combined with reirradiation is considered for patients with locoregional recurrent (LRR) breast cancer. Studies analyzing dichotomized HT thermal dose (TD) parameters suggest that higher TD correlates with better response rates, but evidence quantifying optimal TD levels needed to achieve locoregional control (LRC) is limited. We investigated the continuous TD-effect relationship of LRC in patients with LRR breast cancer treated with postoperative reirradiation and HT.

Methods and materials: In this historical cohort study, 112 patients with LRR breast cancer were treated in 2010-2017 with postoperative reirradiation 8 × 4 Gy (n = 34) or 23 × 2 Gy (n = 78) and 4 to 5 weekly HT sessions, TD was measured using invasive thermometry in the target region. Primary endpoint was the estimated probability of LRC at 5-years. The logarithm of highest ("Best") CEM43T50 (median cumulative equivalent minutes at 43 °C) of all HT sessions was analyzed as TD parameter based on Weibull univariate and stepwise multivariate regression analyses. Additionally, the best fitted Bayesian LRC survival model was analyzed assuming 3 informative priors: age, tumor location (breast/chest wall), and lymph node involvement.

Results: Twenty-four patients developed an infield recurrence; median time to recurrence was 3.4 years (interquartile range, 2.7-4.6 years). Increasing median Best session CEM43T50 TD range from 0.08 to 101.9 minutes was associated with increasing probability of LRC from ∼44% to 94% at 5-years, and over this range a 2-fold TD increase resulted in ∼5% to 10% increasing LRC. The hazard ratio for a subsequent recurrence decreased 48% (95% confidence interval, 18%-84%) with a 2-fold increase in TD over the TD range, P = .001. This effect was confirmed in Weibull multivariate regression analysis and in Bayesian LRC survival regression analysis.

Conclusions: Increasing TD was strongly associated with an improved LRC, showing that adequate TD must be ensured and confirming that HT is essential for strongly sensitizing efficacy of postoperative reirradiation for patients with LRR breast cancer.