In a study reported in Journal of Clinical Oncology, Conforti et al found that invasive disease–free survival with sufficient follow-up may be the best surrogate for overall survival in neoadjuvant randomized trials in early breast cancer.
Study Details
The study involved 11 trials with 15 treatment comparisons among 12,247 patients.
Key Findings
The association between hazard ratios (HRs) for overall survival and odds ratios for pathologic complete response was weak overall (R2 = 0.07, 95% confidence interval [CI] = 0.00–0.48), as well as in the breast cancer subgroups evaluated.
Overall, the R2 for the association between overall survival HR and invasive disease–free survival HR was 0.46 (95% CI = 0.08–0.71), approaching the cutoff value for moderate surrogacy of 0.5.
For the majority of subgroups examined, the R2 for overall survival HR and invasive disease–free survival HR ranged from 0.5 to < 0.7; surrogacy was stronger (R2 ≥ 0.7) for the hormone receptor–negative/HER2-negative subtype, histologic grade 1–2 tumors, and lobular tumors. The surrogacy value of invasive disease survival for overall survival was affected by length of follow-up, with substantial increases in R2 observed up to 36 months and little increase observed after 48 months.
However, surrogacy of invasive disease–survival for overall survival was poor in hormone receptor–positive disease. The investigators stated, “This may be due to the long post-recurrence [overall survival] of patients with [hormone receptor–positive] disease, which hinders the possibility to capture treatment effects on [overall survival] in the [follow-up] framework of neoadjuvant trials. Notably, we observed no association between [invasive disease–free survival and overall survival in this subgroup] despite a median [follow-up] of 70 months….”
Conclusion
The investigators concluded: “…[O]ur results support [invasive disease–free survival], with sufficient [follow-up], as an acceptable surrogate end point to anticipate final [overall survival] results in neoadjuvant [randomized trials] for early [breast cancer]. This holds true across many subgroups, except for [hormone receptor–positive] disease. There is a need to reassess whether [overall survival ] is a valid end point for treatment efficacy measurement in hormone receptor–positive early [breast cancer], given the long survival expected even in patients experiencing distant relapse.”
Fabio Conforti, MD, Medical Oncology Unit, Humanitas Gavazzeni, Bergamo, Italy, is the corresponding author for the Journal of Clinical Oncology article.
Disclosures: For full disclosures of all study authors, visit ascopubs.org.
