22P - Can pathological complete response serve as a surrogate endpoint for survival?

Date 07 May 2015
Event IMPAKT 2015
Session Welcome reception and Poster Walk
Topics Breast Cancer, Metastatic
Pathology/Molecular Biology
Presenter Nisha Hariharan
Citation Annals of Oncology (2015) 26 (suppl_3): 6-9. 10.1093/annonc/mdv115
Authors N. Hariharan, N. Nair, V. Parmar, R. Hawaldar, P. Padmanaban, V. Vanmali, R. Badwe
  • Breast Oncology, Tata Memorial Hospital Centre, 400012 - Mumbai/IN



Introduction: Currently, molecular subtypes of breast cancer (BC) define outcome. Clinical trials analyse BC by these recognized subtypes for survival. Recent studies have demonstrated a possible correlation between pathological complete response (pCR) and survival. Based on available evidence, the US FDA has accepted pCR as a surrogate marker for disease-free and overall survival to allow early approval for drugs. However a recent meta-anlaysis failed to demonstrate this correlation. In light of the ongoing debate, we analyzed the data at our institution to evaluate if pCR can serve as a surrogate endpoint for survival.

Method: We performed a retrospective analysis of women with locally advanced BC (LABC) from 2000-2010. Clinical and tumor related information was obtained from case files. Standard practice at our institution is to sandwich surgery between chemotherapy. To maintain uniformity in the cohort, only patients who received 4 cycles of anthracycline chemotherapy were evaluated. Analysis was done in three groups viz ER and/or PR positive, Her 2 positive irrespective of hormone receptor status and triple-negative breast cancer (TNBC). pCR was defined as no residual invasive or in situ tumor at the primary tumor site and axillary lymph nodes.

Results: We identified 2455 eligible women, of whom 280 (11.4%) achieved pCR. The median follow up was 43 months. pCR was seen in 108/1350 (8%) patients in the ER/PR positive group, 23/255 (9%) in the Her2neu positive group and in 149/850 (17.5%) in TNBC group. (p <0.001). For pCR versus not, the 5-year DFS in ER/PR positive BC was 89.8% versus 68.6% (p < 0.001), in Her2neu positive BC, was 91.3%versus 63.8% (p = 0.007) and in women with TNBC was 85.9% versus 66.5% (p = 0.000). On comparison between subtypes for impact of pCR on disease-free survival (DFS), the test for heterogeneity between the subgroups was not significant (p =0.7, NS), thus indicating that increased pCR was not synonymous with improved DFS.

Conclusion: TNBC subgroup had significantly better pCR rates, but these did not translate into better outcomes. This suggests that pCR is probably not an appropriate surrogate endpoint for DFS.

Disclosure: All authors have declared no conflicts of interest.