Comparison of proportions and prognostic impact of pathological complete response between evaluations of representative specimen and total specimen in primary breast cancer after neoadjuvant chemoradiotherapy: an ancillary study of JCOG0306.

Background: In JCOG0306 trial, a phase II study to examine the efficacy of neoadjuvant chemotherapy followed by radiation therapy (NAC-RT) to primary breast cancer, pathological complete response (pCR) was evaluated from specimens of the representative cross-section including the tumor center that had been accurately marked [representative specimen (RS) method]. In this ancillary study, we examined if the RS method was comparable to the conventional total specimen (TS) method, which is widely employed in Japan, to identify the pCR group showing excellent prognosis.

Methods: We obtained long-term follow-up data of 103 patients enrolled in JCOG0306 trial. As histological therapeutic effect, pCR (ypT0 and ypT0/is) and quasi-pCR [QpCR, ypT0/is plus Grade 2b (only a few remaining invasive cancer cells)] were evaluated with RS and TS methods. Concordance of pCR between these two methods and associations of the pCR with prognosis were examined.

Results: ypT0, ypT0/is, and QpCR were observed in 28 (27.2%), 39 (37.9%), and 45 (43.7%) patients with RS method, whereas these were 20 (19.4%), 25 (24.3%) and 40 (38.9%) with TS method, respectively. Between RS and TS methods, concordance proportions of ypT0 and ypTis were 92.2% and 86.4%, respectively. Risk of recurrence of ypT0/is group was lower than that of non-ypT0/is group (HR 0.408, 95% CI [0.175-0.946], P = 0.037) and risk of death of ypT0/is group was lower than that of non-ypT0/is group (HR 0.251, 95% CI [0.073-0.857], P = 0.027). The ypT0 and ypT0/is groups with RS method showed excellent prognosis similarly with those with TS method, and RS method was able to differentiate the OS and RFS between pCR and non-pCR than TS method significantly even if pCR was classified ypT0 or ypT0/is. With TS method, QpCR criteria stratified patients into the better and worse prognosis groupsmore clearly than pCR criteria of ypT0 or ypT0/is.

Conclusions: RS method was comparable to TS method for the evaluation of pCR in the patients who received NAC-RT to primary breast cancer provided the tumor center was accurately marked. As pCR criteria with RS method, ypT0/is appeared more appropriate than ypT0.