Data on the accuracy of intraoperative resection margin assessment and its influence on survival after neoadjuvant therapy (NT) are scarce. In the present study, we analyzed the accuracy of resection margin assessment by intraoperative frozen section (IFS) in the prospectively randomized neoadjuvant clinical trials ABCSG 24 and ABCSG 34.164 patients with early breast cancer who received NT within the ABCSG 24 and ABCSG 34 trials between 2005 and 2015, and received IFS resection margin assessment were included in our analysis. Resection margin status by IFS was compared to subsequent FFPE resection margin status, which was considered the gold standard. According to the rate of FFPE-based false-negative and false-positive diagnoses, we determined sensitivity and specificity of IFS. Correlation of IFS diagnosis with definitive FFPE assessment was described by Cohen´s kappa coefficient, the distributional probability was analyzed by McNemar test. The association of patient- and tumor-related factors (menopausal status, residual tumor size, tumor grade, lymph node status, multifocality, DCIS component, lymph vessel invasion, ER/PR/HER2/Ki67 expression and RCB category) with correct IFS diagnosis was analyzed by chi2 test and Fisher´s exact test. Recurrence-free survival (RFS) and overall survival (OS) were analyzed using Kaplan-Meier method and log-rank test. Median age of patients was 55 years. IFS diagnosis for negative margins (no tumor on ink) was correct in 131 (80%) cases, translating into a sensitivity of 0.571 and specificity of 0.877 (false negative rate: 43%, false positive rate 12%). Correlation coefficient for IFS and FFPE diagnosis yielded a Cohen´s kappa value of 0.459. 58 patients (35%) were re-excised during the first operation, in 27 patients (17%), a second surgical intervention was performed. Overall re-excision rates in patients with true positive, false positive and false negative IFS diagnoses were 96%, 93% and 78%, respectively. Rate of second surgical interventions in patients with true positive, false positive and false negative IFS diagnoses were 38%, 7% and 44%, respectively. Correct IFS diagnosis correlated with smaller residual tumor size (p=0.032), lymph node status (p=0.007), HER2 negativity (p=0.049), presence of a DCIS component (p=0.013) and absence of lymph vessel invasion (p=0.007). Despite of re-operation resulting in negative margins, false negative IFS diagnosis was associated with worse recurrence-free survival (p=0.0278). The impact of IFS diagnosis on OS did not reach statistical significance (p=0.0995), although patients with false negative IFS diagnosis showed a shorter OS. IFS assessment of margins after NT displayed high specificity and limited sensitivity in the ABCSG 24 and ABCSG 34 trials. Residual tumor size, lymph node status, HER2 positivity, presence of a DCIS component and lymph vessel invasion correlated with accuracy of IFS. Patients with false negative IFS margin status showed a significantly shorter RFS despite of a second intervention yielding negative margins. Larger studies are needed to confirm the effects of IFS margin assessment on OS. Our data could help to lower rates of subsequent surgical interventions for re-excision after NT and identify patients by the above mentioned tumor characteristics who are at an increased risk of recurrence despite negative margin status.

Citation Format: Karin Rokitte, Kristina Tendl-Schulz, Ulrike Heber, Kerstin Wimmer, Rupert Bartsch, Stephanie Kacerovsky-Strobl, Georg Pfeiler, Günther G. Steger, Christian F. Singer, Dominik Hlauschek, Michael Gnant, Florian Fitzal, Zsuzsanna Bago-Horvath. Accuracy and predictive value of resection margin assessment by intraoperative frozen section after neoadjuvant therapy: An analysis of the ABCSG 24 and 34 trials [abstract]. In: Proceedings of the 2021 San Antonio Breast Cancer Symposium; 2021 Dec 7-10; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2022;82(4 Suppl):Abstract nr P1-02-07.