Reproducibility of residual cancer burden for prognostic assessment of breast cancer after neoadjuvant chemotherapy

Abstract

The residual cancer burden index was developed as a method to quantify residual disease ranging from pathological complete response to extensive residual disease. The aim of this study was to evaluate the inter-Pathologist reproducibility in the residual cancer burden index score and category, and in their long-term prognostic utility. Pathology slides and pathology reports of 100 cases from patients treated in a randomized neoadjuvant trial were reviewed independently by five pathologists. The size of tumor bed, average percent overall tumor cellularity, average percent of the in situ cancer within the tumor bed, size of largest axillary metastasis, and number of involved nodes were assessed separately by each pathologist and residual cancer burden categories were assigned to each case following calculation of the numerical residual cancer burden index score. Inter-Pathologist agreement in the assessment of the continuous residual cancer burden score and its components and agreement in the residual cancer burden category assignments were analyzed. The overall concordance correlation coefficient for the agreement in residual cancer burden score among pathologists was 0.931 (95% confidence interval (CI) 0.908-0.949). Overall accuracy of the residual cancer burden score determination was 0.989. The kappa coefficient for overall agreement in the residual cancer burden category assignments was 0.583 (95% CI 0.539-0.626). The metastatic component of the residual cancer burden index showed stronger concordance between pathologists (overall concordance correlation coefficient=0.980; 95% CI 0.954-0.992), than the primary component (overall concordance correlation coefficient=0.795; 95% CI 0.716-0.853). At a median follow-up of 12 years residual cancer burden determined by each of the pathologists had the same prognostic accuracy for distant recurrence-free and survival (overall concordance correlation coefficient=0.995; 95% CI 0.989-0.998). Residual cancer burden assessment is highly reproducible, with reproducible long-term prognostic significance.

Figure 1

Figure 1A) Hazard rates with 95% confidence intervals for prediction of distant recurrence-free survival for pathologists A to E. The hazard rates correspond to 1 unit increase of the continuous RCB score. Figure 1B) Hazard ratios for the categorical RCB classes. The risk for pathologic complete response (pCR)/RCB-I (grey) and RCB-II (black) is reported relative to that for RCB-III. The analysis was adjusted for hormone-receptor status

Figure 2

Concordance of predicted distant recurrence-free survival from the continuous RCB score. The predicted survival proportions per year of observation are derived from Cox models with adjustment for hormone-receptor status and plotted pairwise for the different observers.

Figure 3. Kaplan-Meier plots for distant recurrence-free survival for RCB classes determined by observers A-E