Retrospective analysis of Ki-67 as a predictive and prognostic marker for pathological complete response to neoadjuvant chemotherapy in breast cancer.

Abstract

e12610 Background: According to GLOBOCAN 2020, breast cancer with an estimated 2.3 million new cases (11.7%) is the world's most common cancer. Ki-67 has been established as a proliferation marker in breast cancer. However, there are still some conflicting results on the clinical significance of Ki-67 as predictive and prognostic marker for effectiveness of neoadjuvant chemotherapy. In this retrospective study, Ki-67 is evaluated as predictive and prognostic marker for effectiveness of neoadjuvant chemotherapy (NACT). Methods: Study design: A retrospective observational single institutional study. Place of study: Tata memorial Centre - BBCI, Guwahati. Study period and duration: November 2018 to October 2021 (3 years). Study population: All female patients with biopsy-proven breast cancer. Inclusion criteria: All female patients with the diagnosis of breast cancer, who received NACT. Exclusion criteria: Patients who did not receive NACT, patients who did not undergo surgery after NACT, patients who lost to follow-up after chemotherapy, and patients whose pre-chemo Ki- 67 was not available. Sample size: 2272 patients were reviewed, out of which 127 patients were included in this study. Assuming 30% of patients have a complete response, allowing an error of 8% on these, 127 samples were required for the study at a 95% confidence interval. ROC (receiver operating characteristic) curve analysis was done to know the optimal cut-off value for the Ki-67 labeling index (LI) to discriminate response to treatment. Chi-square test was applied to know the relation between receptor status (ER, PR, HER2neu) and pCR. The pathological response was assessed in reference to Miller-Payne score. Results: Age <40years (22.83%), >40years (77.17%); commonest histology, IDC (95.28%); grade 2 (59.84%), grade 3 (40.16%); ER negative with pCR, 88.2%, p < 0.05; PR negative with pCR, 88.2%, p < 0.05; HER2 positive (treated with trastuzumab) with pCR, 55.8%, p < 0.05; triple negative with pCR, 41.2%, p < 0.05; chemo regimen, 4AC->4T (37%), 4AC->4TH (32.3%); pCR, 26.77% (n=34/127); and Ki-67 LI of 27%, sensitivity 95.4%, specificity 21%. Conclusions: Neoadjuvant chemotherapy is more efficient in tumors presenting with at least 32% Ki-67 expression. A high sensitivity of 95.4% and low specificity of 21% indicate Ki-67 LI is not a reliable predictive and prognostic marker. Other markers like ER, PR, and HER-2 influenced the disease outcome significantly. In our study, ‘PR negative’, ‘ER negative’, ‘HER-2 positive treated with trastuzumab’ and ‘triple negative’ were significantly associated with pCR (p < 0.05). This indicates that pCR rate depends on ER/PR/Triple negativity rates or the rate of HER-2 positivity treated with trastuzumab. Ki67 by itself is poor marker. However, when combined with ER/PR/HER2, it can become predictive of responses at cut off 27% in our study to differentiate the subgroups.