CanAssist-Breast (CAB) is a prognostic test for predicting risk of distant recurrence within five years in hormone receptor positive early stage breast cancer patients. It is unique in that the test uses immunohistochemistry coupled with a support vector machine learning based algorithm to predict risk score and category (High or Low). It has been developed and validated on a mix of Asian and Caucasian patients. The test has been clinically validated in over 1000 retrospective patient samples. In this study, we present for the first time data on the performance of CAB in a single center study from Spain (Vall D’Hebron Institute of Oncology, Barcelona).
Post-surgical FFPE tumor blocks along with patient demographics and clinical follow up data up to a mínimum of five years were obtained from the hospital. CAB was performed on the tumor samples at the CAP and ISO 15189 accredited OncoStem reference laboratory in India. Distant Metastasis Free Survival (DMFS) and Hazard Ratio were used were computed using survival analysis. MedCalc software (Version 18.10.2) was used for all statistical analysis. The negative predictive value (NPV) was computed for establish the accuracy of prediction in the low risk group.
Sixty-two percent of this cohort had stage II disease. Sixty-nine percent and 61% had node negative and Grade 2 disease respectively. The median at onset was 61 years. The DMFS in the low risk category was 98% and 85% for the high risk (P = 0.0032). The Hazard Ratio was 7.04 (95% CI: 1.93-25.73). The NPV was calculated to be 98%. To exclude any confounding effect of chemotherapy, survival analysis was done in the chemotherapy naïve sub-group. In the chemotherapy naïve sub-group, the DMFS for the low risk group was 100% and 85% for the high risk (P = 0.02). These are results of the interim analysis of the study at half way mark and complete study data will be presented at the time ESMO meeting.
CAB performs well in stratifying risk of recurrence in this Spanish cohort with the data matching performance shown earlier with the mix of Asian and Caucasian patients. In the absence of any inter-mediate risk category in CAB, it offers a cost effective alternative to existing prognostic tests providing definitive results to plan treatment of early stage breast cancer patients.
Clinical trial identification
Legal entity responsible for the study
Manjiri M. Bakre.
Has not received any funding.
M. Bakre: Leadership role, Full / Part-time employment, Officer / Board of Directors, CEO and Founder: OncoStem Diagnostics Pvt Limted.