Development of CanAssist Breast
OncoStem Diagnostics has developed an immunohistochemistry-based novel risk classifier called CanAssist Breast (CAB) which combines 5 prognostically relevant biomarkers and 3 clinico-pathological parameters to arrive at probability of distant recurrence within 5 years from diagnosis.
CanAssist Breast is a first of its kind IHC-based prognostic test developed using machine learning-based techniques.
Advantage of advanced machine learning techniques are that they allow assessment of non-linear interactions between biomarkers, which is critical given crosstalk between signalling pathways in cancer.
The CanAssist Breast algorithm produces a risk score ranging from 0-100. A cut-off of 15.5 is applied to stratify patients into either low-risk (score ≤15.5) or high-risk (score >15.5).
CanAssist Breast stratifies patients into 2 distinct risk groups, with no intermediate risk category. There is a significant and clinically relevant difference in recurrence rates between low-risk and high-risk patients.
Clinical Validation
CanAssist Breast has been validated in a retrospective multi-centric clinical study (10+ centres), involving a cohort of close to 1000 patients (85% Asian and 15% Caucasian).
Approximately 40% of the validation cohort was aged 50 or younger, and 66% were Stage II patients, representative of Indian breast cancer patients who tend to be younger and present with more advanced disease than their Western counterparts.
Proportional hazards regression shows CanAssist Breast provides significant prognostic value, superior to that provided by traditional clinico-pathological parameters.
CanAssist Breast also outperforms traditional prognostic markers and tests such as Ki67 and IHC4.
Analytical Validation
CanAssist Breast has undergone rigorous validation to demonstrate precision of the test. The validation strategy was adopted from various international guidelines to assess impact of multiple variables like operator, observer, run, etc.
All IHC protocols used in CanAssist Breast testing are extensively validated and involve stringent QC.
Subjectivity attributable to IHC scoring was evaluated and found to be minimal among OncoStem’s panel of experienced pathologists.
CanAssist Breast is CE marked. It is performed at the reference laboratory at Bangalore, India which is accredited by College of American Pathologists (CAP) and ISO 13485 certified.