Article Text
Abstract
Objectives The aim of this study was to assess the validity of the online PREDICT tool in a population-based cohort of intermediate risk luminal breast cancer.
Methods Among the cohort of breast cancer patients (n=962) treated between 2011–2017, 127 patients considered with intermediate risk RH+/HER2- tumors treated with adjuvant therapy were selected. Patients had at least one factor: 1–3pN+, >2cm, SBR II-III. Observed 5-year overall survival were estimated using the Kaplan–Meier method, and compared with predicted outcomes using PREDICT UK 2.1, in the overall population and in several subgroups.
Results Median age at diagnosis was 51 years old, median tumor size was 28 cm. Node positive disease was observed in 68.5% of cases, grade III in 26.8%, median ki67 was 27. Overall, the PREDICT tool underpredicted 5-year OS by -6.6% (80.8%, 95%CI[70.8%-90.84%] vs 87.4%, 95%CI[86.4%%-92.4%]). This underestimated difference was observed among several subgoups: in pN1–3 group it was -6.4% (78.6% [68.1%-89.1%] vs 85%[81.1%-89.8]), in menopausal women it was -7.9 (77.4% [67.3%-87.4%] vs 85.3% [75.3–95.3]) and it patients who received chemotherapy it was -8.6% (80.9% [71.3%-90.5%] vs 89.5 [86.4%-92.6]). On the other hand, the PREDICT overestimated survival in younger patients ≤40 years old by +6.1% (78.5%, 95%CI [68.5%-88.5%] vs 84.6% 95%CI [75.9%-93.2%]). The ROC analysis of PREDICT showed a medium discrimination value with an AUC of 0.61 (95% CI: 0.51–0.73).
Conclusions PREDICT UK 2.1 showed an under estimation of the 5-year survival of -6.6%, conversely it overestimated it in younger patients by +6.1%. These results highlight the challenge of survival evaluation in RH+/HER2- intermediate risk breast cancer.