Updated PREDICT model identifies low-risk TNBC patients who can skip chemo

The addition of stromal tumour-infiltrating lymphocytes (sTILs) to the PREDICT prognostic model further improves decision-making regarding chemotherapy for early-stage patients with triple-negative breast cancer (TNBC), particularly in identifying those at low risk who may safely waive chemotherapy, according to a study.

A total of 3,698 patients were analysed, of whom 1,806 received chemotherapy (median age 50 years, tumour size 25 mm, one positive lymph node) and 1,892 did not (median age 55 years, tumour size 20 mm, no positive lymph nodes). Some 741 and 860 deaths attributed to breast cancer occurred within 5 and 10 years, respectively.

PREDICT plus sTILs demonstrated a robust internal‒external validity, similar comparable calibration and discrimination at both time points (5 years: pooled observed-to-expected (O/E) ratio, 0.98, 95 percent confidence interval [CI], 0.69‒1.41; area under the curve [AUC], 0.74, 95 percent CI, 0.72‒0.77; 10 years: pooled O/E ratio, 0.99, 95 percent CI, 0.72‒1.35; AUC, 0.74, 95 percent CI, 0.70‒0.78).

PREDICT plus sTILs, compared with PREDICT alone, detected 19‒60 additional net true low-risk patients and 3‒10 additional net true high-risk patients per 1,000 chemotherapy-naïve participants at the risk thresholds of 8 percent to 15 percent at 10 years.

This study updated PREDICT version 2.3 by analysing women with early-stage TNBC, diagnosed between 1979 and 2017, from two pooled cohorts with sTILs scored according to international guidelines. The updated model, PREDICT plus sTILs, is a Cox regression model with sTILs and the PREDICT prognostic index as predictors and breast cancer-specific survival as the outcome.

The investigators conducted internal‒external validation using leave-one-region-out cross-validation, with calibration assessed by the O/E ratio and discrimination by AUC. They then compared the clinical values of PREDICT plus sTILs and PREDICT via decision curve analysis.