Machine Learning Models and Multiparametric Magnetic Resonance Imaging for the Prediction of Pathologic Response to Neoadjuvant Chemotherapy in Breast Cancer.

BACKGROUND: Most breast cancer (BC) patients fail to achieve pathological complete response (pCR) after neoadjuvant chemotherapy (NAC). The aim of this study was to evaluate whether imaging features (perfusion/diffusion imaging biomarkers + radiomic features) extracted from pre-treatment multiparametric (mp)MRIs were able to predict, alone or in combination with clinical data, pCR to NAC. METHODS: Patients with stage II-III BC receiving NAC and undergoing breast mpMRI were retrospectively evaluated. Imaging features were extracted from mpMRIs performed before NAC. Three different machine learning models based on imaging features, clinical data or imaging features + clinical data were trained to predict pCR. Confusion matrices and performance metrics were obtained to assess model performance. Statistical analyses were conducted to evaluate differences between responders and non-responders. RESULTS: Fifty-eight patients (median [range] age, 52 [45-58] years) were included, of whom 12 showed pCR. The combined model improved pCR prediction compared to clinical and imaging models, yielding 91.5% of accuracy with no false positive cases and only 17% false negative results. Changes in different parameters between responders and non-responders suggested a possible increase in vascularity and reduced tumour heterogeneity in patients with pCR, with the percentile 25th of time-to-peak (TTP), a classical perfusion parameter, being able to discriminate both groups in a 75% of the cases. CONCLUSIONS: A combination of mpMRI-derived imaging features and clinical variables was able to successfully predict pCR to NAC. Specific patient profiles according to tumour vascularity and heterogeneity might explain pCR differences, where TTP could emerge as a putative surrogate marker for pCR.

Adjusting the dose of tamoxifen in patients with early breast cancer and CYP2D6 poor metabolizer phenotype.

BACKGROUND: CYP2D6 is a key enzyme in tamoxifen metabolism, transforming it into its main active metabolite, endoxifen. Poor CYP2D6 metabolizers (PM) have lower endoxifen plasma concentrations and possibly benefit less from treatment with tamoxifen. We evaluated tamoxifen dose adjustment in CYP2D6 PM patients in order to obtain plasma concentrations of endoxifen comparable to patients with extensive CYP2D6 metabolism (EM). PATIENTS AND METHODS: Comprehensive CYP2D6 genotyping and plasma tamoxifen metabolite concentrations were performed among 249 breast cancer patients in adjuvant treatment with tamoxifen. Tamoxifen dose was increased in PM patients to 40 mg and to 60 mg daily for a 4-month period each, repeating tamoxifen metabolite measurements on completion of each dose increase. We compared the endoxifen levels between EM and PM patients, and among the PM patients at each dose level of tamoxifen (20, 40 and 60 mg). RESULTS: Eleven PM patients (4.7%) were identified. The mean baseline endoxifen concentration in EM patients (11.30 ng/ml) was higher compared to the PM patients (2.33 ng/ml; p < 0.001). In relation to the 20 mg dose, increasing the tamoxifen dose to 40 and 60 mg in PM patients significantly raised the endoxifen concentration to 8.38 ng/ml (OR 3.59; p = 0.013) and to 9.30 ng/ml (OR 3.99; p = 0.007), respectively. These concentrations were comparable to those observed in EM patients receiving 20 mg of tamoxifen (p = 0.13 and p = 0.64, respectively). CONCLUSION: In CYP2D6 PM patients, increasing the standard tamoxifen dose two-fold or three-fold raises endoxifen concentrations to levels similar to those of patients with EM phenotype.