Extended Data Fig. 1: Palbociclib plus tamoxifen activity against M/D-driven mammary carcinomas is not modulated by CD4+ and CD8+ T cell depletion.
a-f, C57BL/6 J mice bearing M/D-driven tumors were randomly allocated to receive no treatment (control), tamoxifen (T), oral palbociclib (P), focal radiation therapy (RT) followed by (→) T, RT→P (a-c), P + T ± α-CD4 and α-CD8 antibodies (d), or RT→P + T ± an α-PD-1 antibody (e, f). Individual growth curves for primary tumors shown as tumor area relative to the area at the start of treatment (A/A0) (a, e), cumulative relative tumor area (d), survival curves (b, f), and pie charts with drug responses classified at d20 (c), are reported. Pie charts show the percentage of dead mice, objective response (OR; A/A0 ≤ 0.756), stable disease (SD; 0.756 < A/A0 ≤ 1.440), and progressive disease (PD; A/A0 > 1.440), according to RECIST-based criteria. Number of mice, risk ratio (RR) with 95% confidence interval and P values are shown. g, Bubble heatmap of immune cell cluster-defining genes by scRNAseq from M/D-driven tumors treated as reported in Fig. 1f. Data show the mean expression of genes in identified cell clusters and the fraction of cells expressing the gene of interest in 40,853 CD45+ cells (pooled from 27 tumors, with 3 biological replicates/treatment group obtained from independent mice). h, Gini index of TCR γ and δ chain diversity in M/D-driven tumors left untreated (n = 5) or treated with P + T (n = 7) and harvested at endpoint. Box-plots show means ± interquartile range; whiskers extend to 1.5-fold interquartile ranges. P values were determined by linear mixed effects model followed by simultaneous tests of general linear hypotheses (a, d, e), two-sided Gehan-Breslow-Wilcoxon test (b, f), and two-sided Wilcoxon test (h). Part of the results shown in (b) and (f) are also depicted in Fig. 1b, e (control and RT→P + T), and Figs. 2d, 3f, 4i (control).
