Fig. 5: Combining genetic and morphological measurements.
From: Tumor evolution metrics predict recurrence beyond 10 years in locally advanced prostate cancer
a, mPGA is associated with higher continuous Gleason (n = 106 participants, IMRT participants with three or more samples with a PGA of ≥0.01, linear model, two-sided t-test on gradient, estimate = 0.19, s.e. = 0.04, t = 4.4, d.f. = 104). Shaded area represents 95% CI in all scatter plots. b, Twenty-four chromosome arm changes are associated with a change in continuous Gleason (gains are displayed in red, and losses are displayed in blue; n = 62 chromosome arm changes, P values were adjusted using the Benjamini–Hochberg method and are derived from two-sided t-tests on gradient per arm linear mixed effects model; continuous Gleason change derived from gradient estimate). c, The TP53 mutation is associated with higher continuous Gleason (linear mixed effects model, two-sided t-test on gradient, s.e. = 0.06, d.f. = 371, t = 5.1, n = 503 samples). Box plots show center lines as the median and box limits as upper and lower quartiles. Whiskers extend no further than 1.5× interquartile range past the box limits, and points represent outliers. d,e, mPGA (linear model, two-sided t-test on gradient, estimate = 0.23, s.e. = 0.103, t = 2.2, d.f. = 85; d), but not Spearman (estimate = −0.05, s.e. = 0.15, t = −0.3, d.f. = 85; e), is associated with increased mixing of Gleason grades (n = 87 participants, sequencing cohort omitting participants with a Gleason Morisita equal to 0, that is, a homogenous Gleason grade). f, Chromosome 6p loss is uniquely associated with a reduction in Tumor-Immune Morisita (changes are colored and P values were adjusted and derived as in b; n = 62 chromosome arm changes). Samples in b, c and f have a PGA of ≥0.01. g, The most genetically and morphologically heterogeneous tumors are associated with shorter time to recurrence (two-sided log-rank test, χ2 = 13.7, d.f. = 1, n = 106 participants). h, The Joint Diversity metric shows significant association with greater risk of recurrence in a CPH model with clinical covariates. The forest plot shows 95% CI of HRs, and the covariate P values are derived from a Wald test. The HR for Joint Diversity represents the increase in hazard between the 5th and 95th percentile values (within the sequencing cohort, n = 106 participants, *P < 0.05, **P < 0.01, ***P < 0.001). i, Multiplex immunohistochemistry and H&E staining was performed on the same section. Immunohistochemistry experiments were run once following optimization and validation. j, Example of an immune-hot region on matched H&E (left) and multiplex immunohistochemistry (right) images. k, Example of an immune-cold region on matched H&E (left) and multiplex immunohistochemistry (right) images.
