Fig. 5: Relationships among measures of intratumor genetic heterogeneity and their associations with outcome.
From: Inferring early genetic progression in cancers with unobtainable premalignant disease

a, Box plots of MATH scores versus aneuploidy class for N = 522 total participants; red lines indicate the median, whiskers extend to the extreme values within 1.5× IQR of the box delimiting the first and third quartiles. b,c, Individual tumor MATH scores versus the FGA. Colors represent tumor aneuploidy classes. A Student’s t-test for correlation (bivariate data assumed but not tested) was conducted to obtain P values. Data are displayed for HPV– (b) and HPV+ (c) tumors. d, Timing relationship between MATH and WGD. For tumors with timing of WGD, MATH calculated from post-WGD SNV (vertical axis) is plotted against that calculated from pre-WGD SNV (horizontal axis). e–g, Associations of intratumor heterogeneity measures with outcome. Top, Kaplan–Meier survival curves (with 95% log survival CIs) stratified by a heterogeneity measure for the participant subset with survival most associated with intratumor heterogeneity (surgery without adjuvant therapy or therapy involving chemoradiation; excluding HPV+ oropharyngeal tumors to avoid confounding with HPV; N = 153 participants). Numbers at risk are shown over time. Bottom, HR point estimates (with 95% Wald CI) from Cox multiple regression model on 441 participants. The model was stratified by anatomic site and included interactions of the heterogeneity measure with therapy received and with high-risk pathology (evidence of close or positive surgical margins or extranodal tumor extension). HRs are displayed for HPV+ oropharyngeal tumors versus others (HPV+ in OP), smoked within 15 years of diagnosis (recent smoker), 75th and 25th percentiles of age (Age-69:53), T classification greater than 2 (T > 2), N classification greater than 1 (N > 1), presence of high-risk pathology as defined above (high-risk pathology) and surgery without adjuvant radiation and surgery with adjuvant radiation alone versus those receiving chemoradiotherapy (CR) as primary therapy or adjuvant to surgery (Surgery: CR and Adjuvant radiation: CR), with aneuploidy as the heterogeneity measure (e), with high MATH (MATH > 32.7) as the heterogeneity measure (f) and with FGA as the heterogeneity measure (g); high FGA cutoff is at the same percentile among tumors as for MATH.