Fig. 3: Classification of trials with statistical survival, outcome, or response (SOR) comparisons by malignancy with a focus on non-small cell lung cancer (NSCLC) and non-Hodgkin’s lymphoma (NHL).

a Counts of the trials from Fig. 2a by malignancy for the 10 malignancies with the most sex comparisons of SOR; b counts of NSCLC trials from a according to the treatment given; c Analysis of NSCLC SOR trials from a, split by those which used an EGFR inhibitor (EGFRi) and those which did not use an EGFR inhibitor (non-EGFRi). d Classification of NSCLC trials with statistical SOR comparisons by continent; trials on multiple continents were counted once for each continent; no continents have an over-representation of or under-representation of any continent among trials favoring females. e Classification of non-Hodgkin’s lymphoma (NHL) for the five most common treatments in trials with an SOR sex comparison. f Analysis of NHL SOR trials from a, split by those which used rituximab and those which used ‘Other’ non-rituximab treatments. In all panels, color indicates whether these trials show that males have better SOR (blue), females (pink), or that there is no difference (dark gray). Figures a, b and e used two-sided binomial tests to determine whether the proportion of male to female trials was significantly different from 1:1. P-values were corrected for multiple hypothesis testing using the Benjamini–Hochberg method. Figures c and f utilized two-sided hypergeometric tests to assess whether EGFR inhibitor trials or rituximab trials were enriched for improvement in females. Significant FDR values are displayed in the figure. All others are available in Supplementary Tables. Source Data are provided as a Source Data file.