“…In de novo AML, results of such studies are hampered by small subject numbers, not providing Hardy–Weinberg equilibrium (HWE), confounding different leukaemia subentities or by selection of inappropriate source materials for genotyping (Ruan et al , ). A case–control study from India, of 141 AML patients and 245 controls, using unspecified blood samples for genotyping showed a significant association of homozygous TP53Arg72 with AML susceptibility [odds ratio (OR) 1·75, 95% confidence interval (CI) 1·14–2·67] (Dunna et al , ). The largest case–control study in AML to date from China using 307 diagnostic bone marrow samples for genotyping and 560 controls showed significantly inferior treatment outcomes in homozygous TP53Arg72 carriers [2‐year overall survival (OS) Pro/Pro and Pro/Arg 45·7% vs. Arg/Arg 16·5%, P = 0·003] but no association with AML development (Shi et al , ).…”