European Research Initiative on CLL (ERIC). (2015)Recurrent mutations refine prognosis in chronic lymphocytic leukemia. Leukemia, 29, 329-336. Del Giudice, I., Marinelli, M., Wang, J., Bonina, S., Messina, M., Chiaretti, S., Ilari, C., Cafforio, L., Raponi, S., Mauro, F.R., Di Maio, V., De Propris, M.S., Nanni, M., Ciardullo, C., Rossi, D., Gaidano, G., Guarini, A., Rabadan, R. & Fo a, R.(2016) Inter-and intra-patient clonal and subclonal heterogeneity of chronic lymphocytic leukaemia: evidences from circulating and lymph nodal compartments. British Journal of Haematology, 172, 371-383. Diop, F., Moia, R., Favini, C., Spaccarotella, E., De Paoli, L., Bruscaggin, A., Spina, V., Cerri, M., Deambrogi, C., Kodipad, A.A., Favini, S., Sagiraju, S., Jabangwe, C., Mauro, F.R., Del Giudice, I., Forconi, F., Cortelezzi, A., Zaja, F., Visco, C., Chiarenza, A., Rigolin, G. Impact of polymorphisms in apoptosis-related genes on the outcome of childhood acute lymphoblastic leukaemiaDespite intensification of therapy, 20% of children with acute lymphoblastic leukaemia (ALL) relapse (Ceppi et al, 2015).The different response to chemotherapy may be partially explained by inherited genetic variants, e.g. single nucleotide polymorphisms (SNPs) and copy number variants (CNVs). However, clinically useful genetic predictors of ALL treatment response are yet to be identified. We investigated the association between polymorphisms in genes encoding drug-metabolizing enzymes or apoptosis-related proteins and early-treatment response, minimal residual disease (MRD), relapse, overall survival (OS) and event-free survival (EFS) in 173 Caucasian children with ALL. Patients were genotyped for CNVs of CYP2D6, GSTT1, GSTM1, UGT2B17 and SULT1A1, and for SNPs of CYP2D6, SULT1A1 and TP53 (see Supporting Information). This is the first report to seek an association between the exact number of copies (0, 1, 2, ≥3) in selected candidate genes and ALL prognosis, and the first study to analyse the TP53 Arg72Pro polymorphism as a possible predictor of treatment response in childhood ALL.As expected for the Caucasian population, around 50% of children with ALL presented a GSTM1 null genotype, whereas most patients had one copy of GSTT1 and UGT2B17, and two copies of CYP2D6 and SULT1A1. The G allele was predominant in CYP2D6 and SULT1A1 SNPs, and C allele was predominant in TP53 SNP (Tables SI and SII).The association analysis between genotypes and earlytreatment response, MRD or relapse revealed no gene dose or SNP effect with any of the genetic models tested. However, interestingly, patients with 1, 2 or ≥ 3 copies of GSTM1 had worse outcome than patients with GSTM1 null genotype, both when genotypes were analysed separately (OS P = 0Á013, Figure S1A; EFS P = 0Á047, Figure S1B) and when all non-null genotypes were grouped together (OS P = 0Á002, Fig 1A; EFS P = 0Á005, Figure S1C). Differences between null and non-null individuals were confirmed in multivariate analysis for OS (hazard ratio [HR] = 16Á51; 95% confidence interval [CI] 2Á13-128Á18; P = 0Á007...