Clinical Impact of KMT2C and SPRY4 Expression Levels in Intensively Treated Younger Adult Acute Myeloid Leukemia Patients

Background:Recent publications suggest important roles of lysine methyltransferase 2C (KMT2C, located on 7q) and sprouty 4 (SPRY4, located on 5q) as candidate genes in leukemogenesis of acute myeloid leukemia (AML). The prognostic impact of the gene expression levels (ELs) of both genes on outcome in AML patients (pts) is currently unclear.

Aim:To evaluate the prognostic impact of KMT2C and SPRY4 expression in correlation to clinical characteristics and genetic abnormalities assessed at diagnosis in a cohort of intensively treated adult AML pts.

Methods: We retrospectively studied 268 AML pts (median age, 48 years; range, 17-60 years) who had been enrolled on 2 AML SHG trials (0295 and 0199, n=148; only normal cytogenetics pts (CN)) and the SAL-AML2003 trial (n=120; only abnormal cytogenetics pts (CA)). Acute promyelocytic and core-binding factor leukemia pts were excluded. Type of AML was de novo in 235 (88%), secondary in 19 (7%) and therapy-related in 14 (5%) of the 268 pts. Regarding baseline characteristics, CN pts had significantly higher white blood counts (WBC; p=0.001) and blast cells in peripheral blood (p=0.02) as compared to CA pts; all other factors were comparable. Cytogenetic analyses could be performed in 263 (98%) of the 268 pts. Cytogenetic risk classification according to ELN guidelines was intermediate-II in 55 (47%) and adverse in 63 (53%) of the CA pts, respectively. Abnormalities (abn) of 5q were present in 21 (18%) and abn of 7q in 16 (14%) of the CA pts. NPM1 and FLT3-ITD were analyzed in 145 (98%) of the CN pts. Of those, 59 (41%) were only NPM1 positive (pos), 12 (8%) were only FLT3-ITD pos, 34 (23%) were double pos and 40 (28%) were double negative (neg).

KMT2C and SPRY4 ELs, normalized to ABL1and log2-transformed for analysis, were measured in triplets on cDNA obtained at diagnosis by RT-qPCR. Based on cDNA availability, KMT2C ELs could be analyzed in 143 (97%) of the CN and in all of the 120 CA pts, respectively. SPRY4 ELs could be measured in 30 (21%) of the CN and 107 (89%) of the CA pts, respectively.

Results: KMT2C ELs were significantly lower in CN pts with de novo as compared to secondary AML (p= 0.02), whereas there was no difference in CA pts. No significant association was found for SPRY4 and type of AML. KMT2C ELs were significantly lower in FLT3-ITD pos as compared to FLT3-ITD neg CN pts (p=0.046), whereas there was no difference for SPRY4 ELs between the two groups (p=0.57). In addition, there was a significantly lower KMT2C expression in CN pts with intermediate-I risk as compared to NPM1 pos / FLT3-ITD neg pts (p=0.01). Regarding CA pts, there was no difference of KMT2C or SPRY4 ELs in adverse as compared to intermediate-II risk pts (p=0.08; p=0.20, respectively). When focusing on specific subgroups, KMT2C ELs were significantly lower in abn7q CA pts as compared to those without abn7q (p=0.002), whereas there was no difference of SPRY4 ELs in CA pts with or without abn5q (p=0.27). In univariate analysis higher SPRY4 ELs showed a significant favorable impact on relapse-free (RFS, p=0.03) and a trend towards a beneficial impact on overall survival (OS, p=0.06) for CA patients. A similar effect for KMT2C was not observed (RFS, p=0.96; OS, p=0.92). In subgroup analyses of pts with adverse risk cytogenetics, there was no impact of KMT2C or SPRY4 ELs on RFS (p=0.73; p=0.39) or OS (p=0.49; p=0.46), respectively. The same was true for FLT3-ITD pos CN pts (RFS, p=0.73; p=0.37; OS, p=0.91; p=0.36, respectively). In multivariate analyses on RFS and OS in CA pts including age, gender, KMT2C and SPRY4 ELs, logarithm of WBC, blast cells in bone marrow and cytogenetic risk group as variables, only higher age (OS, Hazard ratio (HR),1.28 per 10 years; 95%-confidence interval (CI): 1.02-1.59; p=0.03) and complex karyotype as compared to intermediate-II risk cytogenetics (RFS, HR: 2.25; 95%-CI: 1.20-4.22; p=0.01; OS, HR: 2.97; 95%-CI: 1.65-5.35; p<0.001) had an adverse impact. An effect of KMT2C or SPRY4 on RFS (p=0.84; p=0.16) or OS (p=0.85; p=0.45) was not found in the multivariate setting. In addition, in a multivariate model on CN pts (risk class according to NPM1 and FLT3-ITD mutational status instead of cytogenetic risk class) neither KMT2C nor SPRY4 had an impact on RFS (p=0.13; p=0.39, respectively) or OS (p=0.36; p=0.56, respectively).

Conclusions:Lower KMT2C and SPRY4 ELs are associated with distinct genetic risk groups. An impact on prognosis was evident in univariable analyses for SPRY4 but not for KMT2C ELs in CA pts.