Negative CD19 Expression Is Associated with Inferior Relapse-Free Survival in RUNX1-RUNX1T1-Positive Acute Myeloid Leukemia; The Japanese Pediatric Leukemia/Lymphoma Study Group Experience from the AML-05 Study

Background: Acute myeloid leukemia (AML) with RUNX1-RUNX1T1 and CBFB-MYH11 have been recognized as core-binding factor (CBF) AML accounting for 25% of the pediatric AML patients. CBF-AML patients have been considered to have a good prognosis, but 30 - 50 % of the RUNX1-RUNX1T1-positive AML patients experience relapse. This finding suggests that some population of them have risk factors associated with poor outcome. Previous studies revealed that KIT activating mutation as a predictor of poor outcome. In addition, previous studies of relatively small number of patients also revealed that CD56 positivity or CD19 negativity were also poor prognostic factors. However, the relationship between KIT activating mutation and specific pattern of cell surface antigens has not been fully investigated.

Aim and methods: We performed a retrospective analysis of RUNX1-RUNX1T1-positive AML patients treated in the Japanese Pediatric Leukemia/Lymphoma Study Group (JPLSG) AML-05 protocol to determine risk factors of relapse using the integration of data including pattern of cell surface markers on leukemic cell at diagnosis, genetic abnormalities, and clinical characteristics. Flow cytometric analysis of immunophenotyping was performed in the central laboratories using same panel of antibodies. Conventional cytogenetic analysis using G-banding was performed as part of the routine work-up. Molecular study using quantitative RT-PCR for the detection of RUNX1-RUNX1T1 and PCR for the detection of FLT3-ITD was also performed as part of the routine work-up. Screening of mutation of 8 genes, such as NRAS, KRAS, KIT, WT1, C/EBPA, ASXL1, ASXL2, and CSF3R, was performed by genomic PCR and Sanger sequencing. A total of 106 AML patients enrolled in the JPLSG AML-05 study were RUNX1-RUNX1T1-positive AML, but we could not obtain the data of cell surface marker in 6 of them. Thus, we analyzed 100/106 (94.3%) patients with RUNX1-RUNX1T1-positive AML. Statistical analysis was performed by Kaplan-Meier method with log-rank test. A Cox proportional hazards model was used to determine risk factors for survival and relapse. Results are reported as adjusted odds ratios with 95% confidence intervals. Statistical significance was set at p < 0.05.

Results: In entire study population, 8 of the 100 patients died and 24 of the 100 patients experienced relapse, respectively. The 3-year overall survival (OS) and relapse-free survival (RFS) rates were 91.7 % (95 % CI; 83.2 - 96.0) and 69.5 % (95 % CI; 59.0 - 77.9), respectively. In terms of genetic analysis, 21 / 100 (21.0%) patients had KIT exon 17 mutation, 11/100 (11.0%) had KIT exon 8 mutation, 6/100 (6%) had KRAS mutation, 16/100 (16%) had NRAS mutation, 2/100 (2%) had C/EBPA mutation, 2/100 (2%) had WT1 mutation, 6/100 (6%) had ASXL1 mutation, 9/100 (9%) had ASXL2 mutation, and 6/100 (6%) had CSF3R mutation. FLT3-ITD was also identified in 3 of the 100 (3%) patients. In terms of cell surface marker expression pattern, CD19 expression was negative in 59 / 100 (59.0%) patients, and CD56 expression was positive in 43 / 100 (43.0%) patients. Patients with KIT exon 17 mutation were significantly accumulated in the CD19 negative (CD19 (-)) population (18 / 59 vs. 3 / 41, p < 0.001). Survival analysis revealed that KIT exon 17 mutation and CD19 (-) were associated with inferior 3y-RFS (KIT exon 17 mutation: negative vs positive, 74.6 vs 50%, p<0.01, CD19 expression: positive vs negative, 83.1 vs 59.8 %, p<0.01). On the contrary, CD56 (+) was not associated with poor RFS (positive vs negative, 66.1 vs 74.7%, p=0.15). In addition, CD19 (-) was associated with poor RFS even in the patients without KIT exon 17 mutation (positive vs negative, 85.4 vs 65.2%, p=0.04). Finally, CD19 (-) was the sole significant risk factor of relapse (hazard ratio; 3.09, 95% CI; 1.26 - 7.59, p < 0.01) by multivariate analysis.

Discussions: This study revealed that CD19 negativity might be a distinct character of poor prognostic subgroup of RUNX1-RUNX1T1-positive AML. On the contrary, CD56 positive was not poor prognostic factor of RUNX1-RUNX1T1-positive AML. Although majority of the patients with KIT exon 17 activating mutation belonged to CD19(-) patients, CD19 (-) patients without KIT exon 17 mutation still showed inferior RFS, suggesting biological difference between CD19 (+) and CD19 (-) RUNX1-RUNX1T1 AML should be investigated.