RAS Pathway Mutations As Predictive Biomarker for Treatment Adaptation in Pediatric B-Cell Precursor Acute Lymphoblastic Leukemia

Background

Despite significant improvements in the outcome of children with B-cell precursor acute lymphoblastic leukemia (BCP-ALL), therapeutic strategies for high risk and relapsed patients are limited and cause severe side effects. Reliable risk assessment and new therapeutic targets with high specificity are therefore warranted. The RAS pathway is the most frequently mutated pathway in cancer, and the RAF-MEK-ERK kinase axis is crucial for mediating the oncogenic effects of RAS. We and others have previously shown that in pediatric BCP-ALL, RAS pathway mutations can be retrospectively linked to relapse and chemotherapy resistance. However, data on the frequency of (sub-)clonal mutations at diagnosis and hence information about the prognostic relevance at initial diagnosis is lacking.

Aim

Guide therapy adaptation in pediatric BCP-ALL by evaluating the prognostic relevance of RAS pathway mutations and investigating the sensitivity to MEK inhibition.

Methods

We performed targeted next-generation sequencing of mutational hotspots in 13 RAS pathway genes to determine the frequency and clonality of RAS pathway mutations in a large, clinically and biologically characterized cohort of BCP-ALL patients. Initial diagnosis samples of 461 patients and 19 matched diagnosis-relapse sets were included. Mutations were considered clonal at ≥25% variant allele frequency, and high coverage allowed detection of subclones with down to 1% variant allele frequency. Clinical outcome was evaluated in 244 patients treated according to a contemporary, minimal residual disease (MRD)-based protocol (DCOG ALL10). The evolution of RAS pathway mutations was studied in 19 matched sets from diagnosis and relapse. Ex vivo sensitivity of RAS pathway mutated cells towards chemotherapeutic agents and trametinib was evaluated in an MTT-based cytotoxicity assay.

Results

Variants in RAS pathway genes were observed in 44% of initial diagnosis pediatric BCP-ALL cases, mostly affecting NRAS, KRAS, PTPN11, and FLT3. Clonal and subclonal mutations were found in 24% and 20% of patients, respectively. The mutation frequency was highest in high hyperdiploid, infant t(4;11)-positive, BCR-ABL1-like, and B-other cases (50-70%), whereas mutations were rare in ETV6-RUNX1-positive (27%), TCF3-PBX1-positive (8%) and BCR-ABL1-positive cases (4%). In matched diagnosis-relapse sets, clonal mutations at diagnosis were preserved at relapse, whereas the kinetics of subclones was variable. Interestingly, most RAS pathway mutations at relapse were clonal and exclusive. Cells carrying RAS pathway mutations, especially KRAS G13 mutations, were more often ex vivo resistant to prednisolone and vincristine. No association was found with ex vivo response to daunorubicine, L-asparaginase, 6-mercaptopurine, and 6-thioguanine. Mutant primary leukemic cells were ex vivo sensitive to the MEK-inhibitor trametinib. In addition, trametinib could enhance the cytotoxic effect of prednisolone ex vivo. In DCOG-ALL10 and COALL-97/-03 patients with clonal but not subclonal mutations, MRD levels tended to be more often high compared to wildtype cases (31% vs. 19%, p=0.057), while other risk factors (age, gender, white blood cell count, CNS, prednisone response) where not different. Event-free survival was lower in the standard risk and high risk arms of the DCOG ALL10 protocol (69% vs. 96%, p=0.027 and 56% vs. 100%, p=0.015, respectively).

Conclusions

Collectively, analysis of 461 diagnostic BCP-ALL patient samples identified RAS pathway mutations in 44% of patients, and one out of four carried a clonal mutation. MRD was the only risk factor associated with clonal RAS pathway mutations. MRD is essential to treatment stratification in many contemporary protocols, such as the DCOG ALL10 protocol, where only patients with negative MRD after induction courses are treated with a reduced regimen (standard risk arm). Given their unfavorable event-free survival, therapy should be adapted for mutated patients in future protocols. Since treatment intensification is not feasible for high risk or relapsed cases, addition of MEK inhibitors may be of benefit especially because they enhance the cytotoxicity of prednisolone. RAS pathway mutation status may therefore serve as biomarker to select patients for MEK-inhibitor treatment in new treatment protocols for children with BCP-ALL.