Clinical characterization and prognosis of T cell acute lymphoblastic leukemia with high CRLF2 gene expression in children

It has been reported that overexpression of the CRLF2 gene is associated with poor outcomes in pediatric B cell acute lymphoblastic leukemia (B-ALL), but the incidence rates, clinical characteristics and outcomes of CRLF2 gene overexpression in pediatric T cell ALL (T-ALL) have not been systematically analyzed. In this study, CRLF2 mRNA expression levels and clinical and laboratory parameters in 63 pediatric T-ALL patients were detected at the Children's Hospital of Chongqing Medical University and Children’s Hospital of Xianyang between February 2015 and June 2018. The patients were treated according to the modified St. Jude TXV ALL protocol, and early treatment responses (bone marrow smear and MRD level) and prognoses in the enrolled patients were assessed. CRLF2 overexpression was detected in 21/63 (33.33%) patients. Statistical differences were not found for clinical or laboratory parameters (including sex, age, initial WBC count, incidence mediastinal involvement, abnormal karyotype and fusion genes) between patients with high CRLF2 expression and patients with low expression of CRLF2 (P>0.05). One patient died of tumor lysis syndrome and renal failure, and the treatment response was monitored on day 19 (TP1) of remission in 62 patients. One patient quit treatment because of family decisions, and 61 patients underwent treatment response evaluation on day 46 (TP2) of remission. Significant differences were not found between patients with high CRLF2 expression and patients with low CRLF2 expression in terms of the treatment responses at TP1 or TP2 (P>0.05). Following October 2018, 12 patients among the 61 evaluable patients relapsed (relapse rate: 19.67%), 3 patients died from chemotherapy, and the treatment-related mortality (TRM) rate was 4.92%. Secondary tumors occurred in 1 patient. The 3-year prospective EFS rate was 54.1±11.2% and 77.7±6.6% for the 61 evaluable patients and 58 patients without TRM. Patients with low CRLF2 expression had longer EFS durations than patients with high CRLF2 expression (61 evaluable patients: 35.91± 2.38 months vs 23.43± 2.57 months; 58 patients without TRM: 37.86± 2.08 months vs 24.55±2.43 months, P<0.05). CRLF2 expression levels were also monitored in 13 patients at TP1 and TP2, and the MRD level did not vary with the CRLF2 expression level. Our data suggest that clinical features, laboratory findings and treatment responses in the pediatric T-ALL population do not vary based on the overexpression of CRLF2 but that CRLF2 overexpression can contribute to a high risk of relapse in pediatric T-ALL. Thus, CRLF2 expression levels should not be used as biomarkers for monitoring MRD.