Clinical and molecular prognostic factors in newly diagnosed pediatric T-cell lymphoblastic lymphoma: a prospective, multicenter, single-arm phase 2 clinical trial.

Abstract

Background: Poor early treatment response in T-cell lymphoblastic lymphoma (T-LBL) is associated with an unfavorable prognosis. This multicenter prospective study evaluated the efficacy of the response-adjusted Chinese Children's Cancer Group (CCCG-LBL-2016) protocol for pediatric T-LBL and examined clinical and molecular prognostic factors.

Methods: Clinical and laboratory data from seven pediatric oncology centers were analyzed. A sub-cohort of 23 patients underwent exploratory integrated genomic analysis, including targeted next-generation sequencing, RNA sequencing, and copy-number array analysis. Survival was evaluated using the Kaplan-Meier method, and prognostic factors were analyzed using multivariable Cox proportional hazards regression.

Results: A total of 163 patients (median age: 108 months; 116 males, 47 females) were enrolled, most with advanced disease (stage III: 81.0%; stage IV: 17.8%). Patients were stratified into the low-risk (R1, n = 2) and intermediate-risk groups (R2, n = 161); thirty one patients in the R2 group were escalated to the high-risk intensified regimen (R3) due to poor early response. The 3-year overall survival (OS) was 78.6% ± 3.3% and event-free survival (EFS) was 73.9% ± 3.5%. Outcomes differed by risk group (P < 0.05), with 3-year OS and EFS of 100% and 100% in R1, 82.6% ± 3.4% and 79.5% ± 3.4% in R2, and 58.6% ± 9.1% and 48.3% ± 9.1% in R3. Progression or recurrence occurred in 42 patients (median: 7 months; 3-year OS: 17.1% ± 6.3%). Clinical risk factors included R3 assignment and elevated lactate dehydrogenase. In the exploratory molecular sub-cohort, recurrent alterations included CDKN2A (39.1%), NOTCH1 (26.1%), FBXW7 (21.7%), and MTAP/PIK3R1/NRAS (13.0%). Exploratory multivariable Cox regression analysis identified that CDKN2A alteration was associated with an increased risk of progression or recurrence (hazard ratio = 35.89, 95% confidence interval: 3.07-419, P = 0.004).

Conclusions: Adjusting the risk stratification based on treatment response significantly improved the overall prognosis of T-LBL. However, survival rates remain very low among patients who experience disease progression or recurrence. The preliminarily explored molecular genetic risk factors might contribute to further risk stratification and provide potential therapeutic targets.