Overall, 341 AYAs with Philadelphia chromosome-negative ALL were included in the study. The median age was 23.2 years (IQR 17.2–33.3) with the majority male (n = 212, 62.2%). With a median follow up of 3.8 years (range 0–16), 114 (33.4%) patients experienced at least one VTE event. Patients with vs. without VTE had higher rates of elevated (overweight/obese) BMI (50.9% vs. 39.6%, p = 0.048), lower rates of hyperdiploid karyotype (3.5% vs. 13.2%, p = 0.0038) and were more frequently treated on trial (29.8% vs. 20.3% per protocol, p = 0.049), Table 1. Of note, hyperdiploid karyotype was also associated with younger age and lower BMI (Supplementary Table 1).

In the Consortium cohort, VTE events were documented in 140 instances among 114 patients with events most commonly occurring during consolidation (n = 96, 69%), followed by induction (n = 32, 23%), and uncommonly during continuation/follow-up (n = 12, 8%). Among the 114 patients with VTE, 19 (17%) experienced more than one VTE. In total, 10 patients (8.7% of patients with VTE; 2.9% in the overall cohort) experienced cerebral sinus vein thrombosis (CSVT), with one out of the ten experiencing a second CSVT event. Five of the patients (50%) were rechallenged with ASP and there were no additional VTE events, bleeding or evidence of ALL relapse among re-challenged patients. Further details of the 10 patients with CSVT events are summarized in Supplementary Table 2.

The 1-month, 6-months and 1-year and cumulative incidence of VTE were 6.7% (95% CI: 4.4%, 9.7%), 22.7% (95% CI: 18.4%, 27.2%) and 31.9% (95% CI: 27.0%, 36.9%), respectively. The 2-year cumulative incidence of VTE in the Consortium cohort was 33.5% (95% CI: 28.5%, 38.5%, Fig. 1A) and was higher in patients with overweight/obese vs. normal BMI (39.2% [95% CI: 31.3%, 47.0%] vs. 29.0% [95% CI: 22.7%, 35.6%], p = 0.048, Fig. 1B). In a competing risk regression model, BMI was the only variable associated with higher VTE risk (HR 1.45 [95% CI: 1.01, 2.09], p = 0.047, Supplementary Table 3). Of note, type of ASP (pegylated vs. non-pegylated ASP) did not affect VTE incidence (HR 1.2 [95% CI: 0.82, 1.74], p = 0.35, Fig. 1C).

A Entire cohort. B Stratified by BMI group (normal vs. overweight/obese). C Stratified by type of asparaginase (E. coli vs. peg-asparaginase). VTE venous thromboembolism, BMI body mass index.

Among patients who completed at least 1 year of protocol therapy (n = 220), the 4-year OS was 91.5% [95% CI: 86.7%, 94.6%], and did not differ between patients with vs. without VTE (93.8% [95% CI: 85.6%, 97.4%] vs. 90.0% [95% CI: 83.0%, 94.2%] respectively, p = 0.93, Fig. 2A). Similarly, there were no differences in 4-year EFS (85.1% [95% CI: 75.2%, 91.3%] vs. 86.9% [95% CI: 79.4%, 91.8%], p = 0.63, Fig. 2B), CIR (19.8% [95% CI: 12.2%, 28.8%] vs. 22.6% [95% CI: 16.6%, 29.2%], p = 0.2, Supplementary Fig. 2) and NRM (2.7% [95% CI: 0.7%, 7.0%] vs. 6.9% [95% CI: 3.6%, 11.7%], p = 0.59, Supplementary Fig. 3). In univariate analysis for OS, obese BMI (≥30 vs. <30 kg/m2), higher WBC (as a continuous variable), B-cell immunophenotype, and fewer weeks on ASP therapy were associated with worse OS and included in the multivariable model. All of these covariates except weeks of ASP as a continuous variable remained significant (Table 2). In the uni- and multivariable analyses for EFS, higher WBC, and B-cell immunophenotype were associated with worse EFS (Supplementary Table 4). Notably, VTE was not associated with worse OS or EFS in all regression analyses.

A Overall survival. B Event free survival. OS overall survival, CI confidence interval, VTE venous thromboembolism, EFS event free survival.

In the subgroup treated at DF/HCC (n = 191), 69 patients (36.1%) experienced at least one VTE (total events 82, with 11 experiencing more than one event); those with VTE had a higher rate of elevated BMI (59.4% vs. 43.4%, respectively, p = 0.034), Supplementary Table 5. VTE prophylaxis was given in 71/191 patients (37%), primarily with fixed-dose low molecular weight heparin (LMWH; 69/71, 97%); one person was treated with apixaban and one with fondaparinux. Similar to the Consortium cohort, most events (n = 50, 61%) occurred during consolidation, followed by induction (n = 25, 30%; five patients prior to ASP treatment) and infrequently during continuation/follow-up (n = 7, 9%). With regard to type of VTE, pulmonary embolism (PE) was documented in 27 cases (32.9% of VTE events), CSVT in 7 (8.5% of VTE events) and line-associated VTE in 20 (24.4% of VTE events, Table 3). The 1-month, 6-months and 1-year cumulative incidence of VTE in this cohort were 9.9% (95% CI: 6.2%, 14.7%), 24.2% (95% CI: 18.3%, 30.5%) and 33.8% (95% CI: 27.1%, 40.6%), respectively. The 2-year cumulative incidence of VTE was 36.6% (95% CI: 29.7%, 43.5%, Supplementary Fig. 4A). Like the Consortium cohort, overweight/obese BMI vs. normal BMI was associated with higher cumulative risk of VTE (43.6% [95% CI: 33.4%, 53.4%] vs. 29.6% [95% CI: 20.7%, 39.0%], p = 0.038, Supplementary Fig. 4B). In a Cox univariate regression model for VTE free-time with LMWH prophylaxis as time-varying covariate, both BMI (overweight/obese vs. normal) and immunophenotype (T-ALL vs. B-ALL) were associated with shorter VTE free-time (HR 1.94 [95% CI 1.125-3.35], p = 0.018 and HR 1.62 [95% CI 1.005-2.61], p = 0.048) (Supplementary Table 6). In the multivariable model, only overweight/obese BMI remained significant. Of note, prophylaxis as time-varying covariate did not affect VTE free-time (HR 0.98 [95% CI: 0.56–1.72], p = 0.96). To determine if any subgroup benefited from prophylaxis, we conducted separate regression model for distinct subgroups in an exploratory analysis. The use of prophylaxis was associated with lower risk of VTE in males (HR 0.51 [95% CI 0.26, 0.997], p = 0.049) but not in females (HR 1.24 [95% CI 0.57, 2.67], p = 0.59), Fig. 3. Furthermore, patients with T-ALL and those who were treated with peg ASP who received prophylaxis had lower rates of VTE, although with marginal statistical significance (HR 0.5 [0.24, 1.06], p = 0.07 and 0.61 [95% CI 0.31, 1.1], p = 0.1, respectively). Of note, there was no VTE-reducing benefit of prophylaxis in any of the BMI subgroups.

PPx prophylaxis, HR hazard ratio, CI confidence interval, BMI body mass index, CNS central nervous system, PEG pegylated.