Thromboembolism in ALL Patients 1-45 Years Treated According to the NOPHO-ALL 2008 Protocol

Background. Thromboembolism (TE) contributes to treatment-related morbidity and mortality in Philadelphia-chromosome negative acute lymphoblastic leukemia (ALL) with a reported incidence ranging 4.2-36.7%. No common recommendations for preemptive antithrombotic prophylaxis exist, and identification of ALL patients at high risk of TE, who may benefit from such prophylaxis, is highly warranted.

Methods. We included patients 1-45 years of age diagnosed with either BCP- or T-ALL and treated according to NOPHO-ALL 2008 protocol in Denmark, Sweden, Norway, Finland, Estonia, Lithauania, or Iceland 7/2008-2/2016. Patient demographics, clinical characteristics, ALL biology, therapy, and TE events were prospectively reported to the common Nordic/Baltic ALL database at 3-months intervals throughout the study. Further details on the TE events, incl. treatment and outcome, were subsequently retrospectively collected through questionnaires. End of follow-up for all patients was April 1^st 2017. Of 93 children 1-18.1 yrs of age diagnosed with symptomatic TE in this study, 58 have previously been reported (Tuckuviene, J Thromb Haemost, 2016). In case of TE low molecular weight heparin without truncation of asparaginase is recommended.

Results . A total of 1637 patients with no record of TE and 139 ALL patients diagnosed with first-time TE were included, of whom 90% (N=125) were symptomatic. 102 patients had BCP-ALL (73%) and 37 T-ALL (27%). Male:female ratio 1.7:1. TE patients had a median age of 16y (IQR 6-21); 32% 1-9.9y (N=45), 33% 10-17.9y (N=46), and 35% 18-45y (N=48). 3y cumulative TE incidence was 7.9% (95% confidence interval (CI), 6.6-9.1); 3.9% 1-9.9y (95% CI, 2.7-4.9), 15.6% 10-17.9y (95% CI, 11.5-19.6) and 16.9% 18-45y (95% CI, 12.3-21.2). Time interval between diagnosis of ALL and TE ranged 2-799 days (median 80, IQR 49-127), and 46% of these (N=64) occurred during induction/consolidation 1. 17 patients (12%) had TE at ≥2 sites. 36 cases were cerebral sinus venous thrombosis (26%), 31 cases pulmonary embolism (22%), 54 cases otherwise supra-diaphragmatic (39%), and 36 cases infra-diaphragmatic (26%). 93% were of venous origin. A central venous line was present at the site of thrombosis in 45 patients with symptomatic TE (36%) and in 11 patients with asymptomatic TE (79%). 113 of 139 cases occurred within ≤6 weeks from asparaginase administration, and this treatment was truncated in 24% (N=34) due to TE event. Diagnosis of TE was not associated with increased hazard for mortality or relapse compared to the patients without TE (P 0.29 and 0.15). Risk of TE was significantly increased with presence of mediastinal mass at ALL diagnosis (def: ≥1/3 of thorax width at 5^th thoracic vertebra level) (adjusted TE-specific HR 1.89, P 0.005 (95% CI, 1.21-2.96)), age 10-17.9y (adjusted TE-specific HR 4.16, P <0.0001 (95% CI 2.76-6.28)), and age 18-45y (adjusted TE-specific HR 4.53, P <0.0001 (95% CI, 2.98-6.91)) (Cox proportional hazards model with sex, immunophenotype, CNS status (CNS1-3), (WBC < or > 100K, mediastinal mass, 3 age groups, induction therapy with dexamethasone (T-ALL a/o WBC>100K), and treatment group. In regression model using age groups 1-5.9y, 6-14.9y, 15-20.9y and 21-45y (division into quartiles by TE events) the TE hazard was significantly increased for all 3 older age groups; 6-14.9y (adjusted HR 1.74, P 0.03 (95% CI, 1.07-2.84)), 15-20.9y (adjusted HR 6.38, P <0.0001 (95% CI, 4.00-10.2)), and 21-45y (adjusted HR 4.43, P <0.0001 (95% CI, 2.73-7.20)). The HR of presence of mediastinal mass remained unchanged (adjusted HR 1.92, P 0.004 (95% CI, 1.23-3.00)).

Conclusion . Patients with a mediastinal mass a/o age ≥6-10 yrs at ALL diagnosis constitute a group at high risk for development of TE during chemotherapy with ASP, and routine antithrombotic prophylaxis should be considered for this high risk group.