Early identification of thrombosis risk in adult acute lymphoblastic leukemia Free

Background: Thrombosis is a known complication in patients with acute lymphoblastic leukemia (ALL) that potentially impacts morbidity, treatment intensity, and survival. While risk factors are well characterized in pediatric populations, data in adults remain limited. We aimed to identify clinical and laboratory predictors of thrombosis in adults with newly diagnosed ALL treated at a tertiary referral center.

Methods: We conducted a retrospective analysis of adult patients with newly diagnosed ALL treated between 2017 and 2025 at a referral center in Mexico City. Thrombotic events were radiologically confirmed in the context of clinical suspicion. Variables at diagnosis were assessed using logistic regression. Clinically relevant thresholds for continuous variables were defined using receiver operating characteristic (ROC) curve analysis. The cumulative incidence of thrombosis at 30 and 90 days from diagnosis was estimated.

Results: A total of 164 patients were included. Median age was 33.5 years (IQR: 23.0–48.0). At diagnosis, the median D-dimer was 4,669.5 ng/mL (IQR: 1,784.3–9,419.5), white blood cell (WBC) count was 13.75 × 10⁹/L (IQR: 2.58–89.25), and platelet count was 40 × 10⁹/L (IQR: 17.8–91.5). A pediatric-inspired, asparaginase-based regimen was used in 74.4% of patients. Comorbidities included diabetes (16.5%), and hypertension (17.1%). Thrombotic events occurred in 28 patients (17.1%), with 53.6% during induction, and 67.8% during the first 3 months. Most common sites were deep vein thrombosis (35.7%), pulmonary embolism (25.0%), and portal vein thrombosis (17.9%). Two cases of cerebral venous thrombosis were documented, both occurring in patients treated with asparaginase. Only one thrombosis, a portal venous system event, led to discontinuation of asparaginase therapy. Median time from diagnosis to thrombosis was 46 days (IQR: 21–99), and median follow-up was 15.8 months (IQR: 5.1–32.6).

In multivariable analysis, two variables were independently associated with thrombosis: D-dimer >19,511 ng/ml (OR 7.04; 95% CI 1.15–51; p=0.039), and platelet count ≥106,000/µl (OR 8.12; 95% CI 1.46–55.7; p=0.021). Other variables included in the model such as age ≥55 years (p=0.261), WBC >179,500/µl (p=0.159), body mass index >31.25 kg/m² (p=0.534), diabetes (p=0.213), hypertension (p=0.498), fibrinogen >420 mg/dl (p=0.855), and sepsis during induction (p=0.329) were not significantly associated after adjustment. Treatment with a pediatric-inspired asparaginase-based regimen was not associated with an increased risk of thrombosis (p=0.106).

The cumulative incidence of thrombosis was 6.2% (95% CI: 2.4–9.9%) at 30 days and 12.4% (95% CI: 7.0–17.4%) at 90 days from diagnosis. At 30 days, patients with platelet counts above the ROC-defined cutoff had a cumulative incidence of 11.6% (95% CI: 0.2–21.7%), compared with 4.8% (95% CI: 1.0–8.5%) in those below the cutoff; by 90 days, the incidences increased to 20.9% (95% CI: 5.8–33.6%) and 10.1% (95% CI: 4.5–15.3%), respectively. For D-dimer, the 30-day cumulative incidence was 15.4% (95% CI: 0.0–32.9%) in patients above the cutoff and 6.1% (95% CI: 0.1–11.8%) in those at or below it; by 90 days, these incidences rose to 26.0% (95% CI: 0.0–47.8%) and 11.3% (95% CI: 3.0–18.9%), respectively. Having at least one of these two factors at diagnosis was associated with nearly a 3-fold increased hazard of thrombosis compared with having neither (HR 2.9; 95% CI: 1.11–7.55; p=0.029).Conclusions: Thrombosis occurred in nearly one in five adult patients with ALL, primarily during induction. Markedly elevated D-dimer, and platelets at diagnosis were independently associated with thrombotic events. These laboratory findings may help in identifying high-risk patients who could benefit from closer monitoring and individualized thromboprophylaxis strategies. Importantly, thrombotic events rarely led to treatment discontinuation.