https://orcid.org/0000-0001-8152-1040
In this issue of Blood Advances, Rodriquez et al1 report findings from an a priori planned subgroup analysis of children with obesity and acute lymphoblastic leukemia or lymphoma (ALL/LL) treated with apixaban vs no anticoagulation during induction therapy in the PREVAPIX-ALL randomized clinical trial (RCT).1,2 In contrast to the full trial in which there was not a significant reduction in venous thromboembolism (VTE), analysis of 82 obese children (median age, 9-10.5 years) in this report demonstrated an impressive reduction in VTE in patients treated with apixaban (1/42 [2.8%]) compared to standard of care (SOC; 10/40 [25%]). Bleeding was not increased with apixaban: major bleeding occurred in 2.4% in the apixaban arm and 5% in the SOC arm, with no difference between the composite of major and clinically relevant nonmajor bleeding between the 2 arms.
Inclusion criteria included a new diagnosis of ALL/LL, use of a central venous catheter (CVC), and a chemotherapy regimen that included asparaginase. Obesity was defined as body mass index ≥95th percentile for age and sex in children aged ≥2 years.3 Patients weighing >35 kg received the adult prophylactic dose of 2.5 mg twice daily, and patients weighing <35 kg received weight-based dosing according to the protocol. The protocol included an ultrasound of the extremity with the CVC and an echocardiogram to assess for right atrial thrombus at the end of induction. Participants who had neither an ultrasound nor echocardiogram at the end of the study were considered to not have an event (apixaban arm, n = 6; SOC arm, n = 3), so it is possible there were even more thrombotic events in this subanalysis that were not captured. Nearly all of the VTEs (10/11) were clinically unsuspected (previously termed asymptomatic); there was 1 clinically symptomatic VTE, a cerebral sinovenous thrombosis in the SOC arm.
Children with ALL/LL have long been identified to have increased risk of VTE related to the presence of a CVC as well as the prothrombotic effects of asparaginase administered during induction. Despite this, efforts to prevent thrombosis in children with ALL/LL have been challenging. Although there are clinical trial data to suggest that low-molecular-weight heparin (LMWH) alone or with antithrombin supplementation may reduce VTE in children with ALL/LL, this practice has not gained widespread acceptance, partly due to the painful injections required for LMWH.4,5 One great step forward in the design of the PREVAPIX-ALL trial was the incorporation of an oral anticoagulant, which is far more acceptable to children and their parents than LMWH.
Interestingly, a significant treatment interaction between obesity and apixaban was observed in this subanalysis. Unfortunately, biomarker analysis was unable to identify differences between obese and nonobese patients to explain this treatment effect. Further investigation of this unique treatment interaction is warranted to better understand why apixaban was not as effective in nonobese patients and whether this observation is unique to apixaban. Given this observation, other pediatric thrombosis studies should consider evaluating obese patients as a separate cohort.
Although the results of this subgroup analysis are anticipated to lead to practice changes, there are several nuances to consider. First, nearly all VTEs were clinically unsuspected and identified only by screening. There remains debate regarding the significance of clinically unsuspected CVC-related VTE in children. The natural history of clinically unsuspected VTE in studies of select populations, including neonates and critically ill children, suggest that not treating does not lead to severe outcomes.6,7 The significance of unsuspected VTE in patients with ALL/LL has not been carefully investigated; as the authors point out, rates of postthrombotic syndrome in children with ALL/LL and clinically unsuspected VTE are similar to those with symptomatic VTE. Additional adverse consequences of unsuspected VTE in children ALL/LL could include thrombus progression to become symptomatic, embolization to cause pulmonary embolism, catheter dysfunction, or an increased risk of CVC bloodstream infection. Prospective observational studies of unsuspected VTE in children with ALL/LL are needed to better understand their significance. The PREVAPIX trial would have been the perfect opportunity to do this, because there were 63 patients with clinically unsuspected VTE in the trial whose outcomes have not been reported. Even though we do not use, nor recommend, screening ultrasounds as part of routine clinical care, prevention of unsuspected events using apixaban during induction seems reasonable, given the potential for these to contribute to adverse outcomes, if the bleeding rates are acceptable. The fact that there was not an increase in bleeding in this subgroup analysis, nor in the overall PREVAPIX trial in patients, is reassuring.
Second, incorporation of apixaban thromboprophylaxis outside of a carefully controlled clinical trial, will require close attention by providers to minimize bleeding. In the trial, apixaban was discontinued at least 24 hours before a lumbar puncture and resumed no sooner than 18 to 24 hours after (>48 hours for a traumatic lumbar puncture). The protocol recommended keeping the platelet count >20 x 109/L with platelet transfusions or withholding apixaban if the platelet count was <20 x 109/L to reduce bleeding. The number of platelet transfusions between patients who received apixaban and those receiving SOC was not significantly different in the full trial analysis, but detailed information regarding how often apixaban was held was not included. Providers will need to decide whether they prefer to hold apixaban or administer additional platelet transfusions in patients with thrombocytopenia.
In the trial, apixaban was discontinued upon completion of induction. In practice, the CVC will remain in place, often through the end of therapy, which may be years. Whether discontinuing anticoagulation at day 29 will have implications on the further development of VTE in the setting of ongoing ALL/LL treatment and a CVC is unknown. Future studies should consider following patients after discontinuation of anticoagulation.
Effective and safe thromboprophylaxis strategies in children at high risk of VTE are needed, and the availability of direct oral anticoagulants offers great potential compared to prior agents. This is, to our knowledge, the first RCT subanalysis to demonstrate the superiority of a direct oral anticoagulants over SOC in preventing VTE, without an increase in bleeding. Hopefully, many more will follow.
Conflict-of-interest disclosure: The authors declare no competing financial interests.
