Venous thromboembolism (VTE) in children who are hospitalized represents an important cause of morbidity and resource use. In the general population, VTE in children is rare, but hospitalization increases this incidence 100- to 1000-fold to 30 to 58 events per 10 000 admissions.1 Pediatric VTE is associated with numerous poor outcomes, including congestive heart failure, limb ischemia, all-cause mortality, and increased length of hospital stay.1–3 For reasons not totally understood, rates of pediatric VTE have been increasing in recent years.1,4
Some risk factors for pediatric VTE are frequently reported in the literature. These include a central venous catheter, infection, age <1 year, age >15 years, illness severity, and sepsis.5–7 In this issue of Pediatrics, Goel et al8 present evidence for another risk factor in the pediatric surgical population: red blood cell (RBC) transfusion. This report parallels other pediatric studies of thrombosis risk associated with transfusion3,7 and a previous adult study by the same group.9
The role of RBCs in coagulation and thrombosis is an intriguing concept often overlooked in our medical training but one for which evidence is accumulating.10 RBCs increase blood viscosity and marginate platelets to where they can be more effective. RBCs produce thromboxane and adenosine diphosphate, which increase platelet reactivity. Stored RBCs have decreased deformability and are more likely to undergo hemolysis, particularly within clots, with possible release of procoagulant microparticles.10 The current article includes discussion of these interesting potential mechanisms.8
These mechanistic underpinnings provide an important backdrop for the authors’ main findings: a significant epidemiological risk of RBC transfusion for pediatric VTE. Of note, the authors show that neonates are at highest risk, that a central venous catheter is the greatest additional risk factor, and that a RBC dose response is detected.8 These results extend known associations and strengthen the conclusions related to RBC risk. Yet epidemiological studies carry inherent limitations, and this study is no exception. Notably, the possibility of bias from unspecified confounding variables, the Achilles’ heel of epidemiology, accounts for why causality cannot be established. Appropriately, the authors are careful to point out that they have only established association and that these data should serve as hypothesis generating for future investigations. Furthermore, they emphasize the need for validation of their findings in well-designed studies involving other populations.8
As such, the current study by Goel et al8 serves as a relevant starting point to further refine the role of this unique potential risk factor in pediatric VTE. Specifically, which transfused patients are at greatest risk for VTE? Certainly the role of genetic thrombophilias in pediatric VTE needs clarification in many settings; postoperative and catheter-related VTEs are among these.11 Future trials should therefore quantify the potential thrombotic risk of RBCs in the setting of known genetic variants (factor V Leiden, prothrombin 20210 mutation, methylene tetrahydrofolate reductase, etc). Also, non-RBC components (plasma, platelets, cryoprecipitate), which carry procoagulant activity, frequently accompany RBC transfusion, especially in the setting of massive transfusion.12–15 As the authors suggest, the thrombotic risk of RBCs would be strengthened and clarified by quantifying the contribution of these additional blood products.8
So, if the findings of the current study are indeed validated, then on the spectrum of RBC transfusion hazards, where does thrombosis lie? Goel et al8 report the overall risk in their transfused cohort in the range of 0.6% for children to 1.8% for neonates; adjusted odds ratios for transfusion are reported from 2.2 to 4.1, respectively. The latest guidelines from the American Association of Blood Banks place these risks near those of transfusion-associated circulatory overload (1:100) and allergic reaction (1:250) but less than that of transfusion-related acute lung injury (1:12 000).16 This would render RBC transfusion–associated pediatric VTE among the more common complications of transfusion. As the authors point out, because of the escalation of care and morbidity associated with pediatric VTE, the subject is highly relevant.8
Finally, it is important to remember that these data largely represent general and orthopedic surgical patients, so extension to other populations may not necessarily apply. The relative significance of risk factors varies considerably between different pediatric populations. Cardiac surgery patients, for example, represent a unique high-risk subgroup, possibly owing to a greater use of procoagulant non-RBC blood products, cardiopulmonary bypass, antifibrinolytics, and vascular access devices.3–5,17 The specific role of RBCs in this population is likely different from that in the general or orthopedic surgery populations, in which many of these additional risk factors are usually absent. Because of the high degree of heterogeneity in pediatric VTE, each clinical population and thrombus location may be considered a different disease.18 By further characterizing the various risk factors in different populations, and by designing prospective studies with focused attention to specific risk groups, an improved understanding of the contribution of RBCs to pediatric VTE may be possible.
Opinions expressed in these commentaries are those of the author and not necessarily those of the American Academy of Pediatrics or its Committees.
FUNDING: No external funding.
COMPANION PAPER: A companion to this article can be found online at www.pediatrics.org/cgi/doi/10.1542/peds.2019-2351.
POTENTIAL CONFLICT OF INTEREST: The author has indicated he has no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: The author has indicated he has no financial relationships relevant to this article to disclose.