Preemptive Bone Marrow Transplantation and Event-Free Survival in Fanconi Anemia

Background: Fanconi anemia (FA) is a primarily autosomal recessive bone marrow failure and cancer predisposition syndrome associated with mutations in the FA/BRCA DNA damage response pathway. The median age at diagnosis of FA is 7 years; the diagnosis is often made due to recognition of characteristic birth defects. Over half of patients with FA develop severe bone marrow failure (BMF) by age 50 years, one in ten develop acute myeloid leukemia (AML), and one in four develop a solid tumor (ST) as their first event. Successful allogeneic bone marrow transplantation (BMT) is potentially curative of FA's hematologic manifestations but introduces risks of transplant-related mortality (TRM) and morbidity. We hypothesized that preemptive bone marrow transplantation (PE-BMT) for individuals diagnosed prior to the development of BMF, AML, or ST, would increase event-free survival (EFS) if the risks associated with transplantation were sufficiently low.

Methods: We developed a mathematical decision model (Markov) of EFS with the assumption that successful PE-BMT would eliminate the risks of BMF and AML, but would introduce a procedural risk of TRM. We modeled the EFS of PE-BMT at variable ages at decision ranging from birth to 30 years, and without and with an increase in the rate of ST following BMT above the level in untransplanted patients with FA. We developed our model using empirical estimates of the age-specific conditional probabilities of BMF, AML, and ST (Alter et al, BJH, 2010), and a 4.4-fold estimated increased risk of ST following BMT (Rosenberg et al, Blood, 2005). We tested the sensitivity of the model over a range of values for TRM and an increased risk of ST following BMT, and evaluated the model using TreeAge Pro 2014 (TreeAge Software, Inc, Williamstown MA, http://www.treeage.com).

Results: Children diagnosed at age 7 years receiving standard care could expect to live an additional 16 years before experiencing BMF, ST, or AML, and thus survive free of an event until an average age of 23 years. If those children instead received PE-BMT with a 10% risk of TRM, they could expect to survive an additional 29 years and be cancer-free until an average age of 36 years. However, if PE-BMT were to increase the rate of ST 4.4-fold, PE-BMT would only increase the mean EFS by 2 years over standard care, until an average age of 25 years. PE-BMT would increase the mean EFS at all ages if TRM was ≤10% and the risk of ST was the same as in untransplanted patients. PE-BMT would decrease the mean EFS when performed after age 9 years if there was 10% TRM and a 4.4-fold increased rate of ST. PE-BMT at age 18 years with 10% TRM would increase the mean EFS if it did not affect the trajectory to ST, but would decrease the mean EFS if it modestly increased the rate of ST (≥2.2-fold).

Conclusions: PE-BMT in patients with FA may provide an event-free survival benefit so long as the risk of TRM appears to be low (≤10%) and the regimen has little or no impact on the development of ST. The decision was particularly sensitive to the increase in ST following BMT. Our model suggests that older ages at decision, higher risks of TRM, and greater relative risks of ST following transplant would lead to PE-BMT being a less desirable strategy. Our estimates of event-free survival may be used to inform shared decision making between providers and families, with attention paid to patient values and the morbidity associated with BMT.