Cancer Epidemiology in the National Cancer Institute Inherited Bone Marrow Failure Syndromes Cohort: First Report

Abstract

It is now recognized that Fanconi anemia (FA) is both an inherited bone marrow failure syndrome (IBMFS) and also a highly penetrant cancer susceptibility syndrome associated with leukemias and specific solid tumors. However, the spectrum of cancer susceptibility in other IBMFS remains unclear. The National Cancer Institute (NCI) IBMFS Cohort is the first prospective/retrospective study to follow patients with diverse IBMFS using a comprehensive and unified protocol. During 2002–2007, 66 patients with FA, 55 with Dyskeratosis Congenita (DC), 63 with Diamond-Blackfan anemia (DBA), and 16 with Shwachman-Diamond syndrome (SDS) enrolled in the study and contributed a combined total of 4113 person-years of follow-up. Adverse outcomes ascertained for all patients included bone marrow failure (BMF) leading to death or bone marrow transplant, acute leukemia (AL), myelodysplastic syndrome (MDS), and development of a solid tumor (ST). For each syndrome, we calculated the ratio of observed (O) numbers of cancers versus expected (E) numbers in a demographically matched cohort from the general United States population (O/E ratio). We also calculated cause-specific hazards and cumulative incidence by age of BMF, AL, and ST. For FA, the first adverse event was BMF in 25 patients, AL in 4, and ST in 11; 8 developed MDS. The FA experience was broadly consistent with published reports from our prior North American Survey (NAS) and an independent FA cohort in Germany. In the NCI FA cohort, the O/E ratio was 37 for all ST (versus 48 and 26 respectively in previously analyzed cohorts) and 311 for AL (versus 785 and 868); these increased risks compared with the general population were statistically significant. The O/E ratio for MDS was 4910 (also significant). In time-dependent analysis, the MDS hazard was stable at 0.6%/year. For SDS, no patient has yet developed an adverse event. For DBA, the first adverse event was BMF in 4 and ST in 3 (1 colon and 2 lung cancers); no DBA patient has yet developed AL or MDS. The cumulative incidence of BMF in DBA was 10% by age 30 years. For DC, the first adverse event was BMF in 15, AML in 2, and ST in 5; 5 developed MDS. For DC, the O/E ratio was 10 for all cancers, 7 for all solid tumors, 897 for tongue cancer, and 188 for AL; these increased risks were statistically significant. The O/E ratio was also significantly elevated for MDS (2362). To gain statistical power, we compared DC patients to a pooled cohort of 458 FA from 4 cohorts (NAS, Germany, Israel, and NCI). For DC, the cumulative incidence by age 50 years was 50% for BMF, 10% for AL, and 22% for ST, broadly similar to corresponding values of 56%, 13%, and 30%, respectively, in the pooled FA cohorts. In DC, the cause-specific hazards of ST and AL increased significantly with age, but had a later rise than in FA. The hazard of BMF in DC also increased significantly with age, but lacked the early hazard peak seen in FA. This initial analysis from the NCI IBMFS cohort reveals that as for FA, DC is a highly penetrant bone marrow failure syndrome and a major cancer susceptibility syndrome with numerous events occurring in young adulthood and early middle age. This is the first study to quantify the risk in DC in this way. Continued follow-up of the cohort should clarify the spectrum of cancer susceptibility in each syndrome.