Causes of Death in 184 Patients with Type 1 Gaucher Disease From the United States Who Were Never Treated with Enzyme Replacement Therapy

After IRB approval, COD recorded in death certificates were accessed via the National Death Index for available years 1979–2008. COD prior to 1979 were determined based on autopsy or physician communication with REL and, where possible, by death certificates obtained from State Bureaus of Vital Records. Information on COD (1950–2008) for the general US population is from US National Vital Statistics Reports. The analysis included descriptive statistics, calculation of proportional mortality ratios (PMR), and 2-tailed Fisher exact test calculations based on absolute death numbers in the GD1 and general populations.

COD is unknown for 9 patients who died before 1979. 111 pts were male (60.3%); 124 (67.4%) were Ashkenazi Jewish. Median age at death was 66y (2–97y). Median age at GD1 diagnosis (N=102): 39y (1–83y). Spleen status: Splenectomy 94 (51.1%); Intact 56 (30.4%); Unknown 34 (18.5%). Median age at splenectomy: 36y (1.3–78y). Symptomatic bone disease was present in 74 (40.2%), absent in 7 (3.8%) and undocumented for 103 (60.0%). COD for which the PMR was significantly increased (P<0.01): malignant neoplasms (PMR 1.57), suicide/drug overdose (PMR 3.86), chronic liver disease (PMR 4.76) and septicemia (PMR 9.22). Other COD that were disproportionately high included CNS and gastrointestinal bleeding, post-splenectomy complications, pulmonary hypertension (PHT), and Parkinsonism. Heart disease/atherosclerosis was the only COD for which PMR was very significantly decreased (0.33). PMR for cerebrovascular disease was 0.48 (P=0.027). For 57 pts with a CA COD, PMRs for myeloma (9.66), kidney CA (4.63), liver CA (4.36), NHL (4.13), and all leukemia (3.19) were significantly elevated (P<0.01). Conversely, the PMR for lung CA was 0.32 (P=0.002). There was 1 death from breast CA and none from gynecological CA. PMR for colorectal, pancreatic and prostate CA were not divergent from expected. Compared to the control population, the age distribution of deaths was identical for heart disease, septicemia, and suicide. Age at death was not younger than expected in pts with myeloma but there was a tendency for death at a younger age for GD pts with NHL, chronic liver disease and Parkinsonism. COD significantly more prevalent in surgically asplenic pts included chronic liver disease, septicemia, GI bleeding, PHT and post-splenectomy complications. Spleen status was statistically irrelevant to CA deaths including myeloma except for hepatocellular CA (splenectomy) and CLL (intact spleen).

With earlier diagnosis, improved risk assessment and phase-out of splenectomy, COD that we encountered (chronic liver disease, GI bleeding, septicemia, PHT, suicide and drug dependency) should be increasingly rare with timely institution of appropriate treatment. Our study population of untreated pts (estimated at 5% of all US GD1 deaths from 1950–2008 but a substantially greater percentage of GD1 deaths over age 60y), should serve as a valuable control for future studies of the effect of GD1 treatment on mortality due to malignancy or other later course events.

Weinreb:Genzyme Corporation: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Shire HGT: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Actelion Corporation: Speakers Bureau.