BACKGROUND: Hematopoietic cell transplantation (HCT) survivors have high risks of mortality and late complications compared with the general population. Few analyses have compared the magnitude of these risks against those experienced by a matched non-HCT cancer population where the primary difference may be exposure to HCT.
METHODS: The Fred Hutchinson Cancer Research Center (FHCRC) performs most of the HCTs in Washington State (WA), including all allogeneic HCTs. We linked medical records of all 2+ year HCT survivors who were WA residents treated at FHCRC from 1992-2009 (n=1,929; 53.3% allogeneic; 83.8% hematologic malignancies) to the state’s hospital discharge and death registries. Individuals randomly selected from the state’s cancer registry (n=5,806) and the driver’s license files (n=16,340) who also survived 2+ years formed 2 comparison groups, frequency-matched by age, sex, year of HCT, and underlying diagnosis (HCT and cancer registry subjects only). Based on hospital and death registry coding (per International Classification of Diseases [ICD-9 and 10]), the cumulative incidence and the relative hazards (HR; Cox models) of experiencing specific organ-based complications (time to first event) were estimated for each group. Subanalyses involving HCT survivors alone examined the relationship between adverse outcomes occurring 2+ years after HCT and any chronic graft versus host disease (cGVHD) and/or original disease relapse prior to 2 years.
RESULTS: With a median age at HCT of 43 years (range 0-79), HCT recipients experienced 3,011 hospitalizations and 399 deaths over a median follow-up time of 6.3 years (range 2.0-20.0) following HCT. Compared with non-HCT cancer survivors, HCT survivors had markedly higher rates of hospitalizations or deaths related to infections (10-year cumulative incidence 30.5 vs. 21.5%; HR 1.6, 95% CI 1.4-1.7), in particular respiratory infections (17.7 vs. 11.4%; HR 1.8, 95% CI 1.5-2.0). Among 688 hospitalizations with a primary infection code (ICD9 codes 1-139), a bacterial source was identified in 63% (gram-negative species 16%; staphylococcus 14%; clostridium difficile 7%), virus in 31% (herpes zoster 8%), and fungus in 24% (aspergillus 8%). Among hospitalizations for a respiratory infection (n=511), bacterial pneumonia was common (21%), followed by viral (8%) and fungal (7%) pneumonias; most pneumonias had no pathogen coded (n=228 of 511; 45%). HCT survivors also had greater 10-year cumulative incidence rates for circulatory (25.5 vs. 22.8%), gastrointestinal (19.2 vs. 16.4%), genitourinary (18.1 vs. 15.5%), nervous system (13.5 vs. 11.6%), musculoskeletal (16.9 vs. 13.3%), dermatologic (7.2 vs. 5.6%), and non-infectious respiratory (15.5 vs. 11.8%) complications, with HRs 1.2-1.4 and p<0.05 for all listed categories. Rates of complications related to mental illness were similar in the 2 groups (12.0 vs. 12.7%; HR 1.0, 95% CI 0.9-1.2), while pregnancy-associated hospitalizations were less common following HCT (0.5 vs. 1.2%; HR 0.3, 95% CI 0.1-0.7). In analyses restricted to death categories with ≥1% prevalence, HCT survivors had an increased risk of infection-related death, in particular respiratory causes (HR 1.9, 95% CI 1.3-2.8). Differences were even greater when HCT survivors were compared against the general population, where the 10-year cumulative incidence of most complications was >10 percentage points lower in the general population. Among HCT recipients, history of chronic graft versus host disease (cGVHD) and relapse of one’s original disease were risk factors for most outcomes. Risks among allogeneic and autologous HCT recipients otherwise were similar. Adjustment for cGVHD status attenuated the risks between HCT and non-HCT cancer survivors, but HCT status remained a significant risk factor for many outcomes including infections and respiratory disease.
CONCLUSIONS: Compared with similar non-HCT cancer survivors, 2+ year HCT survivors were at greater risk of multiple adverse late complications, in particular late respiratory infections. Greater awareness and earlier intervention, along with improvements in cGVHD treatment, may reduce some of these disparities.
- ASH 56th (2014)