Cognitive functioning in survivors of hematopoietic stem cell transplantation compared with a matched general population sample: The MOSA study

BACKGROUND: Although cognitive problems can recover over time, a subgroup of hematopoietic stem cell transplantation (HCT) survivors experiences persistent cognitive problems in the long-term. Despite these implications, studies assessing cognitive functioning in HCT survivors are limited.

OBJECTIVES: The aim of the present study was (1) to quantify the prevalence of cognitive impairment in patients treated with HCT who survived at least 2 years, and to compare these with a matched reference group representing the general population; (2) to identify potential determinants of cognitive functioning within the HCT survivor group.

STUDY DESIGN: Within the single-centre Maastricht Observational study of late effects after Stem cell trAnsplantation (MOSA) cognitive performance was assessed by a neuropsychological test battery divided into three cognitive domains respectively memory, information processing speed, and executive function & attention. An overall cognition score was calculated as the average of the domain scores. A total of 115 HCT survivors were group-matched on a 1:4 ratio to the reference group by age, sex, and level of education. Regression analyses adjusted for different sets of covariates including demographic, health- and lifestyle-related factors were used to test for differences in cognition between HCT survivors and the reference group resembling the general population. A limited set of clinical characteristics (diagnosis, type of transplant, time since treatment, conditioning regimen with TBI and age at time of transplant) were assessed as potential determinants of neurocognitive dysfunction among HCT survivors. Cognitive impairment was defined as scores in the cognitive domains < -1.5 SD from what can be expected based on someone's age, sex, and education.

RESULTS: The mean age at time of transplant was 50.2 (SD 11.2) years, and the mean number of years after transplant was 8.7 (SD 5.7) years. The majority of HCT survivors were treated with autologous HCT (N = 73; 64%). The prevalence of cognitive dysfunction in HCT survivors was 34.8%, and 21.3% in the reference group (p = .002.) When adjusted for age, sex, and level of education HCT survivors had a worse overall cognition score (b=-.35, 95%CI=-.55, -.16, p<.001), translating into 9.0 years of higher cognitive age. Analyses of specific cognitive domain scores showed that HCT survivors scored worse on memory (b = -.43, 95%CI = -.73, -.13, p = .005), information processing speed (b = -.33, 95%CI = -.55, -.11, p = .003), and executive function & attention (b = -.29, 95%CI = -.55, -.03, p = .031) than the reference group. The odds of cognitive impairment were on average 2.4 times higher among HCT survivors than the reference group (OR=2.44, 95%CI=1.47, 4.07, p=.001). Within the HCT survivor group none of the tested clinical determinants of cognitive impairment were significantly associated with cognition.

CONCLUSION: This cohort study showed evidence for worse cognitive functioning in HCT survivors encompassing all three cognitive domains, respectively memory, information processing speed, and executive & attention compared to a reference group that represents the general population translating into nine years of faster cognitive ageing in HCT survivors than can be expected based on their chronological age. It is important to increase awareness for signs of neurocognitive dysfunction after HCT in clinicians and HCT survivors.