Self-Endorsed Cognitive Problems Vs. Objectively-Assessed Cognitive Impairment in Blood or Marrow Transplantation (BMT) Recipients — a Longitudinal Study

Abstract

Background: Cognitive impairment in BMT survivors is well-documented, and may include deficits in memory, attention, executive functioning and processing speed. However, the clinical relevance of self-endorsed cognitive problems and their relation to objectively-assessed cognitive impairment is not established.Methods: We assessed cognitive impairment in 379 BMT recipients (median age = 52.2y; 40% females; 68% non-Hispanic whites) at 5 pre-determined timepoints: baseline (pre-BMT), 6mo, 1y, 2y and 3y after BMT. Self-endorsed cognitive problems (0: not at all to 4: extremely) were evaluated using the Neuropsychological Impairment Scale (NIS) - a 95-item survey. Responses were used to derive a Global Measure of Impairment (GMI) T-score; higher scores indicated greater cognitive problems. A standardized 2h battery of objective cognitive testing was also administered at each timepoint to derive a Global Deficit Score (GDS); individuals with GDS >0.5 were considered impaired. Healthy controls (HC: n = 98; median age = 55.1y; 46% females; 62% non-Hispanic whites) also completed the NIS and underwent cognitive testing at the 5 corresponding timepoints. At each timepoint, a multivariable analysis of covariance was performed to examine the magnitude of difference in GMI between BMT recipients and HC. Multilevel modeling was employed to examine participants' rates of change in GMI across time. Among BMT recipients, we used multivariable analyses to identify clinical/demographic variables associated with greater self-endorsed cognitive problems; variables examined included age at BMT, sex, race/ethnicity, education, pre-BMT IQ, fatigue, type of BMT (autologous, allogeneic [myeloablative, reduced-intensity]) and use of total body irradiation. To evaluate the relation between NIS subscales and corresponding objective cognitive domains, within-group first-order partial correlation analyses were conducted at each timepoint. We also examined the association between cognitive impairment and return to work at 3y.Results: Significant differences (p<.001) were observed in mean GMI scores with BMT recipients having consistently higher scores than HC at all timepoints. Multilevel modeling revealed that the rate of change in NIS scores was significantly greater in BMT recipients when compared with HC, after adjusting for age, sex, education level, fatigue and baseline IQ (Fig 1). Among BMT recipients, predictors of endorsement of cognitive problems across all timepoints (p ≤.003) included younger age at study (Wilks' λ = .797), higher pre-BMT IQ (Wilks' λ = .858) and greater fatigue (Wilks' λ = .811). Overall, there was a statistically-significant, modest correlation between GMI and GDS scores (range: .401 to .445, p ≤ .01), with mean GMI scores being generally higher (indicating greater endorsement of cognitive impairment) among those who were found to be cognitively impaired (GDS) via objective testing (Fig 2); these relationships were statistically significant at 6mo, 1y and 2y. Multivariable analysis adjusting for age, sex, education level, fatigue and IQ revealed that higher self-reported cognitive impairment (GMI) and higher objectively assessed cognitive impairment (GDS) were associated with a 3.7-fold (p = .02) and 4.3-fold (p = .02) higher odds of not returning to work at 3y, respectively.Conclusions: BMT recipients are more likely to report cognitive problems as compared with HC. Further, these problems increase with time from BMT. Younger age at BMT, higher pre-BMT IQ and greater fatigue are associated with greater self-endorsed cognitive problems. Further, there is a modest-to-high correlation between self-endorsed cognitive problems and objectively determined cognitive impairment. Finally, both self-endorsed cognitive problems and objectively assessed cognitive impairment are associated with a higher likelihood of not returning to work. Overall, this study demonstrates that self-endorsed cognitive problems can serve as a good screen for identifying those with cognitive impairment. DisclosuresForman:Mustang Therapeutics: Other: Licensing Agreement, Patents & Royalties, Research Funding.