Cognitive Screening in Adults with Sickle Cell Disease: Prospective Validation of the Rowland Universal Dementia Assessment Scale

Background: Adults with sickle cell disease (SCD) are at increased risk of early cognitive decline. The American Society of Hematology recommends surveillance for cognitive impairment. However, optimal screening strategies in this population have not been established. The RUDAS (Rowland Universal Dementia Assessment Scale) is a short screening questionnaire developed for use in culturally diverse patient populations. In adults with SCD, it is less biased by educational attainment than the Montreal Cognitive Assessment (MoCA).

Hypothesis: We hypothesized that the RUDAS is a valid tool to predict neurocognitive disorder (NCD) in adults with SCD.

Aims: 1. Describe trends in RUDAS scores during long-term cognitive surveillance;

2. Evaluate the ability of RUDAS scores to predict the diagnosis of NCD.

Method: A retrospective study was conducted. All patients who had undergone cognitive screening with RUDAS in 2018 and repeat testing as part of routine surveillance in 2021-2022 were identified. Starting in 2021, all adults with SCD are assessed by a multidisciplinary team (neurologist, hematologist, neuropsychologist) for neurocognitive screening and assessment. RUDAS was administered and scored as previously described (Forté et al, JAMA Net Open 2021). A clinical diagnosis of NCD was established by a neurologist and/or neuropsychologist. Repeated measure t-test was performed to compare RUDAS trends over time. Standard t-test was used to compare RUDAS between a priori defined subgroups (with and without NCD). Positive predictive values (PPV) were calculated.

Results: From an initial assessment of 92 patients, 28 adult SCD patients met the inclusion criteria. The median age was 39 years [range: 24-65], 16 (57%) were female and the median educational level was 14 years [range: 10-18]. Genotype distribution was 18 (64%) with the SS or Sβ⁰-thalassemia genotype and 10 (36%) with the SC or Sβ⁺-thalassemia genotype. None of the patients had an established diagnosis of NCD, and 5 (18%) had a prior history of stroke.

The mean (±standard deviation (SD)) RUDAS scores were 26.7±2.0 in 2018 and 27.3±2.7 in 2021-2022. Overall, the scores did not significantly change over time (+0.6±2.8, paired t-test P=0.3). In table 1, we present the results of the subgroup analysis according to NCD diagnosis. In those in whom a diagnosis of NCD was established in 2021-2022, RUDAS score declined by -0.9±2.8 points (P=0.3). In those not meeting the criteria for NCD, the mean change was +1.7±2.4 (P=0.01).

RUDAS scores differed significantly in those with and without NCD (25.5±2.8 and 28.6±1.5, P<0.001). RUDAS scores <28 and even more significantly those <23 were highly predictive of a diagnosis of NCD (PPV=86% and 100%, respectively, P<0.001, table 2).

Conclusion: The results of our study suggest that lower RUDAS scores are associated with the diagnosis of neurocognitive disorder in adults with SCD. Prospective validation is ongoing with a larger sample size. In addition, patient characteristics will be explored as potential predictors of NCD.

Kuo:Agios: Consultancy, Membership on an entity's Board of Directors or advisory committees; Alexion: Consultancy, Honoraria; Apellis: Consultancy; bluebird bio: Consultancy; Celgene/BMS: Consultancy; Pfizer: Consultancy, Research Funding; Novartis: Consultancy, Honoraria; Bioverativ/Sanofi/Sangamo: Membership on an entity's Board of Directors or advisory committees.