Risk factors and prediction model for cancer-related cognitive impairment in thyroid cancer patients.

Abstract

Background: Cognitive impairment is a common, yet often overlooked, complication in thyroid cancer patients, potentially influenced by various demographic, clinical, biochemical, and psychological factors. This study aims to analyze the prevalence and determinants of cancer-related cognitive impairment (CRCI) in thyroid cancer patients.

Methods: A retrospective case-control study was conducted involving 246 thyroid cancer patients treated at our The First Affiliated Hospital of Soochow University from January 2021 to January 2023. Patients were categorized into high cognitive function (n = 125) and low cognitive function groups (n = 121) based on Mini Mental State Examination (MMSE) scores. Data were collected on demographic variables, Charlson Comorbidity Index (CCI), disease duration, clinical stage, blood test results, inflammatory factors (interleukin-6 (IL-6), interleukin-8 (IL-8), tumor necrosis factor-alpha (TNF-α), C-reactive protein (CRP)), psychological status (Self-Rating Depression Scale (SDS), Self-Rating Anxiety Scale (SAS), Self-Esteem Scale (SES)), sleep quality (Pittsburgh Sleep Quality Index (PSQI)), and quality of life (36-item Short-Form Health Survey (SF-36)). Additionally, an external validation set was established, with patients being divided into a high cognitive level group (n = 135) and a low cognitive level group (n = 128), and the model's predictive performance was validated through the external dataset.

Results: Factors significantly associated with lower cognitive function included age (P < 0.001), education level (P < 0.001), CCI scores (P < 0.001), disease duration (P < 0.001), clinical stage (P = 0.003), IL-6 (P < 0.001), IL-8 (P = 0.005), TNF-α (P < 0.001) and CRP (P < 0.001). SDS (P < 0.001), SAS (P < 0.001) and PSQI (P < 0.001) were also associated with reduced cognitive function. The Least Absolute Shrinkage and Selection Operator (LASSO) regression model demonstrated strong predictive performance with an area under the curve (AUC) of 0.903 in the training set and an AUC of 0.835 in the validation set.

Conclusion: CRCI in thyroid cancer patients is multifactorial, with significant contributions from demographic, clinical, inflammatory, and psychological factors. The developed predictive model may serve as a valuable tool in clinical practice for identifying thyroid cancer patients at high risk of cognitive impairment.