Immune Checkpoint Inhibitors and Cancer-Related Cognitive Decline: A Propensity Score Matched Analysis in active Chemotherapy Patients

Guangmin Jian ¹ Jiling Zeng ² Jun Lu ³ Weidong Wang ⁴ Yongluo Jiang ² Tong Huang ⁵ Yu Si Niu ⁶ Zhoufang Chai ⁷ Xin Qi ⁸ Nianqi Liu ⁹ Youlong Wang ⁸ Cantong Liu ⁴ Jiacai Lin ⁸ Guanqing Zhong ² Yiming Li ¹⁰ Pengfei Zhu ¹ Zongqing Zheng ¹¹ Fadian Ding ^11* Xinjia Wang ⁴ Weizhi Liu ¹² Ao Zhang ² Yifei Ma ¹¹

• ¹ First Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan Province, China
• ² Sun Yat-sen University Cancer Center (SYSUCC), Guangzhou, Guangdong Province, China
• ³ First Affiliated Hospital of Shantou University Medical College, Shantou, Guangdong Province, China
• ⁴ Cancer Hospital, College of Medicine, Shantou University, Shantou, Guangdong Province, China
• ⁵ General Hospital of Shenyang Military Command, Shenyang, Liaoning Province, China
• ⁶ Dallas County Health and Human Services, Dallas, United States
• ⁷ Hospital of Maternal and Child Care, JiangShan, Zhejiang Province, Zhejiang, China
• ⁸ Hainan Hospital of People's Liberation Army General Hospital, Hainan, China
• ⁹ Huazhong University of Science and Technology, Wuhan, Hubei Province, China
• ¹⁰ Beijing Tiantan Hospital, Capital Medical University, Beijing, Beijing Municipality, China
• ¹¹ First Affiliated Hospital of Fujian Medical University, Fuzhou, Fujian Province, China
• ¹² Naval Medical University, Shanghai, Shanghai, China

We investigated whether 1-year trajectories of cancer-related cognitive decline (CRCD) would be different in patients with chemotherapy combined with immune checkpoint inhibitors (chemoICI group) as compared with chemotherapy alone (chemo group).Participants scheduled with or without ICI were prospectively recruited from three academic hospitals and followed up for 1 year in four sessions. Subjective and objective CRCD were measured by Perceived Cognitive Impairment (PCI) and Montreal Cognitive Assessment (MoCA), respectively.Primary endpoints were MoCA and PCI score changes and minimal clinically important difference (MCID), which was defined as threshold for meaningful impairment events. Propensity score matching (PSM) was performed for group comparison using logistic regression with covariates including age, cancer stage, and baseline cognitive scores. Linear mixed models adjusted for repeated measures.Out of 1557 recruited patients PSM yielded 460 patient pairs (1:1). Mean PCI and MoCA scores of both groups reached MCID at 12-month session in both groups. In chemoICI, MoCA score changes were significantly lower in the 12-month session, and PCI score changes were lower in the 6, 9, and 12-month sessions than chemo (P<0.05). One-year meaningful impairment events risks were 0.44 and 0.56 in chemoICI, significantly higher than that of chemo (0.35 and 0.38, P<0.01). Significant differences were found in mean event-free survival time in patients with and without irAE in chemoICI subgroup analysis.Our findings suggest that combining chemotherapy with ICIs may exacerbate CRCD compared to chemotherapy alone. However, reliance on screening tools and self-reported measures limits definitive conclusions. Future studies incorporating comprehensive neuropsychological assessments are warranted. This study underscores the importance of using comprehensive cognitive assessments in future research to better understand the impact of ICIs on cognitive function.