(018) Impact of Immune Checkpoint Inhibitors on Psoriasis Incidence: Emulation of a Randomized Target Trial Using Real-world Data

Background: Immune checkpoint inhibitors (ICIs) are novel and revolutionized cancer therapy by enhancing the immune system against tumors. Recently, their clinical application raised concerns about immune-related adverse reactions, including the development of autoimmune conditions. Psoriasis is a prevalent skin disease and also recognized as an autoimmune disease. However, very few studies have investigated the relevant issue.

Objectives: To assess the risk of psoriasis in cancer patients treated with ICIs versus those receiving chemotherapies or targeted therapies (non-ICIs).

Methods: This nationwide study emulated a randomized controlled trial through a target trial design and used data from the Taiwan National Health Insurance Database. Eligibility criteria included patients aged over 18 years, receiving cancer treatment between 2019, and 2021, with prior diagnoses of clinical stage III or stage IV or metastatic cancer. We then identified ICI users as study group and those receiving other cancer therapies as non-ICIs group. The primary endpoint was the incidence of psoriasis during the follow-up period. We applied propensity score methods with stabilized inverse probability of treatment weighting (IPTW) to balance potential confounders and utilized an intention-to-treat approach to establish causality. Cox and Fine-Gray subdistribution hazard models, with death as a competing risk, were used to calculate hazard ratios (HR) for psoriasis risk between groups.

Results: Totally, 3,188 patients were eligible for the ICI groups, while 132,042 patients were eligible for the non-ICI group. Over 197,107 person-years of follow-up, the incidence rates of psoriasis were 5.76 cases per 1,000 person-years in the ICI group and 1.44 cases per 1,000 person-years in the non-ICI group. In the Cox proportional hazards regression model, adjusted for baseline characteristics and comorbidities, ICI-treated patients exhibited a two-fold increase in the incidence rate of psoriasis (IPTW-adjusted HR of 2.97 (95% CI, 1.73 to 5.09) compared to non-ICI users. The results remained robust when using the Fine-Gray subdistribution hazard model, with an IPTW-adjusted sub-distribution HR of 2.26 (95% CI: 1.32 to 3.88). These risk findings were consistently validated in sensitivity analyses and in most subgroup analyses.

Conclusions: This nationwide target trial emulation study found that cancer patients treated with ICIs had a higher risk of developing psoriasis compared to those receiving non-ICI treatments. This study highlighted the important drug safety issue in ICI users. More studies are essential to demonstrate the relationship between ICI use and immune-related adverse reactions in the future.