Clinical and Biological Factors Associated with Pseudoprogression Under Immune Checkpoint Inhibitors: A Retrospective Pan-Tumor Cohort Study

Background: Immune checkpoint inhibitors (ICIs) have transformed cancer treatment but may induce pseudoprogression (PP), an atypical response pattern that may mimic progressive disease (PD). Although iRECIST criteria were developed to characterize such atypical responses, distinguishing PP from PD when progression is first observed remains challenging. We aimed to characterize patients with clinically suspected PP and explore clinical and biological factors associated with subsequent classification as PP versus PD. Methods: We conducted a retrospective single-center study including patients with metastatic cancer treated with ICIs at the Institut de Cancérologie de l’Ouest (ICO) in whom PP was clinically suspected. Clinical, biological, treatment, and outcome data were collected at ICI initiation, PP suspicion, and subsequent PP/PD classification. Associations with PP versus PD were explored using univariable and multivariable logistic regression. Results: Among 123 patients with clinically suspected PP, 56 were subsequently classified as PP and 67 as PD. The most frequent primary tumors were lung (38%), kidney (24%), and bladder (11%) cancers. Median time from ICI initiation to PP suspicion was 79 days. ECOG performance status of 0 at ICI initiation was more frequent in the PP than in the PD group (55% vs. 32%; p = 0.046). At PP suspicion, median LDH levels were lower in the PP than in the PD group (189 vs. 232.5 U/L; p = 0.003). After multivariable adjustment, higher LDH levels at PP suspicion remained associated with PD (adjusted OR 2.78 per 100 U/L increase; 95% CI, 1.19–6.52; p = 0.019). Conclusions: In patients with clinically suspected PP, LDH may provide complementary information but cannot reliably distinguish PP from PD when considered alone. Confirmatory imaging therefore remains essential for treatment decision-making.