Abstract A044: Testing neuro-oncology tenets: are MRI findings and symptoms different with true progression compared with immunotherapy-related treatment effect in patients with primary brain tumors?

Abstract

Abstract Background: Treatment effect (TE) poses a diagnostic challenge in patients treated with immune-checkpoint inhibitors (ICIs). Although TE has been described in patients with primary brain tumors (PBT) treated with chemoradiation as pseudo-progression, it has not been well characterized with ICI therapy. TE is thought to be less symptomatic than true tumor progression (TP), but this has not been well studied. We investigated MRI characteristics and symptoms in PBT patients with histologically confirmed TP or TE. Methods: A retrospective cohort study of 50 PBT patients treated with ICI therapy (single or doublet) with/without chemotherapy that had biopsy proven TP or TE, histologically confirmed with either mitotically active tumor cells or necrosis with macrophages without tumor cells, respectively. Blinded review of pre-operative MRI using T1-post contrast, T2-FLAIR, Perfusion, and Diffuse Weighted Index (DWI) sequences were evaluated for characteristics such as extent of T2-FLAIR, midline shift, and appearance of contrast enhancement. PRO symptom severity had been collected using the M.D. Anderson Symptom Inventory-Brain Tumor (MDASI-BT), and PROMIS Anxiety and Depression 8b short forms. Results: Forty had pathologically confirmed TP and 10 had TE; median age 44, predominantly male (58%), white (80%), non-Hispanic (82%) with GBM (40%). No differences in sex, age, ethnicity, diagnosis, prior treatment, or steroid use at the time of imaging between patients with TP vs TE. TE imaging was more likely to show midline shift (40%) on MRI than TP (9.7%), p-value=0.047. TE lesions had more T2-FLAIR relative to contrast enhancement; the ratio of contrast enhancement/T2-FLAIR hyperintensity was smaller in the TE group vs TP group (0.337 vs 0.569, respectively (p=0.010). Patients with TP had a significantly higher neurologic factor symptom score (1.88 vs 0.35, p-value=0.003). Of the neurologic symptoms, there were statistically significant group differences in numbness/tingling (1.8 vs 0.0, p<0.001) and weakness (2.8 vs 0.0, p<0.001) but not pain (2.1 vs 1.4, p<0.58) and seizures (0.8 vs 0.0, p<0.44). While not statistically significant but with effect sizes (ES) greater than 0.5, patients with TE had worse cognitive symptoms (factor score 2.3 vs 1.2, ES=0.62), difficulty remembering (4.0 vs 1.5, ES=1.03), and difficulty understanding (2.0 vs 0.7, ES=0.79), and worse scores on PROMIS Depression (51.9 vs 47.1, ES=0.51). Conclusions: TE was symptomatic and not uncommon (20%) in this cohort. The symptom profile between TE and TP differed; TE presented with more cognitive and affective symptoms, whereas TP presented with focal neurologic symptoms, including weakness and numbness/tingling. The imaging findings in TE of more midline shift and T2-FLAIR suggest a possible diffuse cytokine-release process causing parenchymal changes rather than tumor invasion. These findings highlight that TE with ICI treatment is common and may have symptoms and imaging changes that are related to the underlying inflammatory process. Confirmatory studies are required. Citation Format: Emma M. Byrne, Maeve Pascoe, Elizabeth Vera, Tito Mendoza, Hope Miller, Tricia Kunst, Kelly Mentges, Kathleen Wall, Matt Lindsley, Anna Choi, Ewa Grajkowska, John Kohrs, Jim Smirniotopoulos, Eric Burton, Byram Ozer, Marta Penas-Prado, Jing Wu, Solmaz Sahebjam, Terri S. Armstrong, Mark R. Gilbert. Testing neuro-oncology tenets: are MRI findings and symptoms different with true progression compared with immunotherapy-related treatment effect in patients with primary brain tumors? [abstract]. In: Proceedings of the AACR Special Conference on Brain Cancer; 2023 Oct 19-22; Minneapolis, Minnesota. Philadelphia (PA): AACR; Cancer Res 2024;84(5 Suppl_1):Abstract nr A044.