Granzyme B PET Imaging Uncovers Dynamic Patterns of Disease Activity and Therapeutic Response in a Murine Colitis Model

The evaluation of therapeutic response is essential in disease monitoring both for disease status and treatment efficacy in inflammatory bowel disease. Here, we focused on the use of positron emission tomography directed towards granzyme B, a serine protease released by activated cytotoxic T cells and natural killer cells, to evaluate the dynamics of therapeutic response in a colitis model. The goal was to explore the use of granzyme B positron emission tomography as a non-invasive biomarker to monitor disease activity and therapeutic response across several treatments in a dextran sulfate sodium-induced colitis model. C57BL/6 interleukin-10 knockout mice were divided into five groups, including a negative control, positive control and three treatment arms (antitumor necrosis factor, prednisolone, and anti-interleukin-23). The negative control group received regular water, while all other groups were induced with colitis via 3% DSS water for 1 week followed by normal water. Treatments were initiated after colitis was induced (anti-TNF antibody, prednisolone, or anti-IL-23 antibody). Positron emission tomography/computed tomography imaging with 68Ga-NOTA-GZP was performed at baseline (after colitis induction, before therapy), and at 1 and 2 weeks after treatment initiation. Histological analyses were also performed at 1 and 2 weeks after treatment initiation. <i>Gzmb</i> expression and histological changes were also assessed with immunofluorescence staining and bulk ribonucleic acid sequencing. <i>Gzmb</i>-targeted PET imaging revealed distinct longitudinal patterns of colonic tracer uptake related to treatment response. In positive control mice with DSS colitis (no treatment), bowel uptake of ⁶⁸Ga-NOTA-GZP increased significantly from baseline to week 2. Anti-TNF treatment reduced granzyme B positron emission tomography uptake significantly at week 2, approaching levels seen in negative controls. In prednisolone-treated mice, ⁶⁸Ga-NOTA-GZP uptake decreased at week 1 but rose significantly by week 2 but still was in normal range. Anti-IL-23 therapy produced a significantly elevated <i>Gzmb</i> PET signal at week 1, followed by a significant decline by week 2 of treatment. The imaging trends were corroborated by tissue analyses and IF staining for <i>Gzmb</i>, which revealed no colonic expression in negative controls and strong <i>Gzmb</i> elevation in positive controls and the prednisolone group but a decreased <i>Gzmb</i> signal in the anti-TNF and late anti-IL-23 groups. Bulk RNA sequencing also supported these findings, with <i>Gzmb</i> gene expression tracking with inflammation severity and NK/T cell abundance and decreasing after effective therapy. <i>Gzmb</i>-targeted PET/CT allows for dynamic and non-invasive assessment of intestinal immune compartment activity and an assessment of therapy in colitis. <i>Gzmb</i> PET was able to detect initial treatment responses of anti-TNF, steroid and anti-IL-23 based on changes in the <i>Gzmb</i> PET signal. This suggests that clinical <i>Gzmb</i> PET imaging may serve as precision imaging for monitoring disease activity with treatment in IBD and help improve patient care by identifying responders and non-responders in real time.