Why This Study Matters
Third- or later-line treatment options for metastatic colorectal cancer (CRC) are extremely limited, with historical response rates of only 1.6%-4.7% and median progression-free survival of 1.9-3.7 months on standard therapies — a cross-study benchmark, not a within-trial comparison. CAR T-cell therapy has transformed treatment of blood cancers, but efficacy in solid tumors has historically been constrained by tumor heterogeneity, stromal barriers, and an immunosuppressive microenvironment. Guanylyl cyclase 2C (GUCY2C) is stably expressed across all stages of CRC and largely restricted to the apical surface of normal intestinal epithelium, making it a rational — if imperfect — CAR T target. This trial reports the first efficacy and safety data for IM96, an autologous GUCY2C-targeted CAR T-cell product, in a heavily pretreated, microsatellite-stable population with very few remaining options.
Study Design
This was an open-label, single-center, investigator-initiated study (ClinicalTrials.gov: NCT05287165) with a 3+3 dose-escalation phase followed by a dose-expansion phase. The primary endpoint was safety and tolerability within 28 days after the first IM96 infusion, including dose-limiting toxicity and maximum tolerated dose. Four dose levels were evaluated: DL1 (3×10⁸ CAR T cells, n=3), DL2 (6×10⁸ cells, n=3), DL3 (12×10⁸ cells, n=11 total — the recommended dose for expansion), and DL4 (20×10⁸ cells, n=3).
Patient Population
Adults aged 18-75 with mismatch repair-proficient (microsatellite-stable) metastatic colorectal cancer and histologically confirmed GUCY2C-positive tumors were enrolled after failing at least second-line treatment. All enrolled patients had microsatellite-stable disease — the genotype that derives the least benefit from immune checkpoint inhibitors and for which third-line options remain sparse.
Primary Endpoint & Efficacy Results
Among all 19 efficacy-evaluable patients, the objective response rate was 26.3% (95% CI, 9.1-51.2) and the disease control rate was 73.7% (95% CI, 48.8-90.9). Results were strongest at the recommended phase II dose: in the DL3 group (n=10), the objective response rate was 40.0% (95% CI, 12.2-73.8), with median progression-free survival of 7.0 months, median duration of response of 5.1 months, and median overall survival not yet evaluable. Among the subset with medium-to-high tumoral GUCY2C expression at DL3 (n=6), the objective response rate rose to 50.0% with a median progression-free survival of 9.0 months.
| Variable | All Evaluable (n=19) | DL3 Group (n=10) | Medium/High GUCY2C Expression (n=14) |
|---|---|---|---|
| Objective response rate, % [95% CI] | 26.3 [9.1–51.2] | 40.0 [12.2–73.8] | 28.6 [8.4–58.1] |
| Disease control rate, % [95% CI] | 73.7 [48.8–90.9] | 60.0 [26.2–87.8] | 78.6 [49.2–95.3] |
| Median PFS, months [95% CI] | 3.0 [1.0–5.0] | 7.0 [0.0–20.7] | 2.0 [0.0–4.1] |
| Median DOR, months [95% CI] | 5.8 [4.3–7.3] | 5.1 [4.3–5.9] | 5.8 [1.2–10.3] |
| Median OS, months [95% CI] | 15.6 [7.9–23.3] | Not evaluable | 24.1 [9.7–38.5] |
Subgroup Analyses
Response appeared strongest in patients who both received the DL3 dose and had medium-to-high tumoral GUCY2C expression, and patients with liver metastases also responded well (40.0% objective response rate). The publication states that the objective response rate "remained consistent across factors, including age, sex, lines of prior therapies, GUCY2C expression, tumor sites, burden, and mutation," but numeric subgroup statistics for RAS/BRAF status, sex, and age were not reported and are not asserted here.
| Subgroup | n | ORR [95% CI] | Median PFS, months [95% CI] |
|---|---|---|---|
| DL3 dose group (12×10⁸ cells) | 10 | 40.0% [12.2–73.8] | 7.0 [0.0–20.7] |
| Patients with liver metastases | 5 | 40.0% [5.3–85.3] | 7.0 [0.0–15.7] |
| All patients, medium/high GUCY2C expression | 14 | 28.6% [8.4–58.1] | 2.0 [0.0–4.1] |
| DL3 patients, medium/high GUCY2C expression | 6 | 50.0% [11.8–88.2] | 9.0 [0.0–24.7] |
Safety Profile
All 20 treated patients experienced at least one treatment-emergent adverse event, and no treatment-related deaths occurred. Cytokine release syndrome occurred in 85.0% of patients (grade 1-2 in 80.0%, grade 3 in 5.0%), with a median time to onset of 2 days. Diarrhea — an expected on-target/off-tumor toxicity from GUCY2C expression in normal intestinal mucosa — occurred in 70.0% of patients, with grade 3 in 55.0%; all cases recovered (median 10 days) without glucocorticoids. Rash occurred in 70.0% (grade 3 in 10.0%) and stomatitis in 35.0%. Grade 3-4 hematologic toxicities were near-universal, including lymphopenia and leukopenia in essentially all patients. One patient each (5.0%) experienced grade 3 immune effector cell-associated neurotoxicity syndrome and immune effector cell-associated hemophagocytic lymphohistiocytosis syndrome.
| Adverse Event | Any Grade, n (%) | Grade 3, n (%) | Grade 4, n (%) |
|---|---|---|---|
| Cytokine release syndrome | 17 (85.0) | 1 (5.0) | 0 (0.0) |
| Diarrhea | 14 (70.0) | 11 (55.0) | 0 (0.0) |
| Rash | 14 (70.0) | 2 (10.0) | 0 (0.0) |
| Stomatitis | 7 (35.0) | 2 (10.0) | 0 (0.0) |
| Leukopenia | 20 (100.0) | 7 (35.0) | 5 (25.0) |
| Lymphopenia | 20 (100.0) | 1 (5.0) | 19 (95.0) |
| Neutropenia | 18 (89.0) | 3 (15.0) | 6 (30.0) |
| Anemia | 12 (60.0) | 3 (15.0) | 0 (0.0) |
| ICANS (neurotoxicity) | 1 (5.0) | 1 (5.0) | 0 (0.0) |
Interpretation and Broader Context
IM96 demonstrated an acceptable safety profile and encouraging antitumor activity at the recommended phase II dose in a microsatellite-stable, third- or later-line CRC population that currently has very limited treatment options. Efficacy did not appear to vary by RAS status, tumor mutation status, or metastatic site, hinting at potential applicability across a broad CRC population — though this reading comes from a small, single-arm, single-center phase I cohort and should be treated as hypothesis-generating.
Historical third-line CRC therapies such as regorafenib, fruquintinib, and TAS-102 have reported objective response rates of roughly 1.6% to 4.7% and median progression-free survival of 1.9 to 3.7 months; TAS-102 plus bevacizumab reported an objective response rate of about 6.1%. The study authors also compare IM96 to a previously reported GUCY2C-targeted CAR T construct, noting a similar 40% objective response rate but a longer median progression-free survival with IM96 (7.0 versus 6.0 months). None of these comparisons come from head-to-head trials, and the authors caution that cross-trial comparisons should be interpreted cautiously given this trial's small, nonrandomized, single-arm design.
The authors state that further studies will be conducted at the DL3 (12×10⁸ cell) dose level. They also note that antigen heterogeneity and loss of GUCY2C expression may drive resistance, and that dual- or multi-targeted CAR T constructs are a potential future strategy to address this.
In a phase I trial of 20 treated patients with refractory, microsatellite-stable metastatic colorectal cancer, the GUCY2C-targeted CAR T-cell product IM96 showed no dose-limiting toxicities and an encouraging efficacy signal, with a 40% objective response rate and 7.0-month median progression-free survival at the recommended phase II dose — well above historical third-line benchmarks, though not from a randomized comparison. Cytokine release syndrome and an expected on-target/off-tumor diarrhea were common but manageable, and no treatment-related deaths occurred. These single-arm, single-center findings will need confirmation in larger studies, which the authors say are planned at the recommended dose.