Pembrolizumab Stabilizes Refractory GIST in Single Case Report

A single patient with metastatic gastrointestinal stromal tumor (GIST) experienced nine months of stable disease after receiving pembrolizumab, an immune checkpoint inhibitor, following progression or intolerance to six different tyrosine kinase inhibitor (TKI) regimens. The case, reported in Cancer Reports, highlights a rare instance of disease control in a heavily pretreated setting where prospective clinical trials have generally shown limited activity for immunotherapy. The patient, a man diagnosed in 2007, had exhausted standard KIT-directed therapies and lacked established predictive biomarkers for checkpoint inhibitor benefit, such as high tumor mutational burden (TMB) or microsatellite instability (MSI).

GISTs are the most common mesenchymal neoplasms of the gastrointestinal tract, with an estimated annual incidence of 10 to 15 cases per million people. These tumors typically arise from interstitial cells of Cajal or related precursors and are driven by activating mutations in the KIT or PDGFRA genes. While the introduction of imatinib substantially improved outcomes for patients with metastatic disease, most eventually develop resistance, often through secondary KIT mutations such as activation-loop mutations in exon 17. Subsequent lines of therapy, including sunitinib, regorafenib, and ripretinib, provide additional disease control but are ultimately limited by resistance or intolerance.

The patient in this report was diagnosed with epithelioid small-bowel GIST in 2007 at age 57. He underwent surgical resection and received 12 months of adjuvant imatinib. Seven years later, in 2014, surveillance imaging revealed metastatic disease involving the liver. Genomic analysis of a resected hepatic lesion in 2019 identified a primary KIT exon 11 D579del mutation, a secondary KIT exon 17 D820G mutation, and deletions of CDKN2A and CDKN2B. The tumor was characterized as having a low tumor mutational burden of 1 mutation per megabase and being microsatellite stable, features that typically predict poor response to immunotherapy.

The patient’s treatment course involved six TKIs across seven treatment courses. After initial control with imatinib, he progressed on regorafenib due to poor tolerability, including anemia and reduced renal function. Ripretinib maintained disease control from 2020 until August 2024, when progression prompted dose escalation and Y-90 radioembolization. Following further progression, he received short courses of sunitinib, cabozantinib, and avapritinib, all discontinued due to adverse effects or radiographic progression. An imatinib rechallenge in mid-2025 also failed, with imaging demonstrating continued progression of abdominal, pelvic, and pulmonary metastases.

By late 2025, the patient had no remaining standard or accessible investigational options. He was screened for a Phase 1a/1b study of ziftomenib, a menin inhibitor, but was ineligible due to anemia and progressive renal dysfunction. With a performance status of ECOG 1, the medical team elected to try off-label pembrolizumab, selected for its comparatively favorable toxicity profile rather than established efficacy in GIST. The decision was made after a 70-day interval without systemic therapy, during which the patient’s renal function began to recover from TKI-associated nephrotoxicity.

Pembrolizumab was initiated on October 28, 2025, at a dose of 200 mg intravenously every three weeks. Response was assessed retrospectively using RECIST 1.1 criteria, with the January 2026 contrast-enhanced CT serving as the measurement baseline because pretreatment imaging lacked intravenous contrast due to renal impairment. Two peritoneal implants were designated as target lesions. The sum of target lesion diameters decreased from 9.9 cm at baseline to 8.0 cm in July 2026, representing a 19.2% reduction. No new lesions appeared, and nontarget disease remained stable to slightly decreased, resulting in a best overall response of stable disease.

Throughout the 13 cycles of pembrolizumab administration, the patient experienced no documented treatment-related or immune-related adverse events. There were no dose interruptions due to toxicity, and the patient reported no fevers, chills, hemoptysis, or bowel or bladder dysfunction. Notably, the patient’s long-standing anemia and kidney function improved during this period. Hemoglobin rose from 8.4 g/dL at imatinib discontinuation to 12.6 g/dL by July 2026, and creatinine decreased from 1.58 mg/dL to 1.16 mg/dL. The authors attribute the renal improvement to recovery from TKI nephrotoxicity and the hemoglobin rise to the resolution of gastrointestinal blood loss previously associated with tumor bleeding and antiangiogenic therapy.

This outcome contrasts with the generally limited activity observed in prospective GIST immunotherapy studies. In a randomized Phase II study, neither nivolumab monotherapy nor the combination of nivolumab and ipilimumab met prespecified response-rate benchmarks. While some patients achieved stable disease, objective responses were rare. The durable stable disease observed in this case resembles the stable-disease signal reported in those trials rather than an objective response, and the authors emphasize that this single case cannot establish causation or generalizability.

The authors propose that sequential TKI exposure may have altered the tumor biology or immune microenvironment, potentially sensitizing the tumor to pembrolizumab. Preclinical models suggest that imatinib can potentiate antitumor T-cell responses, and translational studies indicate that combined PD-1/PD-L1 blockade can enhance imatinib activity. However, without paired tissue or blood-based immune profiling from the time of treatment initiation, it is impossible to determine whether prior TKI exposure was the driving factor. The tumor’s TMB-low and microsatellite-stable status further complicates the interpretation, as these features are not established predictors of checkpoint inhibitor response.

Alternative explanations for the disease stabilization were considered and largely discounted. Prior Y-90 radioembolization could have affected a previously treated hepatic lesion, but this does not account for the concurrent stability of disease outside the radiation field, including peritoneal and pulmonary sites. An indolent natural history is difficult to reconcile with the August 2025 MRI, which documented enlarging peritoneal lesions, new metastases, and increasing ascites just 10 weeks before pembrolizumab initiation. Delayed benefit from prior TKI therapy is also unlikely, as progression occurred during active imatinib exposure.

The study has significant limitations. The imaging assessment was retrospective, and the transition from non-contrast to contrast-enhanced CT precludes technique-matched measurement across the treatment boundary. The lesion-level comparison spanning this transition is reported as exploratory. Furthermore, the lack of contemporaneous PD-L1, CD8, or blood-based immune profiling means the immune context at initiation is unknown. These constraints limit mechanistic inference, but the primary observation remains: disease progressing through TKI exposure stabilized on pembrolizumab for nine months.

This case is described as hypothesis-generating rather than practice-changing. It does not recommend pembrolizumab as a standard treatment for GIST, given the lack of prospective evidence and the absence of predictive biomarkers. However, it supports the need for prospective evaluation of checkpoint blockade in heavily pretreated GIST patients. Ideally, such studies should incorporate paired pre- and on-treatment tissue sampling to characterize the immune microenvironment in patients with prolonged TKI exposure, potentially identifying subgroups that may benefit from immunotherapy.

The genomic profile of this patient’s tumor included deletions of CDKN2A and CDKN2B, alterations that have been associated with malignant progression in KIT-mutant GIST. These genetic changes align with the aggressive disease course documented in the patient’s records, where pretreatment imaging showed a near-tripling of several peritoneal deposits and the enlargement of the dominant hepatic lesion from 7.4 to 11.0 cm over a seven-month period. Despite the presence of these high-risk features, the tumor lacked the molecular characteristics typically associated with favorable responses to immune checkpoint inhibitors. Large molecular series indicate that high tumor mutational burden and high microsatellite instability are uncommon in GIST, making this case an outlier in terms of expected immunotherapy response. The absence of these predictive markers underscores the unusual nature of the observed disease stabilization, as the tumor did not possess the established biological traits that generally correlate with checkpoint inhibitor benefit.

While the authors propose that sequential exposure to multiple tyrosine kinase inhibitors may have altered the tumor’s immune microenvironment, this hypothesis remains unproven in this specific case. Preclinical models suggest that imatinib can potentiate antitumor T-cell responses by inhibiting indoleamine 2,3-dioxygenase, and translational studies indicate that combined PD-1 and PD-L1 blockade can enhance imatinib activity. Additionally, a previous series found an association between exposure to multiple antiangiogenic TKIs and higher expression of PD-1, PD-L1, and other exhaustion markers in advanced GIST specimens. However, that same series demonstrated a predominantly immunosuppressive phenotype, and it was unable to distinguish between treatment effects and natural disease evolution. Without paired tissue or blood-based immune profiling obtained at the time of pembrolizumab initiation, it is not possible to determine whether prior TKI exposure sensitized this specific tumor to immunotherapy.

The clinical decision to initiate pembrolizumab was driven by the exhaustion of standard therapeutic options and the patient’s specific clinical status. After progressing through or failing to tolerate six different TKI regimens, the patient was screened for the Phase 1a/1b KOMET-015 study, which investigated ziftomenib, a menin inhibitor, in combination with imatinib. Although preclinical evidence suggested that menin inhibition could reduce GIST growth and enhance imatinib activity in patient-derived xenograft models, the patient was deemed ineligible for the trial due to anemia and progressive renal dysfunction. His performance status, classified as ECOG 1, did not preclude additional therapy, but the lack of remaining standard or accessible investigational options left the medical team with limited alternatives. The choice to proceed with off-label pembrolizumab was made after a discussion regarding its experimental status in GIST and its comparatively favorable toxicity profile, prioritizing safety in a patient with significant comorbidities.

The retrospective nature of the imaging assessment introduces specific constraints that limit the precision of the reported response. Because pretreatment imaging was performed without intravenous contrast due to renal impairment, the contrast-enhanced CT scan from January 2026 served as the measurement baseline for RECIST 1.1 criteria. This transition from non-contrast to contrast-enhanced techniques precludes technique-matched measurement across the treatment boundary, meaning that the lesion-level comparison spanning this transition is reported as exploratory. While the sum of target lesion diameters decreased by 19.2% from baseline to July 2026, the comparison with the October 2025 PET/CT, which showed a dominant implant measuring 11.0 cm, suggests a potentially larger reduction. However, this broader comparison spans a change in imaging technique and is therefore not considered a definitive measure of response. The primary observation remains that disease progressing through TKI exposure stabilized on pembrolizumab for nine months, a finding that is clinically meaningful in such a heavily pretreated setting but requires prospective validation to establish generalizability.

Subject of Research: Pembrolizumab activity in tyrosine kinase inhibitor-refractory metastatic gastrointestinal stromal tumor

Article Title: Pembrolizumab Activity in Tyrosine Kinase Inhibitor–Refractory Metastatic GIST: A Case Report

Article References: Bandaru, N., Shiach, J., & Agulnik, M. (2026). Pembrolizumab Activity in Tyrosine Kinase Inhibitor–Refractory Metastatic GIST : A Case Report. Cancer Reports, 9(11), Article e70726. https://doi.org/10.1002/cnr2.70726

Image Credits: AI Generated

DOI: 10.1002/cnr2.70726

Keywords: GIST, Pembrolizumab, Tyrosine Kinase Inhibitors, Immunotherapy, Case Report, Metastatic Disease, KIT Mutation, Cancer Treatment, Activity, Tyrosine, Kinase, Inhibitor