Primary Management of Localized Gastrointestinal Stromal Tumors With Imatinib Monotherapy: An Alternative to Resection of Patients Precluded From Surgery

Abstract

Background:

In this study, our goal was to elucidate the role for imatinib monotherapy for treatment of patients with localized gastrointestinal stromal tumors (GISTs) who are precluded from standard-of-care surgical resection due to their medical comorbidities or patient preference.

Methods:

A single-center retrospective study was conducted on a consecutive cohort of adult patients with pathology-confirmed gastrointestinal stromal tumors. The cohort of interest (n = 11) was the subset of patients on imatinib therapy alone with no prior history of curative-intent surgical resection. We analyzed patient demographics, GIST disease characteristics, treatment type, and medical comorbidities at the time of diagnosis. Overall survival (OS) and progression-free survival (PFS) were assessed using Kaplan-Meier univariate analysis.

Results:

Eleven patients met our inclusion criteria. Median age was 77 years (range 65-82) and 7 patients (64%) were men. Eight cases were gastric primary, 2 cases were duodenal, and 1 was esophageal. Four cases had genomics available, 3 of which harbored KIT mutations. For all cases, the documented mitotic rate was less than 5 per 5 mm². Median tumor size was 38 mm (range 20-58 mm). The most common medical comorbidity precluding patients from surgery was cardiac disease. All patients received imatinib as their only treatment modality and median time on treatment was 16 months. Two patients had progression of disease through treatment. Treatment was generally well tolerated with no documented grade III or grade IV adverse events. With a median follow-up of 35 months, the 1-year PFS and OS were 90% and 100%. The 3-year PFS and OS were also 90% and 100%.

Conclusion:

Patients tolerated imatinib monotherapy well and demonstrated robust survival data. Our research highlights a new potential application for imatinib in a patient population that has historically been precluded from standard-of-care therapy for their disease.

Introduction

Gastrointestinal stromal tumors (GISTs) represent the most common mesenchymal tumors of the gastrointestinal tract, arising primarily from the interstitial cells of Cajal or their precursors. ¹ Based on the SEER 17 database, the annual prevalence of GIST in the United States has increased over the last decade from 0.74 to 0.86 per 100 000 person-years with a total of 5625 new diagnoses in adults between the years of 2010 and 2019. ² It is seen in approximately equal numbers between men versus women and most patients were above the age of 50 at diagnosis. The most common primary site was gastric (55.6%) with the next most common primary sites being small intestine (31.8%), colorectal (6.0%), and esophagus (0.5%), respectively. ³ With improvement of noninvasive diagnostic techniques, GISTs are being diagnosed earlier and patients are being started on treatment earlier. ²

Surgical resection has been the cornerstone of treatment for localized GISTs, particularly those demonstrating high-risk features on endoscopic ultrasound (EUS) or biopsy. ⁴ In limited stage disease, high-risk features include an irregular border, cystic space, ulceration, echogenic foci, or tissue heterogeneity. All tumors greater than 2 cm are also recommended for surgery if it can be done without significant morbidity. Otherwise, GISTs under 2 cm in size are candidates for active surveillance, but if the display any of the aforementioned features, then the patient would need surgical evaluation for consideration of resection. ⁵ However, a subset of these patients present with underlying medical comorbidities that preclude them from undergoing surgical intervention, necessitating a pharmacological approach to management.

Imatinib, a tyrosine kinase inhibitor (TKI), is a cornerstone for the management of GISTs, particularly in the metastatic setting or as adjuvant therapy in high-risk disease. ⁶ Despite its established efficacy, the role of imatinib as a primary treatment modality for patients with localized GIST who are not surgical candidates remains underexplored. More recently, there has been an emerging role for neoadjuvant imatinib for patients with localized disease in anatomically challenging locations. In patients with primary tumors greater than 5 cm in size, there is a proven benefit in median overall survival (OS), but the end goal of all patients on neoadjuvant therapy is still surgical resection. ⁷ This study aims to bridge the remaining knowledge gap by evaluating the outcomes of imatinib monotherapy in a patient population with localized GIST who, due to their comorbid conditions, are unable to undergo standard surgical treatment. This retrospective study followed a cohort of adult patients diagnosed with localized GIST between 2014 and 2023 who were not surgical candidates and subsequently treated exclusively with imatinib. We aim to isolate the impact of imatinib on survival outcomes and disease progression in this unique patient demographic.

Methods and Materials

A single-center, retrospective cohort of adult patients diagnosed with GIST at Northwestern Medicine from 2014 to 2023 who had a pathology-confirmed diagnosis prior to initiation of therapy were included. Of the 344 patients who met these criteria, there were 11 cases identified of localized GIST treated with imatinib monotherapy in lieu of their medical comorbidities precluding them from surgery. All patients who underwent curative-intent surgical resection were excluded from this analysis. All patients with a diagnosis of metastatic GIST were excluded from this analysis. Any patient with another concomitant malignancy was excluded. Data were reviewed from the electronic medical record in accordance with the Northwestern University Institutional Review Board. Patient demographics, GIST disease characteristics, treatment type, and medical comorbidities were assessed at the time of diagnosis.

Statistical analysis was performed using GraphPad Prism 10 (Dotmatics, Boston, Massachusetts). Descriptive statistics were used to characterize the patient population. Progression-free survival (PFS) was defined as the time between date of diagnosis until evidence of progression on interval imaging. Overall survival was defined as the time between date of diagnosis and date of death; patients alive at the time of data collection or lost to follow-up were censored at the date of last follow-up. Both OS and PFS were assessed using Kaplan-Meier survival analysis. The Memorial Sloan Kettering Nomogram was used to model PFS at 24 and 60 months had the patients undergoing surgical resection; survival data were calculated based on primary site, tumor size, and mitotic rate (Memorial Sloan Kettering, New York City, New York). The reporting of this study conforms to the STROBE guidelines for reporting observational studies. ⁸ The study was conducted in accordance with the Helsinki Declaration of 1975 as revised in 2024.

Results

Eleven patients with GIST met our inclusion criteria of localized GIST on imatinib monotherapy without any prior history of surgical resection. The median age was 77 years (range 65-82) and 7 patients (64%) were men. Regarding primary site, 8 of the tumors were gastric, 2 were duodenal, and 1 was esophageal. While our current practice is to perform mutation testing for all patients with GIST at the time of diagnosis, only 4 patients had genomics available: 3 had KIT gene sequence variants and 1 was putatively wild type with no evident alterations on sequencing. Among the KIT-mutated tumors, exon 11 was mutated in 2 tumors and exon 17 in 1 tumor (Table 1). For all cases, documented mitotic rates were less than 5 per 5 mm². Median tumor size was 38.0 mm (range 20-58 mm).

Table 1.

Descriptive data for 11 cases of localized GISTs. While most patients had multiple comorbidities precluding them from surgery, the singular most prohibitive condition was listed above for each patient.

Case	Age / Sex	Primary Tumor	Comorbidities	Mitotic Rate (p/hpf)	Tumor Size (mm)	NGS
1	$82/\mathrm{F}$	$\mathrm{Gastric}$	Lung Disease	<5	$58$	N/A
2	$72/\mathrm{M}$	$\mathrm{Duodenum}$	Heart Disease	<5	$41$	N/A
3	$65/\mathrm{M}$	$\mathrm{Gastric}$	Heart Disease	<5	$20$	Wild Type
4	$71/\mathrm{M}$	$\mathrm{Gastric}$	Heart Disease	<5	32	KIT (exon 17)
5	$81/\mathrm{F}$	$\mathrm{Duodenum}$	Obesity	<5	$38$	KIT (exon 11)
6	$75/\mathrm{M}$	$\mathrm{Gastric}$	Heart Disease	<5	30	N/A
7	$82/\mathrm{F}$	$\mathrm{Gastric}$	Heart Disease	<5	$41$	KIT (exon 11)
8		$\mathrm{Esophagus}$	Deconditioning	$\mathrm{N}/\mathrm{A}$	$42$	N/A
9	$60/\mathrm{M}$	$\mathrm{Gastric}$	Substance Use	$\mathrm{N}/\mathrm{A}$	$50$	N/A
10	$71/\mathrm{M}$	$\mathrm{Gastric}$	Heart Disease	<5	$37$	N/A
11	$73/\mathrm{M}$	$\mathrm{Gastric}$	Obesity	<5	$32$	N/A

Two patients were initially followed with surveillance alone, but started imatinib after tumor progression on interval imaging at 4 months and 4 years postdiagnosis, respectively. The remaining patients immediately started imatinib as they had sufficiently high-risk features to qualify for resection had their comorbidities not precluded them from surgery. All patients underwent multidisciplinary consultation by a team with extensive experience with sarcomas. The most common comorbidities included cardiac disease, obesity, lung disease, and substance use disorder. After review of each patient’s surgical oncology evaluation at the time of diagnosis, the singular most prohibitive comorbidity was heart disease (n = 5). All patients received imatinib monotherapy as their only treatment modality. The median time on imatinib was 16 months. Three patients experienced significant enough adverse events (fatigue, lower extremity edema, and diarrhea, respectively) that they elected to discontinue treatment (Table 2).

Table 2.

Imatinib as a treatment for localized GISTs. Of note, case 2 progressed after coming off imatinib while case 9 progressed on imatinib. Memorial Sloan Kettering provides an online nomogram to predict the likelihood of recurrence-free survival after resection at 24 and 60 months; notably, the nomogram is validated in patients who did not receive tyrosine kinase inhibitors. The data displayed above provide comparative insight into a scenario where patient comorbidities did not preclude them from standard-of-care resection. Nomogram data were unable to be calculated for patients with no mitotic rate documented on their diagnostic pathology.

Case	Time on imatinib (months)	Adverse events	Time to progression (months)	MSK nomogram24-month PFS	MSK nomogram60-month PFS
1	15	$\mathrm{None}$	$\mathrm{Did}\mathrm{not}\mathrm{progress}$	94%	89%
2	40	$\mathrm{Fatigue}$	$54$	87%	75%
3	31	$\mathrm{None}$	$\mathrm{Did}\mathrm{not}\mathrm{progress}$	95%	97%
4	2	$\mathrm{None}$	$\mathrm{Did}\mathrm{not}\mathrm{progress}$	97%	93%
5	16	$\mathrm{None}$	$\mathrm{Did}\mathrm{not}\mathrm{progress}$	87%	76%
6	3	$\mathrm{None}$	$\mathrm{Did}\mathrm{not}\mathrm{progress}$	97%	94%
7	11	$\mathrm{LE}\mathrm{Edema}$	$\mathrm{Did}\mathrm{not}\mathrm{progress}$	96%	92%
8	$79$	$\mathrm{None}$	$\mathrm{Did}\mathrm{not}\mathrm{progress}$	N/A	N/A
9	$41$	$\mathrm{Diarrhea}$	$8$	N/A	N/A
10	46	$\mathrm{None}$	$\mathrm{Did}\mathrm{not}\mathrm{progress}$	96%	93%
11	$2$	$\mathrm{None}$	$\mathrm{Did}\mathrm{not}\mathrm{progress}$	97%	93%

Two patients had progression of disease with interval enlargement of the tumor on imaging at 54 months while off imatinib (Table 2, case 2) and at 8 months while on imatinib (Table 2, case 10). With a median follow-up of 35 months, the 1-year PFS and OS were 90% and 100%. The 3-year PFS and OS were also 90% and 100% (Figure 1). Memorial Sloan Kettering nomogram data were calculated for each patient to predict how they would have performed had they been able to tolerate surgical resection. Using tumor size, mitotic rate, and primary site, nomogram data for 24-month PFS ranged from 87% to 97% and 60-month PFS ranged from 75% to 97%. The 2 patients with duodenal primary tumors had the worse predicted PFS data when compared with gastric primary tumors. We were unable to calculate nomogram data for our singular esophageal primary tumor as we did not have a recorded mitotic rate.

Figure 1.

Survival curves for patients with GIST on imatinib monotherapy. Median follow-up time was 35 months. (A) Progression-free survival and (B) Overall survival from date of diagnosis.

Discussion

Gastrointestinal stromal tumors account for approximately 2% of all gastrointestinal malignancies and the overwhelming majority are sporadic occurring in patients with no known risk factors or relevant family history.^2,4 Therefore, an understanding of the molecular alterations specific to each individual tumor is critical to effective clinical management. In totality, about 80% of GISTs have a mutation in the KIT gene, another 10% have a mutation in the PDGFRA (platelet-derived growth factor receptor alpha) gene, and the remaining 10% are a collective variety of other driver mutations. ⁹ Both KIT and PDGFRA are receptor tyrosine kinases responsible for cellular pathways that promote proliferation and suppress apoptosis. ¹⁰ Oncogenic mutation in KIT and PDGFR lead to constitutive activation of their pathways and uncontrolled tumor growth. ¹¹ These oncogenic mutations allow for the use of targeted TKIs in treatment of GIST; the introduction of imatinib for treatment of GISTs drastically improved clinical outcomes and established a principal role for TKIs in management of these tumors.

Imatinib is considered standard therapy for metastatic or unresectable GISTs and was the first targeted therapy approved for treatment of advanced GISTs more than 2 decades ago. ⁶ Mechanistically, imatinib is a 2-phenylaminopyrimidine molecule that selectively inhibits both the BCR/ABL fusion protein in chronic myeloid leukemia and a subset of type 3 transmembrane receptor tyrosine kinases (including KIT and PDGFRA). ¹² Among patients with stage IV GISTs, approximately 70% of patients see a partial response with 20% achieving stable disease. ¹³ The most common adverse events on imatinib include edema, nausea, diarrhea, and musculoskeletal pain, but the drug is generally well tolerated. ¹⁴ Standard dose of imatinib has been determined to be 400 mg daily. In both the landmark EORTC 62005 and SWOG S0033 trials, there was no difference in overall response rate or OS between patients with metastatic or unresectable GISTs receiving 400 mg versus 800 mg of imatinib daily; however, patients in the higher arm dose did experience significantly greater toxicities.^15,16 In an update to the SWOG S0033 trial, the estimated 10-year OS across both study arms was 22%. The strongest prognostic indicator among patients was the molecular characteristics of their tumor; patients with a KIT exon 9 mutation had significantly worse OS than those with the KIT exon 11 mutation or wild-type genotype. ¹⁷

For patients with localized GISTs, complete surgical resection defined as complete removal of the tumor with an intact pseudo-capsule and negative margins (R0 resection) is the standard of care. For localized tumors, resection is indicated if the tumor is greater than 2 cm in size; otherwise, the patient would qualify for endoscopic or radiographic surveillance. ¹⁸ Since GISTs are highly resistant to chemotherapy, adjuvant imatinib has been employed for high-risk tumors (determined via modified National Institutes of Health criteria accounting for size, location and mitotic rate) in 2 randomized phase III clinical trials. The ACOSOG Z9001 trial by DeMatteo et al ¹⁹ demonstrated improved recurrence-free survival for 1 year of adjuvant imatinib compared with placebo in patients (98% vs 83% at 1 year) with GISTs greater than 3 cm in size. The updated SSG trial XVIII analysis by Joensuu et al ²⁰ demonstrated superiority for 3 years of adjuvant therapy versus 1 year with improvement of relapse-free survival to 71.1% from 52.3% at 5 years. Thus, adjuvant imatinib has a clearly defined role in management of localized resected GISTs; subsequently, this has bolstered research into the potential for neoadjuvant imatinib. In patients with locally advanced disease, treatment with neoadjuvant imatinib with the goal of subsequent R0 resection is considered if (a) there is a high risk of tumor rupture or (b) the tumor is in a precarious location such as the gastroesophageal junction, duodenum, or rectum wherein anatomic function is high risk for morbidity with surgery. ⁴ Although it is difficult to predict initial disease response after therapy initiation, the anticipated goal would is approximately 6 to 9 months of imatinib therapy prior to resection. ²¹ Thus, there is precedent for imatinib therapy before resection and while not curative, it may downsize the size of the original tumor allowing for organ-preserving surgery.

In our study, we identified a subset of patients with localized GISTs that would normally proceed to resection in patients who were precluded from surgery due to their medical comorbidities. All 11 patients had tumors that would have qualified for surgical resection on date of diagnosis. To date, this is the first study in the literature to specifically examine the role of imatinib monotherapy in nonmetastatic GISTs that would normally proceed to surgery. The most similar population previously studied would be tumors treated with neoadjuvant therapy that did not achieve sufficient response to ultimately go to surgery; however, those patients presented with significantly more advanced disease with tumors often greater than 10 cm in size to qualify for neoadjuvant treatment in the first place. Our cohort presented with a more moderate tumor burden, and captures a different subset of the GIST population: those with the most significant medical comorbidities. One important limitation to this study is its retrospective and observational nature—power analysis was not conducted a priori as our sample size was fixed to the number of cases identified at our single institution center. In future work, a multicenter analysis would help bolster statistical power and confirm the trends observed in our patient cohort.

The option of medical management with imatinib, if viable, would provide an attractive option to patients after a difficult conversation about their surgical risks precluding them from standard of care. Nine of the 11 patients were immediately started on imatinib after diagnosis as they were still interested in pursuing nonsurgical treatment options. Two of the 11 patients expressed a preference for an initial trial of surveillance at diagnosis, but elected to start imatinib therapy after interval growth on follow-up scans. Using the MSK nomogram data, we estimated the PFS for each case in the event that they had gone for R0 resection with collective results across the 11 tumors yielding anywhere from 87% to 97% of PFS at 2 years post-resection. In comparison, the PFS at 2 years among our data set was 90% The discrepancy in estimated PFS from the MSK nomogram was expected as our patient cohort has a higher comorbidity burden; at baseline, they were not candidates for the surgical resection that patients captured by the MSK nomogram underwent.

Imatinib is generally well tolerated. The most common side effects of edema, nausea, and fatigue are typically mild and self-limited grade I or grade II adverse events. In patients with bulkier tumors, risk of tumor lysis syndrome and hyperuricemia is rare even with rapid tumor shrinkage in strong responders. ²² Among grade III and grade IV adverse events, the most common is myelosuppression. ²³ Le Cesne et al ²⁴ observed severe neutropenia in 7.0% of patients on imatinib for advanced GIST in the EORTC/ISG/AGITG trial. In comparison, DeMatteo et al ¹⁹ observed severe neutropenia in 3.6% of patients on adjuvant imatinib after resection of localized GIST in the Z9001 study. In earlier years, there was also concern for imatinib-mediated cardiotoxicity, but subsequent analysis of clinical trial data demonstrated that heart failure incidence was equivalent in patients on imatinib and the general population. ²⁵ Patients on adjuvant imatinib often finish the full 3 years of therapy without issue; side effects occur in a dose-dependent manner and are more likely to manifest early in the treatment course.^20,25 While more research is needed to determine the optimal length of imatinib therapy for patients in our cohort with localized disease precluded from surgery, there is strong precedent in the literature for patients on imatinib for several years with few long-term side effects.

While TKIs have revolutionized the landscape of GIST treatment, the standard of care is still continuously evolving and advancing. Even decades after it was first approved, imatinib is still under examination. Recently, the IMADGIST trial has shown preliminary evidence now favoring 6 years of adjuvant imatinib over 3 years in KIT-mutated GISTs after R0 or R1 resection. ²⁶ For rarer molecular subtypes of GIST such as the imatinib-resistant PDFGRA D842V mutant, the next-generation TKI avapritinib has been approved for use in the metastatic setting providing an example of where imatinib is no longer the first-line TKI. ²⁷ As more examples of novel TKIs emerge, there will be new research needed to examine their role in the primary, neoadjuvant, and adjuvant therapy. While treatment of advanced GIST understandably receives the majority of the research focus, there is still progress to be made in the optimization of management for localized and early-stage tumors.

Conclusions

For patients with localized GISTs who are precluded from standard-of-care surgical resection due to medical comorbidities, imatinib monotherapy is a viable alternative. With a median follow-up of 35 months, the 1-year PFS and OS were 90% and 100%, respectively. Three-year PFS and OS were also 90% and 100%, respectively. Patients generally tolerated imatinib well. Our research provides clinical evidence supporting the use of imatinib in a patient population that is precluded from standard-of-care resection for the disease.