Metabolic profile alterations as early indicators of gastrointestinal stromal tumors (GISTs) recurrence and imatinib resistance

Introduction

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumor of the gastrointestinal (GI) tract and originate from the interstitial cells of Cajal, cells that serve as the pacemaker of the GI tract. They account for 1–2% of all GI malignancies, and over 95% of GISTs are positive for the KIT protein (CD117) by immunohistochemical staining, and about 80–90% of GISTs harbor a mutation in the KIT and PDGFɑ genes (1).

Depending on the risk of the GISTs, surgery and adjuvant imatinib can be used for treatment (1-3). Due to these advancements, recurrence of GIST for low and intermediate-risk tumors is low, ranging from 0–2.4% and 1.9–20%, respectively (4,5). However, with high-risk tumors, the chance of recurrence is dramatically higher with a 5-year recurrence rate of 47% (6). For some patients, they can develop resistance to imatinib due to new kinase mutations, leading to disease progression (7-9).

Resistance mechanisms can be categorized as primary (intrinsic) or secondary (acquired), with primary resistance often linked to specific KIT and PDGFRɑ (7,8). Secondary resistance typically arises through the emergence of additional point mutations in the ATP-binding pocket or activation loop of the kinase domain, which interfere with drug binding (8,9). The median time to development of secondary resistance is approximately 24 months, highlighting the need for vigilant monitoring strategies to detect disease progression early (7,9).

Of note, imatinib has been shown to have positive effects on lipid metabolism in those taking it for chronic myeloid leukemia (CML) and GIST (10-13). Prior studies have documented improvements in cholesterol, triglycerides (TG), and glycemic markers in patients on imatinib therapy for both CML and GIST (12,13). However, the relationship between changes in these metabolic parameters and disease progression or treatment resistance has not been well characterized. From a mechanistic standpoint, the favorable metabolic effects of imatinib are thought to be mediated through its inhibition of ABL1, PDGFR, and c-KIT signaling pathways, which intersect with lipid homeostasis and insulin signaling. Specifically, imatinib-mediated inhibition of PDGFRα reduces adipocyte differentiation and systemic lipid accumulation, while ABL1 inhibition modulates hepatic lipid metabolism (14,15). When secondary resistance mutations emerge—particularly in the ATP-binding pocket or activation loop of the KIT kinase domain—imatinib can no longer effectively suppress these oncogenic pathways (7-9). The consequent restoration of unopposed KIT/PDGFR signaling may therefore reverse the lipid-lowering and anti-inflammatory effects that were previously maintained under adequate drug suppression, providing a plausible biological link between metabolic deterioration and acquired resistance. Furthermore, tumor progression itself may contribute to systemic metabolic dysregulation through increased inflammatory cytokine production, altered adipokine signaling, and cancer-associated cachexia-related mechanisms, though the relative contribution of these factors in GIST requires further investigation.

In this study, we investigate trends in metabolic profiles for patients who have recurrence of GIST post-resection while taking imatinib. We selected metabolic biomarkers as the focus of this investigation for several reasons. First, while secondary kinase mutations are the established mechanism of imatinib resistance, molecular testing requires tumor tissue and is typically performed only after clinical or radiographic progression is already evident (7-9). Second, standard imaging surveillance is associated with substantial cumulative radiation exposure and cost (2). Metabolic markers, by contrast, are routinely measured in primary care and oncology follow-up visits, are minimally invasive and inexpensive, and can theoretically reflect systemic drug efficacy in real time (12,13). If metabolic deterioration precedes detectable recurrence by weeks to months, these biomarkers could offer a clinically actionable early warning signal that complements rather than replaces imaging surveillance. Furthermore, we hypothesize that these metabolic biomarkers could serve as surrogates indicating recurrence of GIST or resistance to imatinib and therefore, guide management. We present this article in accordance with the REMARK reporting checklist (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-1-0182/rc).

Methods

In this retrospective cohort study, we analyzed patient records from the TriNetX Research Network (TRN) database, which aggregates clinical data for more than 200 individuals across 172 healthcare systems in 18 countries (16). As the TRN utilizes de-identified patient data and Healthcare Organizations (HCOs) in accordance with Section §164.514 of the Health Insurance Portability and Accountability Act (HIPAA) Privacy Rule, the data is not considered protected health information and therefore did not require ethical approval (TriNetX 1). The study adhered to the ethical and reporting standards outlined in the TriNetX “Publication Guidelines” (17). The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Patient selection and inclusion criteria

TrinetX utilizes various coding terms such as International Classification of Diseases, 10^th revision (ICD-10) codes to identify patients with related diagnoses, current procedural terminology (CPT) codes for procedures, and RxNorm codes for medications (see Tables S1-S4 for specific codes used). We included adult patients (≥18 years) with a diagnosis of a GIST who were positive for KIT protein (CD117) from January 1, 2005 to January 1, 2025. Eligible patients underwent surgical resection of their primary GIST and were subsequently treated with imatinib as adjuvant therapy. Patients were then required to have disease recurrence during the study period. This was an exploratory retrospective study utilizing all eligible cases within the TriNetX database. No formal power calculation was performed.

Data collection

Baseline demographic and clinical characteristics were also extracted from the database, including age at time of GIST diagnosis, sex, ethnicity, and race. These variables were analyzed descriptively to characterize the study population.

This study examined several metabolic biomarkers: total cholesterol, TG, high-density lipoprotein (HDL), TG/HDL ratio, low-density lipoprotein (LDL), hemoglobin A1c, body mass index (BMI), and thyroid-stimulating hormone (TSH). Laboratory values were extracted at specific time points relative to the date of recurrence diagnosis: 12 months prior (baseline), 9 months prior, 6 months prior, 3 months prior, 1 month prior, and 1–12 months following recurrence diagnosis. Additionally, we performed one-tailed t-testing with the Pearson correlation coefficient comparing 12 months prior to diagnosis of recurrence versus 9, 6, 3, and 1 months prior to diagnosis of recurrence. These time intervals were selected to capture progressive metabolic changes in the year leading up to recurrence while allowing sufficient time between measurements to detect meaningful trends.

Statistical analysis

Statistical analyses were performed using R. For comparison of metabolic parameters before and after recurrence diagnosis, we performed two-tailed independent t-tests comparing mean values from 12 to 1 month prior to recurrence diagnosis (pre-recurrence period) versus mean values from 1 to 12 months after recurrence diagnosis (post-recurrence period). The data points collected during the time of diagnosis were excluded from the analysis.

To assess temporal trends in metabolic biomarkers leading up to recurrence, we performed one-tailed t-tests with Pearson correlation coefficients comparing values at 12 months prior to recurrence (baseline) versus values at 9, 6, 3, and 1 months prior to recurrence. One-tailed testing was employed based on our hypothesis that metabolic parameters would worsen (increase for cholesterol, LDL, TG, BMI; decrease for HDL, TSH) as patients approached recurrence. The Pearson correlation coefficient (R) was calculated to quantify the strength and direction of linear relationships between time and biomarker values.

A P value of <0.05 was considered statistically significant for all analyses. Continuous variables are reported as means with standard deviations or medians with interquartile ranges as appropriate. Categorical variables are presented as frequencies and percentages. Missing data were handled using complete-case analysis at each time interval. Metabolic biomarkers were analyzed as continuous variables, and no data-driven cutpoints were used.

Results

Study cohort

A total of 36 patients met all inclusion criteria and were included in the final analysis.

Patient characteristics

The mean age of patients at diagnosis was 61.7 years. The mean age of patients at diagnosis was 61.7 years. The cohort consisted of 20 males (55.6%), 15 females (41.7%), and 1 patient with unknown sex (2.8%). Regarding ethnicity, 30 patients (83.3%) were not Hispanic or Latino, 3 (8.33%) were Hispanic or Latino, and 3 (8.3%) had unknown ethnicity. Some patients identified with multiple races, and the racial distribution included 26 White patients (72.2%), 5 Black or African American patients (13.9%), and smaller proportions of American Indian or Alaskan (8.3%), Asian (8.3%), Native Hawaiian or other Pacific Islander (8.3%), other race (8.3%), and unknown race (8.3%) (Table 1).

On average, patients had GIST resection surgery 25.9 months prior to recurrence diagnosis.

Baseline metabolic profiles

At 12 months prior to recurrence diagnosis, the mean baseline metabolic values were as follows: total cholesterol (140 mg/dL), LDL (75.4 mg/dL), HDL (56 mg/dL), TG (96 mg/dL), TG/HDL ratio (2.14), hemoglobin A1c (6.80%), BMI (28.3 kg/m²), and TSH (3.15 mIU/L).

Temporal trends in metabolic parameters leading to recurrence

One-tailed t-testing with Pearson correlation analysis was performed to assess progressive changes in metabolic biomarkers during the 12-month period preceding recurrence diagnosis (Table 2). Total cholesterol demonstrated a very strong positive correlation with time approaching recurrence (R=0.92, P=0.03), rising from 140 mg/dL at 12 months prior to 186 mg/dL at 1 month prior to recurrence, indicating a consistent upward trend in the year leading to recurrence diagnosis (Figure 1). LDL cholesterol similarly showed a strong positive correlation (R=0.87, P=0.050), increasing from 75.4 mg/dL at baseline to 104 mg/dL at 1 month before recurrence (Figure 2). TG exhibited a very strong positive correlation (R=0.93, P=0.02), increasing from 97 to 163 mg/dL over the 12-month pre-recurrence period, demonstrating a marked elevation over time (Figure 3).

Figure 1 Total cholesterol trend relative to diagnosis of recurrence.

Figure 2 LDL trend relative to diagnosis of recurrence. LDL, low-density lipoprotein.

Figure 3 TG trend relative to diagnosis of recurrence. TG, triglycerides.

In contrast, HDL cholesterol showed a moderate negative correlation (R=−0.70, P=0.19), declining from 56.0 mg/dL at 12 months prior to 45.0 mg/dL at 1 month prior to recurrence, although this trend did not reach statistical significance (Figure 4). TSH demonstrated a strong negative correlation (R=−0.88, P=0.049), decreasing from 3.25 to 1.80 mIU/L over the 12-month period (Figure 5). BMI showed minimal correlation with time to recurrence (R=−0.03, P=0.96), remaining relatively stable between 28.4 and 29.7 kg/m² throughout the pre-recurrence period, suggesting that while BMI was altered, it did not demonstrate a consistent progressive temporal trend (Figure 6).

Figure 4 HDL trend relative to diagnosis of recurrence. HDL, high-density lipoprotein.

Figure 5 TSH trend relative to diagnosis of recurrence. TSH, thyroid-stimulating hormone.

Figure 6 BMI trend relative to diagnosis of recurrence. BMI, body mass index.

Hemoglobin A1c and TG/HDL ratio did not show significant correlations in the temporal trend analysis (P>0.05 for both).

Comparison of pre-recurrence versus post-recurrence metabolic parameters

Two-tailed t-test analysis comparing mean metabolic biomarker values from the 12-month pre-recurrence period versus the 1–12-month post-recurrence period revealed significant differences in several parameters (Table 3). Total cholesterol was significantly elevated in the pre-recurrence period (169.4 mg/dL) compared to post-recurrence (141.8 mg/dL), with a mean difference of 27.6 mg/dL (P=0.02) (Figure 1). LDL cholesterol was significantly increased prior to recurrence (91.7 mg/dL) compared to post-recurrence (71.8 mg/dL), with a mean difference of 19.9 mg/dL (P=0.01) (Figure 2). HDL cholesterol was significantly decreased in the pre-recurrence period (49.1 mg/dL) compared to post-recurrence (61.6 mg/dL), with a mean difference of −12.5 mg/dL (P=0.03) (Figure 4). BMI was modestly but significantly elevated prior to recurrence (28.9 kg/m²) compared to post-recurrence (28.6 kg/m²), with a mean difference of 0.3 kg/m² (P=0.047) (Figure 6).

TG showed a trend toward elevation in the pre-recurrence period (132.8 mg/dL) compared to post-recurrence (121.2 mg/dL), with a mean difference of 11.6 mg/dL, though this did not reach statistical significance in the two-tailed test (P>0.05) (Figure 3). TSH also showed a trend toward lower values in the pre-recurrence period (2.25 mIU/L) compared to post-recurrence (2.14 mIU/L), with a mean difference of 0.11 mIU/L (P>0.05) (Figure 5). Hemoglobin A1c and TG/HDL ratio showed no statistically significant differences between pre-recurrence and post-recurrence periods (P>0.05 for both).

Post-recurrence treatment

Following diagnosis of recurrence, treatment regimens were modified for all patients. The majority of patients (31/36, 86.1%) transitioned to sunitinib as second-line therapy. Three patients (8.3%) received regorafenib, one patient (2.8%) received ripretinib, and 1 patient (2.8%) received avapritinib. Following initiation of these alternative tyrosine kinase inhibitors (TKIs), the previously observed upward trends in adverse cardiometabolic markers stabilized. Post-recurrence mean values for total cholesterol (141.8 mg/dL), LDL (71.8 mg/dL), and TG (121.2 mg/dL) remained relatively stable during the 1–12-month post-recurrence period, while HDL improved to 61.6 mg/dL, suggesting restoration of more favorable metabolic profiles with effective second-line therapy.

Discussion

This study found that those patients’ who underwent surgery for GIST followed by imatinib and eventually had GIST recurrence had worsening metabolic profiles until the time of recurrence. Specifically, total cholesterol and TG were significantly trending up, LDL demonstrated an increasing trend, TSH was significantly trending down, HDL demonstrated a non-significant declining trend, and BMI did not demonstrate a significant temporal trend. Notably, once patients began on a further treatment such as sunitinib, these cardiometabolic markers no longer were significantly uptrending and stabilized.

Imatinib has been shown to have significant alterations in lipid and glycemic profiles in patients with GIST (12,13). This occurs through several mechanisms. First, by competing with carnitine at the OCTN2 transporters found in the gut lumen, it impairs intestinal fatty acid absorption and correlates with improved lipid profiles (14). Second, imatinib has also been shown to upregulate LDL receptor expression, reduce uptake of LDL by macrophages and attenuate cholesterol uptake (15). Third, it has been shown to reduce systemic and adipose tissue inflammation, leading to improved insulin sensitivity (18).

The significant worsening of patients lipid profiles leading up to the time of recurrence diagnosis is suggestive of imatinib resistance and increased tumor burden from recurrence. Because BMI did not significantly increase during this period, weight-related metabolic changes are unlikely to confound this trend. Similarly, TSH values were downtrending rather than rising, making thyroid dysfunction an improbable contributor. Given that recurrence occurred at a mean of approximately 26 months post-resection, all patients in this cohort likely developed secondary imatinib resistance. This is consistent with prior studies showing that acquired resistance typically emerges around 24 months and often involves secondary mutations in the KIT kinase domain. The biological significance of the observed correlations warrants consideration. The very strong positive correlation between total cholesterol and time to recurrence (R=0.92) and between TG and time to recurrence (R=0.93) suggests that progressive lipid dysregulation is not a random fluctuation but rather tracks closely with the underlying disease course. This is biologically plausible: as imatinib’s suppression of KIT/PDGFRα signaling wanes due to resistance mutations, the drug’s previously established mechanisms of lipid lowering—including upregulation of LDL receptor expression, inhibition of macrophage cholesterol uptake, and impairment of intestinal fatty acid absorption via OCTN2 competition—are progressively reversed (14,15). The moderate negative correlation of HDL (R=−0.70) is consistent with systemic inflammatory activation, as HDL is an acute-phase reactant that declines in states of chronic inflammation, which may be promoted by expanding tumor burden (18). The strong negative TSH correlation (R=−0.88) is an intriguing observation whose mechanistic basis is less well established; possible explanations include imatinib’s known interactions with thyroid hormone metabolism or, alternatively, a paraneoplastic effect of advancing GIST on the hypothalamic-pituitary-thyroid axis, though this requires dedicated investigation. Importantly, these correlations represent observed associations in a small retrospective cohort and should be interpreted as hypothesis-generating rather than mechanistically conclusive. Nevertheless, the consistency and directionality of the findings across multiple independent biomarkers lend biological credibility to the pattern (7-9).

Therefore, we posit that these cardio-metabolic parameters can potentially be harnessed as tools for early detection of these events. For example, this may prompt an oncologist to consider more frequent imaging studies or switching to another TKI such as sunitinib, at an earlier date (19-22).

2025 National Comprehensive Cancer Network (NCCN) guidelines recommend routine imaging surveillance with computed tomography (CT) abdomen/pelvis with contrast or magnetic resonance imaging (MRI) abdomen/pelvis with and without contrast every 3–6 months for 3–5 years following resection of high-risk GIST (2). This screening represents a substantial cumulative radiation exposure and healthcare cost burden. Per year, this imaging for CT scans ranges from $500–3,000 per scan, and so biannual screening would cost $1,000–6,000 per year in the United States (US) (23). In contrast, routine metabolic testing would cost around $50–120 per bundle, so if applied every 3 months, would cost less than $500 per year (24). Therefore, if this approach were to be validated, incorporating cardiometabolic surveillance could serve as a cost-effective adjunct to imaging. Lastly, as these tests are routinely drawn in primary care, integration of cardiometabolic surveillance would require minimal additional infrastructure, workflow modification, or patient burden, further enhancing its feasibility.

This study has several limitations to note. There may be confounding factors that can be a cause of worsening patient profiles—such patients with recurrence likely began with a higher risk tumor and therefore may be at baseline sicker and likelier to develop such profiles. Additionally, the nature of large retrospective cohort analyses limits the ability to assess patient compliance on imatinib. The most notable limitation is the sample size our study collected. Given the criteria selected, these patients are rare, so such a limitation may not be easily overcome; however, we hope that this study can provide insights into future studies on this population. Lastly, this study lacks a matched control cohort of resected GIST patients receiving imatinib without recurrence, limiting our ability to determine whether similar findings occur in non-recurrent cases. As with all retrospective cohort designs, this study is subject to potential information bias and selection bias that warrant consideration. Information bias may arise from the reliance on administrative and clinical records within the TriNetX database: metabolic laboratory values were not systematically collected at uniform intervals for all patients, and the completeness of data at each time point varied. The use of complete-case analysis at each interval minimizes distortion from missing data but may introduce a selection effect if patients with missing values differ systematically from those with complete records. Recall bias, in the traditional sense, is less applicable here given the use of an electronic health record-based database rather than patient self-report; however, ascertainment bias remains a concern, as more clinically active or closely followed patients may have more complete laboratory records, potentially skewing the sample toward individuals receiving more intensive monitoring. Additionally, confounding by concurrent medications—particularly statins, lipid-lowering agents, or thyroid medications—cannot be excluded, as these data were not systematically captured or adjusted for in the analysis. Future prospective studies should address these limitations through standardized data collection protocols, predefined measurement intervals, and systematic documentation of concomitant therapies.

Prospective studies are warranted to validate these findings and determine clinically meaningful cutoff values for metabolic biomarkers that can guide treatment decisions in GIST patients on imatinib therapy. Future research should evaluate whether incorporating metabolic monitoring into surveillance protocols allows for earlier detection of disease progression compared to imaging alone, and whether this translates to improved clinical outcomes. Additionally, investigating the relationship between specific secondary resistance mutations and patterns of metabolic change may enable more targeted selection of subsequent TKIs. Larger multi-institutional studies will be essential to establish evidence-based guidelines for implementing metabolic biomarker surveillance in routine GIST management. From a practical clinical implementation standpoint, several considerations are worth addressing. First, because lipid panels and TSH are already routinely obtained during oncology and primary care visits, integration of structured metabolic surveillance into existing GIST follow-up protocols would require minimal additional infrastructure. A pragmatic approach would be to standardize the timing and frequency of these measurements—for example, at each scheduled follow-up visit every 3 months in the first 3 years post-resection—and to establish patient-specific baseline values at treatment initiation against which subsequent trends can be compared (2). Second, this study suggests that trajectory, rather than absolute threshold, is the clinically meaningful signal: a sustained upward trend in total cholesterol and TG or a progressive decline in HDL over two or more consecutive visits may be more informative than any single out-of-range value. Defining algorithmically actionable thresholds for such trends is a critical next step. Third, if validated, this approach could be implemented in a tiered manner: metabolic deterioration could prompt more frequent imaging or referral to a GIST specialist, without necessarily triggering immediate empiric therapy changes in the absence of confirmatory imaging (19-22). Finally, an important direction for future research is the development and prospective validation of a composite metabolic risk score integrating multiple biomarkers weighted by their respective correlation strengths, which could provide a more robust and clinically interpretable early warning signal than any single marker alone.

Conclusions

This retrospective cohort study demonstrates that progressive worsening of cardiometabolic parameters—specifically rising total cholesterol, LDL cholesterol (LDL-C), and TG alongside declining HDL cholesterol (HDL-C) and TSH—precedes clinical detection of GIST recurrence in patients receiving adjuvant imatinib following surgical resection. Several of these metabolic trends exhibited significant temporal correlations in the 12 months leading up to recurrence diagnosis and are biologically consistent with the reversal of imatinib’s established lipid-lowering mechanisms as secondary kinase resistance emerges. Importantly, the stabilization and normalization of these parameters following initiation of second-line TKI therapy further supports the mechanistic link between imatinib efficacy, metabolic homeostasis, and disease control.

Given that routine lipid panels and TSH measurements are already embedded in standard oncologic and primary care follow-up, structured metabolic surveillance represents a low-cost, minimally invasive complement to imaging-based monitoring. While this study is limited by its small sample size, retrospective design, absence of a non-recurrent control cohort, and inability to account for concomitant lipid-modifying medications, the consistency and directionality of findings across multiple independent biomarkers provide a compelling hypothesis-generating foundation. Prospective, multi-institutional studies are warranted to validate these findings, define clinically actionable trend-based thresholds, and determine whether early metabolic deterioration can reliably trigger timely therapeutic intervention—ultimately improving outcomes for patients with high-risk GIST.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the REMARK reporting checklist. Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-1-0182/rc

Peer Review File: Available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-1-0182/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jgo.amegroups.com/article/view/10.21037/jgo-2026-1-0182/coif). M.A. reports consulting fees from AADi, Boehringer Ingelheim, SpringWorks Therapeutics, and Daiichi Sankyo; and participation in the Speakers’ Bureau for Deciphera. The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Sanchez-Hidalgo JM, Duran-Martinez M, Molero-Payan R, et al. Gastrointestinal stromal tumors: A multidisciplinary challenge. World J Gastroenterol 2018;24:1925-41. [Crossref] [PubMed]
2. von Mehren M, Kane JM, Riedel RF, et al. NCCN Guidelines® Insights: Gastrointestinal Stromal Tumors, Version 2.2022. J Natl Compr Canc Netw 2022;20:1204-14. [Crossref] [PubMed]
3. Dematteo RP, Ballman KV, Antonescu CR, et al. Adjuvant imatinib mesylate after resection of localised, primary gastrointestinal stromal tumour: a randomised, double-blind, placebo-controlled trial. Lancet 2009;373:1097-104. [Crossref] [PubMed]
4. Nilsson B, Bümming P, Meis-Kindblom JM, et al. Gastrointestinal stromal tumors: the incidence, prevalence, clinical course, and prognostication in the preimatinib mesylate era--a population-based study in western Sweden. Cancer 2005;103:821-9. [Crossref] [PubMed]
5. Tryggvason G, Gíslason HG, Magnússon MK, et al. Gastrointestinal stromal tumors in Iceland, 1990-2003: the icelandic GIST study, a population-based incidence and pathologic risk stratification study. Int J Cancer 2005;117:289-93. [Crossref] [PubMed]
6. Joensuu H, Eriksson M, Sundby Hall K, et al. One vs three years of adjuvant imatinib for operable gastrointestinal stromal tumor: a randomized trial. JAMA 2012;307:1265-72. [Crossref] [PubMed]
7. Maleddu A, Pantaleo MA, Nannini M, et al. Mechanisms of secondary resistance to tyrosine kinase inhibitors in gastrointestinal stromal tumours Oncol Rep 2009;21:1359-66. (Review). [Crossref] [PubMed]
8. Chen LL, Trent JC, Wu EF, et al. A missense mutation in KIT kinase domain 1 correlates with imatinib resistance in gastrointestinal stromal tumors. Cancer Res 2004;64:5913-9. [Crossref] [PubMed]
9. Antonescu CR, Besmer P, Guo T, et al. Acquired resistance to imatinib in gastrointestinal stromal tumor occurs through secondary gene mutation. Clin Cancer Res 2005;11:4182-90. [Crossref] [PubMed]
10. Breccia M, Muscaritoli M, Aversa Z, et al. Imatinib mesylate may improve fasting blood glucose in diabetic Ph+ chronic myelogenous leukemia patients responsive to treatment. J Clin Oncol 2004;22:4653-5. [Crossref] [PubMed]
11. Breccia M, Muscaritoli M, Alimena G. Reduction of glycosylated hemoglobin with stable insulin levels in a diabetic patient with chronic myeloid leukemia responsive to imatinib. Haematologica 2005;90:ECR21.
12. Markovits N, Kurnik D, Friedrich C, et al. Effects of imatinib on glycemic and lipid profiles: a retrospective cohort study. Leuk Lymphoma 2022;63:2224-32. [Crossref] [PubMed]
13. Gómez-Sámano MÁ, Baquerizo-Burgos JE, Coronel MFC, et al. Effect of imatinib on plasma glucose concentration in subjects with chronic myeloid leukemia and gastrointestinal stromal tumor. BMC Endocr Disord 2018;18:77. [Crossref] [PubMed]
14. Burda P, Hlavackova A, Polivkova V, et al. Imatinib therapy of chronic myeloid leukemia significantly reduces carnitine cell intake, resulting in adverse events. Mol Metab 2024;88:102016. [Crossref] [PubMed]
15. Gacic J, Vorkapic E, Olsen RS, et al. Imatinib reduces cholesterol uptake and matrix metalloproteinase activity in human THP-1 macrophages. Pharmacol Rep 2016;68:1-6. [Crossref] [PubMed]
16. TriNetX. Accessed December 23, 2025. Available online: https://live.trinetx.com/
17. TriNetX. Publication guidelines. Accessed December 23, 2025. Available online: https://trinetx.com/real-world-resources/case-studies-publications/trinetx-publication-guidelines/
18. AlAsfoor S, Rohm TV, Bosch AJT, et al. Imatinib reduces non-alcoholic fatty liver disease in obese mice by targeting inflammatory and lipogenic pathways in macrophages and liver. Sci Rep 2018;8:15331. [Crossref] [PubMed]
19. Kee D, Zalcberg JR. Current and emerging strategies for the management of imatinib-refractory advanced gastrointestinal stromal tumors. Ther Adv Med Oncol 2012;4:255-70. [Crossref] [PubMed]
20. Hsu JT, Le PH, Kuo CF, et al. Imatinib dose escalation versus sunitinib as a second-line treatment against advanced gastrointestinal stromal tumors: A nationwide population-based cohort study. Oncotarget 2017;8:71128-37. [Crossref] [PubMed]
21. Platoff RM, Morano WF, Marconcini L, et al. Recurrent Gastrointestinal Stromal Tumors in the Imatinib Mesylate Era: Treatment Strategies for an Incurable Disease. Case Rep Oncol Med 2017;2017:8349090. [Crossref] [PubMed]
22. Nishida T, Blay JY, Hirota S, et al. The standard diagnosis, treatment, and follow-up of gastrointestinal stromal tumors based on guidelines. Gastric Cancer 2016;19:3-14. [Crossref] [PubMed]
23. MDsave. CT scan of abdomen and pelvis with contrast. Accessed December 10, 2025. Available online: https://www.mdsave.com/procedures/ct-scan-of-abdomen-and-pelvis-with-contrast/d781f5c4
24. GoodRx. How much is blood work without insurance? 2024. Accessed December 10, 2025. Available online: https://www.goodrx.com/health-topic/diagnostics/how-much-is-blood-work-without-insurance