GISTs are a subgroup of mesenchymal tumors arising from the gastrointestinal tract and have been considered as smooth muscle tumors (such as leiomyomas or leiomyosarcomas) based on the histologic characteristics (1). Until the discovery of KIT (CD117) expression in GIST (2), the origin of GIST was proposed to be from the interstitial cells of Cajal (ICCs) or precursors (3–5).

GISTs are relatively rare and account for 1%–2% of cases of gastrointestinal malignancy (6). They were most commonly found in the stomach (55.6%), followed by the small intestine (31.8%), colon and rectum (6%), other locations (5.5%) (7), and esophagus (<1%) in a systemic review of 9,747 GISTs. Surgical resection is the main treatment for localized GISTs. The risk of GIST recurrence was estimated using the modified National Institutes of Health consensus classification system based on the tumor size, mitotic count (8), and primary location (gastric vs. non-gastric). Most low- and intermediate-risk GISTs are cured by surgical resection. In contrast, high-risk GISTs should be treated with adjuvant therapy with 3-year imatinib, which can improve recurrence-free survival (RFS) and overall survival. Neoadjuvant therapy may be an alternative for high-risk GISTs, which can potentially increase the complete resection rate and avoid surgical rupture and resection of the involved organs (9). Palliative treatment, mainly targeted therapy, is the standard treatment for advanced/metastatic GISTs, and classical cytotoxic chemotherapy is inactive for GIST.

c-KIT mutations (2, 10) followed by platelet-derived growth factor receptor alpha (PDGFRA) mutations (11) were discovered and small molecule targeted therapies with tyrosine kinase inhibitors (12, 13) were developed in GIST. Imatinib, sunitinib, regorafenib, and ripretinib are approved in advanced GIST without selection for matched druggable targets. The post-hoc analysis of clinical trials found that KIT exon 9 mutation was associated with a lower response rate and progression-free survival (PFS) compared to exon 11 mutation under treatment with 400 mg/day of imatinib, and 800 mg/day of imatinib benefited patients with KIT exon 9 mutation upon progression despite 400 mg/day of imatinib (14). Although most PDGFRA mutations are sensitive to imatinib, D842V is resistant to imatinib and targeted agents other than avapritinib. Currently, genetic testing of KIT and PDGFRA is essential before treatment initiation (15), and comprehensive genomic profiling, such as for lung cancer, should be considered to identify driver mutations (e.g., SDH, BRAF, NF1, NTRK fusion, and FGFR fusion) for GISTs without KIT and PDGFRA mutations.

KIT and PDGFRA are the main driver mutations but insufficient to promote tumor progression from low- to high-risk GISTs. Additional chromosomal aberrations and subsequent dysregulation of the cell cycle are essential for tumor progression. For example, chromosome 9/9p loss, homozygous CDKN2A (located on chromosome 9p) deletion, p14/p16 loss, activation of MDM2/CDK4 leading to RB1/p53 inactivation, and loss of cell cycle control are considered to be the main events for progression to high-risk GISTs (16).

The nature of GISTs became better understood with the identification of KIT (CD117) expression and c-KIT mutations (2, 10). Heinrich et al. investigated possible alternative receptor tyrosine kinase (RTK) oncoproteins using immunoprecipitations with polyclonal panRTK antisera and found hosphor-PDGFRA was the predominant hosphor-RTK in a KIT wild type GIST cell line (GIST478). Activated PDGFRA mutations were subsequently found in 35% (14 of 40) of KIT-WT GIST  (11). Although most GISTs have either mutation of KIT or PDGFRA kinase genes as driver mutations, approximately 10 to 15% of GISTs do not harbor a KIT or PDGFRA mutation, which are collectively grouped as KIT/PDGFRA-WT GIST.

With the understanding of the molecular biology of GIST and discovery of effective targeted therapy against KIT/PDGFRA pathways, the disease has become more manageable, but the treatments are not completely curative. Several small molecule compounds targeting KIT and other tyrosine kinases, such as imatinib (12, 13), sunitinib (17), regorafenib (18) and ripretinib (19) have shown efficacy in advanced GIST and have been approved for the treatment of advanced GIST. In the SWOG phase III S0033 trial imatinib treatment resulted in a median PFS of 18-20 months (13) and 26% of GIST patients survived 8 years or longer (20). Approximately 10% of GIST patients have primary resistance to imatinib and many of these resistant tumors lack mutations in KIT or PDGFRA, or they harbor a PDGFRA exon 18 D842V mutation. Secondary mutation on KIT (exons 13, 14, 17) developed after imatinib treatment leading to resistance to imatinib. Sunitinib and regorafenib were found to overcome such secondary mutation of KIT (21). Sunitinib treatment after failure of imatinib showed a median time to tumor progression of 27.3 weeks in GIST patients (17). Regorafenib treatment resulted in a median PFS of 4.8 months after failure of imatinib and sunitinib (18). Ripretinib produced a median PFS of 6.3 months in GIST patients after failure of imatinib, sunitinib, and regorafenib (19). All the above inhibitors are widely used in routine clinical practice for patients with advanced GIST and significantly improve the overall survival of GIST patients (22–25). In addition, for patients with symptomatic and/or rapidly progressive disease harboring a PDGFRA exon 18 D842V mutation, avapritinib has been suggested over either imatinib or observation in the setting of initial therapy. These tumors often demonstrate primary resistance to imatinib, whereas the response rate with avapritinib was nearly 90% (26).

In KIT/PDGFRA-WT GISTs, SDH deficiency and other genetic alterations, such as BRAF, NF1, and NTRK fusions, were the main alterations according to comprehensive genetic profiling. Therefore, the proportion of GIST without mutation is decreasing. However, quadruple WT GISTs without mutations or alterations in the KIT/PDGFRA, RAS pathway, or SDH complex are the current treatment challenges because of lack of active treatment for such patients and inactivity of transitional cytotoxic chemotherapy for GISTs (27).

In addition, resistance to targeted therapy is another challenge, as these tyrosine kinase inhibitors (TKIs) can cytostatically inhibitor tumor growth but not induce apoptosis leading to cell killing. Therefore, GIST may be incurable once it develops metastasis.

Targeted therapies, either alone or in combination with immune checkpoint inhibitors have been evaluated in clinical trials (53). Most studies have involved small molecule inhibitors approved in GIST, such as imatinib, sunitinib, regorafenib and/or ripretinib. However, no trials have included drugs targeting the p53 pathway, although a few preclinical studies have indicated the efficacy of such compounds.

As MDM2 expression has been associated with poor prognosis GISTSs (48) and MDM2 inhibitors (33) demonstrated to suppress growth and induced apoptosis in WT p53 GIST cells, the use of MDM2 inhibitors may be an additional strategy beyond KIT and PDGFRA in future studies. However, preclinical studies, particularly in vivo studies, to confirm the feasibility of using MDM2 inhibitors in GIST are lacking. As no clinical trials of p53-directed targeted therapy have been performed or are underway in GISTs, only some clinical studies on other malignancies can be summarized to highlight the applicability of such agents in GISTs in the future.

Although several preclinical studies have shown impressive results, the limited clinical trials of MDM2 inhibitors in other cancers have been disappointing. A phase III study, MIRROS (NCT02545283), failed to meet its primary goal of showing the combination therapy of idasanutlin and cytarabine was superior to placebo plus cytarabine regarding improvement of survival in acute myeloid leukaemia (54). Limited efficacy of HDM201 (siremadlin) in combination with ribociclib (a CDK4/6 inhibitor) was observed in patients with locally advanced or metastatic liposarcoma, even though this tumor highly expresses MDM2 and CDK4 (55). As no agents targeting MDM2 have so far been successful, the therapeutic strategy of targeting the p53 pathway in this way has been doubted. The next challenge is to understand why these studies have shown limited efficacy. It should be clarified whether these targeted agents insufficiently reactivate WT-p53 or if the dose limiting hematologic and gastrointestinal toxicity of p53 reactivation is not tolerated. Furthermore, the efficacy of MDM2 inhibitors may be limited by WIP1 phosphatase expression (56). WIP1 inhibition has been shown to enhance the activity of MDM2 inhibitors (28, 45, 57, 58). Unfortunately, due to the lack of suitable compounds for in vivo evaluation, the studies of WIP1 inhibition have been limited to the preclinical setting, however there is a strong case for the development of WIP1 inhibitors suitable for clinic evaluation.

In contrast, MK-1775 is active against p53 mutated GIST cells both in vitro and in vivo. Besides the regulation of KIT expression, synthetic lethality may be the key mechanism in p53 deficient cells (28, 51, 52). Clinical studies of MK-1775 are ongoing and early phase studies in various tumors have reported encouraging results (59–62).

RITA, a p53 activator, induced p53 in only one p53-WT GIST48B (but not GIST430 or GIST48), followed by growth inhibition and apoptosis. GIST-T1, a p53-mutated cell line, was also highly sensitive to RITA without p53 activation. The aforementioned findings suggested the feasibility of a p53-independent mechanism of action for RITA treatment (24). As no biomarkers can predict the response or resistance to RITA, RITA may not be a good targeted agent in GIST.

Based on the above evidence, CDKN2A/MDM2/p53 pathway alterations frequently occur in high-risk GISTs. The clinical application of this knowledge may be a prognostic factor, or a predictive factor to identify patients who would benefit from adjuvant imatinib treatment (63). Therefore, the genetic testing of these p53 pathway components is required and p53 status would be a critical biomarker for risk stratification beyond the current model based on clinical features (location, tumor size, mitosis). In addition, targeted treatment based on p53 status, either as monotherapy or in combination with KIT/PDGFRA targeted therapy, is proposed for metastatic GISTs (Figure 1). Furthermore, if p53-directed targeted therapy shows promising results in metastatic GISTs, the incorporation of p53-directed targeted therapy may be considered for development in the neoadjuvant setting for high-risk GIST, in combination with KIT/PDGFRA targeted therapy to increase response rate and maximize tumor shrinkage (Figure 1). However, due to currently limited studies as described above, preclinical research is warranted to build a case for the evaluation of such a treatment strategy in clinical trials.

CDKN2A/MDM2/p53 pathway alterations in GIST have been associated with high-risk GIST, adverse clinicopathological characteristics and poor patient survival (Figure 2). Preclinical studies have demonstrated the potential activity of MDM2 inhibitors in p53-WT GIST cells and Wee1 inhibitors in p53-mutated GIST cells. Clinical trials are warranted to determine the role of these novel targeted treatments in GIST.

Conceptualization, C-EW, W-KH, C-NY, and JL. Data curation, C-EW and C-PC. Formal analysis, C-PC and Y-RP. Funding acquisition, C-EW and C-NY. Investigation, EA. Methodology, C-EW, and Y-RP. Project administration, C-EW. Supervision, JL. Validation, C-EW, W-KH, C-NY, and JL. Visualization, C-PC and EA. Writing—Original draft, C-EW. Writing—Review and editing, C-NY and JL. All authors contributed to the article and approved the submitted version.

This research was funded by Linkou Chang Gung Memorial Hospital, grant numbers CMRPG3J0971~3, NMRPD1K1332~3, CMRPG3L0911 to C-EW.

This work was supported by grants from Linkou Chang-Gung Memorial Hospital and the Ministry of Science and Technology in Taiwan. Please refer to funding section.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.