This is a non-randomized, open-label, multicenter phase 2 study to evaluate the efficacy and safety of ponatinib in patients with metastatic and/or unresectable GIST after prior failure or intolerability of imatinib. Patients will be enrolled into 1 of 2 cohorts based on absence (Cohort A) or presence (Cohort B) of KIT exon 13 resistance mutations as measured by liquid biopsy. A third cohort (Cohort C) will include patients who have received all approved lines of TKI treatments (imatinib, sunitinib and regorafenib).

Primary Objective

• To assess clinical benefit of ponatinib in patients with KIT or PDGFRA mutant GIST defined as clinical benefit rate (CBR), which is the composite of complete response (CR), partial response (PR) and stable disease (SD) at ≥16 weeks after start of treatment per modified response evaluation criteria in solid tumors (modified RECIST 1.1 [Demetri et al., 2013]) as a measure of disease control
• Two cohorts for second-line patients will be used: Cohort A: patients with secondary resistance mutations in other exons or no resistance mutations (as measured by liquid biopsy in circulating DNA); Cohort B: patients with evidence of secondary resistance mutations in exon 13 as assessed on progressing lesions or in circulating DNA
• One additional Cohort (Cohort C) will include heavily pretreated patients (failure of at least all approved lines of therapy) regardless of secondary mutation Secondary Objectives
• To assess progression-free survival (PFS) in each cohort and in the total patient population
• To assess objective response rate (ORR) in each cohort and in the total patient population
• To assess overall survival (OS) in each cohort and in the total patient population
• To evaluate the safety and tolerability of ponatinib in the total patient population
• To assess Quality of Life (QoL) Exploratory Objectives
• To assess limited elements of pharmacokinetics (PK) in the total patient population
• To explore the relationship between GIST genotype and CBR with ponatinib
• To explore the feasibility of detecting mutations in KIT and possibly other cancer-related genes using circulating nucleic acids derived from blood samples
• To explore the usefulness of "liquid biopsies" to predict treatment response and development of resistance
• To assess duration of follow-up treatment

Primary Endpoint

• CBR consisting of CR+PR+SD by modified RECIST 1.1 (Demetri et al., 2013) at 16 weeks in patients with imatinib-resistant GIST (KIT-mutant) with other or no resistance mutations (Cohort A) and secondary resistance mutation in exon 13 (Cohort B) Secondary Endpoints
• PFS in each cohort and in the total patient population
• ORR (CR + PR) in each cohort and in the total patient population
• OS in each cohort and in the total patient population
• Safety and tolerability of ponatinib
• QoL
• CBR of Cohort C Exploratory Endpoints
• Correlation of plasma levels of ponatinib and response
• Molecular genetic features of GIST at baseline and after treatment with ponatinib
• Correlation of tumor DNA from available paraffin tissue with genotypes of plasma sequencing ("liquid biopsies") and correlation of plasma genotype with treatment response, resistance and duration of follow-up treatment

Inclusion Criteria:

• Male or female patients ≥18 years old
• GIST with failure or intolerance to imatinib or failure / intolerance to all three approved TKIs defined as:
• Histologically confirmed metastatic and/or unresectable GIST (harboring a primary KIT or PDGFRA-mutation) after failure or intolerance of imatinib (cohort A and B) or all three approved TKIs (cohort C). If prior TKI treatment was neoadjuvant therapy, then relapse must have occurred during the neoadjuvant therapy in order to consider it failed therapy
• Patients in Cohort A must have evidence of clinical resistance mutations in any other exon or no resistance mutation but evidence of progression by CT or MRI imaging. Patients in Cohort B must have evidence of an activating resistance mutation in KIT exon 13 (by direct sequencing of progressing lesions or by liquid biopsy).
• Measurable disease per modified RECIST 1.1 (Demetri et al., 2013). A lesion in a previously irradiated area is eligible to be considered as measurable disease as long as there is objective evidence of progression of the lesion prior to study enrollment
• Eastern Cooperative Oncology Group (ECOG) performance status of 0, 1, or 2
• Adequate hepatic function as defined by the following criteria:
• Total serum bilirubin ≤1.5 x upper limit of normal (ULN), unless due to Gilbert's syndrome
• ALT (Alanine Aminotransferase) ≤2.5×ULN or ≤5.0xULN if liver metastases are present
• AST (Aspartate Aminotransferase) ≤2.5×ULN or ≤5.0xULN if liver metastases are present
• Adequate renal function as defined by the following criterion:
• Serum creatinine <1.5×ULN
• Adequate pancreatic function as defined by the following criterion:
• Serum lipase and amylase ≤1.5×ULN
• For patients of childbearing potential, a negative pregnancy test must be documented prior to enrollment
• Female and male patients who are fertile must agree to use an effective form of contraception with their sexual partners from signing of the informed consent form for this study through 4 months after the End-of-Treatment Provision of written informed consent
• Willingness and ability to comply with scheduled visits and study procedures
• Fully recovered (≤ Grade 1 or returned to baseline or deemed irreversible) from the acute effects of prior cancer therapy before initiation of the study drug treatment

Exclusion Criteria:

1. Patients lacking primary mutations of KIT or PDGFRA (including Succinate-Dehydrogenase(SDH)-deficient GIST)
2. Major surgery within 28 days prior to initiating therapy
3. History of bleeding disorder
4. History of acute pancreatitis within 1 year of study or history of chronic pancreatitis
5. History of alcohol and /or drug abuse
6. Uncontrolled hypertriglyceridemia (triglycerides >450 mg/dL)
7. Clinically significant, uncontrolled, or active cardiovascular disease, or other arterial or venous vascular occlusion diseases specifically including, but not restricted to: Myocardial infarction within 6 months prior to enrollment Unstable angina within 6 months prior to enrollment Congestive heart failure within 6 months prior to enrollment, or left ventricular ejection fraction (LVEF) less than lower limit of normal per local institutional standards History of clinically significant (as determined by the treating physician) atrial arrhythmia Any history of ventricular arrhythmia Cerebrovascular accident or transient ischemic attack within 6 months prior to enrollment Any history of peripheral arterial occlusive disease requiring revascularization Venous thromboembolism including deep venous thrombosis or pulmonary embolism within 6 months prior to enrollment
8. Uncontrolled hypertension (diastolic blood pressure >90 mm Hg; systolic >140 mm Hg). Patients with hypertension should be under treatment on study entry to effect blood pressure control
9. Taking medications that are known to be associated with Torsades de Pointes (Appendix A)
10. Taking any medications or herbal supplements that are known to be strong inhibitors of CYP3A4 within at least 14 days before the first dose of ponatinib (Appendix B)
11. Ongoing or active infection. This includes but is not limited to the requirement for intravenous antibiotics
12. Known history of human immunodeficiency virus. Testing is not required in the absence of prior documentation or known history
13. Pregnant or breastfeeding
14. Malabsorption syndrome or other gastrointestinal illness that could affect oral absorption of the study drug
15. Individuals with a history of a different malignancy, other than cervical cancer in situ, basal cell or squamous cell carcinoma of the skin, are ineligible, except if they have been disease-free for at least 5 years, and are deemed by the investigator to be at low risk for recurrence of that malignancy OR if the other primary malignancy is neither currently clinically significant nor requiring active intervention.
16. Use of any approved TKIs or investigational agents within 2 weeks or 6 half-lives of the agent, whichever is longer, prior to receiving study drug
17. Any condition or illness that, in the opinion of the investigator, would compromise patient safety or interfere with the evaluation of the drug
18. History of apoplectic insult