GIST: Second-Line Treatment Options


Margaret von Mehren, MD: The development of drugs for GIST [gastrointestinal stromal tumor] really happened in a very sequential fashion. First, we had imatinib. Then, we had Sutent [sunitinib]. Then, we had regorafenib. And much more recently, we’ve had avapritinib and ripretinib. We often think of them in a sequential fashion, because that’s how they were approved. But there are times when you are making a treatment decision for second-line therapy and you might want to use Sutent rather than regorafenib, or vice versa. Some of that has to do with information about mutations. We know that the majority of patients who progress on imatinib have exon 11 mutations. That group of mutations is the most common, so the greatest number of patients are in that scenario.

These patients, if you look at the disease that’s progressing, will still have that primary exon 11 mutation but usually have additional mutations. Oftentimes, there are additional KIT mutations. They tend to be grouped into 2 areas—exon 13 and exon 14, and exon 17 and exon 18. These 2 different locations are different parts of the kinases, and, interestingly enough, the drugs work better for different types of mutations.

For the patients who have the exon 13 and exon 14 mutations, Sutent is an excellent drug. It tends to control this class of disease. Regorafenib treats some of the exon 17 and exon 18 cases, but not all of them. So that’s probably a better drug for that scenario. I think one of the challenges, however, in making decisions with this type of molecular testing is that oftentimes if we’re taking a biopsy of the tumor, most people have more than 1 site of metastatic disease. Also, we know from patients who have had metastatic lesions removed that if you sample different parts of 1 tumor nodule we may find different mutations. One biopsy may give you 1 answer about what 1 mutation is, but that’s probably not the whole story.

Using blood to look at mutations based on circulating tumor DNA may give us a better sense of the whole array of mutations. This is evolving, in terms of the testing—really understanding what the testing tells us. Certainly, if the mutation is there, it reflects some disease burden that is present. We know from studies that have incorporated this that oftentimes we can see a decrease in specific types of mutations within the blood. But I think as a standard of care in practice, it’s something that is evolving and is still not part of the daily routine in how we take care of these patients.

We have based our decision-making much more on CT scans or MRI scans. I didn’t talk about PET [positron emission tomography] scans, but they can be helpful when you’re looking at a patient who you think may be progressing but you’re not sure. Is this maybe a tumor nodule that may have had some bleeding into it? Is this a tumor that really is having a therapeutic response, but, for whatever reason, the nodule has increased somewhat in size? Using a PET scan may be helpful because it tells you about changes in metabolic response. The other criteria or radiographic criteria that people have used is something called the Choi criteria, which looks at Hounsfield units, or the density of tumor lesions in a contract-enhanced CT scan. If the density of a lesion is decreasing, that typically is consistent with benefit, unless the tumor size is increasing significantly. So if you’re not quite sure about whether a patient is responding, a PET scan can be a helpful adjunct to better understand what the scan is trying to tell you.

Transcript edited for clarity.

Case: A 68-Year-Old Man With Gastrointestinal Stromal Tumor

Initial presentation

  • A 68-year-old man complains of a 5-month history early satiety and vague abdominal pain
  • PMH: hypertension, medically controlled; colonoscopy at age 55 unremarkable; no family history of cancer
  • PE: diffuse abdominal pain on palpation; otherwise unremarkable

Clinical workup

  • Labs: Hb 10.1 g/dL, plt 100 x 109/L; other lab values WNL
  • Endoscopy: showeda submucosal ~ 5 cm mass with ulceration
  • EUS-FNA biopsy: irregular borders on extraluminal surfaces of the stomach with evidence of heterogeneous echogenicity
  • Biopsy showed gastrointestinal tumor with mitotic activity showed >5 mitoses/50 HPFs
  • Abdominal/pelvic CT confirmed a 5.4 cm lesion in the body of the stomach
  • MRI showed evidence of peritoneal metastases
  • IHC and genetic mutational analysis: KIT exon 11 mutation
  • ECOG 0


  • He was started on imatinib 400 mg PO qDay, continued for 26 months until he complained of increased abdominal pain and decreased appetite
    • ECOG 1; Imatinib was discontinued due to progressive disease
  • Treatment with sunitinib 50 mg PO qDay for 4 weeks, with a 2-week drug-free interval was started
    • Treatment was well tolerated for 9 cycles when he developed an altered sense of taste, occasional vomiting and diarrhea; sunitinib was discontinued
  • Regorafenib 160 mg (four 40 mg tablets) PO qDay for the first 21 days of each 28-day cycle was started and tolerated except for hand-foot syndrome
    • Regorafenib was discontinued on progression of disease
  • The patient was started on ripretinib 150 mg PO qDay
Related Videos
Related Content