This article explores the correlation of target kinase genotype with clinical activity of imatinib mesylate (IM) in patients with metastatic GI stromal tumors (GISTs) expressing KIT (KIT+). GIST patients with a particular genetic mutation are more likely to respond to Gleevec than those without the mutation, reported Dr. Michael C. Heinrich at one of the plenary sessions of the annual conference of the American Society of Clinical Oncology, held May 14-17 in Orlando, Florida.

The objective of the study he reported on (CALGB 150105/SWOG S0033) was to correlate molecular abnormalities in pre-treatment GIST specimens with clinical response to Gleevec in patients treated in the S0033 phase III NCI trial. Some 344 patients had sufficient tumor samples to be included in this study. This is about 46 percent of the total number of patients in the phase III S0033 trial.The results confirm previous observations and provide a foundation for molecular testing that can predict who will best respond to treatment with Gleevec.

A professor of medicine at Oregon Health & Science University in Portland, Heinrich’s presentation was titled “Correlation of target kinase genotype with clinical activity of imatinib mesylate (IM) in patients with metastatic GI stromal tumors (GISTs) expressing KIT (KIT+).”

The pathology of these 344 patients was reviewed by Dr. Christopher
Fletcher of the Brigham & Women’s Hospital. Of the 344 patients, 332 of them were confirmed to have GIST, while 12 patients (3.5 percent) were reclassified as non-GIST sarcomas.

Of the 332 GISTs analyzed, 324 (97.5 percent) were KIT-positive GISTs, and eight (2.5 percent) were KIT-negative GISTs. Heinrich also noted that 98 percent of KIT exon 9 mutations have a small intestine origin and 98 percent of PDFRA mutations have a stomach origin. The study data showed that both KIT-positive and KIT-negative GISTs have either KIT or PDGFRA mutations in about 87 percent of cases; however, the ratio of KIT to PDGFRA mutations was different. KIT-positive GISTs had KIT mutations 86 percent of the time and PDGFRA mutations in only 1.5 percent of cases. KIT-negative GISTs had KIT mutations in 50 percent of the cases, and PDGFRA mutations in 37.5 percent of the cases.

Kinase genotyping (checking for mutations in the various exons of c-kit and PDGFRA), was done by Heinrich and Dr. Christopher Corless, also of OHSU in Portland. Using gene expression profiling, three different groups (Subramanian et at, Antonescu et al, and Kang et al) have reported different molecular signatures for KIT exon 9, exon 11, and wildtype GISTs (no KIT or PDGFRA mutations)

The study looked at c-kit positive patients with KIT exon 11 and exon 9 mutations and patients with no mutations to determine the likelihood of an objective response when comparing low-dose (400 mg.) to high-dose (800 mg.) Gleevec. Note that “objective response” measures tumor shrinkage using RECIST criteria, and does not necessarily equate to how long the therapy will be effective (time to progression).

The study found that there was no effect on dose in patients with an exon 11 mutation. The “odds-ratio” of high-dose vs. low-dose in these patients was 1.0 (p=0.96). This means that low-dose patients had the same odds (1.0) of an objective response as high-dose patients. Yet the study also found that the 25 patients with exon 9 mutations were much more likely to have an objective response on 800 mg Gleevec than those on 400 mg, with an odds ratio of 8.0 — that is, eight times the chance of having a response.

“However,” Heinrich noted, “when adjusting for multiple comparisons, and considering the small sample size, this difference was not statistically significant.” Patients with no mutation (n=33) had an odds ratio of 1.5, but this was not statistically significant (p=0.62).

Heinrich presented data showing that patients with exon 11 mutations had a much longer median time to treatment failure (576 days) compared to patients with exon 9 mutations (308 days) or patients with no mutation (251 days).

The study concluded that there is no effect of Gleevec dose on objective response of KIT exon 11 mutant GIST. However, further study is needed to determine if KIT exon 9 mutant GISTs respond better to high-dose imatinib. Also, the study concluded that kinase genotyping is useful for:

— Confirming the diagnosis of GIST, particularly for c-kit negative GIST;
— Predicting response and duration of response to Gleevec;
— The design and interpretation of clinical trials to study new treatments for GIST.