GIST studies receive substantial interest at ASCO '06 conference

By Jerry Call

The American Society of Clinical Oncology (ASCO) website has a large body of information about GIST. This year over thirty reports were added to their GIST database. For a comprehensive look at all of this information, visit the abstracts section of the ASCO website (www.asco.org). All of the presentations, videos and slides from ASCO are available for general viewing as of mid- August.

The following are a few noteworthy studies that were discussed at ASCO.

AMN107

A report of early results of the phase I trial of AMN107 (nilotinib) in combination with Gleevec (imatinib) was presented in a poster by Dr. Peter Reichardt and colleagues. As of February 1, 2006, 37 patients had enrolled in the trial. Follow- up time is short in this trial so the results should be interpreted with caution; 27 patients (73%), achieved some period of stable disease (14 on AMN107 only and 13 on AMN107 plus Gleevec).

Side effects were similar to Gleevec, but elevated liver enzymes were somewhat more common. At the highest doses of Gleevec and AMN107, skin toxicity was the dose-limiting toxicity (side effects during treatment were severe enough to prevent further increase in dosage or strength of treatment). A dose of 800 mg of AMN107 and 800 mg of Gleevec produced excessive skin toxicity; 400 mg of AMN107 and 800 mg of Gleevec was well tolerated according to the poster.

Measurements of GIST treatment response

1. The inadequacy of RECIST (Response Evaluation Criteria in Solid Tumors), the current standard of response measurement, has been known for several years, especially when measuring GIST responses. RECIST measures the efficacy of a treatment by assessing the change in the longest dimension of a tumor(s) size. A patient with greater than a 30 percent reduction in the sum of all target tumors is said to have a “partial response.”

Studying a different form of measurement criteria, Dr. Robert Benjamin from M.D. Anderson Cancer Center gave a presentation supporting the use of response criteria developed at M.D. Anderson, the Choi criteria (named for Dr. Haesun Choi). The Choi criteria classifies responders as patients that achieve either a 10 percent or greater reduction in tumor volume OR a 15 percent or greater reduction in tumor density. These criteria are highly correlated with response by both PET scans and patient history (follow-up). Dr. Benjamin then went on to make a case that RECIST is not only inadequate for GIST, it is inadequate for all solid tumors.

2. Dr. Alex Le Cesne from the Institute Gustave-Roussy in France gave another presentation on measuring response in GIST. Dr. Le Cesne concluded that RECIST could correctly identify nonresponding patients, but not patients likely to respond. The study by Le Cesne and his EORTC (European Organization for Research and Treatment of Cancer) colleagues used the change in tumor size to predict response, but unlike the M.D. Anderson study, they did not use changes in tumor density.

The Le Cesne study used change in tumor size at 2 months, 4 months and 6 months to predict response. In agreement with the M.D. Anderson study, they found that a reduction in size of greater than 10 percent at any time point (2, 4, or 6 months) predicted patients that were likely to receive the greatest benefit. They classified these patients as responders.

Patients with less than a 10 percent reduction and no more than a 20 percent increase were defined as patients with an intermediate sensitivity to treatment only at the 2-month and 4-month treatment points. Patients falling into this size reduction category (less than 10 percent reduction and no more than a 20 percent increase) at the 6-month treatment point, did just as well as those in the “responder” category. Patients having greater than a 20 percent increase in tumor size had poor responses to treatment. This is the only category correctly predicted by the current standard, RECIST.

3. When examining PET scans, Dr. Chandrajit P. Raut and colleagues at Dana-Farber Cancer Institute reported on the patterns of PET response after long-term treatment with Sutent. PET “rebound” (which is higher PET activity) has previously been reported to occur in the two week “off-period” of Sutent treatment. Dana-Farber followed 4 patients for over 2 years on Sutent. Initially all 4 had good PET responses with PET rebound during the off treatment cycle. After 6 months of treatment, all 4 patients had complete suppression of PET activity during both the on treatment cycle and the off treatment cycle. After more than 2 years of treatment, some GIST lesions in two of the patients demonstrated PET rebound during the off period but suppression during the on period. The other two patients did not experience PET rebound, even during the off treatment period. None of the four patients had progressive disease.

Familial GIST

Dr. Eric Kleinbaum and colleagues from M.D. Anderson presented a poster about a family with familial GIST. This family had 15 members with confirmed or suspected GIST. A germ-line mutation, a deletion of codon 579 in exon 11 of the c-kit gene, was found in this family. This makes at least two familial GIST families with this mutation (out of the 10 to 15 known familial GIST families). A deletion of codon 579 is fairly rare in the non-familial GIST population, occurring in about 2 percent of analyzed patients in the phase II and EORTC phase III trials.

Creatine Kinase Increase

Dr. Paolo Allione gave a presentation about the increase in creatine kinase and its correlation with musculoskeletal complaints in GIST patients taking imatinib. These complaints are common in GIST patients affecting 25 to 50 percent of patients. Calcium and magnesium supplements have occasionally provided some relief. Quinine is an alternative option, but side effects can limit its use according to Allione.

An increase in creatine kinase value occurs during muscle damage. Even a single muscle cramp can cause an increase in creatine kinase (CK) values. Allione and colleagues found that a great majority of patients experiencing musculoskeletal complaints had elevated CK levels. They suggested consideration of testing CK levels into standard clinical chemistry workup; this allows clinicians to add an objective measure of musculoskeletal complaints. The authors note that they had to stop imatinib for one week only in one patient (out of 40 treated) for uncontrolled musculoskeletal complaints.

GIST reGISTry

Dr. Jonathan Trent and colleagues presented a poster about the GIST reGISTry. This database is designed to provide information on epidemiology, patterns of diagnosis and management of GIST. Since January 2005, 228 patients have been enrolled.

In summary, Trent reported:

• Most of the patients in this study presented localized disease (78.5%).
• Surgery was first-line treatment in 84 percent of these cases and in 55 percent of patients with metastatic disease.
• Almost 80 percent of patients had a surgical biopsy, 14 percent had an endoscopic biopsy and 10 percent had a biopsy done by interventional radiology.
• CT scans were the most popular type of imaging (60%), followed by endoscopy (19%), PET scans (3%) and CT/ PET (2%).
• Patients were twice as likely (66% vs. 34%) to be treated by a communitybased practice than at a university or academic center.
• While 92.5 percent of patients had testing for c-kit (CD117), only 3.5 percent had genotype testing (mutational analysis).
• Gleevec treatment doses were:
□ 400 mg/day = 71.8%
□ 600 mg/day = 15.5%
□ 800 mg/day = 6.8%
□ Other = 5.8%