Targets, targeted therapy reports abound at AARC

Many speakers mention Gleevec as top example of progress vs. cancer

By David Josephy

I attended the 94th annual meeting of the American Association for Cancer Research held July 11-14 in Washington, D. C. This meeting had been set for April in Toronto, Ontario, but was rescheduled due to the outbreak of severe acute respiratory syndrome (SARS).

It’s the major annual basic science cancer meeting, with more than 10,000 registrants. Compressed into four days, one could only attend a fraction of the many parallel presentations. Lots of papers and posters were relevant to GIST and Gleevec, and I tried to get to as many as I could.

Here is a brief report of some of the highlights, based on the abstracts (available at www.aacr.org) and my rough notes. A warning: I have likely introduced a few errors as I rushed to write down the presentations, many of which were given at a very fast pace.

Friday morning I attended some talks in an “educational session.” Cochairman of the session was to have been Dr. Jonathan A. Fletcher of Dana-Farber Cancer Institute in Boston, a pathologist and leading expert on GIST, but he was unable to attend and Dr. Michael Heinrich from Oregon Health & Science University in Portland took his place. Heinrich gave a talk focused on the molecular analysis of GIST, titled “PDGFRA and KIT gain-of-function mutations are alternative oncogenic mechanisms in GISTs.”

He reviewed the role of c-kit in GIST and the importance of exon 11 mutations. He mentioned a transgenic mouse “knock-in” animal model in which an exon 11 mutation leads to GIST. Then he focused on the subset of GISTs (about 18 percent of patients) in which KIT mutations are NOT detected. He noted the recent study (Gilliland et al., published in Science) that identified PDGFRA mutations in many of these cases.

PDGFRA sequencing analysis was performed on tumor material from 40 c-kit-negative GISTs. He gave the following statistics for PDGFRA mutations: Point mutation D842V, in the kinase activation loop (exon 18), 9 cases; deletions in exon 18, 3 cases; insertions or deletions in the juxtamembrane domain (exon 12): 3 cases; other or not identified, 25 cases. He noted that about 10 percent of GISTS have no identified mutation in either c-kit or PDGFRA, presumably indicating that additional targets remain to be discovered. Some of the PDGFRA mutant proteins are Gleevec-sensitive, but the D842V mutation is relatively resistant, with an inhibitory constant of 2 micromolar

He reviewed the pattern of c-kit mutations in a spectrum of diseases. Exon 8 mutations are associated with leukemia; exon 9 and 11 mutations with GIST, and exon 17 mutations with mast cell disorders (mastocytosis).

Finally, he discussed the recent (published this year) finding that Gleevec is effective in the extremely rare HES (hypereosinophilic syndrome). In this case, a FIPIL1-PDGFRA fusion protein is responsible, and this protein is exquisitely sensitive to inhibition by Gleevec (inhibitory constant of 5 nanomolar.), explaining why Gleevec is effective at almost “homeopathically” small doses.

The Friday afternoon plenary session at AACR was built around a perspective of the impact of the 1953 discovery of the structure of DNA. Attendees received a foldout poster showing a time-line of 50 years of cancer research since that discovery, and the introduction of Gleevec was one of the most recent milestones on the chart. Many speakers mentioned Gleevec as a prime example of progress in the war on cancer.

On Saturday, Dr. Brian Druker of Oregon Health & Science University presented a 7 a.m. “Meet-the Expert Sunrise Session,” titled “Imatinib as a Paradigm of Targeted Therapies.” His focus was on CML but he also discussed GIST. Most of his talk was a review of progress familiar to most LRG members.

After presenting the overall good news of the success of Gleevec therapy for CML, he raised several questions: 1) Is it possible to improve on the present success with treatment of chronic CML patients? 2) Why do some patients relapse? 3) Is the target (Bcr-Abl kinase in CML) actually inhibited? He showed that about half of CML patients who become Gleevecresistant have acquired new mutations in the kinase domain of Abl, with the Y351I (tyrosine to isoleucine) mutation being one of the most troublesome, because the mutant protein is insensitive to Gleevec and also to the investigational drug PD180970. With respect to GIST, he estimated that there are 5,000 new cases of GIST per year in the U.S. — far more than was thought a few years ago. He cited Dr. George Demetri’s work and the recent study of Gilliland et al. (see above).

After his talk, I asked a question about dosing strategies — does he feel that doses should be pushed higher, or dropped lower, in patients who appear to be doing well on Gleevec. He was certainly well aware of the surrounding issues, and mentioned the controversy regarding the comparison of results between the (nominal) 400 and 800 mg. dose regimens for GIST, which has been discussed in these pages, and he agreed that analysis of these results is difficult, in view of the differences between the intent-to-treat doses and the actual delivered doses.

Dr. Andrey Frolov, working with Fletcher and others at Fox Chase and Dana-Farber, gave a talk titled “Gleevec and GISTs: Identification of response markers and the molecular mechanisms of action.” The goal of his work is to identify “surrogate” markers, that is, proteins other than c-kit which might be used to measure a patient’s responsiveness to Gleevec. He compared gene expression in human GIST cell line (GIST882) with and without exposure to Gleevec, looking for proteins whose expression was changed by exposure to the drug. Two proteins were discovered to be strongly down-regulated in response to Gleevec: a signaling molecules called SPRY4A (“sprouty”) and another protein called MAFbx.

Finally, needle biopsies of GIST patients before and after Gleevec treatment were analyzed for expression of these proteins. He concluded that these proteins are highly-reliable predictors of response, because they became down-regulated only in those patients who, clinically, responded well to Gleevec.

Sunday morning, Dr. Sharyn D. Baker gave a sunrise talk, “Pharmacokinetic Variability of Anticancer Agents.” Gleevec was among the many drugs she covered. Her most important message was that there is still no substitute for clinical and analytical observation of the response of the individual cancer patient to a drug. For most studies, there are no significant correlations between response and variables such as body mass, body surface area, age, or polymorphisms of specific enzymes such as P450 3A4. Inter-individual variation in drug metabolism is multi-factorial and usually too complex to be reduced to any single predictive variable. This explains why most doctors believe that it is not useful to adjust Gleevec doses based on factors such as body mass.

Another prominent theme at the AACR meeting was the large number of researchers who are trying to develop new drugs that act by mechanisms similar to that of Gleevec; that is, inhibiting c-kit and other tyrosine kinases. The real number of such ongoing studies must be even larger, because much of the work will be proprietary projects at drug companies, but several academic researchers reported their work at the meeting.

I won’t go into too much detail about these studies, because the new compounds are generally at a very early stage of development. An example was a talk by Daruka Mahadevan (Arizona Cancer Center, Tucson) titled “Structure-based design and synthesis of c-kit tyrosine kinase inhibitors targeting GIST.” He showed a series of new compounds that inhibit c-kit at concentrations similar to that at which Gleevec acts, and can kill GIST882 cells in vitro.

Novartis researchers displayed a poster titled “Antitumor activity of RAD001 in combination with cytotoxic agents” which, although not focused on GIST, may be of interest, since there are ongoing trials of RAD001 for GIST therapy.

I had a chat with the lead author, Dr. Terence O’Reilly, who turned out to be a graduate of the University of Guelph! RAD001 is a derivative of rapamycin. They tested combination therapy of RAD001 with conventional chemotherapy agents such as doxorubicin, cis-platinum and taxol, in tissue culture studies, and found some potentially promising regimens, in terms of synergistic cell killing.

Another poster I looked at was “Plasma and cerebrospinal fluid pharmacokinetics of imatinib in nonhuman primates,” presented by Dr. Merrill J. Egorin of the University of Pittsburgh Cancer Institute. The poster reported studies in monkeys looking at Gleevec’s ability to pass through the blood-brain barrier. Gleevec was detected in the brain after oral dosing, but the levels were much lower than in the peripheral blood. Thus, the brain may be a “sanctuary” site, explaining why CML can recur in the brain after Gleevec treatment. (However, I believe that GIST metastasis to the brain is very rare.)

Egorin told me that his lab ran the analyses of Gleevec serum levels for the NCI clinical trial. He was friendly and approachable and very wellinformed about Gleevec pharmacology. While we were talking, another person stopped to look at the poster, Dr. Anthony Murgo from National Cancer Institute, and we chatted for some time. He is one of the key managers for the NCI Gleevec trial.

A highlight of the meeting was one of the last sessions, a mini-symposium on epidermal growth factor receptor inhibitors Monday afternoon. There were three talks in a row about GIST therapy.

First, Dr. Samuel DePrimo of Sugen Inc., presented “Decreases in circulating levels of soluble KIT in patients with imatinib-resistant GIST receiving the novel kinase inhibitor SU11248: Correlative analysis of blood and plasma biomarkers.” This paper reported a study of a possible biomarker of response to treatment in GIST, carried out in the patients enrolled in the phase I trial of Sugen (SU11248) in Gleevec-refractory, -resistant, or -intolerant GIST patients. The goal is to find markers in the blood which could be used to predict or evaluate response to therapy in GIST patients. C-kit is a membrane-bound protein found in the tumor cells but not circulating in the blood. “S-kit” is a soluble fragment of kit that corresponds to the extra-cellular domain of the protein. It is generated by cleavage of membrane-bound c-kit, carried out by an unidentified protease enzyme. The researchers measured s-kit in the blood of patients in the SUGEN trial. S-kit levels decreased during treatment, and there was a good correlation between the s-kit level and response to therapy. So this may be a promising and rapid new way to assess patient response to treatment.

Dr. Anette Duensing, in collaboration with Drs. Fletcher, Demetri, et al., gave a talk “Oncogenic KIT signaling in GISTs.” They used biochemical techniques to look for proteins involved in c-kit-dependent signaling in GIST patients and in GIST cell lines. This was a rather technical talk and I did not discern any direct implications for GIST therapy.

Dr. George Demetri gave the next talk, “Biological activity of the multitargeted tyrosine kinase inhibitor SU11248 in patients with malignant GIST refractory to imatinib mesylate.” Among the points he made: He reviewed results with Gleevec for GIST, stating that 63 percent of patients show a partial response, 20 percent are stable, and 17 percent fail to respond. However, resistance emerges with time among the responding patients.

So far, median survival time of the metastatic GIST patients in the U.S./ Finland trial is 1.4 years. SUGEN (SU11248 ) is being tested in a phase I dose-escalating study in metastatic GIST patients “with disease that was objectively progressing despite continuation of imatinib therapy, plus one patient who could not tolerate imatinib.”

The study has developed into a phase II trial. Safety and tolerability of the drug is being studied, as well as many correlative science studies (one of which is the DePrimo et al. study mentioned previously).

The SUGEN starting dose is 25, 50, or 75 mg. a day for 14 days, followed by a 14-day rest period per cycle. Patients undergo CT, MRI and PET scans, biopsies and blood work. Forty-five patients were enrolled from April 2002 to March 2003, 40 of whom had developed resistance to Gleevec after an initial response. He showed evidence that kit phosphorylation (activation) and tumor cell proliferation drop dramatically following Sugen treatment. Some 72 percent of patients showed a response, as measured by PET scans. He emphasized one patient who is “doing great” after one year on Sugen.

The overall conclusion is that Sugen shows promising activity. He also showed cases where a dramatic response was evident by PET scans and by evaluation of diminishing proliferative index in biopsy cells, even though there was no clear evidence of response on CAT scans.

A clear lesson was that CAT scans are a poor indicator of response to the Sugen drug. A phase III trial is “coming.”

During questions, the dose schedule was discussed. The two-week-on/twoweek- off regimen was used because previous trials of Sugen for other cancers had shown that patients had trouble tolerating the drug, with a fatigue syndrome developing. The two-week rest period allows patients to recover. However, Demetri said, GIST patients are tolerating the drug better than had been anticipated, and the study may switch to continuous drug dosing.

This was a very encouraging presentation. Thanks to Laura Cooley for letting us know about a report, stemming from Demetri’s talk, available on MSN (http://content.health.msn.com/ content/article/71/81203.htm).

David Josephy, Ph.D., is a professor of chemistry and biochemistry at the University of Guelph, Ontario, Canada.