One of the highlights of the American Society of Clinical Oncology (ASCO) conference was the session “Gastrointestinal Stromal Tumors: Multimodal Approach.” George Demetri M.D. began the discussion with a review of the history and biology of GIST. Dr. Demetri reviewed the current status of GIST and brought up some new challenges ahead.

The following are some questions addressed in the GIST session:

  • What is the role of surgery in managing GIST patients with kinase inhibitors?
  • Is there a role for kinase inhibitor therapy following maximal response?
  • Are there genetic predictive factors which can identify GIST patients most likely to benefit from a given kinase inhibitor?

The following four presentations took place during the “Gastrointestinal Stromal Tumors: Multimodal Approach” session.

1. “Indication and results of surgery following imatinib treatment of locally advanced or metastatic GIST.”Hohenberger et al.

Dr. Peter Hohenberger, a surgeon from the University of Mannheim, Germany, gave this presentation titled “Indication and results of surgery following imatinib treatment of locally advanced or metastatic GIST.” Hohenberger noted that while imatinib (Gleevec) is effective to control advanced GIST, patients eventually develop progression. Imatinib provides an opportunity to resect tumors not amenable to resection earlier.

Dr. Hohenberger presented the results of a surgery survey that covered 113 GIST patients. The indications for doing the resections were grouped as:

  1. To convert a partial response (PR) into no evidence of disease (NED) number of patients (n) = 44
  2. True neoadjuvant n=14
  3. Focal progression n=13
  4. Progressive disease n=42
  5. Emergency n=11

Operation mortality rates were zero percent for patients with a partial response or focal progression. They were 1.7 percent for patients with progressive disease while in the hospital, but 7.7 percent within 30 days.

Surgery to remove all of the tumors of interest with clear margins was successful in the majority of cases in all of the groups (72% to 86%) except progressive disease (26%).

With a median follow-up of 15.8 months, 14 patients with complete tumor removal from groups 1, 2 and 3 stopped Gleevec. Of these patients, 8 of 14 developed a recurrence.

The median progression-free survivals of the groups were the following:

  1. PR to NED = 11 months, but 16 months with clear margins (R0)
  2. Neoadjuvant = median not reached, 72 percent disease-free at one year
  3. Focal progression = 8 months
  4. Generalized progression = 3 months
  5. No data was given for the emergency surgery group

Summary:

  • R0 resection (clean margins) of PR residual GIST after imatinib yields over one year of progression-free survival (PFS).
  • Imatinib should be continued after surgery.
  • This series represents a selected group of patients; it is not a prospective trial and may have bias.
  • Single focus progressive disease does not necessarily indicate wide spread dissemination.
  • If tumors are resectable, do molecular pathology and exon analysis. As Hohenberger states, “…you will get more information about this tumor and you might get some information about which other drug might be effective to control this tumor.”
  • Patients with multifocal progression (generalized progression) cannot be cured by surgery.

Conclusions from the study:

  • Patients with inoperable or locally advanced GIST developing a PR under imatinib should be evaluated for resection as early as possible.
  • Resection of tumor remnants is a safe procedure with a chance for contributing to long-term cure.
  • Randomized trial of early resection vs. delayed resection should be feasible.
  • Timing of surgical intervention is crucial.

2. “The Role of Surgery in Multimodality Therapy for Advanced GIST.” Raut et al.

Expounding on the topic of surgery, Dr. Chandrajit Raut M.D. of Dana-Farber Cancer Institute was the discussant that presented on “The Role of Surgery in Multimodality Therapy for Advanced GIST.”

Dr. Raut noted that the advent of targeted therapy using kinase inhibitors (like Gleevec) has altered the natural history of advanced disease, but that pathologic complete response is rare. He also noted that response to kinase directed therapy is not maintained indefinitely (resistance eventually develops) and that surgery of previously unresectable GIST after treatment with imatinib may be feasible (including complete resection); thus the role of surgery needs to be reevaluated throughout the course of a patient’s treatment.

Dr. Raut presented the surgery experience that the Dana-Farber team had with 69 patients. In most respects, it was similar to the German data presented by Dr. Hohenberger. One difference between the two studies was that 14 patients in the German study did not resume kinase therapy (due to financial reasons, intolerance, etc.) and only one patient did not resume kinase therapy in the American study.

Progression-free survival data between the two groups was very similar; for focal progression it was 8 months for both studies, for generalized progression it was 3 months for both studies, and for surgery after stable disease it was 11 months for the German study and has not been reached for the American study.

Dr. Raut addressed the question of what data currently support surgery for advanced GIST, to which he added:

  • Surgery is feasible.
  • Mortality rates are low.
  • Morbidity (frequency of complications following a surgical procedure or other treatment) rates are high, reflecting the complexity of treating this patient population.
  • PFS and overall survival are encouraging but are they better than kinase inhibitor therapy alone?

Raut noted that there are stumbling blocks to designing a prospective trial to answer questions about surgery. One of these is whether or not patients would consent to a randomized surgery trial. The question to ask is who may benefit from surgery?

• Both Hohenberger and Raut studies found that patients who were stable had the best chance of benefit followed by patients with limited (focal) progression.
• Both studies found that patients with generalized progression typically receive little benefit from surgery.
• Survival is better after complete resection when compared to incomplete resection.

Dr. Raut proposed some guidelines for surgery:

• Operate once maximal response (stable disease after maximum shrinkage) is reached on kinase inhibitor therapy
• En bloc resection without tumor rupture
• Extent of resection

□ Complete resection whenever possible
□ Stable disease

■ Remove all lesions

□ Limited disease progression

■ Remove all progressing lesions
■ Debulk additional lesions as much as possible

□ Generalized progression

■ Resect only if symptomatic or emergency

• Peritonectomy/omentectomy (excision or resection of the peritoneum or omentum; these are thin layers of connective tissue that line the abdomen)

□ When practical, for peritoneal seeding

• Margins

□ Wide margins not generally necessary
□ No data on improved survival with wider margins
□ Radical resection when appropriate

• Lymphadenectomy (lymph node removal) not necessary
• Resume kinase inhibitor therapy postoperatively (very important)

Dr. Raut listed some questions that still remain unanswered, such as “Do individuals with certain mutations derive a greater benefit from surgery?” Dr. Raut noted that “. . . this is a very interesting question and one that we are trying to answer ourselves and one that may be able to generate some data.” Raut noted that the goal must be for multimodality therapy to be alternated with appropriately planned surgery in an effort to increase the time patients with advanced GIST are maintained on individual drug therapies.

3. “Sunitinib (SU) response in imatinib-resistant (IM-R) GIST correlates with KIT and PDGFRA mutation status.” Heinrich et al.

Dr. Michael Heinrich gave a presentation titled “Sunitinib (SU) response in imatinib-resistant (IM-R) GIST correlates with KIT and PDGFRA mutation status.” Sunitinib is the generic name for Sutent. It is approved in the United States and Canada. Sunitinib has direct anti-tumor activity against GIST by inhibition of the KIT and PDGFRA proteins. It also has indirect activity by targeting VEGF receptors and PDGFRB. These are proteins that are important in the formation of the new blood vessels that feed tumors.

This presentation emphasized the importance of genotyping when it comes to GIST.

It has been known since 2001 that genotyping predicts how GIST patients might respond to Gleevec. Recently Dr. Maria Debiec- Rychter and colleagues reported that genotyping can predict which patients benefit the most from higher doses of Gleevec. This current report by Heinrich and colleagues shows that genotyping can also predict which patients are more likely to respond to Sunitinib (Sutent) based not only on their primary mutation but also on their secondary mutations.

Results were based on a Phase I/II study of 97 GIST patients that had failed Gleevec or were unable to tolerate Gleevec. Median PFS and median overall survival for the entire group was 7.8 months and 19.0 months respectively.

The distribution of initial mutations in this study was similar to other studies with 84 percent of patients having a mutation in KIT, 5 percent with PDGFRA mutations and 11 percent with wild-type GIST. Clinical response to Sutent was correlated to the patient’s primary genotype (pre-Gleevec). In this study clinical benefit was defined as stable disease for at least 6 months or a partial response; KIT exon 9 GISTs had the best clinical benefit rate (63%) followed by wild-type (56%), KIT exon 11 (36% ) and PDGFRA (25%). Significant tumor shrinkage with Sutent is more common in KIT exon 9 patients (37% had a partial response) compared to KIT exon 11 patients (5%). Note: The ASCO abstract and presentation incorrectly list the benefit rate for KIT exon 9 as 42 percent. The correct value in this study was 63 percent. A previous study presented by Dr. George Demetri at 2004 ASCO found a benefit rate of 80 percent when looking at 15 patients. This difference might reflect the relatively small sample sizes involved (19 and 15 patients). It should be noted that while 63 percent of exon 9 patients benefited for at least 6 months, a substantial portion, 50 percent of exon 9 patients, benefited for at least 19 months.

In addition to its correlation with primary mutations, Heinrich and colleagues were able to show that Sutent has activity against some secondary mutations but not others. Secondary mutations are the most common cause for resistance to Gleevec. For the first time, Heinrich reported that secondary mutations were more common in Gleevec-resistant patients with primary KIT exon 11 mutations (62%) than in Gleevec-resistant patients with primary KIT exon 9 mutations (16%, P=0.002).

In this study, 29 patients had secondary KIT mutations. These included mutations in exon 13 and exon 14, the portion of the gene that codes for the drug/ ATP binding pocket of the protein. All of the mutations in exon 13 were V654A mutations. This was by far the most common secondary mutation (n=12). Heinrich reported that, in the lab, the exon 13 and exon 14 mutations were resistant to Gleevec but sensitive to Sutent.

Almost half of the secondary mutations occurred in exon 17, the kinase activation loop. Five different exon 17 mutations were reported and one exon 18 mutation was reported. Dr. Heinrich reported that, in the lab, these mutations were resistant to Sutent and Gleevec.

Heinrich and colleagues then demonstrated that the lab tests correlated with benefit in patients. They found that the median PFS in patients with sunitinib sensitive secondary mutations (KIT exons 13 and 14) was 8.1 months compared to only 2.3 months for patients with sunitinib-resistant secondary mutations (KIT exon 17 and 18).

Dr. Heinrich concluded that “sunitinib exhibits significant clinical and biological activity in patients with imatinib resistant GIST.” The “clinical benefit of sunitinib was strongly influenced by both primary and secondary mutations in . . . KIT.” Sunitinib was particularly effective for treatment of wild-type GISTs or KIT exon 9 mutations and sunitinib is more potent than imatinib against secondary mutations involving the KIT/ATP drug binding pocket.

4. “Mutation-directed management of GIST-Should genotyping be part of routine practice?” Judson et al.

Dr. Ian Judson of the Royal Marsden Hospital, London, UK, was the discussant for Dr. Heinrich’s presentation in a talk titled “Mutation-directed management of GIST-should genotyping be part of routine practice?” Dr. Judson noted the “. . . big challenge for people that don’t have access to this technology.”

Dr. Judson began his talk by noting that the initial (primary) genotype can predict not only how patients are likely to respond to Gleevec, but also how likely they are to respond to Sutent as second-line therapy. Judson then reviewed the mechanisms of resistance to Gleevec with a focus on the fact that secondary kinase mutations (the most common reason for resistance) were much more common in GIST with primary KIT exon 11 mutations (62%) than in GISTs with primary KIT exon 9 mutations (16%).

Dr. Judson reviewed the EORTC (European Organization for Research and Treatment of Cancer) Gleevec dose response data by genotype (Debiec- Rychter et al.) noting the superior PFS for KIT exon 9 mutations at higher doses of Gleevec. Also noted was the poor response of wild-type KIT, which was worse at higher doses, prompting Judson to speculate that Gleevec may “. . . be doing nothing at all for these patients” (Note: However, there have been some wild-type responses in the phase III trials). Judson noted that in this analysis (which is a subset of the entire EORTC dataset) there was no significant difference in response by dose when all genotypes were combined or for KIT exon 11 patients.

A graph of the Heinrich presentation on Sutent was reviewed and showed the PFS by primary (pre-imatinib) KIT genotype. Gleevec-resistant patients with wild-type GIST and KIT exon 9 mutations both did quite well on Sutent, with a median PFS of 20.9 months and 19.4 months respectively. Gleevec resistant patients with KIT exon 11 primary mutations had a median PFS of only 5.1 months on Sutent.

Although patients with imatinib resistant GIST and KIT exon 11 mutations progress faster on sunitinib than other types, this does not necessarily mean that the drug is intrinsically less active against exon 11 tumors. Dr. Judson speculated that the exon 11 patients were likely to have had a longer imatinib treatment period (median 2 years), having more time to acquire secondary mutations, some of which are resistant to sunitinib (as well as imatinib).

Dr. Judson’s concluded that genotyping at, or soon after, a diagnosis of advanced GIST would clearly benefit patients with KIT exon 9 mutant GIST. In addition, a trial is indicated to compare sunitinib vs. imatinib 800 mg in exon 9 mutant GIST. There is also a case for a study of sunitinib vs. imatinib for first line therapy since relative inactivity vs. exon 11 mutant tumors may reflect longer duration of imatinib therapy and selection pressure to develop secondary resistance mutations.