Advances in the Care of Patients with GIST: Paths to Progress
Co-chairs: George D. Demetri, MD and Michael C. Heinrich, MD
Nancy Berezin, representing GIST Support International at the 2011 ASCO meeting, prepared the following summary of the talks presented in this satellite symposium held in conjunction with the 2011 ASCO convention and supported by an unrestricted educational grant from Novartis.
After Dr. Demetri welcomed the group, Dr. Heinrich provided a brief overview of landmark events in GIST research – beginning with the 1998 Science paper in which Hirota and colleagues identified positive KIT staining in 94% of GIST tumors. An estimated 5,000 to 6,000 new cases are diagnosed annually in the US. Small GISTs are often asymptomatic or produce vague GI discomfort. Larger tumors may cause hemorrhage, anemia, anorexia, weight loss, nausea, fatigue, or intraperitoneal bleeding or perforation. Morphologically, GIST closely resembles leiomyosarcoma; hence the importance of immunohistochemistry in confirming the diagnosis. On pathology, 70% of GISTs exhibit a spindle cell pattern, roughly 20% are epitheloid in pattern, and a few are of mixed spindle and epitheloid pattern. Epitheloid GISTs can be confusing to pathologists who are inexperienced in sarcoma diagnosis, said Dr. Heinrich. He ended his introduction with a reference to the surprising molecular heterogeneity of GIST – a point that would be repeated throughout the presentations.
Early-stage Disease: New Insights on Adjuvant Therapy
Peter W. T. Pisters, MD
Dr. Pisters began his talk with the still-unanswered question of how microGISTs (0.2 to 10 mm) differ from clinically significant GISTs. MicroGISTs are found in up to a third of patients autopsied for other causes. There are 100 million people in the US and “perhaps 15 to 20 in this room,” he said, with microGISTs in their stomachs. Yet only 5,000 clinically significant GISTs are diagnosed annually in the US. Why? What is the driving force behind malignant transformation? We know it isn’t just KIT, he said, because even small GISTs often have KIT mutations. There must be a secondary event triggering the process.
He then launched into discussion of adjuvant imatinib (IM) therapy -- suggesting that physicians often don’t introduce this topic with patients as skillfully as they could. He emphasized the usefulness of quantitative data (in particular, the AFIP criteria developed by Miettinen and Lasota) in helping patients make informed choices. Risk tolerance varies, said Dr. Pisters, so it’s best to give patients actual numbers to work with rather than descriptive terms like “low risk” or “intermediate risk.” Some will view a 15% risk of recurrence as high and opt for adjuvant IM, while others will focus on the probability of surgical cure and veto drug therapy on the grounds of potential adverse reactions. Personal finances may also factor into the decision, he added.
Surgery remains the only curative therapy for localized, resectable GIST, Dr. Pisters told the audience -- yet general surgeons are often unaware of the basic principles of GIST surgery. For the uninitiated, he described primary GIST resection as equivalent to lumpectomy. The surgical goal should be complete removal of the tumor with maximum preservation of vagal innervation and stomach reservoir capacity. Margins do not have to be wide (as are required for patients with adenocarcinoma of the stomach) in order to be adequate. No lymph node dissection is necessary, because GIST rarely spreads to the lymph nodes. Frozen sections are likewise unnecessary. Dr. Pisters said that too many patients are still subjected to total gastrectomy in settings where there is no oncologic rationale for sacrificing the vagal nerves, and a much smaller operation would suffice. He also stressed the need for close communication between surgeon and pathologist. There is no excuse for failing to order a mitotic count, he said -- yet this still happens when surgeons lack experience with GIST.
On the question of laparoscopic vs open surgery for gastric GIST, Dr. Pisters said that the 2006 management algorithm by Otani et al remains an effective guide: patients with tumors smaller than 2 cm in size can be managed by active surveillance only; tumors between 2 and 5 cm in size can generally be managed laparoscopically; those greater than 5 cm require laparoscopically assisted or open surgery. He noted that different countries use somewhat different cutoff points: in Korea, for example, patients with gastric GISTS as large as 3 cm may be followed endoscopically rather than referred for surgery.
The last portion of Dr. Pisters’ talk consisted of a review of the adjuvant IM trials, concluding with a question on everyone’s minds: For how long should adjuvant therapy be continued? He cautioned against generalizing the findings of the SSG XVIII Phase 3 trial (IM one vs three years), which enrolled only high-risk patients, to those with only moderate risk of progression. He noted that In the ACOSOG Z9001 Phase 3 trial (IM one year vs placebo) by far the greatest gains were seen in patients with larger, more aggressive tumors. Data from a second major Phase 3 trial, EORTC 62024 (IM two years vs observation), will not be forthcoming for some time, Dr. Pisters told the group, because the primary endpoint of that trial is overall survival (OS) rather than recurrence-free survival (RFS). Since EORTC 62024 was designed, he said, “survival of patients with metastatic GIST has changed from being calibrated in months to calibrated in years.” Moreover, European health authorities tend to be more cautious than those in the US when it comes to allocating limited resources for expensive targeted agents. Perhaps, suggested Dr. Pisters, the need to prevent recurrences and allow patients to remain in the workforce may be seen as more urgent in the US -- where health coverage is typically tied to employment -- than in countries where health coverage is universal and the government’s responsibility.
Improving Outcomes for Patients with Recurrent or Metastatic GIST
Michael C. Heinrich, MD
Dr. Heinrich led off with a brief overview of the molecular biology of GIST: Approximately 80 to 85% of GIST tumors involve gain-of-function KIT mutations. Another 5 to 10% involve PDGFRA mutations. A small minority of GIST tumors lack kinase mutations and are referred to as wild type (WT). However, this is a heterogeneous population that probably involves five or more distinct types of disease rather than a single type.
Mutational information is “not just something for molecular biologists,” said Dr. Heinrich. It has become an intrinsic part of diagnosing GIST and optimizing the therapeutic approach. For patients with KIT-mutant Exon 11 GISTs, tumor response has been shown to be similar whether the IM dose is 400 or 800 mg. By contrast, 800 mg is now accepted as the recommended dose for KIT-mutant Exon 9 GISTs. Patients with PDGFRA-mutant GIST have different dosing requirements depending on their mutational status. The 400 mg dose is appropriate for those with mutations at Exon 12 or 14, as well as those with Exon 18 mutations that are D842V negative/IM sensitive. Patients with the D842V mutation or other IM-resistant mutations require higher IM doses (generally 800 mg).
Dr. Heinrich stressed that a good clinical response to therapy did not have to involve dramatic tumor shrinkage. “As long as the tumor is not growing,” he said, “the patient is doing well clinically and is likely to continue to do well.” Because GIST stem and progenitor cells are intrinsically resistant to IM, he said, it is unlikely that any length of treatment can achieve a cure, regardless of how many mature, differentiated cells are “pruned” by the drug. Once therapy is stopped, a new generation of mature GIST cells will expand to fill the void, and the disease will progress. Does this mean that IM therapy should be continued for years without a break? Dr. Heinrich is not convinced that is the case. He reminded the audience of the French data (Study BFR14) showing that control is nearly always achievable in patients who resume IM therapy following a break. Perhaps, he speculated, IM would remain effective for longer periods if the drug were given intermittently, with close surveillance, rather than on a continuous basis.
Dr. Heinrich then identified three potential strategies to optimize frontline TKI therapy in GIST: (1) improve patient compliance, (2) optimize dosing using drug level testing, and (3) surgically resect localized disease in patients with drug-arrested tumors. He pointed out that compliance with oral anticancer medications is surprisingly low: one study in CML patients (Wu et al) showed it to be only 79%. However, counseling can improve compliance, especially when side effects are the main issue.
Drug trough level testing remains the only accurate method for assessing IM pharmacokinetics, said Dr. Heinrich. Studies suggest there is up to a tenfold difference between patients with the lowest and highest IM trough levels – and those with the lowest levels tend to have the shortest time to progression. In a study presented at this year’s ASCO, IM drug levels declined after the first month, suggesting that more than one evaluation may be needed. (Whether the drop in drug levels seen in the study was due to reduced patient compliance or some other cause was not determined.) Who should be tested? He recommended testing patients with unusual or severe side effects and also those with minimal side effects, who may not be taking IM as directed or may simply be unresponsive to it. Other potential candidates include patients with complicating conditions whose response to IM may be hampered by drug-drug interactions. Patients who progress while on IM do not require drug level testing, Dr. Heinrich said. Instead, dosage should be escalated or the patient switched to a different TKI. If tumor growth is localized, surgical resection (with continued TKI therapy) may be a further option.
Updates on the Molecular Basis for Resistance in GIST
Christopher L. Corless, MD, PhD
Dr. Corless showed a dramatic slide series illustrating the capacity of IM to shrink GIST metastases in TKI-naïve patients within only a few days or weeks. But while some patients experience rapid tumor shrinkage followed by long-term stabilization, more typically the effectiveness of IM is limited (and possibility of cure excluded) by development of primary or secondary resistance.
“This is not a simple disease,” said Dr. Corless, explaining that GIST is characterized by much greater variation at the molecular level than had been expected. Patients with KIT-WT GIST in particular may have any of an array of primary mutations including mutations of BRAF and/or KRAS downstream of KIT, increased IGF1R expression, or germline mutations of succinate dehydrogenase (SDH). No agent in existence can adequately address all of these, he said. As for the 5% or so of GIST patients with PDGFRA Exon 18 mutation D842V, they have primary resistance to all of the TKIs, including IM, sunitinib (SU), nilotinib, and sorafenib. The good news for this group, he said, is that crenolanib, a new oral drug being developed by AROG pharmaceuticals, has shown efficacy in blocking the activity of D842V mutant kinases in a Phase 2 clinical trial. (A Phase 3 trial is planned.)
Primary resistance only occurs in about 9% of GIST cases, said Dr. Corless. Most patients initially respond to IM; then, a few years down the road, a new nodule will develop. “It is now accepted,” he said, “that the most common mechanism for the development of such late resistance is the acquisition of a new mutation.” He described two “hot spots” where these acquired mutations tend to cluster. One is located at exons 13 and 14, which encode the ATP binding pocket – also the TKI binding pocket. Replacement of a single enzyme at this point can completely prevent the drug from getting into the pocket. The other hot spot is located in the vicinity of exons 17 and 18 – the activation loop. Mutations here tend to alter the folding pattern of the loop, so drugs have difficulty remaining inside the pocket.
How do current second- and third-line TKIs stack up against these resistance mutations? Dr. Corless said that SU has considerable activity against ATP binding pocket mutations but not against activation loop mutations. Sorafenib shows activity against both, but its activity against the activation loop mutations is typically limited. He noted that secondary cross resistance to SU is much higher among Exon 11 GIST patients than among Exon 9 or WT patients -- hence the observation that Exon 11 patients switched from IM to SU tend to progress more quickly than either Exon 9 or WT patients. Over time, however, the latter groups also become resistant to SU.
The heterogeneity of GIST mutations is such that, even within a single patient, different lesions may carry different secondary mutations. “Depending on where you biopsy,” said Dr. Corless, “you might have very different expectations about that patient’s drug response.” In order to control tumor growth over the long term, he emphasized, a drug would have to target multiple resistance pathways at once. Investigational drugs that may do that include regorafenib (a multikinase inhibitor currently in Phase 3 trials that targets VEGFR1/3 as well as KIT, RET and B-RAF) and the “switch pocket” TKIs that, at least in the laboratory, interfere with the folding of the activation loop at exon 17. Another target of continued interest is heat shock protein (HSP) 90 -- an important stabilizer of mutant forms of KIT. If we can knock out HSP 90, said Dr. Corless, we should see degradation of the KIT molecule and elimination of aberrant signaling.
When the Standard of Care Fails: How to Manage Patients Refractory to Imatinib
George D. Demetri, MD
Dr. Demetri reviewed the management of patients who have progressed on both IM and SU. Ten years down the road, more than 75% of patients with metastatic or unresectable GIST will have “failed” first-line TKI therapy. What is meant by drug failure? Dr. Demetri emphasized that recurrence of GIST after a patient has stopped taking IM as a first-line agent does not count as drug failure. In the absence of outright medical intolerance, such recurrences should always be managed with resumption of IM.
True drug failure he defined as progression that occurs while the patient is taking the agent. Even here, however, Dr. Demetri emphasized that progression may not be multifocal. Localized progression – for example, a single lesion in the liver – often can be eliminated with surgery or radiofrequency ablation (RFA), with continuation of first-line IM to keep the remainder of the disease under control. “If the TKI appears to be working systemically, do not give up on it simply because of a local problem,” he urged. Multifocal progression is more challenging, but surgical resection may still be a possibility if the lesions are clustered in one discrete area.
Dr. Demetri then briefly discussed the structural differences between IM and SU that allow SU, a smaller and more agile molecule, to slip into the drug-binding pocket. The trade-off for patients is, of course, increased toxicities -- including a peculiar tendency to develop gray hair while on the drug. But the biggest frustration of switching from IM to SU is that the duration of response may be disappointingly brief. Dr. Demetri stressed that this was not because SU is intrinsically ineffective, but because patients have already developed cross resistance as a result of their IM therapy.
What happens when both of these TKIs fail? Dr. Demetri expressed optimism about crenolanib and regorafinib, as well the monoclonal antibody IMC-3G3. He added that some patients with TKI-resistant GIST will still respond to nilotinib, although the response is rarely sustained. He believes that the recent trial in which nilotinib failed to show superiority to IM is not the last word on this agent; nor have we seen the last of HSP-90 inhibition. Keep in mind, he said, that “the seeds of secondary mutation are already present in primary GIST.” The idea that a single molecule could tackle all of them was unrealistic. Multiple approaches are needed. And returning to one of his favorite themes, he brought the symposium to a close with a reminder that patient participation in new clinical trials is key to developing the next generation of GIST therapies.