The Era of Molecular Therapy in the Treatment of Gastrointestinal Soft-Tissue Sarcoma: The GIST Paradigm

While first-line intervention for gastrointestinal stromal tumors (GIST) is surgical excision, an 80% recurrence rate exists within the first 24 months. Moreover, GIST are highly resistant to chemotherapy. At a symposium held in March 2004, during the Annual Meeting of the Society of Surgical Oncology, a panel of experts reviewed the latest developments in surgical and medical management of GIST.

This program was supported by an unrestricted educational grant from Novartis Oncology.

Introduction

“Gastrointestinal stomal tumors (GIST) are the most common sarcoma of the GI tract, with an estimated 4,000 to 6,000 new cases per year in the United States,” said George D. Demetri, MD, Associate Professor of Medicine, Harvard Medical School, Boston. They have a mesenchymal origin and strike people of all ages. GIST has only recently (1999) been identified as a distinct clinical pathologic entity; before that time, there was a potential for misclassification.
Patients are often asymptomatic, unless the tumor is large. The incidence of recurrence following initial resection is high, and there is very short survival
following metastatic disease.

“The introduction of the drug imatinib mesylate has revolutionized treatment for GIST,” said Dr. Demetri. He explained that 83% of GIST patients benefit from this targeted therapy.

GIST is a sarcoma that is now reasonably understood at a molecular level. It’s more common than previously thought and should always be viewed as having some malignant potential. There is an effective systemic therapy, which can inhibit the molecular pathways responsible for it.

Differential Diagnosis of Gastrointestinal Soft-Tissue Sarcoma: Identification of GIST

Cristina R. Antonescu, MD, Assistant Attending Pathologist and Director, Electron Microscopy Labor-atory, Memorial Sloan-Kettering Cancer Center, New York, discussed pathologic diagnosis of GIST, immunophenotyping, and types of KIT mutations and links to prognosis.

GIST have historically been classified as smooth muscle tumors, a category of neoplasms including leiomyoma, leiomyosarcoma, and leiomyoblastoma because of histologic similarities such as spindle-shaped and epithelioid cellular morphologies. Up to 72% of tumors currently diagnosed as GIST were
previously (1983 to 2000) misdiagnosed as smooth muscle tumors or other types of cancer.

“The morphologic appearance of GIST are very similar to leiomyoma or leiomyosarcoma, so making the distinction by light microscopy is challenging,” said Dr. Antonescu. “Therefore, the introduction of immunohistochemical staining for KIT, a protein tyrosine kinase whose constitutive activity is
implicated as the basis for most GIST, has provided pathologists with a powerful tool,” she continued. About 95% of GIST express KIT at high levels. Another marker that can be used to confirm the diagnosis is CD34 (seen in two-thirds of cases). Also, GIST are usually negative for S100 protein and desmin.

She explained that the spindle cell morphology is seen in about two-thirds of cases; about 15% to 20% of cases show the epithelioid morphology; and about 10% show mixed morphology. Spindle cell tumors show a strong and diffuse immunoreactivity with KIT staining, while epithelioid tumors show a weaker and more focal staining.

KIT staining will identify most, but not all, GIST. There are some potential pitfalls, such as false negatives due to technical problems or because of the weak and focal staining of epithelioid tumors that may be dismissed as a negative result. Additionally, a KIT-positive tumor may be something besides GIST, such as a desmoid tumor. Also, 5% or fewer GIST are KIT negative, most of which are of epithelioid morphology.

Once the diagnosis of GIST is confirmed, the malignant potential must be assessed. Risk stratification is based mainly on tumor size and mitotic rate.
Dr. Antonescu went on to describe KIT mutations. “The overall mutation frequency is approximately 87.2%,” she said. In KIT, the most common mutations are seen in the juxta membrane domain in exon 11, followed by extracellular domain in exon 9.

Data from the B2222 trial have shown that patients with mutations in the KIT exon 11 have the best response to imatinib mesylate therapy (partial response [PR] 85%); those with KIT exon 9 mutations have an intermediate response (PR 46%), while patients with no kinase mutations have the worst
response (PR 8%).

Surgical Management of GIST

Ronald P. DeMatteo, MD, Assistant Attending Surgeon, Memorial Sloan-Kettering Cancer Center, New York, discussed surgical management of GIST.

He noted that patients with small tumors are often asymptomatic and may go undetected. As the tumors get larger, they may be visible on imaging studies. Eventually, the patient may develop vague GI symptoms. About 20% of patients have GI or intraluminal bleeding or frank tumor rupture.

“Computed tomography (CT) and positron emission tomography (PET) are valuable diagnostic tools for confirming suspected GIST, including determining its extent and resectability,” said Dr. DeMatteo.

Dr. DeMatteo summarized the main surgical principles for GIST. “Completely resect all gross disease and try to get a negative margin,” he said, adding that “it’s not necessary to do a lymphadenomectomy.” However, surgery alone is inadequate. He cited data showing the 5 year survival rate for patients with GIST who undergo complete gross resection is just 54%.

A number of factors will determine how well a patient does following surgery, including tumor size, mitotic index, and location of tumor. “In general, a tumor in the intestine is worse than a similar size tumor in the stomach,” said Dr. DeMatteo. Patients with small tumors (less than 5 cm) tend to do well, while patients with large tumors (over 10 cm) often have early recurrence, and by 5 years only 20% of them are alive.

“Now that we have a drug that can potentially rescue patients once they develop a recurrence, what do we do?” continued Dr. DeMatteo. He recommended following patients outside of a clinical trial closely with CT scans every 3 to 6 months for 5 years.

For most patients who have a recurrence, it will occur in the peritoneum, the liver, or both. “Of all patients with recurrence, we completely resect about half with peritoneal disease and only about one-third with liver metastasis,” said Dr. DeMatteo. Survival for these patients is quite low.

NCCN Guidelines
Recently, the National Comprehensive Cancer Network (NCCN) developed a set of clinical practice guidelines for GIST. Dr. DeMatteo briefly summarized these guidelines, which will be available in greater detail on the NCCN web site.

For patients with primary disease without metastasis, surgery is still the gold standard. “You might consider also giving neoadjuvant imatinib mesylate
or entering the patient into a clinical trial of adjuvant imatinib,” said Dr. DeMatteo.

For patients with a tumor that is marginally resectable or older patients who may not tolerate a large debilitating operation, neoadjuvant imatinib
mesylate is recommended. “This may convert a larger tumor into a slightly smaller one, making the operation easier,” said Dr. DeMatteo. For patients with primary disease that’s completely unresectable or accompanied by synchronized metastases, surgery is not recommended. Instead, imatinib mesylate should be given to try to downstage these patients.

For any metastatic disease or recurrence, the guidelines recommend giving imatinib mesylate, which has an 83% response rate. Dr. DeMatteo remarked that some resistance to imatinib mesylate is starting to develop. “Imatinib is a great drug, but it’s rarely curative,” he said. “In selective cases, we have to go in and take out the resistant tumor.”

Dr. DeMatteo then raised the following question: “What should we do with patients who have a recurrence and who respond to imatinib mesylate? Should we debulk these tumors or should we only operate if we can completely resect them? When should we do it?” He hypothesizes that the chance of a resistant clone is probably proportional to the total tumor load.

Currently, patients who have a recurrence go on imatinib mesylate. Those who respond go into a long asymptomatic holding pattern. Event-ually, many will develop resistance, at which point the drug is stopped, another drug is started, and they may have surgery. Dr. DeMatteo suggested that better outcomes may result if the surgery is done during the period of stable, asymptomatic disease. “Perhaps we should consider these patients for
debulking or even complete resection,” he concluded.

Optimizing Medical Management of GIST: A Multidisciplinary Approach

“Imatinib therapy provides the first effective nonsurgical treatment for GIST,” began George D. Demetri, MD, Associate Professor of Medicine, Harvard Medical School, Boston. “However, there are issues surrounding whether conventional criteria accurately measure clinical benefit and antitumor activity of imatinib mesylate therapy.”

While the results have been dramatic for patients with metastatic or inoperable tumors given imatinib mesylate, more work needs to be done. “We need to cleverly and appropriately integrate surgery with drug therapy in patients with high or intermediate risk disease to cure more patients and prevent the disease from coming back,” said Dr. Demetri.

With imatinib mesylate, 83% of patients have tumor shrinkage, and for about two-thirds of patients tumors will shrink by more than half. “Close to 20% of patients will not have tumor shrinkage, but will have manifestations of response that are not captured by conventional rules,” said Dr. Demetri.

He described the case of a patient with rapidly growing disease and liver metastasis. She was put on imatinib mesylate, which produced a qualitative improvement. However, three years later the lesions hadn’t shrunk. “Something dramatic happened in this patient that was not captured by simply measuring the size of her tumor,” said Dr. Demetri. “You have to look at the qualitative density of lesions.” He noted that oncologists often use tumor shrinkage as a surrogate marker for clinical benefit, which is not necessarily adequate.

It’s also important to be aware that the response to imatinib mesylate is not always immediate. In early studies, the response rate was reported to be about 43%. But with more mature and robust follow-up data, the response rate went up to about 80% at six months.

“What matters even more than response is nonprogression,” said Dr. Demetri. “Arrest the progression of the disease and you’re doing well for the
patient.”

He went on to discuss molecular pathophysiology of the disease, noting that GIST can have mutations in different locations and these may correlate to prognosis. It’s important to keep giving the drug over time. With the drug on board, most of the cells may die, but a small percentage may simply remain static. If the inhibitory effect of the drug is withdrawn, there may be reactivation of these inactive but sensitive clones. “The duration of imatinib treatment is life long for patients with metastatic disease,” said Dr. Demetri.

“In addition, we are now starting to see clonal evolution of GIST and a limited resistance to imatinib therapy,” he continued. For patients with limited clonal resistance, if a progressing lesion is deemed resectable, surgery or a local technique, such as radiofrequency ablation, should be considered, along with continuation of imatinib mesylate to maintain control over the sensitive clones. Optimizing the dosing of imatinib mesylate, consideration of other investigational therapies, and exploring other new agents in combination with imatinib mesylate should also be considered in patients who exhibit some clonal evolution of the disease.

One investigational compound that appears promising is SU11248, which, like imatinib mesylate, is a kinase inhibitor. “We’ll see more such drugs and more combinations,” said Dr. Demetri. “If we cannot resect the resistant clones, we should consider referring these patients to centers that are doing
investigational new drug studies,” he added.

Future Clinical Directions in the Management of GIST

“Burton L. Eisenberg, MD, Professor of Surgery, Dartmouth Medical School, Lebanon, New Hampshire, discussed some of the, as yet, unanswered questions about optimizing therapy for patients with GIST.

He began with a discussion of optimal dosing of imatinib mesylate. In clinical trials, 400 mg per day was used as the starting dose in the U.S., while in Europe 800 mg per day was used.

Two companion Phase III trials have been completed in Europe (European Organisation for Research and Treatment of Cancer [EORTC]) and the United States (Intergroup) to compare efficacy between the two doses. An interim analysis of the EORTC reported improved progression-free survival with the 800 mg per day dose (Proc Am Soc Clin Oncol. 2003;22:814 Abstract 3272). This study had a crossover design (patients who failed on 400 mg were treated with 800 mg); it’s unclear what effect this had on the study.

The U.S. Intergroup study found no difference in progression-free survival between the two doses and no difference in response rate (Proc Am Soc Clin Oncol. 2003;22:814 Abstract 3271). Dr. Eisenberg noted that some progressing patients, however, may benefit from dose escalation.

Nevertheless, surgery remains the major treatment for GIST, especially for patients with primary disease. However, less than 50% of tumors can be completely resected at presentation, and patients with unresectable tumors tend to have poor median overall survival (less than 12 months). Therefore, there is a role for combining imatinib mesylate therapy with surgery for downstaging disease in these patients. Additionally, as microscopic lesions containing viable GIST cells can remain following resection, the combination of imatinib mesylate and surgery may be a treatment option for both resectable and advanced/ recurrent disease.

In the adjuvant setting, imatinib mesylate therapy may reduce the likelihood of recurrence in high-risk patients or patients with positive surgical
margins, as well as manage microscopic sites of disease. This is being evaluated in two American College of Surgeons Oncology Group (ACOSOG) trials.
The Phase II Z9000 study, which has been completed for accrual, included high-risk patients: those with tumors 10 cm or larger, those who had perforation, and those with multiple GIST. They were treated postoperatively with imatinib mesylate 400 mg per day for one year and followed for overall survival.

The Phase III Z9001 study, which is still active, includes patients with tumors of 3 cm or larger and patients with high risk factors, such as perforation, half of whom are given imatinib mesylate 400 mg per day and the other half placebo. Patients in the placebo group who have a recurrence are put on imatinib mesylate. For patients in the treatment group who have a recurrence, the dose is escalated to 800 mg.

Questions also remain as to whether neoadjuvant imatinib mesylate can improve the outcome for patients with advanced primary or recurrent GIST and whether there is a role for surgery in these patients. Dr. Eisenberg emphasized that patients should have surgery if it’s possible, but he said it’s not clear whether it’s sufficient to debulk tumors to the point of minimal disease and then use imatinib mesylate in the postoperative setting.

Potential advantages of neoadjuvant imatinib mesylate therapy include organ sparing, enhanced resectability, the ability to do in vivo drug testing and molecular mechanistic studies pre- and post-imatinib mesylate therapy, and the possibility of obviating resistance.

These issues are being studied in the Radiation Therapy Oncology Group 0132 trial, which includes patients with high-risk primary GIST, recurrent GIST, or metastatic GIST that may potentially undergo surgical debulking. Patients receive 600 mg imatinib mesylate per day for 8 to 10 weeks at which point they have resection if they have stable disease or partial response. They then continue on the drug for two years after surgical resection.

“The surgical experience so far from this trial suggests an acceptable drug-related toxicity and no signs that the drug is adversely affecting bleeding during surgery or wound healing,” said Dr. Eisenberg.

While many advances have been made, some questions surrounding combined-modality therapy for patients with GIST remain unanswered, including the timing of surgery and imatinib mesylate therapy, assessment criteria, duration of imatinib mesylate therapy in the adjuvant setting, and prognostic molecular markers to identify the best candidates for neoadjuvant or adjuvant therapy.