Skip Navigation U.S. Department of Health and Human Services
Agency for Healthcare Research Quality
Archive print banner
Report on the Relative Efficacy of Oral Cancer Therapy for Medicare Beneficiaries Versus Currently Covered Therapy

This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

Part 2. Imatinib for Gastrointestinal Stromal Tumors (GISTs)


Policy Context of the Current Technology Assessment

Section 641 of the Medicare Prescription Drug, Improvement, and Modernization Act (MMA) calls for a demonstration that would pay for drugs and biologicals that are prescribed as replacements for drugs currently covered under Medicare Part B. The demonstration project will be national in scope and will be limited to 50,000 beneficiaries or $500,000,000 in funding, whichever comes first. Forty percent of the funding for this demonstration will be reserved for oral anti-neoplastic drugs.

CMS has requested an assessment of the efficacy of selected oral cancer therapies included in the demonstration relative to drugs currently covered under Medicare Part B. This assessment will provide information that will be used to evaluate the likely effects of the demonstration on patient outcomes and may also provide underlying information to be used for cost-effectiveness analyses that will be completed by CMS.

The scope of the assessment will be limited to the following demonstration drugs and conditions:

  • Imatinib for treatment of chronic myeloid leukemia.
  • Imatinib for treatment of gastrointestinal stromal cancer.
  • Gefitinib for treatment of non-small cell lung cancer.
  • Thalidomide for treatment of multiple myeloma.

This report is responsive to the second item: an assessment of imatinib for the treatment of gastrointestinal stromal tumors (GISTs).

Clinical Context of the Current Technology Assessment

Although relatively rare, GISTs represent the largest subset of mesenchymal tumors of the gastrointestinal (GI) tract and about 5 percent of all sarcomas.1,2 They originate in the stroma, the connective tissue that supports the organs involved in digestion. GISTs are now known to be derived from muscle-like nerve cells called the Interstitial Cells of Cajal, which coordinate the automatic movements of the GI tract. During the last three decades there has been considerable debate about GIST's cells of origin, nomenclature, diagnosis, and prognosis. Many cases have not shown up in cancer registries, have been misclassified as other cancers, or defined as benign.1,3 For this reason, the exact incidence of GIST is unknown, but is approximated at 5,000 to 10,000 cases in the U.S. annually.4 Two recent studies from Scandinavia measure the incidence somewhat lower at 1.1 to 1.4 per 100,000.26,94 Nilsson 2004)

GISTs occur predominantly in middle and older aged individuals; the median age at diagnosis is about 60 years.2 They affect men and women equally.2 The majority (70 percent) of tumors occur in the stomach, with 20-30 percent in the small bowel and < 10 percent elsewhere in the GI tract.1 GISTs vary greatly in size with tumor sizes ranging from 1 to over 20 centimeters (cm).1

Surgical resection is the primary therapy of choice for all localized gastrointestinal stromal cell lesions at the time of initial presentation.3,5 The goal is to completely remove the tumor. GISTs are now classified based on a spectrum of risk, from very low-risk disease (unlikely to recur or metastasize) to high risk disease which would commonly recur and metastasize with life-threatening consequences; the rapidity and timing of tumor progression and metastasis is highly variable.6 Five-year survival rates after complete resection for all patients overall with primary GIST and without accounting for risk stratification are 48-65 percent.5 The ability to resect the tumor is an important prognostic determinant (Table 1). In a multivariate analysis of 200 cases of GIST referred to a major cancer center, important negative predictors of survival included male gender (relative risk (RR) 1.6, confidence interval (CI) 1.0-2.6), size > 5 cm (5-10 cm (RR 2.8, CI 1.3-6.2), > 10 cm (RR 4.4, CI 2.0-9.8)), and incomplete resection of the tumor (RR 3.9, CI 2.4-6.2).2 In this series, over 80 percent were > 5 cm at the time of diagnosis. Other single institution studies report similar findings.7

When complete resection was possible, local recurrence occurred within 2 years in approximately 10 percent of individuals whose primary tumor has been completely resected, metastatic recurrence in approximately 15 percent, and synchronous local and metastatic in 5 percent.2 Complete resection is not possible in up to 40 percent of cases.2,7,8

Pathology is important for differentiating GISTs from other tumors and prognosis (Table 1). Under the light microscope GISTs are composed of spindle-shaped or epithelioid appearing cells; the cells are usually more numerous with less eosinophilic cytoplasm than leiomyomas.9,10 The immunohistochemical staining pattern is critical for diagnosis. Immunohistochemistry is the use of antibodies or antisera to detect a specific marker within a pathological specimen; when a tumor reacts with an immunohistochemical stain, then it is termed "positive" for that antigen and the tumor is recognized as producing the specific marker of interest. Approximately 70 percent of GISTs are positive for CD34, 20-30 percent are positive for smooth muscle actin (SMA), 10 percent are positive for S100 protein, and 5 percent are positive for desmin.10 CD117, the antigen that corresponds with the c-kit proto-oncogene product, is positive in over 90 percent of GISTs.1,11-15 Other tumors are CD117 positive but most of these do not occur in the GI tract and are not part of the differential for GIST.1 The WHO classification of gastrointestinal tumors indicates that the term GIST should be reserved for CD117 positive tumors, but the current literature accepts the concept that there are some CD117 negative GISTs.16

In GIST a specific change or mutation in DNA at the point of the c-kit proto-oncogene causes a cellular enzyme known as KIT to be switched "on" all the time. These activating mutations are most commonly localized on exon 11 (57-71 percent), exon 9, or exon 13 of the KIT gene.1 KIT is a tyrosine kinase enzyme responsible for sending growth and survival signals inside the cell. If it is "on", the cell stays alive and grows or proliferates. The overactive, uncontrolled mutant KIT enzyme triggers the malignant growth of GIST tumor cells. GISTs with mutant KIT are more likely to be aggressive tumors characterized by more frequent recurrence and a higher associated mortality rate. For example, in a Japanese study of 124 patients with GIST, 89 percent were KIT positive, 57 percent of c-kit mutations were missense mutations on exon 11, and the patients with mutation-positive GISTs showed more frequent recurrences (p < 0.001) and higher mortality (p < 0.001) than did those with mutation-negative GISTs.12 Other groups have observed similar findings, but this was not consistent across all studies.17-19 Exon 9 mutations have also been associated with more aggressive tumor characteristics, including larger tumor size and extra-gastric locations.20,21 CD117 negative GISTs may have a mutation in the platelet-derived growth factor receptor (PDGFR) gene that produces a similar tyrosine kinase; this does not appear to be prognostically significant.15,22,23

Pathological characteristics such as sites of origin outside of the stomach, large size, nuclear atypia, high mitotic rate, and mucosal invasion are also associated with the ability of a cancer to spread.1,15,24 For example, in one series of 1,074 cases of GIST, mitotic activity predicted survival such that only 3 percent of patients with low mitotic activity (≤ 5/50 HPFs) died, as opposed to 46 percent with high mitotic activity (> 5/50 HPFs, p < 0.0001).15 Larger tumor size and location (e.g., stomach fundus and gastroesophogeal junction) undermine tumor respectability and hence are negative predictors of prognosis.5,15 Taken together, tumors with high mitotic activity and larger size are most aggressive, such that 16 percent of patients with small mitotically active tumors (2-5 cm, > 5/50 HPFs) died, as opposed to 49 percent with mitotically active 5-10 cm tumors, and 86 percent with mitotically active > 10 cm tumors.15 These two factors (mitotic count and tumor size) were used by and NIH-convened workshop to define the risk of aggressive behavior in GISTs;25 the validity of this scheme was demonstrated in a subsequent population-based study.26

Recent reviews suggest that scientifically advanced genetic markers, including DNA-copy number changes, telomerase activity, KIT mutation status, and KIT mutation type (point vs. deletion) may be useful in more accurately identifying tumors with malignant potential,13,15,24 however, the most important utility of KIT and PDGFR genotyping may be in prediction of drug effects using kinase inhibitors such as imatinib.27

Table 1: Summary of Major Tumor Characteristics Reported To Predict Poor Survival

Most consistently reported characteristics

  • Size > 5 cm.
  • Inability to completely resect the tumor.
  • Exon 11 and 9 KIT mutations.
  • High mitotic rate.

Less consistently reported characteristics

  • Site of origin.
  • Nuclear atypia.
  • Mucosal invasion.
  • DNA copy number changes.
  • Telomerase activity.
  • KIT point vs. deletion mutations.

In the pre-imatinib era, the usual course after complete resection has been vigilant watchful waiting.3,5 Adjuvant chemotherapy after surgery designed to reduce the chance of GIST recurrence has been moderately successful.28,29 In a 1997 meta-analysis of randomized trials of doxorubicin-containing chemotherapy, adjuvant chemotherapy improved recurrence-free survival (hazard ratio (HR) 0.75, CI 0.64-0.87, p 0.0001) but not overall survival (HR 0.89, CI 0.76-1.03, p 0.12).28 Recurrent disease may be treated with repeat surgery when possible.2 The prognosis for unresectable or metastatic GIST has historically been poor, with 5-year survival rates estimated to be lower than 5 percent.1,5 Historically unresectable or metastatic disease was treated with radiotherapy, single-agent doxorubicin, single-agent ifosfamide, and combination chemotherapy including these agents although none of these have clinically meaningful anti-tumor activity.1 Management of advanced tumor and improvements in survival with these interventions has been poor.16 Hence, the usual plan of care in the pre-imatinib era was resection whenever possible with or without adjuvant therapy, surveillance for recurrence and metastatic disease, followed by chemotherapy for advanced disease generally with doxorubicin- and/or ifosfamide-based regimens and possibly radiotherapy for selected cases.

The UK National Institute for Clinical Excellence (NICE) conducted a systematic review of non-imatinib treatments for advanced GIST.16 The interventions reviewed were heterogeneous including novel chemotherapeutics and/or standard sarcoma chemotherapy regimens. The studies included patients with GISTs, non-specific soft tissue sarcomas, GI sarcomas, and leiomyosarcomas. Overall survival was 72 percent (18-100 percent) at 1 year, 40 percent (30-66 percent) at 2 years, and 16 percent (0-40 percent) at 3 years. A total of 13 of 258 patients (5 percent) achieved a partial response, while 64 (24 percent) had stable disease. NICE concluded that the heterogeneity of treatments, small numbers of patients involved, diagnostic difficulties, and uncontrolled study designs made it difficult to interpret these studies as a true historical baseline for imatinib.

The National Cancer Institute (NCI) health professionals guidelines at are a frequent resource for U.S. oncologists.30 Expert reviewers of this systematic summary of the literature highlight that this NCI guideline is very outdated in terms of recommendations for the management of GIST. In recommending chemotherapeutics, this site does not specifically differentiate recurrent or metastatic advanced GIST from other advanced soft-tissue sarcomas. According to this summary, which is related to soft-tissue sarcomas overall and is not focused on GIST, only doxorubicin and ifosfamide show response rates > 20 percent for advanced disease when used as single agents.

  • Doxorubicin: Single-agent doxorubicin response rates vary from 15-34 percent with median 26 percent; the majority of these responses are partial responses only.31 In 1999, Bramwell, et al. of Program in Evidence-based Care of Cancer Care Ontario conducted a systematic review of randomized trials and developed an evidence-based guideline on doxorubicin for advanced soft-tissue sarcomas (available at In this guideline, single-agent doxorubicin response rates ranged from 16-30 percent; meta-analysis did not show a significant response benefit for single-agent doxorubicin nor combination chemotherapy containing doxorubicin.32 Toxicity data were also reviewed in this document. This summary is related to soft-tissue sarcomas overall and is not focused on GIST.
  • Ifosfamide: In a widely-cited review of chemotherapy for advanced soft tissue sarcomas, single-agent ifosfamide response rates vary from 7-38 percent with median 26 percent in patients with previously treated sarcomas.31 Similar estimates were cited by other reviewers.33 Approximately 24 percent of patients who have progressed after doxorubicin can respond to ifosfamide.34 There is a clear dose-response relationship with single agent ifosfamide.35,36 In sequential studies, one group reported response rates of 10 percent at 6 g/m2, 14 percent at 8 g/m2 , 21 percent at 10 g/m2, and 29 percent at 14 g/m2.36 Again, this summary is related to soft-tissue sarcomas overall and is not focused on GIST.

Combination of these drugs may be more effective, especially for younger patients who can tolerate this program; a randomized trial of 340 patients with advanced sarcoma showed a higher response rate (32 percent vs. 17 percent, p < 0.002) and longer time-to-progression (TTP, 6 vs. 4 months, p < 0.02) for doxorubicin, dacarbazine, ifosfamide, and mesna (MAID) vs. doxorubicin and dacarbazine alone.37 According to the NCI Web site, sequential use of doxorubicin followed by ifosfamide or other drugs with each subsequent recurrence is still frequently preferred for older patients.30 GIST is not specified within these trials, and given the difficulty with diagnosis of GIST during the period during which these trials were conducted any specific referrals to GIST outcomes are suspect.

Across this body of work representing the comparative efficacy for single and multi-agent chemotherapy for GIST, there are repeated problems with reliability of the information. Nearly all of this work pre-dated the ability to distinguish GIST from other soft tissue sarcomas. GIST is likely to be more chemoresistant than other soft tissue sarcomas,38,39 and therefore the estimated efficacy of doxorubicin- and/or ifosfamide-based chemotherapy for GIST is likely to be less than the overall efficacy for soft tissue sarcomas reported here. The inclusion of GIST in the original studies about chemotherapy for soft tissue sarcomas may in fact reduce the estimated efficacy of doxorubicin- and/or ifosfamide-based chemotherapy of other non-GIST soft tissue sarcomas.

The Technology

Both the KIT tyrosine kinase and Platelet-Derived Growth Factor Receptor Alpha (PDGFRA) targets of imatinib are receptor tyrosine kinase proteins that are located within the plasma membrane on the surface of both normal cells as well as the cells of GIST.40,41 In the normal state, KIT is stimulated by stem cell factor (SCF) sending a signal inside the cells that tells them to grow only as needed, while the PDGFRA is stimulated by platelet derived growth factor (PDGF) dimers acting as a ligand for the receptor. As described previously, in the vast majority (95 percent) of GIST lesions, the DNA of the c-kit proto-oncogene is mutated and the KIT and PDGFRA proteins are continuously switched "on" in the absence of regulating ligands. This continuous "on" state is called constitutive KIT or PDGFRA protein kinase activity. This constant signal tells the cancerous cells to keep growing and is a critical contributing factor to the development and maintenance of the malignancy.

Imatinib (STI-571, trade name Gleevec [USA] or Glivec [non-US]) is a derivative of 2-phenylaminopyrimidine.16 Imatinib is a tyrosine kinase inhibitor that targets several different tumor proteins, including the KIT and PDGFRA proteins that are the major etiologic factors of GIST.40 It is a competitive inhibitor of the tyrosine kinases associated with PDGFRA, the Abelson (ABL) protein, and the KIT protein. Imatinib works by blocking, or turning off, the message from these relevant target signaling proteins, so that the cancerous cells stop growing.

Imatinib is available as an oral medication and is usually taken once a day at a recommended dose of either 400 milligrams/day (mg/d) or 600 mg/d.42,43 Imatinib should be administered with a meal and a large glass of water. Doses over 600 mg/d should be administered in divided doses, e.g., 400 mg twice daily. Tablets are available in 100 mg and 400 mg forms. Treatment can be continued as long as there is no unacceptable toxicity.

Imatinib was first used in patients with chronic myelogenous leukemia (CML). It was Food and Drug Administration (FDA) approved for the treatment of patients with KIT (CD117) positive unresectable and/or metastatic malignant GIST on February 1, 2001.42,43

Scope and Key Questions

The key questions for this review were developed with experts in the field of oncology, health economics, and health policy. The key questions are as follows:

  1. In patients with GIST, what is the effect of imatinib compared to doxorubicin and ifosfamide on overall survival, disease free survival, time to progression, CR, PR, and quality of life?
  2. In patients with GIST, what is the effect of imatinib compared to doxorubicin (Adriamycin) and ifosfamide on adverse effects, tolerability, and compliance with treatment?
  3. What patient or tumor characteristics distinguish treatment responders from non-responders and have potential to be used to target therapy?

Return to Contents
Proceed to Next Section

The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care