Key Words. Imatinib Gleevec KIT CD117 Gastrointestinal stromal tumors

The Oncologist Regulatory Issues: FDA The Oncologist CME Program is located online at http://cme.theoncologist.com/. To take the CME activity related to this article, you must be a registered user. Approval Summary: Imatinib Mesylate in the Treatment of Metastatic and/or Unresectable Malignant Gastrointestinal Stromal Tumors MARTIN H. COHEN, ANN FARRELL, ROBERT JUSTICE, RICHARD PAZDUR Office of Oncology Drug Products, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, Silver Spring, Maryland, USA Key Words. Imatinib Gleevec KIT CD117 Gastrointestinal stromal tumors Disclosures Martin H. Cohen: None; Ann Farrell: None; Robert Justice: None; Richard Pazdur: None Section editors Patrick G. Johnston, Peter J. O Dwyer, and Richard M. Goldberg have disclosed no financial relationships relevant to the content of this article. The content of this article has been reviewed by independent peer reviewers to ensure that it is balanced, objective, and free from commercial bias. Target audience: Physicians who wish to advance their current knowledge of clinical cancer medicine in gastrointestinal cancer. LEARNING OBJECTIVES 1. Identify the appropriate imatinib dose for patients after diagnosis and patients after disease progression. 2. Differentiate between the FDA s acclerated and regular drug approval processes and their impact on patient care. 3. Demonstrate familiarity with the safety profile of imatinib. CME This article is available for continuing medical education credit at CME.TheOncologist.com. ABSTRACT The purpose of the present application was to fulfill a postmarketing commitment to provide long-term efficacy and safety data on treatment with imatinib mesylate (Gleevec ; Novartis Pharmaceuticals, East Hanover, NJ) in patients with CD117 unresectable and/or metastatic malignant gastrointestinal stromal tumors (GISTs). In addition, this application also provides evidence to support a change in the label to allow for an escalation of imatinib dosing to 800 mg/day for patients with progressive disease on a lower dose. Two open-label, controlled, multicenter, intergroup, international, randomized phase III studies were submitted one conducted by the European Organization for Research and Treatment of Cancer (n 946) and the other by the Southwest Oncology Group (n 746). These studies compared 400 mg/day of imatinib with 800 mg/day of imatinib. A combined analysis of the two studies was prospectively defined and agreed to by both groups. Both protocols allowed patients randomized to the 400-mg/day imatinib arm to cross over to 800 mg/ Correspondence: Martin H. Cohen, M.D., Food and Drug Administration, White Oak Campus, 10903 New Hampshire Avenue, Building 22, Room 2102, Silver Spring, Maryland 20993-0002, USA. Telephone: 301-796-1344; Fax: 301-796-9845; email: martin.cohen@fda.hhs.gov Received November 19, 2008; accepted for publication January 15, 2009; first published online in The Oncologist Express on February 4, 2009. AlphaMed Press 1083-7159/2009/$30.00/0 doi: 10.1634/theoncologist.2008-0255 The Oncologist 2009;14:174 180 www.theoncologist.com

Cohen, Farrell, Justice et al. 175 day imatinib at progression. Objective responses were achieved in >50% of patients receiving either imatinib dose. The median progression-free survival time was approximately 20 months and the median overall survival (OS) time was approximately 49 months. In the combined analysis, 347 patients crossed over to 800 mg/ day imatinib at the time of progression. The median OS time after crossover was 14.3 months. The most common adverse events (AEs) were fluid retention, nausea, fatigue, skin rash, gastrointestinal complaints, and myalgia. The most common laboratory abnormality was anemia. Most often the AEs were of mild-to-moderate severity. Fluid retention events and skin rash were numerically reported more often in the 800-mg/ day treatment cohort of patients. The Oncologist 2009; 14:174 180 INTRODUCTION On September 29, 2008, the U.S. Food and Drug Administration converted a prior accelerated approval of imatinib mesylate (Gleevec ; Novartis Pharmaceuticals, East Hanover, NJ) to full (regular) approval for the treatment of patients with KIT (CD117)-positive unresectable and/or metastatic malignant gastrointestinal stromal tumors (GISTs). Imatinib s accelerated approval, on February 1, 2002, was based on response rate, a surrogate endpoint judged to be reasonably likely to predict clinical benefit [1]. Accelerated approval requires a postmarketing study to demonstrate clinical benefit [2]. The present application was submitted to fulfill the postmarketing commitment by providing long-term efficacy and safety data from two randomized, phase III trials. In addition, the application also provides evidence to support a change in the label to allow for an escalation of imatinib dosing for patients with progressive disease on a lower dose. PATIENTS AND METHODS In 2001, two open-label, controlled, multicenter, randomized phase III studies were started the first led by the European Organization for Research and Treatment of Cancer (EORTC), which included the Australasian Gastro-Intestinal Trials Group and the Italian Sarcoma Group (n 946), and the second led by the Southwest Oncology Group (SWOG), which included the Cancer and Leukemia Group B, the Eastern Cooperative Oncology Group (ECOG), and the National Cancer Institute of Canada (n 746). These studies were designed to compare 400 mg/day versus 800 mg/day of imatinib in patients with advanced inoperable or metastatic GISTs. The studies were planned simultaneously with similar protocols, and a combined analysis was prospectively defined and agreed to by both groups. Both protocols allowed patients randomized to the 400-mg/day imatinib arm who experienced progressive disease (PD) to crossover to the higher, 800 mg/day, dose of imatinib. Eligible patients in both studies were required to have a biopsy-proven diagnosis of GIST that was CD117 by DAKO antibody staining (DAKO, Glostrup, Denmark), that was of visceral or intra-abdominal origin, and that was distantly metastatic or unresectable. Patients must not have received chemotherapy, biologic therapy, or any other investigational drug for any reason within 28 days prior to registration. Patients must not have had major surgery within 14 days prior to registration. Patients had to have an ECOG/Zubrod performance status score of 0 3 and adequate hematologic, hepatic, renal, and cardiac function. Patients with known brain metastases or uncontrolled medical disease were ineligible. The endpoint of the EORTC study was progression-free survival (PFS), whereas the SWOG study evaluated overall survival (OS), with PFS as a secondary efficacy variable. In both studies the evaluation of disease progression was based on the Response Evaluation Criteria in Solid Tumors [3]. Patients were randomized on a 1:1 basis, stratified by performance status (World Health Organization score of 0 2 versus 3) and measurable versus nonmeasurable disease. In both studies, patients in the 400-mg/day arm could cross over to 800 mg/day of imatinib upon documented disease progression as long as they continued to meet the eli- Table 1. Demographic summary: European Organization for Research and Treatment of Cancer study Variable 400 mg/day 800 mg/day Total n 946 Age (yrs) Median 59.0 60.0 59.0 Range 18 91 18 86 18 91 Age category, n (%) 65 yrs 319 (67.4) 299 (63.2) 618 (65.3) 65 yrs 154 (32.6) 174 (36.8) 328 (34.7) Sex, n (%) Male 283 (59.8) 290 (61.3) 573 (60.6) Female 190 (40.2) 183 (38.7) 373 (39.4) www.theoncologist.com

176 Imatinib in Malignant GIST Table 2. Disease characteristics: European Organization for Research and Treatment of Cancer study Variable, n (%) Category gibility criteria required for study entry and in the absence of safety concerns. A combined analysis of the two studies was preplanned. The objectives of this combined analysis were to assess the efficacy and safety of the two imatinib doses, to assess the efficacy and safety of a dose increase from 400 mg/day to 800 mg/day after progression (crossover subset), and to explore the impact of mutation status on efficacy in each of the two dose groups. The latter endpoint was primarily hypothesis generating and is not discussed further here. PFS was defined as the time from the randomization date to the date of the first documented disease progression or death, whichever came first. The time was censored at the last date of disease evaluation for patients who had not experienced disease progression or death. The OS time was defined as the time from randomization to death (for any reason). Patients still alive and on study were considered as censored at the last date on study. The best overall response was assessed locally. For the EORTC study, the best response was either a complete 400 mg/day 800 mg/day Total, n 946 Performance status score 0 221 (46.7) 218 (46.1) 439 (46.4) 1 188 (39.7) 190 (40.2) 378 (40.0) 2 47 (9.9) 47 (9.9) 94 (9.9) 3 17 (3.6) 18 (3.8) 35 (3.7) Type of disease Measurable 453 (95.8) 459 (97.0) 912 (96.4) Nonmeasurable 19 (4.0) 14 (3.0) 33 (3.5) Missing 1 (0.2) 0 1 (0.1) Time since initial diagnosis 1 yr 244 (51.6) 245 (51.8) 489 (51.7) 1 2 yrs 85 (18.0) 73 (15.4) 158 (16.7) 2 yrs 143 (30.2) 154 (32.6) 297 (31.4) Missing 1 (0.2) 1 (0.2) 2 (0.2) Primary site of disease Abdomen 58 (12.3) 71 (15.0) 129 (13.6) Small bowel 177 (37.4) 150 (31.7) 327 (34.6) Stomach 159 (33.6) 157 (33.2) 316 (33.4) Other GI 67 (14.2) 83 (17.5) 150 (15.9) Nonabdominal or GI 12 (2.5) 12 (2.5) 24 (2.5) Tumor size group 100 mm 304 (64.3) 289 (61.1) 593 (62.7) 100 200 mm 143 (30.2) 159 (33.6) 302 (31.9) 200 mm 23 (4.9) 22 (4.7) 45 (4.8) Missing 3 (0.6) 3 (0.6) 6 (0.6) Prior surgery Yes 410 (86.7) 392 (82.9) 802 (84.8) Prior chemotherapy Yes 156 (33.0) 155 (32.8) 311 (32.9) Prior radiotherapy Yes 26 (5.5) 37 (7.8) 63 (6.7) Abbreviation: GI, gastrointestinal. response (CR), a partial response (PR), stable disease (SD), or PD. For the SWOG study, the best response was Table 3. Demographics: Southwest Oncology Group study Variable 400 mg/day n 345 800 mg/day n 349 Total n 694 Age (yrs) Median 61.0 61.0 61.0 Range 18 86 17 94 17 94 Age category, n (%) 65 yrs 197 (57.1) 193 (55.3) 390 (56.2) 65 yrs 148 (42.9) 156 (44.7) 304 (43.8) Sex, n (%) Male 187 (54.2) 189 (54.2) 376 (54.2) Female 158 (45.8) 160 (45.8) 318 (45.8)

Cohen, Farrell, Justice et al. 177 Table 4. Disease characteristics: Southwest Oncology Group study 400 mg/day Variable, n (%) Category n 345 either a CR, a PR, an unconfirmed CR, an unconfirmed PR, stable/no response, increasing disease, or inadequate assessment. To harmonize the coding between the two studies, for the combined analysis the following categories were created: confirmed CR, confirmed PR, not confirmed response (to include all other categories that were not PD), and PD. RESULTS Phase III EORTC Study From February 2001 to February 2002, 946 patients in total with metastatic and/or unresectable GISTs were enrolled in the study by 56 institutions from 13 countries. Demographic and disease characteristics were balanced between the two treatment arms (Tables 1 and 2). In the PFS analysis, there were 332 patients with disease progression in the 400-mg/day imatinib group and 324 patients with disease progression in the 800-mg/day 800 mg/day n 349 Total n 694 Performance status score 0 130 (37.7) 135 (38.7) 265 (38.2) 1 150 (43.5) 159 (45.6) 309 (44.5) 2 45 (13.0) 40 (11.5) 85 (12.2) 3 11 (3.2) 10 (2.9) 21 (3.0) Missing 9 (2.6) 5 (1.4) 14 (2.0) Type of disease Measurable 323 (93.6) 333 (95.4) 656 (94.5) Nonmeasurable 22 (6.4) 16 (4.6) 38 (5.5) Time since initial diagnosis 1 yr 125 (36.2) 132 (37.8) 257 (37.0) 1 2 yrs 81 (23.5) 72 (20.6) 153 (22.0) 2 yrs 125 (36.2) 135 (38.7) 260 (37.5) Missing 14 (4.1) 10 (2.9) 24 (3.5) Primary site of disease Abdomen 78 (22.6) 84 (24.1) 162 (23.3) Small bowel 103 (29.9) 109 (31.2) 212 (30.5) Stomach 129 (37.4) 124 (35.5) 253 (36.5) Other GI 13 (3.8) 8 (2.3) 21 (3.0) Unknown 22 (6.4) 24 (6.9) 46 (6.6) Tumor size group 100 mm 221 (64.1) 210 (60.2) 431 (62.1) 100 200 mm 97 (28.1) 114 (32.7) 211 (30.4) 200 mm 11 (3.2) 13 (3.7) 24 (3.5) Missing 16 (4.6) 12 (3.4) 28 (4.0) Prior surgery Yes 290 (84.1) 297 (85.1) 587 (84.6) Prior chemotherapy Yes 82 (23.8) 69 (19.8) 151 (21.8) Prior radiotherapy Yes 39 (11.3) 32 (9.2) 71 (10.2) Abbreviation: GI, gastrointestinal. imatinib group. The median PFS times were 19.8 and 24.0 months, in the 400-mg/day and 800-mg/day arms, respectively. The hazard ratio (HR), for 800 mg/day imatinib versus 400 mg/day imatinib, was 0.883 (95% confidence interval [CI], 0.757 1.029; p.1106). In the OS analysis, there were 208 deaths among patients in the 400-mg/day imatinib group and 206 deaths among patients in the 800-mg/day imatinib group. The median survival time was 45 months in both the 400 mg/day and 800 mg/day arms (HR, 0.948; 95% CI, 0.782 1.150; p.5896). The CR and PR rates in the 400-mg/day imatinib and 800-mg/day imatinib groups were 5.3% and 5.9%, respectively, and 44.6% and 48.8%, respectively. These differences were not statistically significant (p.0637). Phase III SWOG Study The SWOG study randomized 746 patients with metastatic or unresectable GISTs, between December 2000 and September 2001 at 57 institutions in the U.S. and www.theoncologist.com

178 Imatinib in Malignant GIST Table 5. Adverse events in 10% of patients in the phase III gastrointestinal stromal tumor trials 400 mg/day (n 818) 800 mg/day (n 822) Reported or specified term All grades (%) Grades 3, 4, or 5 (%) All grades (%) Edema 76.7 9.0 86.1 13.1 Fatigue/lethargy, malaise, asthenia 69.3 11.7 74.9 12.2 Nausea 58.1 9.0 64.5 7.8 Abdominal pain/cramping 57.2 13.8 55.2 11.8 Diarrhea 56.2 8.1 58.2 8.6 Rash/desquamation 38.1 7.6 49.8 8.9 Vomiting 37.4 9.2 40.6 7.5 Myalgia 32.2 5.6 30.2 3.8 Anemia 32.0 4.9 34.8 6.4 Anorexia 31.1 6.6 35.8 4.7 Other GI toxicity 25.2 8.1 28.1 6.6 Headache 22.0 5.7 19.7 3.6 Pain (excluding tumor-related pain) 20.4 5.9 20.8 5.0 Other dermatology/skin toxicity 17.6 5.9 20.1 5.7 Leukopenia 17.0 0.7 19.6 1.6 Other constitutional symptoms 16.7 6.4 15.2 4.4 Cough 16.1 4.5 14.5 3.2 Infection (without neutropenia) 15.5 6.6 16.5 5.6 Pruritus 15.4 5.4 18.9 4.3 Other neurological toxicity 15.0 6.4 15.2 4.9 Constipation 14.8 5.1 14.4 4.1 Other renal/gu toxicity 14.2 6.5 13.6 5.2 Arthralgia (joint pain) 13.6 4.8 12.3 3.0 Dyspnea (shortness of breath) 13.6 6.8 14.2 5.6 Fever in absence of neutropenia 13.2 4.9 12.9 3.4 Sweating 12.7 4.6 8.5 2.8 Other hemorrhage 12.3 6.7 13.3 6.1 Weight gain 12.0 1.0 10.6 0.6 Alopecia 11.9 4.3 14.8 3.2 Dyspepsia/heartburn 11.5 0.6 10.9 0.5 Neutropenia/granulocytopenia 11.5 3.1 16.1 4.1 Rigors/chills 11.0 4.6 10.2 3.0 Dizziness/lightheadedness 11.0 4.8 10.0 2.8 Creatinine increase 10.8 0.4 10.1 0.6 Flatulence 10.0 0.2 10.1 0.1 Stomatitis/pharyngitis 9.2 5.4 10.0 4.3 Lymphopenia 6.0 0.7 10.1 1.9 Abbreviations: GI, gastrointestinal; GU genitourinary. Grades 3, 4, or 5 (%) Canada. Of these, 694 were considered evaluable for analysis. The patient demographics and baseline characteristics were well balanced between the two treatment doses (Tables 3 and 4). The median follow-up time at the analysis cutoff date, July 3, 2006, was 44.2 months (range, 0 64 months). About 20% of all patients were still on treatment at the cutoff date. The percentages of patients who discontinued because of toxicity were 7.5%

Cohen, Farrell, Justice et al. 179 and 7.2% in the 400-mg/day and 800-mg/day arms, respectively. One patient in the 400-mg/day group was not treated because of death. The median OS times were 55.1 months and 51.3 months in the 400-mg/day and 800-mg/ day dose groups, respectively (p.5819). The median PFS times were 17.6 and 19.7 months, in the 400-mg/day and 800-mg/day arms, respectively (HR for 800-mg/day imatinib versus 400-mg/day imatinib, 0.891; 95% CI, 0.753 1.055; p.1804). The CR and PR rates in the 400-mg/day imatinib and 800-mg/day imatinib groups were 5.2% and 3.7%, respectively, and 48.1% and 49.0%, respectively (p.2826). There were 1,640 patients in the combined EORTC SWOG dataset. The median PFS time was longer by 4.3 months in patients receiving 800 mg/day of imatinib compared with patients receiving 400 mg/day of imatinib 23.2 months versus 18.9 months (HR, 0.887; 95% CI, 0.792 0.993; p.0374). No difference was observed between the two dose groups with respect to OS (p.9775). The median OS time was 48.8 months in both dose groups. The two dose groups were also nearly identical in terms of the best overall response. Overall, 5.1% of patients achieved a confirmed CR and 47.5% achieved a confirmed PR. Crossover Subset In total, 347 patients originally treated with 400 mg/day of imatinib in the two studies crossed over to the 800-mg/ day dose after progression. The median time on treatment after crossover was 3.7 months. The estimated exposure rates at 1 and 2 years after crossover were 23.4% and 8.7%, respectively. In total, 22 (6.3%) patients continued to be on treatment 2 years after crossover, and overall 39 patients (11.2%) were still on treatment after the data cutoff. The median OS time after crossover was 14.3 months (95% CI, 12.2 16.7). Response data were not collected. Safety In the combined EORTC SWOG analysis, the median duration of drug exposure was 24 months for both imatinib treatment groups. About one third of the patients from either dose group were treated for 36 months. The majority of imatinib-treated patients experienced adverse reactions at some time. The most frequently reported adverse reactions were edema, fatigue, nausea, abdominal pain, diarrhea, rash, vomiting, myalgia, anemia, and anorexia. Most reactions were of mild-to-moderate severity. The drug was discontinued for adverse reactions in a total of 89 patients (5.4%). Adverse reactions, regardless of relationship to study drug, that were www.theoncologist.com reported in at least 10% of the patients in the EORTC and SWOG trials combined (full analysis dataset) are shown in Table 5. Overall, the incidence of all grades of adverse reactions and the incidence of severe adverse reactions (Common Terminology Criteria grade 3) were similar between the two treatment arms except for edema, which was reported more frequently in the 800 mg/day group. There were five grade 5 adverse events in patients receiving 400 mg/day of imatinib and 10 in patients receiving 800 mg/day of imatinib. Three deaths, all in patients receiving 800 mg/day of imatinib, were considered by the investigator to be related to imatinib treatment, including liver dysfunction in one patient, cardiac arrhythmia in one patient, and tumor hemorrhagic necrosis in one patient. Clinically relevant or severe abnormalities of routine biochemistry laboratory values, except creatinine, were not reported or evaluated in the phase III GIST trials. DISCUSSION GISTs are soft tissue sarcomas thought to arise from gastrointestinal tract mesenchymal stem cells. The most common primary tumor locations are the stomach and small intestine. The median survival time from the time of diagnosis of metastatic or recurrent disease, in the preimatinib era, was reported to be 12 19 months [4 6]. The diagnosis of GIST is based on the histologic appearance of the tumor and on the expression of the cell surface marker CD117. CD117 is an epitope on the extracellular domain of the transmembrane tyrosine kinase receptor KIT. The expression of CD117 can be detected on the cell surface of 85% 90% of malignant GISTs [7, 8]. The subset of GISTs that are CD117 can be divided into two groups those with activating mutations of a related tyrosine kinase, platelet-derived growth factor (approximately 5%), and those without an identified kinase mutation [8]. The latter group might include patients with false-negative immunohistochemistry as a consequence of poor tumor fixation or of levels of KIT protein sufficient for signal transduction but too low for immunohistochemistry detection [9]. Until imatinib, chemotherapy had not been proven to be effective in the treatment of GISTs [10, 11]. in addition to inhibiting the Bcr-Abl tyrosine kinase in chronic myelocytic leukemia, also inhibits the receptor tyrosine kinase for stem cell factor (KIT). The latter finding prompted the investigation of imatinib for GISTs [1, 4]. The studies summarized in this manuscript, as well as others, attest to imatinib effectiveness for GIST treatment [12]. As previously indicated, imatinib received accelerated approval for the treatment of advanced/metastatic

180 Imatinib in Malignant GIST GISTs based on the objective response rate, a surrogate endpoint considered reasonably likely to predict clinical benefit. The trials summarized in this application provide evidence of clinical benefit with a median PFS time of approximately 20 months and a median OS duration of approximately 49 months. In addition, there was evidence that, following progression at a lower imatinib dose, institution of an 800-mg/day imatinib dose was beneficial, with a median OS time of 14.3 months. This is of interest because there was no significant difference in outcome from initial treatment with 400 mg/day or 800 mg/day of imatinib in either the EORTC trial or the SWOG trial when analyzed separately. In the combined analysis of the two trials, however, there was a statistically significant longer PFS time, but not OS time, in favor of the 800-mg/day imatinib dose. REFERENCES 1 Dagher R, Cohen M, Williams G et al. Approval summary: Imatinib mesylate in the treatment of metastatic and/or unresectable malignant gastrointestinal stromal tumors. Clin Cancer Res 2002;8:3034 3038. 2 Office of the Federal Register. Code of Federal Regulations. 21 CFR 314.510 and 21 CFR 601.41. 3 Therasse P, Arbuck SG, Eisenhauer EA et al. New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 2000;92: 205 216. 4 Demetri GD, von Mehren M, Blanke CD et al. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 2002;347:472 480. 5 Demetri GD. Gastrointestinal stromal tumors. In: DeVita VT Jr, Hellman S, Rosenberg SA, eds. Cancer: Principles and Practice of Oncology, Seventh Edition. Philadelphia: Lippincott, Williams, and Wilkins, 2005:1050 1060. No new imatinib safety signals were noted in the two reported studies. In general, imatinib was well tolerated. Imatinib was discontinued for adverse reactions in a total of 5.4% of patients. ACKNOWLEDGMENT The views expressed are the result of independent work and do not necessarily represent the views and findings of the U.S. Food and Drug Administration. AUTHOR CONTRIBUTIONS Data analysis: Martin H. Cohen, Ann Farrell, Robert Justice, Richard Pazdur Manuscript writing: Martin H. Cohen, Ann Farrell, Robert Justice, Richard Pazdur Final approval of manuscript: Richard Pazdur 6 DeMatteo RP. The GIST of targeted cancer therapy: A tumor (gastrointestinal stromal tumor), a mutated gene (c-kit), and a molecular inhibitor (STI571). Ann Surg Oncol 2002;9:831 839. 7 Hirota S, Isozaki K, Moriyama Y et al. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 1998;279:577 580. 8 Heinrich MC, Corless CL, Duensing A et al. PDGFRA activating mutations in gastrointestinal stromal tumors. Science 2003;299:708 710. 9 Heinrich MC, Owzar K, Corless CL et al. Correlation of kinase genotype and clinical outcome in the North American Intergroup phase III trial of imatinib mesylate for treatment of advanced gastrointestinal stromal tumor: CALGB 150105 study by Cancer and Leukemia Group B and Southwest Oncology Group. J Clin Oncol 2008;26:5360 5367. 10 Gold JS, van der Zwan SM, Gônen M et al. Outcome of metastatic GIST in the era before tyrosine kinase inhibitors. Ann Surg Oncol 2007;14:134 142. 11 Joensuu H, Fletcher C, Dimitrijevic S et al. Management of malignant gastrointestinal stromal tumours. Lancet Oncol 2002;3:655 664. 12 Blanke CD, Rankin C, Demetri GD et al. Phase III randomized, intergroup trial assessing imatinib mesylate at two dose levels in patients with unresectable or metastatic gastrointestinal stromal tumors expressing the kit receptor tyrosine kinase: S0033. J Clin Oncol 2008;26:626 632.