Whom to treat and how long to treat after resection of GIST Professor John R Zalcberg Chief Medical Officer & Director, Division of Cancer Medicine Peter MacCallum Cancer Centre Chair, Australasian Gastro-Intestinal Trials Group

Disclosures  Novartis  Pfizer  Bayer  Amgen  BMS Research / Travel Support / Advisory Board

Risk of Recurrence After Resection of Primary GIST DeMatteo RP et al. Cancer. 2008;112: Approximately 40% of patients who undergo complete resection of primary GIST have a recurrence within 5 years

Risk Assessment  Accurate assessment of risk of aggressive malignant behaviour in GIST poses a challenge 1  Morphologic features most predictive of outcome 1,2 - Mitotic index - Tumour size  Tumour site and rupture also affect risk of recurrence and progression 2,3  Mutational status is useful in predicting treatment response in the metastatic setting 4,5  ?applicable in the adjuvant setting 1. Fletcher CD et al. Hum Pathol. 2002;33: Demetri GD et al. J Natl Compr Cancer Netw. 2007;5(suppl 2):S1-S Miettinen M, Lasota J. Arch Pathol Lab Med. 2006;130: Debiec-Rychter M et al. Eur J Cancer. 2006;42: Heinrich MC et al. J Clin Oncol. 2003;21:

Primary GIST: Risk Factors for Recurrence After Surgery Adapted with permission from DeMatteo RP et al. Cancer. 2008;112: Rates of RFS were independently predicted by mitotic index, tumour size, and tumour location

Overall Survival by Risk Group AFIP, Armed Forces Institute of Pathology. Adapted with permission from Goh BKP et al. Ann Surg Oncol. 2008;15: Cumulative Survival

Specific KIT Mutations Have Prognostic Importance RFS in 127 patients with completely resected localized GIST based on mutation type Proportion Recurrence-Free Years After Resection P<0.001 KIT exon 9 mutation (n=4) KIT exon 11 DEL557/8 (n=35) No mutation (n=29) KIT exon 11 PM/INS (n=32) Other KIT exon 11 deletion (n=17) PDGFRA mutation (n=8) DeMatteo RP et al. Cancer. 2008;112:

Risk Stratification of Primary GIST: Miettinen (AFIP) Original source: Miettinen M, Lasota J. Semin Diagn Pathol. 2006;23: Data are based on long-term follow-up of 1055 gastric, 629 small intestinal, 144 duodenal, and 111 rectal GISTs. † Denotes small numbers of cases. ≈ Tumour size categories combined for both duodenal and rectal GISTs because of small numbers. ∂ No tumours of such category were included in this study.

Nomogram to Predict RFS Following Complete Surgical Resection of Primary GIST Gold JS et al. Lancet Oncol. 10; , 2009

Risk Classification – Room for Refinement 1. Wardelmann E et al. Virchoves Arch. 2007;451: Joensuu H. Hum Path. 2008;39: Dei Tos AP et al. J Clin Oncol. 2009;27(suppl). Abstract  Additional factors: - Mutational status 1 - Rupture 2 - Necrosis 3

Joensuu H et al. Lancet Oncol, 13; , 2012

Adjuvant Studies of Imatinib TrialNPhaseRegimenSetting Primary Endpoint Status a ACOSOG Z Imatinib 400 mg/dAdjuvantOS 4-year results ACOSOG Z b 3 Imatinib 400 mg/day vs placebo AdjuvantRFS 2-year results Nilsson Imatinib 400 mg/day vs historical control AdjuvantRFS 3-year results LI J 4 105N/A d Imatinib 400 mg/day vs control (refused therapy) AdjuvantRFS 2-year results Kang B Imatinib 400 mg/day (until progression) AdjuvantRFS 2-year results EORTC Imatinib 400 mg/day vs observation AdjuvantTTSR Enrollment Completed SSGXVIII/AIO Imatinib 400 mg/day 12 vs 36 months AdjuvantRFSReported 1. DeMatteo RP et al. ASCO GI Cancers Symposium; Abstract DeMatteo RP et al. J Clin Oncol. 2005;23:818s. Abstract Nilsson B et al. Br J Cancer. 2007;96: Li J et al. J Clin Oncol. 2009;27(suppl). Abstract Kang Y et al. J Clin Oncol. 2009;27(suppl). Abstract e ClinicalTrials.gov. Accessed August 26, 2009.

ACOSOG Z9001: Trial Schema (Phase III) 778 patients Placebo (354 randomised) (345 treated) 87 discontinued treatment early Imatinib (359 randomised) (337 treated) 97 discontinued treatment early 30 events 5 GIST-unrelated deaths 713 patients randomised Phase III, randomised, double-blind, placebo-controlled multi-centre trial IM 400 mg/day or placebo for 1 yr 70 events 5 GIST-related deaths 3 GIST-unrelated deaths DeMatteo RP et al. Lancet. 2009; 373:

ACOSOG Z9001: Study Design/Methods Key Eligibility Criteria: Patients ≥18 years with localised and primary GIST KIT-positive tumours ≥3 cm Complete surgical resection Endpoints: Primary: Recurrence-free Survival (RFS) Secondary: Overall Survival (OS) and safety Other Key Elements: Dose modifications upon grade 3 or 4 events PD patients unblinded: - If placebo  IM 400 mg/day or - If IM 400 mg/day  IM 800 mg/day

Parameters Placebo (n=354) Imatinib (n=359) Tumour size, n (%) >3 and <6 cm149 (42.1%)143 (39.8%) >6 and <10 cm119 (33.6%)123 (34.3%) >10 cm86 (24.3%)93 (25.9%) Margins, n (%) R0330 (93.2%)325 (90.5%) R123 (6.5%)34 (9.5%) Unknown1 (0.3%)0 (0.0%) Tumour origin, n (%) Stomach235 (66.4%)209 (58.2%) Small intestine102 (28.8%)125 (34.8%) Rectum5 (1.4%) Other12 (3.4%)18 (5.0%) Unknown0 (0.0%)2 (0.6%) R0 – negative microscopic margins; R1 – positive microscopic margins Patient characteristics (continued)

Median follow-up: 19.7 months Estimated 1-year RFS (95% CI): Imatinib: 98% (96-100) Placebo: 83% (78-88) HR = 0.35 ( ) p < CI, confidence interval; HR, hazard ratio Events experienced: Imatinib: 8.0% (30) Placebo: 20.0% (70) Recurrence-free Survival (RFS)* *All randomised patients were included in the analysis; recurrence-free survival was defined as the time from patient registration to the development of tumour recurrence or death from any cause. Intention-to-treat analyses were done for recurrence-free survival (ie, analysed patients by randomised group).

Imatinib adjuvant therapy results in significantly longer RFS in each of the tumour size categories compared to placebo Recurrence-free Survival (Tumour size) size >10cm size >3 and <6 cm size >6cm and <10cm

No difference in OS between imatinib and placebo adjuvant therapies Overall Survival (OS)* *All randomised patients were included in the analysis; Overall survival was defined as the time from patient registration to death from any cause. Intention-to-treat analyses were done for overall survival (ie, analysed patients by randomised group).

 Imatinib at 400 mg/day is safe and well tolerated when administered as adjuvant therapy after complete resection of primary GIST  Adjuvant imatinib resulted in an improvement in RFS in patients with all tumour sizes - Especially relevant for high-risk patients (e.g. tumour size ≥10 cm or high mitotic rate) since this patient population has a 50% higher chance of recurrence at 2 years without adjuvant therapy  OS between imatinib and placebo groups comparable at this time  A longer follow-up period is likely required to observe differences  Ongoing trials in the adjuvant setting are under way to determine appropriate treatment duration of imatinib and impact on OS –SSGXVIII/AIO –EORTC Summary

Adjuvant Imatinib: Beyond 1 Year of Treatment

Imatinib 400mg/d for 12 months An open-label Phase III study Imatinib 400mg/d for 36 months Follow-up SSGXVIII: Study design Random assignment 1:1 Stratification: 1) R0 resection, no tumor rupture 2) R1 resection or tumor rupture

SSGXVIII: Objectives  Primary: RFS Time from randomization to GIST recurrence or death  Secondary objectives included: Safety Overall survival

SSGXIII: Key inclusion criteria  Histologically confirmed GIST, KIT-positive  High risk of recurrence according to the modified Consensus Criteria*: –Tumor diameter >10 cm or –Tumor mitosis count >10/50 HPF** or –Size >5 cm and mitosis count >5/50 HPFs or –Tumor rupture spontaneously or at surgery *Fletcher CD et al. Hum Pathol 2002; 33: **HPF, High Power Field of the microscope

Patient disposition Category 12 Months 36 Months No. (%) No. (%) Randomized (Feb 2004 to Sep 2009) Included in ITT Population* No GIST at pathology review 5 (3) 10 (5) - GIST metastases at study entry 13 (7) 11 (6) Included in Efficacy Population Included in Safety Population On treatment at data collection cut-off 0 (0) 19 (10) Discontinued assigned treatment 29 (15) 63 (32) - GIST recurred during treatment 4 (2) 12 (6) - Adverse event 15 (8) 27 (14) - Other reason 10 (5) 24 (12) *3 patients who withdrew consent excluded

Baseline characteristics (ITT) Characteristic 12-Mo group 36-Mo group Median age (range) - years 62 (23-84) 60 (22-81) Male - (%) ECOG performance status 0 - (%) Gastric primary tumor - (%) Median tumor size (range) - cm 9 (2-35) 10 (2-40) Median mitosis count - /50 HPFs 10 (0-250) 8 (0-165) Tumor rupture - (%) GIST gene mutation site - (%)* - KIT exon KIT exon KIT exon PDGFRA (D842V) 13 (10) 12 (8) - wild type 10 8 *Available for 366 (92%) out of the 397 tumors

SSGXVIII: Recurrence-free survival (ITT) No. at risk (n=397) 36 Months of imatinib Months of imatinib % 47.9% 86.6% 65.6% 36 Months 12 Months Hazard ratio 0.46 (95% CI, ) P < % Median follow-up time 54 months Years

Subgroup No. of patients Hazard ratio (95% CI), RFS P value Age ≤ ( ).001 > ( ).01 Sex Male ( ).002 Female ( ).002 Tumor site Stomach ( ).005 Other ( )<.001 Tumor size ≤ 10 cm ( )<.001 >10 cm ( ).002 Mitoses/50 HPF (local) ≤ 10 mitoses ( ).33 > 10 mitoses ( )<.001 Mitoses/50 HPF (central) ≤ 10 mitoses ( ).04 > 10 mitoses ( )<.001 Tumor rupture No ( )<.001 Yes ( ).02 Tumor mutation site KIT exon ( ).34 KIT exon ( )<.001 Wild type ( ).16 Other ( ) mo better 12 mo better

Clinical Risk Factors and Risk-Reduction with 3 Years of Adjuvant Imatinib Risk FactorNo. PatientsHazard Ratio (95% CI, RFS) P-Value TUMOUR SITE Gastric ( )0.006 Non-Gastric ( )<0.001 SIZE <10 cm ( )<0.001 >10 cm ( )0.002 Mitoses/50 HPF < ( )0.03 > ( )<0.001

No. at risk (n=397) 36 Months of imatinib Months of imatinib SSGXVIII: Overall survival (ITT) Hazard ratio 0.45 (95% CI, ) P = %92.0% 94.0% 81.7% 36 Months 12 Months % Years

Treatment safety Category 12-month group (n=194) No. (%) 36-month group (n=198) No. (%) P Any adverse event192 (99) 198 (100).24 Grade 3 or 4 event 39 (20) 65 (33).006 Cardiac event 8 (4) 4 (2).26 Second cancer14 (7)13 (7).84 Death, possibly imatinib-related 1* (1) 0 (0).49 Discontinued imatinib, no GIST recurrence 25 (13) 51 (26).001 *Lung injury

Most frequent adverse events Adverse eventAny GradePGrade 3 or 4P 12 Mo %36 Mo %12 Mo %36 Mo % Anemia Periorbital edema Elevated LDH* Fatigue Nausea Diarrhea Leukopenia Muscle cramps3149<

Conclusions  Compared to 1 year of adjuvant imatinib, 3 years of imatinib improves - RFS - Overall survival as treatment of GIST patients who have a high estimated risk of recurrence after surgery.  Adjuvant imatinib is relatively well tolerated; severe adverse events are infrequent.

Phase 3 Adjuvant Trial (EORTC 62024): Overview EORTC. Objectives Primary Time to secondary resistance Secondary Overall survival Relapse-free survival Relapse-free interval Drug safety Treatment Imatinib 400 mg/day for 2 years Inclusion criteria Intermediate- or high-risk GIST Completely resected KIT-positive GIST

Phase 3 Adjuvant Trial (EORTC 62024): Design Follow for 5 years after treatment to evaluate TTSR, PFS, and OS Observation (for 2 years) Imatinib (400 mg/day for 2 years) a Due to progression or unacceptable toxicity. TTSR, time to secondary resistance; PFS, progression-free survival; OS, overall survival EORTC. Complete resection of primary GIST Discontinued treatment a

Post-Resection Evaluation of Recurrence-Free Survival for Gastro-Intestinal Stromal Tumors Treated with Adjuvant Imatinib: PERSIST-5 Phase II N = 85 patients Primary objective: Recurrence-free survival Imatinib 400 mg/d x 5 years Resected GIST >2 cm and mitotic rate >5 or Non-gastric primary >5 cm Register Adapted DeMatteo

Conclusions from SSGXVIII *  In GIST, 3 years of adjuvant imatinib are better than one in terms of recurrence-free and overall survival  Three years of post-operative imatinib treatment represent the new gold standard for patients with resected “high-risk” GISTs  The overall risk at which adjuvant imatinib should be commenced requires further clarification * Adapted from discussion by Charles Blanke, ASCO 2011