The presence of 18q loss of heterozygosity (LOH) predicts decreased disease-free and overall survival in stage II colon cancer: A study of CALGB Protocol 9581 CALG B 9581: P hase IIIstudy ofm onoclonalantibody 17-1A versus observation follow ing surgery forstage IIcolon cancer Thom as Colacchio,D artm outh-H itchcock,Treatm entStudy PI R obertW arren,C orrelative Science Study PI Donna N iedzwiecki,D onna H ollis,Study Statisticians Accrual began August1997,ended M ay 2002 1738 enrolled;m edian follow -up 7.4 years Treatm enttrial result: no difference in 5-yearD FS orO S betw een the tw o study arm s Colacchio,etal,m anuscriptin prep random ize Surgical resection ofpT3N 0 or pT4bN 0 colon cancer M oAb 17-1A 500 m g IV initial treatm entthen 100 m g IV q28 days x 4 Observation M.M. Bertagnolli 1,4 , D. Niedzwiecki 2 , M. Hall 2 , S.D. Jewell 3 , R.J. Mayer 4 , R.M. Goldberg 5 , T.A. Colacchio 6 , R. S. Warren 7 , M. Redston 1 1 Brigham and Women’s Hospital, Boston, MA; 2 CALGB Statistical Center, Duke University Durham, NC; 3 CALGB Pathology Coordinating Office, Ohio State University, Columbus, OH; 4 Dana Farber Cancer Institute, Boston, MA, 5 University of North Carolina-Chapel Hill, Chapel Hill, NC; 6 Dartmouth-Hitchcock Medical Center, Dartmouth, NH; 7 University of California-San Francisco, San Francisco, CA Overall treatment trial result: no difference in 5-year DFS or OS between the two study arms Colacchio, et al, manuscript in prep 0 2 4 6 8 10 Years from S tudy E ntry 0.0 0.2 0.4 0.6 0.8 1.0 D isease-free Survival Probability 0 2 4 6 8 10 0.0 0.2 0.4 0.6 0.8 1.0 AdjuvantM oAb 17-1A Observation N = 865 N = 873 Events= 210 Events= 215 M edian= NA M edian= NA Chi-square= p-value= 0 0.96 D FS by treatm entfor the entire 9581 cohort (n=1,738). O S by treatm entfor the entire 9581 cohort (n=1,738). 0 2 4 6 8 10 Years from Study Entry 0.0 0.2 0.4 0.6 0.8 1.0 O verall Survival Probability 0 2 4 6 8 0.0 0.2 0.4 0.6 0.8 1.0 AdjuvantM oAb 17-1A Observation N = 865 N = 873 E vents= 157 E vents= 173 M edian= N A M edian= N A Chi-square= p-value= 0.6 0.44 35 . 1 / / 66 . 0 2 # 1 # 2 # 1 # PeakArea PeakArea PeakArea PeakArea N N T T Scoring Algorithm For any marker: If the ratio was ≥0.66 and ≤1.35, then the marker indicated that heterozygosity was present (18q intact) If the ratio is outside this range, then the marker indicated LOH (LOH present) Case calls: If LOH was present at any of the 5 markers, then the patient’s tumor was called LOH+ (LOH present) The case was called s non-informative for 18q if either the non-tumor sample was not heterozygous for the marker or if DNA amplification failure occurred CALGB 9581 Tissue Bank Tumor and normal tissue obtained from CALGB Pathology Coordinating Office Pathology review, dissection to achieve optimal tumor and normal samples, then DNA extraction by standard methods Tumors without either MSI-H or loss of expression of MLH1 or MSH2 genotyped to detect 18qLOH using theD18S55 marker Multiplex PCR using fluorescent dye labeled primers; analysis using 310 Genetic Analyzer with GeneScan/Genotyper software Results Chromosomal location 18q is a common site of loss of heterozygosity in colon cancer 18q contains several genes implicated in tumor progression including: SMAD4 – encodes a nuclear transcription factor mediating TGF- signaling SMAD22 – encodes a gene associated with endodermal differentiation DCC – a site encoding a netrin-1 receptor; loss of expression of DCC protein by immunohistochemistry was associated with poor prognosis in stage II and III colon cancers (Shibata, et al) A retrospective analysis of tissues from randomized phase III cooperative group trials showed that 18qLOH was associated with significantly worse survival in high risk stage II and III colon cancer patients treated with 5- fluorouracil-based chemotherapy (Watanabe et al). No studies to date have prospectively validated the utility of 18qLOH as a prognostic marker for early stage colon cancer. Background Saltz LB, Niedzwiecki D, Hollis D, Goldberg RM, Hantel A, Thomas JP, Fields AL, Mayer RJ. Irinotecan fluorouracil plus leucovorin is not superior to fluorouracil plus leucovorin alone as adjuvant treatment for stage II colon cancer: results of CALGB 89803. J Clin Oncol 2007; 25:3456. Shibata D, Reale MA, Lavin P, Silverman M. Fearon ER, Steele G, Jessup JM, Loda M,Summerhayes IC. The DCC protein and prognosis in colorectal cancer. N Engl J Med 1996; 335:1727. Watanabe T, Wu TT, Catalano PJ, Ueki T, Satriano R, Haller DG, Benson AB 3 rd , Hamilton SR. Molecular patterns of survival after adjuvant chemotherapy for colon cancer. N Engl J Med 2001; 344:1196. References Study Cohort Treatment Trial Result Primary Objective: To evaluate a panel of prognostic markers in order to determine the relationship between pre-treatment tumor marker status, overall clinical outcome, and response to MoAb 17-1A Hypothesis: Patients whose tumors demonstrate 18qLOH will have poorer study outcomes (DFS and OS) than those whose tumors do not demonstrate 18qLOH Other markers tested: Microsatellite instability by genotyping and MLH1, MSH2 immunohistochemistry, Crohn’s-like peritumoral inflammatory cell infiltrate The original study design called for analysis of tumor DCC expression, which was ultimately not completed due to lack of DCC antibody availability Correlative Science Protocol Conclusions D18S55 Marker N= 537 stage II colon cancers: 18q Intact N on-inform ative or insufficient D N A for testing 18q LO H M SI-H, M M R-D DFS: OS: Characteristic Allpatients Patients w ith 18q intact tumors Patients w ith 18qLO H tumors p= N um berofpatients in category 1738 49 101 Treatm ent– no.(% ) 0.388 MoAb17-1A 865 (49.8) 26 (53.0) 46 (45.5) Observation 873 (50.2) 23 (47.0) 55 (54.5) A ge – yrm edian (range) 66 (24-90) 64 (37-89) 68 (33-86) 0.163 G ender– no.(% ) 0.149 Male 901 (51.8) 22 (44.9) 58 (57.4) Female 837 (48.2) 27 (55.1) 43 (42.6) Site oftum or– no.(% ) 1 0.054 Proximal 1048 (60.3) 31 (63.2) 47 (46.5) Distal 678 (39.0) 18 (36.7) 54 (53.5) U nknow n 12 (0.7) 0 0 Tum ordifferentiation – no.(% ) 0.134 Well 146 (8.4) 7 (14.2) 8 (7.9) Moderate 1305 (75.1) 35 (71.6) 86 (85.2) Poor/undifferentiated 268 (15.4) 7 (14.2) 7 (6.9) U nknow n 19 (1.1) 0 (0) 0 (0) Analysis of tumor microsatellite instability status by either genotyping using Bethesda markers to detect high levels of microsatellite instability (termed MSI- H) or immunohistochemistry to detect loss of expression of MLH1 or MSH2 (termed mismatch repair protein deficient, or MMR-D) C ases analyzed for m ism atch repairstatus N =537 M M R-Icases N =413 5-yrD FS=0.79 5-yrOS=0.87 5-yrDFS=0.78 5-yrOS=0.85 5-yrD FS=0.73 5-yrOS=0.82 5-yrD FS=0.92 5-yrOS=0.98 5-yrD FS=0.77 5-yrOS=0.89 5-yrD FS=0.85 5-yrOS=0.91 5-yrD FS=0.80 5-yrOS=0.86 p=0.01,0.04 p=0.03,0.01 M M R -D cases N =124 C ases analyzed for 18qLO H N =239 Non-inform ative N =89 18qLO H + N =101 18q intact N =49 C ases not furthertested N =174 0 2 4 6 8 10 Years from Study Entry 0.0 0.2 0.4 0.6 0.8 1.0 Disease-freeSurvival Probability 0 2 4 6 8 10 0.0 0.2 0.4 0.6 0.8 1.0 LO H18q-Neg LO H18q-Pos N = 49 N = 101 Events= 5 Events= 28 M edian= NA M edian= NA Chi-square= p-value= 4.78 0.029 0 2 4 6 8 10 Years from Study Entry 0.0 0.2 0.4 0.6 0.8 1.0 Overall Survival Probability 0 2 4 6 8 0.0 0.2 0.4 0.6 0.8 1.0 LO H 18q-Neg LO h18q-Pos N = 49 N = 101 Events= 2 Events= 23 M edian= NA M edian= NA Chi-square= p-value= 7.17 0.007 D18S55 Marker N= 537 stage II colon cancers: 18q Intact N on-inform ative orinsufficient D N A for testing 18q LO H M SI-H, M M R-D DFS: OS: C haracteristic Allpatients Patients w ith 18q intact tumors Patients w ith 18qLO H tumors p= N um berofpatients in category 1738 49 101 Treatm ent– no.(% ) 0.388 M oAb17-1A 865 (49.8) 26 (53.0) 46 (45.5) Observation 873 (50.2) 23 (47.0) 55 (54.5) A ge – yrm edian (range) 66 (24-90) 64 (37-89) 68 (33-86) 0.163 G ender– no.(% ) 0.149 Male 901 (51.8) 22 (44.9) 58 (57.4) Female 837 (48.2) 27 (55.1) 43 (42.6) S ite oftum or– no.(% ) 1 0.054 Proximal 1048 (60.3) 31 (63.2) 47 (46.5) D istal 678 (39.0) 18 (36.7) 54 (53.5) U nknow n 12 (0.7) 0 0 Tum ordifferentiation – no.(% ) 0.134 W ell 146 (8.4) 7 (14.2) 8 (7.9) Moderate 1305 (75.1) 35 (71.6) 86 (85.2) P oor/undifferentiated 268 (15.4) 7 (14.2) 7 (6.9) U nknow n 19 (1.1) 0 (0) 0 (0) C ases analyzed for m ism atch repairstatus N =537 M M R-Icases N =413 5-yrD FS=0.79 5-yrO S=0.87 5-yrD FS=0.78 5-yrO S=0.85 5-yrD FS=0.73 5-yrO S=0.82 5-yrD FS=0.92 5-yrO S =0.98 5-yrD FS=0.77 5-yrO S =0.89 5-yrD FS=0.85 5-yrO S=0.91 5-yr D FS=0.80 5-yr O S=0.86 p=0.01,0.04 p=0.03,0.01 M M R -D cases N =124 C ases analyzed for18qLO H N =239 N on-inform ative N =89 18qLO H + N =101 18q intact N =49 C ases not furthertested N =174 0 2 4 6 8 10 Years from Study Entry 0.0 0.2 0.4 0.6 0.8 1.0 Disease-freeSurvival Probability 0 2 4 6 8 10 0.0 0.2 0.4 0.6 0.8 1.0 LO H18q-Neg LO H18q-Pos N = 49 N = 101 Events= 5 Events= 28 M edian= NA M edian= NA Chi-square= p-value= 4.78 0.029 0 2 4 6 8 10 Years from Study Entry 0.0 0.2 0.4 0.6 0.8 1.0 Overall Survival Probability 0 2 4 6 8 0.0 0.2 0.4 0.6 0.8 1.0 LO H18q-Neg LO h18q-Pos N = 49 N = 101 Events= 2 Events= 23 M edian= NA M edian= NA Chi-square= p-value= 7.17 0.007 Preliminary results using D18S55 as a single marker support a role for 18qLOH as a prognostic marker in stage II colon cancer The high rate of non-informative cases in this series likely results from use of a single 18q marker Interrogation of 4 additional 18q markers, including D18S58, D18S61, D18S64, and D18S69, is underway. Tumor Analysis Method