Promoting cancer screening among first degree relatives of colon cancer patients

A442 AGA ABSTRACTS

2427 PRELIMINARY RESULTS OF THE COLORECTAL NEOPLASIA SCREENING IN ASYMPTOMATIC WOMEN AT REGIONAL NAVY/ARMY MEDICAL CENTERS TRIAL: THE CONCERN TRIAL. Brooks D. Cash, Philip S. Schoenfeld, James A. Butler, James W. Kikendall, Javaid A. Shad, William Schindler, David A. Lieberman, National Naval Med Ctr, Bethesda, MD; Walter Reed Army Med Ctr, Washington, DC; Naval Med Ctr San Diego, San Diego, CA; Portland VA Med Ctr, Portland, OR. BACKGROUND: The VA Cooperative Study 380 used screening colonoscopy to determine that approximately 10% of asymptomatic veterans (97% male) referred for colorectal cancer screening had advanced adenomas (i.e., adenoma ~ 10 mrn, villous adenoma, or adenoma with high grade dysplasia). Approximately 20-40% of patients with advanced adenomas had no lesions within reach of the sigmoidoscope, depending on insertion depth of the scope. However, the efficacy of screening colonoscopy has not been adequately examined in women. This trial duplicates the VA Cooperative Study 380 in a female population. METHODS: Prior to undergoing screening colonoscopy, study patients will complete questionnaires about multiple risk factors for adenomas including age > 65 years old. africanamerican race. alcohol use (> 2 ounces/day), tobacco use (>20 pack years), NSAID use (>2 doses/week), and family history of colon cancer. Inclusion criteria: consecutive asymptomatic women referred for colorectal cancer screening. Exclusion criteria: colonoscopy or barium enema within the last 10 years, FS within the last 5 years, hematochezia within the last year, iron deficiency anemia, or positive stool guiaics. All patients undergo screening with CBC and FOBT prior to entry into the trial. Polyp diameter is measured in vivo with a colonoscopic measuring guidewire. Association between risk factors and adenomas are examined with univariate and step-wise multivariate logistic regression analysis. RESULTS: 94.3% (215/ 228) of consecutive women referred for colorectal cancer screening underwent colonoscopy. Demographic data: mean age-58.7 years old (SD= +/- 8), percentage with family history of colon cancer-15.8%, percentage of african american descent-7.9%. 18.6% (40/215) had adenomas found during colonscopy. 4.65% (10/215) had advanced adenomas found during colonoscopy. Among patients with advanced adenomas, 40.0% (4/10) had no adenomas within reach of the FS. In multiple logistic regression analysis, age >65 years old (OR=4.116, 95% CI:1.860 to 9.162) was uniquely associated with the presence of adenomas and alcohol use approached statistical significance as an independent risk factor for adenomas (OR=4.055, 95% CI: 0.726 to 20.439). CONCLUSIONS: Based on preliminary data, colonoscopy may be preferable for colorectal cancer screening in asymptomatic women.

2428 PROMOTING CANCER SCREENING AMONG FIRST DEGREE RELATIVES OF COLON CANCER PATIENTS. Dennis J. Ahnen, Steve Hines, Chrisitne Vanoni, Ned Colange, Pj Flynn, Robert Bresalier, George Triadafilopoulos, Steven Sontag, Hans Berkel, Judith F. Collins, Douglas Nelson, Edmund Bini, Univ of Colorado, Denver, CO; AMC Cancer Research Ctr, Denver, CO; Kaiser Permanente, Denver, CO; Metro Minnesota CCOP, St. Louis Park, MN; Henry Ford Cancer Ctr, Detroit, MI; Stanford Univ, Palo Alto, CA; Dept of Veterans Affairs, Hines, IL; Univ of Louisiana, Shreveport, LA; Dept of Veterans Affairs, Portland, OR; Univ of Minnesota, Minneapolis, MN; Dept of Veterans Affairs, New York, NY. First degree relatives (FDRs) of patients with colorectal cancer are at increased risk of the disease, and it is generally recommended that they have colorectal cancer screening beginning at age 40. Purpose-To design, implement and evaluate a brief, telephone-based, educational and counseling intervention to promote colon cancer screening among FDRs of colorectal cancer patients. Methods-The study was approved by the IRBs of the University of Colorado and all participating sites. Physicians from over 30 hospitals in the U.S. gave permission to contact their patients with colorectal cancer. After patient informed consent was obtained, a telephone interview was conducted to obtain a roster of eligible FDRs (i.e, males and females, 40+ years of age, English-speaking. aware of cancer diagnosis of index case, no personal history of cancer) . Families with ~ I eligible FOR were randomized to intervention or control groups (unit of randomization = the family). Consenting FDRs from both groups received an introductory letter, brochure, and a core baseline computer assisted telephone interview (CATI) that measured colon cancer attitudes, beliefs, risk perceptions and screening practices. The intervention group received a brief, tailored, educational, motivational,telephone barriers counseling module designed to promote colorectal cancer screening, and a tailored, reinforcement mail-out. Both groups had CATI interviews at 3 and 12 months follow-up. Results- 860 colorectal cancer patient interviews were completed and yielded 2069 FDRs, of whom 1266 completed the baseline interview (accrual closed in August 1999). Thus far, 969 FDRs have completed the 3-month interview, and 599 have completed the 12-month interview. The study population is predominantly white (84%), well educated (67% with some college or more) and affluent (63%>$40,000 annual income). At baseline, 45% of all participants reported colorectal screening within the recommended guidelines (FOBT within I year or endoscopic screening within the past 5 years). At the most recent analysis, the rate of adherence to current screening guidelines had increased to 72% in the intervention group and 57% in the control group (p
GASTROENTEROLOGY Vol. 118, No.4

are related to screening behavior. Conclusion- A brief, telephone-based, educational and counseling intervention significantly improved adherence to colorectal cancer screening recommendations in FORs of colorectal cancer patients.

2429 COLORECTALERT: A NOVEL TEST FOR COLORECTAL CANCER SCREENING. Norman E. Marcon, Mike Evelegh, Christine F. Ross, Kujtim Bodinaku, Laila A. Gurney, Gregory B. Haber, Paul P. Kortan, St Michael's Hosp Wellesley Site, Toronto, ON, Canada; Inti Med Innovations, Toronto, ON, Canada. Purpose: The purpose of this study was to evaluate ColorectAlert(CRA) as a screening tool for the early detection of colorectal cancer. CRA quantitatively determines the presence of D-galactose-f3 -[1--c>3]-N-acetyl-Dgalactosamine (GAG) in a rectal mucus sample. GAG is a carbohydrate moiety that has been found to be associated with adenocarcinomas. CRA makes use of a novel color measurement to quantitate the amount of GAG present in a rectal mucus sample. Methods: Mucus samples, obtained during digital rectal examination, were smeared onto CRA membranes and sent to the laboratory for testing. The CRA test results were determined by treating the mucus samples with galactose oxidase and then staining with Schiff's reagent. The results were quantitated by determining the color of the developed sample using a hand-held spectrophotometer. The CRA samples were then treated with periodate to oxidise all carbohydrate residues and restained with Schiff's reagent prior to a second color measurement. Test results were considered positive if the measured color (hue angle) was >370 and negative if less than 350. Samples with a value of 350-370 and a value after periodate treatment of <350 were also considered positive. In a prospective clinical trial we compared CRA to fecal occult blood testing (FOBT) for the detection of colorectal cancer in 601 individuals scheduled for colonoscopy. Patients participating in the study provided informed consent. Stool samples from 3 consecutive bowel movements were obtained following the patient instructions for FOBT (Hemoccult SENSA) and a rectal mucus sample was obtained prior to colonoscopy examination. FOBT and CRA testing were done without knowledge of the colonoscopy outcome. Results: The study population had a mean age of 59 years, was 53% male, and 93% caucasian. 94% of colonoscopies were completed to the cecum or terminal ileum. Based on colonoscopy: 40% of the study population had a normal bowel, 20% had benign bowel disease, 34% had polyps, and 4% had cancer. FOBT and CRA were equally sensitive for cancer detection(8I.3%, 13/16 cancers) but CRA was significantly more specific than FOBT in this study population (75.1% vs 56.5%, X2 p 1.0cm. Conclusion: This novel method, because of its improved specificity, could reduce the cost of screening programs (ie. fewer colonoscopies) without compromising sensitivity.

2430 DOWN REGULATION OF BAK EXPRESSION IN NORMAL ENTEROCYTES BY ANTI-SENSE CONSTRUCT RESULTS IN MALIGNANT TRANSFORMATION. Diana Kaznov, Baruch S. Stern, Marjorie Pick, Varda R. Deutsch, Eli Brazowski, Itzhak Shapira, Ina Fabian, Walter Pyerin, Nadir Arber, Med Ctr Tel-Aviv, Tel-Aviv, Israel; Tel-Aviv Med ce, Tel-Aviv, Israel; TelAviv Univ, Tel-Aviv, Israel; Dkfz, Heidelberg, Germany. Background: In previous studies we found that transfecting a normal enterocyte cell line (IECI8) with an activated K-ras oncogene caused malignant transformation (RI cells) (Arber et al Oncogene 12, 1903-8, 1996). These cells were more resistant to the induction of apoptosis by sulindac sulfide, by downregulating the expression of the pro-apoptotic bak gene (Arber et al Gastroenterology 113, 1892-1900, 1997). Hypothesis: Inhibition of bak expression in IEC18 cells will result in malignant transformation. Material and Methods: Human bak eDNA was cloned into the vector pMVI2 in the anti-sense orientation. IECI8 cells were transfected with the construct, which also encodes for the hygromycin drug resistance selectable marker. 3 resistant clones displaying decreased levels of the BAK protein (similar to those found in RI cells) were isolated from different plates and further characterized. Results: In contrast to the parental IECI8 cells, AS-bak clones displayed a transformed morphology, increased saturation density, plating efficiency, anchorage independent growth in soft agar, aneuploid karyotype and a decrease in the level of alkaline phosphatase. Most importantly they were tumorigenic when injected into nude mice which was confirmed by histology. These clones displayed increased expression of cyclin DI (4-8 folds). Rb, p27 and COX-2 (2-3 folds). There was no change in the level of COX I, total b-catenin and other cell cycle proteins. The clones were more resistant to apoptosis induced by sulindac sulfide, sulindac sulfone (Cell Pathway, USA), yet were more sensitive to nimesulide (Rafa Inc, Israel) a specific COX 2 inhibitor. All cell types were very sensitive to induction of apoptosis by the newer cGMP-PDE inhibitors, CP248 and CP461(Cell Pathway, USA) most probably by a Gz/M block. Conclusions: I. Down regulation of bak expression in normal enterocytes resulted in malignant transformation 2. The AS-bak clones were more resistant to sulindac sulfide and sulfone, and more sensitive to nimesulide. 3. The new cGMPPDE inhibitors were shown to be bak independent and induced a Gz/M arrest. 4. Bak down-regulation caused increased cyclinDI, Rb and COX-2