T2032 Screening for Colorectal Cancer By Immunological Fecal Occult Blood Test: Results of Four...
1
AGA Abstracts T2029 Colonoscopic Evaluation of a Quantitative Immunochemical Fecal Occult Blood Test to Determine Its Optimal Clinical Use Paul Rozen, Zohar Levi, Rachel Hazazi, Amal Waked, Alex Vilkin, Eran Maoz, Shlomo Birkenfeld, Yaron Niv BACKGROUND: Colorectal screening for significant neoplasms (cancer (CRC), advanced adenomas (AAP)) aims for maximum sensitivity with minimal unrewarding follow-up colon- oscopies for positive tests. Quantified, immunochemical, hemoglobin (Hb) specific, fecal occult blood (I-FOBT) measurements are now used for screening & 1 I-FOBT, developed at 100ngHb/mL of buffer threshold, has been used. OBJECTIVES: 1. Correlate presence of CRC & AAP to amount of fecal Hb lost; 2. Evaluate fecal Hb development threshold used & number of I-FOBTs tests prepared to obtain optimal sensitivity & specificity & least number of colonoscopies needed for positive tests/neoplasm detected. METHODS: A pro- spective cross-sectional study on 1682 consecutive ambulatory colonoscopy patients, some above average risk, volunteering to prepare I-FOBTs. Fecal Hb was measured in 3 consecutive samples, analyzed from 50ngHb/mL of buffer, by the OC-MICRO™ automated instrument (Eiken Chemical Co., Tokyo, Japan), & highest result related to endoscopy findings. RESULTS: Colonoscopy identified CRC in 20 patients & AAP in 129. Receiver operator characteristics (ROC) areas under curves (AUC), using the highest of 3 I-FOBTs for CRC was 0.959 (95%CI 0.941, 0.977), for AAP or CRC was 0.793 (95%CI 0.75, 0.836). For higher of first 2 I-FOBTs, AUC for CRC was 0.957 (95%CI 0.935, 0.978) (NS), for AAP or CRC was 0.781 (95%CI 0.736, 0.825) (NS). For CRC or AAP, sensitivity was best using the highest result of all 3 I-FOBTs at 50ngHb/mL threshold, 61.1% (95%CI 53.2, 68.9), specificity 87.8% (95%CI 86.2, 89.4); positive tests would require 205 colonoscopies, 3.1/ positive test, to identify 100% of CRC & 55% of AAP. At this threshold, the higher of first 2 I-FOBTs was 11% less sensitive, 55.0% (95%CI 47.0, 63.0) (P=0.004), specificity 90.2% (95%CI 88.7, 91.7); requiring 165 colonoscopies, 2.8/positive test, to identify 95% of CRC & 48.8% of AAP. With a single test at 100ngHb/mL threshold, sensitivity was significantly lower than with 2 tests (P<0.001), 31.5% (95%CI 24.1, 39.0), specificity 96.4% (95%CI 95.5, 97.3); requiring 61 colonoscopies, 2.2/positive test, to identify 65% of CRC (P=0.031) & 26.4% AAP (P<0.001). CONCLUSIONS: Quantification of I-FOBT, test threshold chosen & number of samples collected (which affects compliance) determine sensitivity & specificity of the I-FOBT for CRC & AAP. These determine number of colonoscopies needed for positive tests & neoplasia yield. In this above-average risk population, one-time pre-colonoscopy I- FOBT screening using the higher of 2 tests developed at 50ngHb threshold, would have reduced the colonoscopy load by 90% & identify 95% of CRC & 48.8% of AAP. T2030 Prevalence and Distribution of Advanced Colorectal Neoplasia Among Whites and Blacks Undergoing Screening Colonoscopy Paul C. Schroy, Timothy Heeren Background: Pooled data from the Clinical Outcomes Research Initiative (CORI) consortium suggest that the prevalence of polyps ≥ 10 mm in size is higher among Blacks than Whites undergoing screening colonoscopy (JAMA 2008;300:1417-22). Polyp size alone, however, is an imprecise surrogate for advanced colorectal neoplasia (ACN) since 10-20% of polyps ≥ 10 mm will be non-neoplastic and 2-10% of polyps < 10 mm in size with have advanced histology. The primary objective of this study was to better define the prevalence and distribution of ACN among non-Hispanic Whites and Blacks undergoing screening colonos- copy in an open access system. Methods: We conducted a retrospective cross-sectional analysis of asymptomatic average-risk Whites and Blacks ≥ 50 years of age who underwent screening colonoscopy between January 2000 and December 2005. Patients with incomplete colonos- copies or poor bowel preparation were excluded. Unlike in the CORI study, race information was ascertained from the patient rather than the endoscopist. ACN was defined as a tubular adenoma ≥ 10 mm in size or an adenoma of any size with villous histology, high grade dysplasia, or cancer. Associations between demographic factors (age, sex and race) and ACN were examined using Chi square and multiple logistic regression analyses. Results: A total of 1481 White and 1329 Black patients met inclusion criteria. The overall prevalence of ACN was higher among Whites than Blacks (6.1% vs. 3.5%; P=0.002). Compared to Whites, the adjusted odds ratio (OR) was 0.63 (95% confidence interval [CI], 0.44-0.91) for Blacks overall, 0.53 (95% CI, 0.29-0.96) for Black women and 0.70 (95% CI, 0.45-1.12) for Black men. Although the overall prevalence of proximal ACN was similar for Whites and Blacks (2.4% vs. 2.1%; P=0.566), proximal ACNs accounted for 60% of ACN among Blacks but only 40% of ACN among Whites (adjusted OR, 2.11; 95% CI, 1.01-4.40; P=0.047). Conclu- sions: Our study finds that race is an independent determinant of ACN, but in contradistinc- tion to the CORI study, we find that the risk of ACN is higher among Whites than Blacks and that there is a predilection for proximal ACN among Blacks. We speculate that differences in the outcome of interest, selection bias, racial misclassification and/or differential exposure to modifiable risk factors may account for the disparate findings between our study and the CORI study. Regardless, our study supports the use of colonoscopy as the preferred screening option for Blacks given the predilection for proximal ACN. T2031 Determinants of the Prevalence of Synchronous Colorectal Cancer; a Dutch Population-Based Study Sanna A. Mulder, M. E. Leerdam van, R. Kranse, R. Damhuis, Ernst J. Kuipers Introduction: Reports on the prevalence and risk factors of synchronous colorectal cancers (S-CRC) vary widely, with rates ranging from 2%-7%. Part of this variation can be explained by limited sample size and by the use of different definitions. We therefore investigated the prevalence and potential risk factors of S-CRC in a large Dutch population. Methods: We performed a multicentre study using data from the Rotterdam Cancer Registry (RCR). This registry covers the southwestern part of the Netherlands, a region with 16 hospitals and 2,3 million inhabitants. All newly diagnosed CRC in this region between 1995 and 2006 were identified using the RCR database. S-CRC was defined as ≥ 2 distinct primary CRC, A-624 AGA Abstracts not due to direct extension or metastasis, diagnosed ≤ 6 months after the initial CRC diagnosis. S-CRC was called unisegmental when occurring in the same segment of colon or rectum, as specified according to the ICD-O subsite definition. Patients with carcinoma in situ or those who did not undergo surgical resection for their CRC were excluded. In patients with S-CRC, the most advanced lesion was considered as the index tumor in the analyses. ICD-O subsite was regrouped as proximal and distal colon (up to splenic flexure). Results: During the 12-year period, 13,943 patients were diagnosed with an invasive CRC. In total, 637 patients (4.5%) were diagnosed with S-CRC. In the S-CRC group, patients were significantly older (71 vs 69 yrs of age [p<0.01]) and more often male (59.5% vs 51.0% [p<0.01]) compared to patients with single CRC. In 269 patients (1.9%) the S-CRC was multisegmental and in 368 (2.6%) unisegmental. Significantly more often, both S-CRCs were located in the distal colon (47% both distal versus 26% both proximal, and 22% both a distal and proximal CRC, p< 0.01). Having at least one proximal CRC was significantly associated with female gender and increasing age (OR 2.14, 95% CI 1.53-3.00 and OR 1.03, 95% CI 1.01-1.04 resp). Overall, S-CRC was significantly associated with proximally located index tumors (OR 1.22, 95% CI 1.03-1.44), age > 70 yrs (OR 1.46, 95% CI 1.19-1.65), male gender (OR 1.49, 95% CI 1.26-1.75) and the presence of distant metastasis (OR 1.68, 95% CI 1.35-2.07). Conclusion: Approximately 5 percent of patients presenting with CRC have a synchronous tumor. Male gender, increasing age, proximal localisation of the index tumor and the presence of metastasis are significant predictors of the presence of S-CRC. The majority of S-CRC are located in the same segment as the index lesion. These results stress the importance of total colon examination when diagnosing a CRC. T2032 Screening for Colorectal Cancer By Immunological Fecal Occult Blood Test: Results of Four Rounds in Two Municipalities of Aosta Valley (Italy) Sergio Crotta, Carlo Senore, Nereo Segnan, Simona Paganin, Bruna Dagnes BACKGROUND: There is a limited experience of screening for colorectal cancer (CRC) by immunological fecal occult blood test (I-FOBT) over several rounds. We report the results of four biennial rounds using I-FOBT in two municipalities of Aosta Valley (Italy). POPULA- TION AND METHODS: Between October 2001 and September 2008 all residents, aged 50 to 74, were invited every two years to perform a 1-day latex I-FOBT (OC Sensor), with a 100 ng/ml cut-off. Invitation letters provided an appointment at a local municipal office where volunteers were involved in the kits' distribution. Patients with positive test were referred for colonoscopy. Interval cancers were identified through a linkage with the hospital discharge records. RESULTS: Out of 2840 people invited in 2001, 2111 (74.4%) performed at least 1 test, 1626 (57.3%) 2 consecutive tests, 1129 (39.8%) 3 consecutive tests and 867 (30.5%) 4 tests. The overall positivity rate among those attending at least once was 10.6% and among those attending four times was 20.1%. The CRC cumulative incidence over a 6.5 years-follow-up was 0.71% among the 2111 attenders (8 screen detected - 4 UICC stage I-II; 7 interval - 2 UICC stage I-II) and 1.51% among never attenders (11/729 - 2 UICC stage I-II CRCs). The detection rate of CRC was 0.19% at the prevalent round, 0.09%, 0.20% and 0.0% among people performing respectively 2, 3 and 4 consecutive tests. The detection rate for advanced adenomas was 0.7% and 1.5% among people who performed 3 and 4 consecutive tests respectively. CONCLUSIONS: I-FOBT still detects a high proportion of advanced adenomas after 3 negative biennial rounds. The CRC cumulative incidence was the half among the attenders than in never attenders. Out of 7 interval cancer 4 were detected among people attending only one test. T2033 Colorectal Cancer Screening in Women with Endometrial Cancer: Are We Following the Guidelines? Marvin Singh, Emily Singh, Heather Miller, Williamson Strum, Walter Coyle Introduction: Colorectal cancer (CRC) is the most common GI malignancy and the guidelines for screening have been updated recently. There is an association between CRC and endomet- rial cancer (EC). Up to 10% of this linkage may be due to HNPCC but in the majority of patients this genetic disorder is not found. The National Comprehensive Cancer Network (NCCN) guidelines on CRC screening since 2005 have suggested that women with endomet- rial or ovarian cancer diagnosed at less than 60 years of age have CRC screening with colonoscopy beginning at age 40 or at time of diagnosis of the gynecologic tumor. We chose to assess our population of women with EC to determine if women were receiving CRC screening after a diagnosis of EC. Methods: Chart review of women diagnosed at our institution with EC in the last ten years was performed. We developed a database in Microsoft Access to assess details such as age at diagnosis, tumor type, BMI, detailed family history of malignancy, CRC screening, and findings at CRC screening. We also sought to assess the Amsterdam and Bethesda criteria in these women. This study was IRB approved. Results: We reviewed the charts of 88 women who developed EC in the last ten years. The average age was 68 and 30.7% of women were less than age 60 at diagnosis of endometrial cancer. Family history of CRC was present in 5 (5.7%) of EC patients. Forty-one (46.6%) of these women had CRC screening, with 9 (21.9%) being screened for CRC within 1 year of diagnosis and 19.5% screened for CRC before diagnosis of endometrial cancer. Forty-seven (53.4%) women did not have CRC screening reported. Of the woman screened, 4 had adenomatous polyps, 1 had tubulovillous histology, and 1 woman had CRC and endometrial cancer. Screening methods included flexible sigmoidoscopy (17.1%), barium enema (14.6%), colonoscopy (53.7%), and FOBT (9.8%). None of the women met criteria for Amsterdam or Bethesda criteria. Conclusions: Our review shows that a minority of women with EC received any screening for CRC. Women who were screened had significant pathology in over 14% of cases and 2.4% had colon cancer. The NCCN guidelines should be more aggressively followed by physicians who care for women. A prospective colonoscopy screen- ing study on these women with EC to assess the yield and utility in screening in this population is needed.
Transcript of T2032 Screening for Colorectal Cancer By Immunological Fecal Occult Blood Test: Results of Four...
AG
AA
bst
ract
sT2029
Colonoscopic Evaluation of a Quantitative Immunochemical Fecal OccultBlood Test to Determine Its Optimal Clinical UsePaul Rozen, Zohar Levi, Rachel Hazazi, Amal Waked, Alex Vilkin, Eran Maoz, ShlomoBirkenfeld, Yaron Niv
BACKGROUND: Colorectal screening for significant neoplasms (cancer (CRC), advancedadenomas (AAP)) aims for maximum sensitivity with minimal unrewarding follow-up colon-oscopies for positive tests. Quantified, immunochemical, hemoglobin (Hb) specific, fecaloccult blood (I-FOBT) measurements are now used for screening & 1 I-FOBT, developedat 100ngHb/mL of buffer threshold, has been used. OBJECTIVES: 1. Correlate presence ofCRC & AAP to amount of fecal Hb lost; 2. Evaluate fecal Hb development threshold used& number of I-FOBTs tests prepared to obtain optimal sensitivity & specificity & leastnumber of colonoscopies needed for positive tests/neoplasm detected. METHODS: A pro-spective cross-sectional study on 1682 consecutive ambulatory colonoscopy patients, someabove average risk, volunteering to prepare I-FOBTs. Fecal Hb was measured in 3 consecutivesamples, analyzed from 50ngHb/mL of buffer, by the OC-MICRO™ automated instrument(Eiken Chemical Co., Tokyo, Japan), & highest result related to endoscopy findings.RESULTS: Colonoscopy identified CRC in 20 patients & AAP in 129. Receiver operatorcharacteristics (ROC) areas under curves (AUC), using the highest of 3 I-FOBTs for CRCwas 0.959 (95%CI 0.941, 0.977), for AAP or CRC was 0.793 (95%CI 0.75, 0.836). Forhigher of first 2 I-FOBTs, AUC for CRC was 0.957 (95%CI 0.935, 0.978) (NS), for AAP orCRC was 0.781 (95%CI 0.736, 0.825) (NS). For CRC or AAP, sensitivity was best usingthe highest result of all 3 I-FOBTs at 50ngHb/mL threshold, 61.1% (95%CI 53.2, 68.9),specificity 87.8% (95%CI 86.2, 89.4); positive tests would require 205 colonoscopies, 3.1/positive test, to identify 100% of CRC & 55% of AAP. At this threshold, the higher of first2 I-FOBTs was 11% less sensitive, 55.0% (95%CI 47.0, 63.0) (P=0.004), specificity 90.2%(95%CI 88.7, 91.7); requiring 165 colonoscopies, 2.8/positive test, to identify 95% of CRC& 48.8% of AAP. With a single test at 100ngHb/mL threshold, sensitivity was significantlylower than with 2 tests (P<0.001), 31.5% (95%CI 24.1, 39.0), specificity 96.4% (95%CI95.5, 97.3); requiring 61 colonoscopies, 2.2/positive test, to identify 65% of CRC (P=0.031)& 26.4% AAP (P<0.001). CONCLUSIONS: Quantification of I-FOBT, test threshold chosen& number of samples collected (which affects compliance) determine sensitivity & specificityof the I-FOBT for CRC & AAP. These determine number of colonoscopies needed for positivetests & neoplasia yield. In this above-average risk population, one-time pre-colonoscopy I-FOBT screening using the higher of 2 tests developed at 50ngHb threshold, would havereduced the colonoscopy load by 90% & identify 95% of CRC & 48.8% of AAP.
T2030
Prevalence and Distribution of Advanced Colorectal Neoplasia Among Whitesand Blacks Undergoing Screening ColonoscopyPaul C. Schroy, Timothy Heeren
Background: Pooled data from the Clinical Outcomes Research Initiative (CORI) consortiumsuggest that the prevalence of polyps ≥ 10 mm in size is higher among Blacks than Whitesundergoing screening colonoscopy (JAMA 2008;300:1417-22). Polyp size alone, however,is an imprecise surrogate for advanced colorectal neoplasia (ACN) since 10-20% of polyps≥ 10 mm will be non-neoplastic and 2-10% of polyps < 10 mm in size with have advancedhistology. The primary objective of this study was to better define the prevalence anddistribution of ACN among non-Hispanic Whites and Blacks undergoing screening colonos-copy in an open access system. Methods: We conducted a retrospective cross-sectional analysisof asymptomatic average-risk Whites and Blacks≥ 50 years of age who underwent screeningcolonoscopy between January 2000 and December 2005. Patients with incomplete colonos-copies or poor bowel preparation were excluded. Unlike in the CORI study, race informationwas ascertained from the patient rather than the endoscopist. ACN was defined as a tubularadenoma ≥ 10 mm in size or an adenoma of any size with villous histology, high gradedysplasia, or cancer. Associations between demographic factors (age, sex and race) and ACNwere examined using Chi square and multiple logistic regression analyses. Results: A totalof 1481 White and 1329 Black patients met inclusion criteria. The overall prevalence ofACN was higher among Whites than Blacks (6.1% vs. 3.5%; P=0.002). Compared to Whites,the adjusted odds ratio (OR) was 0.63 (95% confidence interval [CI], 0.44-0.91) for Blacksoverall, 0.53 (95% CI, 0.29-0.96) for Black women and 0.70 (95% CI, 0.45-1.12) for Blackmen. Although the overall prevalence of proximal ACN was similar for Whites and Blacks(2.4% vs. 2.1%; P=0.566), proximal ACNs accounted for 60% of ACN among Blacks butonly 40% of ACN among Whites (adjusted OR, 2.11; 95% CI, 1.01-4.40; P=0.047). Conclu-sions: Our study finds that race is an independent determinant of ACN, but in contradistinc-tion to the CORI study, we find that the risk of ACN is higher among Whites than Blacksand that there is a predilection for proximal ACN among Blacks. We speculate that differencesin the outcome of interest, selection bias, racial misclassification and/or differential exposureto modifiable risk factors may account for the disparate findings between our study and theCORI study. Regardless, our study supports the use of colonoscopy as the preferred screeningoption for Blacks given the predilection for proximal ACN.
T2031
Determinants of the Prevalence of Synchronous Colorectal Cancer; a DutchPopulation-Based StudySanna A. Mulder, M. E. Leerdam van, R. Kranse, R. Damhuis, Ernst J. Kuipers
Introduction: Reports on the prevalence and risk factors of synchronous colorectal cancers(S-CRC) vary widely, with rates ranging from 2%-7%. Part of this variation can be explainedby limited sample size and by the use of different definitions. We therefore investigated theprevalence and potential risk factors of S-CRC in a large Dutch population. Methods: Weperformed a multicentre study using data from the Rotterdam Cancer Registry (RCR). Thisregistry covers the southwestern part of the Netherlands, a region with 16 hospitals and2,3 million inhabitants. All newly diagnosed CRC in this region between 1995 and 2006were identified using the RCR database. S-CRC was defined as ≥ 2 distinct primary CRC,
A-624AGA Abstracts
not due to direct extension or metastasis, diagnosed ≤ 6 months after the initial CRCdiagnosis. S-CRC was called unisegmental when occurring in the same segment of colon orrectum, as specified according to the ICD-O subsite definition. Patients with carcinoma insitu or those who did not undergo surgical resection for their CRC were excluded. In patientswith S-CRC, the most advanced lesion was considered as the index tumor in the analyses.ICD-O subsite was regrouped as proximal and distal colon (up to splenic flexure). Results:During the 12-year period, 13,943 patients were diagnosed with an invasive CRC. Intotal, 637 patients (4.5%) were diagnosed with S-CRC. In the S-CRC group, patients weresignificantly older (71 vs 69 yrs of age [p<0.01]) and more often male (59.5% vs 51.0%[p<0.01]) compared to patients with single CRC. In 269 patients (1.9%) the S-CRC wasmultisegmental and in 368 (2.6%) unisegmental. Significantly more often, both S-CRCswere located in the distal colon (47% both distal versus 26% both proximal, and 22% botha distal and proximal CRC, p< 0.01). Having at least one proximal CRC was significantlyassociated with female gender and increasing age (OR 2.14, 95% CI 1.53-3.00 and OR 1.03,95% CI 1.01-1.04 resp). Overall, S-CRC was significantly associated with proximally locatedindex tumors (OR 1.22, 95% CI 1.03-1.44), age > 70 yrs (OR 1.46, 95% CI 1.19-1.65),male gender (OR 1.49, 95% CI 1.26-1.75) and the presence of distant metastasis (OR 1.68,95% CI 1.35-2.07). Conclusion: Approximately 5 percent of patients presenting with CRChave a synchronous tumor. Male gender, increasing age, proximal localisation of the indextumor and the presence of metastasis are significant predictors of the presence of S-CRC.The majority of S-CRC are located in the same segment as the index lesion. These resultsstress the importance of total colon examination when diagnosing a CRC.
T2032
Screening for Colorectal Cancer By Immunological Fecal Occult Blood Test:Results of Four Rounds in Two Municipalities of Aosta Valley (Italy)Sergio Crotta, Carlo Senore, Nereo Segnan, Simona Paganin, Bruna Dagnes
BACKGROUND: There is a limited experience of screening for colorectal cancer (CRC) byimmunological fecal occult blood test (I-FOBT) over several rounds. We report the resultsof four biennial rounds using I-FOBT in two municipalities of Aosta Valley (Italy). POPULA-TION AND METHODS: Between October 2001 and September 2008 all residents, aged 50to 74, were invited every two years to perform a 1-day latex I-FOBT (OC Sensor), with a100 ng/ml cut-off. Invitation letters provided an appointment at a local municipal officewhere volunteers were involved in the kits' distribution. Patients with positive test werereferred for colonoscopy. Interval cancers were identified through a linkage with the hospitaldischarge records. RESULTS: Out of 2840 people invited in 2001, 2111 (74.4%) performedat least 1 test, 1626 (57.3%) 2 consecutive tests, 1129 (39.8%) 3 consecutive tests and 867(30.5%) 4 tests. The overall positivity rate among those attending at least once was 10.6%and among those attending four times was 20.1%. The CRC cumulative incidence over a6.5 years-follow-up was 0.71% among the 2111 attenders (8 screen detected - 4 UICC stageI-II; 7 interval - 2 UICC stage I-II) and 1.51% among never attenders (11/729 - 2 UICCstage I-II CRCs). The detection rate of CRC was 0.19% at the prevalent round, 0.09%,0.20% and 0.0% among people performing respectively 2, 3 and 4 consecutive tests. Thedetection rate for advanced adenomas was 0.7% and 1.5% among people who performed3 and 4 consecutive tests respectively. CONCLUSIONS: I-FOBT still detects a high proportionof advanced adenomas after 3 negative biennial rounds. The CRC cumulative incidence wasthe half among the attenders than in never attenders. Out of 7 interval cancer 4 were detectedamong people attending only one test.
T2033
Colorectal Cancer Screening in Women with Endometrial Cancer: Are WeFollowing the Guidelines?Marvin Singh, Emily Singh, Heather Miller, Williamson Strum, Walter Coyle
Introduction: Colorectal cancer (CRC) is the most common GI malignancy and the guidelinesfor screening have been updated recently. There is an association between CRC and endomet-rial cancer (EC). Up to 10% of this linkage may be due to HNPCC but in the majority ofpatients this genetic disorder is not found. The National Comprehensive Cancer Network(NCCN) guidelines on CRC screening since 2005 have suggested that women with endomet-rial or ovarian cancer diagnosed at less than 60 years of age have CRC screening withcolonoscopy beginning at age 40 or at time of diagnosis of the gynecologic tumor. We choseto assess our population of women with EC to determine if women were receiving CRCscreening after a diagnosis of EC. Methods: Chart review of women diagnosed at ourinstitution with EC in the last ten years was performed. We developed a database in MicrosoftAccess to assess details such as age at diagnosis, tumor type, BMI, detailed family historyof malignancy, CRC screening, and findings at CRC screening. We also sought to assess theAmsterdam and Bethesda criteria in these women. This study was IRB approved. Results:We reviewed the charts of 88 women who developed EC in the last ten years. The averageage was 68 and 30.7% of women were less than age 60 at diagnosis of endometrial cancer.Family history of CRC was present in 5 (5.7%) of EC patients. Forty-one (46.6%) of thesewomen had CRC screening, with 9 (21.9%) being screened for CRC within 1 year ofdiagnosis and 19.5% screened for CRC before diagnosis of endometrial cancer. Forty-seven(53.4%) women did not have CRC screening reported. Of the woman screened, 4 hadadenomatous polyps, 1 had tubulovillous histology, and 1 woman had CRC and endometrialcancer. Screening methods included flexible sigmoidoscopy (17.1%), barium enema (14.6%),colonoscopy (53.7%), and FOBT (9.8%). None of the women met criteria for Amsterdamor Bethesda criteria. Conclusions: Our review shows that a minority of women with ECreceived any screening for CRC. Women who were screened had significant pathology inover 14% of cases and 2.4% had colon cancer. The NCCN guidelines should be moreaggressively followed by physicians who care for women. A prospective colonoscopy screen-ing study on these women with EC to assess the yield and utility in screening in thispopulation is needed.
