Screening for Colorectal Cancer in the Elderly Charles J. Kahi, MD, MSCR Indiana University School of Medicine Richard L. Roudebush VA Medical Center Indianapolis, Indiana ACG Regional Midwest Course Symposium Indianapolis, Indiana August 24, 2014 The Broad Perspective Age is an important risk factor for CRC: - Incidence per 100,000: 74.5 at ages 50-64 290 after age 75 - Elderly are greatest proportion of new CRC diagnoses Lukejohn et al. Am J Gastroenterol. 2011; 106 (7):1197-1206 Between 2000 and 2010, the population 65 years and over increased at a faster rate (15.1%) than the total U.S. population (9.7%) High rate of colonoscopy utilization in older age groups: - Analysis of 1.4 M reports from CORI - 13% 75 yo, 23% 70 yo - Most common indication in older patients is surveillance of polyps Lieberman et al. Gastrointest. Endosc 2014; In press. 1
Population 65 Years and Older by Size and Percent of Total Population: 1900 to 2010 (For more information on confidentiality protection, nonsampling error, and definitions, see www.census.gov /prod/cen2010/doc/sf1.pdf) Number (in millions) Percentage (of total population) Millions Percent 45 14 40 35 30 25 12 10 8 20 6 15 4 10 5 2 0 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 0 Sources: U.S. Census Bureau, decennial census of population, 1900 to 2000; 2010 Census Summary File 1. Expectation of Life at Birth Year Total Male Female 1970 70.8 67.1 74.7 1980 73.7 70.0 77.4 1990 75.4 71.8 78.8 2000 76.8 74.1 79.3 2008 78.0 75.5 80.5 2015 (proj) 78.9 76.4 81.4 2020 (proj) 79.5 77.1 81.9 U.S. Census Bureau, Statistical Abstract of the United States: 2012 2
(In)appropriate Colonoscopy in the Elderly 75,000 Medicare beneficiaries in Texas - Inappropriate colonoscopy based on age or occurrence too soon after colonoscopy with negative findings - Rate about 40% for 76-85, 25% for 86 year-olds Scheffield et al. JAMA Intern Med 2013; 173: 542-50 24,000 Medicare patients with negative screening colonoscopy - 23.5% underwent colonoscopy within 7 years without clear indication - Higher risk if male, more comorbidity, high-volume endoscopist Goodwin et al. Arch Intern Med 2011; 171: 1335-43. 13,000 Medicare patients after colonoscopy with polypectomy - 45.7% underwent surveillance colonoscopy at 5 years Cooper et al. Cancer 2013; 119 (10): 1800-7. USPSTF GUIDELINES Continued screening in 75-year-old persons after consecutive negative screenings singe age 50 is of little benefit For individuals older than age 85, competing causes of mortality preclude a mortality benefit that outweighs the harms For adults (75-85) who have not been previously screened, decisions about first-time screening in this age group should be made in the context of the individual s health status and competing risks Zauber et al. Annals Int Med 2008; 149(4): 659-69. 3
When 900 years you will reach, look as good, you will not. - Jedi Master Yoda. THE ELDERLY ARE A HETEROGENEOUS GROUP, AND THERE IS NO SINGLE AGE THRESHOLD BEYOND WHICH SCREENING WILL SUDDENLY LOSE ITS BENEFITS, OR BECOME MORE HARMFUL, EQUALLY FOR ALL PATIENTS Walter et al. Am J Med. 2005; 118 (10): 1078-86. 4
FACTORS INVOLVED Patient chronological age and life expectancy Comorbidity, functional status Magnitude of protective effect of screening Natural history of colon neoplasms Individual risk for colonic neoplasia Prior screening/surveillance history Procedure-related related harm Patient preference and beliefs Physician preference and practice pattern. YIELD STUDIES High prevalence of CRC and advanced polyps Yield highest for patients with symptoms Low complication rates Many conclude (incorrectly) that high prevalence of advanced colon neoplasms is reason enough to screen Sardinha et al. Int J Colorectal Dis. 1999; 14(3): 172-6 Arora et al. Gastrointest. Endosc. 2004; 60(3): 408-13 Feingold et al. Am J Surg. 2003; 185 (4): 297-300 Stevens et al. Am J Gastroenterol. 2003; 98 (8):1881-85 Ure et al. Surg Endosc. 1995; 9 (5): 505-8 Duncan et al. Dis Colon Rectum. 2006; 49(5): 646-51. 5
COLONOSCOPY RISK IN THE ELDERLY Not as straightforward as in younger patients - Higher rate of poor bowel preparation - Procedure takes longer to complete - Risk of incomplete/aborted procedures is higher Froehlich et al. Gastrointest Endosc. 2005; 61 (3): 378-84. Increased major complication risk - 53,220 Medicare beneficiaries age 66-95, outpatient colonoscopy (matched beneficiaries without colonoscopy) - Higher complication rate, increased with age, comorbidity - Risk per 1,000 procedures: 0.6 for perforation, 8.7 for postpolypectomy hemorrhage, 19.4 for cardiovascular events Warren et al. Ann Intern Med 2009; 150(12): 849-57. LIFE EXPECTANCY AND SCREENING OUTCOMES Screening for cancer aims to prolong life through prevention or early detection Framework based on quantitative estimates of life expectancy, risk of CRC death, screening outcomes, harm: Significant variation in survival benefit for patients with similar ages but varying life expectancy. Walter and Covinsky. JAMA 2001; 285(21): 2750-6 Meta-analysis of 4 FOBT screening RCTs: - Average 10.3 years (95% CI 6.0-16.4) before one CRC death was prevented for 1000 patients screened Lee et al. BMJ 2012; 345:e8441. 6
Cumulative Colorectal-Cancer Mortality. Long-Term Mortality after Screening for Colorectal Cancer Shaukat A et al. N Engl J Med 2013;369:1106-14. AGE OF SCREENING CESSATION BASED ON MODELS OF HARM AND BENEFIT Microsimulation models to estimate harms and benefits of having one cancer screen in regularly screened cohorts aged 66 to 90 years by comorbid condition level CRC screening using FIT Harms and benefits compared with those of average-health cohort screened at age 74 Harms and benefits in comparison 74 yo cohort similar to: - 76 yo with no comorbidity - 72 yo with moderate comorbidity - 66 yo with severe comorbidity Lansdorp-Vogelaar et al. Ann Int Med 2014; 161: 104-12. 7
COST-EFFECTIVENESS Model with previously unscreened average-risk persons 76-90 yo CRC screening using FIT, sigmoidoscopy, colonoscopy At threshold of $100,000 per QALY, CRC screening was cost-effective in unscreened elders until age: Colonoscopy Sigmoidoscopy FIT No comorbidity Severe comorbidity 83 84 86 77 78 80 van Hees et al. Ann Int Med 2014;160:750-9. BEYOND MODELS: REAL WORLD CLINICAL IMPLICATIONS Relationship between the prevalence of colorectal neoplasms and impact of screening on life expectancy, and long-term outcome after screening Cross-sectional study conducted among 1244 asymptomatic individuals who underwent screening colonoscopy Age Group N Prevalence of Neoplasia Extension of life expectancy (Mean, SD) 50-54 1034 13.8% 0.85 (3.40) 75-79 147 26.5% 0.17 (0.49) 80 63 28.6% 0.13 (0.30) Lin et al. JAMA 2006; 295 (20): 2357-65. 8
Survival of Elderly Persons after Colonoscopy Retrospective cohort of 404 Veterans aged 75 - Total deaths: 167 (41%). - Mean survival: 4.1 ± 0.1 years. - Most common cause of death: CV (35%) Mortality predicted by: - Age (HR 1.16 for each year increase beyond age 75) - Charlson score (HR 8.3 for each point increase) - Colonoscopy indication and advanced adenoma NOT predictive Median survival of patients 75-7979 was > 5 years if Charlson 4 Among patients 80, median survival was < 5 years regardless of Charlson score. Kahi et al. Gastrointest. Endosc 2007; 66 (3): 544-50. QUANTIFYING THE IMPACT OF COLONOSCOPY ON CRC IN THE ELDERLY Effectiveness of lower GI endoscopic procedures to prevent CRC in geriatric populations: Understudied Complicating issue: Colonoscopy associated with decreased protection against right-sided CRC Case-control study, VA-Medicare, patients aged 75-623 cases with CRC, 1869 controls without CRC - Exposure to lower GI endoscopy associated with 42% CRC reduction (aor 058 0.58, 95%CI0480 0.48-0.69) - Colonoscopy associated with significant reductions in distal CRC (aor 0.45, 0.32-0.62) and proximal CRC (aor 0.65, 0.46-0.92) Kahi et al. Gastroenterology 2014; 146(3): 718-25. 9
Screening/Surveillance Decisions Assessment of life expectancy by clinicians should be main driver of CRC screening/surveillance decisions Healthy older patients are not receiving screening while older patients with significant comorbidity are receiving it regardless: - VA-Medicare study of 27,068 Veterans 70 years - Rate of screening: No comorbidity 5-year mortality = 19% 47% Severe comorbidity 5-year mortality = 55% 41% Walter et al. Ann Internal Med 2009; 150(7): 465-73. Screening/Surveillance Decisions Surveys of providers regarding CRC screening practices: - Providers incorporate patient age, comorbidity, life expectancy, and prior screening history when making decisions - Significant prevalence of inappropriate screening Kahi et al. J Gen Int Med 2009; 24(12): 1263-8 Cooper et al. Arch Int Med 1997; 157(17): 1946-50 15-21% would screen a 75-year-old with active non-colon malignancy, severe CHF, or severe COPD Kahi et al. J Gen Int Med 2009; 24(12): 1263-8. 10
Screening/Surveillance Decisions Reasons for inappropriate screening unclear, but probably multifactorial: - Organizational pressures (performance measures) - Fear of litigation - Oversimplification of benefits of CRC screening - Erroneous impression that CRC screening benefits are universal - Reliance on heuristics, intuitive decision-making Kahi et al. J Gen Int Med 2009; 24(12): 1263-8 Hoffman and Walter. J Gen Int Med 2009; 24(12): 1336-7. Screening/Surveillance Decisions: Practical Guide Use USPSTF guidelines as basic framework Communicate with patient t and PCP Base decisions on health-adjusted estimate of life expectancy, NOT just chronological age If LE < 5-7 years, stop! Compare LE to recommended colonoscopy surveillance interval: Stop if LE < surveillance interval. 11
Screening/Surveillance Decisions Based on LE Women Men 21.3 18 17 15.7 13 11.9 9.5 9.6 8.6 6.8 5.9 6.8 46 4.6 3.9 2.9 1.8 70 75 80 85 90 Top 25th percentile 50th percentile Lowest 25th percentile 14.2 12.4 10.8 9.3 7.9 6.7 6.7 4.9 4.7 3.3 2.2 5.8 3.2 1.5 70 75 80 85 90 Top 25th Percentile 50th Percentile Lowest 25th Percentile Walter and Covinsky. JAMA 2001; 285(21): 2750-6. 12