The study, which appears in the November American Journal of Roentgenology, supports the practice of three-year surveillance with virtual colonoscopy (VC or CT colonography).
Colonoscopic referral for polyps 5-mm and smaller is ineffective due to the low likelihood of advanced neoplasia and high cost of removal; however, the management of 6- to 9-mm lesions remains controversial, in part due to a perceived lack of data regarding the natural history of these lesions, wrote Dr. Perry Pickhardt and Dr. David Kim from the University of Wisconsin, who worked with colleagues at several other institutions.
"Although the current clinical practice is generally to offer polypectomy for small (6- to 9-mm) polyps detected at [CT colonography (CTC)], many patients will choose short-term CTC surveillance if presented with the option," they wrote. "To date, no detrimental effect has been shown from longitudinal follow-up of small colorectal polyps in previous endoscopic and barium enema studies, suggesting that this may be a reasonable clinical approach. In fact, most small polyps will either remain stable or decrease in size over time, which agrees with our preliminary experience with CTC surveillance of 6- to 9-mm polyps" (AJR, November 2008, Vol. 191:15, pp. 1509-1516).
The University of Wisconsin team -- along with Dr. Andrea Laghi and Dr. Franco Iafrate from the University of Rome La Sapienza, and Dr. Cesare Hassan, Dr. Angelo Zullo, and Dr. Sergio Morini from Santa Margharita Hospital, also in Rome -- created a decision analysis model to evaluate available data with regard to the expected clinical course of small polyps over a period of three years.
The model analyzed VC surveillance, immediate colonoscopy with polypectomy, or neither approach in a hypothetical population of 100,000 60-year-old asymptomatic adults with 6- to 9-mm polyps detected at CTC screening. It accounts for the expected prevalence of advanced neoplasia and the frequency of growth, the prevalence and risk of colorectal cancer, and costs related to screening and treatment of colorectal cancer.
Colorectal cancer risk was assumed to be independent of adenoma size, which intentionally overstates the risk of cancer, the authors noted. The baseline assumption of a 1.0% annual transition rate from a small advanced adenoma to colorectal cancer was also used to intentionally overstate the risk.
Notably, the model applied results from another project, Pickhardt and colleagues' ongoing study on the natural history of polyps, which demonstrate a baseline frequency of 10% of 6- to 9-mm polyps showing measurable growth at CTC follow-up in such a cohort. In line with clinical experience and available literature, all histologically advanced lesions were assumed to show interval growth.
"To define the risk associated with 6- to 9-mm polyps, we related polyp prevalence by age to the risk of developing cancer within a 5-year time interval," the group wrote. "We assumed that all potentially preventable [colorectal cancers (CRCs)] arose from advanced adenomas, which are defined by histologic features of a prominent villous component or high-grade dysplasia, or by size ≥ 10 mm."
VC's performance metrics were based on the use of current techniques, including primary 3D reading, and oral contrast tagging. Costs for both VC and colonoscopy were derived from costs realized at the Wisconsin facility.
The results showed that without any intervention, the estimated five-year colorectal cancer death rate from 6- to 9-mm polyps in the cohort was 0.08%, a sevenfold decrease from the 0.56% colorectal cancer risk from the general screening population, which includes larger lesions.
"The death rate was further reduced to 0.03% with the CTC surveillance strategy and to 0.02% with immediate colonoscopy referral," they wrote. "However, for each additional cancer-related death prevented with immediate polypectomy versus CTC follow-up, 9,977 colonoscopy referrals would be needed, resulting in 10 additional perforations and an incremental cost-effectiveness ratio of $372,853."
The results suggest that the short-term colorectal cancer risk for patients harboring 6- to 9-mm lesions is many times lower than the baseline colorectal cancer rate of 8.5% in the general population. According to Surveillance, Epidemiology, and End Results (SEER) data, the 0.56% risk associated with the general population is almost entirely due to the 4% to 5% of patients with large polyps or masses, Pickhardt and colleagues wrote.
The high cost and complications of colonoscopy follow-up also appeared to underscore the benefits of VC surveillance. Short-term surveillance of small polyps was far more cost-effective, with an incremental cost-effectiveness ratio for immediate polypectomy of $372,853 per life-year-gained, representing about 10,000 additional colonoscopy exams, they wrote.
"In addition, any gain in clinical efficacy related to CRC mortality reduction would be largely negated by the endoscopic complications resulting from such a large number of invasive examinations," they wrote. "In fact, when the estimated 5-year CRC death rate from unresected 6- to 9-mm polyps in our study is applied to the general screening population as a whole, the absolute death rate (0.007%) is similar to the expected death rate related to complications at optical colonoscopy (0.006%)."
The low clinical yield of colonoscopy for small lesions is another reason to seek out a less aggressive strategy, they wrote.
Another similar model, by Hur et al (Clinical Gastroenterology and Hepatology, February 2007, Vol. 5:3, pp. 237-244), estimated far more cases of colorectal cancer for patients harboring 6- to 9-mm lesions: 773 per 100,000 patients. However, the results were based on old data from a high-risk, symptomatic cohort, Pickhardt and his team wrote.
The present study has limitations in that its accuracy is highly dependent on the underlying input assumptions, the team wrote. However, by intentionally biasing some of the predictions in favor of higher cancer risk, the authors said they hoped to counteract any inherent underestimation of risk.
Based on a hypothetical population of asymptomatic 60-year-old adults, "our analysis shows that the very low [colorectal cancer] risk associated with 6- to 9-mm polyps detected at CTC screening argues against immediate colonoscopy referral," they concluded. "... the high costs and additional complications associated with immediate colonoscopy referral for all 6- to 9-mm polyps also support the practice of CTC surveillance. ... the most important factor for CRC risk reduction at CTC screening appeared to be exclusion of large polyps or masses at the baseline evaluation."
By Eric Barnes
AuntMinnie.com staff writer
November 4, 2008
VC finds perforations from incomplete colonoscopy, October 24, 2008
Follow-up care lacking for many older colon cancer survivors, September 11, 2008
Highest colonoscopy screening rates are in low-risk patients, May 29, 2008
Model finds polyp surveillance effective for cancer prevention, January 21, 2008
Screening colonoscopy findings linked to incident lesions during surveillance, November 8, 2007
Copyright © 2008 AuntMinnie.com