Motion – Colonoscopic surveillance is more cost effective than colectomy in patients with ulcerative colitis : Arguments for the motion

This article was originally presented at a symposium entitled, “Controversies in Gastroenterology”, sponsored by Axcan Pharma, Toronto, Ontario, April 8 to 10, 2002 Center for Inflammatory Bowel Disease, Cleveland Clinic Foundation, Cleveland, Ohio, USA Correspondence: Dr BA Lashner, Center for Inflammatory Bowel Disease, Cleveland Clinic Foundation, 9500 Euclid Avenue, 1A30, Cleveland, Ohio 44195, USA. Telephone 216-444-6524, fax 216-444-6305, e-mail lashneb@ccf.org BA Lashner. Motion – Colonoscopic surveillance is more cost effective than colectomy in patients with ulcerative colitis: Arguments for the motion. Can J Gastroenterol 2003;17(2):119-121.

P rotocols for the detection of neoplastic lesions by surveil- lance colonoscopy have evolved over the past 30 years, and have become the standard of care in high risk patients with ulcerative colitis (UC).Throughout the world, colonoscopy is routinely undertaken to detect dysplasia and asymptomatic cancers.Efforts to minimize the cost of this expensive form of cancer surveillance would allow it to be offered to as many patients as possible within the financial constraints of the health care system.

COLORECTAL CANCER IN UC
Recent studies from Sweden, Denmark, Greece, Germany and Japan have found that the cumulative incidence of colorectal cancer (CRC) in UC patients is between 5% and 13% (1)(2)(3)(4)(5).While these figures are not very different from the CRC risk in the general population, the younger age at onset of cancer in UC patients means that the age specific relative risk is greater than three.
As with sporadic CRC, cancer-related mortality in UC patients with CRC is approximately 50%.Some investigators, however, have reported that the mortality rates from rectal cancer in UC patients are greater than in sporadic rectal cancer (6,7).On the other hand, while incidence rates for CRC in patients with UC have been rising, mortality rates from this tumour have been decreasing (8).The divergence of incidence and mortality curves suggests that UC patients are enjoying longer life expectancies, which is likely related to improvements in treatment and cancer surveillance.
Patients with longstanding and anatomically extensive disease are especially likely to develop CRC.There is also evidence that a young age at disease onset is an independent risk factor (9).Primary sclerosing cholangitis (PSC) has recently been identified as an additional risk factor.The small proportion of UC patients with PSC (approximately 2%) has a relative risk for CRC of at least three, compared with UC patients without PSC (10).Patients with PSC should, therefore, be made aware of the magnitude of this risk and undergo either frequent surveillance colonoscopies or prophylactic colectomy.One study has shown that the incidence of dysplasia and cancer can be reduced by treatment with ursodeoxycholic acid, which alters bile salt composition in PSC patients (11).Chemoprevention is not a substitute for cancer surveillance, however.

EFFECTIVENESS OF CANCER SURVEILLANCE
Surveillance colonoscopy reduces the cancer-related mortality in patients with UC, as has been shown in two case control studies (12,13).Moreover, increased frequency of surveillance is associated with a larger reduction in mortality.
The strongest evidence for the effectiveness of cancer surveillance comes from a decision analysis.In this study, Provenzale et al (14) applied the best available data to a hypothetical cohort of 30-year-old patients with a 10-year history of ulcerative pancolitis without PSC.They compared 17 different cancer surveillance strategies with either prophylactic colectomy or no surveillance (ie, colectomy only when symptomatic cancer is detected).The estimated cumulative incidence of cancer in the 'no surveillance' group was 6.67%.The strategy of performing surveillance colonoscopy every three years, and colectomy for patients with even low grade dysplasia, resulted in a decrease in cumulative cancer incidence to 0.47%.Of course, prophylactic colectomy eliminated the risk of CRC.The remaining life expectancy in the cohort was 46.31 years with no surveillance, 47.19 years with surveillance every three years, and 47.45 years with prophylactic colectomy.The authors concluded that the cancer risk and life expectancy could be dramatically improved by colonoscopic surveillance, and that the results were almost as good as with prophylactic colectomy.

COSTS OF CANCER SURVEILLANCE
The benefits of cancer surveillance and prophylactic colectomy must be weighed against its costs, including increased frequencies of colonoscopy and surgery (both elective and urgent) and reduced quality of life following colectomy.Table 1 shows data from the decision analysis undertaken by Provenzale et al (14).It is apparent that both 'surveillance' and 'prophylactic colectomy' strategies are both more effective and more costly than 'no surveillance'.

COST EFFECTIVENESS OF CANCER SURVEILLANCE
Provenzale et al (15) later calculated cost effectiveness ratios from their model.They found that all surveillance strategies had lower costs and superior effectiveness than no surveillance.Therefore, a 'no surveillance' policy is unacceptable.Moreover, strategies using low grade dysplasia as a criterion for performing a colectomy were both more effective and less costly than were those that used high grade dysplasia.The incremental cost effectiveness ratios for various colonoscopic surveillance strategies are listed in Table 2.These ratios compare favourably to those derived from widely accepted surveillance regimens, such as mammography for breast cancer and Papanicolaou smears for cervical cancer.Specifically, colonoscopic surveillance every three years or at a variable interval, as defined in Table 2, are worthwhile.This means that the decrease in cancer risk and increase in life expectancy justify the costs of testing and surgery and the decreases in quality of life that follow colectomy.
Another decision analysis has recently been published (16).It considered the costs of testing, surgery and terminal care for patients dying from CRC, as well as the number of life-years saved.The authors found that a strategy of colonoscopy every two years would become cost effective if the cumulative risk of developing CRC were 27%.The threshold CRC risk was 19% for testing every three years, and 14% for testing every four years.In this model, surveillance was not cost effective compared with no surveillance.The poor effectiveness of the 'no surveillance' strategy, however, makes this option unacceptable.Neither of these two cost effectiveness analyses (15,16) explored the option of prophylactic colectomy.

IMPROVING COST EFFECTIVENESS
Cancer surveillance colonoscopy regimens are cost effective compared with no surveillance, and are generally regarded as the standard of care for eligible patients.It is, therefore, incumbent on gastroenterologists to recommend surveillance, with parameters that maximize life expectancy at a reasonable cost.
Clinicians should pay close attention to patient selection, the criteria for recommending colectomy and the testing interval.Cost effectiveness ratios could be reduced by limiting surveillance to high risk patients, such as UC patients who are at high risk for CRC, have pancolitis of at least seven years' duration or PSC.Because patients with left-sided UC, short duration of disease or Crohn's colitis are less likely to develop CRC, surveillance is associated with higher cost effectiveness ratios.Performing colectomy in patients with low grade dysplasia, as  Data from Provencale et al (14) opposed to high grade dysplasia, results in a higher cost (increased number of colectomies) but a disproportionately higher effectiveness (increased number of life-years saved), and thus a lower cost effectiveness ratio.As already stated, the marginal cost effectiveness ratios are reasonable when colonoscopy is performed every three years, or when the testing interval is adjusted according to the patient's individual risk (17).Annual testing is not cost effective for patients with moderately high risk, nor is it an efficient use of scarce resources, but is appropriate for patients who are at very high risk for CRC, such as those with either PCS or extensive disease for more than 30 years.

SURVEILLANCE VERSUS PROPHYLACTIC COLECTOMY
A perfectly effective cancer surveillance colonoscopy program does not exist.For a variety of reasons, some patients develop cancer even though dysplasia was not detected at earlier colonoscopy.Patients who are concerned about the risk of cancer and cannot accept the limitations of cancer surveillance should undergo prophylactic colectomy.The ability to eliminate the risk of cancer should be weighed against the decrease in quality of life after colectomy.This personal decision is not necessarily based on cost effectiveness calculations.Variable interval* versus every three years $155,400 Every three years versus every four years $111,600 Every four years versus every five years $83,700 Every five years versus no surveillance $4,700 Data from Provenzale et al (15).*Colonoscopy every three years for the first 20 years of disease, every two years for the next eight years, and annually thereafter

Colonoscopy is more cost effective than colectomy in UC
Can J Gastroenterol Vol 17 No 2 February 2003 121