The impact of colonoscopy on colorectal cancer incidence and mortality is one of the great recent success stories in medicine. Although the major impact of colonoscopy to date has been in colorectal cancer prevention through the practice of screening and surveillance, advances in endoscopic resection and closure techniques provide expanded opportunities for colonoscopy as destination therapy for early-stage colorectal neoplasia. In this issue of Gastrointestinal Endoscopy, Currie et al
1
report the results of a decision analysis intended to determine the preferred treatment strategy for hypothetical cohorts of patients with early colon cancer.The analysis compares the effectiveness of endoscopic versus surgical resection for T1 colon cancer using a Markov simulation model. This mathematical model is able to define outcomes as patients transition between health/disease states and undergo therapeutic interventions. The “effectiveness” of an intervention is influenced both by the probabilities of disease-specific outcomes after an intervention and by what is known as utility assessment, assigning a value between 0 and 1, where 0 indicates death and 1 indicates perfect health, meant to reflect a patient’s quality of life within a specific health/disease state.
Decision analysis studies are valuable in instances when randomized prospective data comparing competing treatment strategies are not available. Design and completion of a randomized controlled trial comparing endoscopic versus surgical therapy for T1 colorectal cancer would require considerable effort and expense given the large number of patients and duration of follow-up required to detect a difference in meaningful clinical endpoints (cancer morbidity and cancer death). Both physicians and patients may be loath to participate in such a trial, concerned about making a critical cancer decision based on chance. Moreover, high rates of post-randomization dropout might be anticipated among those allocated to the more-invasive surgical arm.
Currie et al
1
designed their model to include 4 different base case hypothetical scenarios: a 65-year-old healthy (fit) man, a 65-year-old man with significant comorbid disease, an 80-year-old healthy man, and an 80-year-old man with significant comorbid disease. T1 cancers were stratified as either low-risk or high-risk lesions. The results of the analysis suggest that surgical resection is the preferred strategy in 65-year-old healthy men with high-risk T1 lesions; endoscopic therapy is the preferred strategy in all other patient groups, including across a range of possible treatment outcomes as assessed by univariate sensitivity analysis.Any given decision analysis model is constrained by the details of its design and the quality of the model parameters or input data. This model assumes that accurate staging and risk stratification of T1 colon cancer is established by techniques including narrow-band imaging, chromoendoscopy, and miniprobe ultrasonography. As the authors acknowledge, this may not reflect standard practice outside specialized academic or research settings. The scenario also does not account for the possibility that a T1 cancer may be incidentally detected in post-resection histopathologic assessment, as may occur even in a small percentage of adenomas smaller than 10 mm in size.
2
The model assumes that palliative care is the only therapeutic option for distant recurrence, when in practice some instances of distant recurrence (ie, hepatic metastases) may be treated with surgical resection. The utility values used in the model are nearly 2 decades old, and thus may not accurately reflect patient utility states with current surgical practice, and were derived from a survey of healthy patients younger than any of the base case scenarios in this model.3
In many respects, the analysis presented by Currie et al
1
mirrors the recent evolution in the treatment approach to early-stage esophageal neoplasia. Whereas endoscopic therapy for Barrett’s esophagus with high-grade dysplasia or T1 cancer was once reserved for patients unfit for surgical esophagectomy on the basis of age or comorbid illness, the accumulation of data demonstrating efficacy of endoscopic therapy with increasing follow-up duration has resulted in an increase in the number of patients undergoing endoscopic therapy in clinical practice.4
The decision to offer endoscopic resection rather than esophagectomy (or colectomy) to an 85-year-old patient with significant comorbid disease may be an easy decision. The decision to offer endoscopic resection as destination curative therapy to a 65-year-old (or younger) patient who is otherwise a fit surgical candidate with decades of life expectancy is much more difficult and nuanced.In such cases, individual treatment decisions are in part the product of numerous subjective factors that cannot be accounted for in a mathematical model: physician and patient bias. From the perspective of physician bias, again using the example of esophageal cancer, data suggest that whether a patient with esophageal neoplasia associated with Barrett’s esophagus undergoes esophagectomy or endoscopic therapy may be heavily influenced by whether the patient is seen by a surgeon or a gastroenterologist at the initial consultation.
5
And from the perspective of patient bias, any clinician who has discussed the option of colectomy with patients will readily acknowledge that apprehension regarding the potential for colostomy may be an insurmountable hurdle for some patients, utility assessment and all other treatment and prognostic factors aside.The study by Currie et al
1
provides modeling data to inform acceptable (even preferred) options available for the effective treatment of T1 colorectal cancer across a range of clinic scenarios. Yet how we arrive at the better option requires that we engage patients in candid discussions regarding possible treatment outcomes. High rates of decision confidence and low rates of decision regret may be experienced by both patients with esophageal cancer who undergo endoscopic therapy and those who undergo surgery.6
Such outcomes should be attainable for our patients with early colorectal cancer if we as health care providers effectively engage in a shared, patient-centered decision-making process.Disclosure
All authors disclosed no financial relationships relevant to this publication.
References
- The potential impact of local excision for T1 colonic cancer in elderly and comorbid populations: a decision analysis.Gastrointest Endosc. 2016; 84: 986-994
- Prevalence of clinically important histology in small adenomas.Clin Gastroenterol Hepatol. 2006; 4: 343-348
- Utility valuations for outcome states of colorectal cancer.Am J Gastroenterol. 1999; 94: 1650-1657
- Survival of patients with superficial esophageal adenocarcinoma after endoscopic treatment vs surgery.Clin Gastroenterol Hepatol. 2013; 11: 1424-1429
- Treatment of Barrett’s esophagus with high-grade dysplasia or cancer: predictors or surgical versus endoscopic therapy.Clin Gastroenterol Hepatol. 2008; 6: 1206-1211
- Patient decision-making and clinical outcomes following endoscopic therapy or esophagectomy for Barrett’s neoplasia.Gastroenterology. 2014; 150: S266
Article info
Identification
Copyright
Copyright © 2016 by the American Society for Gastrointestinal Endoscopy
ScienceDirect
Access this article on ScienceDirectLinked Article
- The potential impact of local excision for T1 colonic cancer in elderly and comorbid populations: a decision analysisGastrointestinal EndoscopyVol. 84Issue 6
- PreviewPopulation-based bowel cancer screening has resulted in increasing numbers of patients with T1 colonic cancer. The need for colectomy in this group is questioned due to the low risk of lymphatic spread and increased treatment morbidity, particularly for elderly, comorbid patients. This study examined the quality-of-life benefits and risks of endoscopic resection compared with results after colectomy, for low-risk and high-risk T1 colonic cancer.
- Full-Text
- Preview
