Background: Current guidelines stratify patients with a personal history of adenomas as low risk (ie, 1-2 small [<10 mm] adenomas at index colonoscopy) or high risk (> or =3 small adenomas or advanced adenoma at index colonoscopy) for recurrent advanced adenomas. Guidelines recommend longer intervals between surveillance colonoscopies for low-risk patients, but physicians frequently perform surveillance colonoscopy at shorter intervals for these patients.
Objective: Our purpose was to perform a meta-analysis about the incidence of advanced adenomas at 3-year surveillance colonoscopy among high- and low-risk patients.
Methods: Computer searches of MEDLINE, PREMEDLINE, and EMBASE were performed to identify appropriate studies. Study selection criteria were (1) study design--prospective or registry-based study, (2) study population--patients with a personal history of adenomas, and (3) intervention--completion of surveillance colonoscopy at an interval of > or =2 years. Data were extracted on (1) incidence of advanced adenomas at surveillance colonoscopy, (2) interval between colonoscopies, and (3) risk factors associated with recurrent adenomas. After the validity of study design was assessed and independent, duplicate data extraction was performed from selected trials, summary relative risks (RR) for the incidence of advanced adenomas at 3-year colonoscopy were calculated.
Results: Fifteen studies met study selection criteria, but only 5 studies stratified surveillance colonoscopy results according to findings at the index colonoscopy. Patients with > or =3 adenomas at index colonoscopy were more likely to have recurrent advanced adenomas than were patients with 1 to 2 adenomas: RR 2.52, 95% CI 1.07-5.97. Patients with adenomas with high-grade dysplasia at index colonoscopy were also at increased risk for recurrent advanced adenomas: RR 1.84, 95% CI 1.06-3.19. In the individual studies, increasing size of adenomas and increasing number of adenomas at index colonoscopy were the most commonly reported risk factors associated with recurrent advanced adenomas. No studies stratified surveillance colonoscopy results according to the definitions of low risk and high risk used in current guidelines.
Conclusion: Few published studies stratify the incidence of advanced adenomas at surveillance colonoscopy according to index colonoscopy findings. In the future, large prospective studies or studies using pooled data from existing randomized controlled trial databases or polyp registries should be used to better define which patients are at low risk for advanced adenoma recurrence.