Metachronous colorectal cancer: Is it all about colonoscopy quality?

Colonoscopy has the potential to reduce mortality by detecting premalignant precursor lesions and resecting them in a timely manner. Besides being used as a screening method, regular colonoscopy is the method of choice for patients at an increased risk for colorectal cancer (CRC) (eg, those with a familial history or known hereditary risk or those with previous adenomas of CRC). The purpose of regular surveillance colonoscopies is to protect these high-risk persons from the development of (metachronous) CRC. Although colonoscopy is the criterion standard for the detection and prevention of CRC, it is not perfect. Despite clear surveillance guidelines, the cumulative incidence of metachronous CRCs is approximately 3% after 10 years in the Western world, rising to 7% after 20 years.^,  The study by le Clercq and colleagues in this issue describes a retrospective cohort of 5157 patients with primary CRC, in 1.8% of whom metachronous CRC developed during a mean of 6 years of follow-up. The authors have put much effort into disentangling the causes of those metachronous cancers. They conclude that most metachronous lesions result from lesions that were missed during colonoscopy and from nonadherence to surveillance guidelines. This study gives important insights into an area with scarce data. The retrospective nature of this study, however, relies heavily on overlapping definitions regarding the cause of metachronous CRC. This gives food for thought.

According to the criteria published by Moertel and colleagues as early as 1958, the definition of a metachronous CRC is (1) a pathologically proven adenocarcinoma, (2) distinctly separated from the previous line of anastomosis, (3) diagnosed at a minimum interval of 6 months after the initial CRC. However, the time interval between the initial and secondary primary CRC, chosen to distinguish between synchronous and metachronous CRCs, varies in the medical literature between 2 months and 3 years, only accentuating the difficulty in comparing the existing data and understanding the origin and cause of these lesions. Taking into account a polyp dwell time of 10 to 15 years for the majority of CRCs, this suggests that clearing of all lesions at the time of the primary CRC should in theory protect a person for a new metachronous CRC for 10 years.

When the timing of the diagnosis of metachronous CRC is examined, it is striking that in the study by le Clercq and colleagues, 40.8% of these metachronous CRC occurred in the first 3 years after the primary diagnosis of CRC. This is in line with another Dutch study showing that the risk for the development of a metachronous CRC after resection of a primary CRC was significantly increased during the first 3 years of follow-up when compared with the incidence of CRC in an age-matched and gender-matched general population (standardized incidence ratio [SIR] 1.4; 95% confidence interval, 1.1-1.8). After 3 years of follow-up, there was no difference in CRC risk compared with the general population. This means that for the prevention of metachronous CRC, much can be gained in these first 3 years after the primary CRC.

To effectively reduce the incidence of metachronous CRCs, it is important to understand the origin and cause of this second tumor. A metachronous CRC could (1) be missed at the time of diagnosis of the primary CRC (actually a synchronous CRC), (2) have resulted from a precursor lesion that was not detected at the time of diagnosis, or (3) be a second primary that was not present at the time of the diagnosis of the primary CRC, not even as a precursor lesion, and thus a fast-growing lesion. However, all lesions should ideally have been picked up in a timely manner by a scheduled surveillance colonoscopy. More detailed information on patients with metachronous CRCs should help to identify the cause of those unwanted CRCs: information on the quality of the colonoscopy at the time of diagnosis, information on the surveillance interval after diagnosis and treatment for CRC, and, last but not least, more knowledge about the biology of those metachronous cancers.

In recent years, data on the importance of the quality of colonoscopy to optimize its protective potential for CRC have accumulated in the literature. A landmark study from Poland demonstrated a clear correlation between the adenoma detection rate and the incidence of interval carcinomas in the years after the colonoscopy. It is now clear that not only adenomas, but also large and especially right-sided serrated lesions, can be precursor lesions for CRC, and those are more difficult to detect because of their morphology—usually flat elevated—and innocuous color. Their detection rates vary widely among endoscopists. Although no data on the association with interval carcinomas are available, this suggests that besides adenoma detection rates, serrated polyp detection rates should be monitored also. Besides the detection of those relevant lesions, radical endoscopic resection is also an issue. The CARE study demonstrated that 10.1% of polyps up to 2 cm are not completely resected; the rate of incomplete resection was highest for sessile serrated adenomas at 31%. Among other important quality indicators for colonoscopy are quality of the bowel preparation and cecal intubation rate. It is important that all quality determinants are reported in every colonoscopy report, ideally facilitated by a quality-assured endoscopy reporting system that can be used for regular quality assessment, benchmarking, feedback, and improvement. However, because most studies of metachronous cancers are retrospective in nature, and quality is not reported routinely, it is difficult to ascertain the quality of colonoscopy at the time of primary CRC diagnosis. It is quite possible that this was not optimal (eg, because of the stenosing character of the tumor, emotional stress in the patient and/or the endoscopist upon endoscopic detection of CRC leading to less-optimal bowel preparation or less attention during inspection). If the diagnostic endoscopy was not of high quality—or perhaps, in the case of an acute obstruction, not complete—guidelines advise additional preoperative CT colonography or a complete colonoscopy 3 months after surgery to detect potential synchronous tumors.

To cover for the potential miss of a synchronous CRC or precursor lesion and to detect newly developing metachronous lesions in time, surveillance guidelines have been developed. Different schedules exist in the literature, and because randomized studies are lacking, the optimal interval after the first diagnosis of CRC is not well known. Retrospective data have shown a relatively high incidence of metachronous CRCs in the first 3 years after diagnosis, but as just discussed, this could be partly due to the quality of the colonoscopy at the time of diagnosis. However, to understand the cause of metachronous CRC it is important to be informed about the surveillance interval being used. In most studies, again because of the retrospective nature of the studies, it is difficult to ascertain the exact surveillance schedule. Besides these guidelines, adherence to colonoscopy surveillance advice varies depending on patient factors (eg, burden of colonoscopy, fear of the result), medical factors (eg, age or comorbidity of the patient), and logistic factors (eg, no organized recall system).

The cause of a metachronous CRC can also lie in the biology of the tumor. Although the conventional adenoma–carcinoma pathway has an estimated polyp dwell time of 10 to 15 years, microsatellite instable cancers (an estimated 15% of all CRCs) are estimated to develop in only 3 to 5 years. Besides, uncertainty exists about the progression time of a sessile serrated polyp becoming invasive. Thus, metachronous cancers can indeed be newly developed and fast-growing tumors. To elucidate the role of biology in metachronous CRCs, studies on molecular profiling of these lesions are of utmost importance. Besides, information is lacking on the specific patient and tumor characteristics at the time of the primary CRC diagnosis that are associated with an increased risk for the development of a metachronous CRC.

To truly disentangle the cause of metachronous CRCs and ultimately reduce their incidence, we need large prospective cohorts of patients undergoing high-quality colonoscopy, with standardized data on the patients and on the quality of the procedures, standardized surveillance intervals, and coupling with a regional or national cancer registry. This would facilitate the determination of safe and tailor-made surveillance intervals, based on true risk stratification, ideally not aiming to detect all small and irrelevant lesions but only those lesions at a high risk for their eventual development into CRCs.