Colonoscopy surveillance in Lynch syndrome is burdensome and frequently delayed

Individuals with Lynch syndrome, the most common hereditary colorectal cancer (CRC) syndrome, have a high lifetime risk for CRC, which varies between the different causal germline mutations from 15 to 70% [1]. Early detection of CRC, but also detection and removal of premalignant adenomas, reduces incidence, morbidity and mortality of CRC in Lynch [2]. Therefore, these individuals are advised to undergo biennial colonoscopy, starting from age 25 or 35 depending on mutated gene variant [1, 3‐5]. Since the implementation of this surveillance program, the prognosis of individuals with Lynch syndrome has improved substantially [6, 7].

However, the effectiveness of a surveillance program for high-risk individuals is dependent on both the accuracy of the surveillance method as well as the adherence. Previous research has demonstrated that adherence to timely colonoscopy surveillance in individuals with Lynch syndrome is suboptimal [8‐11]. As these studies are slightly outdated and do not reflect current, intensified surveillance recommendations, non-adherence rates might be even greater in clinical practice. Suboptimal compliance to colonoscopy surveillance can partly be explained by insufficient interval recommendations and suboptimal logistics, and by surveillance being performed in a non-specialty clinic and/or a clinic without coordination by a centralised (colorectal) cancer registry [8‐10]. However, of greater influence are probably patients’ perceived barriers to invasive and regular colonoscopy surveillance, such as discomfort, embarrassment and surveillance being time-consuming [9, 10, 12, 13].

Multiple studies in average-risk individuals showed that colonoscopy and the required bowel preparation clearly come with a burden [14‐16]. This burden might be even greater among individuals with Lynch syndrome, given that this high-risk population faces regular and lifelong colonoscopies and that surveillance for cancer on itself has a negative impact on psychological well-being and health related quality of life [17‐20]. However, in this population no comprehensive study has yet investigated the perception of colonoscopy surveillance or factors that prompt or deter individuals to adhere to surveillance. Such detailed information is important to design targeted interventions to improve satisfaction with colonoscopy surveillance and lower its burden and non-compliance rates.

Therefore, the aim of the current study was to investigate the perception and preferences regarding different aspects of colonoscopy surveillance in a large cohort of individuals with Lynch syndrome. Secondly, we aimed to further explore surveillance behaviour in this population, including up-to-date non-compliance rates and predictors of (non-)compliance.

This study evaluated the impact of colonoscopy surveillance in 197 individuals with Lynch syndrome, a population with increased CRC risk requiring strict lifelong colonoscopy surveillance. We showed that surveillance was perceived as impacting quality of life in 21% of individuals and as moderately to extremely burdensome in 57%, particularly in those below age 40. Specifically burdensome were the volume and taste of the laxative and the laxation-related bowel movements. Other studies also indicated bowel preparation as the most burdensome aspect, independent of colonoscopy indication [9, 13, 15, 16]. Regarding the colonoscopy itself, we showed that perceived embarrassment was low whereas perceived pain modest, which is in line with previous research [14‐16, 26]. Over 80% of our cohort did yet receive (adequate) sedation, hence, to reduce the pain scores the procedure itself needs to be further optimised. Lastly, we further explored surveillance behaviour in Lynch syndrome, including up-to-date non-compliance rates. We found that 28% of individuals had delayed colonoscopy surveillance and that an additional 10% considered quitting or postponing surveillance. Based on our results, we believe an effort should be made to improve acceptance of and compliance with colonoscopy surveillance in individuals with Lynch syndrome.

To potentially lower the burden of colonoscopy surveillance, participants prioritised more acceptable bowel preparation, shorter waiting times on the day of the colonoscopy, and a more personal and respectful approach of endoscopic staff. The vast majority of our cohort received bowel preparation by ascorbic acid-enriched polyethylene glycol (Moviprep®) as this was standard care in our centre, however, it has been shown that in both surveillance and other populations some might better tolerate the smaller volume preparations Picoprep® (magnesium citrate plus picosulphate) and Pleinvue® (polyethylene with higher ascorbate concentration), with non-inferior bowel cleansing quality and adenoma detection rates [27]. Hence, substituting Moviprep® by Picoprep® or Pleinvue® may be considered in clinical practice. Other studies also underlined the importance of a respectful and personal approach during the procedure, irrespective of its indication [13‐15, 28, 29]. Familiarity with the endoscopist on the other hand seems to be particularly of value in populations requiring regular (surveillance) colonoscopies [13, 29].

A substantial proportion of our cohort (28%) showed non-compliance with 2-yearly colonoscopy surveillance, which was slightly higher than in other studies extracting “objective” compliance behaviour from medical files [8, 9, 11]. However, these studies did not take multiple surveillance rounds into account, used a less stringent definition for non-compliance or included patients with 2–3 yearly interval recommendations [8, 9, 11]. In our study, 76% of surveillance delays were patient-related. On multivariable analysis, patient-related surveillance delay was not associated with surveillance being experienced as burdensome, even though this was the main reason to consider quitting or postponing surveillance. Of the individuals considering quitting/postponing surveillance, only 50% actually had a patient-related delay. These findings might imply that individuals with Lynch syndrome regard colonoscopy as a life necessity and something that has to be endured, as has also been observed in patients with inflammatory bowel disease under strict surveillance [29]. Further, surveillance behaviour in Lynch syndrome, amongst other populations, can be explained and predicted using the Health Belief Model [30]. According to this model, surveillance behaviour is influenced by a patient’s 1) perceived susceptibility to the disease, 2) perceived severity of the disease, 3) perceived benefits of surveillance, 4) perceived barriers to action, 5) exposure to cues to action and 6) confidence in capability to succeed. Thorough exploration of these themes was beyond the scope of our study, though would be informative, hence qualitative research in this field is warranted.

In our study we gained insight into independent factors that prompt or deter individuals with Lynch syndrome to adhere to colonoscopy surveillance. First, patient-related delay was correlated with a history of ≤ 4 colonoscopies (OR 20.68 – 34.77). This might indicate that the first experiences with surveillance may determine whether or not individuals will participate in successive surveillance rounds, making it important that clinicians strive for a satisfactory first experience and that they provide clear information and interval recommendations. Secondly, non-compliance was associated with receiving surveillance at location AMC rather than location VUmc (OR 3.49). A major difference between both locations was that VUmc did have an automated recall system whereas AMC did not (see method section). We also found that of the patients having a patient-related delay, 50% indicated they never considered delaying surveillance. These findings suggest that individuals might not be aware of the recommended surveillance interval or of their surveillance being delayed, and that proactive management by a hospital recall system – for example coordinated by a centralised cancer registry such as the Netherlands Foundation for Detection of Hereditary Tumours – seem to facilitate better compliance. According to the previously mentioned Health Belief Model [30], such a reminder can serve as a “cue to action”. This also holds true for visiting a clinician who discusses the importance of surveillance, making it important that clinicians other than gastroenterologists (e.g. gynaecologists or general practitioners) recommend colonoscopy surveillance during consultation with someone having Lynch syndrome. The third identified predictor of non-compliance concerned low and medium educational level (OR 7.41 and OR 2.82, respectively), which is consistent with other studies on CRC screening [31‐33]. Further research should explore why this sub-set of individuals shows lower compliance rates, and which strategies could enhance their compliance.

Another approach that could potentially improve compliance with colorectal surveillance in Lynch syndrome, as well as quality of life, is a less-invasive screening method that guides the need and/or timing of colonoscopy. Besides reducing the frequency of invasive and burdensome colonoscopic examination and bowel preparation, such a test would potentially reduce the interval CRC rate by selecting those requiring a colonoscopy at a shorter time interval. Our recent systematic review on non-endoscopic colorectal surveillance in Lynch syndrome concluded that non-invasive biomarkers may hold potential [34]. In the current study, participants expressed a clear preference for alternative less-invasive surveillance modalities, of which biomarkers were particularly favoured. Based on these observations, we believe (pre-)clinical studies in this field should be conducted.

To the best of our knowledge, this is the first study to date providing a comprehensive assessment of the impact of colonoscopy surveillance for Lynch syndrome in a large cohort of 197 individuals. Other strengths of our study lie in having used validated surveys which were formally re-validated in our population of interest, the high response rate of 68%, and 98% rating the survey as average or easy to perform. Moreover, non-compliance was not self-reported but extracted from medical files, and was determined based on three most recent colonoscopies; as the risk of CRC in Lynch syndrome is lifelong, CRC prevention is only effective if individuals adhere to multiple surveillance rounds over time. Only a small proportion of our cohort (20%) had undergone just one or two colonoscopies; these individuals were not excluded as this might have introduced selection bias, and we valued to also investigate surveillance burden and compliance in those just having started surveillance.

Several limitations should be acknowledged as well. First, our data were derived from a sample of mainly Caucasian individuals under surveillance in a single academic hospital, hence, may not be generalisable to other Lynch syndrome populations. Second, we did not gain insight into (factors predicting) uptake of the first colonoscopy following Lynch syndrome diagnosis. The third limitation concerns the possibility of selection bias, as non-compliers to surveillance are likely overrepresented among those who decline to participate in survey-based research [35]. This might imply that non-compliance rates were actually even greater than 28% in our practice.

In conclusion, our large study highlights that colonoscopy surveillance in Lynch syndrome is often experienced as burdensome, particularly in individuals below age 40. Moreover, we showed that colonoscopy surveillance is frequently delayed, and that the vast majority of the delays were for patient-related reasons. An effort should be made to improve acceptance of and compliance with colonoscopy surveillance in this population; potential interventions requiring further evaluation include a personalised bowel preparation regimen, personalised approach of endoscopic staff, short waiting times, clear information provision and interval recommendations, a hospital recall system and non-invasive biomarkers.