Risk of total metachronous advanced neoplasia at surveillance colonoscopy after detection of serrated lesions: a matched case–cohort study

Abstract

Background Serrated lesions are potential colorectal cancer precursors. This study evaluated the presence of total metachronous advanced neoplasia (T-MAN) at follow-up in patients with index serrated lesions compared with a matched cohort without serrated lesions.

Methods Patients aged 45–74 years with serrated lesions were matched 2:1 by sex, age, synchronous polyps, and timing of index colonoscopy, to patients without serrated lesions. The primary outcome was T-MAN (advanced adenoma or high-risk serrated lesion) at follow-up. Secondary outcomes included presence of T-MAN stratified by synchronous polyps and serrated lesion characteristics.

Results 1425 patients were included (475 patients, 642 serrated lesions; 950 controls; median follow-up 2.9 versus 3.6 years). Patients with serrated lesions had greater risk of T-MAN than those without (hazard ratio [HR] 6.1, 95 %CI 3.9–9.6). Patients with serrated lesions and high-risk adenoma (HRA) had higher risk of T-MAN than those with HRA alone (HR 2.6, 95 %CI 1.4–4.7); similarly, patients with serrated lesions plus low-risk adenoma (LRA) had higher risk than those with LRA alone (HR 7.0, 95 %CI 2.8–18.4), as did patients with serrated lesions without adenoma compared with no adenoma (HR 14.9, 95 %CI 6.5–34.0). Presence of proximal sessile serrated lesion (SSL; HR 9.3, 95 %CI 5.4–15.9), large SSL (HR 17.8, 95 %CI 7.4–43.3), and proximal large SSL (HR 25.0, 95 %CI 8.8–71.3), but not distal SSL, were associated with greater risk for T-MAN.

Conclusion Patients with serrated lesions had higher risk for T-MAN regardless of synchronous adenomas. Patients with serrated lesions and HRA, and those with large or proximal SSLs, were at greatest risk.

Introduction

Colonoscopy-based screening and surveillance programs have been established to reduce the morbidity and mortality associated with colorectal cancer (CRC), with multiple guidelines establishing follow-up recommendations based on number, size, and histology of colorectal polyps at index examination [1] [2] [3] [4] [5] [6].

Serrated lesions, including sessile serrated lesions (SSLs) and traditional serrated adenomas (TSAs) have become the subject of increased interest for their role as CRC precursors [7]. These polyps are frequently found during colonoscopy, with studies showing serrated polyp detection rates of between 10 % and 20 % [8] [9]. The optimal timing of surveillance colonoscopies after diagnosis of serrated lesions remains controversial, as studies looking into surveillance timing have been lacking. The US Multi-Society Task Force on Colorectal Cancer (USMSTF) 2020 guideline recommends a 5–10-year surveillance interval for detection of 1–2 SSLs, 3–5-year intervals for 3–4 SSLs, and 3-year intervals following detection of > 4 SSLs or TSAs [10]. These intervals are longer than the recommended surveillance intervals in the previous 2012 guideline, where surveillance intervals of either 3 or 5 years, but not > 5 years, were recommended [4]. In contrast to the USMSTF guideline, the 2020 European Society of Gastrointestinal Endoscopy (ESGE) guideline states that serrated lesions of < 10 mm do not require further colonoscopy follow-up [11]. Studies have shown increased risk of follow-up findings in patients with index detection of serrated lesions [12] [13] [14] [15]. However, these studies have been limited by small sample sizes, inclusion of distal small hyperplastic polyps (HPPs), and exclusion of advanced serrated lesions as outcomes of interests, explaining the varying assessments of serrated lesion risk compared with studies performed in patients with CRC and the resulting discrepancy in serrated lesion surveillance recommendations in guidelines [4] [11].

Given the uncertainty surrounding optimal surveillance intervals after a diagnosis of serrated lesions, and the emerging evidence that such lesions might be less benign than previously thought, we were interested in studying the risk of total metachronous advanced neoplasia (T-MAN) found in surveillance colonoscopies in patients with a previous diagnosis of serrated lesions, compared with a matched cohort of patients without serrated lesions at index colonoscopy.

Methods

This study was reported according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) checklist for cohort studies [16] (see online-only Supplementary material).

Study design, patient selection, and matching criteria

We conducted a retrospective, single-center, matched case–cohort study of screening patients aged 45–74 years who underwent a screening, surveillance, or diagnostic colonoscopy from 2010 to 2019 at Montreal University Hospital Center (CHUM). Inclusion criteria was at least one serrated lesion (SSL or TSA) diagnosed at index colonoscopy, and available follow-up within 10 years of index. The study protocol was approved by the Montreal University Hospital Research Center (CRCHUM) institutional review board (CER 20.354). Patients were matched individually in a 2:1 ratio with patients without serrated lesions at index, based on sex, age (± 5 years), colonoscopy date (± 2 years), and synchronous polyp status (high-risk adenoma [HRA], low-risk adenoma [LRA], no adenoma).

Exclusion criteria were: 1) inflammatory bowel disease; 2) hereditary CRC syndromes; 3) personal history of CRC; 4) serrated polyposis syndrome; 5) synchronous CRC at index colonoscopy; 6) no follow-up colonoscopy after index. Using the search term “serrated” in the CHUM pathology database, we identified all consecutive patients diagnosed with a colorectal serrated lesion during the period 2010–2019. Patient electronic medical records were accessed to determine inclusion/exclusions. To identify the non-serrated lesion group, the list of all patients who underwent endoscopy at CHUM was consulted systematically using year of index colonoscopy in the serrated lesion group as a starting point, and patients were matched consecutively according to the prespecified criteria. During the inclusion period, dedicated pathologists at CHUM who exclusively subspecialized in gastrointestinal pathology determined the diagnoses for all polypectomy samples. The pathologists were therefore considered experts in differentiating SSLs from HPPs, although knowledge of SSLs was not as prevalent in the literature during the earlier part of the inclusion period.

Data collection and outcomes

Individual patient data were collected from patient electronic medical records at CHUM. Data collected included: patient age and sex; past medical history; family history of CRC; number, pathology, and size of polyps at index and follow-up colonoscopies; presence of high grade dysplasia on pathology reports; bowel preparation quality at index and follow-up colonoscopies; timing between index and follow-up colonoscopies; completeness of index and follow-up colonoscopies; adenoma detection rate (ADR) at index and follow-up colonoscopies. Data were entered onto a database by four researchers (R.D., M.L.L., T.D., E.L.), and a quality review of the entered data was performed during the study to ensure consistency and reduce bias in data entry between researchers. Disagreement over data collection was resolved by a fifth researcher (D.vR.). There were no differences in data collection or assessment methods between the serrated lesion group and the matched group.

The primary outcome was the detection of T-MAN on a per-patient basis in the serrated lesion group (SSL and TSA) compared with the non-serrated lesion group. Secondary outcomes were: the detection of T-MAN in patients according to index findings (HRA, LRA, no adenoma); the risk of T-MAN, metachronous advanced adenoma, and metachronous advanced serrated lesion according to index findings; serrated lesion size, location, and number.

T-MAN was defined as either metachronous advanced adenoma or metachronous advanced serrated lesion. Advanced adenoma was defined as adenoma ≥ 10 mm in size, adenoma with tubulovillous/villous histology, adenoma with high grade dysplasia, or CRC. Advanced serrated lesion was defined as serrated lesions ≥ 10 mm in size (including HPPs ≥ 10 mm), or serrated lesion with dysplasia. HRA was defined as either ≥ 3 adenomas, adenoma ≥ 10 mm in size, adenoma with tubulovillous/villous histology, adenoma with high grade dysplasia, or CRC as per the 2020 USMSTF guideline [10]. LRA was defined as 1–2 nonadvanced adenomas < 10 mm in size. Serrated lesion was defined as either SSL or TSA; however, HPPs ≥ 10 mm were not included in the definition, as identifying all patients with HPPs at our center would have yielded too large a sample to be able to reasonably verify all HPP sizes for inclusion. Proximal serrated lesion or HPP was defined as proximal to the rectum.

Index colonoscopy was defined as the first colonoscopy that detected serrated lesions for the serrated lesion group. Follow-up colonoscopy was defined as any colonoscopy that occurred 12 months or more after the index colonoscopy to allow sufficient time for patients to develop new findings. Findings for colonoscopies performed within 12 months of index colonoscopy were combined as though from one examination if the initial examination had been incomplete, had poor bowel preparation, or if the second colonoscopy was performed to remove an unresected polyp identified during the initial examination.

Poor bowel preparation was defined for both index and follow-up colonoscopies as a Boston Bowel Preparation Scale score of < 6 or a score of < 2 in any bowel segment, or mention of inadequate bowel preparation in the colonoscopy report with a recommendation for earlier surveillance due to the poor bowel preparation.

Adequate follow-up according to guidelines was assessed according to the 2012 USMSTF guideline and was considered adequate if it occurred within 6 months of recommendations for 3-, 5-, and 5–10-year intervals, and within 1-year for recommended 10-year intervals [4]. For serrated lesions, adequacy was calculated from October 2013, as the guideline was published in September 2012, thus allowing enough time for guideline integration in routine practice.

Sample size and statistical analysis

For the study, 10 T-MAN cases per variable included would be needed in order to perform multivariate analyses. Descriptive analysis with presentation of crude numbers, proportions, medians with interquartile ranges (IQRs), or means with SDs were used to present patient, procedure, and polyp outcomes.

Distribution of characteristics between the serrated lesion and non-serrated lesion groups were compared for imbalance using standardized mean differences, with cutoff levels of 0.2, 0.5, and 0.8 being indicative of small, medium, and large differences, respectively, between the two groups [17]. The risk of T-MAN was compared between the serrated lesion group and non-serrated lesion group using a Cox proportional hazard model stratified by the matched pairs. Comparisons were presented as hazard ratios (HRs) with 95 %CIs.

We performed multivariate regression analysis to determine the effects of index and follow-up confounders on development of T-MAN. Our model was adjusted for colonoscopy preparation quality (adequate vs. inadequate), incomplete resection status, index indication, and colonoscopy completeness (cecal intubation vs. non-intubation) at index and follow-up, as these factors could represent significant confounders in the risk of T-MAN detection. All HRs reported in our results are adjusted using this model. A two-tailed P < 0.01 was considered statistically significant for our univariate and multivariate model after pragmatic correction for multiple testing. T-MAN proportions were computed at the follow-up colonoscopy time point. Patients with missing data were excluded from the analyses dependent on the missing data, as almost no patients had missing data in our study. Findings for the serrated lesion group and patients with loss to follow-up were compared using the chi-squared test and t test to assess for possible confounders in patient selection affecting T-MAN risk. A two-tailed P < 0.05 was considered statistically significant for these analyses. We performed all analyses using Stata version 16 (StataCorp, College Station, Texas, USA).

Results

Patient and polyp characteristics

A total of 2428 patients with a colorectal serrated lesion diagnosed at CHUM between 2010 and 2019 were identified. After applying the exclusion criteria, 475 patients (mean age 61.9 years; 46.7 % female; median follow-up 2.9 years) remained and were matched with 950 patients without index serrated lesions (median follow-up 3.6 years) ([Fig. 1]). In the serrated lesion group, 103 patients had HRA, 102 had LRA, and 270 had no adenoma. Successful cecal intubation was achieved in 98.9 % of patients in the serrated lesion group and 98.6 % in the non-serrated lesion group, and 96.0 % of patients in both groups had adequate bowel preparation. A total of 642 serrated lesions were detected in the serrated lesion group; 455 patients had index SSLs and 20 had index TSAs. Median serrated lesion size was 8 mm (IQR 5–10 mm) ([Table 1]). There was no statistically significant difference in most patient and polyp characteristics between the serrated lesion group and patients who were lost to follow-up, except for synchronous HRA findings and colonoscopy indication (Table 1 s).

Fig. 1 Study inclusion flow chart. IBD, inflammatory bowel disease; CRC, colorectal cancer.

Table 1 Patient and polyp characteristics at index colonoscopy.

	Serrated lesion group (n = 475)	Non-serrated lesion group (n = 950)	SMD
Age, mean (SD), years	61.9 (7.2)	61.5 (7.1)	0.04
Sex, female, n (%)	222 (46.7)	444 (46.7)	0.00
ASA class, n (%)			0.23
1	267 (56.2)	598 (64.7)
2	173 (36.4)	297 (32.1)
3 +	35 (7.4)	29 (3.1)
Family history of CRC, n (%)	114 (24.0)	285 (30.0)	0.14
Procedure timeframe, n (%)			0.13
2010–2011	51 (10.7)	78 (8.2)
2012–2013	88 (18.5)	212 (22.3)
2014–2015	139 (29.3)	268 (28.2)
2016–2017	95 (20.0)	203 (21.4)
2018–2019	102 (21.5)	189 (19.9)
Difference from date of matched index, median, (IQR), months	–	5.0 (2.2–8.3)
Index colonoscopy indication, n (%)			0.12
Screening	146 (30.7)	243 (25.6)
Surveillance	220 (46.3)	458 (48.2)
Diagnostic	109 (22.9)	249 (26.2)
Cecal intubation, n (%)	470 (98.9)	937 (98.6)	0.03
Adequate bowel preparation, n (%)	456 (96.0)	912 (96.0)	0.00
ADR, n (%)	206 (43.4)	412 (43.4)	0.00
Proximal HPP, n (%)	107 (22.5)	55 (5.8)	0.49
Synchronous proximal HPP + adenoma, n (%)	57 (12.0)	24 (2.5)	0.37
Largest HPP size, n (%)			0.37
1–5 mm	89 (18.7)	115 (12.1)
6–9 mm	32 (6.7)	35 (3.7)
≥ 10 mm	16 (3.4)	0 (0.0)
SSL, n (%)	455 (95.8)	–	–
TSA, n (%)	20 (4.2)	–	–
Serrated lesions, n	642	–	–
Size, median (IQR), mm	8 (5–10)	–	–
Location, proximal to rectosigmoid, n (%)	359 (75.6)	–	–

SMD, standardized mean difference; ASA, American Society of Anesthesiologists; CRC, colorectal cancer; IQR, interquartile range; ADR, adenoma detection rate; HPP, hyperplastic polyp; SSL, sessile serrated lesion; TSA, traditional serrated adenoma.

High-risk findings in patients with vs. without serrated lesions

A total of 105 T-MANs were detected in the serrated lesion group vs. 62 in the non-serrated lesion group. Overall, 10.3 % of patients in the serrated lesion group had metachronous advanced adenomas vs. 5.8 % in the non-serrated lesion group, and 13.5 % vs. 0.9 %, respectively, had metachronous high-risk serrated lesions ([Table 2]).

Table 2 Patient and polyp characteristics at follow-up colonoscopy.

	Serrated lesion group n = 475	Non-serrated lesion group n = 950
Interval between index and follow-up, mean (IQR), years	2.9 (1.5–3.7)	3.6 (2.7–5.0)
Follow-up[1], n (%)
Appropriate	115 (33.0)	338 (35.6)
Shorter	192 (55.0)	496 (52.2)
Longer	42 (12.0)	116 (12.2)
Indication, n (%)
Surveillance	444 (93.5)	763 (80.3)
Diagnostic	31 (6.5)	187 (19.7)
Cecal intubation, n (%)	465 (97.9)	937 (98.6)
Adequate bowel preparation, n (%)	457 (96.2)	912 (96.0)
ADR, n (%)	167 (35.2)	310 (32.6)
Metachronous advanced adenoma, n (%)	49 (10.3)	55 (5.8)
Subtype of metachronous advanced adenoma[2], n (%)
Adenoma ≥ 10 mm	44 (9.3)	45 (4.7)
TVA	11 (2.3)	18 (1.9)
Villous adenoma	2 (0.4)	2 (0.2)
HGD	6 (1.3)	4 (0.4)
CRC	1 (0.2)	0 (0.0)
Metachronous high-risk serrated lesions, n (%)	64 (13.5)	9 (0.9)
T-MAN, n (%)	105 (22.1)	62 (6.5)

IQR, interquartile range; ADR, adenoma detection rate; TVA, tubulovillous adenoma; HGD, high grade dysplasia; CRC, colorectal cancer; T-MAN, total metachronous advanced neoplasia.

¹ From October 2013 for serrated lesion group.

² Multiple findings can coexist in the same patient.

The serrated lesion group had statistically significantly higher detection of T-MAN compared with the non-serrated lesion group (22.1 % vs. 6.5 %; HR 6.1, 95 %CI 3.9–9.6). Patients with index serrated lesions and HRA combination had higher risk of T-MAN than patients with HRA alone (32.0 % vs. 14.1 %; HR 2.6, 95 %CI 1.4–4.7). Patients with index serrated lesions and LRA combination had higher risk of T-MAN than patients with LRA alone (20.5 % vs. 6.9 %; HR 7.0, 95 %CI 2.7–18.4). Patients with index serrated lesions without adenoma had higher risk of T-MAN than patients with neither index serrated lesions nor adenoma (18.9 % vs. 3.5 %; HR 14.9, 95 %CI 6.5–34.0) ([Table 3], [Fig. 2]).

Table 3 Hazard ratios for total metachronous advanced neoplasia between serrated lesion group and matched non-serrated lesion group.

Index findings	Serrated lesion group, N	Non-serrated lesion group, N	TMAN, n (%)		Unadjusted HR (95 %CI)	Adjusted HR[1] (95 %CI)
			Serrated lesion group	Non-serrated lesion group
Serrated lesion	475	950	105 (22.1)	62 (6.5)	6.1 (4.1–9.1)	6.1 (3.9–9.6)
HRA	103	206	33 (32.0)	29 (14.1)	2.7 (1.6–4.7)	2.6 (1.4–4.7)
LRA	102	204	21 (20.5)	14 (6.9)	7.0 (2.7–18.4)	7.0 (2.7–18.4)[2]
No adenoma	270	540	51 (18.9)	19 (3.5)	14.8 (6.3–35.1)	14.9 (6.5–34.0)
SSL	455	910	104 (22.9)	58 (6.4)	6.6 (4.4–10.1)	6.8 (4.2–11.0)
SSL ≥ 10 mm	217	434	63 (29.0)	25 (5.8)	13.3 (6.5–27.4)	17.9 (7.4–43.3)
Proximal[3]	209	418	63 (30.1)	22 (5.3)	18.6 (7.9–43.9)	25.0 (8.8–71.3)
SSL < 10 mm	238	476	41 (17.2)	33 (6.9)	3.3 (2.0–5.6)	3.2 (1.9–5.6)
Proximal[4]	209	418	39 (18.7)	29 (6.9)	4.4 (2.5–7.8)	4.3 (2.3–8.0)
SSL location[4]
Proximal	418	836	102 (24.4)	51 (6.1)	8.8 (5.5–14.2)	9.3 (5.4–15.9)
Distal	36	72	2 (5.6)	7 (9.7)	0.5 (0.1–4.4)5	0.5 (0.1–4.4)[2] , [5]
SSL with dysplasia	18	36	4 (22.2)	3 (8.3)	2.3 (0.6–9.3) 5	2.3 (0.6–9.3)[2] , [5]
SSL number[4]
1–2	416	832	90 (21.6)	51 (6.1)	6.7 (4.3–10.5)	6.8 (4.0–11.3)
≥ 3	38	76	13 (34.2)	7 (9.2)	5.8 (2.0–17.3)	5.8 (2.0–17.3)[2]

TMAN, total metachronous advanced neoplasia; HR, hazard ratio; HRA, high risk adenoma; LRA, low risk adenoma; SSL, sessile serrated lesion.

¹ Adjusted for index and follow-up bowel preparation quality, incomplete resection status, index indication, and cecal intubation rate.

² Univariate due to low case numbers in both groups.

³ No T-MAN in serrated lesion group for distal SSL ≥ 10 mm.

⁴ Missing data for one patient.

⁵ Not significant.

Fig. 2 Hazard ratios (with 95 %CIs) for total metachronous advanced neoplasia based on synchronous findings for the serrated lesion group compared with their matched patients without serrated lesions, adjusted for index and follow-up bowel preparation quality and cecal intubation rate. SL, serrated lesion; HRA, high-risk adenoma; LRA, low-risk adenoma.

One case of interval CRC was detected in the serrated lesion group. Characteristics of the patient with interval CRC can be found in Table 2 s.

Risk of T-MAN according to index serrated lesion characteristics

Overall, 22.9 % of patients with index SSLs had T-MAN compared with 6.4 % for the non-serrated lesion group (HR 6.8, 95 %CI 4.2–11.0). SSLs ≥ 10 mm were associated with significantly higher risk of T-MAN compared with the non-serrated lesion group (29.0 % vs. 5.8 %; HR 17.9, 95 %CI 7.4–43.3). SSLs < 10 mm were associated with significantly higher risk of T-MAN compared with the non-serrated lesion group (17.2 % vs. 6.9 %; HR 3.2, 95 %CI 1.9–5.6). Proximal SSLs of all sizes were associated with significantly higher risk of T-MAN compared with the non-serrated lesion group (24.4 % vs. 6.1 %; HR 9.3, 95 %CI 5.4–15.9); however, distal SSLs of all sizes were not (5.6 % vs. 9.7 %; HR 0.5, 95 %CI 0.1–4.4) ([Table 3], [Fig. 3]). When including sigmoid SSLs in the definition of distal, sensitivity analysis showed increased risk of T-MAN that became statistically significant.

Fig. 3 Hazard ratios (with 95 %CIs) for total metachronous advanced adenoma stratified by index sessile serrated lesion characteristics for the serrated lesion group compared with their matched patients without serrated lesions, adjusted for index and follow-up bowel preparation quality and cecal intubation rate. SSL, sessile serrated lesion.

SSLs with dysplasia had a higher percentage of T-MAN compared with the non-serrated lesion group (22.2 % vs. 8.3 %); however, this did not reach statistical significance. Index findings of ≥ 3 SSLs had higher risk of T-MAN compared with the non-serrated lesion group (34.2 % vs. 9.2 %; HR 5.8, 95 %CI 2.0–17.3).

Discussion

In this matched cohort study, we found that patients with serrated lesions had almost 6-times the risk of developing T-MAN compared with patients without serrated lesions. Furthermore, we found that finding synchronous serrated lesions and HRA at index colonoscopy was associated with 2.6-times the risk of T-MAN (32 %) compared with patients with HRA alone (14.1 %). Furthermore, we demonstrated significantly higher risk of T-MAN for proximal SSLs ≥ 10 mm and proximal SSLs < 10 mm, but not for distal SSLs of all sizes. These findings are important within the context of current CRC screening programs. With efforts being made toward complete polyp resection and adherence to guidelines as a means of reducing interval CRC, current guidelines are trending toward longer surveillance intervals for patients with index serrated lesions [18] [19]. The USMSTF 2020 guideline recommends 5–10-year surveillance intervals for detection of 1–2 SSLs and 3–5-year intervals for 3–4 SSLs, whereas a maximum of 5-year intervals were recommended for small serrated lesions in the 2012 guideline [10]. However, there was a distinct lack of data to support surveillance interval recommendations for ≥ 3 SSLs detected at index colonoscopy. In our study, patients with ≥ 3 index SSLs had higher rates of T-MAN when compared with patients with 1–2 index SSLs (34.2 % vs. 21.6 %); however, there were few patients with more than two index SSLs included in the study. T-MAN was mainly driven by the high rates of advanced serrated lesions for patients with ≥ 3 SSLs (28.9 % vs. 12.5 %) compared with the lower rates of metachronous advanced adenoma (7.9 % vs. 10.8 %) (Table 3 s). This underscores the findings that metachronous advanced adenoma alone may be inadequate to ascertain the entire risk of interval CRC in patients with index serrated lesions. The inclusion of advanced serrated lesions in the definition of advanced neoplasia can be an interesting proxy for the risk of CRC. Patients with serrated lesions are likely to develop further clinically significant serrated lesions at follow-up, including high-risk serrated lesions, which are very rarely encountered in patients with only index conventional adenomas [12].

Using metachronous advanced adenoma as a proxy for CRC risk without including metachronous advanced serrated lesions could therefore misjudge the risk of certain index findings in patients with serrated lesions. Serrated polyps can progress toward CRC through the serrated pathway with BRAF mutations as opposed to the conventional pathway [20]. This could explain why studies using proxies for CRC risk, such as “advanced adenoma at follow-up” have detected an increased risk for SSL + HRA combinations but failed to find increased risk for SSLs with no adenoma, whereas large studies in CRC patients do detect an increased risk. One study found significant increased likelihood of detecting HRA at follow-up colonoscopy after resection of synchronous HRA and SSLs at index (odds ratio 16.0) [13]. Large serrated polyps and serrated lesions without adenoma were not associated with increased risk for metachronous HRA; however, they were associated with increased risk for metachronous large serrated lesions [13]. A small prospective study (75 patients with serrated lesions) found higher incidence of metachronous HRA in the group with both serrated lesions and HRA at index compared with patients with HRA alone, but failed to find a difference in patients with serrated lesions + LRA or no adenoma [15]. A recent meta-analysis confirmed these findings in patients with index HRA using metachronous advanced adenoma as an outcome, but did not find any increased risk in patients with serrated lesions + LRA or no adenoma [21]. In contrast, a large study in patients with CRC did find increased CRC risk in patients with SSLs alone and SSLs + adenomas, including patients with unspecified serrated polyps [22]. Another study showed a similar trend toward increased CRC but did not investigate synchronous findings [23]. Our study confirms the findings in CRC studies, using T-MAN as a means to counter the limitations of the studies performed using traditional CRC risk proxies. We similarly found significantly increased risks in patients with LRA and no adenoma at index. Studying serrated lesion risk using CRC data is difficult as relatively few CRC cases are detected for a given patient group, limiting the ability to conduct analyses for certain index findings. Future studies should consider the use of T-MAN as an end point, particularly when substratifying index findings in patients with serrated lesions.

The optimal timing of follow-up colonoscopies after detection of serrated lesions remains controversial within guidelines. In contrast to the USMSTF guidelines, the 2020 ESGE guideline states that serrated lesions < 10 mm do not require further colonoscopy follow-up. It is interesting to note that patients in our study with SSLs < 10 mm had significant risk of T-MAN, driven by proximal SSLs < 10 mm (HR 4.3). There is conflicting evidence in the literature on the follow-up risks after detection of serrated lesions < 10 mm. One study including 56 low-risk SSLs found no metachronous advanced adenoma on follow-up [24]. A large study found no increased risk of CRC after detection of small serrated lesions; however, distal HPPs < 10 mm were included in the serrated lesion definition, with many HPPs included in the study, in contrast to another study finding a trend toward increased CRC for proximal small SSLs [22] [23]. There is increasing evidence that proximal small HPPs contribute toward metachronous advanced findings, with a recent study finding increased T-MAN for small SSLs and proximal HPPs, and another finding a trend toward increased CRC incidence for proximal small SSLs [22] [25]. Further studies are required to determine the true risk of small proximal HPPs and to define adequate follow-up durations for these findings. Our findings suggest that the risk of T-MAN in patients with serrated lesions is likely to be most affected by synchronous findings, size, and location of polyps, with synchronous HRA, large SSLs, proximal SSLs of all sizes, and SSL dysplasia driving most of the observed high risk. Modifications in recommended follow-up surveillance intervals are therefore needed, with location, size, and synchronous findings with index SSLs determining optimal follow-up instead of number of SSLs exclusively. This could improve stratification of high-risk patients who would undergo earlier surveillance and free-up endoscopy resources by reducing the follow-up burden of patients with low-risk SSLs.

The strength of our study lies in the inclusion of a large and adequately powered sample size of patients with serrated lesions, allowing for granular substratification of index findings. The relatively low number of T-MAN in the control group led to lower precision for 95 %CIs; however, many studies include HPPs < 10 mm or distal HPPs in their definition of serrated lesion, include very small (50–150) sample sizes for SSLs, and do not incorporate advanced serrated lesions into outcomes, which could limit the interpretation of their findings [12] [14] [15] [23] [25]. The use of T-MAN as a primary outcome more adequately reflects the CRC potential of serrated lesions and allowed us to demonstrate that many findings were less benign than previously suggested in studies only utilizing metachronous advanced adenoma as the outcome. The ability to capture all patients with serrated lesions during a 10-year period at a tertiary center with a large colonoscopy volume (> 10 000 per year) is also a strength. Patients in Quebec’s single payer healthcare system almost exclusively undergo follow-up at their original screening center; it is therefore unlikely that these patients had follow-up elsewhere.

There are however some weaknesses to our study. The retrospective nature of our study limits some of its interpretation as some patient and polyp data were missing; however, this affected a small minority of patients. The significant loss to follow-up is also a consideration in our study, although patient polyp and colonoscopy characteristics were not significantly different in our sensitivity analysis except for index HRA rates, a variable controlled for through matching, and index indication, with loss to follow-up having significantly higher proportion of diagnostic colonoscopies at index. The effects of loss to follow-up on the results of our study are therefore likely to be limited; however, it is possible that this population could present low-risk features not accounted for in our sensitivity analysis, which could explain the short follow-up duration in the included population. It is also important to note that most indications for index colonoscopy were for surveillance, which may signify that these patients are at increased risk for high-risk lesions at follow-up; however, this was controlled for in our analyses. We chose to limit our study to SSLs and TSAs, as identifying HPPs through our pathology department database would have yielded such a high number of patients as to render data collection unfeasible. For this same reason, and given the complexity of determining ADR for multiple periods and many endoscopists, we could not evaluate the impact of endoscopist ADR on outcomes at follow-up. Future studies should consider including this variable as a potential confounder. At our center, overall ADR was sampled over multiple periods and was found to be consistently > 30 % [26] [27] [28]. Pathology-based diagnosis of serrated lesions is also subject to variability in interobserver agreement, which limits the interpretation of all serrated lesion studies [29]. The inclusion of patients from 2010 when interest and knowledge of serrated lesions was lower would have likely led to some serrated lesions being diagnosed as HPPs and therefore not included in our study. However, when analyzing periods with low (2010–2013), moderate (2014–2017), and high (2018–2019) serrated lesion detection (Fig. 1 s), we found no differences in risk of T-MAN (Table 4 s).

In conclusion, serrated lesions were significantly associated with a higher risk of T-MAN. Index findings of serrated lesions regardless of synchronous findings significantly increased the risk of T-MAN. Patients with serrated lesions detected at index examination should receive earlier follow-up, especially those with large SSLs, proximal SSLs regardless of size, or synchronous HRA.

Acknowledgments

The authors would like to thank the American College of Gastroenterology for funding the study (ACG Medical Resident Clinical Research award). The authors also thank Maureen Fontaine, MSc, research coordinator, and Ekaterina Lebedeva, MSc, research assistant, for their help in conducting the study.

Competing Interests

R. Djinbachian was supported by a resident clinical research award from the American College of Gastroenterology for the conduction of this study. The study was performed independently from the funder. Daniel von Renteln is supported by a “Fonds de Recherche du Québec Santé” (FRQS) career development award and has received research funding from ERBE, Ventage, Pendopharm, and Pentax and is a consultant for Boston Scientific and Pendopharm. M.L. Lafontaine, J. Anderson, H. Pohl, T. Dufault, M. Boivin, and M. Bouin declare that they have no conflict of interest.

• References

• 1 Zauber AG, Winawer SJ, O’Brien MJ. et al. Colonoscopic polypectomy and long-term prevention of colorectal-cancer deaths. N Engl J Med 2012; 366: 687-696
• 2 Winawer SJ, Zauber AG, Ho MN. The National Polyp Study Workgroup. et al. Prevention of colorectal cancer by colonoscopic polypectomy. N Engl J Med 1993; 329: 1977-1981
• 3 Leddin D, Enns R, Hilsden R. et al. Colorectal cancer surveillance after index colonoscopy: guidance from the Canadian Association of Gastroenterology. Can J Gastroenterol 2013; 27: 224-228
• 4 Lieberman DA, Rex DK, Winawer SJ. et al. Guidelines for colonoscopy surveillance after screening and polypectomy: a consensus update by the US Multi-Society Task Force on Colorectal Cancer. Gastroenterology 2012; 143: 844-857
• 5 Hassan C, Quintero E, Dumonceau JM. et al. Post-polypectomy colonoscopy surveillance: European Society of Gastrointestinal Endoscopy (ESGE) Guideline. Endoscopy 2013; 45: 842-851
• 6 von Karsa L, Patnick J. European Colorectal Cancer Screening Guidelines Working Group. et al. European guidelines for quality assurance in colorectal cancer screening and diagnosis: overview and introduction to the full supplement publication. Endoscopy 2013; 45: 51-59
• 7 Rex DK, Ahnen DJ, Baron JA. et al. Serrated lesions of the colorectum: review and recommendations from an expert panel. Am J Gastroenterol 2012; 107: 1315-1329
• 8 Anderson JC, Butterly LF, Goodrich M. et al. Differences in detection rates of adenomas and serrated polyps in screening versus surveillance colonoscopies, based on the New Hampshire colonoscopy registry. Clin Gastroenterol Hepatol 2013; 11: 1308-1312
• 9 Liang J, Kalady MF, Appau K. et al. Serrated polyp detection rate during screening colonoscopy. Colorectal Disease 2012; 14: 1323-1327
• 10 Gupta S, Lieberman D, Anderson JC. et al. Recommendations for follow-up after colonoscopy and polypectomy: a consensus update by the US Multi-Society Task Force on Colorectal Cancer. Gastroenterology 2020; 158: 1131-1153
• 11 Hassan C, Antonelli G, Dumonceau JM. et al. Post-polypectomy colonoscopy surveillance: European Society of Gastrointestinal Endoscopy (ESGE) Guideline – Update 2020. Endoscopy 2020; 52: 687-700
• 12 Macaron C, Vu HT, Lopez R. et al. Risk of metachronous polyps in individuals with serrated polyps. Dis Colon Rectum 2015; 58: 762-768
• 13 Anderson JC, Butterly LF, Robinson CM. et al. Risk of metachronous high-risk adenomas and large serrated polyps in individuals with serrated polyps on index colonoscopy: data from the New Hampshire Colonoscopy Registry. Gastroenterology 2018; 154: 117-127
• 14 Melson J, Ma K, Arshad S. et al. Presence of small sessile serrated polyps increases rate of advanced neoplasia upon surveillance compared with isolated low-risk tubular adenomas. Gastrointest Endosc 2016; 84: 307-314
• 15 Pereyra L, Zamora R, Gómez EJ. et al. Risk of metachronous advanced neoplastic lesions in patients with sporadic sessile serrated adenomas undergoing colonoscopic surveillance. Am J Gastroenterol 2016; 111: 871-878
• 16 von Elm E, Altman DG, Egger M. et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007; 370: 1453-1457
• 17 Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med 2009; 28: 3083-3107
• 18 Djinbachian R, Iratni R, Durand M. et al. Rates of incomplete resection of 1- to 20-mm colorectal polyps: a systematic review and meta-analysis. Gastroenterology 2020; 159: 904-914
• 19 Djinbachian R, Dubé AJ, Durand M. et al. Adherence to post-polypectomy surveillance guidelines: a systematic review and meta-analysis. Endoscopy 2019; 51: 673-683
• 20 Leggett B, Whitehall V. Role of the serrated pathway in colorectal cancer pathogenesis. Gastroenterology 2010; 138: 2088-2100
• 21 Jung YS, Park JH, Park CH. Serrated polyps and the risk of metachronous colorectal advanced neoplasia: a systematic review and meta-analysis. Clin Gastroenterol Hepatol 2022; 20: 31-43
• 22 Li D, Doherty AR, Raju M. et al. Risk stratification for colorectal cancer in individuals with subtypes of serrated polyps. Gut 2021;
• 23 He X, Hang D, Wu K. et al. Long-term risk of colorectal cancer after removal of conventional adenomas and serrated polyps. Gastroenterology 2020; 158: 852-861
• 24 Jin EH, Seo JY, Bae JH. et al. Small sessile serrated polyps might not be at a higher risk for future advanced neoplasia than low-risk adenomas or polyp-free groups. Scand J Gastroenterol 2022; 57: 99-104
• 25 Hamoudah T, Vemulapalli KC, Alsayid M. et al. Risk of total metachronous advanced neoplasia in patients with both small tubular adenomas and serrated polyps. Gastrointest Endosc 2022; 96: 95-100
• 26 Djinbachian R, Marchand E, Pohl H. et al. Optical diagnosis of colorectal polyps: a randomized controlled trial comparing endoscopic image-enhancing modalities. Gastrointest Endosc 2021; 93: 712-719
• 27 Taghiakbari M, Pohl H, Djinbachian R. et al. The location-based resect and discard strategy for diminutive colorectal polyps: a prospective clinical study. Endoscopy 2022; 54: 354-363
• 28 von Renteln D, Djinbachian R, Benard F. et al. Incomplete resection of 4–20 mm colorectal polyps when using cold snare and associated factors. Endoscopy 2022;
• 29 Vennelaganti S, Cuatrecasas M, Vennalaganti P. et al. Interobserver agreement among pathologists in the differentiation of sessile serrated from hyperplastic polyps. Gastroenterology 2021; 160: 452-454

Publikationsverlauf

Eingereicht: 25. April 2022

Angenommen nach Revision: 26. Januar 2023

Accepted Manuscript online:
26. Januar 2023

Artikel online veröffentlicht:
09. März 2023