Introduction
The definition of attenuated adenomatous polyposis has traditionally been based on
the occurrence of 10–99 cumulative lifetime adenomas in the colorectum [1 ]. This definition, regardless of family history, is known to cover a wide spectrum
of situations, from patients with multiple sporadic adenomas to those with genetically
inherited polyposis syndromes due to APC or MUTYH pathogenic germline variants [2 ].
Current guidelines recommend that patients with 10 or more cumulative adenomatous
colorectal polyps should receive genetic counseling, as well as intensive endoscopic
surveillance in specialized colorectal cancer (CRC) high risk units [1 ]
[3 ]
[4 ]
[5 ]
[6 ]
[7 ]
[8 ]
[9 ]. Given the growing volume of screening colonoscopies and the increasing improvements
in technology and polyp detection [10 ]
[11 ], more and more individuals are being diagnosed with multiple colorectal adenomas,
representing a meaningful burden for high risk clinics [12 ].
The diagnostic yield of genetic testing in this population (especially in those with
10–19 adenomas) is very low [2 ]
[13 ]. Moreover, in the setting of organized CRC screening programs aimed at individuals
≥ 50 years of age without a relevant family history of this neoplasia, the yield of
germline genetic testing is likely to be even lower. With this in mind, Spanish guidelines
for CRC diagnosis and prevention establish the lower limit for genetic testing as
20 lifetime adenomas, except in those cases diagnosed at a young age or where a family
history of CRC or polyposis is documented [14 ].
On the other hand, there is scarce information to support the efficacy of intensive
surveillance in patients with non-hereditary attenuated adenomatous polyposis. Some
of these patients may carry a low risk of metachronous neoplasia and are currently
over-surveilled, representing a significant burden on endoscopy units.
The prevalence of patients with ≥ 10 adenomas and the proportion of inherited polyposis
syndromes in fecal immunochemical test (FIT)-based CRC screening programs remain unclear
[15 ]. More importantly, for those patients without a hereditary cause, the risk of metachronous
neoplasia has been poorly described. Given the above mentioned uncertainties, the
aims of the current study were: (i) to evaluate the prevalence of patients with ≥ 10
adenomas in a FIT-based CRC screening program; (ii) to describe the prevalence of
hereditary syndromes; and (iii) to investigate the incidence of CRC and advanced neoplasia
during follow-up of non-inherited cases.
Methods
Study design and patients
This observational retrospective cohort study included individuals from the FIT-based
organized Barcelona–Eixample–Esquerra population CRC Screening Program, in which all
individuals aged 50–69 are invited to participate. A personal history of CRC, adenoma,
or inflammatory bowel disease; a family history of CRC (defined as those individuals
with two first-degree relatives [FDRs] with CRC or one diagnosed before the age of
60); known hereditary CRC syndromes; severe coexisting illness; colonoscopy performed
within the last 5 years; previous colectomy; or contraindication for colonoscopy are
considered definitive or temporary exclusion criteria for screening.
For the present study, participants from the first to fourth screening round of the
program (from January 2010 to June 2018) who had a positive FIT result (cutoff ≥ 20
μg hemoglobin/g feces) followed by a complete colonoscopy with ≥ 10 adenomas found
and who had undergone at least one surveillance colonoscopy at an interval consistent
with current recommendations were considered eligible to enter the study [3 ]
[4 ]
[9 ]
[14 ]. Individuals with ≥ 10 adenomas who met either the WHO 2010 or 2019 criteria of
serrated polyposis syndrome (SPS) [16 ] were also excluded.
All colonoscopies were performed at the Hospital Clinic of Barcelona, a tertiary academic
center that follows high quality standards [17 ]
[18 ]. All participating endoscopists had an adenoma detection rate ≥ 40 % in FIT-positive
patients. In the setting of the screening program, all colonoscopies and pathology
reports were reviewed weekly by a committee composed of expert gastroenterologists,
endoscopists, and nurses before follow-up recommendations were given.
Participants’ baseline demographic data were prospectively recorded in the CRC screening
program database. Other clinical data, such as cigarette smoking history, body mass
index (BMI) or cardiovascular disease (i. e. diabetes, hypertension, dyslipidemia,
ischemic heart disease), were obtained, when available, from hospital medical records
and Catalonia’s National Health Service database.
Only histologically confirmed adenomas were counted for diagnosis. Patient selection
was based on endoscopic and histopathological reports from all polyps removed at baseline
screening colonoscopy.
Patients with ≥ 10 adenomas were offered an appointment at the CRC high risk unit.
Based on Spanish Guidelines, genetic testing was offered in the following situations:
(i) individuals with ≥ 20 lifetime adenomas; (ii) individuals with ≥ 10 adenomas detected
before the age of 40; (iii) individuals with ≥ 10 adenomas and a personal or family
history of CRC before the age of 60; (iv) ≥ 10 adenomas and a family history of adenomatous
polyposis [14 ].
Germline genetic testing was carried out by Sanger sequencing and multiplex ligation-dependent
probe amplification (MLPA) of APC and MUTYH genes until 2016, and afterwards by a multigene cancer panel that included the APC and MUTYH genes (Trusight Cancer v1; Illumina Inc., San Diego, California, USA). All germline
genetic reports from both the hospital medical records and the high risk clinic’s
database were reviewed. Patients with germline pathogenic variants were excluded for
the analysis of the risk of CRC and advanced neoplasia during surveillance.
The study was approved by the institutional review board (IRB) of our Institution
and was carried out respecting the fundamental principles established in the Declaration
of Helsinki.
Endoscopic characteristics and treatment
Data on the baseline and surveillance colonoscopies, such as the number and size of
adenomas and serrated lesions detected, and the grade of dysplasia, were retrieved
from endoscopy and pathology reports.
When several baseline colonoscopies were needed, for example for inadequate bowel
preparation, a high burden of polyps, or complex polypectomy, the baseline colonoscopy
was set as the date of the last colonoscopy that completely scrutinized the entire
colon and/or achieved a clear colon. For analysis purposes, the findings of any repeated
colonoscopies were compiled into one and the date of the last colonoscopy was used.
After a first surveillance colonoscopy, all consecutive surveillance procedures performed
were counted until October 2019. Incomplete procedures and those with inadequate colon
cleansing, as well as those performed outside of the recommended interval, were excluded.
The type of any colorectal surgery and cause for referral were documented. Patients
referred for surgery after baseline colonoscopy owing to CRC, an unresectable polyp,
or the polyp burden were included if a segmental colectomy had been performed, while
those who underwent a total colectomy were excluded.
Data was registered and stored in an anonymous database.
Histopathological records
Polyp histology was evaluated by expert pathologists dedicated to gastrointestinal
oncology. Adenomas were histologically classified according to the Vienna classification
[19 ]. Advanced adenomas were defined as adenomas ≥ 1 cm in size, and/or with villous
component, and/or showing high grade dysplasia. Serrated lesions were classified as
hyperplastic polyps, sessile serrated lesions [SSLs], and traditional serrated adenomas
[TSAs], based on the current WHO classification criteria [20 ]. Cytological dysplasia among serrated lesions was analyzed both as the presence/absence
of dysplasia, and as the presence of low and high grade dysplasia. Advanced serrated
lesions were defined as serrated lesions ≥ 1 cm in size, and/or with dysplasia (i. e.
SSLs with dysplasia and TSAs). Neoplastic extension vertically into the submucosal
layer or beyond was classified as invasive cancer. Advanced neoplasia was defined
as CRC, advanced adenoma, or advanced serrated lesion.
Outcome measures during surveillance
The main outcome was the detection during surveillance of: (i) CRC; and (ii) advanced
neoplasia. CRCs detected at baseline screening colonoscopy were classified as prevalent
and those diagnosed during surveillance as post-colonoscopy CRC. Post-colonoscopy
CRC was defined as a CRC diagnosed after the performance of a colonoscopy without
cancer. Following World Endoscopy Organization (WEO) consensus, post-colonoscopy CRC
was subdivided into “interval CRC” (detected before the recommended surveillance interval)
and “non-interval CRC” (detected at the recommended surveillance interval [non-interval
type A], after the recommended surveillance interval [non-interval type B], or when
no surveillance interval had been recommended [non-interval type C]) [21 ].
Statistical analysis
A description of colorectal lesions per patient identified at baseline and during
colonoscopy surveillance is presented (per-patient analysis). Percentages were used
for categorical data, using median and interquartile range (IQR) for non-normally
distributed variables, and mean and SD for normally distributed variables. When information
was missing, the denominator accounted for patients with available data.
The cumulative incidences of CRC, advanced adenoma, and advanced neoplasia were calculated
by Kaplan-Meier survival analysis. Because the surveillance times were different for
each patient, the incidence rate, expressed in new cases per 100 person-years under
surveillance, was also calculated to describe the risk of post-colonoscopy CRC and
metachronous lesions. The denominator of this rate was obtained from the sum of the
time each person was observed, totaled for all persons. This denominator represents
the total time the population was at risk of and being watched for each event.
We performed bivariable analyses to explore baseline factors potentially associated
with metachronous advanced neoplasia at first surveillance endoscopy. Quantitative
variables were analyzed using Student’s t test, and qualitative variables were analyzed using the chi-squared test. Next, a
multivariable logistic regression analysis was performed, including clinically relevant
variables and those with P values of < 0.10 obtained on bivariable analysis. We included odds ratios (ORs) with
95 %CIs to quantify the magnitude of the association. Statistical analysis was performed
using SPSS 22.0 (IBM Corp., Armonk, New York, USA).
Results
Prevalence of adenomatous polyposis and hereditary syndromes
A total of 215 of 9582 FIT-positive individuals (2.2 %) had 10–99 adenomas on their
baseline screening colonoscopy ([Fig. 1 ]). Of these, 23 patients were excluded owing to incomplete surveillance data, but
this group did not differ from the patients who did undergo surveillance (Table 1s , see online-only Supplementary material).
Fig. 1 Study flowchart showing the inclusion and exclusion of individuals from a fecal immunochemical
test (FIT)-positive colorectal cancer (CRC) screening program, which resulted in data
being analyzed from a final group of 161 patients.
Germline genetic analysis was performed in 60 cases (27.9%), comprising 57 of 62 individuals
with ≥ 20 adenomas (92 %) and three patients with 10–19 adenomas. The indications
for germline testing in the latter group were: a family history of CRC in a second-degree
relative younger than 60 years (n = 2); and a family history of CRC in an FDR < 60
years who was diagnosed after the inclusion of the patient in the screening program
(n = 1).
Pathogenic mutations were found in two of the 60 individuals (3.3 %): a biallelic
pathogenic variant in the MUTYH gene – Y179C (573A > G) and G396 D (1187 G > A) – detected in a 57-year-old woman
with 52 adenomas; and a pathogenic variant in the APC gene (c.6742A > T; P . K2248X) found in a 68-year-old man with > 70 adenomas.
Study population baseline characteristics
A total of 161 patients with at least one surveillance colonoscopy and no suspicion
of a hereditary syndrome were included as the study population ([Fig. 1 ]). The median age at index colonoscopy was 61 years, with a male predominance (124/161
[77 %]). With regard to environmental risk factors, 65 % (87/134) were either active
(60/134 [44.8 %]) or former (27/134 [20.1 %]) smokers, and 82 % of patients (132/161)
had at least one cardiovascular risk factor (including being overweight, or having
dyslipidemia, hypertension, diabetes, or a history of ischemic heart disease). Detailed
demographic and clinical characteristics are represented in [Table 1 ].
Table 1
Clinical features and colonic phenotype in the 161 patients with ≥ 10 adenomas.
Demographic and clinical features
Age at baseline colonoscopy, median (IQR), years
61 (57–65)
Sex, female, n (%)
37/161 (23.0)
BMI, median (IQR), kg/m2
28.7 (26.1–32.4)
Cardiovascular risk factors present, n (%)
87/134 (64.9)
44/53 (83.0)
40/137 (29.2)
60/140 (42.9)
66/141 (46.8)
Ischemic heart disease, n (%)
11/132 (8.3)
Any cardiovascular risk factor[1 ], n (%)
132/161 (82.0)
Family history of colorectal cancer in a first- or second-degree relative, n (%)
27/161 (16.8)
Personal history of any extracolonic cancer, n (%)
23/161(14.3)
6 (26.1 %)
5 (21.7 %)
3 (13.0 %)
2 (8.7 %)
1 (4.3 %)
6 (26.1 %)
Endoscopic phenotype at index colonoscopy (per patient)
Colorectal cancer, n (%)
8/161 (5.0)
At least one advanced adenoma, n (%)
128/161 (79.5)
Any serrated lesion, n (%)
84/161 (52.2)
At least one sessile serrated lesion, n (%)
34/161 (21.1)
At least one advanced serrated lesion, n (%)
17/161 (10.6)
At least one advanced neoplasia, n (%)
130/161 (80.7)
IQR, interquartile range; BMI, body mass index.
1 Cardiovascular risk factors: diabetes, dyslipidemia, hypertension, smoking history,
and BMI ≥ 25 kg/m2 .
A family history of an FDR with CRC ≤ 60 years was reported in only one case, which
occurred after the patient had entered the screening program. In 14 individuals (8.7 %),
a CRC history in FDRs older than 60 was notified. A personal history of extracolonic
malignancy was reported in 23 individuals (14.3 %), with lung and prostate cancers
being the most frequent.
Findings at baseline colonoscopy
At baseline colonoscopy, a median of 10 adenomas (IQR 10–13) and one serrated lesion
(IQR 0–2) were detected per patient. The majority of patients had 10–19 adenomas (147;
91.3 %), 13 (8.0 %) had 20–29 adenomas, and one patient (0.6 %) had 30 adenomas. Advanced
adenomas, advanced serrated lesions, and advanced neoplasia were found in 128 (79.5 %),
17 (10.6 %), and 130 patients (80.7 %), respectively ([Table 2 ]).
Table 2
Colorectal lesions identified during colonoscopy surveillance (per-patient analysis).
Index colonoscopy
Surveillance colonoscopy number
1
2
3
4
5
Number of patients
161
161
106
62
43
34
Time since previous colonoscopy, median (IQR), months
–
14 (12–17)
13 (12–24)
13 (12–18.5)
13 (12–20.7)
15 (12–24)
Patients with invasive CRC[1 ]
8 (5.0)
0 (0)
0 (0)
0 (0)
1 (2.3)
0 (0)
Patients with ≥ 1 adenoma, n (%)
161 (100)
122 (75.8)
82 (78.8)
49 (79.0)
26 (60.0)
12 (35.3)
Number of adenomas, total; median (IQR)
1946; 10 (10–13)
461; 2 (1–4)
315; 2 (1–4)
200; 2 (1–4)
85; 1 (1–3)
38; 1 (0–2.7)
Patients with ≥ 10 adenomas, n (%)
161 (100)
8 (5.0)
4 (3.8)
3 (4.8)
1 (2.3)
0 (0)
Patients with at least one advanced adenoma, n (%)
128 (79.5)
35 (21.7)
14 (13.3)
9 (14.5)
1 (2.3)
0 (0)
Patients with at least one adenoma ≥ 1 cm, n (%)
124 (77)
25 (15.5)
13 (12.5)
5 (8.1)
1 (2.3)
0 (0)
Patients with at least one adenoma ≥ 2 cm, n (%)
46 (28.6)
6 (3.1)
1 (0.9)
1 (1.6)
0 (0)
0 (0)
Patients with at least one serrated lesion, n (%)
84 (52.2)
62 (38.5)
45 (43.3)
25 (40.3)
12 (27.9)
5 (141)
Patients with at least one SSL, n (%)
34 (21.1)
10 (6.2)
12 (11.3)
6 (9.6)
1 (2.3)
3 (8.8)
Patients with at least one advanced serrated lesion, n (%)
17 (10.6)
7 (4.3)
3 (2.9)
3 (4.8)
0 (0)
0 (0)
Patients with at least one advanced neoplasia, n (%)
130 (80.7)
39 (24.2)
17 (16.2)
9 (14.5)
1 (2.3)
0 (0)
CRC, colorectal cancer; SSL, sessile serrated lesion.
1 Only endoscopic data from surveillance procedures are shown in this table. For this
reason, only one of two post-colonoscopy CRCs is included (the other case was diagnosed
at a colonoscopy performed because of symptoms).
Prevalent CRCs were detected in eight patients (5.0 %) at a median age of 62.5 years
(range 51–67), with six patients (75%) being male. Most tumors were diagnosed at early
stages (six stage I–II [75 %]), including four pT1 CRCs. Two of these pT1 CRCs (50 %)
were able to be successfully managed by endoscopic resection owing to the absence
of pathological risk factors for lymph node metastasis.
With regard to the clinical management, 10 patients (6.2 %) were referred for surgery
after their baseline colonoscopy because of either severe polyposis (n = 1), an unresectable
polyp (n = 2), or CRC (n = 6); the remaining patient underwent surgery owing to a
post-colonoscopy perforation (n = 1). Right hemicolectomy was performed in six patients
and sigmoidectomy in four patients.
Findings during surveillance
Polyposis phenotype
A total of 427 surveillance colonoscopies (median 2 per patient [IQR 2–4]) were performed
at a median follow-up time of 3 years (IQR 1–5). The median interval time between
procedures was 14 months (IQR 12–17). Only eight patients (5.0 %) showed ≥ 10 adenomas
at first surveillance colonoscopy. Cumulatively, 110 individuals (68.3 %) displayed
10–19 adenomas and 51 (31.7 %) developed ≥ 20 adenomas. None of the 161 patients displayed
a classic phenotype (≥ 100 adenomas).
Incidence of post-colonoscopy colorectal cancer
During follow-up, 48 individuals (29.8 %) developed advanced adenomas (3-year cumulative
incidence 15.7 %, 95 %CI 12.3 %–19.1 %); and nine (5.6 %) developed advanced serrated
lesions (3-year cumulative incidence 5.0 %, 95 %CI 3.0 %–7.2 %). The corresponding
incident rate figures were 8.3 and 1.5 new cases per 100 person-years under surveillance
for advanced adenomas and advanced serrated lesions, respectively.
Two CRCs were diagnosed during surveillance (3-year cumulative incidence 1.0 %, 95 %CI
0.89 %–1.9 %); incidence rate 0.3 new cases per 100 person-years under surveillance.
In one case, an interval-type post-colonoscopy CRC, TNM stage IIIa (T3N2M0), was detected
in a 69-year-old man who had undergone colonoscopy because of weight loss and increased
carcinoembryonic antigen 14 months after a previous high quality surveillance procedure.
An ulcerated lesion of 5 cm in size, with infiltrative appearance, was observed in
the hepatic flexure. In this patient, 11 non-advanced adenomas had been previously
removed: 10 at baseline colonoscopy and only one diminutive adenoma in the previous
surveillance procedure. The other case of CRC was a pT1 CRC detected in a 63-year-old
woman who had accumulated 60 adenomas over 7 years. A slightly elevated polyp of 1 cm
was detected over a scarred base in the sigmoid colon during scheduled surveillance
colonoscopy (a non-interval type A post-colonoscopy CRC).
Incidence of advanced neoplasia in subsequent colonoscopies
With regard to the incidence of polyps during follow-up, 122 patients (75.8 %) had
adenomas at their first surveillance colonoscopy, whereas 39 (24.2 %) showed no polyps.
Advanced neoplasia during follow-up was detected in 52 patients (32.3%) The 3-year
cumulative incidence rate was 16 % (95 %CI 12.6 %–19.4 %), with an incidence rate
of nine new cases per 100 person-years under surveillance, the great majority of these
being found at the first surveillance colonoscopy (39 [75%]). In 28 patients (53.8 %),
advanced neoplasia was found only once during surveillance.
The proportion of patients with advanced adenomas, advanced serrated lesions, and
advanced neoplasia progressively decreased throughout successive follow-up colonoscopies.
The per-patient distribution of lesions at each surveillance colonoscopy is shown
in [Table 2 ] and [Fig. 2 ].
Fig. 2 Colorectal lesions identified during colonoscopy surveillance (per-patient analysis).
In terms of the clinical management, six patients (3.7 %) were referred for surgery
during surveillance: four for unresectable polyps and two because of CRC.
Risk factors of advanced neoplasia at first surveillance colonoscopy
At first surveillance colonoscopy, advanced neoplasia was diagnosed in 39 patients
(24.2 %). Clinical and phenotypical characteristics of the patients with and without
advanced neoplasia at first surveillance colonoscopy are summarized in [Table 3 ]. Bivariable analysis and subsequent multivariable regression analysis revealed that
the presence of an advanced adenoma at baseline was independently associated with
a higher risk of advanced neoplasia at first surveillance colonoscopy (OR 3.91, 95 %CI
1.12–13.62; P = 0.03).
Table 3
Factors associated with advanced neoplasia at first surveillance colonoscopy in patients
with ≥ 10 baseline adenomas on bivariable and multivariable analysis.
Advanced neoplasia at first surveillance colonoscopy
No (n = 122)
Yes (n = 39)
P value
Adjusted odds ratio (95 %CI)
Adjusted P value
Baseline characteristics
Age at baseline colonoscopy, median (IQR), years
61 (56.6–65)
62 (59–65)
0.23
Sex, female, n (%)
28 (22.9)
9 (23)
0.98
BMI, median (IQR), kg/m2 (n = 58)
29.4 (26.8–33.6)
26.8 (24.8–31.5)
0.19
Cardiovascular risk factors present, n (%)
Smoking history
66 (54.1)
21 (53.8)
0.65
Overweight/obese (BMI ≥ 25 kg/m2 )
31 (86)
13 (33.3)
0.38
Diabetes (n = 154)
31 (29.5)
9 (25.5)
0.87
Dyslipidemia (n = 159)
46 (43.4)
14 (35.8)
0.82
Hypertension (n = 159)
47 (43.9)
19 (48.7)
0.22
Ischemic heart disease, n (%) (n = 150)
9 (9.0)
2 (5.1)
0.62
Any cardiovascular risk factor, n (%)
101 (82.7)
31 (79.4)
0.64
Family history of CRC, n (%)
19 (15.5)
8 (20.5)
0.47
Personal history of extracolonic tumor, n (%)
18 (14.2)
5 (12.8)
0.76
Phenotype at the index colonoscopy
Number of adenomas, median (IQR)
10 (10–13)
10 (10–13)
0.97
Patients with at least one advanced adenoma; n (%)
92 (75.4)
36 (92.3)
0.03
3.91 (1.12–13.62)
0.03
Patients with at least one adenoma ≥ 1 cm, n (%)
90 (73.7)
34 (87.1)
0.08
0.54 (0.09–2.97)
0.48
Patients with at least one adenoma ≥ 2 cm, n (%)
33 (27.0)
13 (33.3)
0.45
Patients with at least one serrated lesion, n (%)
59 (48.3)
25 (64.1)
0.08
1.79 (0.84–3.81)
0.13
Patients with at least one SSL, n (%)
24 (19.6)
10 (25.6)
0.42
Patients with at least one advanced serrated lesion, n (%)
12 (9.8)
5 (12.8)
0.59
Patients with advanced adenoma and advanced serrated lesions, n (%)
11 (9.0)
5 (12.8)
0.48
Interval time between procedures, median (IQR), months
14 (12–17)
14 (12–17)
0.65
IQR, interquartile range; BMI, body mass index; CRC, colorectal cancer; SSL, sessile
serrated lesion.
Discussion
In this study, we have shown that the prevalence of ≥ 10 adenomas (i. e. adenomatous
polyposis) in a FIT-based CRC screening program is 2.2 %, with a very low frequency
of germline pathogenic variants among the patients with an indication for germline
genetic testing. A considerable percentage of individuals (32 %) develop advanced
neoplasia during follow-up, most of them (75 %) found at the first surveillance colonoscopy.
On the other hand, post-colonoscopy CRC is rare (1 %) in our series and both the percentage
of patients with advanced and non-advanced neoplasia tends to progressively decrease
throughout the successive follow-up colonoscopies. Overall, our results suggest that
adenomatous polyposis within a FIT-based program is mainly sporadic and surveillance
intervals could probably be lengthened after the first surveillance colonoscopy.
This work is, to our knowledge, the first study that describes the rate of polyposis
within a FIT-based CRC screening cohort. Our results suggest a minor occurrence of
hereditary polyposis in CRC screening populations, given the low rate of germline
pathogenic variants detected. However, germline genetic testing was not performed
in the whole cohort, in line with the current guidelines. Moreover, only adenomatous
polyposis genes were investigated in the majority of cases (mainly APC and MUTYH ). Previous evidence has suggested that the diagnostic yield of testing for pathogenic
variants in patients with 10–19 adenomas is low and always influenced by a referral
bias [2 ].
Recently, Stanich et al. [13 ] described the prevalence of polyposis- and CRC-associated inherited gene mutations
in patients with ≥ 10 colorectal polyps (including adenomas and hamartomas) who underwent
multigene panel testing. Within the adenoma cohort (almost 3200 patients [median age
58.7 years]), the prevalence of pathogenic mutations in adenomatous polyposis genes
(APC, bi-allelic MUTYH, POLE, POLD1 ) was only 2.3 % in individuals with 10–19 adenomas, 8.5 % in those with 20–99 adenomas,
and 41 % in patients with ≥ 100 adenomas. Nevertheless, a key limitation of the Stanich
et al. study was the referral bias, because patients with ≥ 10 polyps were selected
from a database of individuals who had undergone genetic testing for several reasons
(personal and/or family history of CRC or other tumors), meaning not all patients
with ≥ 10 colorectal adenomas were tested. Accordingly, considering the average-risk
population of our study, we expected the true prevalence of inherited syndromes to
be very low. Furthermore, we found a strong association with clinical and environmental
factors in our cohort. A clear predominance of male sex (77 %) was observed and both
smoking history and cardiovascular risk factors (65 % and 82 %, respectively) were
common, reinforcing the involvement of these factors in colorectal carcinogenesis
[14 ].
Our results provide evidence reinforcing the recommendation of a 1-year surveillance
colonoscopy in individuals with ≥ 10 adenomas within a FIT-based screening program.
Nearly 25 % of this population displayed an advanced neoplasia at 1 year, and the
3-year risk of CRC was up to 1 % (0.3 new cases per 100 person-years under surveillance).
A previous Korean study evaluated the incidence of advanced adenoma in a cohort of
214 individuals with ≥ 10 adenomas within a colonoscopy-based screening program [22 ]. In this work, no post-colonoscopy CRC was observed and a 3-year cumulative incidence
of 7 % was reported for advanced adenoma. This lower percentage could be explained
because, in the Asian study, lesions detected within 2.5 years following the index
colonoscopy were counted as prevalent and merged to the date of screening colonoscopy.
Therefore, the authors only considered advanced adenoma as incident when it occurred
afterwards.
Taking into account our results, the consideration of individuals with ≥ 10 adenomas
as a high risk population, after (if indicated) hereditary causes have been ruled
out, is basically justified by the meaningful risk of advanced neoplasia at the 1-year
surveillance colonoscopy. However, it should be kept in mind that the risk of advanced
neoplasia remains high (> 10 %) until the third surveillance colonoscopy. Afterwards,
the risk of advanced neoplasia considerably decreases, suggesting that surveillance
intervals could be lengthened.
These data are interesting as no surveillance recommendation has been established
in patients with non-hereditary polyposis after the first surveillance colonoscopy.
Actually, in patients with < 10 adenomas who require surveillance, evidence of the
benefit of a second surveillance colonoscopy in terms of CRC risk is still unclear.
Even in the updated European guidelines, for the high risk group (≥ 5 polyps or size
≥ 2 cm), surveillance recommendations after the first surveillance colonoscopy are
based on low quality evidence [9 ]. Our results suggest that, after the first surveillance colonoscopy, the follow-up
could be reassessed on the basis of what was found during that examination.
In our work, the presence of an advanced adenomas at baseline was a clear independent
risk factor for developing advanced neoplasia at first surveillance colonoscopy. This
finding is not unexpected, taking into account that both the incidence and mortality
of CRC is higher in individuals with advanced adenomas [23 ]
[24 ]. In our opinion, given the scarce evidence to tailor colonoscopy surveillance intervals
in patients with ≥ 10 adenomas, this factor becomes quite relevant and could be considered
in the design of future prospective studies comparing different personalized surveillance
strategies.
Lastly, no association was observed in relation to the number of baseline adenomas,
consistent with the current evidence suggesting the minor role of multiplicity by
itself in post-colonoscopy CRC risk [23 ]
[24 ]
[25 ]. This is an important point owing to increasing colonoscopy screening activity and
the improvement in endoscopy equipment and ancillary techniques, which result in more
and more individuals who are diagnosed with many diminutive polyps and are then referred
for surveillance.
Our study has several strengths. First, this is the first work to estimate the prevalence
of inherited and non-inherited adenomatous polyposis syndromes within a FIT-based
CRC screening program; second, although retrospective, we focused on an unselected
population of individuals with a shared surveillance program.
Nevertheless, several limitations should be acknowledged. First, as pointed out previously,
germline genetic analysis testing was not performed in the whole cohort, so the percentage
of patients with hereditary syndromes could have been underestimated. It is important
to note that the figures observed in our cohort cannot be generalized to the average
risk population, because they are restricted to FIT-based CRC screening cohorts.
Second, the number of patients decreased from the second surveillance colonoscopy
onwards, so the incidence of lesions described during follow-up could be overestimated.
To minimize this bias, only the findings at the first surveillance colonoscopy were
included when analyzing the potential risk factors of advanced neoplasia. Taking into
account that 75 % of patients who developed advanced neoplasia in the follow-up did
so by their first surveillance colonoscopy, our observations are probably an adequate
estimate of reality.
Finally, advanced neoplasia was selected as an end point instead of CRC owing to the
small number of post-colonoscopy CRCs detected. Nevertheless, surveillance should
aim to prevent, rather than detect, CRCs and the recent recommendations from the WEO
[10 ] recognized the need for possible surrogate measures in surveillance studies. In
this regard, the rate of “advanced colorectal polyps” (defined as an advanced adenoma
or advanced serrated lesion) represents an acceptable surrogate outcome that is less
prone to overdiagnosis or lead-time bias, as compared with using any polyp as an outcome
[12 ]. Given all of this, we think that there is enough evidence to support advanced neoplasia,
as a precursor condition of invasive cancer, being a good surrogate end point.
In conclusion, in a FIT-based CRC screening scenario, 10 or more adenomas are found
in a small proportion of patients and inherited adenomatous polyposis syndromes seem
to be rare within this population. A low rate of post-colonoscopy CRC is observed,
but there is a substantial risk of advanced neoplasia, especially at the first surveillance
colonoscopy in those individuals with advanced adenomas on their baseline colonoscopy.
It is important to point out that, after the first surveillance colonoscopy, the proportion
of patients with advanced neoplasia tends to progressively decrease through successive
follow-up colonoscopies.
Based on our findings, in patients with ≥ 10 baseline adenomas, we therefore recommend
genetic counseling to assess the indication for genetic testing and 1-year interval
endoscopic surveillance. Once an inherited cause has been ruled out, the following
surveillance intervals may be based on the findings of each successive colonoscopy,
as is recommended for those patients with < 10 adenomas. Future studies should focus
on patients with ≥ 10 adenomas undergoing standardized protocols in order to define
the best management and surveillance strategies for these patients.