Surveillance Colonoscopies of Synchronous Colorectal Cancer: What Should We Do?

Background: The aim of this study is to investigate the occurrence of metachronous neoplasms at 2-year surveillance colonoscopy for synchronous colorectal cancer patients and the relative risk factors. Methods: Synchronous colorectal cancer patients who underwent surgery or endoscopic resection for colorectal cancer between January 2008 and December 2019 were enrolled. All patients underwent surveillance colonoscopies at least twice within 2 years after operation. Univariate and multivariate analyses were conducted to assess the risk factors for the metachronous neoplasms. Results: Totally 38 patients (male/female: 26/12) were included, with an average age of 64.6 years (±11.5 years) and a mean surveillance interval of 23.47 ± 4.39 months. In 21 of 38 patients (55.3%), metachronous adenoma was detected, including 6 metachronous advanced adenomas. Two patients were detected with metachronous carcinomas. In univariate analysis, male sex, elderly age at diagnosis, and the presence of synchronous adenomas/synchronous advanced adenoma at baseline colonoscopy were associated with the development of metachronous adenoma (P = .037, .047, .013, .039), but not associated with metachronous advanced adenoma (P = 0.455, .746, .503, .269). Patients tends to occur less metachronous advanced adenoma if index colorectal tumors were treated by endoscopic resection (P = .010), but the tendency was not discovered in metachronous adenoma (P = .289). Tumor location (with/without rectum cancer) was not associated with the development of metachronous lesions (P = .526, .382). On multivariate analysis, the presence of synchronous adenomas at baseline colonoscopy was an independent risk factor for MA during follow-up (odds ratio = 15.0; 95% CI: 1.55-145.22). Conclusion: For postoperative synchronous colorectal cancer patients, doctors should design individual surveillance strategies according to sex, baseline colonoscopy, and operative (or endoscopic) approach of resection.


INTRODUCTION
Colorectal cancer (CRC) is the third most common cancer worldwide. 1 Synchronous colorectal carcinoma (SCRC) refers to more than 1 primary colorectal carcinoma detected in a single patient at initial presentation. The prevalence of SCRC is approximately 3.5% of all colorectal carcinomas. 2 Over recent years, with the improvement of colonoscopy technology, many early-stage colorectal cancers were diagnosed and treated by endoscopy (endoscopic mucosal resection [EMR] or endoscopic submucosal dissection [ESD]). Studies showed that, different from solitary colorectal carcinoma, SCRC was significantly associated with poor prognosis. 2 Though postoperative endoscopic surveillance has become a routine part of clinical practice, studies are lacking for those SCRC patients. Considering that most epidemiological studies indicate a higher incidence of metachronous adenoma (MA) in the initial 2-3 years following surgery, and significantly decreased thereafter, 3,4 we aimed to assess the occurrence of metachronous neoplasms at 2-year surveillance colonoscopy for SCRC patients and the relative risk factors.

MATERIALS AND METHODS Study Population
In this retrospective and prospective cohort study, SCRC patients who underwent surgery or endoscopic treatment (EMR or ESD) for colorectal cancer (CRC) between January 2008 and December 2019 were enrolled. The study was approved by the ethical committee of Beijing Shijitan Hospital, Capital Medical University (No: sjtkylllx-2021(49) and informed consent was collected from the patients. All patients had undergone (1) a complete preoperative colonoscopy with cecum intubation for non-obstructing tumors or index colonoscopy within 6 months after curative resection for obstructing tumors; (2) clearance of all synchronous lesions; (3) twice followup surveillance colonoscopies (an interval of ≥6 months) in 24 months after surgery.
Patients were excluded for these reasons: (1) patients diagnosed with inflammatory bowel disease, familial adenomatous polyposis, or Lynch syndrome; (2) patients who underwent colectomy for other diseases; (3) patients with poor bowel preparation at the baseline or surveillance colonoscopy; (4) patients with incomplete medical records. Poor bowel preparation was defined as "poor" or "inadequate" according to the Aronchick Bowel Preparation Scale or if the Boston score was <2 for each colonic segment. 5,6 Synchronous colorectal carcinoma was defined as 2 or more histologically distinct CRCs that are diagnosed simultaneously or in a period less than 6 months after the first diagnosis. 7 When multiple tumors are located in the same colon segment, lesions should be distinctly separated by at least a 4-cm distance between each other and should not consist of submucosal spread. In synchronous cases, index lesions were defined as the most pathologically advanced tumors or as the largest lesion when 2 or more lesions were in an identical pathological stage. 8

Colonoscopic and Histological Examinations
All colonoscopic examinations were performed by professional endoscopists (accomplished more than 1000 colonoscopies individually) using the EVIS LUCERA CV-260 or CV-290 colonoscope (Olympus Medical Systems, Tokyo, Japan). All detected polyps were removed via biopsy, EMR, or ESD, and the specimens were histologically assessed by experienced pathologists.
The size of each polyp was estimated using open biopsy forceps or measured directly after polypectomy.
Synchronous adenomas (SA) and synchronous advanced adenomas (SAA) were defined as adenomas or advanced adenomas that are discovered in baseline colonoscopy. A baseline colonoscopy was defined as a preoperative colonoscopy for non-obstructing tumors and as colonoscopy at 6 months after curative resection for obstructing tumors, 5 while advanced adenomas were defined as highgrade dysplasia of tubular and serrated adenomas, adenomas with at least 25% villous component, and adenomas larger than 10 mm according to the United States Multi-Society Task Force guidelines. 9

Data Collection
The following clinical traits, cancer-related variables, and baseline colonoscopy data were collected: gender, age, preoperative carcinoembryonic antigen (CEA) level, comorbidities, tumor location (tumor located from cecum to splenic flexure were defined as right-sided colon cancer, while tumor located from splenic flexure to sigmoid was defined as left-sided colon cancer), tumor stage, tumor pathology (differentiation rate, mucinous component, lymphatic invasion, venous invasion, nervous invasion), and baseline colonoscopy data (presence or not of SA/SAA). Tumor staging was based on the eighth edition of the American Joint Committee on Cancer Board staging criteria. 10

Outcome Measures
The primary outcome of this study was the occurrence of MA at the 2-year surveillance colonoscopy after operation, while the secondary outcome was the occurrence of metachronous advanced adenomas (MAA).

Statistical Analyses
All the statistical analyses were performed by using Statistical Package for the Social Sciences (SPSS) 20.0 (IBM Corp.; Armonk, NY, USA). Continuous variables were expressed as mean ± standard (mean ± SD) and categorical variables as frequencies or percentages. Based on clinical knowledge and the literature, potential baseline risk factors associated with the detection of metachronous neoplasia at follow-up evaluation were selected. Univariate analysis was conducted with the independent-sample t-test and the χ 2 -test, as appropriate. Multivariate analysis was performed with logistic regression to assess the independent risk factors for MA (MAA), and the adjusted odds ratio (OR) along with 95% CIs were calculated.
Of the SCRC patients, the majority (n = 33, 86.8%) had 2 tumors (a total of 66 tumors), while 5 patients had 3 tumors (a total of 15 tumors). Table 1 shows the anatomical distribution and stage of the index and concurrent lesions of patients with SCRCs.

Comparison Between Index Lesions and Concurrent Lesions
Most patients (n = 32, 84.2%) were diagnosed with III/IV stages in the index tumor, while 5 patients were diagnosed with I/II stages. Index lesions were more frequently moderately differentiated, more deeply penetrated, and more often ulcerated in morphology than concurrent lesions ( Table 2). In total, 19 tumors were diagnosed as stage T1 or Tis, of which 11 tumors (57.9%) were resected under colonoscopy, with a curative resection rate of 100%.
In 21 of 38 patients (55.3%), MA was detected, including 6 MAA. Two patients had MC (1 patient had MC detected at 9 months after surgery and 1 patient had MC detected at 17 months after surgery). The mean number of MA and MAA per patient followed was 2.69 and 0.31, respectively, and men's MA detection rate was significantly higher (P = .034). Eleven tumors (9 patients) were curatively resected by colonoscopy. The relationship between ER (which means EMR or ESD) and metachronous lesions was inconsistent. Patients tended to have less MAA if index CRCs were treated by ER (P = .010), but the tendency was not discovered in MA (P = .289).

Risk Factors for Metachronous Neoplasms
On multivariate analysis, the presence of SA at baseline colonoscopy was an independent risk factor for MA during the follow-up (OR = 15.0; 95% CI: 1.55-145.22), which means that there was a trend for a higher rate of development of MA for those patients.

Radiological Examination
All patients underwent enhanced computed tomography (CT) in addition to colonoscopy in the postoperative period. Distant metastasis was found in 9 (23.7%) patients during the follow-up, including 2 cases of hepatic metastasis, 3 cases of pulmonary metastasis, and 4 cases of lymph node metastasis. However, we did not find a correlation between radiological findings and the followup surveillance colonoscopy.

DISCUSSION
The present study collected the clinicopathological data and 2-year postoperative endoscopic surveillance, investigating for the relative risk factors of metachronous neoplasms in SCRC patients.
Lam et al 2 reported an overall SCRC incidence of 3.5% (3667/105686) based on the data from 39 studies (ranged from 1.1% to 8.1%, neither in Asia nor in Europe). The prevalence of SCRC was 4.6% in the current study, consistent with previous studies. Synchronous colorectal carcinoma was more often seen in men, 11 the male to female ratio was 1.8 : 1.0 (2260 men vs. 1241 women) based on 38 studies. 2 In this study, the sex ratio was 2.17 : 1 (male/ female: 26/12), similar to the literature. In univariate analysis, the male sex was associated with the development of MA (P = .037), but not associated with MAA (P = .455).
Male sex tends to have a higher incidence of MA (70.8% in men vs. 33.3% in women, P < .05). However, male sex was not an independent risk factor for MA on multivariate analysis. Large-scale studies will be required to verify the relationship between male sex and metachronous lesions.
By pooling the data from 32 series in analysis, the mean age at SCRC diagnosis was 63 years (ranged from 47 to 79 years) in Lam's study. 2,12 In the current study, the age at diagnosis for SCRC was 64.6 years (±11.5 years), similar to the present research. Elderly age at diagnosis was associated with the development of MA (P = .047) but was not an independent risk factor. This demonstrated that surveillance colonoscopy should be carried out regardless of age.
Carcinoembryonic antigens were useful prognostic factors in CRC. 13 Compared to solitary CRC, abnormal CEA was more commonly seen in patients with SCRC. 14 Preoperative CEA positivity was a significant predictor of SA for SCRC patients (P = .047) but was not associated with metachronous lesions (P = .187, .302, respectively). Endoscopists should pay more attention to preoperative colonoscopy for those SCRC patients with abnormal CEA.
The anatomical distribution of SCRC is inconsistent in different reports. 2,15,16 In some studies, SCRC occurred in the same segment or close to each other; nevertheless, some SCRCs are noted in different segments of the large intestine in the other studies. 11 Lam et al 2 considered the common sites of SCRC as the sigmoid colon or rectum. In this study, 22.2% (18/81) tumors occurred in the same segment of the colon, while 19.8% (16/81) tumors were situated in the rectum. However, tumor location (with/without rectum cancer) was not significantly associated with the development of metachronous  Studies have shown that some tumor morphology is more common in patients with synchronous tumors. 2,16,17 Index tumors of synchronous cases were more frequently moderately differentiated, mucinous adenocarcinoma, more deeply penetrated, and more often ulcerated, while concurrent lesions were less advanced. 18  For SCRC that is distributed in multiple segments, the required extent of the procedure for tumors has remained controversial in the past decades. 17,19 Existing research demonstrated that there is no benefit in the survival of subtotal colectomy; therefore, segmental resection should be performed whenever possible. 20 In our study, most patients underwent multiple segment resection or extended resection, consistent with the previous studies. 17 Population-based studies have shown that colorectal neoplasms without signs of deep submucosal invasion can be treated by ER (ESD or EMR), which could reduce the incidence and mortality of CRC. In a multicenter study conducted in Japan, colorectal polyps (2 cm and larger) were removed by ER, 9.9% of which were staged as T1. In our study, 11 (16.2%) T1/Tis tumors were treated by ER with an en-bloc resection rate of 100%. Interestingly, although ER was not associated with MA, more ER indicated less MAA (P < .05). In a prope nsity -scor e-mat ched study, Yamashita et al analyzed the influence of preceding ESD combined with surgical operation on the prognosis of T1 stage colorectal cancer. In these patients, preceding ESD with histological en-bloc resection did not affect their post-operational oncologic behavior adversely. In our study, patients with ER of T1/Tis stage tumors had significantly lower occurrence rates of MAA. Combining ER and surgery could improve the discovery rate of SCRC and offer significant surveillance colonoscopy benefits. The treatment strategy in patients with SCRC is suggested in Figure 4.
Several previous studies 17  According to the National Comprehensive Cancer Network Guidelines, 23 enhanced CT is the preferred diagnostic method for colorectal cancer evaluating distant metastasis after surgery. Survival rates are higher in CRC patients who undergo CT than in those who do not undergo these investigations. 24 However, intensive surveillance with CT does not improve overall survival and recurrence-free survival in a retrospective observational study. 25 Thus, doctors should seek the balance between terms of oncologic benefit and medical economics. In our study, distant metastasis was found in approximately a quarter of the patients (23.6%), but a relationship was not found between endoscopic and radiological findings. One possible explanation was that the sample size is not large enough to be significant.
There were a few limitations associated with our study. First, as a single-center, retrospective, and prospective cohort study, selection bias was inevitable. Second, we did not take into account the impact of molecular biology (microsatellite instability, p53 mutation, or K-ras mutation). Considering that the molecular biology of SCRC tumors is not completely consistent with solitary colorectal carcinoma, 2 large-sample clinical research should be carried out to derive more precise conclusions. Despite these limitations, our study met with an important need that there have been few studies on the surveillance colonoscopy of SCRC to date.

CONCLUSION
For postoperative SCRC patients, individual surveillance strategies should be designed according to sex, baseline colonoscopy, and operative (or endoscopic) approach of resection.