Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA
Share the Article
Song JH, Shin JE, Kim DS, Lee YJ, Yang HC, Kim KO, Kim JH, Kang GH, Shim SR, Huh KC. Clinicopathologic and biomolecular features of interval colorectal cancers in Korean patients: A KASID multicenter prospective study. World J Gastroenterol 2027; 33(2): 127120 [DOI: 10.3748/wjg.127120]
Joo Hye Song, Internal Medicine, Konkuk University Medical Center, Seoul 03050, South Korea
Jeong Eun Shin, Internal Medicine, Dankook University College of Medicine, Cheonan 03050, South Korea
Dae Sung Kim, Department of Gastroenterology, Konyang University College of Medicine, Daejeon 05505, South Korea
Yoo Jin Lee, Department of Internal Medicine, Keimyung University School of Medicine, Daegu 42601, South Korea
Hee Chan Yang, Internal Medicine, Biomedical Research Institute, Chonbuk National University Hospital and Medical School, Jeonju-si 54907, Jeollabuk-do, South Korea
Kyeong Ok Kim, Internal Medicine, Yeungnam University College of Medicine, Daegu 42115, South Korea
Jae Hyun Kim, Internal Medicine, Kosin University College of Medicine, Busan 03050, South Korea
Gyeong Hoon Kang, Department of Pathology, Seoul National University Hospital, Seoul National University College of Medicine, Seoul 03080, South Korea
Sung Ryul Shim, Biomedical Informatics, Konyang University, Daejeon 03050, South Korea
Kyu Chan Huh, Department of Internal Medicine, Konyang University Hospital, Konyang University College of Medicine, Daejeon 03050, South Korea
Kyu Chan Huh, Myunggok Medical Research Center, Konyang University College of Medicine, Daejeon 03050, South Korea
Author contributions: Song JH contributed to study concept and design, data analysis and interpretation, and drafting of the manuscript; Shin JE, Kim DS, Lee YJ, Yang HC, Kim KO, and Kim JH contributed to data acquisition and critical revision of the manuscript for important intellectual content; Kang GH contributed to data analysis, material support, and critical revision of the manuscript for important intellectual content; Shim SR contributed to data and statistical analysis; Huh KC contributed to study concept and design, data acquisition, analysis and interpretation, drafting of the manuscript, and study supervision.
AI contribution statement: Portions of this manuscript were edited using AI tools solely for language refinement. The authors carefully reviewed and verified all AI-assisted outputs and take full responsibility for the scientific content of the manuscript.
Supported by the National Research Foundation of Korea grant funded by the Korea government, No. NRF-2017R1A2B2006514.
Institutional review board statement: This study was approved by the Institutional Review Boards of Konyang University Hospital (IRB No. 2017-06-007) and the remaining five participating hospitals.
Clinical trial registration statement: This study conformed to the ethical guidelines of the 1975 Declaration of Helsinki. This study was an observational study and did not involve prospective assignment of interventions to human subjects. Therefore, it did not meet the ICMJE criteria requiring clinical trial registration.
Informed consent statement: Informed consent was obtained from all patients.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
Data sharing statement: Data are available from the corresponding author on reasonable request.
Corresponding author: Kyu Chan Huh, MD, PhD, Professor, Department of Internal Medicine, Konyang University Hospital, Konyang University College of Medicine, 158 Gwanjeodong-ro, Seo-gu, Daejeon 03050, South Korea. kchuh@kyuh.ac.kr
Received: August 24, 2026 Revised: September 22, 2026 Accepted: October 29, 2026 Published online: January 14, 2027 Processing time: 82 Days and 0.4 Hours
Abstract
BACKGROUND
Interval colorectal cancer (CRC) refers to CRC diagnosed after an index colonoscopy before the next recommended surveillance examination. Our previous multicenter retrospective study showed that, unlike Western studies, interval CRCs developed more often in men and on the left side. Few studies on interval CRC in Asia have focused on its biomolecular features.
AIM
To identify the clinicopathological and biomolecular features of Korean patients with interval CRCs.
METHODS
A prospective multicenter cohort study involving 107 patients newly diagnosed with sporadic or interval CRC from six referral hospitals in Korea was conducted. Propensity score weighting was used to balance patient characteristics.
RESULTS
In contrast with Western studies, Korean interval CRCs were more common in male patients and in the left colon. The most commonly suspected etiology of CRCs is missed lesions, followed by incomplete resection, incomplete colonoscopy, and newly developed cancer. In multivariable Cox proportional hazard analysis, advanced stages [stage III: Adjusted hazard ratio (aHR), 9.58; 95% confidence interval (CI): 1.97-46.51, stage IV: AHR, 25.36; 95%CI: 4.09-157.10 vs stage I] and CpG island methylator phenotype (+) (aHR, 45.15; 95%CI: 8.09-251.9) were risk factors for survival. On the other hand, body mass index (aHR, 0.82; 95%CI: 0.70-0.96), microsatellite instability (aHR, 1.27e-8; 95%CI: 2.00e-9-8.12e-8), BRAF (+) (aHR, 0.08; 95%CI: 0.01-0.44) were protective factors for survival.
CONCLUSION
To the best of our knowledge, this is the first study on the clinicopathological and biomolecular features of Korean interval CRCs. There were several differences in the clinicopathological and biomolecular features of interval CRCs between Korean and Western populations. Further large-scale multicenter studies are required.
Core Tip: This is the first study regarding the clinicopathological and biomolecular features of Korean patients with interval colorectal cancers (CRC). In our study, interval CRC were more common in males and located in the left colon, consistent with previous Asian, but in contrast to Western studies. The risk factors associated with survival were advanced stages and CpG island methylator phenotype (CIMP) (+). On the other hand, microsatellite instability (MSI) and BRAF (+) were protective factors for survival. Similar to western studies, CIMP was higher in interval CRCs. However, MSI, and BRAF mutations were not higher in interval CRCs, than sporadic CRCs, in contrast to Western studies.
Citation: Song JH, Shin JE, Kim DS, Lee YJ, Yang HC, Kim KO, Kim JH, Kang GH, Shim SR, Huh KC. Clinicopathologic and biomolecular features of interval colorectal cancers in Korean patients: A KASID multicenter prospective study. World J Gastroenterol 2027; 33(2): 127120
Colorectal cancer (CRC) is the second most fatal and the third most prevalent cancer diagnosed globally[1]. The major pathway of CRC development from adenoma is the adenoma-carcinoma sequence[2]. Colonoscopy is the most sensitive tool and reference standard for detecting adenomas. Also, it increases the chance of preventing and detecting early CRC that can be removed endoscopically[3-5]. Individuals with colorectal polyps are at higher risk of developing colorectal polyps and CRC in the future; thus, proper and qualified surveillance with colonoscopy after colorectal polyp resection is very important[6,7]. Postpolypectomy surveillance is the most important strategy for preventing and managing CRC[8-11]. Given Korea’s relatively low medical costs and high accessibility to healthcare services, colonoscopic surveillance is recommended to be carried out five to ten years after polypectomy if there are no high-risk findings related to CRC at the index colonoscopy performed by a qualified endoscopist and under adequate bowel preparation[8,9]. The high-risk findings for metachronous CRC following polypectomy are as follows: (1) Adenomas ≥ 10 mm in size; (2) 3–5 (or more) adenomas; (3) Tubulovillous or villous adenomas; (4) Adenomas containing high-grade dysplasia; (5) Traditional serrated adenomas; (6) Sessile serrated lesions (SSLs) containing any grade of dysplasia; (7) Serrated polyps of at least 10 mm in size; and (8) 3–5 (or more) SSLs[8,9].
However, CRC can still be diagnosed after an index colonoscopy, before the next recommended surveillance examination. This type of CRC, known as interval CRC, accounts for 1.8%-9.0% of all CRC cases. Interval CRC originates from missed lesions, partially or incompletely removed lesions, or de novo cancers. Interval CRC can be minimized by improving colonoscopy quality[12,13]. Most studies on the characteristics of interval cancer have been conducted in Western countries. In a meta-analysis of 12 studies with 7912 interval CRCs, interval CRC was 2.4 times more likely to arise in the proximal colon and was not related to sex[14]. Risk factors associated with interval CRC, included patient factors (old age, diverticulum-related diseases, comorbidity), and colonoscopic factors (low polypectomy rate, low procedure completion rate, and procedure performance by non-gastroenterologists)[14]. In respect of biomolecular features, CpG island methylator phenotype (CIMP), microsatellite instability (MSI), and BRAF mutation were more common in interval CRCs, while KRAS and chromosomal instability were more common in sporadic CRCs[14-19].
Few studies have been conducted on interval CRC in Asian populations. Studies regarding the biomolecular features of Asian interval CRCs have not yet been addressed except one study that addressed the status of RAS and BRAF mutations in patients with metastatic interval CRC in Japan[20]. Our previous multicenter retrospective study with 66016 follow-up colonoscopies for 38412 patients from 11 tertiary hospitals in Korea demonstrated that the incidence of interval CRCs was 0.095%[21]. Unlike in Western studies, interval CRCs developed more often in men and on the left side, while tumor size, cancer staging, and overall survival (OS) were not significantly different compared to non-interval CRCs. However, previous studies were limited by the retrospective nature of the design and the possibility of missing interval CRCs, and did not evaluate some risk factors such as family history and molecular profiles. Therefore, we conducted a prospective study to determine the clinicopathological and biomolecular features of interval CRCs in Korea. Considering the incidence of interval CRC, we performed a propensity score weighting (PSW) analysis to compare these features between interval and sporadic CRCs, adjusting for confounders and balancing patient characteristics.
MATERIALS AND METHODS
Study population
This prospective, multicenter cohort study was conducted in six tertiary hospitals in Korea. The data were collected between October 2017 and December 2021. Patients who met the following criteria were included: (1) Patients with newly diagnosed CRC under the age of 85 years; and (2) Patients who underwent a colonoscopy within the last 5 years. All index colonoscopic findings included negative results or polyps that were subsequently removed. The exclusion criteria were as follows: (1) A history of colectomy for CRC, inflammatory bowel disease, or familial adenomatous polyposis; and (2) A history of CRC at index colonoscopy or incomplete histological information. This study conformed to the ethical guidelines of the 1975 Declaration of Helsinki and was approved by the Institutional Review Boards of Konyang University Hospital (IRB No. 2017-06-007) and the remaining five participating hospitals. Informed consent was obtained from all patients. This study was an observational study and did not involve prospective assignment of interventions to human subjects. Therefore, it did not meet the ICMJE criteria requiring clinical trial registration.
Clinicopathologic and biomolecular features of enrolled patients
Clinicopathologic characteristics included patient factors and lesion characteristics, such as location, size, morphology, differentiation, and staging. The molecular biological characteristics included CIMP, KRAS, BRAF, and MSI status. The central coordinating center received the de-identified data from each center.
Definition of interval CRC
In this study, interval CRC was defined as colorectal adenocarcinoma developed within 5 years of index colonoscopy or 3 years after endoscopic removal of preexisting high-risk adenoma, advanced adenoma, or ≥ 3 lesions[12]. Index colonoscopy was defined as colonoscopy performed before the diagnostic colonoscopy for CRC. Sporadic CRC was defined as CRC that did not meet the interval CRC criteria. Right-sided colon cancer was defined as CRC located proximal to the splenic flexure. To determine the causes of interval CRC, each case was sorted into four groups: Missed lesions, incomplete colonoscopy, incomplete removal, and newly developed cancer. Missed lesions at index colonoscopy were defined as cancers detected at different locations from prior lesions diagnosed within 30 months after index colonoscopy, or after 30 months, characterized by having all advanced CRC features such as 2 cm or larger and advanced stage (III and IV)[13]. Considering the natural history of adenomas, CRC was unlikely to appear de novo unless after 30 months[22]. Incomplete colonoscopy was defined as failure to reach or poor bowel preparation at index colonoscopy. Incomplete removal indicated cancer detected at the same site as in the previous polypectomy. Newly developed cancers were defined as cancers detected at least 30 months after the index colonoscopy, with no or only one feature of advanced cancer[13]. Board-certificated gastroenterologists at each center independently reviewed the index colonoscopy and assigned one of the four suspicious etiologies. The disagreement was resolved by consensus. Initial suspicious etiology was discussed at central. Finally, suspicious etiologies were confirmed at central. Follow-up information was obtained at the end of study (January 2, 2025) and survival was analyzed.
DNA extraction and DNA methylation analysis
Glass tissue slides were microscopically reviewed, and tissue blocks representative of tumor histology with the highest tumor purity were selected from each CRC case. The tumor portions were marked on glass slides, and the corresponding areas were dissected from the unstained tissue slides after deparaffinization in xylene. The dissected tissues were collected into microtubes containing tissue lysis buffer and proteinase K and kept at 55 °C for 1 day. After heating at 95 °C for 5 minutes and subsequent centrifugation, the supernatant was transferred to a new microtube and kept in a deep freezer.
The EZ DNA Methylation kit (Zymo Research, Orange, CA, United States) was used for bisulfite modification of genomic DNA samples. The modified DNA was subjected to an Alu-based MethyLight control reaction to assess the amount of modified DNA and was diluted with distilled water if the cross-threshold value was below 18. MethyLight assays were performed using the following eight CIMP markers: CACNA1G, CDKN2A, CRABP1, IGF2, MLH1, NEUROG1, RUNX3, and SOCS1[23]. The percentage of methylated reference (PMR) at a particular locus was calculated as 100 × (methylated reaction in the GENE/control reaction at Alu)sample/(methylated reaction in the GENE/control reaction at Alu)M.SssI-treated genomic DNA. The MethyLight assay for each CIMP marker was independently repeated three times, and the median value was regarded as the representative value of the methylation level of each marker. Markers with a median PMR > 4 were considered methylated. CIMP status was defined based on the number of methylated markers as follows: CIMP-high (> 4 methylated markers), CIMP-low (1-4 methylated markers), and CIMP-0 (no methylated markers). We defined CIMP positivity as CIMP-high and CIMP-negative as CIMP-low or CIMP-0.
KRAS and BRAF mutation detection and MSI analysis
Direct sequencing of KRAS codons 12 and 13 and allele-specific PCR for BRAF codon 600 were performed using the fluorescent multiplex PCR method with NCI-recommended microsatellite markers (BAT25, BAT26, D2S123, D5S346, and D17S250). The MSI status of each case was classified into two categories: MSI-high (two or more unstable markers among the five) and MSI-low/microsatellite-stable (MSS) (one unstable marker and no unstable markers among the five markers, respectively).
Statistical analysis
To address selection bias, PSW was performed to adjust for baseline differences between the interval and sporadic CRC groups. We performed inverse probability of treatment weighting (IPTW) using stabilized weights (SW) as one PSW analysis. IPTW with SW was selected to retain the full sample size and balance covariates effectively, simulating the conditions of a randomized controlled trial[24]. Propensity scores were calculated using a logistic regression model to estimate the probability of each patient being classified as either having interval CRC or sporadic CRC based on a set of covariates. These covariates included continuous variables such as age at diagnosis, body mass index (BMI), and tumor size and categorical variables, including smoking history, alcohol consumption history, comorbidity, family history, and tumor location, tumor shape, and differentiation. The dependent variable in the propensity score model was the type of CRC (interval CRC vs sporadic CRC). Propensity scores, ranging from 0 to 1, represent the likelihood of each patient being classified as having either interval or sporadic CRC, given their baseline characteristics.
To evaluate the effectiveness of PSW in balancing the baseline characteristics between interval CRC and sporadic CRC groups, we compared the covariates before and after weighting. The balance of covariates was assessed using standardized mean differences (SMDs), with an SMD of less than 0.2 considered indicative of a good balance between classifications. After PSW, we performed χ2 tests and t-tests to confirm that there was no significant selection bias at baseline between the interval and sporadic CRC groups.
Following the application of PSW, univariable and multivariable Cox proportional hazard analyses were conducted to identify clinical factors associated with survival in CRC. The cumulative OS rate was calculated from the day of diagnosis using the Kaplan–Meier method.
Data analyses related to PSW were performed using R version 4.3.1 (The R Foundation for Statistical Computing, Vienna, Austria), with the ‘survey’ package employed for PSW and covariate adjustment. For cases with missing data on key outcome variables, a complete case analysis was performed and cases with incomplete data were excluded from the final analysis to ensure statistical consistency. All statistical analyses were two-tailed and a P value < 0.05 was considered statistically significant.
RESULTS
Clinicopathologic features of enrolled patients with interval and sporadic CRCs
A total of 119 patients with CRC were eligible for this study. Twelve patients were excluded due to incomplete histological information. Finally, 107 patients (28 with interval CRC and 79 with sporadic CRC) were analyzed (Supplementary Figure 1). A comparison of the baseline characteristics of the sporadic and interval CRC groups before and after PSW is shown in Table 1. The PS distributions before and after PSW are shown in Figure 1. Before PSW, the mean propensity scores were 0.26 in the interval CRC group and 0.91 in the sporadic CRC group, and the values differed significantly between the two groups (P < 0.001), indicating that variables were not similarly distributed. However, after PSW, both groups’ PS distributions were comparable, and there were no statistically significant differences between the interval and sporadic CRC groups at baseline, with well-adjusted means and standard deviations (0.69 ± 0.37 and 0.86 ± 0.19, respectively; P = 0.106).
Figure 1 Box plots of the differences in variables between the interval and sporadic colorectal cancer groups before and after propensity score weighting.
A: Interval colorectal cancer group before propensity score weighting (PSW); B: Sporadic colorectal cancer group before PSW; C: Interval colorectal cancer group after PSW; D: Sporadic colorectal cancer group after PSW. Box-and-whisker plot: First quartile (Q1 or 25th percentile), the bottom line in the box; Median (Q2 or 50th percentile), the middle line in the box; Third quartile (Q3 or 75th percentile), the top line in the box. The whiskers represent the minimum and maximum observations.
Table 1 Comparison of clinicopathological features between interval and sporadic colorectal cancers, before and after propensity score weighting, n (%).
Even after PSW, a family history of CRCs was more common in interval CRC than in sporadic CRC (2.2% vs 0%, P = 0.049). Interval CRCs were most commonly found in the sigmoid colon (10.5 cases), followed by the rectum (8.4 cases), and ascending colon (4.5 cases). In contrast, sporadic CRCs were most commonly found in the sigmoid colon (24 cases), followed by the rectum (19.2 cases). Although the most common location was different between interval and sporadic CRCs, interval CRCs and sporadic CRCs both occurred more commonly in the left-side colon than in the right-side colon (80.0% and 66.2%, respectively). The mean lesion size in the interval and sporadic CRC groups was 39.28 ± 16.57 mm and 45.43 ± 19.45 mm, respectively (P = 0.117). The balance for individual variables before and after PSW is shown in Supplementary Figure 2.
Suspected etiologies of enrolled patients with interval CRCs
The suspected etiologies of interval CRC were most often missed lesions (15 cases, 53.6%) followed by incomplete resection (5 cases, 17.8%), incomplete colonoscopy during the previous colonoscopy (4 cases, 14.3%), and newly developed cancer (4 cases, 14.3%) (Supplementary Figure 3).
Biomolecular features of enrolled patients with interval and sporadic CRCs
After PSW, there were no differences in biomolecular features between the two groups. In the biomolecular analysis, CIMP was 6.2% positive in the interval CRC group and 4.5% positive in the sporadic CRC group (P = 0.670) (Table 2). KRAS mutations were observed in 25.4% of patients with interval CRC and 34.2% of those with sporadic CRC (P = 0.641). The detailed genetic analysis results for the enrolled patients are shown in Supplementary Figure 4.
Table 2 Comparison of biomolecular features between interval and sporadic colorectal cancers, before and after propensity score weighting, n (%).
Clinical outcomes of patients with interval and sporadic CRCs
The first-line treatment modality did not differ significantly between the interval and sporadic CRC groups (P = 0.250) (Table 3). Endoscopic treatment or chemoradiation therapy was performed more frequently in patients with interval CRCs. In contrast, colectomy or radiation therapy was more frequently performed for sporadic CRCs. Stage I CRC was more common in interval CRCs than in sporadic CRCs. However, this difference was not statistically significant (P = 0.177).
Table 3 Comparison of clinical outcomes between interval and sporadic colorectal cancers, before and after propensity score weighting.
During a median follow-up of 69 months [interquartile range (IQR), 52-80], mortality was observed in 5.1 patients (21.4%) with interval CRC and 17.9 patients (26.5%) with sporadic CRC. The OS (month) showed a statistically significant difference: 87.96 ± 41.20 months in the interval CRC group and 61.17 ± 26.5 months in the sporadic CRC group (P = 0.032). The cumulative survival rates of patients with interval and sporadic CRCs are shown in a Kaplan–Meier graph (Figure 2). The 5-year cumulative survival rate was 83.0% in patients with interval CRC and 76.3% in those with sporadic CRC. However, there was no statistically significant difference in survival rates between patients with interval CRC and sporadic CRC (P = 0.400).
Figure 2
Cumulative survival rates in patients with interval colorectal cancer and sporadic colorectal cancer.
Factors associated with OS of patients with CRC
In the multivariable Cox hazard proportional analysis, advanced stage [stage III: Adjusted hazard ratio (aHR), 9.58; 95% confidence interval (CI), 1.97-46.51, P = 0.005 and stage IV: AHR, 25.36; 95%CI: 4.09-157.10, P < 0.001, respectively, vs stage I], and CIMP positivity (aHR, 45.15; 95%CI: 8.09-251.9, P < 0.001) were risk factors for survival in patients with CRC (Table 4). On the other hand, BMI (aHR, 0.82; 95%CI: 0.70-0.96, P = 0.014), MSI (aHR, 1.27e-8; 95%CI: 2.00e-9-8.12e-8, P < 0.001), and BRAF mutation (aHR, 0.08; 95%CI: 0.01-0.44, P = 0.003) were protective factors for survival in patients with CRC.
Table 4 The risk factors for survival of patients with colorectal cancer, after propensity score weighting.
Interval CRC was defined as CRC diagnosed after a negative colonoscopy. Most interval CRCs originate from missed or incompletely removed lesions. Although interval CRC could be minimized by fully understanding the quality indicators and implementing colonoscopy accordingly, interval CRC could not be completely prevented. Therefore, it is important to understand the characteristics of interval CRCs, identify the risk factors, and establish strategies to predict interval CRCs.
To the best of our knowledge, this is the first study on the clinicopathological and biomolecular features of Korean patients with interval CRCs. Our prospective multicenter study included 107 patients with interval and sporadic CRCs with a median follow-up of 69 months (IQR, 52-80). During the follow-up period, the OS (months) in interval CRCs tended to be longer than that in sporadic CRCs; however, the difference was not statistically significant (83.0 vs 68.0, P = 0.400). The OS rates at 1 year, 3 years, and 5 years were 98.9%, 95.1%, and 83.0% in interval CRC, and 94.9%, 84.0%, and 76.3% in sporadic CRC, respectively. The risk factors associated with survival were advanced stages (stage III: AHR, 9.58; 95%CI: 1.97-46.51, stage IV: AHR, 25.36; 95%CI: 4.09-157.10 vs stage I) and CIMP (+) (aHR, 45.15; 95%CI: 8.09-251.90). On the other hand, BMI (aHR, 0.82; 95%CI: 0.70-0.96), MSI (aHR, 1.27e-8; 95%CI: 2.00e-9-8.12e-8), and BRAF (+) (aHR, 0.08; 95%CI: 0.01-0.44) were protective factors for survival. However, considering CI, careful interpretation was needed.
A previous meta-analysis of 12 Western studies with 7912 interval CRCs revealed that the risk factors associated with interval CRCs were old age [age > 65-70 years vs < 65-70 years: Odds ratio (OR), 1.15; 95%CI: 1.02-1.30], diverticulum-related disease (OR, 4.25; 95%CI: 2.58-7.00), and comorbidity (high Charlson comorbidity index: OR, 2.00; 95%CI: 1.77-2.27)[14]. In our study, the presence of comorbidities and old age were not independent risk factors for interval CRCs, and the association with diverticulum-related disease was not evaluated. Samadder et al’s United States population-based study[25] showed that interval CRCs were associated with earlier-stage cancer, and family history of CRC (OR, 2.27; P = 0.008), similar to our study. Also, our study showed that endoscopic treatment was more frequently performed in interval CRCs, while colectomy was more frequently performed in sporadic CRCs. They also revealed that the proximal colon (OR, 2.24; P < 0.001) and a lower risk of death (HR, 0.64; P < 0.001) were associated with interval CRCs[25]. Similarly, OS in our study was longer in interval CRC than in sporadic CRC.
Regarding the location of interval CRCs, they were more common in the proximal (right-sided) colon in Western studies, whereas in Asian studies, they were more common in the distal (left-sided) colon[21,26-29]. Consistent with an Asian study, our study demonstrated that interval CRCs were more commonly located in the left colon. Interval CRCs were most commonly found in the sigmoid colon (10.5 cases), and there was no statistically significant difference in location between interval and sporadic CRCs. Inspection of the right colon was closely related to the cecal intubation rate and risk of interval CRCs. Several studies found that failure to verify the cecal landmarks for complete inspection was a factor contributing to missed lesions located in the right colon[30-32]. In Korea, most reports showed relatively higher cecal intubation rates (> 90% to 95%) than those in Western countries (88% to 97%)[33-36]. We assumed that this was one of the possible reasons why interval CRCs in Western countries were more common in the right colon. A Danish nationwide population-based cohort study showed that female sex, localized stage at diagnosis, proximal tumor location, and high comorbidity burden were factors independently associated with interval CRC[37]. However, including our present study, Asian studies showed that most interval CRCs were more common in males, in contrast to Western studies. This might be explained by the fact that male patients tend to undergo screening and/or surveillance colonoscopy more frequently than female patients, in Asia[26-29].
Although we could not provide an accurate mechanism, we suggest the possibility of ethnic differences in the tumor location and sex of interval CRCs. Kim et al[37] demonstrated evidence of ethnic differences in the location and sex proportion ratio by comparing differences in sessile serrated adenoma/polyps (SSA/Ps) between Koreans and Americans. According to this study, SSA/Ps without dysplasia tended to occur more frequently in females (63% vs 37%), in American patients, while SSA/Ps with or without dysplasia developed in more than three-fourths of males in Korean patients. In addition, Koreans were more likely to have SSA/Ps with dysplasia and precancerous lesions in the left colon (58.3% vs 41.7%).
With respect to the biomolecular features of interval CRCs, our study revealed no significant difference in CIMP positivity between the two groups. In contrast to our study, Arain et al’s study[15] showed that interval CRCs were 2.5 times more likely than non-interval CRCs to be CIMP+ (OR, 2.41; 95%CI: 1.2-4.9). Interestingly, in the present study, CIMP + was an independent risk factor for OS of patients with CRC (aHR, 45.15; 95%CI: 8.09-251.9), in contrast with Arain et al’s study[15] which revealed that there was no difference in OS between CIMP + and CIMP- cancers. Conflicting studies existed about the impact of CIMP on patient OS; the majority of the studies reported either no benefit to OS or worse OS when CIMP was present[15,38-41]. However, one report linked CIMP to better patient OS[42]. These discrepancies were caused in part by heterogeneity within the populations being studied and inconsistencies in CIMP analysis. Furthermore, additional molecular changes involving mutation in MSI, BRAF, and KRAS were observed. However, our study conducted CIMP analysis using eight markers and strictly defined CIMP+ as the methylation of > 4 genes of interest. Additionally, we analyzed the association between CIMP and OS after multivariable adjustment for KRAS, MSI, and BRAF mutation statuses. Therefore, our results suggest that CIMP has an exploratory association with OS in patients with CRCs. In contrast with Western studies, MSI and BRAF mutations were not higher in interval CRCs, than in sporadic CRCs[14-19]. Similar to our study, Kim et al[37] showed that BRAF mutations and hMLH1 methylation were less frequent in Korean patients than in American patients. Because, these factors are well known etiologies of interval CRCs in Western population, we suggest that ethnic differences influenced not only clinicopathological characteristics, but also biomolecular features. However, further studies are needed to investigate the association with ethnicity.
In the present study, patients with MSI had a lower risk of mortality than those with MSS. Several large population-based and retrospective cohorts and meta-analysis have shown that MSI status is associated with better OS in early-stage disease compared to MSS tumors[43-46]. However, even after PSW, the number of patients with MSI remained low, so careful interpretation is needed considering CIs. In the present study, BRAF mutations were associated with better OS in present study. Generally, BRAF mutation showed poor OS and poor response to chemotherapy in advanced cancer[47]. However, BRAF mutation with MSI has a better prognosis than BRAF mutation with MSS[48]. Therefore, we interpreted that the effect of the BRAF mutation on the OS was attenuated by MSI and early-stage CRC (most patients undergo surgical treatment), rather than the protective factor itself. In the present study, BMI was associated with better OS. Obesity is an established risk factor for CRC. As an example of the obesity paradox-the inverse relationship between obesity and mortality in some populations, recent research has revealed that CRC patients with elevated BMI had improved cancer-specific survival and lower mortality[49]. Being moderately overweight (BMI, 25-30 kg/m2) was sometimes linked to better outcomes than normal weight, particularly in colon cancer, although severe obesity (BMI, > 35 kg/m2) often increases mortality[50].
Missed lesions were the most common etiology of interval CRC, as observed in our study. Many factors are associated with missed lesions, including technical colonoscopy limitations, patient characteristics, previous history of abdominal or pelvic surgery, diverticulosis, flat lesions, and lesions located in the right colon. Similar to our study, Erichsen et al[36] found that interval CRCs’ clinical features and OS rates suggested that most of them were missed lesions rather than aggressive biology. Also, interval CRC was detected in earlier stages and was more frequently treated endoscopically than sporadic CRC.
Our study had several strengths. First, this was a prospective, multicenter study. With prospective follow-up of the enrolled patients, we were able to increase the accuracy of diagnosing and evaluating the etiology of interval CRCs by reviewing index colonoscopies (endoscopic pictures and results). Second, we identified clinicopathological and biomolecular features of Korean patients with interval CRCs that are not well known.
However, our study has some limitations. First, the number of prospectively enrolled patients with interval CRCs was relatively small. Even though we enrolled patients at six tertiary hospitals over four years, it was very difficult to enroll patients during the coronavirus disease 19 pandemic period. In addition, there is an inherent numerical imbalance between interval and sporadic CRCs. Some variables showed greater SMDs than the prespecified 0.2 threshold, while total post-weighting SMD was relatively well balanced after PSW. This imbalance may result in inherently skewed outcomes. Even after the PSW analysis, there was a possibility of overfitting and low statistical power. So, cautious interpretation is needed. Second, this study was conducted in Korea; therefore, it is difficult to generalize the results to Asian populations. Therefore, multinational multicenter studies in Asia, including Korea, are needed to confirm the differences in clinicopathological and biomolecular features between Western and Asian patients with interval CRCs. Third, our study did not include some risk factors already known as risk factors for interval CRCs, such as endoscopist proficiency (experts vs trainees vs general physicians), diverticular disease, and quality indicators (adenoma detection rate, bowel preparation, cecal intubation rate, etc.). Therefore, there is the possibility of uncontrolled confounding factors. Fourth, there was a possibility that the cause of interval CRC might be misclassified. Although endoscopic pictures and the results of the index colonoscopy were reviewed, detailed information regarding the index colonoscopy was not included in the analysis. Fifth, many cases were confined to the early stages of CRC, and the number of deaths was very low in our study. Therefore, this might introduce bias into the Cox proportional hazards analysis for survival models with instability. In that point of view, parsimonious model had the strong advantage of using minimal variables, making it easy to interpret and preventing overfitting. However, in our study it might reduce the statistical power and clinical importance. Because, parsimonious model minimized variables focused only the statistical significance (P value) of variables, it might have risks dropping variables that are clinically or domain-wise very important due to sample size or noise. Therefore, a careful interpretation is required.
In the present study, interval CRCs in Korean patients were more common in the left-sided colon. However, the difference in location between the two groups was not statistically significant. Interval CRC was more frequent in patients with a family history of CRC than in those with sporadic CRC, even after PSW. A family history of CRC was not associated with mortality in patients with CRC. In the biomolecular analysis, there were no significant differences in the expression of CIMP, KRAS mutations, MSI, or BRAF mutations between the two groups. In our study, BMI, CIMP, MSI, and BRAF showed exploratory associations with OS in patients with CRCs. The most commonly suspected etiology of CRC in our study was missed lesions, followed by incomplete resection and incomplete colonoscopy, which were correctable etiologies. Therefore, the incidence of interval CRC can be reduced by efforts to improve the quality of colonoscopy through proper bowel preparation, improved adenoma or polyp detection rates, and complete resection of neoplastic lesions. However, quality indicators and detailed information from the index colonoscopy were not included in the analysis, and cautious interpretation is needed.
CONCLUSION
There are several different characteristics of interval CRC in Korea compared to those in Western countries. This might be related to ethnic differences, but also to other factors not included in present study, such as colonoscopy quality, healthcare system, and screening strategies based on regional guidelines. Considering the study limitations, the present study showed exploratory associations between some risk factors and survival in interval CRCs in Korea; therefore, careful interpretation is needed. A better understanding of the clinicopathological behavior and biomolecular characteristics of interval CRC in Asian populations, which differ from interval CRC in Western populations would minimize the incidence and reduce the risk of CRC and its associated mortality of interval CRCs in clinical practice. Further large-scale, standardized Asian and multinational studies are needed to identify the differences in the clinicopathological and biomolecular characteristics of interval CRC between Asian and Western populations.
Robertson DJ, Greenberg ER, Beach M, Sandler RS, Ahnen D, Haile RW, Burke CA, Snover DC, Bresalier RS, McKeown-Eyssen G, Mandel JS, Bond JH, Van Stolk RU, Summers RW, Rothstein R, Church TR, Cole BF, Byers T, Mott L, Baron JA. Colorectal cancer in patients under close colonoscopic surveillance.Gastroenterology. 2005;129:34-41.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 349][Cited by in RCA: 308][Article Influence: 14.7][Reference Citation Analysis (15)]
Kim SY, Kwak MS, Yoon SM, Jung Y, Kim JW, Boo SJ, Oh EH, Jeon SR, Nam SJ, Park SY, Park SK, Chun J, Baek DH, Choi MY, Park S, Byeon JS, Kim HK, Cho JY, Lee MS, Lee OY; Korean Society of Gastrointestinal Endoscopy; Korean Society of Gastroenterology; Korean Association for the Study of Intestinal Diseases. [Korean Guidelines for Postpolypectomy Colonoscopic Surveillance: 2022 Revised Edition].Korean J Gastroenterol. 2022;80:115-134.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 7][Cited by in RCA: 6][Article Influence: 1.5][Reference Citation Analysis (2)]
Kim SY, Kwak MS, Yoon SM, Jung Y, Kim JW, Boo SJ, Oh EH, Jeon SR, Nam SJ, Park SY, Park SK, Chun J, Baek DH, Choi MY, Park S, Byeon JS, Kim HK, Cho JY, Lee MS, Lee OY; Korean Society of Gastrointestinal Endoscopy; Korean Society of Gastroenterology; Korean Association for the Study of Intestinal Diseases. Korean Guidelines for Postpolypectomy Colonoscopic Surveillance: 2022 revised edition.Intest Res. 2023;21:20-42.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 15][Cited by in RCA: 18][Article Influence: 6.0][Reference Citation Analysis (1)]
Hassan C, Antonelli G, Dumonceau JM, Regula J, Bretthauer M, Chaussade S, Dekker E, Ferlitsch M, Gimeno-Garcia A, Jover R, Kalager M, Pellisé M, Pox C, Ricciardiello L, Rutter M, Helsingen LM, Bleijenberg A, Senore C, van Hooft JE, Dinis-Ribeiro M, Quintero E. Post-polypectomy colonoscopy surveillance: European Society of Gastrointestinal Endoscopy (ESGE) Guideline - Update 2020.Endoscopy. 2020;52:687-700.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 425][Cited by in RCA: 413][Article Influence: 68.8][Reference Citation Analysis (3)]
Weisenberger DJ, Siegmund KD, Campan M, Young J, Long TI, Faasse MA, Kang GH, Widschwendter M, Weener D, Buchanan D, Koh H, Simms L, Barker M, Leggett B, Levine J, Kim M, French AJ, Thibodeau SN, Jass J, Haile R, Laird PW. CpG island methylator phenotype underlies sporadic microsatellite instability and is tightly associated with BRAF mutation in colorectal cancer.Nat Genet. 2006;38:787-793.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 1638][Cited by in RCA: 1516][Article Influence: 75.8][Reference Citation Analysis (11)]
Sekiguchi M, Kishida Y, Ikematsu H, Konno M, Mizuguchi Y, Hotta K, Imai K, Ito S, Takada K, Shiomi A, Yasui H, Tsukamoto S, Hirano H, Kobayashi N, Saito Y, Inaba A, Shinmura K, Konishi J, Ozawa H, Fujita S, Murakami Y, Matsuda T. Proportions and characteristics of interval cancer in annual fecal immunochemical test screening and postcolonoscopy colorectal cancer: Results from a Japanese multicenter prospective study using questionnaires, the C-DETECT study.Dig Endosc. 2024;36:1140-1151.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 1][Cited by in RCA: 3][Article Influence: 1.5][Reference Citation Analysis (0)]
Schoen R. Surveillance after positive and negative colonoscopy examinations: issues, yields, and use.Am J Gastroenterol. 2003;98:1237-1246.
[PubMed] [DOI] [Full Text]
Kim DH, Lee SY, Choi KS, Lee HJ, Park SC, Kim J, Han CJ, Kim YC. The usefulness of colonoscopy as a screening test for detecting colorectal polyps.Hepatogastroenterology. 2007;54:2240-2242.
[PubMed] [DOI]
Van Rijnsoever M, Elsaleh H, Joseph D, McCaul K, Iacopetta B. CpG island methylator phenotype is an independent predictor of survival benefit from 5-fluorouracil in stage III colorectal cancer.Clin Cancer Res. 2003;9:2898-2903.
[PubMed] [DOI]
Samowitz WS, Curtin K, Ma KN, Schaffer D, Coleman LW, Leppert M, Slattery ML. Microsatellite instability in sporadic colon cancer is associated with an improved prognosis at the population level.Cancer Epidemiol Biomarkers Prev. 2001;10:917-923.
[PubMed] [DOI]
Gallois C, Ambrosini M, Lonardi S, Alouani E, Guimbaud R, Overman MJ, Sinicrope F, Mazard T, Decraecker M, Ros J, Elez E, Pernot S, Cremolini C, Boilève A, Parent P, Jayachandran P, Dutherage M, Mercier M, Flecchia C, Fakih M, Sclafani F, Pietrantonio F, Taieb J, Tougeron D. Prognostic impact of the BRAF V600E mutation in patients with MSI-high metastatic colorectal cancer treated with immune checkpoint inhibitors.Eur J Cancer. 2025;227:115645.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 12][Reference Citation Analysis (0)]
