Risk factors for pathological upgrading pre- and post-endoscopic resection of colorectal adenoma: A retrospective study

Abstract

BACKGROUND

Colorectal adenoma (CRA) is a precancerous lesion of colorectal cancer. Pathological upgrading often occurs when comparing preoperative endoscopic biopsy results with postoperative pathological findings in certain adenomas.

AIM

To investigate the risk factors associated with pathological upgrading after endoscopic resection of CRAs.

METHODS

This retrospective study collected 200 cases of CRAs between January 2023 and December 2025. All adenomas underwent complete endoscopic resection and were divided into pathological upgrading and non-upgrading groups according to postoperative pathological findings. Baseline clinical data and endoscopic features of the adenomas were compared between the two groups to identify factors correlated with pathological upgrading.

RESULTS

Pathological upgrading was observed in 49 of the 200 included CRAs. No statistically significant differences were found between the two groups in age, gender, body mass index, smoking history, drinking history, diabetes history, personal tumor history, fasting blood glucose level, total cholesterol level, triglyceride level, platelet count, carcinoembryonic antigen level, and adenoma location. Significant intergroup differences were observed in hemoglobin level, adenoma diameter, Yamada classification, and mucosal hyperemia. Multivariate logistic regression analysis demonstrated that pathological upgrading incidence was significantly higher in adenomas with a diameter > 20 mm than in those ≤ 20 mm (odds ratio = 23.237, 95%CI: 5.868-92.018, P < 0.001). Adenoma diameter was an independent influencing factor for pathological upgrading of CRAs.

CONCLUSION

CRAs with a diameter > 20 mm are more susceptible to pathological upgrading and carry an elevated risk of undetected high-grade dysplasia or early carcinoma missed by preoperative biopsy.

Key Words: Colorectal adenoma; Colorectal cancer; Pathological upgrading; Endoscopic resection; Risk factor; Retrospective study

Core Tip: Colorectal adenomas (CRAs) are the predominant precancerous lesions of colorectal cancer. Assessing their pathological upgrading risk is critical for early diagnosis and intervention. In this retrospective study, pathological upgrading incidence was significantly higher in adenomas with a diameter > 20 mm than in those ≤ 20 mm. Adenoma diameter was an independent influencing factor for the pathological upgrading of CRAs. CRAs with a diameter > 20 mm are more susceptible to pathological upgrading and carry an elevated risk of undetected high-grade dysplasia or early carcinoma missed by preoperative biopsy. Early and complete endoscopic resection is recommended to prevent progression to colorectal cancer. These findings will enhance clinicians’ diagnostic capabilities and assist in formulating appropriate treatment strategies for CRAs.

INTRODUCTION

Colorectal cancer is one of the most prevalent malignancies in China, with its incidence and mortality rates showing a continuous upward trend. According to the GLOBOCAN 2020 data[1], newly diagnosed colorectal cancer cases accounted for 12.2% of all malignancies in China in 2020, ranking second among all malignant neoplasms and first among gastrointestinal tumors. Colorectal cancer accounted for 9.53% of all cancer-related deaths. Colorectal adenomas (CRAs) represent the most common precancerous lesions of the colorectum and are histologically classified into three main subtypes: Tubular adenomas, tubulovillous adenomas, and villous adenomas. The prevalence of CRAs increases progressively with advancing age. Malignant transformation generally follows the well-established adenoma-carcinoma sequence[2]. Among the three histological types, villous adenomas carry the highest risk of malignant transformation, whereas tubular adenomas have the lowest transformation rate. Colonoscopy is considered the gold standard for the diagnosis of early colorectal cancer and precancerous lesions, as well as a crucial therapeutic modality for colorectal lesions[3,4].

Pathological examination of biopsy specimens is essential for determining lesion nature and plays an indispensable role in the diagnosis and clinical management of gastrointestinal diseases. Biopsy only obtains partial tissue from the target lesion; thus, histological discordance frequently exists between endoscopic biopsy samples and radically resected specimens, a phenomenon defined as pathological upgrading. For example, lesions diagnosed as low-grade intraepithelial neoplasia by biopsy may harbor undetected foci of high-grade intraepithelial neoplasia or even early invasive carcinoma[5].

Pathological upgrading is also frequently observed between pre- and post-endoscopic histological examination of CRAs. Exploring the risk associated with pathological upgrading after endoscopic resection of CRAs is of clinical significance. This study aimed to explore the clinical factors associated with pathological upgrading in specific subtypes of CRAs, including tubular adenomas, tubulovillous adenomas, and villous adenomas.

MATERIALS AND METHODS

Study design

This retrospective study aimed to identify factors associated with pathological upgrading between preoperative biopsy and postoperative pathological findings after endoscopic resection of CRAs. We collected 200 cases of CRAs at Suining Central Hospital treated between January 2023 and December 2025. This study was approved by the Ethics Committee of Suining Central Hospital (approval No. KYLLKS20250132) and registered in the Medical Research Registration Information System (https://www.medicalresearch.org.cn/Registration No. MR-51-25-054893). All patient data were anonymized to protect individual privacy.

The inclusion criteria were as follows: (1) There were no restrictions regarding age or gender distribution, and no upper limit was set for the number of adenomas eligible for inclusion from a single patient; (2) First-onset sporadic CRAs confirmed via endoscopic biopsy, encompassing tubular adenomas, tubulovillous adenomas, and villous adenomas; (3) All lesions underwent complete endoscopic resection with no partial or segmental resection performed; intact tissue specimens were retrieved, which ensured no distortion of pathological outcomes; (4) Full clinical, laboratory and endoscopic data retrievable; and (5) No restrictions on patients with prior tumor-related medical treatment history.

The exclusion criteria were: (1) History of prior intestinal polyps or previous endoscopic polyp resection; (2) Incomplete lesion resection or insufficient pathological specimens; (3) Laterally spreading tumors, adenomatosis, and polyposis syndromes; (4) Diagnosis of inflammatory bowel disease; and (5) Lesions outside the study’s scope.

Regarding the definition of pathological upgrading, preoperative biopsies of CRAs revealed only three histological subtypes: Tubular adenoma, tubulovillous adenoma, and villous adenoma. Pathological upgrading was defined as the detection of additional pathological findings beyond these three subtypes in the corresponding postoperative resection specimen of a single CRA, including low- or high-grade dysplasia, low- or high-grade intraepithelial neoplasia, intramucosal carcinoma, and other malignant or premalignant alterations. Illustrative examples include: (1) Tubular adenoma progressing to low/high-grade dysplasia; (2) Tubular adenoma progressing to low/high-grade intraepithelial neoplasia; and (3) Tubular adenoma progressing to intramucosal carcinoma.

Data collection

We collected clinical baseline data, including gender, age, body mass index (BMI), smoking history, drinking history, history of diabetes, and personal tumor history; serological parameters including routine blood test results, carcinoembryonic antigen (CEA) level, fasting blood glucose level, total cholesterol level, and triglyceride levels; and endoscopic characteristics of CRAs, including lesion location, Yamada classification, adenoma diameter, and mucosal surface hyperemia. Adenomas located in the cecum, ascending colon, and transverse colon were categorized into Group A, while lesions in the descending colon, sigmoid colon and rectum were assigned to Group B.

Preoperative biopsy and postoperative pathological findings following endoscopic resection of colorectal polyps were collected. Postoperative pathological findings were regarded as the gold standard for determining the nature of adenomas. All adenomas were assigned to either the pathological upgrading or non-pathological upgrading group according to the presence or absence of pathological upgrading on postoperative pathological examination, respectively. Further analysis was performed to identify factors associated with postoperative pathological upgrading of CRAs.

First colonoscopy for patients with CRAs

All patients received oral compound polyethylene glycol electrolyte powder for bowel preparation before colonoscopy. A Boston Bowel Preparation Score of 9 was achieved to ensure adequate visualization of intestinal lesions during endoscopic examination. All adenomas were evaluated under white-light endoscopy in terms of location, diameter, surface appearance, and morphological classification according to the Yamada classification system. Endoscopic biopsy was performed to obtain adequate adenoma tissue specimens. All biopsy samples were fixed in 10% neutral buffered formalin for subsequent pathological examination.

Endoscopic treatments for CRAs

Endoscopic treatment was performed for all lesions pathologically confirmed as adenomas by biopsy. All patients received either intravenous painless anesthesia or endotracheal intubation anesthesia. The endoscopic therapeutic modality was selected according to the clinical characteristics of the CRAs. Cold snare polypectomy was performed for CRAs with a diameter < 10 mm. Endoscopic mucosal resection (EMR) was applied for some broad-based or subpedunculated CRAs < 10 mm. EMR was also performed for sessile or subpedunculated CRAs with a diameter of 10-20 mm. Endoscopic submucosal dissection was used for CRAs > 20 mm or those with thick pedicles, whereas high-frequency electric resection was adopted for long-pedunculated CRAs. Complete endoscopic resection of all CRAs was ensured intraoperatively, and all resected specimens were sent for pathological examination.

Procedures of cold snare polypectomy: The snare tip was closely pressed against the mucosal surface at the adenoma margin with slight compression. Normal mucosa within 1-2 mm surrounding the adenoma was incorporated into the snare. After complete encirclement of the lesion, the snare was gradually tightened to achieve complete resection. The wound surface was carefully examined for signs of perforation or active bleeding. Additional endoscopic intervention was administered when necessary.

Procedures of high-frequency electrosurgical resection: The pedicle base of long-pedunculated adenomas was pre-occluded with titanium clips. The lesion pedicle was then captured with an endoscopic snare, followed by complete resection using 40 W high-frequency electrocautery.

Procedures of EMR: Saline solution was injected into the submucosa beneath the base of the adenoma to elevate the lesion and separate it from the muscularis propria, thereby reducing the risk of muscular injury and intraprocedural perforation. The elevated lesion was subsequently captured with an endoscopic snare; the snare loop was gradually tightened, and the lesion was completely resected via 40 W high-frequency electrocautery. Following resection, titanium clips were applied to close the mucosal defect and achieve hemostasis.

Procedures of ESD: Circumferential marking was performed 5 mm outside the lesion border with a Dual knife or Golden knife based on the adenoma characteristics. Submucosal injection of methylene blue-saline solution was delivered to sufficiently elevate the lesion. The mucosa was incised along the marked margins to form an initial dissection entrance, followed by stepwise dissection within the submucosal layer. Intraoperatively, repetitive submucosal injection and meticulous microvascular management were implemented, with timely hemostasis performed as required until complete en bloc resection of the lesion was achieved. After dissection, electrocoagulation was applied to cauterize exposed vessels on the wound surface to prevent bleeding, and the resulting mucosal defect was closed with titanium clips.

Statistical analysis

All data were analyzed by SPSS version 26.0 software (IBM Corp., Armonk, NY, United States). Continuous variables are expressed as mean ± SD and compared using t test. Categorical variables are presented as n (%) and compared using χ² test. Variables with P < 0.05 were included in multivariate logistic regression analysis to identify factors associated with pathological upgrading. Results were presented as odds ratio (OR) with 95%CI. Statistical significance was set at P < 0.05.

RESULTS

General histological distribution and overall pathological upgrading rate of CRAs

A total of 200 cases of CRAs were included in the present study, consisting of 150 tubular adenomas, 42 tubulovillous adenomas, and 8 villous adenomas. Among all lesions, pathological upgrading occurred in 49 adenomas, corresponding to an overall pathological upgrading incidence of 24.5%. Stratified by histological subtype, the pathological upgrading rate varied significantly among different CRAs. Tubular adenomas showed a pathological upgrading rate of 16.7% (25/150), tubulovillous adenomas had a rate of 42.9% (18/42), and villous adenomas exhibited the highest upgrading rate of 75.0% (6/8) (Table 1).

Table 1 Comparison of preoperative biopsy and postoperative pathological findings stratified by colorectal adenoma histological subtype, n (%).

Variables	Total	Non-upgrading group (n = 151)	Upgrading group (n = 49)
Tubular adenoma	150	125 (83.3)	25 (16.7)
Tubulovillous adenoma	42	24 (57.1)	18 (42.9)
Villous adenoma	8	2 (25.0)	6 (75.0)

Percentages indicate the proportion of non-upgrading/upgrading lesions within each adenoma subtype.

Univariate comparison of baseline endoscopic and laboratory features between the two adenoma groups

Based on postoperative pathological diagnoses, the 200 adenomas were categorized into the non-upgrading group (n = 151) and the pathological upgrading group (n = 49). Demographic, serological, and endoscopic characteristics were compared between the two groups (Table 2). No significant intergroup differences were found in lesion-related indicators, including age, gender distribution, smoking history, drinking history, diabetes history, personal tumor history, BMI, platelet count, CEA level, fasting blood glucose, serum total cholesterol, triglyceride levels, and lesion anatomical location (all P > 0.05). Four indicators differed significantly between the two groups: Hemoglobin concentration, adenoma diameter, Yamada morphological classification, and mucosal hyperemia (all P < 0.05).

Table 2 Baseline characteristics of colorectal adenomas stratified into non-upgrading and pathological upgrading groups.

Variables	Non-upgrading group (n = 151)	Pathological upgrading group (n = 49)	t/χ2	P value
Age (years)	58.83 ± 9.39	61.20 ± 11.70	1.289	0.202
BMI (kg/m2)	24.33 ± 3.28	23.76 ± 3.39	-1.063	0.289
Hb (g/L)	138.13 ± 14.59	131.25 ± 21.67	-2.075	0.042
PLT count (× 109/L)	198.15 ± 62.15	201.88 ± 81.11	0.337	0.736
CEA (ng/mL)	1.89 ± 1.05	2.34 ± 1.89	1.594	0.116
Cholesterol (mmol/L)	5.11 ± 1.15	4.77 ± 1.32	-1.752	0.081
Triglycerides (mmol/L)	1.97 ± 1.53	1.71 ± 1.10	-1.11	0.268
Glucose (mmol/L)	5.91 ± 2.52	5.97 ± 2.04	0.149	0.882
Gender			0.044	0.834
Male	93 (61.6)	31 (63.3)
Female	58 (38.4)	18 (36.7)
Smoking			0.77	0.781
Yes	40 (26.4)	12 (24.5)
No	111 (73.6)	37 (75.5)
Drinking			0.081	0.775
Yes	34 (22.5)	12 (24.5)
No	117 (77.5)	37 (75.5)
Diabetes			0.042	0.838
Yes	17 (11.3)	5 (10.2)
No	134 (88.7)	44 (89.8)
Tumor history			2.793	0.095
Yes	10 (6.6)	7 (14.3)
No	141 (93.4)	42 (85.7)
Adenoma location			0.937	0.333
A group	61 (40.4)	16 (32.7)
B group	90 (59.6)	33 (67.3)
Adenoma diameter (mm)			51.144	0.000
≤ 20	148 (98.0)	30 (61.2)
> 20	3 (2.0)	19 (38.8)
Yamada classification			8.153	0.043
I	30 (19.9)	4 (8.2)
II	72 (47.7)	22 (44.9)
III	29 (19.2)	18 (36.7)
IV	20 (13.2)	5 (10.2)
Mucosal surface hyperemia			10.269	0.001
No	86 (57.0)	15 (30.6)
Yes	65 (43.0)	34 (69.4)

BMI: Body mass index; Hb: Hemoglobin; PLT: Platelet; CEA: Carcinoembryonic antigen.

Multivariate logistic regression analysis of independent risk factors for pathological upgrading

All variables with P < 0.05 in the univariate analysis were included in the multivariate logistic regression model to screen for independent risk factors of pathological upgrading (Table 3). Adenoma diameter > 20 mm was an independent risk factor for pathological upgrading of CRAs. Lesions with a diameter > 20 mm were associated with a significantly increased risk of pathological upgrading (OR = 23.237, 95%CI: 5.868-92.018, P < 0.001). In contrast, hemoglobin level, mucosal hyperemia, and Yamada classification did not demonstrate independent predictive value for pathological upgrading (all P > 0.05).

Table 3 Multivariate logistic regression analysis of factors influencing pathological upgrading of colorectal adenoma.

Variables	OR	95%CI	P value
Hb	0.987	0.964-1.011	0.277
Diameter (mm)
≤ 20	Reference
> 20	23.237	5.868-92.018	< 0.001
Mucosal surface hyperemia
No	Reference
Yes	1.825	0.793-4.201	0.158
Yamada classification
I	Reference
II	1.489	0.410-5.403	0.545
III	2.735	0.703-10.637	0.146
IV	0.747	0.126-4.421	0.748

Hb: Hemoglobin; OR: Odds ratio.

DISCUSSION

CRAs are the most common precancerous lesions of the colorectum, and adenoma resection represents the primary strategy for colorectal cancer prevention. Close endoscopic surveillance is particularly essential for patients with advanced or high-risk CRAs after lesion resection[6]. Accumulating studies have demonstrated that the risk of malignant transformation varies markedly across different adenoma subtypes[7,8]. Colonoscopy remains the mainstay for detecting early colorectal cancer and precancerous lesions, with conventional white-light endoscopy as the basic modality. The entire colorectal mucosa should be observed comprehensively and carefully during both insertion and withdrawal of the endoscope. Once lesions or suspected lesions are identified, auxiliary endoscopic techniques such as chromoendoscopy, electronic enhanced endoscopy, and magnifying endoscopy should be adopted to further evaluate lesion size, scope, invasion depth, and other morphological characteristics. In recent years, artificial intelligence-based computer-assisted diagnosis (AI-CAD) technology has been increasingly applied in medicine. AI-CAD has shown promising application value in colonoscopy, as it can objectively evaluate bowel preparation quality and colonoscopy withdrawal time, and effectively detect colorectal lesions[9,10]. Although continuous advances in endoscopic techniques have facilitated the diagnosis of colorectal precancerous lesions, endoscopic biopsy combined with histopathological examination remains the gold standard for diagnosing early colorectal cancer and precancerous lesions, consistent with the diagnostic criteria for other gastrointestinal tumors.

Due to the inherent limitations of endoscopic biopsy, pathological discordance frequently exists between biopsy specimens and completely resected lesions[5]. This retrospective study confirms that adenomas > 20 mm are prone to pathological upgrading. Histological subtype stratification showed rising upgrading rates from tubular (16.7%) to tubulovillous (42.9%) and villous adenomas (75.0%). Univariate analysis found significant intergroup differences in hemoglobin, adenoma diameter, Yamada morphological classification, and mucosal hyperemia. Multivariate logistic regression further identified adenomas with a diameter > 20 mm as the sole independent risk factor. Supported by the above results, we recommend early complete endoscopic resection for large adenomas to avoid missed high-grade lesions on preoperative biopsy. A previous study has demonstrated that a larger colonic polyp diameter is associated with a higher likelihood of advanced pathological features, which is consistent with the findings of our study[8]. Epidemiological evidence has demonstrated that advanced age, male gender, smoking, heavy alcohol consumption, obesity, diabetes mellitus, inflammatory bowel disease, family history of colorectal neoplasms, and long-term consumption of red and processed meat are risk factors for colorectal cancer[11-16]. However, in the present study, no significant differences were observed between the two groups in terms of age, gender, BMI, smoking history, alcohol drinking history, and history of malignancy. It is well established that CEA level serves as a core serological biomarker and plays a critical role in the diagnosis of colorectal cancer[17]. However, all subjects in the present study had CEA levels within the normal reference range, and no significant difference in CEA levels was detected between the pathological upgrading and non-upgrading groups. These findings suggest that CEA has limited diagnostic efficacy for CRAs and precancerous lesions, and cannot reflect the pathological upgrading risk of adenomas.

CRAs represent the predominant precancerous lesions of colorectal cancer, and accurate assessment of their pathological upgrading risk is essential for early clinical diagnosis and targeted intervention. The present study confirmed that the pathological upgrading risk of CRAs was positively correlated with lesion size. Compared with adenomas with a diameter of ≤ 20 mm, adenomas larger than 20 mm yielded an OR of 23.237 for pathological upgrading. It has been shown that adenoma diameter is associated with long-term cancer risk in a dose-dependent manner. Taking adenomas < 5 mm as the reference, the relative risk was 2.3 for adenomas 10-19 mm in diameter and 3.0 for those ≥ 20 mm, indicating gradual elevation in malignant potential with increasing adenoma size[18]. Hu et al[19] also demonstrated that adenoma diameter was the strongest independent predictor of high-grade dysplasia and early invasive carcinoma. Relative to lesions ≤ 1 cm, the ORs for canceration were 13.9 for 1-2 cm adenomas, 59.3 for 2-3 cm adenomas, and 161.6 for adenomas > 3 cm. Notably, cancer risk increased sharply for adenomas > 3 cm. Shaukat et al[20] reported that the 20-year cumulative incidence of colorectal cancer was 5.5% in patients with advanced adenomas ≥ 10 mm, which was markedly higher than the 3.9% in patients with non-advanced adenomas and 1.8% in the general population. High-risk adenomas are closely associated with the risk of colorectal cancer and cancer-related mortality[21]. Large-sized CRAs carry a significantly elevated risk of pathological upgrading, and routine biopsy alone may fail to reflect their true pathological features. Timely radical endoscopic resection or surgical intervention is therefore required to avoid missed diagnoses and misdiagnoses.

Despite our efforts to comprehensively analyze risk factors for pathological upgrading of CRAs, this study has several limitations. First, this was a single-center, retrospective study, which may have introduced selection bias and restricted the generalizability of our results. Second, this study only explored risk factors for pathological upgrading of routine sporadic CRAs, the number of subgroups was small, and there was potential selection bias. The relatively modest sample size may have limited the statistical power to detect all relevant risk factors. Third, complete standardized endoscopic data (such as the Paris classification and NBI/pit patterns) were unavailable. These markers were not routinely documented or fully archived at our center during the study period, preventing confounding adjustment. Our analysis was restricted to basic, routinely documented clinical and endoscopic parameters; only histological findings of adenomas were available for evaluation. Future prospective studies with unified endoscopic recording standards should collect these morphological parameters to build more accurate risk prediction models. Therefore, studies with larger sample sizes and prospective designs are needed to validate and extend our conclusions.

CONCLUSION

CRAs with a diameter > 20 mm are more susceptible to pathological upgrading and carry an elevated risk of undetected high-grade dysplasia or early carcinoma, which are often missed by preoperative biopsy. Early and complete endoscopic resection is strongly recommended for such large lesions to interrupt the adenoma-carcinoma sequence and reduce colorectal cancer progression risk.