Endoscopic follow-up in colorectal cancer: Comparisons and critical issues across different guidelines

Abstract

Endoscopic surveillance plays a fundamental role in the long-term management of patients treated for colorectal cancer. Despite significant advances in surgical techniques, systemic therapies, and molecular diagnostics, recurrence and metachronous neoplasia remain clinically relevant challenges, particularly within the first three years following curative-intent treatment. This review summarizes current evidence and major guideline recommendations on endoscopic follow-up after colorectal resection, endoscopic removal of early-stage tumors, and non-operative management strategies. We describe the natural history of precursor lesions and their variable malignant potential, emphasizing the importance of accurate characterization, complete resection, and risk-adapted surveillance. The growing adoption of organ-preserving approaches, such as the watch-and-wait strategy following total neoadjuvant therapy for rectal cancer, and immunotherapy-induced complete responses in microsatellite instability-high/deficient mismatch repair tumors, necessitates more intensive and prolonged monitoring, including structured endoscopic follow-up. As these oncologic surveillance strategies become increasingly widespread, clearly defined clinical pathways and standardized protocols will be essential. We highlight how risk-stratified endoscopic follow-up is critical for the early detection of recurrence, prevention of new primary tumors, and the safe implementation of non-surgical management pathways. The aim is to define a clear algorithm supporting follow-up across distinct clinical contexts: Post-surgical, post-endoscopic resection, and in patients who are candidates for non-operative management. Future research will need to integrate molecular biomarkers, refine risk stratification models, and harmonize surveillance algorithms to ensure effective and personalized follow-up for patients with colorectal cancer.

Key Words: Colon cancer; Surveillance; Endoscopy; Adenomas; Recurrence

Core Tip: Endoscopy plays a central role in the long-term follow-up of patients treated for colorectal cancer, enabling early detection of recurrence and prevention of new primary tumors. Surveillance strategies should be risk-adapted, accounting for the natural history of precursor lesions, resection quality, and the distinct clinical settings of surgery, endoscopic resection, and non-operative management. The growing adoption of organ-preserving approaches, such as watch-and-wait after total neoadjuvant therapy and immunotherapy-induced complete responses in microsatellite instability-high/deficient mismatch repair tumors, requires more intensive and prolonged endoscopic monitoring. The goal is to develop clear, standardized algorithms that integrate clinical factors and emerging molecular biomarkers to ensure effective, personalized, and safe follow-up for patients with colorectal cancer.

INTRODUCTION

Colorectal cancer (CRC) represents one of the leading causes of cancer incidence and mortality worldwide: In 2020, approximately 1.9 million new cases and 935000 deaths were reported, with estimates continuing to rise[1]. Five-year survival rates have progressively improved over the past two decades, although they remain stage-dependent and vary by geographical origin, ranging from approximately 60%-70% in high-income countries[2,3]. Nevertheless, recurrence after curative resection remains a major issue; indeed, the five-year recurrence risk is up to 20% for stage II and 35% for stage III disease[4].

From a therapeutic perspective, adjuvant chemotherapy has been the cornerstone strategy for reducing recurrence risk. The QUASAR trial documented a survival benefit for fluoropyrimidines in selected stage II patients[5], while the MOSAIC study demonstrated the advantage of adding oxaliplatin (FOLFOX or CAPOX) in stage III patients[6], later complemented by less intensive regimens in “low-risk” stage III patients (T1-3, N1)[7]. This evolution reflects how modern oncology is increasingly oriented toward a personalized approach, moving away from the “one size fits all” paradigm in favor of tailored treatments. The DYNAMIC study demonstrated the feasibility of a circulating tumor DNA (ctDNA)-guided strategy, capable of reducing chemotherapy use without worsening disease-free recurrence[8]. Further evidence from CIRCULATE-Japan (GALAXY trial) confirmed the prognostic and predictive value of postoperative ctDNA[9,10]. In this context, the phase III SAGITTARIUS trial is randomizing high-risk stage II and stage III patients to ctDNA-guided vs standard adjuvant management, aiming to demonstrate clinical benefit through treatment personalization, intensifying therapy where needed and de-escalating when appropriate, while also reducing chemotherapy-related adverse events[11]. The goal is to identify patients at the highest risk of recurrence in order to improve postoperative care and reduce recurrence rates.

In parallel, personalized oncological follow-up, with closer surveillance for higher-risk patients, plays a crucial role in the early detection of recurrence, enabling treatment of isolated events such as single liver metastases. Equally important, however, is the prevention of new primary CRCs and anastomotic recurrences (ARs). Endoscopic follow-up is therefore of paramount importance within oncological surveillance, not only to detect local recurrences but also to identify new adenomas or precancerous polyps in predisposed individuals, thereby reducing the risk of new colorectal carcinomas. The optimal strategy for such follow-up has not yet been fully established. The following sections outline current guideline recommendations, highlight key critical issues, and provide a practical algorithm for clinicians. Finally, an aspect that should not be overlooked in the context of endoscopic surveillance is patient compliance: Observational studies report suboptimal adherence to surveillance programs, driven by clinical and organizational factors, which may ultimately reduce their overall effectiveness[12].

NATURAL HISTORY OF CRC: POLYPS AND PRECANCEROUS LESIONS

Adenomatous polyps (conventional adenomas)

Adenomatous polyps are the most common type of colonic polyp, accounting for approximately 60%-70% of all colorectal polyps. They are usually asymptomatic and identified incidentally during routine CRC screening[13]. These lesions represent the precursors of the majority of CRCs, with over 90% of cases arising from adenomatous lesions that progressively undergo malignant transformation[14].

The adenoma-carcinoma sequence generally spans approximately 10 years[15]. Recent modeling and surveillance data suggest that many advanced polyps exhibit a period of slow initial growth followed by more rapid expansion[16]. The malignant potential of colorectal adenomas is size-dependent: The risk of carcinoma is estimated at 10%-20% in adenomas larger than 2 cm, approximately 5% in those measuring 1-2 cm, and rare in lesions smaller than 1 cm[15,17]. A recent observational study analyzing colorectal polyps in young patients found that approximately one-third of cases with advanced histological features occurred in polyps smaller than 10 mm, supporting the recommendation to resect all neoplastic polyps regardless of size, as even small lesions may, albeit rarely, harbor malignant potential[16].

Dysplasia is a universal feature of conventional adenomas. Grading dysplasia is fundamental to guiding follow-up and treatment decisions; however, significant interobserver variability remains a recognized limitation. The risk of malignant transformation increases with the severity of dysplasia[18,19]. Dysplasia is categorized as low-grade or high-grade[20]. Establishing standardized diagnostic criteria is therefore essential to support risk stratification and surveillance strategies while minimizing both overdiagnosis and undertreatment[21].

According to the World Health Organization (WHO) histological classification, conventional adenomas are divided into three subtypes based on the proportion of villous architecture present, each carrying a distinct malignant potential. Tubular adenomas are the most common subtype (approximately 80%), characterized by less than 25% villous component and a low risk of malignant transformation (under 5%). Tubulovillous adenomas account for approximately 10%-15% of cases, contain 25%-75% villous component, and carry an intermediate risk (20%-25%). Villous adenomas are the least common (5%-10%), harbor more than 75% villous component, and are considered the highest-risk subtype, with the highest reported rates of harboring carcinoma[22-24].

Colorectal serrated lesions: Clinic, histologic and endoscopic features

The term “serrated polyp” refers to a lesion characterized by a sawtooth appearance of the epithelial crypts on microscopy. The WHO classification identifies four categories of serrated lesions: Hyperplastic polyps (HPs), sessile serrated lesions (SSLs), previously known as sessile serrated adenoma/polyp, with or without dysplasia, traditional serrated adenomas (TSAs), and unclassified serrated adenomas, the latter referring to ambiguous colorectal polyps displaying both dysplasia and serrated architecture that cannot be clearly classified as SSL, TSA, or conventional adenoma[25]. The anatomical location of serrated polyps is associated with the risk of colorectal carcinoma[26]; accelerated progression to CRC is more commonly observed in high-risk serrated lesions (≥ 10 mm or with dysplasia) located in the proximal colon[27].

HPs

HPs represent the most common type of serrated lesion, accounting for approximately 80% of all serrated polyps. They are generally considered to carry negligible malignant potential and are most frequently located in the rectosigmoid colon[28]. Morphologically, HPs are small, round, and typically flat or only slightly elevated, often becoming even flatter with insufflation. Narrow-band imaging (NBI) can be used to distinguish HPs from conventional adenomas[29].

SSLs

SSLs are the second most common type of serrated lesion, accounting for up to 20% of all serrated polyps. According to the updated WHO criteria, the presence of a single unequivocally distorted crypt is considered diagnostic for SSL[25]. Their characteristic morphology, sessile or flat[16,30], with less distinct borders compared to conventional adenomas, contributes to an increased risk of both missed detection and incomplete resection. Over a prolonged growth period, typically 10-15 years, SSLs may develop areas of dysplasia, becoming SSLs with dysplasia, which can progress rapidly to CRC[31,32]. These features are believed to play a central role in the development of interval CRC, defined as a cancer diagnosed after a screening or surveillance colonoscopy negative for malignancy and arising before the next recommended endoscopic assessment[33,34].

An important endoscopic marker is the mucus cap. A recent meta-analysis demonstrated that SSLs without dysplasia are more frequently associated with a mucus cap than those with dysplasia or carcinoma[35,36]. For small serrated lesions, evidence from a large prospective cohort indicates no significant increase in CRC risk compared with individuals without polyps (hazard ratio 1.25 for lesions < 10 mm). In contrast, large serrated lesions (≥ 10 mm) are associated with a markedly higher risk (hazard ratio = 3.35)[37]. This size-dependent gradient is supported by a recent systematic review and meta-analysis, which reported CRC incidence rates of 0.50 per 1000 person-years for non-advanced serrated lesions, 2.09 per 1000 person-years for advanced serrated lesions (defined as ≥ 10 mm or with dysplasia), and 0.44 per 1000 person-years following a normal colonoscopy[38].

While the risk of developing dysplasia has been linked to increasing patient age and polyp size, evidence from Liu et al[39] indicates that dysplastic change is not confined to large lesions. In their series, the median size of SSLs with dysplasia was 12 mm, with 40% of cases measuring less than 10 mm, underscoring that size alone cannot be considered a reliable marker of malignant potential.

SSLs with dysplasia

SSLs with dysplasia represent a distinct histological subgroup of SSLs and are considered the most clinically relevant lesion within this category. The prevalence of dysplasia among SSLs is approximately 4%-8%. SSLs with dysplasia are more frequently located in the proximal colon and show a female predominance[40]. The duration of the SSL phase before the onset of dysplasia is estimated at approximately 17 years; however, once dysplasia develops, progression to carcinoma tends to occur rapidly[32].

TSAs

TSAs are rare, accounting for approximately 1% of all serrated polyps. They are typically large lesions, with a mean size of approximately 19 mm, and are most frequently located in the distal colon and rectum[41]. Both males and female are affected, usually in sixth or seventh decade of life[32]. Morphologically, TSAs are often pedunculated and display a villous architecture with conventional dysplasia, which confers significant malignant potential[25,40,42]. Morphology varies with size: Smaller TSAs are typically sessile or pedunculated, whereas larger lesions tend to be lobulated, irregular, or carpet-like. At the molecular level, TSAs are most commonly associated with BRAF mutations and, less frequently, with KRAS mutations[43].

RECURRENCES AT THE SURGICAL ANASTOMOSIS SITE

AR is a rare but clinically significant complication following curative surgery for CRC. It is defined as the reappearance of tumor at the anastomotic line and requires pathological confirmation, via biopsy or resected specimen, to distinguish true malignant recurrence from benign postoperative changes, such as granulation tissue, fibrosis, or inflammatory polyps, and to differentiate recurrence from metachronous neoplasia[44-46]. Historically, AR rates were higher; however, with the introduction of intraoperative irrigation techniques, AR is now considered a rare event, occurring in approximately 1%-3% of cases and most commonly within the first 2-3 years after surgery[47].

The pathogenesis of AR is multifactorial, involving an interplay of tumor-, patient-, and surgery-related determinants. Among tumor-related factors, advanced T stage (pT4), lymph node metastasis (LNM), short distal resection margins, and adverse histology, particularly poorly differentiated and mucinous subtypes, are associated with an increased risk of recurrence. Tumor location also plays a role, with rectal cancer showing a significantly higher recurrence rate than colon cancer[48,49]. Patient-related factors such as male sex, smoking, malnutrition, and preoperative radiotherapy have also been linked to increased risk, primarily through their association with anastomotic leakage (AL), which predisposes to local recurrence and reduces five-year disease-free survival[49,50].

Consequently, short-interval surveillance during the first two postoperative years is crucial, particularly in patients with high-risk features[51]. AR is frequently detectable at colonoscopy; therefore, vigilant endoscopic surveillance is essential to maximize the chances of curative salvage resection[52].

Perioperative perturbations of the gut microbiota, caused by chemotherapy, radiotherapy, fasting, bowel preparation, and antibiotic exposure, lead to microbial dysbiosis characterized by reduced diversity, overgrowth of opportunistic pathogens, impaired mucosal barrier function, and bacterial translocation, all of which contribute to the pathogenesis of AL[53].

In recent years, the intestinal microbiota has emerged as a key determinant of outcomes after colorectal surgery. Among the microbial species implicated in AL, Pseudomonas aeruginosa and Enterococcus faecalis are most strongly associated with AL and subsequent recurrence[54,55]. In particular, Enterococcus faecalis tends to colonize the anastomotic site, where it secretes collagenases and proteases that degrade healing tissue, thereby promoting leakage. This effect appears independent of traditional risk factors but may act synergistically with them to increase recurrence risk[55,56].

The local abundance of Enterococcus faecalis in mucosal biopsy samples has been proposed as a potential biomarker for predicting both leakage and recurrence. Although other bacteria, such as Bifidobacterium, have been statistically associated with AL, Enterococcus faecalis is unique in its pathogenic role owing to its enzymatic activity rather than its abundance alone[56].

Microbiome-targeted interventions, such as intraluminal lavage with warm saline or povidone-iodine solution before anastomosis, selective antimicrobial prophylaxis, or probiotic supplementation, may represent future strategies to reduce anastomotic complications and AR[47,50]. A recent systematic review confirmed that preoperative oral antibiotic prophylaxis can reduce AL rates, supporting its routine use[57]. Given the fundamental role of the gut microbiota in anastomotic healing, targeted preoperative bowel decontamination with specific antibiotics could significantly influence AL rates and indirectly reduce AR risk[58,59].

Perioperative and surgical factors also independently contribute to AR pathogenesis. Intraoperative blood loss has been identified as a significant predictor of AR and reduced long-term survival, underscoring the importance of minimizing bleeding during surgery; its reduction should therefore be a specific goal of perioperative management[48,60]. Systematic reviews have confirmed that AL is a well-established risk factor for local recurrence[49,50], whereas adequate lymph node harvest is protective, reflecting the oncological quality of resection[52].

Mechanisms of AR

Several mechanisms have been proposed to explain malignant recurrences at the anastomotic site, including unrecognized lymphatic metastasis, positive resection margins, shedding of cells from the primary lesion, and mishandling of the tumor during surgery[61].

A widely discussed hypothesis is the implantation of exfoliated cancer cells at the staple line during anastomosis. This mechanism is supported by histopathological findings of well-differentiated adenocarcinoma with negative margins and no lymphovascular invasion, suggesting that recurrence may arise from local implantation rather than incomplete resection or nodal spread[62]. Prophylactic lavage of the bowel lumen with ≥ 2000 mL of warm saline or povidone-iodine solution before anastomosis has been shown to effectively reduce exfoliated cancer cells, particularly in rectal cancer with short distal margins or large tumors[63].

Overall survival benefit of oncological follow-up

The rationale for oncological follow-up rests on the premise that earlier detection of recurrence enables timely treatment initiation, with the ultimate goal of improving survival in patients whose disease unfortunately reappears. However, in oncology, earlier diagnosis of recurrence does not always translate into a survival benefit.

The first randomized trials investigating earlier recurrence detection during follow-up, conducted in the 1990s, did not demonstrate any overall survival advantage[64-66]. By the late 1990s, preliminary meta-analyses suggested a modest OS benefit, approximately 10% at five years, associated with intensive follow-up regimens, likely mediated by a higher rate of curative resections[67]. With the advent of systematic reviews, a British Medical Journal meta-analysis concluded that intensive follow-up anticipates recurrence diagnosis and may improve survival, although larger trials were needed to clarify which individual components are truly effective. Other studies emphasized the importance of risk-adapted approaches and the quality of the tests performed[68,69].

Further investigations demonstrated that systematic carcinoembryonic antigen (CEA) monitoring and routine computed tomography (CT) increase the proportion of recurrences treated surgically with curative intent, but do not improve overall survival compared with less intensive follow-up, suggesting the need for more personalized strategies[70-73]. In summary, current evidence supports a strong rationale for oncological follow-up, early diagnosis and increased rates of curative resection, but a consistent overall survival benefit in randomized trials remains limited. Modern strategies therefore focus on risk personalization, genetic syndromes, pathway quality, and integration with endoscopy and molecular markers of disease, rather than indiscriminate intensification of surveillance schedules.

Special population: Lynch syndrome and other genetic syndromes

Lynch syndrome is the most common hereditary CRC syndrome, caused by pathogenic variants in mismatch repair (MMR) genes (MLH1, MSH2, MSH6, PMS2, EPCAM). The lifetime risk of CRC ranges from 30% to 76% by age 70, depending on the specific gene, sex, and population, with onset typically occurring earlier than in sporadic cases[74,75]. It is characterized by a right-sided predominance, multiple synchronous or metachronous tumors, and a rapid adenoma-carcinoma sequence.

American and European guidelines[76-79] recommend intensive colonoscopic surveillance: Every 1-2 years from age 20-25 for MLH1/MSH2 carriers, and every 1-3 years from age 30-35 for MSH6/PMS2 carriers, reflecting their lower risk and later disease onset[80]. Regular surveillance reduces CRC incidence by more than 50% and CRC-specific mortality by up to 72% compared with no surveillance[81,82]. Shorter intervals (≤ 2 years) improve early-stage detection, especially in MLH1 and MSH2 carriers; biennial colonoscopy is cost-effective, while annual examinations provide limited additional benefit[83,84].

Despite close follow-up, some CRCs in Lynch syndrome arise through non-adenomatous or rapidly progressing pathways, underscoring the need for individualized risk assessment. Universal tumor testing for MMR deficiency in all newly diagnosed CRC cases is recommended to identify Lynch syndrome and enable cascade testing among family members[74,81,85].

Both United States and European guidelines[76-79] recommend colectomy with ileorectal anastomosis as the treatment of choice for colon cancer in Lynch syndrome, given the high rate of metachronous CRC (up to 62% at 30 years) after segmental resection. Subtotal colectomy is preferred in younger or higher-risk patients, while segmental resection may be considered in older patients or those with functional limitations[86]. Following surgery, lifelong colonoscopic surveillance every 1-2 years is advised, or every 6-12 months after subtotal colectomy, to detect new neoplastic lesions[76-79]. Intensive surveillance combined with extended resection substantially reduces the incidence and mortality of metachronous CRC in Lynch syndrome, with regular colonoscopy lowering CRC risk by over 50% and mortality by up to 72%[82].

Familial adenomatous polyposis (FAP) is an autosomal dominant syndrome caused by pathogenic APC variants, characterized by the development of hundreds to thousands of colorectal adenomas and an almost universal risk of CRC without prophylactic intervention. Attenuated FAP presents with fewer polyps and later onset. MUTYH-associated polyposis (MAP) is an autosomal recessive condition, typically presenting with fewer polyps and a lower, but still markedly elevated - CRC risk[87]. Serrated polyposis syndrome and hamartomatous polyposis syndromes also carry a substantial risk of CRC[88].

American guidelines recommend annual colonoscopy starting at age 10-12 for classic FAP, with colectomy indicated when polyp burden becomes unmanageable or high-grade dysplasia is detected. After colectomy, lifelong surveillance of the rectal stump or ileal pouch is required. For attenuated FAP and MAP, colonoscopy every 1-2 years from early adulthood is advised, with colectomy reserved for cases in which endoscopic control is inadequate[89]. European Society of Gastrointestinal Endoscopy (ESGE) guidelines are consistent with these recommendations, emphasizing early and frequent colonoscopic surveillance across all adenomatous polyposis syndromes[90].

Endoscopic follow-up after colectomy is essential due to the ongoing risk of metachronous adenomas and carcinoma in the retained rectum or ileal pouch. Regular endoscopic surveillance and polypectomy can effectively control rectal adenoma progression, resulting in low rates of secondary proctectomy (4%) and rectal cancer (0.5%) over a median follow-up of 8.6 years, with most patients maintaining rectal preservation[91]. In MAP, regular endoscopic surveillance and polypectomy frequently allow colectomy to be avoided or delayed, with no CRCs reported in expert centers[92]. For patients with ileorectal anastomosis, flexible sigmoidoscopy every 6-12 months indefinitely is recommended; for those with ileal pouch-anal anastomosis, pouch endoscopy every 1-2 years is advised, reduced to every 6 months if advanced adenomas are present[89,93].

For serrated polyposis syndrome, recent prospective studies support individualized surveillance intervals of 1-2 years based on polyp burden, extending to 2 years in low-risk patients. This approach maintains a low five-year CRC incidence (1%-1.3%) while reducing colonoscopy frequency without compromising patient safety[94,95].

Adherence to follow-up

Patient adherence to oncological follow-up programs, and in particular to endoscopic surveillance, is a crucial aspect that should not be underestimated. Periodic colonoscopy after resection for CRC is recommended for the early detection of recurrences and metachronous neoplasms; however, not all patients accept or comply with the recommended schedule, often due to reluctance, fear of the procedure, comorbidities, or organizational barriers.

To better understand reduced adherence to endoscopic follow-up in patients with a history of surgically treated CRC, it is useful to broaden the perspective to upstream adherence within CRC screening pathways. A randomized study conducted in China evaluated a tailored communication intervention via WeChat in first-degree relatives of patients with CRC. The intervention resulted in a significant increase in colonoscopy uptake, primarily by enhancing perceived susceptibility, reinforcing cues to action, and reducing perceived barriers, suggesting that obstacles are not only structural but also informational and psychosocial[96]. In contrast, a United States study conducted in rural populations found no additional benefit from sending a gender-specific educational flyer together with mailed FIT kits, indicating that simple informational interventions are insufficient to improve adherence[97].

Beyond logistical barriers, psychological factors play a relevant role. Analyses conducted in the United States have shown that many patients with a positive fecal test do not proceed to diagnostic colonoscopy, with fear, anxiety, embarrassment, and risk perception identified among the main reasons[98,99]. A more recent study confirmed that psychological barriers often outweigh structural ones, ultimately leading patients to forgo surveillance[100]. Similar trends are observed in the endoscopic follow-up of non-oncological conditions, particularly among older patients and those with limited access[101]. One might expect patients with a prior diagnosis of CRC to show better adherence, but this is not always the case: In a study by Kupfer et al[102], adherence to post-diagnosis endoscopic surveillance was below 50%.

In summary, non-adherence substantially reduces the effectiveness of follow-up programs, undermining part of the expected benefit of endoscopic surveillance. To address this, several studies highlight the need for a multifactorial approach. Clear and personalized physician-patient communication is essential to explain the importance of endoscopic follow-up and the risks associated with recurrence. Practical tools such as reminder systems, telephone, email, or app-based, may help reduce delays and forgetfulness. Simplification of access pathways to endoscopy also plays an important role, as do targeted and culturally sensitive educational programs designed to address individual fears and resistance. Finally, the involvement of primary care physicians and a multidisciplinary team ensures more comprehensive and continuous patient management. The integration of these strategies, applied synergistically, likely represents the most effective way to bridge the gap between guidelines and clinical practice, and to ensure that patients derive maximum benefit from endoscopic follow-up.

POST-SURGICAL SURVEILLANCE

After curative-intent resection, surveillance aims to detect recurrences and metachronous neoplasms at an early stage, when curative treatment remains feasible. Colonoscopy at one year after surgery is a consistent recommendation across major guidelines[103-106]. However, the traditional “1-3-5 year” schedule is no longer uniformly endorsed. Current guidelines favor a risk-adapted approach, with surveillance intervals guided by the completeness and quality of perioperative colonoscopy, baseline findings, and patient-specific risk factors.

In this context, relevant differences emerge across major societies. National Comprehensive Cancer Network and European Society for Medical Oncology recommend tailoring follow-up based on findings and examination quality[106], while ESGE and United States Multi-Society Task Force emphasize the importance of complete perioperative colonoscopy and risk stratification according to adenoma and serrated lesion features[103-105]. British Society of Gastroenterology places greater weight on risk stratification and allows return to routine screening in low-risk patients, whereas Italian Association of Medical Oncology broadly aligns with European Society for Medical Oncology with less granular stratification. Additional distinctions apply to specific clinical settings, particularly rectal cancer, where most guidelines recommend more intensive local surveillance during the first 2-3 years owing to the higher risk of local recurrence. In patients with incomplete preoperative colonoscopy, for example, in the setting of obstruction, completion colonoscopy is recommended within the first year after surgery or at the end of adjuvant treatment[106]. Overall, there is clear convergence toward individualized, risk-based surveillance strategies rather than fixed schedules, and endoscopic surveillance represents a key component of this framework.

A practical distinction applies between patients treated with surgery alone and those who receive adjuvant chemotherapy. In the former, follow-up begins immediately according to standard schedules for clinic visits, CEA, and chest-abdomen-pelvis CT: Several guidelines propose visits every 3-6 months for the first three years, then every 6-12 months until year five; CEA every 3-6 months for two years, then every six months up to five years in candidates for curative-intent treatment; and CT every 6-12 months for three years in higher-risk cases[106-108]. In patients receiving adjuvant chemotherapy, typically all stage III and high-risk stage II patients, oncology visits during treatment provide closer clinical and hematological monitoring, and CEA measurements are often incorporated into treatment visits. Colonoscopy timing remains anchored to the date of surgery, while CT can be scheduled at the end of adjuvant therapy and thereafter according to risk[108]. In essence, adjuvant therapy does not overturn recommended schedules but complements them with closer clinical contact during the first year; once treatment is completed, patients return to the standard risk-adjusted scheme. An overview of the recommendation by major guidelines, is summarized in Table 1.

Table 1 Comprehensive tests for oncologic follow-up based on National Comprehensive Cancer Network, European Society for Medical Oncology and Italian Association of Medical Oncology guidelines.

Exam/assessment	Post-surgical patients (colon/rectum)
Clinical visit + medical history + physical examination	Every 3-6 months for 3 years, then every 6-12 months up to 5 years
Serum CEA	Every 3-6 months for 2 years, then every 6 months up to 5 years (only if candidates for curative surgery)
Colonoscopy	At 1 year after surgery, then at 3 years, then every 5 years (more frequently if advanced polyps or incomplete resections)
Rectoscopy/flexible sigmoidoscopy	In patients with rectal resection, every 6-12 months for 2-3 years
Pelvic MRI	Not routine; reserved for high-risk rectal tumors
Chest-abdomen-pelvis CT scan	Every 6-12 months for 3 years, then annually up to 5 years (for high-risk stage II and stage III)
Emerging biomarkers (ctDNA)	Not routinely recommended; use in clinical trials

CEA: Carcinoembryonic antigen; MRI: Magnetic resonance imaging; CT: Computed tomography; ctDNA: Circulating tumor DNA.

Risk stratification remains crucial: The five-year cumulative probability of recurrence is higher in stage III than stage II, and within stage II certain features, pT4, low lymph node yield, perforation, margin involvement, MSI status, and others, define higher-risk subgroups[4]. In these patients, closer radiological and laboratory monitoring is justified (e.g., CT and CEA at the lower end of recommended intervals), whereas in lower-risk profiles, intervals can be extended without loss of overall effectiveness. It should be emphasized that when preoperative colonoscopy has not been performed, for example, in presentations with bowel obstruction, it should be completed at the end of adjuvant therapy and in any case within 12 months of resection[109].

Regarding the question of who should lead surveillance, the literature shows that different models, oncologist-led, surgeon-led, or general-practitioner-led with shared protocols, can achieve comparable outcomes when pathways are well structured[68]. More recent data confirm that shared-care models can maintain quality of life and outcomes, provided roles and responsibilities are clearly defined and recommendations adhere to guidelines[110,111]. American Society of Clinical Oncology/Cancer Care Ontario documents recommend that the specialist define a follow-up plan and formalize handoff to primary care when appropriate, specifying tests, timing, and criteria for re-evaluation[109]. In practice, near the time of surgery and during adjuvant therapy, leadership is typically oncological or surgical; in the medium-to-long term, a shared model with the primary care physician is reasonable and sustainable without loss of quality, provided that schedules are followed and easy re-entry to specialist care is available in the event of new symptoms or a rising CEA. The choice of care leader can therefore be flexible: What matters is a clearly codified pathway, centered on individual risk and aligned with guidelines, with clear communication among oncology, surgery, and primary care. This approach preserves the effectiveness of follow-up, supports adherence, and enables timely intervention when needed.

Surveillance according to precancerous lesions detected during CRC surveillance

During oncological surveillance, endoscopists frequently encounter precancerous lesions that can be resected and submitted for histopathological evaluation. The subsequent follow-up strategy depends on several factors, including the histological type of the resected polyps, their size, and the resection technique employed. Surveillance intervals and management recommendations vary across international guidelines[112-114].

Adenomas - follow-up post polypectomy

Recent evidence indicates that the removal of small, diminutive adenomas, fewer than four, each < 10 mm, without advanced histological features, is associated with a CRC risk comparable to that of individuals without adenomas at baseline, supporting a return to routine screening rather than intensive surveillance[115,116].

Patients with ≥ 5 adenomas, at least one adenoma ≥ 10 mm, or any adenoma with high-grade dysplasia are considered high-risk and require surveillance colonoscopy at three-year intervals. Among high-risk patients not undergoing surveillance, the cumulative 10-year CRC incidence was 3.3% [95% confidence interval (CI): 2.5%-4.3%], significantly higher than in the general population (standardized incidence ratio = 1.30, 95%CI: 1.03-1.62)[117]. Evidence from the United States healthcare system further demonstrated that adenoma size ≥ 10 mm was independently associated with a 3.6-fold increased risk of advanced adenoma [odds ratio (OR) = 3.6; 95%CI: 2.8-4.5] and a 5.2-fold increased risk of CRC during follow-up (OR = 5.2; 95%CI: 1.8-15.1)[118]. However, a more recent retrospective study suggested that only adenomas ≥ 20 mm were associated with a significantly increased risk of CRC incidence and CRC-related mortality[116]. Consistently, a meta-analysis involving 936540 patients with a mean follow-up of 2.9 ± 5.4 years reported that patients with baseline adenomas ≥ 20 mm or with high-grade dysplasia had the greatest risk of metachronous CRC or advanced adenomas, while those with adenomas ≥ 10 mm or with a villous component also exhibited a significantly elevated CRC risk[119].

According to major gastroenterological and oncological guidelines, the detection of ≥ 10 adenomas warrants referral for genetic counseling to evaluate for potential cancer-predisposing syndromes[90,112,114,120,121]. British Society of Gastroenterology guidelines further refine these recommendations by incorporating patient age: Individuals under 60 years with a lifetime total of ≥ 10 adenomas, or those aged ≥ 60 years with either ≥ 20 adenomas or > 10 adenomas in the presence of a family history of CRC, should be offered germline testing[122]. In the largest study to date of individuals with 10-19 adenomas, 3789 patients with ≥ 10 colorectal polyps underwent constitutional testing with a 17-gene panel. The diagnostic yield of pathogenic variants remained above 5% across all age groups, although prevalence decreased with advancing age[123]. These findings underscore the importance of thorough clinical history-taking to appropriately identify patients who may benefit from genetic testing.

For adenomas ≥ 20 mm removed by piecemeal resection, current guidelines recommend repeat colonoscopy within 3-6 months with targeted biopsies of the scar site, then at one year after the first surveillance examination, then at three years after the second. Most recurrences are detected during the first or second surveillance colonoscopy, and shorter intervals do not improve detection rates[124-126]. If recurrence is found, endoscopic resection should be performed, with subsequent examinations at 6-month or 12-month intervals until the site is clear. Shorter intervals are favored for recurrences ≥ 1 cm or with high-grade dysplasia, given that piecemeal resection is an independent risk factor for recurrence: Local recurrence occurs in approximately 3% of en bloc resections compared with 20% of piecemeal resections following endoscopic mucosal resection (EMR) of non-pedunculated colorectal lesions[125]. More recent data suggest that recurrence risk may be low in selected cases, particularly for lesions 20-30 mm without high-grade dysplasia, supporting the possibility of extending the first surveillance interval to 12 months after piecemeal EMR[127]. Several predictive models have been developed to estimate the risk of recurrence after EMR, although their external validity remains limited[128-130]. An overview of the recommended surveillance intervals for adenomas, as outlined by major guidelines, is summarized in Table 2.

Table 2 Different surveillance interval recommendations for conventional adenomas in Western guideline.

Colonoscopy findings	ESGE	BSG/ACPGBI/PHE	USMSTF
1-2 tubular adenomas < 10 mm	Return to screening	Return to screening	7-10 years
3-4 tubular adenomas < 10 mm	Return to screening	Return to screening	3-5 years
5-10 tubular adenomas < 10 mm	3 years	3 years	3 years
≥ 10 adenomas	Genetic counseling	Refer to BSG hereditary CRC guidelines1	1 year and consider genetic counseling
Adenoma with villous histology < 10 mm	Return to screening	Return to screening	3 years
Adenomas ≥ 10 mm	3 years	3 years	3 years
Adenoma with high-grade dysplasia	3 years	3 years	3 years
Piecemeal resection of adenoma ≥ 20 mm	3-6 months	2-6 months	6 months

¹If > 60 years old and lifetime history > 10 adenomas removed or if ≥ 60 years old and lifetime total ≥ 20 adenomas or ≥ 10 and family history colorectal cancer.

ESGE: European Society of Gastrointestinal Endoscopy; BSG: British Society of Gastroenterology; ACPGBI: Association of Coloproctology of Great Britain and Ireland; PHE: Public Health England; USMSTF: United States Multi-Society Task Force.

SSLs - follow-up post polypectomy

The follow-up strategy for SSLs largely parallels that for adenomas. Patients with large serrated polyps (≥ 10 mm) are at significantly increased risk of CRC after resection, with hazard ratios of approximately 3.3 reported across different studies[37,131]. Similarly, SSLs with dysplasia are associated with a nearly fivefold higher odds of CRC (OR = 4.76; 95%CI: 2.59-8.73) compared with patients without polyps, supporting guideline recommendations for post-polypectomy surveillance at three years[132]. Of note, a recent meta-analysis demonstrated that patients with synchronous SSLs and high-risk adenomas have a greater likelihood of developing metachronous advanced colorectal neoplasia than those with high-risk adenomas alone, suggesting that more intensive surveillance may be warranted in this subgroup[133] (Table 3).

Table 3 Different surveillance interval recommendations for sessile serrated lesions in Western guideline.

Colonoscopy findings	ESGE	BSG/ACPGBI/PHE	USMSTF
1-2 SSP < 10 mm	Return to screening	Return to screening	5-10 years
3-4 SSP < 10 mm	Return to screening	Return to screening	3-5 years
≥ 5 SSP < 10 mm	3 years	3 years	3 years
SSP ≥ 10 mm	3 years	3 years	3 years
SSP with dysplasia	3 years	3 years	3 years

SSP: Sessile serrated polyps; ESGE: European Society of Gastrointestinal Endoscopy; BSG: British Society of Gastroenterology; ACPGBI: Association of Coloproctology of Great Britain and Ireland; PHE: Public Health England; USMSTF: United States Multi-Society Task Force.

POST-RESECTIVE ENDOSCOPIC SURVEILLANCE

CRC most often develops from precursor lesions that progress from low-grade dysplasia to invasive carcinoma[134]. Once a neoplastic lesion invades the submucosa (pT1)[135], the risk of lymphovascular dissemination becomes clinically significant, requiring a balance between endoscopic and surgical management. Deep submucosal invasion (dSMI), defined as ≥ 1000 μm beyond the muscularis mucosae[136,137], has traditionally been used as a key threshold for this decision. However, increasing evidence suggests that dSMI alone may not fully capture the risk of LNM, highlighting the need for a more comprehensive risk stratification approach. The primary clinical challenge is to accurately distinguish lesions that can be safely treated endoscopically from those requiring surgery, in order to avoid both undertreatment, inadequate resection of high-risk cancers, and overtreatment, unnecessary surgery for low-risk lesions[138,139]. Several macroscopic and microscopic features, including lesion morphology, location, and findings from optical classification systems, have been validated as predictors of dSMI.

Depressed lesions (0-IIc or 0-IIa+IIc according to the Paris classification)[140] carry the highest risk of invasive carcinoma[141-143]. Large lateral spreading tumors (LSTs), defined as lesions ≥ 10 mm[144], particularly the non-granular pseudodepressed type, are strongly associated with deep invasion[145-147]. Both rectosigmoid location and increasing lesion size are independent predictors of dSMI: Each 10 mm increase in diameter is associated with an OR of 1.12, while rectosigmoid location carries an OR of 1.91[148]. Surface ulceration, irregular or raised borders, and converging or retracted folds are all associated with a higher likelihood of malignancy[149,150]. The non-lifting sign after submucosal injection remains an important intraprocedural indicator: In treatment-naive lesions, failure to lift strongly suggests deep invasion, as tumor infiltration into the submucosa prevents adequate tissue elevation[151,152]. High-definition endoscopy, including virtual chromoendoscopy such as NBI, has markedly improved real-time histological assessment of colorectal lesions, particularly for predicting dSMI[153]. The most widely adopted classification systems are the NBI International Colorectal Endoscopic (NICE) classification[154,155] and the Japanese NBI Expert Team (JNET) classification[156,157]. Both NICE type 3 and JNET type 3 patterns are highly specific for identifying dSMI, enabling confident identification of lesions at high risk of LNM[158,159].

From a clinical standpoint, these optical criteria support a pragmatic dichotomous risk stratification. High-risk lesions, suggestive of dSMI and LNM, include those classified as NICE type 3 or JNET type 3, LSTs of the non-granular pseudodepressed subtype, and lesions exhibiting ulceration or a non-lifting sign. Low-risk lesions, consistent with superficial invasion, are characterized by the absence of high-risk features and include those with JNET type 2 patterns. This distinction directly informs the choice between endoscopic and surgical management. A major limitation, however, remains the suboptimal sensitivity of optical diagnosis, meaning that some deeply invasive cancers may still be misclassified as superficial.

Treatment selection

Current ESGE and United States Multi-Society Task Force on CRC guidelines[138,139] converge on a risk-adapted strategy based primarily on optical diagnosis. Lesions classified as high-risk should be referred directly to surgery or multidisciplinary evaluation, as endoscopic resection is unlikely to be curative and risks incomplete resection with missed LNM. Conversely, lesions assessed as low-risk are appropriate candidates for en bloc endoscopic resection (EMR or ESD), which serves both diagnostic and therapeutic purposes. This strategy is critical to avoid overtreatment, as a substantial proportion of T1 cancers, particularly those without additional high-risk histological features, carry a very low risk of LNM and can be definitively cured endoscopically.

The primary goal of endoscopic therapy is to achieve R0 resection, defined by histologically negative margins, while minimizing the risk of LNM. R0 resection rates are highest with ESD, particularly for lesions > 20 mm or those with superficial submucosal invasion. Multiple studies have demonstrated that ESD achieves en bloc and R0 resection rates of approximately 90%-94% and 89%-90%, respectively[160-162], significantly higher than those obtained with EMR. Meta-analyses comparing ESD and EMR have shown that ESD yields en bloc resection rates approximately seven times higher and R0 resection rates three to four times higher[163,164]. A recently published Dutch study showed that between 2015 and 2022, the proportion of primary surgery for T1 colon cancers decreased from 53.2% to 29.7%, without compromising five-year cancer-specific or overall survival[165].

Histological evaluation of malignant polyps provides essential information for estimating the risk of LNM and guiding subsequent management. The strongest predictors of LNM have been consistently identified across large meta-analyses: Submucosal invasion beyond 1000 μm, vascular invasion, lymphatic invasion, poor differentiation, and tumor budding are all associated with significantly increased LNM risk[166,167]. Poorly differentiated adenocarcinoma is a strong independent predictor of LNM, with an OR of 8.27[167]. Another review confirmed its association with both residual disease (OR = 2.2) and LNM (OR = 3.9)[168]. Depth of submucosal invasion remains a debated factor: Some studies report that invasion ≥ 1 mm triples the risk of LNM (OR = 3.00)[167], confirming the clinical significance of the 1000 μm cutoff (OR = 3.53)[169]. However, other authors have found that dSMI alone, in the absence of other risk factors, confers a modest absolute LNM risk (2.6%) and is not a strong independent predictor in multivariate models (OR = 1.73)[170,171]. Lymphovascular invasion is among the most powerful prognostic indicators, with ORs ranging from 4.7 to 7.0 in large analyses[167,168,172]. High-grade tumor budding also confers a substantially increased risk, with an OR of 4.59[167]. In surgical series, tumor budding was identified in 42% of LNM-positive cases compared with 18% of LNM-negative cases (OR = 2.3)[173].

Margin status is another critical determinant of oncological outcomes following endoscopic resection of pT1 CRC. A positive or indeterminate (R1/Rx) vertical margin is associated with a markedly increased risk of local recurrence and residual carcinoma. Recurrence or residual tumor rates after resection with a margin < 1 mm have been reported as high as 21%-33%, with correspondingly worse outcomes including residual disease, recurrence, LNM, and mortality, while clear margins (≥ 1 mm) are associated with recurrence rates of 0-2%[168]. The definition of a “positive margin” is not universally standardized, and polypectomy-related artifacts can lead to overestimation: In one study, 63% of resected malignant polyps had positive margins, yet only 2.8% of subsequent colectomy specimens contained residual carcinoma[174]. Notably, many local recurrences are amenable to further endoscopic management. Positive horizontal margins do not appear to increase the risk of LNM, supporting a conservative observational approach when other high-risk histological features are absent[175]. In contrast, positive vertical margins are more strongly associated with residual invasive disease and may justify additional surgery, especially in the presence of other adverse prognostic factors[176-178].

Histological evaluation following endoscopic resection of T1 CRC should be used to stratify patients into low-risk and high-risk categories for LNM, in accordance with current guideline recommendations. Low-risk T1 cancers are defined by all of the following criteria: En bloc R0 resection, absence of lymphovascular invasion, well or moderately differentiated histology, low-grade tumor budding, and superficial submucosal invasion. In these cases, the estimated risk of LNM is very low (< 1%-2%), endoscopic resection is considered curative, and no additional surgery is indicated[179]. High-risk T1 cancers are defined by the presence of one or more adverse histological features, lymphovascular invasion, poor differentiation, high-grade tumor budding, positive or indeterminate vertical resection margins, or dSMI in combination with other risk factors. These lesions carry a clinically relevant risk of LNM, and additional surgical resection with lymph node dissection is recommended in patients fit for surgery.

Post ESD/EMR follow-up

After R0 resection achieved by ESD or EMR, particularly in the absence of high-risk histopathological features, endoscopic surveillance alone yields excellent long-term outcomes. Five-year recurrence-free survival exceeds 94% and overall survival surpasses 97%[180]. The risk of local recurrence is primarily influenced by incomplete resection (R1/Rx), tumor budding, and comorbidity burden, rather than by the resection technique, provided en bloc resection is achieved[181].

Recent prospective cohort studies and systematic reviews have confirmed that, after curative ESD for T1 colorectal carcinoma, both local and distant recurrence risks are extremely low, with local recurrence rates ranging from 0.5% to 2.2% and metachronous cancer rates between 0.22% and 1.5% at five years[126,182-184]. Most recurrences occur within the first six years, with over 95% detected within 72 months[179]. For patients achieving en bloc R0 resection without adverse histopathological features, evidence supports performing the first surveillance colonoscopy at one year, given the low risk profile and cost-effectiveness of this approach[185]. Surveillance should include high-definition inspection of the resection site and the entire colon to exclude synchronous or metachronous neoplasia, as most metachronous cancers are detected during scheduled follow-up. Subsequent colonoscopy intervals should follow standard post-polypectomy and CRC surveillance guidelines[104].

For patients with non-curative endoscopic resection, defined by R1 margins, dSMI, lymphovascular invasion, poor differentiation, or high-grade tumor budding, the risk of residual disease and LNM increases substantially, and additional surgery is generally recommended. However, multidisciplinary risk assessment is essential, as surgical overtreatment remains a concern: Over 90% of high-risk T1 CRC patients show no nodal involvement at colectomy[186]. The recurrence rate after endoscopic treatment is approximately 7% in high-risk T1 CRC, comparable to rates observed after surgical resection, with both local and distant recurrences occurring with similar frequency and the vast majority developing within 72 months[179]. These findings support surveillance strategies that include monitoring for both local and distant disease for at least six years following resection. In patients with major comorbidities or high surgical risk, close endoscopic and radiological surveillance may represent a reasonable alternative, as disease-specific survival does not differ significantly between surveillance and surgical management in carefully selected high-risk groups[181,187].

The high-risk T1 CRC category is markedly heterogeneous, and not all risk features carry the same prognostic weight. This variability limits the applicability of uniform surveillance protocols and highlights the importance of individualized follow-up tailored to the patient’s clinical status and preferences. At present, there is no robust evidence that intensive surveillance strategies, colonoscopy at 3, 6, and 12 months, then every 6 months until year two, then annually until year five, improve survival outcomes[73,126,188]. Piecemeal resection of T1 CRC precludes accurate histological assessment and is an independent risk factor for local recurrence; accordingly, T1 cancers resected piecemeal should generally be referred for surgery owing to the risk of understaging and residual disease[189]. In patients unfit for or refusing surgery, intensive endoscopic surveillance, colonoscopy within six months with advanced imaging of the resection site, followed by close monitoring at 12 months and 24 months, represents a reasonable alternative within a multidisciplinary decision-making framework.

POST-CONSERVATIVE MEDICAL TREATMENT SURVEILLANCE

The possibility of avoiding surgery in selected patients with CRC represents a recent and still evolving development. It is essential, however, to distinguish two distinct scenarios. In rectal cancer, the watch-and-wait strategy is now supported by extensive international experience: In patients who achieve a clinical complete response (cCR) after neoadjuvant therapy, opting not to proceed with resection can allow organ preservation without compromising oncological outcomes, provided an intensive surveillance program is adopted[190-193]. In colon cancer, by contrast, surgery remains the standard of care, and a conservative approach is reserved for highly selected contexts, particularly patients with microsatellite instability-high (MSI-H)/deficient mismatch repair (dMMR) tumors treated with immunotherapy who achieve complete radiological and clinical responses[194].

In locally advanced rectal cancer, the introduction of total neoadjuvant therapy (TNT) has further increased the likelihood of achieving a cCR. TNT refers to an approach combining systemic intravenous chemotherapy (typically FOLFOX or CAPOX, administered as induction) with preoperative chemoradiotherapy, aiming to maximize pathological complete response and bring systemic treatment forward into the preoperative setting. Studies such as the OPRA trial have shown that roughly half of patients achieving a complete response can avoid surgery, maintaining survival rates comparable to those after total mesorectal excision, provided surveillance is stringent and salvage intervention is feasible in the event of progression[192,195].

Follow-up in these patients is inevitably more intensive than after standard surgical management. Since most recurrences occur within the first two to three years, surveillance should be particularly stringent during this period. In patients with rectal cancer managed with watch-and-wait after cCR, current practice in expert centers includes clinical evaluation with digital rectal examination and endoscopy every 3-4 months, pelvic magnetic resonance imaging every six months, regular CEA monitoring, and annual chest-abdomen CT or earlier if clinically indicated. This intensive approach is recommended for at least the first three years, when the risk of local regrowth is highest, and should be conducted in centers with expertise in early detection and prompt salvage intervention. Thereafter, endoscopic surveillance can be progressively de-intensified, with examinations every six months between years three and five, and annually beyond year five.

A similar surveillance strategy may be adopted in patients with locally advanced rectal cancer treated with TNT who achieve a cCR and are managed non-operatively. In this context, strict adherence to follow-up is essential, as early detection of regrowth significantly increases the likelihood of successful salvage and long-term organ preservation[196,197].

In parallel, for MSI-H/dMMR tumors, particularly of the rectum but also of the colon, immunotherapy has opened entirely new clinical scenarios. In the phase II study published by Cercek et al[194], all patients treated with an anti-programmed death-1 agent achieved a cCR without the need for surgery. Although these represent limited series with short follow-up and non-standardized surveillance protocols, they raise important questions: How long to maintain intensive surveillance in responders, which diagnostic tools to integrate (such as ctDNA), how to reliably distinguish fibrosis from residual disease, and how to manage potential resistance or late recurrences.

Looking ahead, surveillance programs are likely to become increasingly personalized[198]. In patients with rectal cancer treated with TNT and watch-and-wait, as well as in those with MSI-H tumors treated with immunotherapy, the challenge will be to define shared criteria for determining the optimal duration of close monitoring, the conditions for its de-intensification, and the integrative role of molecular biomarkers[199]. Until more robust data are available, the most prudent strategy remains tight surveillance conducted in experienced centers with rapid restaging pathways, to ensure timely diagnosis and the possibility of salvage intervention when needed. An overview of the non-surgical management strategies in CRC is summarized in Table 4.

Table 4 Non-surgical management strategies in colorectal cancer.

Setting	Key evidence	Criteria to avoid surgery	Recommended follow-up	Key recommendations
Rectal cancer - “watch and wait” strategy	Large international experience; organ preservation possible without compromising oncologic outcomes	cCR after neoadjuvant therapy. Assessment in expert centers. Ability to perform salvage surgery if needed	Clinical visit + digital rectal exam every 3-4 months. Endoscopy every 3-4 months. Pelvic MRI every 6 months. CEA monitoring. Annual chest-abdomen CT	Use only after strict confirmation of cCR. Intensive surveillance for at least 2-3 years. Requires multidisciplinary management
Locally advanced rectal cancer treated with TNT	TNT increases cCR rates and allows organ preservation (e.g., OPRA: Approximately 50% of complete responders avoid surgery)	Clinical complete response after TNT (chemo + chemoradiotherapy). Patient commitment to intensive follow-up	Same as above: Intensive surveillance concentrated in first 2-3 years	Real possibility to avoid TME. Early detection of recurrence is crucial for successful salvage
MSI-H/dMMR tumors (rectum and colon) treated with immunotherapy	Early studies (e.g., Cercek): All patients achieved clinical complete response without surgery after anti-PD1 therapy	Confirmed MSI-H/dMMR. Complete clinical/radiologic response after immunotherapy	Not yet standardized; likely prolonged intensive surveillance. Possible integration of ctDNA	Monitoring strategy still evolving. Watch for late recurrences or resistance. Should be managed only in highly experienced centers
Colon cancer (non-MSI-H/dMMR)	Surgery remains the standard of care; conservative management only in extremely selected cases	Rare complete response, generally only in MSI-H tumors treated with immunotherapy	Standard oncologic follow-up	Avoid non-surgical strategies outside specialized protocols or exceptional cases

cCR: Clinical complete response; CEA: Carcinoembryonic antigen; MRI: Magnetic resonance imaging; CT: Computed tomography; TNT: Total neoadjuvant therapy; TME: Total mesorectal excision; MSI-H: Microsatellite instability-high; dMMR: Deficient mismatch repair; ctDNA: Circulating tumor DNA; PD1: Programmed death 1.

CLINICAL CARE PATHWAY

In Figure 1, we propose a pragmatic and risk-adapted clinical care pathway integrating oncological follow-up with endoscopic surveillance across three pathways: (1) Post-surgical surveillance after curative resection; (2) Post-resective endoscopic surveillance after EMR/ESD of pT1 CRC; and (3) Non-operative management (watch-and-wait) after cCR to neoadjuvant therapy in rectal cancer. Cross-cutting modules address the management of precancerous lesions detected during surveillance, hereditary syndromes, and adherence and coordination of care.

Figure 1 Proposed clinical care pathway for endoscopic follow-up in colorectal cancer. IPAA: Ileal pouch-anal anastomosis; IRA: Ileorectal anastomosis; CS: Colonoscopy; sSMI: Superficial submucosal invasion; LVI: Lymphovascular invasion; dSMI: Deep submucosal invasion; cCR: Clinical complete response; MRI: Magnetic resonance imaging; MSI: Microsatellite instability; MMR: Mismatch repair; ctDNA: Circulating tumor DNA.

CONCLUSION

Recent advances in endoscopic and histopathological evaluation have significantly improved the detection and management of colorectal neoplasia. Accurate assessment of invasion depth, lymphovascular involvement, and resection margins is essential to guide treatment decisions and avoid unnecessary surgery. Endoscopic resection offers curative potential for early lesions when R0 resection is achieved and no high-risk features are present. Growing evidence supports tailored surveillance strategies and organ-preserving approaches in selected patients, particularly after curative endoscopic resection or in rectal cancer managed with TNT and watch-and-wait. Immunotherapy has further expanded therapeutic options for MSI-H/dMMR tumors. Ongoing research should aim to refine risk stratification, integrate molecular biomarkers, and optimize surveillance protocols to ensure safe, effective, and personalized CRC care.