INTRODUCTION
Rectal cancer is a major global health burden and is one of the most commonly diagnosed malignancies worldwide. According to recent global cancer statistics, colorectal cancer ranks among the top three cancers in terms of incidence and remains a leading cause of cancer-related mortality, with rectal cancer accounting for a substantial proportion of cases[1]. Despite improvements in screening, surgical techniques, and multimodal therapy, significant geographic variation persists in the incidence, mortality, and access to optimal treatment, particularly among high-, low-, and middle-income countries[2].
The management of rectal cancer has shifted from a surgery-centered paradigm to a decision-making model based on the treatment response and risk stratification. Although the traditional approach of high-quality total mesorectal excision combined with preoperative long-course radiotherapy and adjuvant chemotherapy has significantly reduced local recurrence rates, distant metastasis remains the major limiting factor for long-term survival[3,4]. Total neoadjuvant therapy (TNT) integrates full-dose chemotherapy and radiotherapy preoperatively, which improves treatment completion rates, promotes profound tumor regression, and reduces the risk of metastasis-related failure. This makes the depth and durability of the treatment response the core basis for individualized decision-making[5,6].
Accordingly, the focus of preoperative imaging assessment has expanded from initial staging and resectability evaluation to response stratification and recurrence risk reconstruction during restaging after neoadjuvant therapy. It covers high-risk signs, including the scope of residual tumor activity, spatial relationship with the mesorectal fascia (MRF), circumferential resection margin (CRM) risk, extramural venous invasion (EMVI) status, lymph node characteristics, tumor deposits, and pelvic sidewall invasion[7]. Magnetic resonance imaging (MRI) has become a pivotal tool because of its high soft-tissue resolution and multiparametric imaging capability. However, treatment-induced fibrosis, inflammation, and mucin pool formation lead to signal overlap, significantly increasing the uncertainty of image interpretation[8]. Therefore, rigorous imaging–pathology correlation studies are crucial for clarifying the predictive endpoints of MRI, identifying sources of bias, and establishing their association with clinical outcomes, ultimately serving as the scientific foundation for constructing a precise risk stratification system (Figure 1).
Figure 1 Research progress on magnetic resonance imaging evaluation and postoperative pathology comparison after neoadjuvant therapy for rectal cancer.
MRI: Magnetic resonance imaging; W&W: Watch-and-wait; EMVI: Extramural venous invasion; CRM: Circumferential resection margin.
EVOLUTION OF NEOADJUVANT THERAPY AND THE DEMAND FOR EFFICACY ASSESSMENT
The core effects of neoadjuvant therapy are not only reflected in tumor volume reduction but also involve decreased tumor cell density, disruption of glandular structures, and remodeling of the tumor-stroma microenvironment, ultimately forming a fibrosis-dominant tumor bed, often accompanied by pathological changes, such as mucin pools, inflammatory infiltration, and necrosis[9]. The degree of treatment response is affected by multiple factors, including tumor biological heterogeneity, treatment regimen composition, radiotherapy dose-fractionation mode, chemotherapy intensity, and interval from treatment to surgery, showing significant interindividual variability and time dependence[10,11]. Thus, efficacy assessment should not be limited to static imaging at a single time point, but rather to a comprehensive judgment integrating the dynamic trajectory of tumor regression and the laws of biological evolution.
Postoperative pathological stage (ypTNM) and pathological tumor regression grade (pTRG) remain the gold standards for efficacy evaluation. Although pathological complete response (pCR; ypT0N0) is associated with a lower risk of local recurrence and a more favorable prognosis, it does not equate to “risk elimination”. Clinical decision-making still requires individualized risk management combined with lymph node status, vascular/nerve invasion, and other high-risk factors[12,13]. Notably, there are threshold differences among different tumor regression grade systems (e.g., Mandard, Dworak). In addition, pathological sampling density, section interval, full-thickness sampling strategy of the tumor bed, and staining methods affect the detection of minimal residual disease and the stability of grading[9]. In cases of near-complete response, residual tumors often present as scattered cancer nests or intrastromal infiltration; insufficient sampling tends to underestimate the lesion burden, leading to an apparent imaging–pathology inconsistency of false-positive results[11]. With the popularization of organ-sparing strategies, the availability of postoperative pathological specimens has decreased, and the evidence weight of imaging and multimodal clinical assessment in treatment decision-making has increased significantly.
RECTAL MRI EXAMINATION NORMS AND REPORTING ELEMENTS AFTER NEOADJUVANT THERAPY
Imaging features and sequence integration
The core challenge of MRI assessment after neoadjuvant therapy stems from the imaging phenotypic overlap caused by histological remodeling and the physical limitations of spatial resolution in detecting minimal residual lesions[14]. T2-weighted imaging (T2WI) can sensitively detect fibrosis, scarring, and edema. However, scattered micronodular residual lesions are susceptible to partial volume effect interference and signal masking. Diffusion-weighted imaging (DWI) reflects cellular density through the diffusion characteristics of water molecules, providing a functional basis for viable residual tumor; however, inflammatory infiltration, hemorrhage or artifacts can induce nonspecific hyperintensity, thereby increasing the uncertainty in the interpretation of “residual tumor”[15,16]. Dynamic contrast-enhanced sequences help identify tissue characteristics and define tumor bed boundaries in specific cases and have been used in some research settings to improve the discriminatory ability associated with postoperative pathological stratification; however, their added value usually relies on strict indication selection and multisequence comparative interpretation[17]. Clinical practice must strictly adhere to the principle of multisequence evidence integration and avoid a definitive diagnosis based on a single sequence.
Standardized scanning protocol and quality control
Standardized MRI acquisition is essential for a consistent assessment. High-resolution oblique axial T2WI serves as the core sequence, supplemented by sagittal and coronal planes, to adequately cover the pelvic anatomy. DWI should balance geometric distortion and signal-to-noise ratio using optimized b values (e.g., 0, 800, 1000 seconds/mm²) and ensure accurate spatial alignment with T2WI to minimize misregistration and misinterpretation of minimal residual disease[18].
In low rectal cancer, imaging should clearly delineate the anal canal, sphincter complex, and levator ani to support the assessment of sphincter-preserving feasibility[19,20]. Image quality is influenced by bowel preparation, antiperistaltic agents, and acquisition parameters, and inter-center variability remains a major barrier to comparability. Therefore, standardized acquisition protocols and quality control checklists are necessary[15,18]. Multicenter data show that experienced radiologists achieve higher agreement in evaluating complete response, MRF involvement, and EMVI, underscoring the importance of structured training and double-reading systems[21].
Structured reporting for clinical decision-making
The timing of posttreatment MRI critically affects its reliability. Because tumor regression is often delayed, early reassessment may overestimate the residual disease. Extending the interval or performing sequential imaging improves the differentiation between fibrotic and viable tumor[22].
Structured reports should systematically address the following: (1) Location, extent, and distance from the anal verge of residual nodular lesions or focal restricted diffusion within the tumor bed; (2) Shortest distance between the tumor bed and the MRF for CRM risk stratification[23]; (3) Number, distribution, short-axis diameter, and signal characteristics of mesorectal and lateral lymph nodes[24]; (4) Posttreatment EMVI status[25]; and (5) High-risk features, including tumor deposits, pelvic sidewall invasion, peritoneal reflection involvement, and adjacent organ invasion[26].
This structured approach prioritizes actionable risk factors and directly informs surgical plane optimization, combined resection decisions, lateral pelvic lymph node management, and follow-up strategies, thereby facilitating the translation of imaging findings into individualized surgical planning.
ADVANCES IN MRI ASSESSMENT INDICES AND IMAGING-PATHOLOGY CORRELATION STUDIES AFTER NEOADJUVANT THERAPY
Core objectives of imaging–pathology correlation studies and systematic assessment framework
Imaging-pathology correlation studies aim to systematically evaluate the predictive efficacy of MRI for key pathological endpoints and their clinical relevance, rather than pursuing precise morphological “reproduction”. The post-neoadjuvant therapy assessment framework includes indices such as ymrT, ymrN, mrTRG, ymrMRF, and ymrEMVI; however, there is significant heterogeneity in their imaging–pathology consistency. MRI provides a reliable assessment of macroscopic anatomical relationships (e.g., the spatial relationship between the tumor bed and the key structures), whereas the identification of minimal residual lesions and micrometastases is limited by physical resolution. Clarifying such discrepancies is a prerequisite for the rational translation of imaging findings into evidence for clinical decision-making[27,28].
Imaging features and interpretative limitations of restaging for the primary tumor bed after treatment
The accuracy of posttreatment ymrT staging is generally lower than that of initial staging, mainly due to the blurring of intestinal wall layer structures caused by fibrotic and inflammatory remodeling, which invalidates the traditional basis of “layer-based staging”. Clinical decision-making should focus more on the presence of extramural invasion and the distance between the tumor bed and the MRF to complete CRM risk stratification. As an imaging inference of the pathological regression degree, mrTRG is correlated with pTRG but without a linear correspondence and exhibits high specificity but low sensitivity in predicting pCR. When MRI shows typical cicatricial fibrosis and no nodular solid components or focal restricted diffusion on DWI, the possibility of a complete response is high; however, mild nonspecific signal abnormalities cannot rule out minimal residual disease[29,30]. Therefore, for watch-and-wait (W&W) strategies or local excision decisions, MRI should be integrated with digital rectal examination, endoscopy, and biopsy in a multimodal approach, and dynamic re-examinations should be performed to reduce the risk of missed diagnoses[31]. Although quantitative parameters, such as DWI/apparent diffusion coefficient (ADC) values, can reflect changes in cellular density, they are affected by inflammatory and mucinous components. Recent studies tend to adopt combined “morphological + functional imaging” criteria or introduce machine learning/radiomics frameworks to improve the stability and reproducibility of predictions[32-34].
Restaging uncertainties of regional lymph nodes and vascular invasion and the significance for risk stratification
ymrN staging is a relatively weak aspect of MRI assessment. Neoadjuvant therapy may cause metastatic lymph nodes to shrink or undergo fibrosis, making sole reliance on short-axis diameter thresholds prone to underestimating residual metastases. Meanwhile, reactive hyperplasia increases the false-positive rate. Pathological “positivity” often includes microscopic lesions, which do not necessarily correspond to macroscopic imaging features. Thus, a negative ymrN result should be regarded as “risk reduction” rather than “absolute exclusion”. Especially in W&W decision-making, assessment of the extramesorectal and lateral pelvic regions should be strengthened[35,36]. ymrEMVI has evolved from a staging element to an important imaging biomarker; its persistent positivity indicates a high systemic risk, and risk stratification should not be downgraded owing to local tumor regression[37]. Notably, pathological EMVI is susceptible to sampling bias, whereas MRI reflects macroscopic vascular abnormalities, and the two are not synonymous. Clinically, it is advisable to integrate ymrEMVI and CRM risk into a dual-axis assessment framework of “local radical cure-systemic control”. In the Asian population, the assessment of lateral pelvic lymph nodes is particularly important. Even if nodules shrink after treatment, persistent abnormal morphology or signal still suggests the possibility of metastasis, and decisions should be made in conjunction with the surgical scope, radiotherapy target volume, and follow-up strategies in multidisciplinary team (MDT) discussions.
Clinical translational value of CRM status assessment and integration into multidisciplinary decision-making
The MRF/CRM status is the most directly correlated parameter between MRI assessment and surgical decision-making. Post-treatment assessment should not only document whether the MRF has converted to negative but also clarify the sources of threat (primary lesion, lymph nodes, tumor deposits, or EMVI-related vascular abnormalities) to guide individualized surgical strategies: Threats from the primary lesion require optimization of the resection plane or consideration of combined organ resection; threats from lymph nodes emphasize local dissection; persistent EMVI or tumor deposits indicate a high risk of systemic dissemination[38]. Compared with ymrT and ymrN, MRF/CRM assessment has higher clinical “usability” and should be highlighted as a core entry in structured reports to provide precise anatomical evidence for multidisciplinary decision-making[39].
MECHANISTIC BASIS AND INFLUENCING FACTORS OF IMAGING–PATHOLOGY INCONSISTENCY
Multidimensional pathogenic mechanisms of imaging-pathology inconsistency
Inconsistencies between imaging and pathological findings are relatively common and mainly associated with histological heterogeneity, physical constraints of imaging, variability in diagnosis and treatment processes, and subjectivity in image interpretation[40]. After neoadjuvant therapy, the tumor bed is a mixture of fibrosis, inflammation, necrosis, mucin pools, and residual tumors. Tumor regression is time-dependent, with imaging phenotypes changing with the treatment course and assessment intervals. On T2WI, mucin/edema typically shows hyperintensity, fibrosis presents hypointensity, and residual tumor mostly exhibits an intermediate signal intensity that easily overlaps with inflammatory changes; DWI can indicate restricted diffusion, yet inflammatory infiltration may also lead to false-positive results[41,42]. Therefore, posttreatment MRI reports should describe the probability of residual tumors based on the weight of the evidence-probability stratification approach, adopt mutual verification of T2WI and DWI findings with attention to longitudinal changes, avoid simple binary classification, and classify uncertain results into dynamic follow-up and re-evaluation protocols.
Limitations of spatial resolution and challenges in detecting minimal residual lesions
MRI spatial resolution is limited by voxel size and partial volume effect; minimal residual lesions may be signal-“diluted” or masked by background tissues, and scattered cancer nests are particularly difficult to detect[43]. Micrometastases in lymph nodes may retain tumor cells after shrinkage and fibrosis, with no stable or specific imaging signs[44]. Pathological examination excels in microscopic observation and systematic sampling, whereas MRI is advantageous in preoperative global anatomical assessment, margin evaluation, and related risk stratification. Clinically, it is advisable to prioritize high-impact parameters such as CRM and combine endoscopic/clinical information with multimodal assessment to reduce deviations in surgical strategies caused by missed diagnoses[45].
Impact of variability in diagnosis and treatment processes on assessment consistency
Inter-center inconsistencies in scanning parameters, DWI quality control, patient preparation, and image interpretation training can lead to fluctuating results. The combined criterion of “T2WI + DWI” integrated with structured reporting key points yields more robust results, but inflammation, mucin, and artifacts may still form gray zones of interpretation[46]. Thus, it is necessary to unify the scanning protocols, strengthen DWI quality control and specialist training, and conduct MDT reviews when necessary to improve cross-institutional comparability. In addition, pathological correlation is affected by sampling density and regression grading systems: Intensive full-thickness sampling achieves higher sensitivity, whereas rough sampling tends to underestimate residual tumor burden, seemingly amplifying the proportion of false-positive imaging results[47].
Time-dependent tumor regression and dynamic assessment strategies
Tumor regression is a time-dependent process, with postradiotherapy fibrotic remodeling and tumor clearance persisting for weeks to months. Suspected signals detected by early MRI may evolve into stable scar tissue over time, whereas true residual tumors may develop into more distinct nodules or foci with restricted diffusion[45]. A single-time-point assessment is prone to systematic bias, and dynamic follow-up is more consistent with the biological processes of tumor regression. Especially in organ preservation/W&W scenarios, sequential imaging combined with clinical information is required for the interpretation of gray zone findings such as “mucin residue and mild signal abnormalities”[48]. For patients with a near-complete response, a stratified strategy of delayed re-evaluation and close follow-up is superior to a “one-size-fits-all” decision, which can reduce the risk of misjudgment and optimize the robustness of clinical decision-making. Integrating imaging features at multiple time points can improve the accuracy of treatment response assessment[41].
IMPLICATIONS FOR SURGICAL PROGNOSTIC JUDGMENT AND TREATMENT STRATEGIES
Posttreatment MRI plays a pivotal role in individualized surgical planning after neoadjuvant therapy. Its clinical value can be summarized into three core domains: (1) Assessment of CRM risk; (2) Support for organ preservation strategies; and (3) Prognostic stratification to guide follow-up and adjuvant management.
CRM risk assessment and optimization of radical resection strategy
Accurate evaluation of CRM risk is fundamental to surgical planning. MRI-based determination of the ymrMRF status directly influences the choice of surgical approach. When the MRF is threatened, modification of the resection plane, consideration of multivisceral resection, and targeted management of the suspected invaded structures are required[49].
For low rectal tumors, high-resolution assessment of the anal canal and sphincter complex provides an anatomical basis for determining the feasibility of inter-sphincteric resection and sphincter preservation[50]. Thus, MRI findings directly translate into decisions regarding the extent and complexity of radical surgeries.
Imaging support for organ preservation strategies
In organ-preserving approaches, MRI is a critical component of multimodal response assessment. Safe implementation of W&W strategies requires strict imaging criteria: Absence of nodular solid residue in the tumor bed; no focal restricted diffusion on DWI; no suspicious mesorectal or lateral pelvic lesions; nonthreatened MRF status; no persistent EMVI[51].
Given the inherent limitations of imaging, W&W strategies must be embedded within a standardized and intensive follow-up protocol, particularly during the first 2 years. For patients with near-complete response but persistent nodules or unexplained restricted diffusion, delayed reassessment, local excision, or radical surgery should be considered rather than premature enrollment in a W&W model[52]. Therefore, MRI acts as both a selection tool and a safety checkpoint for organ preservation.
Prognostic stratification and individualized follow-up
Posttreatment MRI enables the early identification of imaging biomarkers associated with recurrence risk. Persistently positive ymrEMVI, tumor deposits, or pelvic sidewall invasion indicate an elevated risk of systemic dissemination. In contrast, poor mrTRG response, residual MRF involvement, and nodal persistence suggest an increased risk of local recurrence and warrant intensified pelvic surveillance[53,54].
These findings should be translated into tailored management strategies, including adjustment of follow-up frequency, enhanced metastatic screening, and optimization of adjuvant therapy.
Multidisciplinary integration and dynamic feedback
Within a multidisciplinary framework, a closed-loop system linking preoperative MRI assessments, intraoperative findings, and postoperative pathology can refine interpretative standards and reduce systematic bias. With the widespread adoption of TNT, follow-up imaging has become increasingly important for monitoring local regrowth[55].
Accordingly, a standardized pathway centered on specialized MRI assessment-coordinated with endoscopic reassessment and surgical decision-making-is essential to ensure the safe implementation of organ-preserving strategies. By identifying high-risk features and guiding intervention timing, posttreatment MRI serves as a critical bridge between neoadjuvant therapy and individualized surgical management.
FUTURE DIRECTIONS OF IMAGING ASSESSMENT AFTER NEOADJUVANT THERAPY FOR RECTAL CANCER
Future research on post-neoadjuvant imaging for rectal cancer should focus on advancing its clinical utility in guiding individualized surgical decision-making. Building on the current challenges, four key directions are particularly relevant: Quantification, standardization, integration, and intelligent imaging.
Quantitative imaging biomarkers
Quantitative MRI parameters, such as the ADC, intravoxel incoherent motion, diffusion kurtosis imaging, and dynamic contrast-enhanced MRI, offer objective measures of tumor response, thus reducing reliance on subjective interpretation. Longitudinal assessment of these parameters in patients with a near-complete response can better capture the dynamics of tumor regression than single time-point measurements, supporting more precise decisions on organ preservation or the timing of surgery[56,57].
Standardization and quality control
To improve reproducibility and multicenter comparability, future studies should focus on unifying scanning protocols, optimizing DWI quality, popularizing structured reporting, and implementing formalized radiologist training. Standardization will minimize interobserver variability and ensure that imaging-derived assessments of CRM, EMVI, and tumor regression can reliably guide surgical planning[58].
Multimodal integration
Imaging should be integrated with pathology, endoscopy, digital rectal examination, and molecular biomarkers within a multidisciplinary framework. For W&W strategies, combining follow-up MRI with endoscopic and clinical findings can provide dynamic risk models that optimize surveillance intervals and intervention timing[59]. Integration also enables a more precise risk stratification for local recurrence and systemic metastasis, directly informing individualized surgical strategies.
Intelligent imaging and artificial intelligence applications
Radiomics and artificial intelligence offer the potential to extract high-dimensional imaging features and map them to pathological and clinical endpoints. Current limitations include feature instability, data heterogeneity, limited external validation, and interpretability challenges[60,61]. Future research should focus on validating artificial intelligence (AI)-based models in standardized multicenter cohorts, targeting clinically meaningful endpoints, such as local regrowth, disease-free survival, and CRM involvement. The ultimate goal is to embed AI tools into structured reporting and MDT workflows, transforming them from research prototypes into actionable clinical decision support systems.
CONCLUSION
In the clinical practice of neoadjuvant therapy for rectal cancer, posttreatment MRI has become a core tool for surgical prognostic assessment and optimization of treatment strategies. There are inherent structural discrepancies between imaging assessments and postoperative pathological findings. MRI has a limited ability in identifying minimal residual lesions and lymph node micrometastases, but it offers unique value in predicting CRM risk, assessing MRF involvement status, and identifying EMVI and high-risk lateral pelvic regions.
The rational clinical approach is not to pursue a complete replacement of pathology by MRI but to translate imaging findings into risk stratification parameters with surgical decision-making value based on a full understanding of its technical limitations through standardized high-quality scanning, structured reporting systems, multimodal imaging integration, and dynamic sequential assessment. With the clinical popularization of organ function preservation strategies and the W&W model, in-depth imaging-pathology correlation studies will continuously optimize the accuracy of assessment.
The ultimate goal is to construct an integrated and precise surgical management system centered on the paradigm of treatment response, recurrence risk, intervention strategy, and individualized follow-up to maximize the benefits of organ function preservation and long-term survival for patients while ensuring oncological safety.
Peer review: Externally peer reviewed.
Peer-review model: Single blind
Specialty type: Gastroenterology and hepatology
Country of origin: China
Peer-review report’s classification
Scientific Quality: Grade B
Novelty: Grade C
Creativity or Innovation: Grade B
Scientific Significance: Grade C
P-Reviewer: Tougeron D, Assistant Professor, PhD, Researcher, France S-Editor: Qu XL L-Editor: A P-Editor: Zhao YQ
