Watch & Wait - Post Neoadjuvant Imaging for Rectal Cancer

Abstract

Rectal cancer management has evolved over the past decade with the emergence of total neoadjuvant therapy (TNT). For select patients who achieve a clinical complete response following TNT, organ preservation by means of the watch-and-wait (WW) strategy is an increasingly adopted alternative that preserves rectal function and quality of life without compromising oncologic outcomes. Recently, published 5-year results from the OPRA trial demonstrated that organ preservation can be achieved in approximately half of patients managed with the WW strategy, with most local regrowth events occurring within two years. Considering the potential for local regrowth, the implementation of the WW strategy mandates rigorous clinical and radiographic surveillance. Magnetic resonance imaging (MRI) serves as the conventional imaging modality for local staging and surveillance of rectal cancer given its excellent soft-tissue resolution. This review will discuss the current evidence for the WW strategy and the role of restaging rectal MRI in determining patient eligibility for this strategy. Restaging rectal MRI acquisition parameters and treatment response assessment, including important factors to assess, pitfalls, and classification systems, will be discussed in the context of the WW strategy.

Introduction

The incidence of rectal cancer is rapidly rising worldwide, particularly among young adults [1]. The past decade has seen an evolution in the treatment paradigm for rectal cancer, as total neoadjuvant therapy (TNT) has demonstrated superior oncologic outcomes (treatment response, 3-year survival) and treatment compliance compared to the previous standard of neoadjuvant chemoradiation, total mesorectal excision (TME), and adjuvant chemotherapy [2-4]. Importantly, the adoption of TNT has been demonstrated to facilitate organ preservation by means of the watch-and-wait (WW) strategy, or use of non-operative management, when a clinical complete response (cCR) is observed after neoadjuvant therapy [5]. As such, this strategy limits the morbidity associated with TME and allows the chance for quality-of-life improvement.

Candidacy for the WW strategy relies on excellent response to neoadjuvant treatment, determined by both endoluminal (i.e., no residual tumor) and extraluminal (i.e., no residual lymph node) assessments [6]. WW implementation relies on strict clinical and radiographic surveillance, as approximately one quarter of patients can be expected to develop local regrowth (LrG), most of which occur within two years [7-9]. Magnetic resonance imaging (MRI) assessment of the pelvis is the standard imaging modality for local staging and surveillance given its excellent soft-tissue resolution. The detection of extraluminal and nodal regrowth requires timely MRI assessment of the mesorectum and of the pelvis, as clinical examination and endoscopy are unlikely to diagnose these classes of regrowth. The use of rectal MRI in conjunction with endoscopy and clinical exam optimizes regrowth detection [10], which can be successfully salvaged with resection in most instances [7-9].

Rectal cancer treatment has seen progressive deintensification over the past century. At the start of the 20^th century, Professor William Ernest Miles recognized that radical surgery decreased local recurrence and therefore pioneered abdominoperineal resection [11]. While oncologic outcomes improved to a degree, abdominoperineal resection was associated with considerable morbidity and mortality. The concept of sphincter preserving surgery via low anterior resection was shown to be oncologically acceptable by Dr. Claude Dixon toward the middle of the 20^th century, sparing many patients from excessive resections [12]. Rectal cancer resection was performed along non-anatomic planes leading up to the 1980s, combined with adjuvant therapy (chemotherapy and/or radiation), which reduced pelvic recurrence rates and enhanced survival [13, 14]. The recognition of the mesorectal plane by Professor William Heald led to the adoption of TME, which continues to be the surgical standard of care [15, 16]. More recently, the use of neoadjuvant therapy supplanted adjuvant therapy, as the former has demonstrated improved pelvic control and sphincter preservation, while reducing morbidity [17]. Advances in treatment paradigms have led to the realization that some patients may achieve pathological complete response (pCR) after neoadjuvant therapy, thereby potentially obviating the need for TME altogether [18]. Candidates for non-operative management of early rectal cancer based on clinical response were first described by Dr. Angelita Habr-Gama, who is regarded as the pioneer of the WW strategy [19]. Despite the initial skepticism, the WW strategy has been adopted globally, presenting many patients the opportunity to maintain their rectum and quality of life.

This review will discuss the evidence to date for the WW strategy, emphasizing the role of restaging rectal MRI in determining patient eligibility for this strategy. Restaging rectal MRI acquisition parameters and treatment response assessment, including important factors to assess, pitfalls, and classification systems, will be discussed in the context of the WW strategy.

Evidence for the Watch-and-Wait Strategy

Several studies have demonstrated the benefits of TNT, which have led TNT to become widely used in patients with locally advanced rectal cancer. Cercek et al. in a seminal study in 2018 demonstrated the benefits of TNT compared to conventional treatment of neoadjuvant chemoradiotherapy and adjuvant chemotherapy, showing that TNT was associated with a greater complete response rate than conventional treatment (36% vs 21%), although differences in long-term survival are still unknown [2-4]. This work as well as others also demonstrated that TNT is associated with improved patient compliance [20, 21]. Additionally, two large randomized clinical trials, RAPIDO and PRODIGE-23, validated the benefit of TNT compared to the prior standard of care in the setting of surgery [2, 3]. PRODIGE-23 demonstrated superior 3-year disease-free survival (75.7% vs. 68.5%) and pCR rates (27.8% vs. 12.1%) in the TNT arm compared to the conventional arm, respectively [3]. RAPIDO similarly demonstrated improved pCR rates in the TNT arm compared to the conventional arm (28.4% vs. 14.3%), with decreased disease-related treatment failure (23.7% vs. 30.4%) [4]. Altogether, given the superior response and tolerability associated with TNT, TNT was introduced into the National Comprehensive Cancer Network (NCCN) guidelines and is used in modern practice for the treatment of locally advanced rectal cancer patients.

Meanwhile, early in the evolution in rectal cancer management, Dr. Habr-Gama in 2004 demonstrated that patients could be candidates for non-operative management as long as there is an absence of residual tumor following neoadjuvant chemoradiation (stage T0). Her seminal study demonstrated that 27% of patients treated with neoadjuvant chemoradiation achieved a cCR [19]. Those managed with the WW strategy had excellent 5-year overall and disease-free survival of 100% and 92%, respectively. Additionally, the LrG rate was exceptionally low at 3%. Although initially met with criticism, Dr. Habr-Gama’s observations have been validated by several other studies, leading to the implementation of the WW strategy worldwide.

The International Watch-and-Wait Database (IWWD) study reviewed the outcomes of 880 rectal cancer patients from 47 institutions who achieved cCR following neoadjuvant therapy and who subsequently underwent WW management [8]. The study demonstrated a 24% LrG rate, with 85% of the LrG events occurring within 24 months of TNT completion. Additionally, the vast majority of regrowths were endoluminal and were successfully salvaged with resection.

Similarly, Smith et al. reviewed the outcomes of 113 cCR patients managed by WW management and reported an absolute LrG rate of 19.5% with a median time to regrowth of 11.2 months, corresponding to an actuarial 5-year LrG rate of 21.4% [9]. Again, the majority of regrowths were endoluminal and amenable to salvage surgery. Importantly, organ preservation was achieved in 82% of patients, with 5-year overall survival and disease-free survival of 73% and 75%, respectively. The observed LrG patterns underscore the importance of MRI surveillance of the rectal wall, when the WW strategy is employed, particularly in the first two years.

The Organ Preservation for Rectal Adenocarcinoma (OPRA) trial is the largest prospective study to investigate organ preservation. 324 patients with stage II/III rectal cancer were randomized to receive either induction chemotherapy followed by chemoradiotherapy (INCTCRT) or chemoradiotherapy followed by consolidation chemotherapy (CRT-CNCT), followed by surgery or WW management based on tumor response. 72% of INCT-CRT patients and 76% of CRT-CNCT patients were eligible for WW management following the completion of TNT. Long-term follow-up demonstrated improved five-year TME-free survival in patients treated with CRT-CNCT compared to INCT-CRT (54% vs. 39%, P = 0.012), but no differences were observed in the primary endpoint of disease-free survival [7]. Additionally, 94% of LrG events occurred within two years, adding to the body of evidence that the vast majority of regrowths occur within this timeframe.

Notably, the optimization of treatment strategies that facilitate organ preservation remain areas of ongoing research, with the Janus Rectal Cancer Trial (NCT05610163), the Japanese Ensemble Trial (NCT05646511), and the German ACO/ARO/AIO-18.1 Rectal Cancer Trial (NCT04246684) being among the current prospective trials being conducted globally. With mounting evidence supporting the oncologic safety and efficacy of the WW strategy, this approach, as with TNT, has been incorporated in the NCCN guidelines [22]. However, the importance of the multidisciplinary team cannot be overstated when implementing the WW strategy, as intensive surveillance is essential to monitor for potential tumor regrowth or progression.

Restaging Rectal MRI Protocol

Given the critical role of MRI in the staging and treatment response assessment of rectal cancer, the Society of Abdominal Radiology (SAR) Colorectal and Anal Cancer Disease-Focused Panel (DFP) have published recommendations for standardizing rectal MRI protocols [23]. The restaging (as well as staging) MRI protocol used at our institution, based on these recommendations, is summarized in Table 1. The key sequences are: true axial T2-weighted imaging of the entire pelvis from the aortic bifurcation/inferior mesenteric artery origin through the sphincter complex to include all pelvic nodal stations; sagittal T2-weighted imaging; coronal oblique T2-weighted imaging; axial small field-of-view oblique T2-weighted imaging through the tumor and sphincter complex; and diffusion-weighted imaging (DWI), both including a true axial plane acquisition with a large field-of-view and b-value of 800 s/mm² and an acquisition centered over the tumor with a small field-of-view and b-value of 1500 s/mm². Notably, intravenous contrast administration is not recommended [24] although it may be valuable in assessing EMVI. Additionally, fat suppression is not included in the standard restaging MRI protocol.

Table 1:

MRI protocol for rectal cancer staging/restaging at our institution.

Sequences	1	2	3	4	5	6	7
Series descriptor	Axial T2	Sagittal T2	Oblique axial T2	Oblique coronal T2 through tumor	Oblique coronal T2 sphincter	Axial DWI b800 MUSE	Focus DWI b1500
Generic sequence name	FSE T2	FSE T2	FSE T2	FSE T2	FSE T2	DWI	Focus Diffusion
Plane	Axial	Sagittal	Oblique axial	Oblique coronal	Oblique coronal	Axial	Oblique axial
Field-of-view (cm)	20-24	18	18	18	18	Fit to anatomy	16
Slice thickness (mm)	5	4	3	3	3	5	3
Gap (mm)	1	1	1	1	1	1	1
b-value						800	1500
Saturation pulse	S/I/A	A	S/I/A	A	A	Special	N/A
TR	4000-6000	4000-6000	4000-6000	4000-6000	4000-6000	8000	6000
TE1 / TE2	110	120	120	120	120	Min	Min
Flip angle	90	90	90	90	90	N/A	N/A
Bandwidth (kHz)	62	62	62	62	32	N/A	N/A
ETL	24	24	24	24	24	N/A	N/A
NEX	1	4	4	4	4	16	16
Frequency steps	384	384	384	384	384	128	140
Phase-encoding steps	256	256	256	256	256	128	70
Frequency direction	A/P	A/P	A/P	A/P	A/P	R/L	R/L
DL factor	medium	medium	medium	medium	medium

Abbreviations: A/P, Anterior/Posterior; DWI, diffusion-weighted imaging; ETL, echo train length; FSE, fast spin echo; MUSE, Multiplexed sensitivity-encoding diffusion-weighted imaging; NEX, number of excitations; R/L, Right/ Left; S/I/A, superior/inferior/anterior; TE, echo time; TR, repetition time.

Consensus recommendations by the SAR and the European Society of Gastrointestinal and Abdominal Radiology (ESGAR) have noted the significance of T2-weighted imaging in the assessment of the treated tumor site [25, 26]. On T2-weighted imaging, it is vital to go beyond assessing reductions in tumor size from baseline and evaluate morphologic changes. However, it is not enough to rely on T2-weighted imaging alone, as many instances of complete response can be overlooked. A recent meta-analysis highlighted that adding DWI to T2-weighted imaging significantly enhances the sensitivity (from 62% to 94%) for identifying complete response, without notably affecting specificity (from 91% to 84%) [27]. During treatment, decreased apparent diffusion coefficient (ADC) values is associated with the progression of apoptosis. Following the end of treatment, decreased diffusion restriction is associated with increased interstitial fibrosis, with low ADC values in the treated tumor site indicating nonviable tumor areas [28]. However, DWI is limited in evaluating treatment response in mucinous tumors adenocarcinomas due to the "T2-shine through" effect [29].

Glucagon is a widely used spasmolytic agent in abdominopelvic MRI protocols. Initial use of this agent dates back 30 years when studies supporting its use in the imaging of gastric and pancreatic carcinomas were put forth [30]. In rectal MRI, evidence suggest that it may reduce peristalsis, improving the diagnostic quality of motion-sensitive sequences such as DWI [26]. However, there is no consensus regarding its use. The SAR Rectal and Anal Cancer DFP recommends the use of a microenema. At our institution, most patients are offered a microenema (hyperosmolar docusate sodium laxative in 5 ml liquid [Enemeez®]) which can be easily self-administered to evacuate rectal contents including air within several minutes and which has been shown to significantly reduce image artifacts on DWI [31]. In cases of anal stricture, rectal abscess/fistula, anorectal pain or patient refusal, a microenema may be omitted.

Assessment of Treatment Response on Restaging Rectal MRI

To determine eligibility for the WW strategy, treatment response assessment on restaging rectal MRI after neoadjuvant therapy is critical, along with clinical and endoscopic assessments. However, rectal MRI assessment can be challenging. Proper treatment response assessment requires the assessment of various factors, such as rectal wall involvement, extramural vascular invasion (EMVI), circumferential resection margin (CRM) involvement, anterior peritoneal reflection (APR) involvement, anal sphincter complex involvement, presence of tumor deposits, and lymph node involvement. The radiologist needs to be aware of the characteristics of the different morphologic changes (fibrosis, desmoplastic reactions, and mucin formation) that can occur in response to treatment as well as various pitfalls that can lead to inaccurate staging.

Rectal Wall Involvement

The extent of rectal wall involvement is crucial for post-treatment MRI-based T-staging (ymrT). ymrT is associated with favorable or unfavorable pathologic tumor response grade (TRG) in locally advanced rectal cancer patients [32]. To date, ymrT-staging as described in the American Joint Committee on Cancer staging manual at post-treatment MRI is no different from that at pre-treatment MRI [33]. However, at post-treatment MRI, the presence of fibrosis, which can be difficult to distinguish from residual tumor, poses a challenge [27]. Khabibi et al. showed that while ymrT had moderate diagnostic accuracy overall, it had significantly diminished accuracy in good responders who had predominant fibrosis than in poor responders with predominant tumor [34]. Regarding the rectal wall, one study showed that T3a tumors (≤ 1 mm beyond the muscularis propria layer of the rectal wall) on post-treatment MRI had a similar prognosis as T2 tumors [35]. Elsewhere, one study showed that ultrasmall superparamagnetic iron oxide (USPIO)-enhanced MRI features can predict downstaging to ypT0–2 tumor confined to rectal wall in locally advanced rectal cancer [36].

The MR tumor regression grade (mrTRG) system was introduced based on the pathologic TRG to classify the extent of response to neoadjuvant therapy in locally advanced rectal cancer patients [37, 38]. There are five mrTRG grades corresponding to the relative amounts of remaining tumor and the extent of morphological changes: Grade 1 – complete response (normalization of rectal wall or thin scar); Grade 2 – near-complete response (thick scar with no definite areas of intermediate signal intensity); Grade 3 – partial response (50% fibrosis and 50% tumor of intermediate signal intensity); Grade 4 – poor response (predominant tumor with minimal fibrosis); and Grade 5 – no response (tumor appearing unchanged or increased). In a retrospective study, mrTRG grades 1 and 2 achieved a sensitivity of 84% and a negative predictive value of 94% for detecting pathologic complete response; however, specificity was only 56% [39]. This study concludes that radiologists should shift their focus from ymrT-staging to detecting gross residual (and progressive) disease and identifying potential candidates for organ preservation who would benefit from further clinical and endoscopic evaluation to guide final treatment planning.

In the first international consensus recommendations for key outcome measures of organ preservation in 2021, [40], cCR is defined as no palpable tumor on digital rectal examination, no residual macroscopic tumor growth on proctoscopy (with scar and/or erythematous ulcer being normal post-treatment findings), and no evidence of viable tumor or adenopathy on MRI. Patients with a cCR would qualify for the WW strategy, whereas patients with incomplete response or near complete response that fails to evolve to a cCR would not qualify. Of note, the MSK Regression Schema, described in 2015, was used in OPRA to assess response post-TNT and classifies patients according to the tripartite features as assessed by clinical examination, endoscopy, and MRI into those with cCR, near complete response, or incomplete response [5, 6].

In 2023, the SAR Rectal and Anal Cancer DFP published an updated lexicon to describe and grade post-treatment morphologic findings [41]. Based on this lexicon, treatment response is classified as complete response (CR), near-complete response (nCR), or incomplete response (iCR). CR is defined as resolved rectal tumor with unremarkable wall morphology, linear or crescent-shaped partial or transmural scar on T2-weighted imaging, the lack of DWI signal, and low ADC value (Figure 1). nCR is characterized by a marked decrease in tumor size with minimal indeterminate residual signal abnormalities. The typical imaging appearance on T2-weighted imaging includes a low signal intensity scar or normalization of rectal wall with a residual focus of intermediate signal. On DWI, diffusion restriction is markedly decreased from baseline, with minimal residual focal areas of high signal on DWI and low ADC value [42]. iCR is defined as obvious residual tumor and/or no regression of lymph nodes albeit there is some reduction in primary tumor size from baseline. On T2-weighted imaging, there is remaining intermediate signal. On DWI, there is insignificant signal regression, with corresponding low ADC value [6] (Figure 2). Hupkens et al. showed that 90% of patients who achieved nCR 8–10 weeks after chemoradiotherapy achieved CR on MRI 6–12 weeks later [42]. These data support the notion that waiting for a longer period could allow more patients to be eligible for WW management (Figure 3).

Figure 1.

57-year-old male with rectal cancer who underwent total neoadjuvant therapy. Baseline rectal MRI (1A–D) shows a tumor with T2 intermediate signal (1A, arrow, axial oblique T2-weighted imaging), high DWI signal (IB, arrow, b-value 1000), and low ADC signal (1C, arrow) in the lower rectum, as well as irregular and T2 heterogenous suspicious mesorectal lymph node measuring 0.5 cm in the short axis (ID, axial T2 T2-weighted imaging). A year after the completion of total neoadjuvant therapy and 1.5 years after baseline MRI, there is no evidence of tumor regrowth (2A–D), with a low T2 signal scar in the tumor bed (2A, arrow, T2 axial oblique) without associated diffusion restriction (2B–C, arrows). The previously suspicious lymph node has decreased in size and T2 signal intensity, measuring 0.2 cm in the short axis (2D, T2 axial).

Figure 2.

61-year-old male with rectal adenocarcinoma who underwent total neoadjuvant therapy. Baseline rectal MRI shows a circumferential mass with intermediate signal on T2-weighted imaging (A) and diffusion restriction on DWI/ADC (B, C). Restaging rectal MR demonstrates a decrease in the size of the mass, with few areas of fibrosis and areas of viable tumor with intermediate signal intensity on T2-weighted imaging that the rectal tumor has (D), and a focal area of diffusion restriction on DWI/ADC (E, F), consistent with incomplete response. Follow-up MRI 2 months later shows a scar on T2-weighted imaging (G) with persistent diffusion restriction on DWI/ADC (H, I), consistent with continued incomplete response. Subsequent pathologic analysis of the resection specimen showed residual adenocarcinoma exhibiting changes consistent with treatment effect.

Figure 3.

50-year-old male with rectal adenocarcinoma who underwent total neoadjuvant therapy. Baseline MRI shows a semicircumferential rectal mass (arrows) with intermediate signal on T2-weighted imaging (A) and diffusion restriction on DWI/ADC (B, C). Restaging rectal MRI depicts a marked decrease in the size of the mass with T2 hypointense fibrosis on T2-weighted imaging (D) and a focal area of diffusion restriction (arrows) on DWI/ADC (E, F), consistent with near complete response. Follow-up MRI 3 months later shows scar on T2-weighted imaging (G) with resolution of the diffusion restriction on DWI (FI) and ADC (I), consistent with complete response.

CRM Involvement

The updated lexicon by the SAR Rectal and Anal Cancer DFP distinguishes between the terms “mesorectal fascia” (MRF) and “circumferential resection margin” (CRM), with MRF denoting the tissue surrounding the rectum anatomically and CRM denoting the tissue surrounding the tumor after surgery.

Regarding MRF, the MERCURY study has defined positive involvement as tumor extent within 1 mm of the fascia on MRI [43]. MRI is limited in assessing the MRF, especially at the anterior wall and in proximity to the anal verge [44]; while MRI has high negative predictive value and sensitivity (100%), specificity and positive predictive value are low to moderate (32–59% and 57–68%, respectively) and inter-reader agreement is merely fair (k = 0.38) [45]. Post-treatment changes, e.g., fibrosis, decrease the accuracy of MRI in determining MRF involvement and result in overstaging [45, 46]. According to ESGAR, a preserved fat plane between the tumor and MRF on MRI can be regarded as uninvolved [26]. Extramural tumor, EMVI, the presence of tumor deposits, and lymph nodes with a disrupted capsule can involve or threaten the MRF, based on their distance from the MRF: “involved” (< 0.1 cm), “threatened” (0.1–0.2 cm), and “clear” (> 0.2 cm) [41, 44, 47, 48].

Regarding CRM, distance from the MRF is the main criterion to determine CRM involvement. MRF thickening and diffusion restriction are additional criteria that can improve the specificity of MRI to determine CRM involvement [49]. The MERCURY study showed that MRI has high specificity (92%) for predicting negative CRM after CRT [50]. Positive CRM involvement on restaging rectal MRI has been associated with lower 5-year survival and higher recurrence rates [35].

EMVI

EMVI is defined as tumor spread to the veins beyond the muscularis propria [51]. On T2-weighted imaging, EMVI typically manifests as a tubular or varicose structure with intraluminal intermediate signal in the course of a perirectal vein and in the vicinity of the rectal tumor [51]. The use of contrast-enhanced Tl-weighted imaging in addition to T2-weighted imaging has been shown to improve the accuracy of MRI in diagnosing EMVI [51]. On contrast-enhanced T1 imaging, EMVI presents as a filling defect or as “tumor in vein” with enhancement [52]. As post-treatment changes (e.g., fibrosis and inflammation) can limit the accuracy of T2-weighted imaging in diagnosing EMVI [53], it is useful to add DWI to yield higher accuracy for this purpose [54]. Kim et al. showed that DWI had high specificity and moderate sensitivity for detecting viable EMVI [55]. Patients with persistent EMVI post-neoadjuvant CRT have an associated significantly higher risk of recurrence [56]. MRI is also helpful in the scenario of EMVI tumor regrowth, which is not visible on endoscopy [53].

APR Involvement

The APR marks the border between the peritonealized and non-peritonealized portions of the rectum and is best identified on sagittal T2-weighted imaging [57, 58]. APR involvement has an impact on surgical decision-making and has been shown to be a negative prognostic factor [58, 59]. On imaging, it classically presents as T2 intermediate nodular thickening of the peritoneal reflection in the vicinity of the anterior rectal attachment [60]. Sim et al. showed a diagnostic accuracy of 74.6% for APR involvement prediction in patients undergoing MRI for preoperative staging or restaging after neoadjuvant concurrent chemoradiotherapy [58].

Anal sphincter complex involvement

For rectal tumors located in the lower-third portion of the rectum, an additional sequence parallel to the anal canal is recommended to evaluate sphincteric involvement [25]. If the levator ani and/or external sphincter is involved and the tumor is located < 1 cm from the anal verge, standard abdominoperineal resection is performed, which includes sphincter complex resection and permanent colostomy [61]. If the tumor is located in proximity to the anorectal ring and the intersphincteric space is not involved, intersphincteric abdominoperineal resection is performed, sparing the external sphincter [62]. More extensive surgical options include extralevator abdominoperineal resection, which includes a wider resection and is performed when the intersphincteric space and external sphintcer and/or levator ani are involved by tumor [61]. MRI has been found to have a negative predictive value of 100% for anal sphincter involvement (100%) post neoadjuvant chemoradiotherapy in two studies, indicating its reliability for this purpose [63, 64].

Involvement of Adjacent Compartments

The most challenging task on restaging rectal MRI is the delineation of residual tumor and fibrosis from adjacent structures to guide surgical resection. Making a checklist of anatomic landmarks to assess (i.e., MRF, sphincter complex, genitourinary organs, adjacent small and large bowel loops, pelvic sidewall, vessels, and nerves) is essential to produce a clinical report that leads to optimal patient care and outcomes.

Presence of Tumor Deposits

Tumor deposits on MRI (mrTDs) typically appear as irregular nodules in the mesorectum; these nodules are not contiguous with the primary tumor and occur along the course of the vasculature draining the tumor [65]. MRI features that separate mrTDs from lymph nodes are the inability to separate the nodules from vasculature on two orthogonal planes, and a classical “comet tail” appearance tapering into the vein [41]. Compared with lymphadenopathy, mrTDs has been identified as an important prognostic factor indicating a worse prognosis [65, 66]. TDs have also been associated with EMVI [66].

Lymph Node Involvement

In the post-treatment setting, size is the main criterion in evaluating lymph nodes. After neoadjuvant therapy, both reactive and malignant lymph nodes decrease or disappear [24]. The accuracy of post-treatment MRI is up to 95% in detecting nodal metastases [67].

Locoregional lymph nodes are described as mesorectal or lateral pelvic based on their location. Regarding mesorectal lymph nodes, a size < 0.5 cm on the short axis indicates a negative status [68]. Regarding lateral pelvic lymph nodes, internal iliac and obturator lymph nodes > 0.4 cm and > 0.6 cm on the short axis, respectively, are suspicious for recurrence [68], implying that radiation to lateral lymph nodes or nodal dissection as the next step [69].

Pelvic non-regional lymph nodes are described as common iliac, external iliac, or inguinal. If the lower rectal tumor extends below the dentate line, inguinal lymph nodes become regional, considering the lymphatic drainage [53]. The only criterion raising suspicion, despite lacking specificity or sensitivity, is a short axis measurement > 1.0 cm. Other clues (e.g., interval change, rectal tumor location, irregular border, atypical shape, and signal) can also be helpful [68].

Fibrosis, Desmoplastic Reaction, and Mucin Formation

Fibrosis appears on T2-weighted imaging as decreased tumor size and signal intensity. Regarding signal intensity, fibrosis in the tumor and the rectal wall appears as T2 hypointense signal, similartothe signal of the muscularis propria, while residual tumor typically displays intermediate signal intensity, similar to its original signal on pre-treatment MRI. Additionally, fibrosis often presents with irregular and somewhat linear edges, while residual tumor tends to have a more nodular and structured appearance [70].

Desmoplastic reaction is characterized by linear, hypointense streaks extending into the mesorectal fat, perpendicular to the treated tumor, on T2-weighted imaging [43]. It is important to distinguish desmoplastic reaction from remaining tumor which tends to have a more nodular appearance and a more intermediate signal intensity, in order to avoid incorrectly classify the tumor as more advanced than it is.

Mucin formation can be a response to treatment or a characteristic of the tumor itself. Nonmucinous tumors can produce mucin in response to treatment; this is seen as a further increase in signal intensity in tumor regions that originally exhibited intermediate signal intensity on T2-weighted imaging [43]. Within these areas, spots of mixed intermediate signal intensity possibly indicate a mixture of malignant cells, cords, or blood vessels [43]. Meanwhile, mucinous tumors can transition to a state of acellular mucin in response to treatment; this is seen as decreased tumor size and increased hyperintense homogeneous signal accompanied by decreased heterogeneous intermediate signal from cellular components on T2-weighted imaging [71]. Notably, the detection of acellular mucin is not reliable without pathologic confirmation. When mucinous tumors do not respond to treatment, high signal with intermediate-signal-intensity components persists from baseline through post-treatment T2-weighted imaging; these tumors are associated with a worse prognosis and a higher risk of local recurrence [72]. However, Judge et al. evaluated the oncologic outcomes of 19 patients managed by the WW strategy with mucin present on posttreatment MRI and reported a 21% LrG rate and 11% distant failure, suggesting that the presence of post-treatment mucin should not preclude WW eligibility [73].

Pitfalls

A frequent T2 sequence pitfall is a pseudotumor caused by rectal wall thickening adjacent to the tumor bed, containing a high T2 signal, preserved wall stratification, and lacking diffusion restriction. A review of baseline tumor location prior to analyzing the post-treatment MRI is helpful to overcome this pitfall [68].

When assessing post-treatment DWI, the radiologist needs to be familiar with two phenomena: “T2 shine-through” and “T2 dark-through.” T2 shine-through refers to high DWI signal corresponding to intermediate or high ADC signal. Long T2 decay of the post-treatment tumor bed, in the setting of the treated tumor, inflammation, or cellular and acellular mucin can explain the finding. Intraluminal fluid usually has a tridentate configuration known as the “Mercedes-Benz” sign [53, 70]. T2 dark-through represents low DWI and ADC signal due to post-treatment fibrosis containing an abundance of collagen with intrinsic low T2 signal [53].

Use of MRI in Alongside Clinical and Endoscopic Findings in Assessing Response

It is important to consider imaging results alongside clinical and endoscopic findings. When there is no residual malignancy detected across all post-treatment evaluation methods, including digital rectal examination, endoscopy, and MRI, it is possible to accurately predict pathologic complete response in as many as 98% of cases, with a sensitivity of 71% and specificity of 97% [10].

Endoscopy remains the essential modality to assess the degree of mucosal response following neoadjuvant therapy and aid in determining the subsequent treatment strategy. Although rigid proctoscopy was described in the early literature [74], current practice favors the use of flexible proctosigmoidoscopy [75].

In the OPRA trial, a three-tiered regression schema was developed to establish uniform criteria for tumor response across multiple institutions and continues to be a reference for modern watch-and-wait practice [5]. Endoscopic features of cCR include the presence of a flat white scar without mucosal nodularity or ulceration. Any minor mucosal irregularity or nodularity, superficial ulceration, or scar erythema are considered nCR, which may evolve with later reassessment. Visualization of any obvious, viable tumor is consistent with an iCR, and many studies consider the presence of large (> 3 cm), deep or necrotic ulceration as indicative of persistent disease [76] (Figure 4). The international expert consensus statement published 2021 agreed upon similar definitions to describe luminal response but considered the presence of small erythematous ulcers as acceptable for endoscopic cCR [40]. Although widely considered to be a hallmark of cCR, the presence of a flat white scar has been shown to only carry a positive predictive value of 70–80% [77]. Endoscopic biopsy is not a requisite to define cCR or nCR, and the international consensus does not advise the routine use of endoscopic biopsy given the lack of additional value and high potential for false-negative results [40, 78]. Endoscopic findings should be taken into accordance with T2-weighted imaging and DWI, as the combination of modalities have been demonstrated to predict complete tumor response with a post-test probability of 98% [10].

Figure 4.

Endoscopic appearance of complete/near complete and incomplete response. Endoscopic images prior to neoadjuvant treatment (left) and upon initial response assessment after the completion of neoadjuvant treatment (right). Yellow arrows highlight the significant findings at the site of the primary tumor.

WW Protocol for Eligible Patients

Given the time-dependent increase in tumor response with longer intervals to follow-up [79-81], the initial endoscopic response assessment typically occurs 6–8 weeks following the conclusion of neoadjuvant treatment. In the OPRA trial, the investigators established an initial assessment timeline of 8–12 weeks post-TNT [5, 6], and this is followed in ongoing prospective trials including the Janus Rectal Cancer trial (NCT05610163), the Japanese Ensemble Trial (NCT05646511), and the German ACO/ARO/AIO-18.1 Rectal Cancer Trial (NCT04246684). For patients with nCR on the initial assessment who wish to be considered for organ preservation, close endoscopic surveillance should be implemented with reassessments every 4–8 weeks until tumor response ceases or cCR is achieved. However, it is critical to note that a tumor that stops responding should not be watched, and the maximum recommended time for surveillance of an nCR is 6 months after initial assessment (Figure 5). LrG events can be expected in 25–30% of patients, with the majority occurring with 24 months of establishing cCR [8]. Approximately 90% of regrowth events have an endoluminal component which may appear as an interval mucosal irregularity or ulceration at the site of previous flat white scar visualization, underscoring the importance of mucosal surveillance [82].

Figure 5.

Role of endoscopy in watch-and-wait management. Top panel demonstrates a case of near complete response evolving to clinical complete response. Bottom panel demonstrates a case of clinical complete response which later demonstrated local regrowth at the site of the tumor scar.

Regarding the surveillance protocol, patients managed with the WW strategy in the OPRA trial followed a standardized protocol of flexible sigmoidoscopy and digital rectal examination every 4 months for the first two years, and then every 6 months for the following three years. Pelvic MRI was obtained every 6 months for the first 2 years, and then every year for years 3–5 [5]. The current NCCN guidelines (version 1.2024) recommend a similar surveillance approach, with proctoscopy and physical exam every 3–6 months for the initial 2 years after the completion of neoadjuvant treatment, followed by every 6 months for the subsequent 3 years [22]. However, pelvic MRI is only obtained every 6 months for up to three years. It is strongly recommended that rectal cancer patients who are eligible for the WW strategy be managed on a clinical trial protocol if at all possible, to maximize the multidisciplinary resources needed for both short- and long-term surveillance.

For select patients with robust clinical responses to neoadjuvant therapy but exhibiting small (≤ ycT1) residual peri-mucosal lesions in the absence of extramural or transmural disease, minimally invasive interventions such as surgical local excision or endoscopic submucosal dissection may be considered by experienced providers. Despite the lack of prospective data, limited single institutional experiences suggest that local excision may have a role in safely extending organ preservation for properly selected cases [83, 84]. Although these adjuncts have the potential to extend WW candidacy, careful patient selection and continuation of rigorous surveillance are essential to ensuring safe oncologic outcomes.

WW is considered on the basis of achieving a cCR to TNT, which is the treatment indicated for locally advanced rectal cancer, defined as stage II (T3-4, NO) and stage III (T any, N1 or greater) disease. There is evidence to suggest that increasing cT stage is associated with risk for local regrowth, as a systematic review demonstrated 2-year cumulative incidences of local regrowth of 19% for stage cT1 and cT2 tumors, 31% for cT3 tumors, and 37% for cT4 tumors (p = 0.0330) [85]. No associations between cN stage and incidence of local regrowth were noted. However, analysis from the OPRA trial suggests that nodal involvement (Hazard Ratio [HR] 1.89; 1.23–2.90) was negatively associated with organ preservation [86], and that baseline cT3 (HR 2.05; 1.06–3.97) and cT4 (HR 1.84; 1.22–2.78) disease were associated with time to TME, relative to cT1 or cT2 disease [5]. These findings suggest that baseline disease status should be accounted for by providers when discussing the WW approach with patients. It is our strong preference to manage rectal cancer patients who desire a WW approach on trial, if possible.

Conclusion

The WW strategy in rectal cancer is a feasible and oncologically safe approach that can potentially spare patients from surgery and its associated morbidities. This strategy mandates intensive surveillance, with frequent rectal MRI assessments serving a critical role alongside clinical exam and endoscopy. Ultimately, successful WW implementation cannot be achieved without a multidisciplinary team comprising, but not limited to, colorectal surgeons, medical and radiation oncologists, radiologists, and pathologists.

Highlights.

• The watch-and-wait (WW) strategy, a non-operative strategy in selected locally advanced rectal cancer patients who achieve clinical complete response to neoadjuvant therapy, avoids the potential morbidity associated with total mesorectal excision and allows the chance for quality-of-life improvement.

• Data has emerged to support total neoadjuvant therapy as a viable approach to achieve durable clinical complete response, determined based on clinical examination, endoscopy, and magnetic resonance imaging (MRI) findings.

• MRI features of clinical complete response include the presence of a low T2-weighted scar without restricted diffusion and resolution of nodal disease following neoadjuvant therapy.