Pelvic MRI for Guiding Treatment Decisions in Rectal Cancer

Table 1: Histopathologic Grading of the Quality and Completeness of the Mesorectum in a Total Mesorectal Excision Specimen

Table 2: MRI Grading of Extramural Venous Invasion

Table 3: Guidelines Regarding Imaging With MRI and Treatment for Rectal Cancer

Table 4: Proposed Mid Rectal Cancer Risk Categorization Based on MRI and Clinical Risk Factors

Table 5: Low Rectal Cancer (0–5 cm From Anal Verge) Possibly Requiring Abdominoperineal Resection

Fluoropyrimidine-based chemoradiation (CRT) is used routinely for locally advanced rectal cancer to shrink the tumor preoperatively, improve lateral surgical clearance at total mesorectal excision, prevent local recurrence, and preserve organ function. In Northern Europe, short-course preoperative radiotherapy (SCPRT) is preferred to achieve locoregional control. However, with recent improvements in the quality of surgery, in magnetic resonance imaging (MRI), and in pathologic reporting, we question whether “routine” CRT or SCPRT should be offered indiscriminately for all patients.

MRI is considered the optimal modality for locoregional staging and evaluation of the potential for an involved circumferential resection margin. MRI also provides detailed anatomic information for surgical planning, and may identify poor prognostic features, which influence the way in which the pathologist processes specimens. MRI can predict the likelihood of good/poor tumor response to neoadjuvant CRT and can categorize responders/nonresponders following treatment.

Using MRI to define the risk of both local recurrence and metastatic spread allows clinicians to determine which patients might benefit from or safely avoid neoadjuvant treatment. We have arrived at these views after comparing data from published observational studies, results from randomized trials, and outcome analyses of the Norwegian National Cancer Registry.

Introduction

Background on locally advanced rectal cancer

Rectal cancer is a common disease with a high rate of mortality, but outcomes have improved in the past 2 decades, as a result of improvements in surgery, imaging, pathologic evaluation, and oncologic treatment. The long-term success of surgery depends in part on accurate preoperative staging and on choosing the most appropriate surgical technique. Early-stage (T1) tumors may be amenable to local excision or transanal endoscopic microsurgery (TEM), whereas locally advanced rectal cancers (LARC) extending close to the mesorectal fascia (MRF) may require neoadjuvant chemoradiation therapy (CRT) followed by total mesorectal excision (TME) or abdominoperineal excision of the rectum (APER).[1]

Although there is no standard definition of LARC, most phase III trials have addressed clinically or pathologically staged (T3/4 and/or N1/2) tumors according to TNM staging in the American Joint Committee on Cancer Staging Manual (7th edition). In the 1970s, nonrandomized observational studies highlighted the poor prognosis of rectal cancer in terms of local control and overall survival, and defined clinical and histopathologic factors that could predict local and systemic recurrence. In the 1990s, randomized trials[2] established the role of short-course preoperative radiotherapy (SCPRT) in reducing the risk of local recurrence (LR). During the same period, postoperative fluorouracil (5-FU)–based CRT for patients with stage II or III rectal cancer was extrapolated to the preoperative setting. Randomized trials of preoperative 5-FU–based CRT also demonstrated improved locoregional control,[3,4] and in more advanced unresectable/borderline cases, CRT improved resectability and disease-free survival.[5]

Many oncologists believe all patients with rectal cancer should be treated with similar combinations of chemotherapy, radiotherapy, and surgery in a one-size-fits-all approach. Hence, preoperative CRT and surgery form the standard of care in the United States and parts of Europe for patients with clinical stage II and III rectal cancer. However, improvements in surgical technique have resulted in a reduction in the LR rate from 30% in the 1980s to between 5% and 10% at present. Even lower rates are achieved currently at some specialist centers.[6,7] Many still argue that good local control is a legitimate outcome of different interventions in rectal cancer, even if overall survival is not improved. However, this blanket strategy of preoperative CRT for all has to be weighed against the costs in terms of late effects of radiation, increased surgical morbidity, and financial cost to the healthcare system. New data are emerging[8] demonstrating that it is possible to substitute radiotherapy for chemotherapy in selected cases.

Can we improve patient selection for neoadjuvant therapy? We should also ask: In which patients can we safely omit preoperative CRT altogether? And in which patients can we replace CRT with chemotherapy alone?

Many factors modify outcomes. The outcome for a group of cT3 cancers will depend on: 1) their site in the upper, middle, or low rectum, and whether they are located anteriorly or posteriorly; 2) the percentage of the circumference involved; 3) tumor morphology, including whether it is exophytic (polypoid or sessile) or ulcerated, mobile or fixed; 4) tumor extension above or below the peritoneal reflection, or above or below the levators; 5) patient demographics, including sex and age (< 40 years or older).

Following neoadjuvant CRT, important prognostic factors include the pathologically assessed circumferential resection margin (CRM), quality of surgery (ie, the plane of surgery), and ypT and ypN classifications. However, these data are only obtained postoperatively.

MRI with a phased-array surface coil provides excellent soft-tissue contrast at high spatial resolution. The ability to visualize the tumor and surrounding structures in coronal, axial, and sagittal planes means that MRI is currently the optimal method for staging rectal cancer and evaluating the potential CRM. A caveat is in order regarding the use of this modality to evaluate levator involvement in low rectal cancer and nodal involvement, which remain difficult even with newer MRI sequences, such as diffusion-weighted imaging. The challenge for an imaging study is not just to be able to use it to determine resectability, but also to be able to use it to predict the likelihood of LR and distant metastases-and thereby define the optimal neoadjuvant strategy, and determine whether the tumor will respond to standard fluoropyrimidine-based CRT.

MRI facilitates identification of high-risk groups that may benefit from preoperative treatment. Optimized MRI offers excellent soft-tissue contrast and high spatial resolution within a tolerance of 0.5 mm.[9] In current practice, MRI relies on morphologic changes, but functional MRI techniques are emerging as useful imaging tools in oncology. These tools include perfusion MRI, diffusion MRI, and MR spectroscopy; they show tissue properties such as angiogenesis, cellular density, and proliferation. In the future, these tools may improve prognostication and prediction/assessment of response.

Following neoadjuvant treatment (and prior to surgery), it is now possible with MRI to define different types of response (fibrous, desmoplastic, and colloid), predict histopathologic TN downstaging, and define a radiologic tumor regression grade (TRG). Where downstaging/downsizing is confirmed, modification of the initial treatment plans can be made. In contrast, a lack of response could prompt intensification of treatment at this point, prior to surgery, rather than in the postoperative setting. MRI can also be useful in monitoring the pelvis when there is clinical complete response following CRT, in a “watch-and-wait” approach, and in detecting an early salvageable recurrence.

In the following sections, we discuss the features that predict LR and distant metastasis, and we review the features that can be imaged on MRI to allow decision making regarding the best neoadjuvant treatment in LARC.

Features That Predict Local Recurrence and Metastatic Disease

Features that predict local recurrence

Rectal cancer is clearly not a single disease entity but has several different modes of spread. Randomized and nonrandomized studies have provided a number of surgical histopathologic prognostic indicators, which appear to differ in importance depending on whether or not preoperative radiotherapy or CRT has been administered.

These histopathologic features include pathologic TNM stage,[10] T substage,[11] CRM status,[12,13] adequacy of lymph node assessment,[14] number of involved lymph nodes,[15] extracapsular extension,[16] the presence of extranodal deposits, peritoneal involvement, tumor perforation, quality of mesorectal excision (whether within the muscularis plane) (Table 1),[17] tumor differentiation, lymphovascular invasion, extramural vascular invasion (EMVI),[18-20] and perineural invasion.[21] The risk of lymph node positivity significantly increases with increasing depth of invasion of tumor through the bowel wall (ie, with increasing T substage).[22]

The risks of local failure are much higher for cancers below the peritoneal reflection, and for those in the lower rectum below or involving the levators. Low rectal cancers that are anteriorly located and fixed are more difficult to resect and appear to have a higher risk of LR.[23,17] Different individuals also may have variable sensitivities to chemotherapy and radiotherapy. An early prospective multicenter observational study (Study Group Colo-Rectal Carcinoma [SGCRC]) examined 1,121 patients and showed tumor spillage; the treating institution and the individual surgeon were also independent factors with an impact on LR.[24]

An exploratory subset analysis of T3 tumors demonstrated LR rates for SCPRT vs selective postoperative CRT of 2.7% vs 5.8% in T3a (≤ 1 mm) tumors; 2.9% vs 9.6% in T3b (> 1–5 mm) tumors; and 9.6% vs 21.9% in T3c (> 5–15 mm) tumors.[25] Other groups have also shown that T3c and T3d rectal cancers have markedly worse progression-free and cancer-specific survival compared with T3a and T3b cancers.[26]

The CRM is pivotal to the management decision making process, because the histopathologic finding of a close or involved CRM is a dominant influence on LR. In a landmark study, rigorous histopathologic analysis revealed 27% of all rectal cancers demonstrated an occult positive CRM (< 1 mm) after potentially curative surgery.[12]

Involved regional lymph nodes at histopathology after radical surgery for T3/4 cancers confer a high risk of locoregional recurrence[15] of up to 15% for pN1 tumors, and of up to 32% for pN2 tumors. In the Trans-Tasman Radiation Oncology Group (TROG) trial,[27] positive resection margins, histopathogically involved lymph nodes following radiotherapy in the resected specimen, and baseline carcinoembryonic antigen level were independently associated with LR. In the Cancer Research UK 07 (CR07) trial, significant risk factors included anterior quadrant tumor involvement, extramural vascular invasion, and N stage (N2 > N1).[25] All the above need to be taken into consideration before deciding to recommend postoperative CRT, and the current variability in outcomes means that MRI results should always be discussed with the patient with a view to informed consent and shared decision making.

Features that predict metastatic disease

Many of the features that predict a high risk of LR also predict a risk of metastases-particularly an involved CRM.[28] EMVI also predicts a high risk of metastatic disease-tripling the risk of subsequent metastases within 12 months.[29] In a recent small retrospective series from the Medical University of Graz (Austria), actuarial 5-year progression-free survival rates of 22% and 69% were observed for those with and without vascular invasion, respectively.[30]

The Features We Can Image

TNM stage

TNM staging is based on the size and extent of the primary tumor (T), the absence or extent of involvement of the relevant regional lymph nodes (N), and the presence or absence of distant metastases (M). The system is practical, simple, and versatile in that it allows for both clinical (cTNM) and pathologic assessment (pTNM), and even has a category for patients who have received prior treatment, to designate potential modification (ypTNM). The T-stage, N-stage, and M-stage can be combined into stage groupings (stage I to IV), which give a better-defined risk stratification and prognosis. Data from trials that provide data on outcome continue to be incorporated on a regular basis in an ever-developing, revised, updated, and refined algorithm. Hence the prognostic accuracy of the TNM system continues to improve.

Nevertheless, there are limitations to current T staging. Approximately 75% of patients with rectal cancer are initially clinically staged as cT3; this large group of patients encompasses wide variability and much heterogeneity, and a “cT3” classification thus fails to accurately predict outcomes for the individual. Adverse features found on rectal MRI can identify patients at increased risk for synchronous and subsequent metastatic disease.[31] The Mercury Study Group used preoperative MRI to extend the clinical subclassification of T3 into four groups: “a” (< 1 mm outside the wall), “b” (1–5 mm), “c” (5–15 mm), and “d” (> 15 mm).

Lymph node involvement (LNI) is conventionally the most important factor. The majority of mesorectal nodes are located in the upper-posterior part of the mesorectum, and a high proportion of these nodes are found in the outer layer of the mesorectum. Careful resection is required, as it may be easy for the surgeon to come out of the correct plane posteriorly and laterally, particularly before he or she reaches the level of the tumor.

The lymph node status may be refined in more detail and with greater prognostic accuracy by characterizing the total number of involved lymph nodes, the ratio of involved lymph nodes vs the total number of nodes analyzed, the absolute number of identified lymph nodes, the presence of extracapsular spread,[32] or lymph node micrometastases in the surgical specimen. Data from the University of Erlangen (Germany) showed a higher risk of LR after primary surgery for patients with at least four involved regional lymph nodes.[33] Extracapsular LNI was found in 54 patients (23%), while in 41 patients with a positive lymph node status (18%), no extracapsular LNI was detected.[34]

Size criteria, with additional impressions regarding shape, border, and pattern, are usually used in standard MRI to distinguish between benign and malignant nodes. Involved nodes outside the MRF (not usually resected in the West) appear uncommon in the mid and upper rectum, and are found mainly in patients with distal rectal cancers close to the anal verge with visible nodal metastases within the mesorectum.[35]

A meta-analysis of MRI staging of rectal cancer revealed the sensitivity and specificity of N staging to be 77% and 71%, respectively. Hence there are limitations to assessing mesorectal lymph nodes with MRI. At present there are no validated management strategies for selecting treatment based on findings of macroscopic nodal involvement, microscopic involvement, or the presence of a few isolated cells; thus, it is questionable whether we should focus overmuch on this.

Circumferential resection margin

MRI can accurately predict CRM status and direct risk-stratified management strategies so as to select patients appropriate for preoperative CRT. MRI prediction of CRM as compared with histopathology suggested the use of a wider threshold with MRI (2 mm) than with pathology[36]; however, the MERCURY group based their analysis on an MRI cutoff of < 1 mm,[9] and a prospective study also showed that a 1-mm cutoff on MRI predicts clear margins in 96.7% of cases.[37] A multicenter observational prospective study aimed to determine the minimal distance of tumor from the MRF using a cutoff of ≤ 1 mm for predicting involvement of the CRM. An 82% agreement (266/325 cases; 95% confidence interval [CI], 77%–85%) between MRI and histopathologic assessment of CRM status was observed. (CRM involvement defined as tumor with ≤ 1 mm extension into the MRF). In the CR07 trial, the rate of finding a positive CRM fell from 21% in 1998 to a low of 10% in 2005,[17] mirrored by an improvement in the quality of the mesorectal plane of excision (predominantly in patients undergoing anterior resection rather than APER). The preoperative results of MRI were compared with the histopathologic findings in the resected specimens, and showed an accuracy for preoperative MRI of 90.9% (139/153).[38]

Extramural vascular invasion

EMVI is defined as the invasion of carcinoma cells into surrounding vascular structures, which has been suggested as a stage-independent adverse prognostic factor.[18] EMVI is observed in about 40% of patients. Both LR and liver metastases develop more frequently when EMVI is present, which significantly worsens 5-year disease-free and overall survival. There is good concordance between macroscopic EMVI predicted on MRI (Table 2), and the pathology specimen.[39] In an analysis of 48,894 patients with colorectal cancer from a prospective German multicenter observational study, venous invasion appeared highly important for the development of distant metastases.[40]

Differentiation

Histologic tumor grade is prognostic, with poorly differentiated tumors having a worse prognosis; however, the majority of tumors are moderately differentiated. There is no clear evidence currently that MRI can discriminate between tumor grades; there are no specific texture features on standard images, and no clear relationship with diffusion-weighted MRI in the small studies published to date.

Mucinous tumors

Mucinous adenocarcinoma of the rectum is uncommon, accounting for 7% to 20% of all cases of rectal cancer treated with surgery alone[41-43]; however, it is a more aggressive histopathologic subtype, conventionally defined as tumors that express extracellular mucin in more than 50% of the stroma.[44] The lack of precision in this definition may account for some disparity in outcomes.

Both radiotherapy and CRT drive mucinous differentiation in rectal cancers. The evidence comes from the Dutch TME study: tumors that were irradiated were more frequently mucinous (13%, compared with 7% of tumors that were not irradiated). Also, in the German pre- vs postoperative chemoradiation study (Sauer et al, 2004),[3] there was a higher proportion of mucinous tumors in patients treated with CRT preoperatively than in those treated postoperatively. The Dutch TME study results[45] suggest that there are two distinct classes of mucinous carcinoma: those that are pre-existing (ie, uninduced) and mucinous carcinomas that are induced by preoperative radiotherapy. The paper suggested that the prognosis of patients with induced mucinous carcinoma is significantly better than that of patients with pre-existing mucinous carcinoma (91% vs 39% recurrence-free survival at 2 years).

It has been hypothesized that mucinous tumors respond poorly to CRT.[46] A small study of 70 patients with rectal cancer who received preoperative CRT followed by surgery and who were evaluated over a 5-year period found this to be the case. If there really is less downstaging in mucinous carcinomas, one would expect to see a difference between the mucinous and nonmucinous groups in the proportion of patients in whom there was a positive CRM.

Guidelines

There are now a number of guidelines available to oncologists, including those of the National Comprehensive Cancer Network (NCCN),[47] the American College of Radiology (ACR),[48,49] the European Society for Medical Oncology (ESMO),[50] and the UK National Institute for Health and Care Excellence (NICE) guidelines (Table 3).[51] Current guidelines are based on historical data, prior to the use of MRI, TME, and modern pathology reporting. The guidelines are only risk-adaptive with regard to LR.[50] They are outdated because they do not account for individual patient risk factors and preferences. There is no internationally agreed-upon definition of LARC, nor any consensus regarding which patients require preoperative radiotherapy and which do not require it.

NCCN guidelines recommend neoadjuvant long-course CRT for all clinically staged T3/T4 tumors irrespective of CRM involvement. However, in the United Kingdom, CRT is often reserved for only those T3 tumors with a threatened CRM. Preoperative radiotherapy improves local disease control in some, and in a small proportion may also increase overall survival,[52] but the benefits are not shared by all patients. There is therefore the need for an update of the imaging information and some harmonization of existing guidelines.

Influencing Clinical Practice

Utility for the multidisciplinary team

Exchange of information between the different disciplines responsible for decisions within the multidisciplinary team (MDT) allows audit recommendations based on MRI, and integrates surgical findings in a feedback loop that ultimately leads to a better understanding of the pathology. It is not sufficient to use the images to characterize and describe the anatomy and the tumor’s relationship to the circumferential margin. It is also crucial to know the prognostic value, in terms of the risk of LR, the risk of metastatic disease, and the prospect of long-term survival, because all these predictions have an impact on informed decision making involving both the surgeon and the patient.

How these features can help the surgeon choose the appropriate operation

Where MRI predicts a surgically clear CRM in a high or mid rectal cancer, there is greater than 90% confidence that a complete resection can be performed via standard TME.[9,37] For low rectal cancer, MRI can also distinguish between patients who should undergo low anterior resection and those who should undergo standard APER or extralevator abdominoperineal excision, based on the relation of the tumor to the levator ani and anal sphincter muscles.[53] Tumors located above and below the puborectalis sling and anteriorly at the level of the prostate may be better treated by exenteration.[54]

How these features can help the pathologist be more accurate

Preoperative MRI may increase identification of EMVI and of the potential involvement of the MRF/CRM and nodes, which may influence the way in which the pathologist processes specimens and may shed light on where to sample. The detection of vascular invasion is related to the number of examined tissue blocks. MRI may increase the proportion of patients in whom these features are identified and reduce the percentage of patients classed as having achieved a pathologic complete response.[55] However, the question remains whether this is merely a “Will Rogers” phenomenon in which a pathologic “upstage migration” gives rise to improved outcomes in both groups but does not necessarily reflect the efficacy of the type of treatment with regard to long-term survival of the patient.

Predicting the likelihood of good or poor response to CRT

Pretreatment MRI findings of greater depth of tumor penetration into the mesorectum, lymph node involvement, and EMVI all appear associated with poor response to CRT.[56] Others have suggested that mucinous tumors, especially those with signet ring pathology, respond poorly.[46]

How can we evaluate good or poor response to CRT?

Quantitative (change in tumor volume) and qualitative (grade of tumor response) MR assessment can distinguish good responders from poor responders following neoadjuvant CRT. The most important prognostic factors in rectal cancer include pathologically assessed CRM, the quality of surgery (ie, the plane of surgery), and ypT and ypN classifications. Some also argue that the histologic regression grade within primary tumor and regional lymph nodes influence outcome, in particular the LR rate.[57] However, these data are only obtained postoperatively and rely on an estimate of initial appearances. MRI-based TRG assessment that compares the pre- and post-CRT scans may allow the MDT to make decisions preoperatively.

Good radioresponse

Tumor downstaging is usually overestimated by conventional MRI evaluation, but cancers will often shrink as a result of CRT, and occasionally they may completely disappear. If the surgeon is amenable, a good clinical response offers the opportunity to change the original plan for surgery, and to perform a less mutilating resection if an adjacent organ was originally involved-and a good clinical response could facilitate sphincter-sparing surgery in distal tumors, or allow either a local excision or even a nonoperative approach in a patient who achieves a complete clinical response.

Poor radioresponse

Radioresponse to CRT varies significantly. Downstaging/downsizing is achieved in approximately 50% of patients with LARC, with up to 20% achieving pathologic complete response. The other 50% of patients have tumors that are relatively resistant to CRT and that change little. Larger T3/T4 tumors tend to respond by fragmentation and can be difficult to assess after CRT, whereas smaller T2/T3a tumors tend to shrink back to the endoluminal point of origin. It is therefore important to discriminate clearly between these populations if we are ever to individualize therapy. Individual lymph node size is a poor predictor of nodal metastases following CRT,[58] although there is a high negative predictive value if previously imaged involved nodes return to normal.

Patients with little or no response may benefit from intensified strategies at this point. Hence, accurate assessment of tumor response following neoadjuvant CRT is crucial to the determination of optimal management. If it is possible to stratify patients to these low- and high-risk groups after neoadjuvant treatment, but before surgical resection and pathologic evaluation, intensified strategies may be planned.

Limitations of MRI

A current limitation of MRI for pretreatment evaluation is the significant learning curve associated with image evaluation; there is also the uncertainty regarding nodal involvement. The widespread use of preoperative CRT (leading to frequent nodal downstaging) offers limited opportunities to the radiologist in the MDT to audit performance in pretreatment MRI assessment. As we look to the future, functional MRI is now being transitioned into everyday practice, and this will increasingly allow us to assess perfusion, diffusion, cellular density, and membrane proliferation, in addition to standard morphology. In the future, these quantitative techniques may assist with the differentiation of tumor stage, help to discriminate nodal stage for clinical decision making, and improve imaging phenotyping in terms of prognosis and prediction of treatment response. A recent study showed that there is good interobserver correlation of apparent diffusion coefficient and perfusion values in lymph nodes and primary tumors of patients with rectal cancer, and these functional parameters, in particular perfusion fraction, may be useful in determining tumor stage.[59] However, in the patients who proceeded to surgery without neoadjuvant CRT, the authors could find no statistically significant differences in the assessed functional parameters as these related to pathologic stage.

Follow-up after surgery

LR historically reflected inadequate mesorectal resection.[60] In a Danish study, residual mesorectal tissue was commonly inadvertently left in the patient and can be observed on postoperative MRI (54 of 136 patients [40%]), particularly after partial mesorectal excision,[61] which may help to determine which patients are at greater risk for LR, and who might require more frequent imaging or more intensive postoperative treatment.

Cost-Effectiveness

A selective approach to SCPRT or CRT using MRI and other factors, rather than blanket use of SCPRT or CRT (as advocated by some radiation oncologists), is likely to be more cost-effective. However, there have been few economic assessments of radiotherapy,[62-64] and we were unable to find any modern evaluation integrating good-quality TME, MRI, and 5-FU–based CRT-particularly CRT that included intensity-modulated radiotherapy. Historical studies such as the Swedish Rectal Cancer Trial[2] did not use good-quality radiotherapy or TME; hence the LR rate was high, and the reduction in LR led to an improvement in survival that has not been observed in more modern trials.

From the healthcare perspective, costs include imaging, primary treatment, supportive treatment, and continuing care (including support for stoma care, surveillance, and treatment of any recurrence). From the societal perspective, additional costs include productivity losses (for paid work), informal care costs, and travel costs. The cost of SCPRT (ie, 5 fractions of 5 Gy) was estimated at $2,600 in 2004.[64] Long-course CRT would be substantially more expensive. The cost-effectiveness in the Swedish Rectal Cancer Trial[62] was estimated at $3,700 per quality-adjusted life-year (QALY) saved. Their cost-effectiveness ratio of $25,100 per QALY is less favorable. Their model, based on the Dutch TME trial data,[64] estimated that the loss of quality of life due to preoperative radiotherapy is outweighed by the gain in life expectancy; however, the assumption that that the lower LR rate leads to a survival advantage is flawed, because long-term data do not show any survival gain.[65]

The cost per QALY is likely to be substantially higher than the $25,000 proposed by Van den Brink and would therefore be unlikely to be endorsed if a formal cost-effectiveness evaluation were undertaken today. The Dutch model also did not take into account the role of CRT in sphincter-sparing surgery (because SCPRT does not achieve downstaging), nor the impact of postoperative adjuvant chemotherapy on local control,[66] since in the Dutch TME study, only 4.9% of patients randomized to surgery alone received postoperative chemotherapy.[64] On the other hand, the cost of omitting CRT may involve more frequent follow-up, endoscopic surveillance, and imaging.

Conclusion

MRI captures the whole tumor, its characteristics, and its relationship to pelvic structures, unlike preoperative biopsy, which only samples a small proportion of the tumor. The anatomical accuracy of MRI for identifying the precise depth of penetration into and beyond the muscularis propria, and for predicting the histopathologic characteristics preoperatively, offers the prospect of tailoring patient treatment preoperatively. Using a cut-off for selection of CRT of 2 mm from the MRF, LR is no longer the overriding problem, and the risk of metastases predominates.[67] MRI is also useful in this setting because it can help to distinguish between the risk of local vs distant recurrence, and can direct more intensive neoadjuvant chemotherapy.

MRI can predict the likelihood of Reponse Evaluation Criteria in Solid Tumors (RECIST) response and MRI TRG regression. Restaging MRI 6 to 8 weeks after completion of neoadjuvant CRT can also accurately determine tumor response-because scans can be compared side by side with the original MRI (unlike pathologic tumor regression grading). Hence, initial assessment with MRI and subsequent reassessment provide individualization and decision making flexibility within the MDT regarding the options for neoadjuvant treatment (Tables 4 and 5).

Diffusion-weighted imaging and dynamic contrast-enhanced MRI may offer additional information for predicting/detecting response to CRT, but multicenter prospective studies are needed to corroborate present research data.

The use of CRT needs to be balanced against the acute tolerability and late functional consequences, which should be selected for each patient according to his or her individual risk of both local and distant recurrence. The argument that even if overall survival is not improved, reducing LR as far as possible with preoperative CRT/SCPRT is a legitimate and worthwhile aim per se should be increasingly questioned in units where good-quality surgery is being performed. Good communication between radiologists, pathologists, and surgeons is critical to the improvement of overall quality of care for rectal cancer patients.

We accept that the algorithm we are proposing does go against the conventional wisdom that all patients with T3 rectal cancer should receive CRT. This has worked well for the past 2 decades; however, in our view, in 2014 it now needs re-evaluation.

Financial Disclosure:The authors have no significant financial interest in or other relationship with the manufacturer of any product or provider of any service mentioned in this article.

References:

1. Salerno G, Sinnatamby C, Branagan G, et al. Defining the rectum: surgically, radiologically and anatomically. Colorectal Dis. 2006;8(Suppl 3):5-9.

2. Swedish Rectal Cancer Trial. Improved survival with preoperative radiotherapy in resectable rectal cancer. Swedish Rectal Cancer Trial. N Engl J Med. 1997;336:980-7.

3. Sauer R, Becker H, Hohenberger W, et al. Preoperative versus postoperative chemoradiotherapy for rectal cancer. N Engl J Med. 2004;351:1731-40.

4. Bosset JF, Collette L, Calais G, et al. Chemotherapy with preoperative radiotherapy in rectal cancer. N Engl J Med. 2006;355:1114-23.

5. Braendengen M, Tveit KM, Berglund A, et al. Randomized phase III study comparing preoperative radiotherapy with chemoradiotherapy in nonresectable rectal cancer. J Clin Oncol. 2008;26:3687-94.

6. Taylor FG, Quirke P, Heald RJ, et al; for the MERCURY study group. Preoperative high-resolution magnetic resonance imaging can identify good prognosis stage I, II, and III rectal cancer best managed by surgery alone: a prospective, multicenter, European study that recruited consecutive patients with rectal cancer. Ann Surg. 2011;253:711-9.

7. Frasson M, Garcia-Granero E, Roda D, et al. Preoperative chemoradiation may not always be needed for patients with T3 and T2N+ rectal cancer. Cancer. 2011;117:3118-25.

8. Schrag D, Weiser MR, Goodman KA, et al. Neoadjuvant chemotherapy without routine use of radiation therapy for patients with locally advanced rectal cancer: a pilot trial. J Clin Oncol. 2014;32:513-8.

9. MERCURY Study Group. Diagnostic accuracy of preoperative magnetic resonance imaging in predicting curative resection of rectal cancer: prospective observational study. BMJ. 2006;333:779.

10. Gunderson LL, Sargent DJ, Tepper JE, et al. Impact of T and N stage and treatment on survival and relapse in adjuvant rectal cancer: a pooled analysis. J Clin Oncol. 2004;22:1785-96.

11. Merkel S, Mansmann U, Siassi M, et al. The prognostic inhomogeneity in pT3 rectal carcinomas. Int J Colorectal Dis. 2001 Sep;16:298-304.

12. Quirke P, Durdey P, Dixon MF, Williams NS. Local recurrence of rectal adenocarcinoma due to inadequate surgical resection. Histopathological study of lateral tumor spread and surgical excision. Lancet. 1986; 2:996-9.

13. Nagtegaal ID, Quirke P. What is the role for the circumferential margin in the modern treatment of rectal cancer? J Clin Oncol. 2008;26:303-12.

14. Tepper JE, O’Connell MJ, Niedzwiecki D, et al. Impact of number of nodes retrieved on outcome in patients with rectal cancer. J Clin Oncol. 2001;19:157-63.

15. Stocchi L, Nelson H, Sargent DJ, et al. Impact of surgical and pathologic variables in rectal cancer: a United States community and cooperative group report. J Clin Oncol. 2001;19:3895-902.

16. Brabender J, Bollschweiler E, Hölscher AH, et al.The prognostic impact of extracapsular lymph node involvement in rectal cancer patients: implications for staging and adjuvant treatment strategies. Oncol Lett. 2012;3:825-30.

17. Quirke P, Steele R, Monson J, et al; MRC CR07/NCIC-CTG CO16 Trial Investigators; NCRI Colorectal Cancer Study Group. Effect of the plane of surgery achieved on local recurrence in patients with operable rectal cancer: a prospective study using data from the MRC CR07 and NCIC-CTG CO16 randomised clinical trial. Lancet. 2009;373:821-8.

18. Talbot IC, Ritchie S, Leighton MH, et al. The clinical significance of invasion of veins by rectal cancer. Br J Surg. 1980;67:439-42.

19. Maughan NJ, Morris E, Forman D, Quirke P. The validity of the Royal College of Pathologists’ colorectal cancer minimum dataset within a population. Br J Cancer. 2007; 97:1393-8.

20. Sebag-Montefiore D, Stephens RJ, Steele R, et al. Preoperative radiotherapy versus selective postoperative chemoradiotherapy in patients with rectal cancer (MRC CR07 and NCIC-CTG C016): a multicentre, randomised trial. Lancet. 2009;373:811-20.

21. Peng J, Sheng W, Huang D, et al. Perineural invasion in pT3N0 rectal cancer: the incidence and its prognostic effect. Cancer. 2011;117:1415-21.

22. Sitzler PJ, Seow-Choen F, Ho YH, Leong AP. Lymph node involvement and tumor depth in rectal cancers: an analysis of 805 patients. Dis Colon Rectum. 1997;40:1472-6.

23. den Dulk M, Collette L, van de Velde CJ, et al; EORTC Radiation Oncology Group. Quality of surgery in T3-4 rectal cancer: involvement of circumferential resection margin not influenced by preoperative treatment. Results from EORTC trial 22921. Eur J Cancer. 2007;43:1821-8.

24. Hermanek P, Wiebelt H, Staimmer D, Riedl S. Prognostic factors of rectum carcinoma-experience of the German Multicentre Study SGCRC. German Study Group Colo-Rectal Carcinoma. Tumori. 1995;81(Suppl 3):60-4.

25. Sebag-Montefiore D, Steele R, Grieve R, et al. Mature follow up of the MRC CR07 NCIC CO16 trial after a median of 8 years. In: National Cancer Research Institute (NCRI) Cancer Conference 2013; Liverpool, UK. NCRI; Nov 2013. Abstract LB73.

26. Pollheimer MJ, Kornprat P, Pollheimer VS, et al. Clinical significance of pT sub-classification in surgical pathology of colorectal cancer. Int J Colorectal Dis. 2010;25:187-96.

27. Ngan SY, Burmeister B, Fisher RJ, et al. Randomized trial of short-course radiotherapy versus long-course chemoradiation comparing rates of local recurrence in patients with T3 rectal cancer: Trans-Tasman Radiation Oncology Group Trial 01.04. J Clin Oncol. 2012;30:3827-33.

28. Hall NR, Finan PJ, al-Jaberi T, et al. Circumferential margin involvement after mesorectal excision of rectal cancer with curative intent. Predictor of survival but not local recurrence? Dis Colon Rectum. 1998;41:979-83.

29. Bugg WG, Andreou AK, Biswas D, et al. The prognostic significance of MRI-detected extramural venous invasion in rectal carcinoma. Clin Radiol. 2014 Feb 26. [Epub ahead of print]

30. Betge J, Pollheimer MJ, Lindtner RA, et al. Intramural and extramural vascular invasion in colorectal cancer: prognostic significance and quality of pathology reporting. Cancer. 2012;118:628-38.

31. Hunter CJ, Garant A, Vuong T, et al. Adverse features on rectal MRI identify a high-risk group that may benefit from more intensive preoperative staging and treatment. Ann Surg Oncol. 2012;19:1199-205.

32. Wind J, Lagarde SM, Ten Kate FJ, et al. A systematic review on the significance of extracapsular lymph node involvement in gastrointestinal malignancies. Eur J Surg Oncol. 2007;3:401-8.

33. Hermanek P, Merkel S, Fietkau R, Rödel C, Hohenberger W. Regional lymph node metastasis and locoregional recurrence of rectal carcinoma in the era of TME [corrected] surgery. Implications for treatment decisions. Int J Colorectal Dis. 2010;25:359-68.

34. Brabender J, Bollschweiler E, Hölscher AH, et al. The prognostic impact of extracapsular lymph node involvement in rectal cancer patients: implications for staging and adjuvant treatment strategies. Oncol Lett. 2012;3:825-830.

35. Engelen SM, Beets-Tan RG, Lahaye MJ, et al. Location of involved mesorectal and extramesorectal lymph nodes in patients with primary rectal cancer: preoperative assessment with MR imaging. Eur J Surg Oncol. 2008;34:776-81.

36. Beets-Tan RG, Beets GL, Vliegen RF, et al. Accuracy of magnetic resonance imaging in prediction of tumor-free resection margin in rectal cancer surgery. Lancet. 2001;357:497-504.

37. Taylor FG, Quirke P, Heald RJ, et al; MERCURY Study Group. Preoperative high-resolution magnetic resonance imaging can identify good prognosis stage I, II, and III rectal cancer best managed by surgery alone: a prospective, multicenter, European study. Ann Surg. 2011;253:711-9.

38. Ruppert R, Ptok H, Strassburg J, et al. [Quality indicators of diagnosis and therapy in MRI-based neoadjuvant radiochemotherapy for rectal cancer-interim analysis of a Prospective Multicentre Observational Study (OCUM)]. Zentralbl Chir. 2013;138:630-5.

39. Smith NJ, Shihab O, Arnaout A, et al. MRI for detection of extramural vascular invasion in rectal cancer. Am J Roentgenol. 2008;191:1517-22.

40. Mantke R, Schmidt U, Wolff S, et al. Incidence of synchronous liver metastases in patients with colorectal cancer in relationship to clinico-pathologic characteristics. Results of a German prospective multicentre observational study. Eur J Surg Oncol. 2012;38:259-65.

41. Minsky BD, Mies C, Rich TA, et al. Colloid cancer of the colon and rectum. Cancer. 1987;60:3103-12.

42. Nagtegaal ID, Marijnen CA, Kranenbarg EK, et al. Circumferential margin involvement is still an important predictor of local recurrence in rectal carcinoma: not one millimeter but two millimeters is the limit. Am J Surg Pathol. 2002;26:350-7.

43. Kang H, O’Connell JB, Maggard MA, et al. A 10-year outcomes evaluation of mucinous and signet-ring cell carcinoma of the colon and rectum. Dis Colon Rectum 2005;48:1161-8.

44. Jass JR, Sobin LH (editors). Histological typing of intestinal tumors: World Health Organization. 2nd ed. New York, NY: Springer-Verlag NY Inc; 1989.

45. Nagtegaal I, Gaspar C, Marijnen C, et al. Morphological changes in tumour type after radiotherapy are accompanied by changes in gene expression profile but not clinical behaviour. J Pathol. 2004;204:183-92.

46. Sengul N, Wexner SD, Woodhouse S, et al. Effects of radiotherapy on different histopathological types of rectal carcinoma. Colorectal Dis. 2006;8:283-8.

47. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Rectal cancer. Version 2.2014. Available from: http://www.nccn.org/professionals/
physician_gls/f_guidelines.asp#site. Accessed April 15, 2014.

48. Dewhurst C, Rosen MP, Blake MA, et al. ACR Appropriateness Criteria®. Pretreatment staging of colorectal cancer. J Am Coll Radiol. 2012;9:775-81.

49. Jones WE 3rd, Thomas CR Jr, Herman JM, et al. ACR Appropriateness Criteria®. Resectable rectal cancer. Radiat Oncol. 2012;7:161.

50. Glimelius B, Tiret E, Cervantes A, Arnold D; ESMO Guidelines Working Group. Rectal cancer: ESMO Clinical Practice Guidelines for diagnosis, treatment and follow-up. Ann Oncol. 2013;24(Suppl 6):vi81-8.

51. National Institute of Clinical Excellence (NICE). Colorectal cancer: the diagnosis and management of colorectal cancer. Clinical Guidelines, CG131. Issued: November 2011. Availble from: http://guidance.nice.org.uk/CG/Wave16/2. Accessed July 11, 2014.

52. Maas M, et al. Long-term outcome in patients with a pathological complete response after chemoradiation for rectal cancer: a pooled analysis of individual patient data. Lancet Oncol. 2010;11:835-44.

53. Shihab OC, Moran BJ, Heald RJ, et al. MRI staging of low rectal cancer. Eur Radiol. 2009;19:643-50.

54. How P, West NP, Brown G. An MRI-based assessment of standard and extralevator abdominoperineal excision specimens: time for a patient tailored approach? Ann Surg Oncol. 2014;21:822-8.

55. Messenger DE, Driman DK, Kirsch R. Developments in the assessment of venous invasion in colorectal cancer: implications for future practice and patient outcome. Hum Pathol. 2012;43:965-73.

56. Yu SK, Tait D, Chau I, Brown G. MRI predictive factors for tumor response in rectal cancer following neoadjuvant chemoradiation therapy-implications for induction chemotherapy? Int J Radiat Oncol Biol Phys. 2013;87:505-11.

57. Ptok H, Ruppert R, Stassburg J, et al. Pretherapeutic MRI for decision-making regarding selective neoadjuvant radiochemotherapy for rectal carcinoma: interim analysis of a multicentric prospective observational study. J Magn Reson Imaging. 2013;37:1122-8.

58. Perez RO, Pereira DD, Proscurshim I, et al. Lymph node size in rectal cancer following neoadjuvant chemoradiation-can we rely on radiologic nodal staging after chemoradiation? Dis Colon Rectum. 2009;52:1278-84.

59. Attenberger UI, Pilz LR, Morelli JN, et al. Multi-parametric MRI of rectal cancer-do quantitative functional MR measurements correlate with radiologic and pathologic tumor stages? Eur J Radiol. 2014;83:1036-43.

60. Syk E, Glimelius B, Nilsson PJ. Factors influencing local failure in rectal cancer: analysis of 2315 patients from a population-based series. Dis Colon Rectum. 2010;53:744-52.

61. Bondeven P, Hagemann-Madsen RH, Bondeven P, et al. Extent and completeness of mesorectal excision evaluated by postoperative magnetic resonance imaging. Br J Surg. 2013;100:1357-67.

62. Dahlberg M, Stenborg A, Pahlman L, et al. Cost-effectiveness of preoperative radiotherapy in rectal cancer: results from the Swedish Rectal Cancer Trial. Int J Radiat Oncol Biol Phys. 2002;54:654-60.

63. Durand-Zaleski I. [Economic evaluation of radiotherapy: methods and results.] Cancer Radiother. 2005;9:449-51.

64. Van Den Brink M, Van Den Hout WB, Stiggelbout AM, et al; Dutch Colorectal Cancer Group. Cost-utility analysis of preoperative radiotherapy in patients with rectal cancer undergoing total mesorectal excision: a study of the Dutch Colorectal Cancer Group. J Clin Oncol. 2004;22:244-53.

65. van Gijn W, Marijnen CA, Nagtegaal ID, et al; Dutch Colorectal Cancer Group. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer: 12-year follow-up of the multicentre, randomised controlled TME trial. Lancet Oncol. 2011;12:575-82.

66. Bosset JF, Calais G, Mineur L, et al; EORTC Radiation Oncology Group. Fluorouracil-based adjuvant chemotherapy after preoperative chemoradiotherapy in rectal cancer: long-term results of the EORTC 22921 randomised study. Lancet Oncol. 2014;15:184-90.

67. Engelen SM, Maas M, Lahaye MJ, et al. Modern multidisciplinary treatment of rectal cancer based on staging with magnetic resonance imaging leads to excellent local control, but distant control remains a challenge. Eur J Cancer. 2013;49:2311-20.