The tumour is localised using a T2-weighted turbo (or fast) spin echo (TSE or FSE) sequence in the sagittal plane. High resolution (3.0mm) T2-weighted images are then acquired perpendicular to the tumour to allow assessment of the circumferential extent of the lesion. A pelvic phased array coil is required for images of suitable quality. Endo-rectal coils are not required and may have inherent problems (see below). Smooth muscle relaxants improve image quality, though not necessarily staging accuracy. Similarly, intravenous contrast enhancement and endo-rectal contrast agents may increase the conspicuity of the tumour though not influence tumour staging.
There is no routine indication for MR imaging in the staging of colon cancer. However, pre-operative staging of rectal cancer with pelvic phased array coil MR imaging is rapidly evolving as a routine procedure. MR accurately determines the proximity of tumours to the meso-rectal fascia, allowing identification of patients who may benefit from neo-adjuvant radiotherapy. Conversely, MR accurately defines earlier stages of disease, preventing unnecessary use of neo-adjuvant radiotherapy in these patients.
Staging accuracy of up to 82% for local tumour staging has been reported. This, however, takes into account studies that used inherently low-resolution body coil techniques. More recently, 100% correlation with histological staging was demonstrated using a pelvic phased array coil and high-resolution technique. Local staging accuracy of 84% has been quoted for endo-rectal coil techniques. However, acquisition of an image plane perpendicular to the tumour may not always be possible using endo-anal MRI and technical difficulties may also arise with high rectal tumours or a tightly stenotic lesion. Near-field flaring may also occur with endo-rectal techniques, although most manufacturers have developed software to overcome this problem.
Sensitivity for nodal infiltration has been reported as 65%, though this may improve with higher resolution techniques and detailed assessment of nodal morphology. Research is ongoing into the efficacy of super-paramagnetic iron oxide contrast agents in identifying malignant nodes in a variety of pelvic tumours including rectal cancer.
MRI has been shown to be superior to CT for the identification and characterisation of malignant liver lesions, when current techniques have been compared in most reports since 1990. A recent study comparing super-paramagnetic iron oxide MRI and dual-phase CT demonstrated a better sensitivity for MR (80.6%) than CT (73.5%) in the demonstration of colorectal liver metastases. In addition, MRI has been shown to be more sensitive than CT in the identification of extra-hepatic disease.
There are five layers of the normal rectal wall, namely: mucosa, muscularis mucosa, submucosa, circular and longitudinal muscle (muscularis propria). Although all five layers may be seen on high resolution T2W1, this is rarely the case. In most cases, the mucosa and submucosa are seen as a single high signal layer deep to the low signal muscularis propria. The normal rectal wall should be no more than 6mm thick in the distended state. Rectal tumours are usually of intermediate intensity compared with the muscularis propria of the bowel wall on T2W1. Mucinous tumours return a high signal due to high fluid content. Important features to be assessed are:
• proximity of tumour to the meso-rectal fascia;
• proximity of tumour to the anal sphincter;
• presence of meso-rectal nodes and peri-rectal tumour deposits and their proximity to the meso-rectal fascia;
• peri-vascular infiltration.
In common with various other tumour types, enlarged nodes in the presence of rectal cancer may be simply inflammatory and non-enlarged nodes may contain tumour. Pointers towards malignancy include an overtly infiltrative gland margin, nodal necrosis and signal intensity similar to the primary tumour (such as high signal nodes infiltrated by mucinous tumour). Proximity of suspicious nodes to the meso-rectal fascia should be noted as this may compromise the surgical plane and the patient may benefit from neo-adjuvant radiotherapy.
Radiotherapy. The effects of rectal irradiation depend upon the total radiation dose, and will be more pronounced following a long (five weeks) course of treatment compared with the shorter five day regime. In the acute and sub-acute post treatment phases (up to twelve months post therapy) changes on MRI reflect cell death and inflammation. Hence, there may be visceral mural thickening (e.g. bowel, bladder, vagina, urethra), increased mucosal and submucosal signal on T2W1, and mucosal enhancement with intravenous contrast agents. The chronic effects of radiotherapy reflect fibrosis. Therefore, there may still be visceral mural thickening, though this will usually return low signal on T1 and T2W1 and show little or no intravenous contrast enhancement. Other changes include peri-rectal fascial and peritoneal thickening, inflammation and atrophy of adjacent pelvic muscles and fatty marrow change (high signal on T1W1) in involved bone. More severe changes may include strictures of bowel and fistulation (e.g. recto-vaginal or recto-vesical).
Surgery. Both anterior resection and abdomino-perineal (APR) total meso-rectal excision may produce pre-sacral and perirectal fibrotic change. This is usually more pronounced following an APR. Also, post APR, there may be a pre-sacral 'pseudo-mass' on CT and T1W1 due to the uterus, prostate or seminal vesicles occupying the void in the rectal bed. This should be readily discernible on T2W1.
Pelvic recurrence has been reported in up to 50% of patients with rectal cancer, though following TME with negative circumferential resection margins,this figure falls to less than 10%. Recurrent tumour usually manifests as a mass returning intermediate signal intensity on T2W1 compared with muscle. However, radiation fibrosis may also return intermediate signal especially within the first two years following irradiation. Also, although fibrotic tissue returns low signal on T1 and T2W1, the same may apply to desmoplastic foci of tumour. Dynamic intravenous contrast-enhanced MR imaging may add to the specificity of diagnosing recurrent disease. MR imaging is accurate in assessing the extent of proven recurrent tumour and can significantly aid the planning of further ' salvage' surgery (see chapter 14).
• Tumour identification on the planning T2-weighted sagittal view: This may be difficult, though it is crucial to further imaging. Knowledge of the tumour site at sigmoidoscopy is very important and should be included in the clinical information. Tumour is usually of intermediate signal intensity relative to the muscularis propria and flatus. At the site of tumour.there is often loss of detail of the different layers of the rectal wall, which is thickened, except in the case of T1 lesions. Other useful pointers may be the presence of blood vessels entering the tumour, or occasionally, bowel wall retraction at the tumour site. Smooth muscle relaxants and intravenous contrast enhancement may increase tumour conspicuity, but, with experience, are seldom required. There may be synchronous tumours in the rectum or colon. Careful review of the sagittal MR image should exclude the former, though the patient will also require formal colonic evaluation by colonoscopy ('virtual' or fibre-optic) or barium enema.
Over-staging due to peri-tumoural fibrosis: This appears as linear, low signal stranding into the meso-rectal fat, as opposed to tumour, which has a nodular interface with the meso-rectum, and usually returns intermediate signal. Over-staging due to partial volume artifact: This should be minimised by meticulous attention to detail when planning the high-resolution sequence perpendicular to the tumour.
Intra-vascular tumour infiltration: This is defined as tumour tissue extending to infiltrate a vessel or as satellite deposits in the meso-rectal fat.
Adjacent visceral infiltration: This is clear evidence of the primary tumour extending into the adjacent viscera. Contiguity of the tumour with an adjacent organ does not necessarily indicate invasion and peri-tumoural fibrosis or inflammation may cause over-staging in these cases.
Nodal disease: Mr imaging remains inaccurate in the assessment of meso-rectal lymph node metastasis, though there are some useful discriminating features such as round shape, irregular margin or increased signal intensity on T2W1 indicating central nodal necrosis or a mucinous tumour deposit.
Recurrent disease: Radiation effect and/or surgical changes may give rise to an intermediate signal pre-sacral mass, mimicking tumour recurrence. Recurrent disease may return low signal due to desmoplasia, mimicking simple fibrosis. Biopsy may be required to clarify this although sampling errors can be problematical. Alternatively, repeat interval imaging and observation for recurrent disease outside of the pre-sacral area is valuable.
Intussuscepted tumour makes it difficult to precisely T stage the tumour as the interface with adjacent fat is not evident. However, an intussuscepted tumour cannot be fixed and should be evident on the sagittal view. The key feature on the perpendicular image is the presence of several layers of muscularis propria.
Linitis Plastica. This is a rare form of rectal cancer where malignant 'signet ring' cells or undifferentiated carcinoma diffusely infiltrate the submucosa and muscularis propria, with relative sparing of the mucosa. On T1W1 there is marked thickening of the rectal wall and meso-rectal fascia. On T2W1 there may be a ring pattern in the rectal wall due to a combination of tumour infiltration and fibrosis around intact layers of the muscularis propria. Early recognition is important as it may guide the physician to perform deep biopsies (if the mucosa is spared). Also, because the prognosis is very poor, the patient may be spared unnecessary major surgery.
1. Skibber JM, Minsky BD and Hoff PM. (2001) Cancer of the colon. In: Cancer. Principles and Practice of Oncology, 6th Edn (eds DeVita VT Jr., Hellman S, Rosenberg SA). Lippincott, Williams and Wilkins, Philadelphia, Pensylvania, USA, pp. 1216-1271. This is a definitive, comprehensive and up to date text on all aspects ofrectal cancer and its management.
2. Souhami R and Tobias J. (1995) Tumours of the small and large bowel. In: Cancer and its management, 2nd Edn. Blackwell Science, Oxford, UK, pp. 305-316. A more concise though equally authoritative chapter on all aspects of rectal cancer.
3. Heald RJ. (1995) Total mesorectal excision is optimal surgery for rectal cancer:a Scandinavian Consensus. Br.J. Surg. 82:1297-1299. This paper defines the best surgical practice for patients with operable rectal cancer.
4. Adam IJ, Mohamdee MO, Martin IG, Scott N, Finan PJ, Johnston D, Dixon MF, Quirke P. (1994) Role of circumferential margin involvement in the local recurrence of rectal cancer. Lancet 344: 707-711. This paper highlights the increased risk of local tumour recurrence in patients with positive circumferential resection margins following rectal cancer surgery.
5. Bissett IP, Fernando CC, Hough DM etal. (2001) Identification of the fascia propria by magnetic resonance imaging and its relevance to pre-operative assessment of rectal cancer. Dis. Colon Rectum 44(2): 259-265. Elegant confirmation of the MR appearances of the meso-rectal fascia by cadaveric imaging.
6. Kapiteijn E, Marijnen CAM, Nagstegaal ID et al. (2001) Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N. Engl.J. Med 345(9): 638-646. Large controlled trial proving that short-course pre-operative radiotherapy reduces the risk oflocal recurrence in patients with rectal cancer who have had a standardised TME.
7. Beets-Tan RG, Beets GL, Vliegen RF et al. (2001) Accuracy of MR imaging in the prediction of tumour-free resection margins in rectal cancer surgery. Lancet 357: 497-504. This paper demonstrates that the distance of tumour from the meso-rectal fascia can be accurately predicted on pre-operative MR imaging, allowing identification ofpatients at risk for positive resection margins.
8. Brown G, Richardson CJ, Newcombe RG etal. (1999) Rectal carcinoma: thin-section MR imaging for staging in 28 patients. Radiology 211 (I): 215-222. Seminal paper on the method and accuracy of high resolution MR imaging in defining the stage and depth of extramural invasion of rectal cancer extent using surface phased array coils.
9. KwokH, BissettIP and Hill GL. (2000) Pre-operative staging of rectal cancer. Int.J. Colorectal Dis. 15:9-20. A large meta-analysis of almost 5000 patients in 83 studies comparing CT, MRI and endoanal ultrasound for staging rectal cancer.
10. Okizuka H, Sugimura K, Yoshizako et al. (1996) Rectal Carcinoma: Prospective comparison of conventional and gadopentate dimeglumine enhanced fat-suppressed MR imaging. J. Magn. Reson. Imaging 6:465-471. A study of thirty two patients showing that fatsuppressed, Gd-enhanced TI-weighted images give excellent rectal cancer detection though do not improve staging accuracy compared with non-enhanced images.
11. Dicle O, Obuz F and Cacmakci H. (l999) Differentiation of recurrent rectal cancer and scarring with dynamic MR imaging. Br.J. Radiol. 72:1155-1159. This small study (n=19) supports the use of quantitative dynamic contrast enhanced MR imaging in distinguishing locally recurrent rectal cancer from benign tissue.
12. Hawnaur JM, Zhu XP and Hutchinson CE. (1998) Quantitative dynamic contrast-enhanced MRI of recurrent pelvic masses in patients treated for cancer. Br.J. Radiol. 71 (851): 1136-1142. This study (n=32) showed that the combination of quantitative dynamic MR Imaging, signal intensity and tissue morphology on conventional MR images may be helpful in distinguishing recurrent pelvic tumour from benign disease.
13. Robinson P, Carrington BM, Swindell R et al. (2002) Recurrent or residual bowel cancer: Accuracy of MRI Local Extent Before Salvage Surgery. Clin. Radiol. 57:514-522. The most thorough assessment to date of the ability of MRI to detail the extent of recurrent tumour, and its usefulness in pre-operative surgical planning.
14. Semelka RC and Helmberger TK. (2001). Contrast agents for MR Imaging of the liver. Radiology 218:27-38. Excellent review of state-of-the-art MRI contrast liver imaging.
Figure 9.2. Normal rectum.
(a) Sagittal T2W1 and (b) Transaxial T2W1 of normal rectum. Note signal void due to gas in the rectal lumen (R), high signal combination of mucosa and submucosa (arrowheads) and the low signal muscularis propria (arrows). Symphysis pubis (SP). Intermediate signal endometrial tumour (asterisk), cervix (C).
Figure 9.3. Lateral view of a surgical specimen from an abdominoperineal TME.
Note the peritoneal reflection (arrowheads) and the Pouch of Douglas (asterisk), the uterus having been removed. Peritoneum covers the anterior and lateral aspects of the upper two thirds of the rectum. The lower third of rectum is circumferentially invested by meso-rectal fascia the visceral surface of which (arrow) invests the surgical specimen. Note the anus (AC) and the ischio-anal fossa fat (IAF).
Figure 9.4. T1 Rectal cancer.
Transaxial T2W1 of the lower third of rectum. Note the intact muscularis propria (arrow). The tumour (T) can be distinguished from faecal residue by the fact that there is a vessel entering the mass (arrowheads). Bladder (B), obturator internus muscles (asterisk) and sacrum (S).
Figure 9.5. T2 Rectal cancer.
T2W1 perpendicular to an upper rectal tumour (T). Note the muscularis propria (arrow), which is focally infiltrated (asterisk) by the tumour. Note the peritoneal reflection (curved arrow) and piecemeal depiction of the mesorectal fascia (arrowheads). Bladder (B)
Figure 9.6. T3 Lower third rectal cancer.
(a) Sagittal T2W1 demonstrating a posterior wall lower third rectal tumour (T) and (b) high resolution T2W1 perpendicular to the tumour. Note the axis planned for the subsequent high-resolution off axis sequence (line).There is a broad front of infiltration into the meso-rectal fat (arrowheads) making this a T3 lesion, though the meso-rectal fascia (circumferential resection margin) is well clear (arrows). Note the peritoneal reflection anteriorly (open arrows), rectal lumen (L), obturator internus muscle (O), coccyx (C) anal canal (AC), levator ani (curved arrow), prostate gland (PG), bladder (B), sacrum (S), pubic symphysis (PS).
Figure 9.7. T3 Mid-rectal cancer.
High resolution T2W1 perpendicular to a tumour (T) of the anterior and right mid-rectal walls. The tumour demonstrates central ulceration (U) and is contiguous anteriorly with the retro-prostatic or Denonvilliers' fascia (arrows) making this T-stage 3 with an involved anterior margin. This patient would benefit from a long course of pre-operative radiotherapy to reduce the risk of a positive anterior resection margin. Note how little space there is between the rectal wall and the meso-rectal fascia anteriorly compared with laterally, making anterior tumours at particular risk for resection margin involvement. Meso-rectal fascia (arrowheads), rectal lumen (L), bladder (B), prostate central
T2W1 perpendicular to an extensive T3 stage mid-rectal tumour (T). There is a broad front of nodular tumour extending into the right side of the mesorectum with a tumour nodule (arrow) infiltrating the right meso-rectal fascia. This nodule may in fact reflect an infiltrated lymph node. This patient would require a long course of neo-adjuvant radiotherapy to reduce the risk of a positive lateral resection margin. Note the normal meso-rectal fascia elsewhere (arrowheads). Bladder (B), prostate central zone (PCZ), obturator internus muscle (O). An infiltrative tumour edge such as this carries a poorer prognosis than tumours with a better-defined margin such as in Figure 9.7.
T2W1 perpendicular to a mid-third rectal tumour (T). Intact muscle is shown around the tumour (closed arrowheads), though there is histologically proven intra-vascular infiltration on the left side (open arrowheads) making this a T3 stage tumour. The intra- vascular infiltration describes a cigar shape extending around the blood vessel. This is in contrast to the more usual broad front oftumour infiltration seen in Figure 9.8 and the nodular tumour infiltration and fine stranding of peri-tumoural fibrosis seen in Figure 9.10. Seminal vesicle (SV), bladder (B), piriformis muscle (P), coccyx (C), obturator internus muscle (O).
Figure 9.10. T3 Rectal cancer contrasted against peri-tumoural fibrosis.
(a) Sagittal T2W planning image through a bulky upper and mid- third rectal tumour (T) and (b) high resolution T2W1 perpendicular to the tumour. Note focal infiltration dorsally into the meso-rectum (arrowheads). There is quite extensive fine stranding (crosssed arrows) extending into the dorsal mesorectal fat. This has been shown to represent peri-tumoural fibrotic change. Additionally however there is a nodular front of infiltrative tumour from the right and left walls (curved arrows) making this a stage T3 lesion. Note the anterior peritoneal reflection (straight arrows), and the meso-rectal fascia (open arrows), which are not infiltrated. Seminal vesicle (SV), prostate central zone (PCZ), prostate peripheral zone (PPZ), bladder (B), levator ani muscle (curved open arrow), anal canal (AC), sacrum (S),piriformis muscle (P).
(a) Sagittal T2W1 showing an unusual tumour (T) anteriorly at the ano-rectal junction and (b) high resolution T2W1 perpendicular to the tumour. The tumour has infiltrated anteriorly into the plane between the vagina (arrowheads) and the anal canal (AC).The vagina (V) is not overtly infiltrated so this is an advanced T3 stage lesion. Urethra (Ur), bladder (B), levator ani (curved arrow), superior pubic ramus (SPR), lower rectal lumen (L), ischio-anal fossa (IAF), inferior pubic ramus (IPR),anal canal (AC). Note the incidental pre-sacral liposarcoma (LS).
T2W1 perpendicular to a tumour of the left side of lower rectum (T) just above the ano-rectal junction. Tumour extends anteriorly to infiltrate the left posterior vaginal wall where the muscle layer is destroyed (arrows). Note the intact posterior vaginal muscle on the right (arrowheads). This is therefore a T4 lesion. There is a broad base of contiguity with the left levator ani muscle (curved arrows), which was also infiltrated. Ischio-anal fossa (IAF), right superior pubic ramus (SPR), bladder neck (BN).
(a) Sagittal T2W1 through an extensive mid- and upper third rectal tumour (T) and (b) T2W1 perpendicular to the more caudal part of the tumour. The peritoneal reflection (arrowheads) is thickened with focal nodular peritoneal deposits (arrows) making this a T stage 4 lesion. Note the enlarged mesorectal node
(asterisk), Prostate gland (PG), anal canal (AC), symphysis pubis (SP), bladder (B), sacrum (S). left obturator internus muscle (O).
Coronal T2W1 of a large lower rectal tumour (T) that infiltrates through the right rectal wall (arrowheads) and is contiguous with the right levator ani muscle (curved arrow). However, it retains its normal morphology and is not infiltrated. Left ischio-anal fossa (IAF) and left levator ani muscle (arrow).
Coronal T2W1 showing a large lower rectal tumour (T) which clearly infiltrates the right levator ani muscle (curved arrow). Left ischio-anal fossa (IAF) and left levator ani muscle (straight arrow).
Off-axis T2W1 of a large, high signal mucinous tumour (T) which has formed a sinus track (straight arrow)
through the left levator ani muscle. Left ischioanal fossa (IAF), left levator ani muscle (curved arrow).
(a) T2W1 perpendicular to a mucinous mid-rectal tumour (T) and (b) sagittal T2W1 of the tumour. Note posterior extension into a lower left sacral foramen (curved arrow).Anteriorly, the tumour is contiguous with the posterior bladder wall (straight arrow) and there is gas in the bladder (asterisk) indicating recto-vesical fistulation (F).The tumour is contiguous over a wide area with the prostate (P) and bladder (B). Anal canal (AC), right superior pubic ramus (SPR).
Figure 9.18. Enlarged benign meso-rectal lymph node.
Transaxial T2W1 through the mid rectum (R) showing a clearly enlarged left meso-rectal lymph node (arrow). This was benign on histology. Bladder (B) and sacrum (S).
(a) Sagittal T2W1 showing a high signal mucinous tumour (T) on the posterior upper rectal wall and (b) T2W1 perpendicular to the tumour. The signal intensity of the tumour is similar to that of adjacent faecal residue (FR), though the tumour is defined by adjacent mural retraction (arrowheads). There is focal infiltration of the muscularis propria (arrows) making this a T2 lesion. In addition, there is a large right-sided meso-rectal lymph node (curved arrow). This returns high signal, which is abnormal, and is consistent with mucinous tumour infiltration. Bladder (B), prostate gland (PG).
Figure 9.20. N2 Rectal cancer.
Transaxial T2W1 showing a left sided mid-rectal tumour (T). The muscularis propria (curved arrow) is poorly defined though there is no overt meso-rectal infiltration. There are enlarged nodes in the left side of the meso-rectum and right internal iliac territory (asterisks). The right internal iliac node demonstrates inhomogeneous signal intensity and the meso-rectal node on the left demonstrates an irregular margin (arrowheads) both highly suggestive of lymph node tumour infiltration, which was subsequently proven. Bladder (B), right seminal vesicle (SV), coccyx (C).
(a) Transaxial gradient echo proton density image (GEPDI) of liver and (b) transaxial gradient echo (GE) T2W1 of liver showing two separate metastatic deposits (M) in segment 4A. Also note amorphous increased signal in segment 8 (asterisk). (c) Transaxial GEPDI and (d) transaxial GET2W1 of liver obtained 60 minutes following infusion of super-paramagnetic iron oxide particles (SPIO) (Endorem 15mmol kg-1). There is marked increase in contrast between the metastases (M) and background liver. The lesion in segment 8 (asterisk) is now seen to represent further metastatic disease.
SPIOs are taken up by Kupffer cells, which are present in normal liver but absent from metastatic deposits.
Therefore SPIOs cause signal drop out in normal liver but not in metastases. The lesion/background contrast is greatest using moderately T2W/proton density images at high field strengths of 1-1.5 T. Gallbladder (GB).
Figure 9.22. Advanced local lymph node metastasis.
Transaxial T1W1 showing multiple enlarged lymph nodes (arrowheads) in the upper meso-rectum in the distribution of the superior rectal vessels. Note the T3 rectal tumour (T), sacrum (S) and piriformis muscle (P).
Figure 9.23. Intussuscepted tumour.
(a) Sagittal T2W1 through an intussuscepted mid-rectal tumour (T) and (b) T2W1 perpendicular to the lesion.
This was pathological Stage T2 disease though the intussusception makes it very difficult to accurately define the stage on MR imaging. Note the multiple layers of muscularis propria (arrowheads) and intussuscepted meso-rectal fat (asterisk). Anal canal (AC), urethra (Ur), bladder (B), superior pubic ramus (SPR), sacrum (S), coccyx (C).
Was this article helpful?
Learning About 10 Ways Fight Off Cancer Can Have Amazing Benefits For Your Life The Best Tips On How To Keep This Killer At Bay Discovering that you or a loved one has cancer can be utterly terrifying. All the same, once you comprehend the causes of cancer and learn how to reverse those causes, you or your loved one may have more than a fighting chance of beating out cancer.