First published online September 27, 2006
British Journal of Radiology (2007) 80, 242-247
© 2007 British Institute of Radiology
doi: 10.1259/bjr/66311683
Variation of mesorectal volume with abdominal fat volume in patients with rectal carcinoma: assessment with MRI
S D Allen, MBBS, FRCR
V Gada, MBBS, FRCR
and
D M Blunt, MBBS, FRCR
Department of Radiology, Hammersmith Hospitals NHS Trust, Charing Cross Hospital, Fulham Palace Road, London W6 8RF, UK
Correspondence: Dr Steven D Allen, Department of Radiology, Hammersmith Hospitals NHS Trust, Charing Cross Hospital, Fulham Palace Road, London W6 8RF, UK. E-mail: stevendallen{at}hotmail.co.uk
 |
Abstract
|
|---|
The purpose of this study was to assess the variability in the volume of the mesorectum in patients with rectal carcinoma. A retrospective review was made of pelvic MRI studies in 30 patients (mean age 64 years, range 34–88 years, 18M:12F) with histologically proven rectal carcinomas that were confined to the mesorectum. The outer low signal margin of the mesorectum was traced, over at least 10 consecutive 10 mm contiguous slices, until its disappearance. The visceral fascial compartment, body cross-sectional area and body mass index were measured, on a solitary slice, at the level of the L5/S1 disc. Linear regression was calculated for independent determinants of the mesorectal volume. Mean mesorectal volume in males was 227.5 cm3 (95%CI 191.6–263.4), and in females was 157.5 cm3 (95%CI 129.3–185.7). The difference in mesorectal volume between men and women was statistically significant (p<0.001). Mean visceral compartment area in males was 18.4 cm2 (95%CI 16.3–20.5) and in females was 14.6 cm2 (95%CI 12.8–16.4). Visceral compartment area correlated with mesorectal area and volume in females (p<0.05), and extremely well in males (p<0.005). Body cross-sectional area, body mass index and age did not correlate with mesorectal size. The correlation of visceral compartment area with mesorectal volume and mean area suggests that the mesorectum is determined in a similar way to other body fat compartments, with a similar anatomical variation. This significant variation in size and volume may have an important prognostic implication in patients with rectal carcinoma. This volume can be measured and reported on pre-operative MRI scanning and may be communicated to surgeons and radiotherapists.
|
Introduction
|
|---|
Total mesorectal excision (TME) surgery is now the standard surgical technique for rectal cancer and has reduced local recurrence rates considerably, due to minimizing the risk of leaving an involved circumferential resection margin (CRM) [1–4]. With TME surgery, the entire mesorectal compartment (or mesorectum) is removed intact. This consists of the rectum, the surrounding perirectal fat (or mesorectal fat), which contains lymph nodes, blood vessels and nerves, and a thin enveloping fascial layer termed the mesorectal fascia. The CRM in TME surgery is thus taken to be the mesorectal fascia.
Pre-operative staging with MRI has been optimized by the widespread use of phased-array pelvic coils [5–7]. The role of imaging has been increasingly focused on delineation of the relationship of tumour to the mesorectal fascia, and MRI has been shown to be the most accurate modality in this role [8]. This information has significant impact on the subsequent surgical procedure. If the pre-operative MRI predicts an involved CRM, then the patient may benefit from neoadjuvant therapy prior to TME surgery, in order to reduce the risk of a positive surgical margin [9].
Only one study to date has assessed the shape and volume of the mesorectum using MRI measurements, although not in relation to abdominal fat volume [10]. Another study has used MRI measurements of the dimensions of the bony pelvis to help predict resectability of rectal cancer [11]. No work to our knowledge has specifically assessed the axial size and volume of the mesorectum as an entity in patients with rectal carcinoma, with respect to abdominal fat. Axial T1 weighted MRI has been well validated as an accurate and reliable method of quantifying abdominal fat volumes [12–15]. This may have particular relevance in undernourished patients, with rectal carcinoma, who may subsequently be at a greater risk of CRM involvement with only relatively small tumours due to a small mesorectal compartment. The purpose of our study was to show that mesorectal axial size and volume varies with body habitus and sex, as assessed by measurements of body habitus and total visceral compartment area.
|
Methods and materials
|
|---|
Subjects
A retrospective review of 30 patients (mean age 64 years, range 34–88 years, 18 male and 12 female) with biopsy proven rectal carcinoma and adequate high-resolution pelvic MRI for local staging was undertaken in the study. Patients who had received previous pelvic radiotherapy or previous pelvic surgery were excluded. Patients with significant extramural disease (advanced T3 disease) or whose disease had spread locally beyond the mesorectum were also excluded (T4 disease), as it was felt that tumour morphology in these cases could have influenced mesorectal volume measurements. We did not adjust for rectal filling or state of evacuation as the axial boundaries of the mesorectum are relatively fixed structures in the pelvis. Mesorectal fat will be displaced superiorly with increased rectal filling, but with relatively little effect on the dimensions of the compartment as measured with our technique.
Our institutional review board does not require approval of retrospective reviews of patient data when conducted for audit purposes.
MRI technique
Pelvic MRI examinations were performed on a 1.5 T whole body imager (Magnetom avanto; Siemens Medical Systems, Erlangen, Germany) using an 8 channel body matrix coil and a pelvic phased array coil. Patients were investigated supine with both arms placed parallel to the body. Following initial localization images, examination of the pelvis and rectum was performed as follows. Imaging began by acquiring an axial whole pelvis spinecho T1 weighted and T2 weighted sequence from the L4/L5 disc to cover the whole pelvis. The imaging parameters for the T1 weighted sequence were repetition time (TR): 845 ms, echo time (TE): 10 ms, flip angle: 150°, echo, train length: 3, a slice thickness of 6 mm, field of view 350 cm and matrix size 256x512.
Smaller field of view T2 weighted sequences were then performed in coronal, sagittal and axial planes, but these images were not used for this analysis.
MRI interpretation
The axial whole pelvis spin echo T1 weighted sequence was selected for morphometric analysis. These axial images with 10 mm interval were used for quantitative measurements. The analysis was performed on a dedicated workstation (Centricity system; General Electric Healthcare, Chalfont St Giles, UK), with all measurements made to the nearest millimetre. All measurements were performed by consensus by two readers (DMB and SDA) with 7 years and 2 years of pelvic MRI experience, respectively.
Quantitative measurements
On each axial image a measurement of the whole mesorectal fascial area was performed by planimetry. The mesorectal fascia was defined on MRI as the fine linear structure enveloping the mesorectum, hypointense on the axial whole pelvis spin echo T1 weighted sequence [7]. Employing workstation specific software, this measurement was obtained using an electronic cursor and a mouse to draw a freehand region of interest (ROI) along the mesorectal fascia, which was located by eye (Figure 1
). The caudal aspect of the mesorectum, and hence the lowest slice where the mesorectal area was measured, was the first slice where fatty tissue was seen to surround the rectum. This was assessed to represent the first cranial slice above the anorectal junction. The most cranial part of the mesorectum, hence the highest slice where the mesorectal area was measured, was the junction between the rectum and sigmoid colon. This was assessed to represent the first slice where the low signal mesorectal fascia was not visualized. From the summation of these measured areas and multiplying by the slice thickness of 10 mm, a total mesorectal volume was derived. An average axial mesorectal area was also calculated.
View larger version (61K):
[in this window]
[in a new window]
|
Figure 1. Axial MR whole pelvis spin echoT1 weighted sequence images in a patient with rectal cancer. (a) The outer margin of the mesorectal fascia was defined on MRI as the fine hypointense linear structure (arrows) enveloping the mesorectum. (b) Measurement of the whole mesorectal fascial area was performed by planimetry. Employing workstation specific software, this measurement was obtained using an electronic cursor and a mouse to draw a freehand region of interest (ROI) along the mesorectal fascia, which was located by eye.
|
|
A measurement of body cross-sectional area was obtained using the electronic cursor and a mouse to draw a freehand ROI along the entire abdominal circumference (Figure 2). An axial T1 weighted image at the level of the L5/S1 disc space was selected for this measurement in each case. A measurement of the total visceral fat area was obtained using the same technique on the same axial image (Figure 2). The L5/S1 disc level was selected to assess visceral and body cross-sectional area as this was the highest bony landmark uniformly included on all of our standard pelvic MRI studies. While we accept that both the L4/5 disc level and the level of the umbilicus have most frequently been used in prior studies to assess abdominal fat on MRI, the lower L5 level has been validated as having accuracy in this regard. Han et al although concluding that a single measurement at L2/L3 has the closest accuracy in predicting abdominal fat, also stated that the lower L5 level correlated highly with no significant statistical difference [15].
View larger version (49K):
[in this window]
[in a new window]
|
Figure 2. On an axial MR whole pelvis spin echoT1 weighted sequence image at the level of the L5/S1 disc space, a measurement of total abdominal area was obtained using the electronic cursor and a mouse to draw a freehand region of interest (ROI). (a) This was performed along the entire abdominal circumference, and also (b) the visceral compartment circumference.
|
|
Body habitus was also assessed by performing a measurement of body mass index (BMI) for each patient. This was calculated by dividing patient weight (kg) by their height squared (m2).<</p>
Statistical analysis
For statistical analysis, Microsoft Office XP Excel was used. Total mesorectal volumes and mean axial mesorectal areas were assessed, with respect to sex, age, visceral compartment area and measurements of body habitus (body cross-sectional area and BMI). A linear regression was calculated for significant independent determinants of mesorectal area and volume. An analysis of covariance was performed to assess whether visceral compartment area and/or total body area had an effect on mesorectal area. Statistical significance was assigned at the 5% probability level. An unpaired t-test was also used to compare male and female patients, where data were normally distributed.
|
Results
|
|---|
A total of 326 mesorectal axial areas were measured. The mean volume of the mesorectum was 201.2 cm3 with a standard deviation of 74.2 cm3. The mean volume of the mesorectum in males was 227.5 cm3 (95%CI 191.6–263.4), and in females was 157.5 cm3 (95%CI 129.3–185.7). The difference in mesorectal volume between men and women was statistically significant (p<0.001). The sex variation in visceral compartment volume was also statistically significant (p = 0.016). Mean mesorectal axial area ranged from 10.3 cm2 to 43.4 cm2, with a mean of 19.7 cm2 (males 22.2 cm2 (95%CI 18.8–25.6), females 15.4 cm2 (95%CI 12.5–18.3)).
Mean visceral compartment area in males was 18.4 cm2 (95%CI 16.3–20.5) and in females was 14.6 cm2 (95%CI 12.8–16.4). Visceral compartment area correlated well with mesorectal area and volume in females (p<0.05), and extremely well in males (p<0.005) (Figures 3 and 4). On analysis of covariance, the relationship of visceral compartment area (both sexes) and mesorectal area was confirmed (p<0.001). This also showed that body cross-sectional area did not correlate with mesorectal area (p = 0.88) and nor did body mass index (p = 0.57). Age was also shown to have no correlation with mesorectal area (p = 0.26).
View larger version (6K):
[in this window]
[in a new window]
|
Figure 3. (a) Correlation between average mesorectal axial area and visceral compartment area in females (r = 0.65, p = 0.002). (b) Correlation between mesorectal volume and visceral compartment area in females (r = 0.65, p = 0.002).
|
|
View larger version (6K):
[in this window]
[in a new window]
|
Figure 4. (a) Correlation between average mesorectal axial area and visceral compartment area in males (r = 0.67, p = 0.018). (b) Correlation between mesorectal volume and visceral compartment area in males (r = 0.68, p = 0.016).
|
|
|
Discussion
|
|---|
Outcomes following surgery for rectal cancer have not really attended to the role of patient factors and mesorectal morphology. Locally advanced cancer is clearly predictive of a difficult surgical resection, but patient size and in particular mesorectal proportions have not really been considered. Methods of assessing whether surgery will be difficult are, to some extent, subjective. This assessment may vary from surgeon to surgeon, with none of the factors such as height, weight, body mass and build having been shown to be accurate in the assessment of the size of the mesorectum. A recent study has shown that in certain patients with rectal cancer, CRM positivity may be predicted from the pre-operative MRI pelvic measurements [11]. However, there are no reported studies assessing the pelvic or mesorectal size with regards to surgical outcome.
With standardization of surgical practice, there has also been a drive to increase the accuracy of assessment of tumour size, degree of extramural spread and relationship of the tumour to the CRM. We suggest that patient anatomical factors, in particular mesorectal volume and area, should also be considered. Our findings show there is a wide anatomical variation. Mesorectal area and volume was significantly larger in males, as was visceral compartment area. While the distance of the tumour to the CRM on MRI continues clearly to be the most important predictor of a possible positive surgical margin, there may still be occasions where there is debate about the need for pre-operative chemoradiation. Especially if the rectum is very distended, doubt about the likely CRM status may be such that the size of the mesorectal compartment could have an influence on this decision. Slater et al measured on MRI using insufflation, the relationship of rectal distension and mesorectal width, as an implication for the pre-operative prediction of a tumour-free CRM [16]. Rectal distension before MRI significantly reduced the distance between the rectal wall and mesorectal fascia. This also supported our contention of not adjusting mesorectal measurements for rectal filling, as the axial boundaries of the mesorectum do not distend, thus being relatively fixed structures in the pelvis.
Torkzad and Blomqvist retrospectively assessed the size and configuration of the mesorectum using MRI in 25 patients who were imaged for a variety of indications, but did not have rectal cancer or any other known rectal disease. They showed a great individual variation in the amount of mesorectal fat, and in the morphometric parameters between the two sexes. Males were shown to have significantly more fat posterior to the rectum than females. They also suggested that the mesorectal fascia was subject to impression by other nearby visceral organs, though this was not quantified [10]. They were comprehensive in their anatomical assessment of the mesorectum on MRI, though they did not look at the relationship of the mesorectal size to other aspects of the patient's habitus such as visceral compartment area, BMI and body cross-sectional area. Our study differed significantly as we assessed the variability of mesorectal size in patients with rectal cancer that was not locally advanced, and hence assessed this parameter in its clinical context. However, we corroborated their general findings of individual and sex variation in the amount of mesorectal fat.
In our study, mesorectal volume and mean area were considered as separate entities simply because mesorectal compartments varied slightly in craniocaudal length. Although axial area seems more clinically useful, we would only be presenting part of the anatomical information, and not fully quantifying the longer mesorectal compartments. As it happens, our results showed that there was considerably less variability in craniocaudal length, compared with axial size. This meant that mean mesorectal area and mesorectal volume similarly correlated with visceral compartment area.
It has long been known that there is sex variation in visceral compartment area, although the correlation of sizes of the visceral compartment and mesorectum have not been previously shown. This suggests that the mesorectum, like the visceral fat compartment is an anatomical compartment that to an extent is predetermined in its size. The clinical relevance of this is that it may also suggest that some patients with rectal cancer will therefore be more likely than others to have an involved or an endangered CRM for the exact same degree of extramural tumour spread (T3 disease), quite simply because they have an anatomically predetermined smaller mesorectum. For example, 5 mm of extramural disease in the same anatomical position in two different patients could mean either a straightforward primary surgical resection (TME) with a highly likely clear surgical and pathological resection margin, or a likely involved surgical margin, and the patient undergoing a long course of pre-operative chemoradiotherapy in an attempt to downstage the tumour. The exactness of the estimation of mesorectal size per se is not particularly clinically relevant, rather the appreciation that it varies significantly and perhaps why it varies is more relevant.
A simpler workable method than we utilized to assess mean mesorectal size would be to perform a single measurement of axial area at a mid-rectal level (4–6 cm above anal verge), and a single measurement at a low rectal level (just above the anal verge). The mean of these two measurements would approximate the overall mean mesorectal area, and this value could be compared with our study mean and range, to assess relative size (males = 22.2 cm2 (95%CI 18.8–25.6), females = 15.4 cm2 (95%CI 12.5–18.3)). Mesorectal area in the upper rectum is less relevant, as patients are less often considered for pre-operative chemoradiation with tumours at this site, even if the outer mesorectum is potentially involved, as there is often surrounding presacral fascia or pelvic peritoneal fat that will allow a clear CRM.
The lack of correlation of mesorectal area and volume with body cross-sectional area and BMI suggests that body habitus does not influence mesorectal size. Although a study with two separate MR examinations at a suitable time interval would be required to confirm this, we have shown a lack of relationship of these sizes in a single "snapshot". However, this is an interesting finding, as it may further support the suggestion that the mesorectum is largely anatomically predetermined, and not influenced by patient body habitus. If this is the case, it seems that the variability in mesorectal size which will determine whether a patient will have a clear, an endangered or an involved CRM for the exact same degree of extramural tumour spread (T3 disease), is more "genetic" than "environmental". It could also suggest that patients with an endangered margin would not benefit from an intensive high calorie diet pre-operatively in an attempt to "boost" their mesorectal size. The likely anaesthetic detriment of the extra weight would render this significantly disadvantageous.
We did not find that older patients had smaller mesorectal compartments, so we did not show that there is age related compartment atrophy. Our patient age range of 34–88 years was broad enough to appropriately consider this.
Our study excluded patients with significant extramural disease (advanced T3 disease), or whose disease had spread locally beyond the mesorectum (T4 disease), as it was felt that tumour morphology in these cases could have influenced mesorectal volume measurements. A prospective study assessing the exact effect of tumour size on the size of the mesorectal compartment would examine the variation in mesorectal compartment size further, and determine whether this is an independent predictive factor for a positive surgical margin. Tumour size presumably would have a similar effect to insufflation quantity and degree of rectal distension in reducing the distance between the rectal wall and mesorectal fascia as demonstrated by Slater et al [16].
In conclusion, observed mesorectal size variability is not related to body habitus, but is related to the visceral compartment size, hence suggesting a significant genetic predisposition to an anatomically small or large mesorectum. This in turn implies that for a given degree of extramural tumour, in the same anatomical position, some patients will have a clear CRM, but others will have an endangered CRM, and due to the predisposition of compartment size, an increase in body habitus will not cause a beneficial increase in mesorectal size. Mesorectal volume can be easily determined from the current standard pre-operative MRI images, and can be included in discussion with surgeons and radiation oncologists during evaluation of pre-operative staging investigations.
No funding or financial assistance was applied for or utilized in the preparation of this paper.
Received for publication April 24, 2006.
Revision received July 8, 2006.
Accepted for publication August 15, 2006.
|
References
|
|---|
- Adam IJ, Mohamdee MO, Martin IG, et al. Role of circumferential margin involvement in the local recurrence of rectal cancer. Lancet 1994;344:707–11.[CrossRef][Medline]
- Quirke P, Durdey P, Dixon MF, Williams NS. Local recurrence of rectal adenocarcinoma due to inadequate surgical resection: histopathological study of lateral tumour spread and surgical excision. Lancet 1986;2:996–9.[CrossRef][Medline]
- Heald RJ, Ryall RDH. Recurrence and survival after total mesorectal excision for rectal cancer. Lancet 1986;1:1479–82.[CrossRef][Medline]
- Martling AL, Holm T, Rutqvist LE, Moran BJ, Heald RJ, Cedemark B. Effect of a surgical training programme on outcome of rectal cancer in the County of Stockholm. Stockholm Colorectal Cancer Study Group, Basingstoke Bowel Cancer Research Project. Lancet 2000;356:93–6.[CrossRef][Medline]
- Hadfield MB, Nicholson AA, MacDonald AW, et al. Preoperative staging of rectal carcinoma by magnetic resonance imaging with a pelvic phased array coil. Br J Surg 1997;84:529–31.[CrossRef][Medline]
- Brown G, Richards CJ, Newcombe RG, et al. Rectal carcinoma: thin-section MR imaging for staging in 28 patients. Radiology 1999;211:215–22.[Abstract/Free Full Text]
- Beets-Tan RG, Beets GL, Vliegen RF, et al. Accuracy of magnetic resonance imaging in prediction of tumour-free resection margin in rectal cancer surgery. Lancet 2001;357:497–504.[CrossRef][Medline]
- Beets-Tan RG, Beets GL. Rectal cancer: review with emphasis on MR imaging. Radiology 2004;232:335–46.[Abstract/Free Full Text]
- Kapetejin E, Marijnen CA, Nagtegaal ID, et al. Preoperative radiotherapy combined with total mesorectal excision for resectable rectal cancer. N Engl J Med 2001;345:638–46.[Abstract/Free Full Text]
- Torkzad MR, Blomqvist L. The mesorectum: morphometric assessment with magnetic resonance imaging. Eur Radiol 2005;15:1184–91.[CrossRef][Medline]
- Boyle KM, Petty D, Chalmers AG, et al. MRI assessment of the bony pelvis may help predict resectability of rectal cancer. Colorectal Dis 2005;7:232–40.[CrossRef][Medline]
- Poll LW, Wittsack HJ, Koch JA, et al. A rapid and reliable semiautomated method for measurement of total abdominal fat volumes using magnetic resonance imaging. Magn Reson Imaging 2003;21:631–6.[CrossRef][Medline]
- Eliakim A, Burke GS, Cooper DM. Fitness, fatness, and the effect of training assessed by magnetic resonance imaging and skinfold-thickness measurements in healthy adolescent females. Am J Clin Nutr 1997;66:223–31.[Abstract/Free Full Text]
- Gronemeyer SA, Steen RG, Kauffman WM, Reddick WE, Glass JO. Fast adipose tissue (FAT) assessment by MRI. Magn Reson Imaging 2000;18:815–18.[CrossRef][Medline]
- Han TS, Kelly IE, Walsh K, Greene RM, Lean ME. Relationship between volumes and areas from single transverse scans of intra-abdominal fat measured by magnetic resonance imaging. Int J Obes Relat Metab Disord 1997;21:1161–6.[CrossRef][Medline]
- Slater A, Halligan S, Taylor SA, Marshall M. Distance between the rectal wall and mesorectal fascia measured by MRI: effect of rectal distension and implications for preoperative prediction of a tumour-free circumferential resection margin. Clin Radiol 2006;61:65–70.[CrossRef][Medline]
Top
Abstract
Introduction
