- Email: [email protected]
Perioperative Colonic Evaluation in Patients with Rectal Cancer; MR Colonography Versus Standard Care
Perioperative Colonic Evaluation in Patients with Rectal Cancer; MR Colonography Versus Standard Care Michael Patrick Achiam, PhD, MD, Vibeke Løgager, MD, Vera Lund Rasmussen, MD, Cecilie Okholm, Talie Mollerup, MD, Henrik S. Thomsen, MD, DMSc, Jacob Rosenberg, MD, DMSc Abbreviations CC conventional colonoscopy MRC magnetic resonance colonography PCE perioperative clean-colon evaluation
Rationale and Objectives: Preoperative colonic evaluation is often inadequate because of cancer stenosis making a full conventional colonoscopy (CC) impossible. In several studies, cancer stenosis has been shown in up to 16%–34% of patients with colorectal cancer. The purpose of this study was to prospectively evaluate the completion rate of preoperative colonic evaluation and the quality of perioperative colonic evaluation using magnetic resonance colonography (MRC) in patients with rectal cancer. Materials and Methods: Patients diagnosed with rectal cancer were randomized to either group A: standard preoperative diagnostic work-up or group B: preoperative MR diagnostic work-up (standard preoperative diagnostic work-up + MRC). A complete and adequate perioperative clean-colon evaluation (PCE) was defined as either a complete preoperative colonic evaluation or a complete colonic evaluation within 3 months postoperatively. Results: Twenty-eight patients were randomized to group A and 28 to group B. Complete preoperative colonic evaluation with CC was achieved in 39% patients in group A and 93% for group B (Fisher’s exact test, P < .001). PCE with CC was achieved in 64% and 93% in groups A and B, respectively (Fisher’s exact test, P = .02). In group A, one synchronous cancer was found by CC. However, the location was misjudged as a sigmoid cancer. In group B, two synchronous cancers were found in the same patient who had an insufficient preoperative CC due to an obstructing rectal cancer. Conclusions: MRC is a valuable tool and is recommended as part of the standard preoperative evaluation for patients with rectal cancer. Key Words: MR colonography; colonic evaluation; preoperative diagnostic work-up; synchronous; rectal cancer. ªAUR, 2015
INTRODUCTION
C
olorectal cancer is the third most common cancer in men and the second in women worldwide. At this time, it is estimated that 1.361 million new cases and 694 million deaths are caused every year by colorectal cancer, and most patients die from synchronous and metachronous lesions (1). Thus, a thorough preoperative (PREOP) diagnostic work-up is necessary. It is often
Acad Radiol 2015; 22:1522–1528 From the Department of Surgical Gastroenterology, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark (M.P.A., C.O.); Department of Diagnostic Radiology, Herlev Hospital, University of Copenhagen, Herlev, Denmark (V.L., V.L.R., H.S.T.); and Department of Surgical Gastroenterology, Herlev Hospital, University of Copenhagen, Herlev, Denmark (T.M., J.R.). Received February 4, 2015; accepted August 23, 2015. Address correspondence to: M.P.A. e-mail: michael.patrick. [email protected] ªAUR, 2015 http://dx.doi.org/10.1016/j.acra.2015.08.019
1522
comprised of a computed tomography (CT) of the thorax and abdomen along with colonic evaluation. In case of rectum cancer, an evaluation of the tumor by transrectal ultrasound or magnetic resonance (MR) is also performed. The PREOP diagnostic work-up is performed to find synchronous lesions; liver metastases are found in up to 19% of patients (2), colon cancer in 2%–11% and up to 61% of patients with polyps (3). The diagnosis of synchronous liver metastases is primarily used for determination of resectability, whereas the diagnosis of synchronous colonic lesions is for planning of surgical strategy. The presence of synchronous colonic lesions in patients with colon cancer was described already in 1922 (4); more than 30 years ago, it was recommended to do PREOP colonic evaluation (3,5). For the time being, it recommended that the colonic evaluation is done by conventional colonoscopy (CC) because of the high sensitivity, direct visualization of the mucosa, and therapeutic ability for resecting polyps. However, inadequate colonic evaluation is often seen because of lack of resources, acute surgery, poor bowel
Academic Radiology, Vol 22, No 12, December 2015
preparation, adhesions, pain, and most importantly, cancer stenosis making a full CC impossible. In several studies, cancer stenosis has been shown in up to 16%–34% of patients with colorectal cancer (6,7). The purpose of this study was to prospectively evaluate the completion rate of PREOP colonic evaluation. Furthermore, to evaluate the quality of perioperative clean-colonic evaluation (PCE) using either standard perioperative colonic evaluation versus magnetic resonance colonography (MRC) in patients with rectal cancer.
MATERIALS AND METHODS This study was approved by the local ethics committee for medical research, and written informed consent was obtained from all patients. Patients
From October 2010 to February 2013, patients diagnosed with rectal cancer were consecutively invited to enroll in the study at the first visit to the outpatient clinic. Patients were selected based on the referral papers among those with biopsy-proven cancer. The exclusion criteria for participation in the trial were 1) withdrawal of consent, 2) contraindications to MR, 3) benign tumor in the rectum, 4) nonrectal colonic cancer, 5) liver metastases, 6) inoperable cancer, 7) patients aged younger than 18 years, and 8) patients not eligible for surgery. The patients were randomized to either group A: standard PREOP diagnostic work-up (CT of thorax/ abdomen, MR of rectum, transrectal ultrasound, and CC) or group B: PREOP MR diagnostic work-up (CT of thorax/abdomen, MR of rectum, MRC, transrectal ultrasound, and CC). Randomization was performed using the Web site www.random.org. A computer-generated randomization schedule using fixed size blocks assigned patients to either group A or group B using a 1:1 ratio. Sample size was calculated to be 50 patients in a group to ensure at least one synchronous cancer in each group with a 95% likelihood, assuming that the prevalence of synchronous cancer is 6%. To further strengthen the study, the number was increased to 75 patients in each group. Patient Bowel Preparation
Patients undergoing MRC and CC underwent standard bowel purgation (Moviprep, Norgine, Denmark) 1 day before surgery along with a fiber-restricted liquid diet.
PERIOPERATIVE EVALUATION WITH MR COLONOGRAPHY
enema tube was placed. After this, the colon was filled by approximately 2 L of warm tap water using hydrostatic pressure (1.5-m water column). The distension of the colon continued until all the water had been administered or the patient halted the filling process because of the increasing abdominal pressure. MRC was performed using a Philips Achieva 1.5-T scanner (Philips Medical Systems, Best, the Netherlands). Examination started with a fluid-weighted sequence (turbo field echo) to obtain overview of the static fluid in the abdomen and as a control for inflammatory bowel disease. The parameters were 3.5 ms, TE = 1.76 ms, FOV = 450 mm, TFE factor = 110, FA = 90, matrix = 236 214, and slice thickness = 5 mm. Afterward, the ‘‘dark lumen MRC’’ method with T1-weighted sequence was applied. Parameters for the T1-weighted sequences were optimized to achieve the thinnest slices with best resolution. The T1-weighted sequence was applied before and after intravenous contrast administration. Gadoterate meglumine intravenous contrast agent 0.2 mL/kg (Dotarem 279.3 mg/ mL; Guerbet, Roissy CDG, France) was injected at a flow rate of 2 mL/sec followed by 10 mL of isotonic saline using a powerinjector (MedRad, Warrendale, PA). The postcontrast scanning took place approximately 75 seconds after the contrast injection. Data Analyses
All MR data sets were transferred to a workstation (View-forum; Phillips Medical Systems). All MRC images were anonymized and randomized by a computer before being blindly evaluated by two readers. Reader I was a board-certified abdominal radiologist with more than 15 years of experience in MR abdominal imaging including reading more than 200 MRCs. Reader II was a gastrointestinal surgeon who had experience from reading more than 200 MRCs. MR data sets were analyzed by scrolling through coronal planes with multiplanar reconstruction for problem solving, which permitted scrolling through all three orthogonal planes. The colon was evaluated by starting from rectum moving through to the cecum and then back again. The colon was divided into six segments: cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum. Lesions were recorded according to morphology, size, and location. All colon segments were rated for distention (1, optimal; 2, adequate; and 3, inadequate) and artifacts (1, none; 2, some, but suitable for diagnostic evaluation; 3, too many artifacts, unsuitable for diagnostic evaluation). Conventional Colonoscopy
MR Imaging
To minimize bowel peristalsis and colonic spasm, the patients were given butylscopolamine intravenously (40 mg of Buscopan; Boehringer, Ingelheim, Germany) before the MRC. The patients were placed in the prone position, and a rectal
CC was considered as the gold standard for colonic evaluation and was either performed preoperatively or postoperatively. After the previously mentioned bowel purgation, CC was performed using standard equipment (Olympus Endoscope Type CFQ160 DL, Tokyo, Japan). 1523
ACHIAM ET AL
The performing gastroenterologist was unaware of the MR findings. All patients were offered sedation with midazolam (Dormicum; F. Hoffmann-LaRoche Ltd, Basel, Switzerland) and fentanyl (Janssen-Cilag, Birkerød, Denmark) or propofol. Colorectal lesions were noted on basis of their size and position (the gastroenterologist judging each colonic segment using a magnet for localization). All CC procedures and findings were recorded and compared to MRC. Synchronous cancer was defined as a secondary colon cancer diagnosed preoperatively or within 6 months postoperatively. For those patients who did not have a complete PREOP or postoperative CC within 6 months, lesions reported on MRC were compared to subsequent findings from CC or operations at any time in the course of the follow-up period. A lesion on MRC was classified as true positive if two criteria were fulfilled: 1) lesions reported on MRC were within a 50% margin of size of the lesion reported on CC or in the resected specimen and 2) lesions were within the same or adjacent segment. Complete Colonic Evaluation
A complete and adequate PCE was defined as either a complete PREOP colonic evaluation or a complete colonic evaluation within 3 months postoperatively. Furthermore, PCE could be performed with either a complete CC from rectum to the cecum or a complete MRC evaluation. An MRC evaluation was considered complete with a maximum grade of 2 in both the distention and the artifact grading from rectum to cecum. Furthermore, both readers should have scored grade 3 in the same patient and category (artisfacts or distention) for an MRC to be considered incomplete. RESULTS Patient Population
In a period of 28 months, 75 patients were enrolled in the trial, 37 in group A, and 38 in group B. The trial was ended prematurely in February 2013 because of lack of MR scanning capacity. Nineteen patients were later excluded, nine in group A, 10 in group B (Fig. 1). Of the remaining 56 patients, 28 were randomized to group A (18 male and 10 female) and 28 to group B (22 male and six female). The median age was 65 years (range, 52–87 years) in group A and 64 years (range, 45–85 years) in group B. Evaluation
Colon Evaluation. Complete PREOP colonic evaluation with CC was achieved in 11 of 28 patients (39%) in group A. In group B, using only CC without the benefit of MRC, colonic evaluation was achieved in 10 of 28 patients (36%; Fisher’s exact test, P = 1.0). However, considering a complete MRC as a complete PREOP colonic evaluation, the completion rate for group B was 93% (26 of 28 patients) and signifi1524
Academic Radiology, Vol 22, No 12, December 2015
cantly higher than in group A (Fisher’s exact test, P < .001). The two incomplete MRCs were caused by a large obstructing tumor preventing insertion of the rectal probe in one patient and by grade 3 artifacts evaluation by both readers in the other patient. The various PREOP colonic evaluations were performed according to Table 1. PCE with CC was achieved in 64% (18 of 28 patients) in group A, leaving 36% of the patients without a PCE. In group B, using only CC without the benefit of MRC, 59% (16 of 27 patients, one patient did not have CC due to recurrence) underwent PCE (Fisher’s exact test, P = .785). However, with the benefit of MRC, PCE was achieved in 93% (Fisher’s exact test, P = .02) as shown in Figure 2. Of the 10 patients in group A, who did not have a PCE, four patients later underwent CC without pathologic findings (median, 5.5 months after surgery; range, 4–9 months). Of the 11 patients in group B, who did not have a complete PREOP colonic evaluation with CC, five patients later underwent CC with no pathologic findings (median, 12 months after surgery; range, 9–17). MR Colonography. MRC images were evaluated at a median time of 8 minutes and 4 minutes by reader I and II, respectively. Artifacts were rated a median of 1 with a range of 1–3. Grade 3 artifacts were found in three patients by each reader, but in only one patient 1 of 28 (4%) by both readers, subsequently categorizing the MRC as incomplete. Colonic distention was rated at a median of 1 by both readers. Both readers scored it between 1 and 2 except in the sigmoid colon, where the range was 1–2 and 1–3 for reader I and II, respectively. Grade 3 distention was found only in two patients and only by reader II. No patients had grade 3 distention by both readers. Reader I reported four possible cancers (6 mm, 20 mm flat tumor, 30 mm, and 30 mm) in three patients; two cancers in the same patient were true positive findings and one false positive findings in two patients. Reader II reported two possible cancers (40 mm and 45 mm) in the same patient, both true positive findings. None of the synchronous cancers were missed. Thus, the per-patient sensitivity and/or specificity for synchronous cancers was 100% of 89% and 100% of 100% for reader I and II, respectively, when comparing with the 19 patients in group B who had a complete CC or operations at any time in the course of the follow-up period. Colonic Findings. Three synchronous cancers were found in the 56 patients (5%). In group A, one synchronous cancer was found by CC. However, the location was misjudged as a sigmoid cancer. Intraoperatively, the synchronous tumor was located in the midtransverse colon. No other synchronous cancers were found, neither preoperatively nor postoperatively in that group. In group B, two synchronous cancers were found in the same patient by MRC (see Fig. 3). The same patient had an insufficient PREOP CC due to an obstructing rectal cancer. No other synchronous cancers were found on MRC or on the CC, neither preoperatively nor postoperatively in group B.
Academic Radiology, Vol 22, No 12, December 2015
PERIOPERATIVE EVALUATION WITH MR COLONOGRAPHY
Pa ents randomized to standard preopera ve diagnos c evalua on Group-A (n =37)
Included
Pa ents randomized to preopera ve MR diagnos c evalua on Group-B (n =38)
Pa ents excluded due to: 6 not undergoing opera on 3 due to benign tumor 3, due to liver metastasis 2 withdrawal of consent 1 due to MR missing capacity 1 due to sigmoid cancer 1 due to preopera ve death 1 due to CT technical difficul es 1 due to MR technical difficul es (n = 19)
Exclusion
RandomizaƟon
SelecƟon
Pa ents included (n = 75)
Pa ents included Group-A (n = 28)
Pa ents included Group-B (n = 28)
Figure 1. Flowchart of patients included. CT, computed tomography; MR, magnetic resonance. (Color version of figure is available online.)
TABLE 1. Modality Used for Preoperative Colonic Evaluation
Group A Rectoscopy Sigmoidoscopy Colonoscopy Rectocopy due to stricture Sigmoidoscopy due to stricture Total Group B Rectoscopy Sigmoidoscopy Colonoscopy Rectoscopy due to stricture Sigmoidoscopy due to stricture Total
Frequency
Percent
1 12 11 2 3 28
3.6 42.9 39.3 7.1 10.7 100.0
2 10 10 3 3 28
7.1 35.7 35.7 10.7 10.7 100.0
DISCUSSION In the present study, only 39% of the patients in group A had a full PREOP colonic evaluation with CC compared to 93% of
the patients in group B who had a full PREOP colonic evaluation with MRC. Moreover, MRC found two synchronous cancers proximal to an obstructing rectal cancer, which had prevented further colonic evaluation with CC in the patient. Full PREOP colonic evaluation has been proposed several times (5,8). The guidelines from our national surgical society recommends a full PREOP colonic evaluation or if incomplete, a full colonic evaluation within 3 months postoperatively. This is primarily due to synchronous cancers, which have been shown in the two largest retrospective studies to have a prevalence of 2% and 7% (9,10). The rationale behind the PCE guidelines is a quick postoperative diagnose and treatment for those synchronous cancers that were missed due to incomplete PREOP colonic evaluation to prevent disseminated disease. Only few prospective studies exist regarding synchronous cancers, and these are all smaller studies (11,12). The importance of PREOP colonic evaluation and subsequent treatment and/ or procedures was shown in a study reporting that patients who had PREOP CC had fewer local recurrences, fewer distant metastases, and a longer survival time (13). This was preceded by another study showing fewer metachronous colon cancers in patients who had PREOP CC (14). In the 1525
ACHIAM ET AL
Academic Radiology, Vol 22, No 12, December 2015
PreoperaƟve colonic evaluaƟon
Included
Pa ents included Group-A (n = 28)
Conven onal colonoscopy 39% (n = 11/28)
<3 month perioperaƟve colonic evaluaƟon >3 month perioperaƟve colonic evaluaƟon
Yes
Conven onal colonoscopy (n = 4)
n=1
No
Not evaluated 36% (n = 10/28)
No
n=2
Not evaluated 7% (n = 2/28)
MR colonography 93% (n = 26/28)
Not evaluated 61% (n = 17/28)
Yes
Conven onal colonoscopy 64% (n = 18/28)
Pa ents included Group-B (n = 28)
Not evaluated due to recurrence (n = 3)
Not evaluated due to 1 due to benign stricture 3 not referred
No
MR colonography 93% (n = 25/27)
No
Not evaluated due to: 1 due to comorbidity
Not evaluated 7% (n = 2/27)
Yes
Conven onal colonoscopy (n = 1)
Figure 2. Flowchart of perioperative colonic evaluation of the patients included. MR, magnetic resonance. (Color version of figure is available online.)
present study, a complete PREOP colonic evaluation was done in 39% of the patients in group A and in 93% of the patients in group B with MRC (P < .001). Without MRC, the PREOP colonic evaluation rate would have been 36% in group B. Furthermore, only 41% (7 of 17 patients) of the patients, who did not receive a full PREOP colonic evaluation in group A, had PCE within 3 months. In group B, the PCE rate remained 93% with MRC. The literature on the subject of complete preoperative or postoperative evaluation within 3 months remains scarce, and to our knowledge, only one study has published results in this area. It involved 534 patients undergoing operation for colorectal cancer; 22% had a full PREOP colonic evaluation (9). Postoperatively, only 62% of the patients who received inadequate PREOP colonic evaluation had a PCE, a total of 71% of all the patients. Although the percentage of completed PREOP colonic evaluations is higher in this prospective study for group A (39%), the overall percentage of complete PCE within 3 months is lower in group A (64%). As shown in Table 1, an incomplete PREOP 1526
colonic evaluation due to tumor stenosis was recorded in 18% and 21% of the patients in group A and group B, respectively. Impassable tumor stenosis was present in 23% in the retrospective study (9). These rates are in line with the literature where obstructing tumors are reported to occur in as many as 34% of patients (6). Other studies have reported incomplete PREOP colonic evaluation with CC in as many as 45%–58% of patients with colorectal cancer (8,15); most were because of obstructing tumors. Although CC remains the gold standard of colonic evaluation at this time, the method may not be the most optimal. Besides the high rate of incomplete procedures and the serious complications (perforation, bleeding, and mortality), the procedure have been shown to misjudge tumor localization preoperatively in up to 11%–14% (16,17) in accordance with the findings in the present study. In another study, 6.5% of colorectal surgeons have reported making the choice of a wrong colonic segment for laparoscopic colectomy requiring conversion to standard laparotomy and resection (18).
Academic Radiology, Vol 22, No 12, December 2015
PERIOPERATIVE EVALUATION WITH MR COLONOGRAPHY
Figure 3. (a) ‘‘Dark lumen’’ MR colonography in patient with occlusive rectal cancer. The images are T1-weighted MRI after intravenous contrast administration. White arrow: synchronous cancer located orally from the occlusive rectal cancer. (b) Fluid-weighted sequence (turbo field echo) of the same synchronous cancer. MRI, magnetic resonance imaging.
Both CT and MR colonography offers an alternative to PREOP colonic evaluation with high completion rates. In the present study, completion rate was 93%; in another study, 98% of the patients who were not offered CC or had an incomplete CC had a full colonic examination by MRC (19). Similarly, several studies have shown feasibility and high completion rates with CT colonography in patients with stricturing tumors. The completion rates have been shown to be 90%–97% (12,20). MR of the rectum for PREOP staging is currently often a part of the standard work-up for patients with rectal cancer. Thus, it would be advantageous to use the opportunity of MRC, when the patient is in the MR scanner anyway. The specificity for lesions >1 cm with MRC is high. A recent systematic review found a per-patient sensitivity for lesions of 88% (95% confidence interval [CI], 63%–97%) and a specificity of 99% (95% CI, 95%–100%) (21). In the same review, the sensitivity of MRC in detecting CRC was 100%. The present study used water for distention and dark lumen MRC technique, but other techniques have also been developed. Recently, MRC with automated CO2 insufflation has found a per-patient sensitivity for lesions >10 mm to be 91.7% and specificity to be 96.5% for an experienced reader (22). As shown in the present study, the time for evaluation was between 4–8 minutes median for experienced readers and the sensitivity was 100%, making MRC a time-efficient and effective examination. The aim of the present study was not to evaluate the sensitivity and specificity of MRC regarding all lesions, but primarily to detect synchronous cancers. An economical study in progress is aimed at determining whether the costs of an additional MRC preoperatively in patients with rectal cancers undergoing MR of the rectum is financially sound. However, PREOP MRC may be an important supplement if not replacement for the PREOP CC or the PCE in patients with rectal cancer.
One of the limitations of this study is the small number of patients included in the study. This was primarily because of a recruiting problem in the project group, which was not able to attend the outpatients facility every day. Thus, many days passed with no recruiting. The other reason for the small number of patients included was the lack of MR scanning capacity, which was announced on short notice. However, the main purpose of the study was to evaluate the completion rate of PREOP colonic evaluation and the quality of perioperative colonic evaluation, which was achieved. The power calculation was made to ensure at least one synchronous cancer in each group to emphasize the importance of PREOP colonic evaluation. The fact, that synchronous cancers were found in each group anyway is fortunate as it underlines the limitations of the existing standard colonic evaluation compared to MRC. A strength of the present study is, however, that both readers were experienced readers of MRC. MRC is a procedure that requires dedicated personnel. A weakness of the study was the inability to evaluate all MRCs preoperatively, which would have given the surgeons the possibility of an intraoperative colonic evaluation in case of positive findings. The retrospective nature of the evaluation also led to the decision not to include polyps. The sensitivity and specificity polyp detection are well described, and smaller polyps (<10 mm) are rarely malignant lesions. However, most polyps need to be removed or followed up because more than 90% of colorectal cancers are believed to arise from smaller adenomas (23). Even so, the main focus of this article was the perioperative colonic evaluation and thus the consequences missed synchronous cancers. The fact that not all patients underwent postoperative colonic evaluation, just as 10 of 19 patients were excluded from the study due to disseminated disease or comorbidity, may have led to an overestimation of the sensitivity of the MRC and underestimation of overlooked synchronous cancers. However, the study reflects the daily clinical life in a Danish colorectal surgical department 1527
ACHIAM ET AL
and may contribute to estimating the benefits of implementing MRC in PREOP evaluation. Furthermore, another strength of the study is the randomized prospective design and the blinded evaluation. In conclusion, MRC is a valuable tool for PREOP colonic evaluation in patients with rectal cancer. Furthermore, we have shown that both the preoperative and postoperative colonic evaluation has serious lacks due to logistical challenges, but also due to tumor stenosis which prevent complete colonic evaluation. Thus, we recommend the use of MRC as part of the standard PREOP evaluation for all patients with rectal cancer.
ACKNOWLEDGMENTS This study was supported by a grant by the Lundbeckfonden (grant number: R112-A9278). The funding played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the article for publication.
REFERENCES 1. International agency for research on cancer. GLOBOCAN 2012: Estimated cancer incidence, mortality and prevalence worldwide in 2012; 2014. Ref Type: Online Source. 2. Scheele J, Stangl R, Altendorf-Hofmann A. Hepatic metastases from colorectal carcinoma: impact of surgical resection on the natural history. Br J Surg 1990; 77(11):1241–1246. 3. Heald RJ, Bussey HJ. Clinical experiences at St. Mark’s Hospital with multiple synchronous cancers of the colon and rectum. Dis Colon Rectum 1975; 18(1):6–10. 4. Gordon-Watson C. Specimen showing carcinoma of the pelvic colon and rectum co-existing and causing acute obstruction. Proc R Soc Med 1922; 15(Surg Sect):92–93. 5. Maxfield RG. Colonoscopy as a routine preoperative procedure for carcinoma of the colon. Am J Surg 1984; 147(4):477–480.
1528
Academic Radiology, Vol 22, No 12, December 2015
6. Carraro PG, Segala M, Cesana BM, et al. Obstructing colonic cancer: failure and survival patterns over a ten-year follow-up after one-stage curative surgery. Dis Colon Rectum 2001; 44(2):243–250. 7. Phillips RK, Hittinger R, Fry JS, et al. Malignant large bowel obstruction. Br J Surg 1985; 72(4):296–302. 8. Isler JT, Brown PC, Lewis FG, et al. The role of preoperative colonoscopy in colorectal cancer. Dis Colon Rectum 1987; 30(6):435–439. 9. Achiam MP, Burgdorf SK, Wilhelmsen M, et al. Inadequate preoperative colonic evaluation for synchronous colorectal cancer. Scand J Surg 2009; 98(1):62–67. 10. Adloff M, Arnaud JP, Bergamaschi R, et al. Synchronous carcinoma of the colon and rectum: prognostic and therapeutic implications. Am J Surg 1989; 157(3):299–302. 11. Askew A, Ward M, Cowen A. The influence of colonoscopy on the operative management of colorectal cancer. Med J Aust 1986; 145(6):254–255. 12. Fenlon HM, McAneny DB, Nunes DP, et al. Occlusive colon carcinoma: virtual colonoscopy in the preoperative evaluation of the proximal colon. Radiology 1999; 210(2):423–428. 13. Howard ML, Greene FL. The effect of preoperative endoscopy on recurrence and survival following surgery for colorectal carcinoma. Am Surg 1990; 56(3):124–127. 14. Dasmahapatra KS, Lopyan K. Rationale for aggressive colonoscopy in patients with colorectal neoplasia. Arch Surg 1989; 124(1):63–66. 15. Tate JJ, Rawlinson J, Royle GT, et al. Pre-operative or postoperative colonic examination for synchronous lesions in colorectal cancer. Br J Surg 1988; 75(10):1016–1018. 16. Vignati P, Welch JP, Cohen JL. Endoscopic localization of colon cancers. Surg Endosc 1994; 8(9):1085–1087. 17. Cho YB, Lee WY, Yun HR, et al. Tumor localization for laparoscopic colorectal surgery. World J Surg 2007; 31(7):1491–1495. 18. Wexner SD, Cohen SM, Ulrich A, et al. Laparoscopic colorectal surgery–are we being honest with our patients? Dis Colon Rectum 1995; 38(7):723–727. 19. Achiam MP, Holst Andersen LP, Klein M, et al. Preoperative evaluation of synchronous colorectal cancer using MR colonography. Acad Radiol 2009; 16(7):790–797. 20. Morrin MM, Farrell RJ, Raptopoulos V, et al. Role of virtual computed tomographic colonography in patients with colorectal cancers and obstructing colorectal lesions. Dis Colon Rectum 2000; 43(3):303–311. 21. Zijta FM, Bipat S, Stoker J. Magnetic resonance (MR) colonography in the detection of colorectal lesions: a systematic review of prospective studies. Eur Radiol 2010; 20(5):1031–1046. 22. van der Paardt MP, Zijta FM, Boellaard TN, et al. Magnetic resonance colonography with automated carbon dioxide insufflation: diagnostic accuracy and distension. Eur J Radiol 2014; 83(5):743–750. 23. Leslie A, Carey FA, Pratt NR, et al. The colorectal adenoma-carcinoma sequence. Br J Surg 2002; 89(7):845–860.
Rationale and Objectives: Preoperative colonic evaluation is often inadequate because of cancer stenosis making a full conventional colonoscopy (CC) impossible. In several studies, cancer stenosis has been shown in up to 16%–34% of patients with colorectal cancer. The purpose of this study was to prospectively evaluate the completion rate of preoperative colonic evaluation and the quality of perioperative colonic evaluation using magnetic resonance colonography (MRC) in patients with rectal cancer. Materials and Methods: Patients diagnosed with rectal cancer were randomized to either group A: standard preoperative diagnostic work-up or group B: preoperative MR diagnostic work-up (standard preoperative diagnostic work-up + MRC). A complete and adequate perioperative clean-colon evaluation (PCE) was defined as either a complete preoperative colonic evaluation or a complete colonic evaluation within 3 months postoperatively. Results: Twenty-eight patients were randomized to group A and 28 to group B. Complete preoperative colonic evaluation with CC was achieved in 39% patients in group A and 93% for group B (Fisher’s exact test, P < .001). PCE with CC was achieved in 64% and 93% in groups A and B, respectively (Fisher’s exact test, P = .02). In group A, one synchronous cancer was found by CC. However, the location was misjudged as a sigmoid cancer. In group B, two synchronous cancers were found in the same patient who had an insufficient preoperative CC due to an obstructing rectal cancer. Conclusions: MRC is a valuable tool and is recommended as part of the standard preoperative evaluation for patients with rectal cancer. Key Words: MR colonography; colonic evaluation; preoperative diagnostic work-up; synchronous; rectal cancer. ªAUR, 2015
INTRODUCTION
C
olorectal cancer is the third most common cancer in men and the second in women worldwide. At this time, it is estimated that 1.361 million new cases and 694 million deaths are caused every year by colorectal cancer, and most patients die from synchronous and metachronous lesions (1). Thus, a thorough preoperative (PREOP) diagnostic work-up is necessary. It is often
Acad Radiol 2015; 22:1522–1528 From the Department of Surgical Gastroenterology, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark (M.P.A., C.O.); Department of Diagnostic Radiology, Herlev Hospital, University of Copenhagen, Herlev, Denmark (V.L., V.L.R., H.S.T.); and Department of Surgical Gastroenterology, Herlev Hospital, University of Copenhagen, Herlev, Denmark (T.M., J.R.). Received February 4, 2015; accepted August 23, 2015. Address correspondence to: M.P.A. e-mail: michael.patrick. [email protected] ªAUR, 2015 http://dx.doi.org/10.1016/j.acra.2015.08.019
1522
comprised of a computed tomography (CT) of the thorax and abdomen along with colonic evaluation. In case of rectum cancer, an evaluation of the tumor by transrectal ultrasound or magnetic resonance (MR) is also performed. The PREOP diagnostic work-up is performed to find synchronous lesions; liver metastases are found in up to 19% of patients (2), colon cancer in 2%–11% and up to 61% of patients with polyps (3). The diagnosis of synchronous liver metastases is primarily used for determination of resectability, whereas the diagnosis of synchronous colonic lesions is for planning of surgical strategy. The presence of synchronous colonic lesions in patients with colon cancer was described already in 1922 (4); more than 30 years ago, it was recommended to do PREOP colonic evaluation (3,5). For the time being, it recommended that the colonic evaluation is done by conventional colonoscopy (CC) because of the high sensitivity, direct visualization of the mucosa, and therapeutic ability for resecting polyps. However, inadequate colonic evaluation is often seen because of lack of resources, acute surgery, poor bowel
Academic Radiology, Vol 22, No 12, December 2015
preparation, adhesions, pain, and most importantly, cancer stenosis making a full CC impossible. In several studies, cancer stenosis has been shown in up to 16%–34% of patients with colorectal cancer (6,7). The purpose of this study was to prospectively evaluate the completion rate of PREOP colonic evaluation. Furthermore, to evaluate the quality of perioperative clean-colonic evaluation (PCE) using either standard perioperative colonic evaluation versus magnetic resonance colonography (MRC) in patients with rectal cancer.
MATERIALS AND METHODS This study was approved by the local ethics committee for medical research, and written informed consent was obtained from all patients. Patients
From October 2010 to February 2013, patients diagnosed with rectal cancer were consecutively invited to enroll in the study at the first visit to the outpatient clinic. Patients were selected based on the referral papers among those with biopsy-proven cancer. The exclusion criteria for participation in the trial were 1) withdrawal of consent, 2) contraindications to MR, 3) benign tumor in the rectum, 4) nonrectal colonic cancer, 5) liver metastases, 6) inoperable cancer, 7) patients aged younger than 18 years, and 8) patients not eligible for surgery. The patients were randomized to either group A: standard PREOP diagnostic work-up (CT of thorax/ abdomen, MR of rectum, transrectal ultrasound, and CC) or group B: PREOP MR diagnostic work-up (CT of thorax/abdomen, MR of rectum, MRC, transrectal ultrasound, and CC). Randomization was performed using the Web site www.random.org. A computer-generated randomization schedule using fixed size blocks assigned patients to either group A or group B using a 1:1 ratio. Sample size was calculated to be 50 patients in a group to ensure at least one synchronous cancer in each group with a 95% likelihood, assuming that the prevalence of synchronous cancer is 6%. To further strengthen the study, the number was increased to 75 patients in each group. Patient Bowel Preparation
Patients undergoing MRC and CC underwent standard bowel purgation (Moviprep, Norgine, Denmark) 1 day before surgery along with a fiber-restricted liquid diet.
PERIOPERATIVE EVALUATION WITH MR COLONOGRAPHY
enema tube was placed. After this, the colon was filled by approximately 2 L of warm tap water using hydrostatic pressure (1.5-m water column). The distension of the colon continued until all the water had been administered or the patient halted the filling process because of the increasing abdominal pressure. MRC was performed using a Philips Achieva 1.5-T scanner (Philips Medical Systems, Best, the Netherlands). Examination started with a fluid-weighted sequence (turbo field echo) to obtain overview of the static fluid in the abdomen and as a control for inflammatory bowel disease. The parameters were 3.5 ms, TE = 1.76 ms, FOV = 450 mm, TFE factor = 110, FA = 90, matrix = 236 214, and slice thickness = 5 mm. Afterward, the ‘‘dark lumen MRC’’ method with T1-weighted sequence was applied. Parameters for the T1-weighted sequences were optimized to achieve the thinnest slices with best resolution. The T1-weighted sequence was applied before and after intravenous contrast administration. Gadoterate meglumine intravenous contrast agent 0.2 mL/kg (Dotarem 279.3 mg/ mL; Guerbet, Roissy CDG, France) was injected at a flow rate of 2 mL/sec followed by 10 mL of isotonic saline using a powerinjector (MedRad, Warrendale, PA). The postcontrast scanning took place approximately 75 seconds after the contrast injection. Data Analyses
All MR data sets were transferred to a workstation (View-forum; Phillips Medical Systems). All MRC images were anonymized and randomized by a computer before being blindly evaluated by two readers. Reader I was a board-certified abdominal radiologist with more than 15 years of experience in MR abdominal imaging including reading more than 200 MRCs. Reader II was a gastrointestinal surgeon who had experience from reading more than 200 MRCs. MR data sets were analyzed by scrolling through coronal planes with multiplanar reconstruction for problem solving, which permitted scrolling through all three orthogonal planes. The colon was evaluated by starting from rectum moving through to the cecum and then back again. The colon was divided into six segments: cecum, ascending colon, transverse colon, descending colon, sigmoid colon, and rectum. Lesions were recorded according to morphology, size, and location. All colon segments were rated for distention (1, optimal; 2, adequate; and 3, inadequate) and artifacts (1, none; 2, some, but suitable for diagnostic evaluation; 3, too many artifacts, unsuitable for diagnostic evaluation). Conventional Colonoscopy
MR Imaging
To minimize bowel peristalsis and colonic spasm, the patients were given butylscopolamine intravenously (40 mg of Buscopan; Boehringer, Ingelheim, Germany) before the MRC. The patients were placed in the prone position, and a rectal
CC was considered as the gold standard for colonic evaluation and was either performed preoperatively or postoperatively. After the previously mentioned bowel purgation, CC was performed using standard equipment (Olympus Endoscope Type CFQ160 DL, Tokyo, Japan). 1523
ACHIAM ET AL
The performing gastroenterologist was unaware of the MR findings. All patients were offered sedation with midazolam (Dormicum; F. Hoffmann-LaRoche Ltd, Basel, Switzerland) and fentanyl (Janssen-Cilag, Birkerød, Denmark) or propofol. Colorectal lesions were noted on basis of their size and position (the gastroenterologist judging each colonic segment using a magnet for localization). All CC procedures and findings were recorded and compared to MRC. Synchronous cancer was defined as a secondary colon cancer diagnosed preoperatively or within 6 months postoperatively. For those patients who did not have a complete PREOP or postoperative CC within 6 months, lesions reported on MRC were compared to subsequent findings from CC or operations at any time in the course of the follow-up period. A lesion on MRC was classified as true positive if two criteria were fulfilled: 1) lesions reported on MRC were within a 50% margin of size of the lesion reported on CC or in the resected specimen and 2) lesions were within the same or adjacent segment. Complete Colonic Evaluation
A complete and adequate PCE was defined as either a complete PREOP colonic evaluation or a complete colonic evaluation within 3 months postoperatively. Furthermore, PCE could be performed with either a complete CC from rectum to the cecum or a complete MRC evaluation. An MRC evaluation was considered complete with a maximum grade of 2 in both the distention and the artifact grading from rectum to cecum. Furthermore, both readers should have scored grade 3 in the same patient and category (artisfacts or distention) for an MRC to be considered incomplete. RESULTS Patient Population
In a period of 28 months, 75 patients were enrolled in the trial, 37 in group A, and 38 in group B. The trial was ended prematurely in February 2013 because of lack of MR scanning capacity. Nineteen patients were later excluded, nine in group A, 10 in group B (Fig. 1). Of the remaining 56 patients, 28 were randomized to group A (18 male and 10 female) and 28 to group B (22 male and six female). The median age was 65 years (range, 52–87 years) in group A and 64 years (range, 45–85 years) in group B. Evaluation
Colon Evaluation. Complete PREOP colonic evaluation with CC was achieved in 11 of 28 patients (39%) in group A. In group B, using only CC without the benefit of MRC, colonic evaluation was achieved in 10 of 28 patients (36%; Fisher’s exact test, P = 1.0). However, considering a complete MRC as a complete PREOP colonic evaluation, the completion rate for group B was 93% (26 of 28 patients) and signifi1524
Academic Radiology, Vol 22, No 12, December 2015
cantly higher than in group A (Fisher’s exact test, P < .001). The two incomplete MRCs were caused by a large obstructing tumor preventing insertion of the rectal probe in one patient and by grade 3 artifacts evaluation by both readers in the other patient. The various PREOP colonic evaluations were performed according to Table 1. PCE with CC was achieved in 64% (18 of 28 patients) in group A, leaving 36% of the patients without a PCE. In group B, using only CC without the benefit of MRC, 59% (16 of 27 patients, one patient did not have CC due to recurrence) underwent PCE (Fisher’s exact test, P = .785). However, with the benefit of MRC, PCE was achieved in 93% (Fisher’s exact test, P = .02) as shown in Figure 2. Of the 10 patients in group A, who did not have a PCE, four patients later underwent CC without pathologic findings (median, 5.5 months after surgery; range, 4–9 months). Of the 11 patients in group B, who did not have a complete PREOP colonic evaluation with CC, five patients later underwent CC with no pathologic findings (median, 12 months after surgery; range, 9–17). MR Colonography. MRC images were evaluated at a median time of 8 minutes and 4 minutes by reader I and II, respectively. Artifacts were rated a median of 1 with a range of 1–3. Grade 3 artifacts were found in three patients by each reader, but in only one patient 1 of 28 (4%) by both readers, subsequently categorizing the MRC as incomplete. Colonic distention was rated at a median of 1 by both readers. Both readers scored it between 1 and 2 except in the sigmoid colon, where the range was 1–2 and 1–3 for reader I and II, respectively. Grade 3 distention was found only in two patients and only by reader II. No patients had grade 3 distention by both readers. Reader I reported four possible cancers (6 mm, 20 mm flat tumor, 30 mm, and 30 mm) in three patients; two cancers in the same patient were true positive findings and one false positive findings in two patients. Reader II reported two possible cancers (40 mm and 45 mm) in the same patient, both true positive findings. None of the synchronous cancers were missed. Thus, the per-patient sensitivity and/or specificity for synchronous cancers was 100% of 89% and 100% of 100% for reader I and II, respectively, when comparing with the 19 patients in group B who had a complete CC or operations at any time in the course of the follow-up period. Colonic Findings. Three synchronous cancers were found in the 56 patients (5%). In group A, one synchronous cancer was found by CC. However, the location was misjudged as a sigmoid cancer. Intraoperatively, the synchronous tumor was located in the midtransverse colon. No other synchronous cancers were found, neither preoperatively nor postoperatively in that group. In group B, two synchronous cancers were found in the same patient by MRC (see Fig. 3). The same patient had an insufficient PREOP CC due to an obstructing rectal cancer. No other synchronous cancers were found on MRC or on the CC, neither preoperatively nor postoperatively in group B.
Academic Radiology, Vol 22, No 12, December 2015
PERIOPERATIVE EVALUATION WITH MR COLONOGRAPHY
Pa ents randomized to standard preopera ve diagnos c evalua on Group-A (n =37)
Included
Pa ents randomized to preopera ve MR diagnos c evalua on Group-B (n =38)
Pa ents excluded due to: 6 not undergoing opera on 3 due to benign tumor 3, due to liver metastasis 2 withdrawal of consent 1 due to MR missing capacity 1 due to sigmoid cancer 1 due to preopera ve death 1 due to CT technical difficul es 1 due to MR technical difficul es (n = 19)
Exclusion
RandomizaƟon
SelecƟon
Pa ents included (n = 75)
Pa ents included Group-A (n = 28)
Pa ents included Group-B (n = 28)
Figure 1. Flowchart of patients included. CT, computed tomography; MR, magnetic resonance. (Color version of figure is available online.)
TABLE 1. Modality Used for Preoperative Colonic Evaluation
Group A Rectoscopy Sigmoidoscopy Colonoscopy Rectocopy due to stricture Sigmoidoscopy due to stricture Total Group B Rectoscopy Sigmoidoscopy Colonoscopy Rectoscopy due to stricture Sigmoidoscopy due to stricture Total
Frequency
Percent
1 12 11 2 3 28
3.6 42.9 39.3 7.1 10.7 100.0
2 10 10 3 3 28
7.1 35.7 35.7 10.7 10.7 100.0
DISCUSSION In the present study, only 39% of the patients in group A had a full PREOP colonic evaluation with CC compared to 93% of
the patients in group B who had a full PREOP colonic evaluation with MRC. Moreover, MRC found two synchronous cancers proximal to an obstructing rectal cancer, which had prevented further colonic evaluation with CC in the patient. Full PREOP colonic evaluation has been proposed several times (5,8). The guidelines from our national surgical society recommends a full PREOP colonic evaluation or if incomplete, a full colonic evaluation within 3 months postoperatively. This is primarily due to synchronous cancers, which have been shown in the two largest retrospective studies to have a prevalence of 2% and 7% (9,10). The rationale behind the PCE guidelines is a quick postoperative diagnose and treatment for those synchronous cancers that were missed due to incomplete PREOP colonic evaluation to prevent disseminated disease. Only few prospective studies exist regarding synchronous cancers, and these are all smaller studies (11,12). The importance of PREOP colonic evaluation and subsequent treatment and/ or procedures was shown in a study reporting that patients who had PREOP CC had fewer local recurrences, fewer distant metastases, and a longer survival time (13). This was preceded by another study showing fewer metachronous colon cancers in patients who had PREOP CC (14). In the 1525
ACHIAM ET AL
Academic Radiology, Vol 22, No 12, December 2015
PreoperaƟve colonic evaluaƟon
Included
Pa ents included Group-A (n = 28)
Conven onal colonoscopy 39% (n = 11/28)
<3 month perioperaƟve colonic evaluaƟon >3 month perioperaƟve colonic evaluaƟon
Yes
Conven onal colonoscopy (n = 4)
n=1
No
Not evaluated 36% (n = 10/28)
No
n=2
Not evaluated 7% (n = 2/28)
MR colonography 93% (n = 26/28)
Not evaluated 61% (n = 17/28)
Yes
Conven onal colonoscopy 64% (n = 18/28)
Pa ents included Group-B (n = 28)
Not evaluated due to recurrence (n = 3)
Not evaluated due to 1 due to benign stricture 3 not referred
No
MR colonography 93% (n = 25/27)
No
Not evaluated due to: 1 due to comorbidity
Not evaluated 7% (n = 2/27)
Yes
Conven onal colonoscopy (n = 1)
Figure 2. Flowchart of perioperative colonic evaluation of the patients included. MR, magnetic resonance. (Color version of figure is available online.)
present study, a complete PREOP colonic evaluation was done in 39% of the patients in group A and in 93% of the patients in group B with MRC (P < .001). Without MRC, the PREOP colonic evaluation rate would have been 36% in group B. Furthermore, only 41% (7 of 17 patients) of the patients, who did not receive a full PREOP colonic evaluation in group A, had PCE within 3 months. In group B, the PCE rate remained 93% with MRC. The literature on the subject of complete preoperative or postoperative evaluation within 3 months remains scarce, and to our knowledge, only one study has published results in this area. It involved 534 patients undergoing operation for colorectal cancer; 22% had a full PREOP colonic evaluation (9). Postoperatively, only 62% of the patients who received inadequate PREOP colonic evaluation had a PCE, a total of 71% of all the patients. Although the percentage of completed PREOP colonic evaluations is higher in this prospective study for group A (39%), the overall percentage of complete PCE within 3 months is lower in group A (64%). As shown in Table 1, an incomplete PREOP 1526
colonic evaluation due to tumor stenosis was recorded in 18% and 21% of the patients in group A and group B, respectively. Impassable tumor stenosis was present in 23% in the retrospective study (9). These rates are in line with the literature where obstructing tumors are reported to occur in as many as 34% of patients (6). Other studies have reported incomplete PREOP colonic evaluation with CC in as many as 45%–58% of patients with colorectal cancer (8,15); most were because of obstructing tumors. Although CC remains the gold standard of colonic evaluation at this time, the method may not be the most optimal. Besides the high rate of incomplete procedures and the serious complications (perforation, bleeding, and mortality), the procedure have been shown to misjudge tumor localization preoperatively in up to 11%–14% (16,17) in accordance with the findings in the present study. In another study, 6.5% of colorectal surgeons have reported making the choice of a wrong colonic segment for laparoscopic colectomy requiring conversion to standard laparotomy and resection (18).
Academic Radiology, Vol 22, No 12, December 2015
PERIOPERATIVE EVALUATION WITH MR COLONOGRAPHY
Figure 3. (a) ‘‘Dark lumen’’ MR colonography in patient with occlusive rectal cancer. The images are T1-weighted MRI after intravenous contrast administration. White arrow: synchronous cancer located orally from the occlusive rectal cancer. (b) Fluid-weighted sequence (turbo field echo) of the same synchronous cancer. MRI, magnetic resonance imaging.
Both CT and MR colonography offers an alternative to PREOP colonic evaluation with high completion rates. In the present study, completion rate was 93%; in another study, 98% of the patients who were not offered CC or had an incomplete CC had a full colonic examination by MRC (19). Similarly, several studies have shown feasibility and high completion rates with CT colonography in patients with stricturing tumors. The completion rates have been shown to be 90%–97% (12,20). MR of the rectum for PREOP staging is currently often a part of the standard work-up for patients with rectal cancer. Thus, it would be advantageous to use the opportunity of MRC, when the patient is in the MR scanner anyway. The specificity for lesions >1 cm with MRC is high. A recent systematic review found a per-patient sensitivity for lesions of 88% (95% confidence interval [CI], 63%–97%) and a specificity of 99% (95% CI, 95%–100%) (21). In the same review, the sensitivity of MRC in detecting CRC was 100%. The present study used water for distention and dark lumen MRC technique, but other techniques have also been developed. Recently, MRC with automated CO2 insufflation has found a per-patient sensitivity for lesions >10 mm to be 91.7% and specificity to be 96.5% for an experienced reader (22). As shown in the present study, the time for evaluation was between 4–8 minutes median for experienced readers and the sensitivity was 100%, making MRC a time-efficient and effective examination. The aim of the present study was not to evaluate the sensitivity and specificity of MRC regarding all lesions, but primarily to detect synchronous cancers. An economical study in progress is aimed at determining whether the costs of an additional MRC preoperatively in patients with rectal cancers undergoing MR of the rectum is financially sound. However, PREOP MRC may be an important supplement if not replacement for the PREOP CC or the PCE in patients with rectal cancer.
One of the limitations of this study is the small number of patients included in the study. This was primarily because of a recruiting problem in the project group, which was not able to attend the outpatients facility every day. Thus, many days passed with no recruiting. The other reason for the small number of patients included was the lack of MR scanning capacity, which was announced on short notice. However, the main purpose of the study was to evaluate the completion rate of PREOP colonic evaluation and the quality of perioperative colonic evaluation, which was achieved. The power calculation was made to ensure at least one synchronous cancer in each group to emphasize the importance of PREOP colonic evaluation. The fact, that synchronous cancers were found in each group anyway is fortunate as it underlines the limitations of the existing standard colonic evaluation compared to MRC. A strength of the present study is, however, that both readers were experienced readers of MRC. MRC is a procedure that requires dedicated personnel. A weakness of the study was the inability to evaluate all MRCs preoperatively, which would have given the surgeons the possibility of an intraoperative colonic evaluation in case of positive findings. The retrospective nature of the evaluation also led to the decision not to include polyps. The sensitivity and specificity polyp detection are well described, and smaller polyps (<10 mm) are rarely malignant lesions. However, most polyps need to be removed or followed up because more than 90% of colorectal cancers are believed to arise from smaller adenomas (23). Even so, the main focus of this article was the perioperative colonic evaluation and thus the consequences missed synchronous cancers. The fact that not all patients underwent postoperative colonic evaluation, just as 10 of 19 patients were excluded from the study due to disseminated disease or comorbidity, may have led to an overestimation of the sensitivity of the MRC and underestimation of overlooked synchronous cancers. However, the study reflects the daily clinical life in a Danish colorectal surgical department 1527
ACHIAM ET AL
and may contribute to estimating the benefits of implementing MRC in PREOP evaluation. Furthermore, another strength of the study is the randomized prospective design and the blinded evaluation. In conclusion, MRC is a valuable tool for PREOP colonic evaluation in patients with rectal cancer. Furthermore, we have shown that both the preoperative and postoperative colonic evaluation has serious lacks due to logistical challenges, but also due to tumor stenosis which prevent complete colonic evaluation. Thus, we recommend the use of MRC as part of the standard PREOP evaluation for all patients with rectal cancer.
ACKNOWLEDGMENTS This study was supported by a grant by the Lundbeckfonden (grant number: R112-A9278). The funding played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the article for publication.
REFERENCES 1. International agency for research on cancer. GLOBOCAN 2012: Estimated cancer incidence, mortality and prevalence worldwide in 2012; 2014. Ref Type: Online Source. 2. Scheele J, Stangl R, Altendorf-Hofmann A. Hepatic metastases from colorectal carcinoma: impact of surgical resection on the natural history. Br J Surg 1990; 77(11):1241–1246. 3. Heald RJ, Bussey HJ. Clinical experiences at St. Mark’s Hospital with multiple synchronous cancers of the colon and rectum. Dis Colon Rectum 1975; 18(1):6–10. 4. Gordon-Watson C. Specimen showing carcinoma of the pelvic colon and rectum co-existing and causing acute obstruction. Proc R Soc Med 1922; 15(Surg Sect):92–93. 5. Maxfield RG. Colonoscopy as a routine preoperative procedure for carcinoma of the colon. Am J Surg 1984; 147(4):477–480.
1528
Academic Radiology, Vol 22, No 12, December 2015
6. Carraro PG, Segala M, Cesana BM, et al. Obstructing colonic cancer: failure and survival patterns over a ten-year follow-up after one-stage curative surgery. Dis Colon Rectum 2001; 44(2):243–250. 7. Phillips RK, Hittinger R, Fry JS, et al. Malignant large bowel obstruction. Br J Surg 1985; 72(4):296–302. 8. Isler JT, Brown PC, Lewis FG, et al. The role of preoperative colonoscopy in colorectal cancer. Dis Colon Rectum 1987; 30(6):435–439. 9. Achiam MP, Burgdorf SK, Wilhelmsen M, et al. Inadequate preoperative colonic evaluation for synchronous colorectal cancer. Scand J Surg 2009; 98(1):62–67. 10. Adloff M, Arnaud JP, Bergamaschi R, et al. Synchronous carcinoma of the colon and rectum: prognostic and therapeutic implications. Am J Surg 1989; 157(3):299–302. 11. Askew A, Ward M, Cowen A. The influence of colonoscopy on the operative management of colorectal cancer. Med J Aust 1986; 145(6):254–255. 12. Fenlon HM, McAneny DB, Nunes DP, et al. Occlusive colon carcinoma: virtual colonoscopy in the preoperative evaluation of the proximal colon. Radiology 1999; 210(2):423–428. 13. Howard ML, Greene FL. The effect of preoperative endoscopy on recurrence and survival following surgery for colorectal carcinoma. Am Surg 1990; 56(3):124–127. 14. Dasmahapatra KS, Lopyan K. Rationale for aggressive colonoscopy in patients with colorectal neoplasia. Arch Surg 1989; 124(1):63–66. 15. Tate JJ, Rawlinson J, Royle GT, et al. Pre-operative or postoperative colonic examination for synchronous lesions in colorectal cancer. Br J Surg 1988; 75(10):1016–1018. 16. Vignati P, Welch JP, Cohen JL. Endoscopic localization of colon cancers. Surg Endosc 1994; 8(9):1085–1087. 17. Cho YB, Lee WY, Yun HR, et al. Tumor localization for laparoscopic colorectal surgery. World J Surg 2007; 31(7):1491–1495. 18. Wexner SD, Cohen SM, Ulrich A, et al. Laparoscopic colorectal surgery–are we being honest with our patients? Dis Colon Rectum 1995; 38(7):723–727. 19. Achiam MP, Holst Andersen LP, Klein M, et al. Preoperative evaluation of synchronous colorectal cancer using MR colonography. Acad Radiol 2009; 16(7):790–797. 20. Morrin MM, Farrell RJ, Raptopoulos V, et al. Role of virtual computed tomographic colonography in patients with colorectal cancers and obstructing colorectal lesions. Dis Colon Rectum 2000; 43(3):303–311. 21. Zijta FM, Bipat S, Stoker J. Magnetic resonance (MR) colonography in the detection of colorectal lesions: a systematic review of prospective studies. Eur Radiol 2010; 20(5):1031–1046. 22. van der Paardt MP, Zijta FM, Boellaard TN, et al. Magnetic resonance colonography with automated carbon dioxide insufflation: diagnostic accuracy and distension. Eur J Radiol 2014; 83(5):743–750. 23. Leslie A, Carey FA, Pratt NR, et al. The colorectal adenoma-carcinoma sequence. Br J Surg 2002; 89(7):845–860.