Advertisement

Familial Cancer

, Volume 9, Issue 4, pp 555–561 | Cite as

Magnetic resonance colonography for colorectal cancer screening in patients with Lynch syndrome gene mutation

  • Eu Jin LimEmail author
  • Christopher Leung
  • Alex Pitman
  • Damien L. Stella
  • Gregor Brown
  • Masha Slattery
  • Kaye Marion
  • Finlay Macrae
Article

Abstract

Lynch syndrome gene carriers have a 50–80% risk of colorectal cancer (CRC). Current guidelines recommend yearly colonoscopy, with associated procedure-related risks. Magnetic resonance colonography (MRC) was evaluated as a non-invasive alternative for CRC screening in this high-risk population. Adult Lynch syndrome gene carriers underwent both screening procedures on the same day. MRI radiologists read the scans and rated image quality. Endoscopists performed colonoscopy unaware of MRC findings until after procedure completion. If lesions were detected, their number, size and location were noted. Post-procedure, patients compared discomfort and inconvenience of MRC and colonoscopy on a visual analogue scale. Thirty patients were recruited. 83% of the MRC scans were of adequate to good quality. MRC detected three lesions in three patients (70, 36, 17 mm). All 3 were independently detected on colonoscopy, excised and found to be CRC. MRC failed to detect a 3 mm CRC found on colonoscopy. CRC prevalence was 13%. Colonoscopy detected a further 30 polyps, all <10 mm. Of these, 17 were hyperplastic polyps and 10 normal mucosa. Colonoscopy had a false positive rate of 32% as defined by histology. MRC failed to detect any polyp <10 mm. Mean patient discomfort scores were 20% for MRC and 68% for colonoscopy, P = 0.003. Mean patient inconvenience scores were 54% for MRC and 52% for colonoscopy, P = 0.931. MRC was reliable in detecting large polyps, potentially CRC. However MRC currently has poor sensitivity in detecting small polyps, limiting its utility in adenoma screening at this time. MRC was associated with less discomfort than CC.

Keywords

Lynch syndrome Magnetic resonance colonography Colonoscopy Colorectal cancer screening 

Abbreviations

CRC

Colorectal cancer

CTC

Computed tomography colonography

DNA

Deoxyribonucleic acid

MLH1

Human MutL Homolog-1

MMR

Mismatch repair

MRC

Magnetic resonance colonography

MRI

Magnetic resonance imaging

MSH2

Human MutS Homolog-2

MSH6

Human MutS Homolog-6

Introduction

Lynch syndrome is the most common type of hereditary colon cancer, accounting for 20–35% of such cancers and 3–5% of all colorectal cancer (CRC) [1]. It is an autosomal dominant cancer predisposition syndrome caused by germ line mutations in the DNA mismatch repair (MMR) gene system [2]. The MMR proteins function to maintain fidelity in DNA replication and mutations in this system results in single base pair mismatches and small-mispaired loops of DNA not being corrected, causing replication errors producing microsatellite instability [3]. Accumulation of mistakes in DNA replication ultimately results in oncogenesis. Affected individuals have a higher risk of developing colorectal cancer (CRC) and other associated cancers, such as endometrial, ovarian and renal tract cancer. CRCs tend to develop at an early age, arise in the proximal colon and be multiple (either synchronous or metachronous) [4].

Although multiple colonic polyps are not a feature, CRCs in Lynch syndrome are thought to arise from adenomatous polyps. Therefore, a surveillance program is important to detect precursor lesions and early stage cancers [5]. Regular colonoscopy is currently the recommended tool for CRC screening in these individuals, and colonoscopic screening has been shown to reduce the incidence as well as mortality rate of CRC in Lynch syndrome gene carriers in non-randomised controlled trials [5]. Although relatively safe in capable hands, colonoscopy does require sedation and bowel preparation, and is occasionally associated with complications [6].

As Lynch syndrome gene carriers require regular and repeated screening procedures throughout their lives, it is important to provide a safe and reliable method of detecting polyps and early CRCs with the least amount of procedure-related morbidity possible. Computed tomography colonography (CTC) has been studied in this regard [7]. CTC has the benefit of not being an invasive procedure, with no sedation required and no documented risk of bowel perforation associated. It is also able to visualize the rest of the abdomen and pelvis outside of the bowel and potentially detect Lynch syndrome-associated extracolonic disease [8]. However CTC is associated with high doses of ionizing radiation and exposing patients already at risk of developing cancer to such radiation would be undesirable. Furthermore multiple studies have shown that CTC has poor sensitivity for detecting small polyps [9].

Our pilot study evaluates the utility of magnetic resonance colonography (MRC), which has recently been developed to accurately visualize the structure of the colonic wall. Like CTC, MRC offers detailed images of the abdomen in any plane, but without the hazard of ionizing radiation and can therefore be used repeatedly in Lynch syndrome gene carriers without adding risk to a field already primed for oncogenesis by virtue of the monoallelic mismatch repair mutation. We assess whether MRC is an accurate and feasible screening technique for the detection of polyps and cancers in Lynch syndrome gene carriers in comparison to colonoscopy.

Methods

A prospective study of adult patients with Lynch syndrome gene mutations confirmed on genetic testing was undertaken. Participants were being followed up in the familial cancer clinic of a tertiary teaching hospital in Victoria, Australia, with surveillance managed by a dedicated surveillance service. These patients undergo yearly CRC screening with colonoscopy and have regular genetic counseling. Patients who were pregnant or had contraindications to MRI scanning (e.g. cardiac pacemakers, metal implants) were excluded. Both screening procedures were performed on the same day after a single bowel preparation, with the MRC undertaken in the morning and colonoscopy in the afternoon. Prior to the screening procedures, patients were required to consume low fiber food and avoid foods with high manganese content such as coffee, tea, chocolate, and fruit two days before examination. Standard bowel preparation for colonoscopy (e.g. Picoprep, Pharmatel) was provided the day before the procedures.

The MRC was performed on a 1.5 Tesla MRI (Magnetom Sonata; Siemens Medical Solutions, Erlangen, Germany) unit equipped with an advanced performance gradient system. No sedation or analgesics were required for the MRC. Optimal bowel distention was achieved with enemas of 1,500–2,000 ml of water. Scopolamine 20 mg was given intravenously to minimize patient discomfort and slow bowel motility. Intravenous contrast (Gadolinium 20 ml) was provided, and then imaging of the entire abdomen and pelvis was performed initially acquiring two dimensional T2 weighted fast field gradient echo data in the coronal position, then acquiring T2 weighted three dimensional gradient echo acquisition sequences of four millimetre thickness. Image interpretation and manipulation was performed on a workstation using GE Signa Echospeed (Software Version 9.1). Two consultant radiologists experienced in reading MRI then read the MRC scans independently and rated the images of the bowel as good, adequate or poor depending on the quality of bowel preparation. Firstly, the 2D images of each patient were evaluated to identify polyps or cancers, based on the angle, orientation, location and the segment in the colon. Then, the 3D images were used as a second reference if abnormal findings were identified in the 2D images. If lesions are found on MRC, their number, size and location within the bowel were documented. Polyp size was divided into small (≤5 mm), medium (6–10 mm) and large (>10 mm). For the purpose of determining lesion location, the bowel was divided into six parts—caecum, ascending, transverse, descending and sigmoid colons, and rectum. Extra-colonic findings were also recorded.

Colonoscopy was done after the MRC for each patient. Sedation was provided by a consultant anaesthetist and comprised of intravenous propofol and midazolam. A consultant gastroenterologist experienced in colonoscopy performed the procedure without prior knowledge of the MRC findings. At colonoscopy, if lesions were found, their number, size and location within the bowel were documented. Lesion size was determined by comparison to an open biopsy forceps (7 mm). Lesion location was recorded using the same system as MRC. All polyps detected at colonoscopy were resected for histopathological assessment. At the end of the colonoscopy, the results of that patient’s MRC was revealed to the endoscopist and, if any lesions on the MRC were not found on colonoscopy, the procedure was repeated with special attention paid to involved sections of bowel. The results of the MRC procedures were compared with the colonoscopy procedures to evaluate the sensitivity of MRC at detecting polyps.

Post-procedure, the patients were asked to compare the discomfort and inconvenience of MRC compared to colonoscopy on a visual analogue scale to determine whether patients preferred the non-invasive MRC as a screening test over colonoscopy. The lengths of the mark made by patients on the scoring sheets (Fig. 1) were converted into percentages from 0% (no discomfort or inconvenience at all) to 100% (very uncomfortable or very inconvenient). The Wilcoxon signed rank test was applied to evaluate the significance of differences in discomfort and inconvenience data sets between colonoscopy and MRC. A P value of <0.05 was considered statistically significant. The study protocol was approved by the Melbourne Health Human Research Ethics Committee.
Fig. 1

Patient questionnaire done post-procedure to evaluate the level of discomfort and inconvenience between MRC and colonoscopy

Results

Thirty patients with Lynch syndrome gene mutations were recruited from 2005 to 2009. Eleven patients (44%) had MSH2 mutation, 3 (12%) had MSH6 and 11 (44%) had MLH1. Seventeen patients (57%) were female and 13 (43%) were male, with a mean age of 48 (range 22–70 years old). Ten patients (33%) had between 1 and 7 polyps detected and excised on previous surveillance colonoscopies. Two patients (7%) had previous colorectal cancers, which were resected.

On review of the MRC scans, the radiologists deemed 19 scans (63%) to be of good image quality, 6 (20%) of adequate quality and 5 (17%) of poor quality due to insufficient bowel preparation, inadequate colonic distension or the presence of air-bubbles resulting in image artifact. MRC detected a total of 3 lesions in 3 patients—2 in scans deemed of adequate image quality and 1 in a scan deemed of poor quality. All of these lesions were large, measuring 70, 36 and 17 mm and were detected by both radiologists independently. MRC images of the largest lesion can be seen in Figs. 2 and 3. The lesions were found in the ascending colon in 2 patients and in the transverse colon in 1 (Table 1). These 3 large lesions were independently detected on colonoscopy and excised or biopsied. On histological examination, all 3 were found to be adenocarcinoma.
Fig. 2

T2-weighted coronal section (a) and Post-contrast 3D T1-weighted sequence (b) showing the 70 mm stenosing tumour in the transverse colon

Fig. 3

Magnetic resonance colonography luminal views showing a normal section of the distal transverse colon (a) and the stenosing lesion in the proximal transverse colon (b)

Table 1

Lesions detected on MRC

Lesion

Size (mm)

Location

Image quality

1

70

Transverse colon

Poor

2

36

Ascending colon

Adequate

3

17

Ascending colon

Adequate

At colonoscopy, caecal intubation was achieved in 100% of patients. Colonoscopy detected a total of 33 lesions in 11 patients, with 28 (85%) small polyps (≤5 mm), 2 (6%) medium-sized polyps (6–10 mm), and 3 (9%) large polyps (>10 mm). All polyps found on colonoscopy were excised or biopsied. Information regarding the polyps detected on colonoscopy can be seen in Tables 2, 3 and 4. On histological examination, 4 of these lesions detected in 4 separate patients (measuring 75, 25, 11 and 3 mm) were found to be adenocarcinoma, giving a CRC prevalence in this series of 13%. Two of these cancers were found in the ascending colon and 2 in the transverse colon. As such, MRC failed to detect the 3 mm CRC in the transverse colon of one patient. Two tubulovillous adenomas were detected in 2 patients, both in the transverse colon, measuring 4 mm and 3 mm. In one patient, a sessile serrated adenoma was detected in the caecum measuring 8 mm. Seventeen hyperplastic polyps were found in seven patients measuring 2–6 mm. One hyperplastic polyp was found in the caecum, 4 in the transverse colon, 2 in the descending colon, 3 in the sigmoid colon, and 7 in the rectum. Ten “polyps” measuring 2–5 mm, resected from the transverse colon, sigmoid colon and rectum of seven patients, revealed normal mucosa on histological examination.
Table 2

Large lesions detected on colonoscopy

Lesion

Location

Size (mm)

Histology

1

Transverse colon

75

Adenocarcinoma

2

Ascending colon

25

Adenocarcinoma

3

Ascending colon

10

Adenocarcinoma

Table 3

Medium-sized lesions detected on colonoscopy

Lesion

Location

Size (mm)

Histology

1

Caecum

8

Sessile serrated adenoma

2

Rectum

6

Hyperplastic polyp

Table 4

Small lesions detected on colonoscopy

No. of lesions

Location

Size (mm)

Histology

1

Caecum

4

Hyperplastic polyp

1

Transverse colon

3

Adenocarcinoma

2

Transverse colon

3–4

Tubulovillous adenomas

2

Transverse colon

3–4

Hyperplastic polyps

2

Transverse colon

3

Normal mucosa

2

Descending colon

3–5

Hyperplastic polyps

3

Sigmoid colon

2–3

Hyperplastic polyps

1

Sigmoid colon

2

Normal mucosa

8

Rectum

2–5

Hyperplastic polyps

6

Rectum

2–4

Normal mucosa

Although MRC detected all 3 adenocarinomas that were larger than 10 mm, it failed to detect 1 adenocarcinoma, 1 sessile serated adenoma and 2 tubulovillous adenomas, all smaller than 10 mm. As such, the sensitivity of MRC for detecting all-sized adenoma/carcinoma in our study was 43%. MRC detected extra-colonic findings in nine patients (36%)—renal cysts were detected in five patients and liver cysts in four patients. None of these findings had any clinical consequence. None of the four patients with colorectal cancer determined on histology had any evidence of metastatic disease on MRC. MRC did not detect any non-colorectal Lynch syndrome-related cancers in our study.

The mean patient discomfort score for MRC was 20%, which was less than the mean patient discomfort score for colonoscopy (68%), P = 0.003. The mean patient inconvenience score for MRC (54%) was comparable to that for colonoscopy (52%), P = 0.931.

Discussion

Lynch syndrome gene carriers have a 50–80% lifetime risk of developing CRC [10], with the associated morbidity and mortality. Hence an effective CRC screening protocol, which is acceptable to these patients, is of paramount importance. The current gold standard for CRC surveillance in Lynch syndrome is full colonoscopy because of the predominance of right-sided tumours in these patients, with over 50% of CRCs located in the right colon [5]. However colonoscopy requires sedation, causes patient discomfort and may be associated with procedure-related complications including abdominal pain, bowel perforation (0.1%), bleeding especially after polypectomy (0.21%) and even death (0.01%) [11]. Current guidelines recommend a colonoscopy every 1–3 years starting at age 20–25 for individuals with Lynch syndrome-associated mutations or 5 years earlier than the earliest age of cancer onset in the family, which ever is earlier [12]. In our patients who require frequent, repeated and ongoing surveillance colonoscopies, there is a real risk that an adverse event will eventually occur. Hence there is a need for an acceptable non-invasive screening method with high sensitivity.

Although CTC provides a non-invasive method to image the bowel, the lack of sensitivity in detecting polyps under 10 mm significantly reduces its efficacy in CRC screening. Furthermore the associated high doses of ionizing radiation could theoretically cause mutations in DNA that would not be corrected due to the defective MMR system in our patients, hence leading to procedure-related increased cancer risk. This has led us to evaluate the feasibility of using MRC to screen for CRC. Although other studies have evaluated the utility of MRC for detecting CRC in patients with bowel symptoms [13], or screening for bowel cancer in patients with a positive faecal occult blood test or a family history of such cancers [14], there have been no studies to date addressing the efficacy of MRC in screening Lynch syndrome mutation carriers, a high-risk population for developing CRC. We decided to compare MRC against conventional colonoscopy as conventional colonoscopy has a high sensitivity for detecting small and large polyps and is the current recommended modality.

In our study, MRC did not miss any large lesion found on colonoscopy. MRC detected 3 of the 4 CRCs (the 3 that were larger than 10 mm), only missing the small adenocarcinoma in the transverse colon measuring 3 mm. However, MRC did not detect any of the lesions under 10 mm, missing 1 adenocarcinoma, 1 sessile serrated adenoma and 2 tubulovillous adenomas (all 3–8 mm). The sensitivity of MRC for detecting all-sized adenoma/carcinoma in our study was 43%. Conventional colonoscopy did not miss any of the large polyps detected on MRC and detected a further 30 small to medium-sized polyps in 11 patients. However, 10 out of the 28 (36%) small polyps detected at colonoscopy were found on histology to be normal mucosa, probably just prominent mucosal folds. Also 1 of the medium-sized polyps and 16 of the small polyps found at colonoscopy were hyperplastic polyps, with no clinical consequence as they would not develop into CRC. This means that 26 of the 28 (93%) small polyps detected by colonoscopy would have had no malignant potential, giving colonoscopy a high false positive rate of 93% for detecting small adenomas, and 79% for detecting adenomas in all-sized polyps.

Our findings are similar to other studies comparing MRC to colonoscopy for the detection of polyps and CRC. Luboldt et al. found that of 122 small polyps ≤5 mm detected by colonoscopy, none were detected by MRC. However, all of 60 larger polyps (≥5 mm) in the colonoscopy results were detected by MRC screening (the sensitivity for large polyps was 100%) [15]. The study by Ajaj et al. also supports the sensitivity of MRC in screening for large polyps in that although the MRC did not detect 30 small polyps (≤5 mm) detected by colonoscopy, MRC did detect 18 of 20 polyps ≥5 mm identified by colonoscopy (sensitivity 93%) [16].

Although an advantage of MRC is the ability to detect extra-colonic pathology, in our study no relevant extra-colonic lesions were detected. In particular, the three patients with large polyps detected at MRC (later determined to be adenocarcinoma) did not have evidence of regional lymphadenopathy or distant metastases within the abdomen. And after confirmation of the other patient with colon cancer, review of the MRC films did not reveal the presence of extra-colonic abnormalities. In one study, MRC screening of 132 patients detected 5 liver metastases, 1 prostate cancer, 1 hepatocellular carcinoma, and 3 ovarian cancers [15]. MRC certainly has the potential to detect extracolonic tumours such as uterine, ovarian and renal tract cancers associated with Lynch syndrome, which would not be seen on colonoscopy. However another recent study suggests that the diagnostic yield of MRC for detecting extracolonic findings may be limited even in a population suspected of having colonic pathology [17]. Another benefit of MRC is the ability to detect synchronous CRCs preoperatively, especially in patients with impassable stenoses on colonoscopy, thus enabling planning for optimal surgical therapy [18].

An issue with MRC is the image quality as determined by the adequacy of bowel preparation. We note that other centres employ the faecal tagging method for MRC, but there have been concerns regarding suboptimal image quality using faecal tagging [19]. As such we decided to use the same bowel preparation as colonoscopy for the MRC in order to obtain adequate quality images and reduce the inconvenience to patients. Overall, 80% of the MRC scans were deemed adequate or good by the reviewing radiologists, and even though one of the scans was deemed to be of poor quality, a large lesion of 70 mm, later determined to be adenocarcinoma, was still detected in the transverse colon. This shows that MRC is reliable in detecting large polyps, which have the potential for being adenocarcinomas, and as such can be used to search for colorectal cancers in patients unwilling or unable to undergo colonoscopy. However, MRC currently has poor sensitivity for detecting small to medium-sized polyps, thus limiting its usefulness in screening for polyps to prevent progression to adenocarcinoma in a high-risk population at this stage. Recently, a study using a 3.0T MRI-unit for MRC showed that it was able to detect 100% of polyps >5 mm and 50% of polyps 5 mm or less compared to colonoscopy, indicating that with improvement in technology, the accuracy of MRC is expected to improve [20]. Another study showed that 3.0T MRI scanners were superior to 1.5T scanners for detecting smaller polyps [21].

In order for MRC to be accepted as a screening test by patients, it should cause less discomfort and less inconvenience than the current method of colorectal cancer screening. We found that MRC was indeed associated with significantly less discomfort than colonoscopy, but the level of inconvenience associated with MRC was comparable to colonoscopy. MRC was well tolerated and did not require sedation, in contrast to colonoscopy. This is consistent with the findings by Pappalardo et al. which found that of 70 patients, 15 felt severe discomfort during colonoscopy, but none felt severe discomfort during MRC [13].

Conclusion

MRC was reliable in detecting large polyps, which have the potential to be CRC. However MRC currently has poor sensitivity in detecting small polyps, limiting its utility in adenoma screening in high-risk populations at this time. MRC provides better localisation information than colonoscopy and has the potential to detect extra-colonic Lynch syndrome-associated cancers. MRC was associated with less discomfort than colonoscopy.

References

  1. 1.
    Annie Yu HJ, Lin KM, Ota DM et al (2003) Hereditary non-polyposis colorectal cancer: preventive management. Cancer Treat Rev 29(6):461–470CrossRefPubMedGoogle Scholar
  2. 2.
    DeFrancisco J (2003) Diagnosis and management of hereditary non-polyposis colon cancer. Gastrointest Endosc 58(3):390–408CrossRefPubMedGoogle Scholar
  3. 3.
    Chung DC, Rustgi AK (2003) The hereditary non-polyposis colorectal cancer syndrome: genetics and clinical implications. Ann Intern Med 138(7):560–570PubMedGoogle Scholar
  4. 4.
    Lynch HT, De la Chapelle A (2003) Hereditary colorectal cancer. N Engl J Med 348(10):919–932CrossRefPubMedGoogle Scholar
  5. 5.
    Järvinen HJ, Aarnio M, Mustonen H et al (2000) Controlled 15-year trial on screening for colorectal cancer in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 118(5):829–834CrossRefPubMedGoogle Scholar
  6. 6.
    Purkayastha S, Tekkis PP, Athanasiou T et al (2005) Magnetic resonance colonography versus colonoscopy as a diagnostic investigation for colorectal cancer: a meta-analysis. Clin Radiol 60(9):980–989CrossRefPubMedGoogle Scholar
  7. 7.
    Rockey DC, Paulson E, Niedzwiecki D et al (2005) Analysis of air contrast barium enema, computed tomographic colonography, and colonoscopy: prospective comparison. Lancet 365(9456):305–311PubMedGoogle Scholar
  8. 8.
    Renkonen-Sinisalo L, Kivisaari A, Kivisaari L et al (2007) Utility of computed tomographic colonography in surveillance for hereditary nonpolyposis colorectal cancer syndrome. Fam Cancer 6(1):135–140CrossRefPubMedGoogle Scholar
  9. 9.
    Cotton PB, Durkalski VL, Pineau BC et al (2004) Computed tomographic colonography (virtual colonoscopy) a multicenter comparison with standard colonoscopy for detection of colorectal neoplasia. JAMA 291(14):1713–1719CrossRefPubMedGoogle Scholar
  10. 10.
    De Vos Tot Nederveen Cappel WH, Nagengast FM, Griffioen G et al (2002) Surveillance for hereditary nonpolyposis colorectal cancer: a long-term study on 114 families. Dis Colon Rectum 45(12):1588–1594CrossRefPubMedGoogle Scholar
  11. 11.
    Viiala CH, Zimmerman M, Cullen DJE et al (2003) Complication rates of colonoscopy in an Australian teaching hospital environment. Intern Med J 33:355–359CrossRefPubMedGoogle Scholar
  12. 12.
    Mecklin JP, Jarvinen HJ (2005) Surveillance in Lynch syndrome. Fam Cancer 4(3):267–271CrossRefPubMedGoogle Scholar
  13. 13.
    Pappalardo G, Polettini E, Frattaroli FM et al (2000) Magnetic resonance colonography versus conventional colonoscopy for the detection of colonic endoluminal lesions. Gastroenterology 119(2):300–304CrossRefPubMedGoogle Scholar
  14. 14.
    Hartmann D, Bassler B, Schilling D et al (2006) Colorectal polyps: detection with dark-lumen MR colonography versus conventional colonoscopy. Radiology 238(1):143–149CrossRefPubMedGoogle Scholar
  15. 15.
    Luboldt W, Bauerfeind P, Wildermuth S et al (2000) Colonic masses: detection with MR colonography. Radiology 216(2):383–388PubMedGoogle Scholar
  16. 16.
    Ajaj W, Pelster G, Treichel U et al (2003) Dark lumen magnetic resonance colonography: comparison with conventional colonoscopy for the detection of colorectal pathology. Gut 52:1738–1743CrossRefPubMedGoogle Scholar
  17. 17.
    Yusuf E, Florie J, Nio CY et al. (2009) Incidental extracolonic findings on bright lumen MR colonography in a population at increased risk for colorectal carcinoma. Eur J Radiol Oct 15. Epub ahead of printGoogle Scholar
  18. 18.
    Achiam MP, Holst Andersen LP, Klein M et al (2009) Preoperative evaluation of synchronous colorectal cancer using MR colonography. Acad Radiol 16(7):790–797CrossRefPubMedGoogle Scholar
  19. 19.
    Kuehle CA, Langhorst J, Ladd SC et al (2007) Magnetic resonance colonography without bowel cleansing: a prospective cross sectional study in a screening population. Gut 56(8):1079–1085CrossRefPubMedGoogle Scholar
  20. 20.
    Saar B, Gschossmann JM, Bonel HM et al (2008) Evaluation of magnetic resonance colonography at 3.0 Tesla regarding diagnostic accuracy and image quality. Invest Radiol 43(8):580–586CrossRefPubMedGoogle Scholar
  21. 21.
    Lauenstein TC, Saar B, Martin DR (2007) MR colonography: 1.5T versus 3T. Magn Reson Imaging Clin N Am 15(3):395–402CrossRefPubMedGoogle Scholar

Copyright information

© Springer Science+Business Media B.V. 2010

Authors and Affiliations

  • Eu Jin Lim
    • 1
    Email author
  • Christopher Leung
    • 1
  • Alex Pitman
    • 2
  • Damien L. Stella
    • 3
  • Gregor Brown
    • 1
  • Masha Slattery
    • 1
  • Kaye Marion
    • 4
  • Finlay Macrae
    • 1
  1. 1.Department of Gastroenterology and Clinical Nutrition ServiceRoyal Melbourne HospitalParkvilleAustralia
  2. 2.Department of RadiologySt. Vincent’s Hospital MelbourneFitzroyAustralia
  3. 3.Department of RadiologyRoyal Melbourne HospitalParkvilleAustralia
  4. 4.School of Mathematical and Geospatial SciencesRoyal Melbourne Institute of TechnologyMelbourneAustralia

Personalised recommendations