Computerized tomography: an unreliable method for accurate staging of bladder tumors in patients who are candidates for radical cystectomy

A total of 164 patients with bladder tumors underwent preoperative staging by computerized tomography. All patients were previously untreated, or had undergone only transurethral biopsy or resection of the tumor before computerized tomography. The post-cystectomy histological stage was compared to the preoperative computerized tomography stage. Computerized tomography accuracy according to the tumor, nodes and metastasis classification was only 32.3%, whereas overstaging was found in 39.6% and understaging in 28.1% of the cases. In untouched tumors or after transurethral resection computerized tomography accuracy demonstrated no significant difference. Only 2 of 19 true positive lymph nodes were staged correctly. Of 10 suspicious nodes results of computerized tomography were false positive in 8. Computerized tomography is an unreliable method for accurate preoperative staging of bladder carcinoma. The indication for either an operation, chemotherapy or radiotherapy for the treatment of bladder neoplasms should not be based on computerized tomography findings.     

The role of preoperative transarterial embolization in spinal tumors. A large single-center experience

Background contextPatients with spinal tumors are often referred for preoperative angiography and embolization before surgical resection to minimize intraoperative bleeding.PurposeThe purpose of the present study was to investigate the angiographic appearance of a variety of spinal tumors, assess the safety and efficacy of preoperative embolization in relation to the amount of intraoperative blood loss, and correlate intraoperative tumor histology with the degree of gadolinium enhancement on spinal magnetic resonance imaging (MRI) and tumor vascularity visualized during angiography.Study design/settingRetrospective and single-institution cohort study.Patient sampleOne hundred four patients with spinal tumors referred for preoperative embolization.Outcome measuresEffectiveness of preoperative embolization in relation to intraoperative blood loss and number of transfused packed red blood cell units in perioperative period (72 hours).MethodsFrom 2000 to 2009, 104 patients with spinal tumors underwent 114 spinal angiographies with the intent to embolize feeder vessels before surgery. The effectiveness of embolization was compared with the documented intraoperative blood loss. Angiographic tumor vascularity was graded from 0 (avascular) to 3 (highly vascular). Ninety-four patients had a pre– and post–gadolinium-enhanced MRI of the spine before transarterial embolization. Magnetic resonance imaging vascular enhancement was classified as Grade 3 (avid contrast enhancement), Grade 2 (moderate), or Grade 1 (mild).ResultsTransarterial tumor embolization was angiographically complete in 63 (66%) and partial in 33 procedures (34%). In 18 cases, the target was not deemed suitable for embolization. A limited statistical analysis did not reveal a statistical difference in documented intraoperative blood loss between patients with complete versus partial embolization for the entire cohort or when stratified into renal cell carcinoma (RCC; p=.64), multiple myeloma (p=.28), malignant (p=.17) and benign tumor groups (p=.26). There were no clinical complications associated with embolization. There was poor correlation between MRI enhancement and angiographic vascularity.ConclusionsPreoperative embolization was angiographically effective in most cases. Avid gadolinium enhancement (Grade 3) on MRI was not predictive of hypervascularity on angiography. Furthermore, hypervascularity was not restricted to classically vascular tumors, such as RCC, as it was noted in some patients with breast and prostate cancer. However, with the available numbers, the quality of preoperative embolization did not significantly affect intraoperative blood loss. A future prospective randomized controlled study may be warranted to better characterize the benefits of preoperative embolization for spinal tumors.     

Staging urinary bladder cancer with dynamic MR imaging

This article reviews the magnetic resonance (MR) staging of bladder cancer. The multiplanar and soft-tissue characterization capabilities of MR imaging make it a valuable diagnostic tool to image the urinary bladder. Recent advances of MR imaging such as fast imaging, pelvic phased array coil, and dynamic imaging improve the image quality and diagnostic accuracy for staging bladder cancer. Some patient-related factors are also important for optimal imaging of the urinary bladder, especially motion artifacts from the gastrointestinal tract and the degree of bladder distension. An anticholinergic agent should be used for suppressing the motion artifacts. Optimal bladder filling can be achieved by asking patients to void and drink water 1 hour before examinations. Scanning perpendicular to the bladder wall is necessary for optimal evaluation for staging bladder cancer. Oblique scanning is needed in cases when a tumor is not located on the dome, base, anterior wall, posterior wall, or lateral walls. The early phase image of dynamic imaging is most useful for staging tumors. Better contrast between tumor and bladder wall on dynamic images provides high staging accuracy, especially in differentiation between superficial tumors and tumors with muscle invasion. MR imaging is comparable to computed tomography (CT) in the evaluation of lymph nodes. Although MR imaging currently is not appropriate for screening for bladder cancer and detecting small tumors, it has been proved to be most useful in the staging of bladder cancer.     

Multiple imaging procedures including MRI in bladder cancer

Endoscopic photography, double contrast cystography, transurethral echography, X-ray CT scan, and MRI (magnetic resonance imaging) were utilized for the staging diagnosis of the four patients with carcinoma of the bladder. In the first case, a 70-year-old man, since all of the five imaging procedures suggested a superficial and pedunculated tumor, his bladder cancer was considered T1. The classification of stage T3 carcinoma was made for the second 86-year-old male. Because all of his imaging examinations showed a tumor infiltrating deep muscle and penetrating the bladder wall. The third case was a 36-year-old male. His clinical stage was diagnosed as T2 or T3a by cystophotography, double contrast cystogram, ultrasonography, and X-ray CT scan. However, MRI showed only thickened bladder wall and the infiltrating tumor could not be distinguished from the hypertrophic wall. The last patient, a 85-year-old female, had a smaller Ta cancer. Her double contrast cystography revealed the small tumor at the lateral bladder wall. But, the tumor could not be detected by transaxial, sagittal and coronal scans. Multiple imaging procedures combining MRI and staging diagnosis of the bladder carcinoma were discussed.     

Efficacy of gadolinium-diethylenetriaminepentaacetic acid-enhanced magnetic resonance imaging for differentiation between superficial and muscle-invasive tumor of the bladder: a comparative study with computerized tomography and transurethral ultrasonography

Gadolinium-labeled diethylenetriaminepentaacetic acid (Gd-DTPA)-enhanced magnetic resonance imaging (MRI) was evaluated in an effort to clarify whether MRI could replace or be proved to be superior to computerized tomography (CT) and/or transurethral ultrasonography. A total of 57 bladder cancer patients was evaluated. MRI was performed with a superconducting magnet operating at 1.5 Tesla. The images acquired were multisections, having a fast spin-echo pulse sequence of less than a 14-second breath holding. Serial scans were performed before and immediately after Gd-DTPA venous injection. The findings on different imaging techniques were compared with the histological stagings. A proper diagnosis was made in 42 of 57 cases (73.7%) by Gd-DTPA-enhanced MRI, in 27 of 57 (47.4%) by CT and in 31 of 57 (54.4%) by transurethral ultrasonography when comparing the histological findings. The sensitivity and specificity for differentiating superficial and muscle-invasive tumor of each imaging method were, respectively, 96.2 and 83.3% in Gd-DTPA-enhanced MRI, 96.0 and 58.3% in CT, and 88.0 and 66.7% in transurethral ultrasonography. These data suggest that the staging of bladder cancer by Gd-DTPA-enhanced MRI appears to be superior and more accurate than the staging obtained by CT and transurethral ultrasonography.     

Is a preoperative multidiagnostic approach to predict surgical resectability of periampullary tumors still effective?

BACKGROUND: Multimodality staging is recommended in patients with periampullary tumors to optimize preoperative determination of resectability. We investigated the potency of currently used diagnostic procedures in order to determine resectability. METHODS: Ninety-five consecutive patients with periampullary tumors prehospitally staged resectable underwent preoperative diagnostic tests: helical-computed tomography (CT) with maximum intensity projection of arterial vessels (MIP), magnetic resonance imaging (MRI), magnetic resonance cholangiopancreaticography (MRCP), endoscopic ultrasonography (EUS), endoscopic retrograde cholangiopancreaticography (ERCP), digital subtraction angiography (DSA), and positron emission tomography (PET). Diagnoses were verified by surgery and histopathology. RESULTS: In 45 patients with benign and 50 patients with malignant periampullary tumors sensitivity for tumor diagnosis was 89% to 96% in CT, MRI, EUS, and PET. Small tumors were best diagnosed by EUS (100%). Diagnosis of malignancy was made with 85% (EUS), 83% (CT), 82% (PET), and 72% (MRI) accuracy. Arterial vessel infiltration was best predicted by CT/MIP with an accuracy of 85%. For venous vessel infiltration MRI reached 85% accuracy. Accuracy rates for local nonresectability were 93% (EUS), 92% (MRI), and 90% (CT). Two and 4 of 8 patients with distant metastases were identified by CT and PET, respectively. The correct diagnosis of malignancy and determination of resectability was made by CT in 71% and by MRI in 70%. Biliary stenting reduced accuracy of CT diagnosis of malignancy from 88% to 73%. CONCLUSIONS: CT obtained before stenting was the single most useful test, providing correct diagnosis in 88% and resectability in 71% of patients. If no tumor is depicted in CT, EUS should be added. Uncertain venous vessel infiltration can be verified by MRI or EUS. Angiography should no longer be a routine diagnostic procedure. Equivocal tumors or possible metastasis may be further examined with PET.     

The effects of neoadjuvant chemotherapy and chemo-radiation therapy on MRI staging in invasive bladder cancer: comparative study based on the pathological examination of whole layer bladder wall

Objective

We evaluated the correlation of radiological findings obtained by MRI study with pathological diagnosis in invasive bladder cancer treated with neoadjuvant chemotherapy, with or without radiation.

Design, Setting, and Participants

Twenty-seven patients, who underwent total or partial cystectomy for invasive bladder tumors, were enrolled into the present study. Eight cases had received neoadjuvant chemotherapy following the staging biopsy (group A), ten cases had received chemo-radiation therapy following the staging biopsy (group B), and nine cases had received preoperative staging biopsy alone (group C). As a final treatment, 12 of the 27 patients underwent total cystectomy and the other 15 patients underwent partial cystectomy. MRI was conducted prior to total or partial cystectomy in each case. The pathological stage was assessed by histological examination of the entire layer of the bladder wall.

Results and Limitations

Tumor stage assessed by MRI was consistent with pathological findings in 16 of the 27 cases (59.3%), while MRI produced over-staging in 7 cases and under-staging in 4 cases. The accuracy of staging was 75.0, 30.0, and 77.8% in groups A, B, and C, respectively. The accuracy of MRI staging in group B was lower than that in group C (P < 0.05). There was no difference in the accuracy of MRI staging between groups A and C.

Conclusion

MRI is useful for the staging of bladder cancer. However, care needs to be taken when staging invasive bladder tumors treated with neoadjuvant chemo-radiation therapy, because inflammatory infiltrations and/or fibrous changes caused by the chemotherapy or chemo-radiation therapy make precise staging with MRI difficult.     

Staging of bladder carcinoma by magnetic resonance imaging. A preliminary report

Magnetic resonance imaging (MRI) of bladder carcinoma was performed in 10 patients who subsequently underwent operations for tumor staging. MRI was performed with a 0.1 T resistive magnetic resonance unit. Images were obtained in the transverse, coronal and sagittal directions with different repetition and echo times. Simultaneously, computed tomography (CT) was performed in 9 of the 10 patients. The accuracy of MRI and CT staging was then evaluated by comparison with the staging from pathologic diagnosis. All of the tumors, which were 2.3-6.3 cm in diameter, were readily depicted by both imaging techniques. The overall accuracy of MRI staging for the 10 patients in whom staging had been pathologically confirmed was 90%, while the accuracy of CT staging for the 9 patients was 55%. Although no statistically significant differences in accuracy between CT and MRI staging can be established on the basis of the above results because of the small number of patients included in this study, MRI appears to be a very useful modality for staging bladder carcinoma.     

Magnetic resonance imaging of the bladder and prostate

Summary Thirteen patients with midline pathology of the bladder or prostate and two normal volunteers were examined by Magnetic Resonance Imaging (MRI) using a variety of pulse sequences. The MRI results were compared with computed tomography (CT) and transrectal or transurethral ultrasound (US). This study demonstrates that MRI is capable of visualizing pathology of the bladder and prostate at least as well as CT and US. However, MRI seems to be more advantageous in identifying lesions of tumors in the roof and base of the bladder because of flexible imaging planes. MRI is also very promising in the staging of pelvic malignancies because of excellent contrast between tumor and nearby important anatomical structures. This indicates that MRI might become the modality of choice in urological disease of the bladder and prostate.     

螺旋CT仿真内窥镜技术在膀胱肿瘤分期诊断中的应用

目的 探讨螺旋CT仿真内窥镜(CTVE)对膀胱肿瘤分期诊断的作用。方法 对10例正常对照和40例经手术病理检查证实的膀胱肿瘤患者术前采用多层面螺旋CT机进行容积扫描,将所得扫描数据转入工作站利用三维重建检查软件进行后处理,获得膀胱CTVE三维图像。将重建图像分别与冠状或轴位CT扫描、膀胱镜检查所见、手术中所见及病理检查结果进行比较分析。结果 CTVE对膀胱肿瘤检出率及分期准确率分别为98％(39／40)和85％(33／39),直径≥0．5cm的肿瘤显示率为100%。可观察尿道内口及膀胱颈部病变,弥补膀胱镜视野盲区,但不能显示膀胱黏膜的表浅病变及进行活检。结论 CTVE具有无创优点,在膀胱肿瘤临床分期及膀胱镜视野盲区的肿瘤诊断方面是膀胱镜较好的替代和补充方法。     

Lymph node staging for urinary bladder carcinoma with positron emission tomography (PET)

PET is a new method in nuclear medicine which examines the metabolism and not the morphology. Tumors show a higher rate of glycolysis than benign tissue and hence can be detected by radioactive glucose. This method has proved good for various tumors. In this study the lymph node staging of bladder cancer by PET was investigated. In 64 patients a PET of the pelvis after injection of fluorodeoxyglucose (FDG) was carried out preoperatively; the PET-results were compared with the histology of the OR specimen after classical pelvic lymphadenectomy. For lymph node staging positive nodes were found in 14 patients which was correct; a false-negative result was obtained in 7 patients. In 37 patients the PET-result was true-negative and in 6 patients false-positive resulting in a sensitivity of 67 %, a specificity of 86 % and an accuracy of 80 %. Therefore, our PET results are encouraging and seem to be better than those obtained by classical staging procedures such as CT or MRI.     

Lymph node staging for urinary bladder carcinoma with positron emission tomography (PET)

Summary PET is a new method in nuclear medicine which examines the metabolism and not the morphology. Tumors show a higher rate of glycolysis than benign tissue and hence can be detected by radioactive glucose. This method has proved good for various tumors. In this study the lymph node staging of bladder cancer by PET was investigated. In 64 patients a PET of the pelvis after injection of fluorodeoxyglucose (FDG) was carried out preoperatively; the PET-results were compared with the histology of the OR specimen after classical pelvic lymphadenectomy. For lymph node staging positive nodes were found in 14 patients which was correct; a false-negative result was obtained in 7 patients. In 37 patients the PET-result was true-negative and in 6 patients false-positive resulting in a sensitivity of 67 %, a specificity of 86 % and an accuracy of 80 %. Therefore, our PET results are encouraging and seem to be better than those obtained by classical staging procedures such as CT or MRI.      

Preoperative lymph-node staging of invasive urothelial bladder cancer with 18F-fluorodeoxyglucose positron emission tomography/computed axial tomography and magnetic resonance imaging: correlation with histopathology

OBJECTIVE: The treatment and prognosis of bladder cancer are based on the depth of primary tumour invasion and the presence of metastases. A highly accurate preoperative tumour, node, metastasis (TNM) staging is critical to proper patient management and treatment. This study retrospectively investigated the value of 1?F-fluorodeoxyglucose (FDG) positron emission tomography/computed axial tomography (1?F-FDG PET/CT) and magnetic resonance imaging (MRI) for preoperative N staging of bladder cancer. Material and methods. From June 2006 to January 2008, 48 consecutive patients diagnosed with bladder cancer were referred to preoperative staging including MRI and 1?F-FDG PET/CT. Eighteen out of 48 patients underwent radical cystoprostatectomy including removal of lymph nodes for histology, and were included in the study. Values of 1?F-FDG PET/CT and MRI for regional N staging were compared to histopathology findings, the gold standard. Results. 1?F-FDG PET/CT and MRI were performed in 18 patients. The specificities for detection of lymph-node metastases for MRI and 1?F-FDG PET/CT were 80% (n = 15) and 93.33% (n = 15), respectively. The negative predictive values were 80% (n = 15) and 87.5% (n = 16) for MRI and 1?F-FDG PET/CT, respectively. The differences in specificity and negative predictive values were not statistically significant. Conclusions. No significant statistical difference between 1?F-FDG PET/CT and MRI for preoperative N staging of urothelial bladder cancer was found in the study. However, the trend of the data indicates an advantage of 1?F-FDG PET/CT over MRI. Larger prospective studies are needed to elucidate the role of 1?F-FDG PET/CT in N staging of bladder cancer.     

Staging of bladder carcinoma by magnetic resonance imaging compared with computed tomography and transurethral echography

Various techniques are used for tumor staging of bladder carcinoma such as endoscopic photography, transurethral echography, and computed tomographic (CT) scan. Endoscopic photography enables easy observation of tumor morphology, but we utilized other approaches to determine the presence or absence of tumor infiltration. Magnetic resonance imaging (MRI) has recently found widespread use in the staging of tumors. We first employed this approach at our hospital in February 1987. We compared the diagnostic accuracy rate of MRI with that of transurethral echography and CT scan in 10 cases of bladder tumor. The diagnostic accuracy rate of MRI was 70%, while that of transurethral echography and CT scan was 40% and 50%, respectively. We concluded that MRI was very useful in the diagnostic staging of bladder tumor.     

Preoperative balloon occluded arterial infusion chemotherapy for locally invasive bladder cancer--accurate staging for bladder preservation

The possibility of bladder preservation by preoperative balloon occluded arterial infusion (BOAI) chemotherapy was studied in 111 patients with locally invasive bladder cancer. BOAI was performed by blocking the blood flow of the internal iliac artery and by performing intra-arterial infusion of adriamycin (50 mg/body) and cisplatin (100 mg/body). Before BOAI the clinical diagnosis was T2 in 36, T3a in 29, T3b in 27, T4 in 11 and after BOAI it was T0 in 1, T1 in 27, T2 in 25, T3a in 20, T3b in 20, and T4 in 10. Down staging was observed on diagnostic images in 46.6%. Thirty patients (27.0%) received transurethral resection of bladder tumor (TUR-Bt) and their bladder could be preserved. The 5-year cancer-specific survival rate was 100% in pT0 (n = 9), 97.5% in pT1 (n = 47), 79.9% in pT2 (n = 21), 80.0% in pT3a (n = 6), 39.9% in pT3b (n = 18) and 51.9% in pT4 cases (n = 9). For the bladder preservation, accurate staging diagnosis is required. Since 1992, endorectal magnetic resonance imaging (MRI) has been used in addition to imaging diagnosis for improving the accuracy of staging diagnosis. The accuracies of staging diagnosis with and without endorectal MRI were 62.5% and 44.0%, respectively. BOAI as a neoadjuvant chemotherapy has the possibility of bladder-preserving therapy in locally invasive bladder cancer. Also, the endorectal MRI can improve the accuracy of staging diagnosis, which is important for the bladder preservation.     

Magnetic resonance imaging with magnetic resonance cholangiopancreatography accurately predicts resectability of pancreatic carcinoma

Accurate preoperative staging of pancreatic malignancy aids in directing appropriate therapy and avoids unnecessary invasive procedures. We evaluated the accuracy of magnetic resonance imaging (MRI) with magnetic resonance cholangiopancreatography (MRCP) in determining resectability of pancreatic malignancy. Twenty-one patients with suspected pancreatic malignancy underwent dynamic, contrast-en-hanced breath-hold MRI with MRCP prior to surgical evaluation. Results of this study were correlated with operative results and pathologic findings. The sensitivity, specificity, and accuracy of MRI with MRCP in detecting a mass, determining the nature of the mass, and predicting lymph node involvement and resectability were determined. MRI with MRCP correctly identified the presence of a pancreatic mass in all 21 of these patients. Following pathologic correlation, it was determined that MRI with MRCP was 81 % accurate in determining the benign or malignant nature of the pancreatic mass and 43% accurate in predicting lymph node involvement. In predicting resectability, MRI with MRCP had a sensitivity of 100%, specificity of 83%, positive predictive value of 94%, negative predictive value of 100%, and accuracy of 95%. MRI with MRCP is an accurate, noninvasive technique in the preoperative evaluation of pancreatic malignancy. Information obtained from MRI with MRCP including identification of a mass and predicting tumor resectability may be of value in staging and avoiding unnecessary invasive diagnostic procedures in patients with pancreatic cancer. Presented at the Thirty-Ninth Annual Meeting of The Society for Surgery of the Alimentary Tract, New Orleans, La., May 17–20, 1998.     

Differential diagnosis of bone and soft tissue tumors by MRI

Musculoskeletal neoplasms are rare, and both the medical history and complaints of the patients are usually uncharacteristic and of limited information. After a clinical evaluation and biplane conventional radiography, the clinician must classify the patient as having a nonprogressive or progressive primary benign, primary malignant, or metastatic bone tumor. In the case of a probably benign, nonprogressive bone tumor, the patient has to be observed continuously or an additional biopsy should be performed. In the case of a probably malignant lesion, the patient should be referred for further staging and treatment to an orthopedic oncologist. Conventional biplane radiography, scintiscan, computed tomography scan, and magnetic resonance imaging (MRI) are indispensable in staging and treatment planning for patients with musculoskeletal tumors. For limb salvage procedures, delineation of the tumor from adjacent tissue structures is crucial. Hence, MRI of the entire anatomic structure involved, together with adjacent joints, is of the utmost importance, both in the coronal and axial planes. The significance of MRI in clinical follow-up depends on keeping the sequences and imaging planes used constant. Differentiating pseudotumors from true neoplasms still poses a challenge. The cellular pattern and matrix characteristics of a lesion cannot definitely be identified as neoplastic even with application of all imaging modalities including MRI. Information on epidemiology, clinical picture, radiology, and histology of the lesion is necessary to draw a firm conclusion. Biopsy is still the first choice in making the diagnosis.     

Can dynamic gadolinium-enhanced magnetic resonance imaging with chemical shift studies predict the status of adrenal masses?

Endoscopic adrenalectomy represents the new gold standard in the surgical treatment of benign adrenal lesions up to 6 cm. In some cases lesions larger than 10 cm have been removed laparoscopically to offer the patient the advantages of the minimally invasive technique. The larger the diameter of an adrenal lesion, the greater the probability of malignancy. In a prospective study 130 consecutive patients (88 women, 42 men; mean age 47.8 years) with 137 adrenal lesions earmarked for surgery underwent preoperative gadolinium-enhanced magnetic resonance imaging (MRI) with chemical shift studies (CSS). The aim of this study was to predict the status (benign, borderline, malignant) of adrenal lesions by MRI irrespective of tumor size. There were 14 patients with malignant tumors, 3 had borderline tumors (epithelial tumors with high malignant potential), and the remaining 120 had benign adrenal lesions. Five malignant lesions (36%) had a diameter < 6 cm. MRI correctly predicted 11 of 14 malignant tumors (1 malignant pheochromocytoma and 2 adrenocortical carcinomas had false-negative results), 117 of 120 benign lesions, and 2 of 3 borderline lesions. All but two malignant tumors were operated on using open surgery; 82 (68%) of 120 benign adrenal lesions were treated using the transperitoneal laparoscopic approach. Tumor size alone is not suitable for predicting the status of adrenal lesions. Dynamic gadolinium-enhanced MRI with CSS can predict the status of at least 95% of adrenal lesions. Tumors > 6 cm classified as benign by preoperative MRI may be removed laparoscopically by endocrine surgeons experienced in endoscopic adrenalectomy.     

Magnetic resonance imaging of soft tissue masses

Twenty-nine soft tissue masses were studied with magnetic resonance imaging (MRI) which proved to be useful in the preoperative evaluation of these lesions. Other imaging modalities employed had significant limitations. Plain films were of little value because of the intrinsically low contrast of soft tissues. Angiography was not necessary unless MRI suggested a vascular lesion or proximity to major blood vessels. Computed tomography (CT) and MRI both readily identified fatty lesions and their relationship to adjacent structures. Some soft tissue tumors could not be delineated from normal muscle with CT, but were easily seen with MRI. MRI is ideally suited for the study of suspected soft tissue tumors because of its excellent soft tissue contrast and its ability to image directly in any plane. Optimum evaluation required imaging in at least two planes with spin echo sequences chosen to bring out both T1 and T2 features.     

Preoperative magnetic resonance staging of rectal cancer with an endorectal coil and dynamic gadolinium enhancement

BACKGROUND: The aim of the study was to assess the value of endorectal coil magnetic resonance imaging (MRI) with gadolinium enhancement in the preoperative staging of rectal cancer. METHODS: In addition to standard evaluation, patients with rectal lesions were assessed by MRI obtained with a pelvic phased-array coil in combination with an endorectal coil. RESULTS: The study group comprised 29 patients with rectal cancer staged with an endorectal coil who had surgery without preoperative adjuvant therapy. In addition to standard T1- and T2-weighted images, dynamic contrast-enhanced images were acquired in all patients. Considerable interobserver variation was noted, particularly for pathological tumour stage pT1 or pT2 (kappa = 0.36). Compared with pathological findings, endorectal MRI correctly staged nine patients, overstaged 16 and understaged four. Whilst lymph node metastases were accurately detected in 70 per cent of patients, the positive predictive value was only 58 per cent. CONCLUSION: MR staging of rectal cancer with an endorectal coil and gadolinium enhancement is inaccurate for early tumours (stage T1 or T2) and is associated with a considerable degree of interobserver variation for individual scan sequences.