Effect of superparamagnetic iron oxide on bone marrow

The goal of this study was to compare the effects of SPIO particles on the signal intensity of the bone marrow of the vertebra spine in patients with and without liver cirrhosis. Forty-eight patients with normal liver tissue and 56 patients with liver cirrhosis were examined before and after intravenous SPIO administration, using a 1.5-T system (Magnetom Vision, Siemens, Erlangen, Germany) with a semiflexible cp-array coil. Three different pulse sequences were applied: a T1-weighted gradient-echo sequence, a T2-weighted fast spin-echo sequence with spectral fat suppression and a T2^*-weighted gradient-echo sequence. The signal-to-noise ratio (SNR) of the liver, vertebra bone and paraspinal muscle were obtained. The SNR value change in each patient group and the SNR value difference between the two groups were evaluated. For assessment of statistical significance, Student's t-test with a level of p < 0.05 was applied. No significant differences in the SNR values of the liver and bone marrow between the two groups could be seen with any of the three sequences precontrast. Using the T1-weighted gradient-echo sequence in the noncirrhotic liver group, pre- and postcontrast comparisons of the SNR values of the liver and bone marrow indicated a decrease of approximately –44.3 % (p = 0.02) and increase of approximately 15.3 % (p = 0.04), respectively. No significant change was seen in the cirrhotic liver group. With the T2-weighted fast spin-echo sequence, a significant decrease of the SNR value of the liver and the bone marrow in both groups was seen. With the T2^*-weighted gradient-echo sequence, the signal intensity decrease of the normal liver tissue was approximately –65.6 % (p = 0.00), in cirrhotic liver tissue the decrease was –29.9 % (p = 0.02). The SNR values of the bone marrow showed a decrease of –27.8 % (p = 0.04) in the noncirrhotic liver group, whereas in the cirrhotic liver group it was only –11.3 % and statistically not significant. The effect of SPIO particles on the liver and bone marrow is significantly less in patients with liver cirrhosis. Received: 2 April 1999; Revised: 5 October 1999; Accepted: 2 February 2000     

Superparamagnetic iron oxide-mediated hepatic signal intensity change in patients with and without cirrhosis: pulse sequence effects and Kupffer cell function

PURPOSE: To analyze superparamagnetic iron oxide (SPIO)-mediated hepatic signal intensity change in cirrhotic and noncirrhotic liver and to investigate the relationship between pulse sequence effects in SPIO-enhanced magnetic resonance (MR) imaging for hepatic cirrhosis. MATERIALS AND METHODS: Twelve patients with and 12 patients without cirrhosis underwent T2-weighted fast spin-echo, T2*-weighted gradient-echo (GRE), and T1-weighted GRE MR imaging before and twice (early and late phase) after SPIO administration. To assess the effect of SPIO, postcontrast relative signal-to-noise ratio (SNR) changes were statistically analyzed with repeated measurements analysis of variance for each pulse sequence. RESULTS: No interaction was shown between groups and data time points for any pulse sequence. There was no significant difference in mean hepatic relative SNR change on T2-weighted fast spin-echo images between the cirrhotic group and noncirrhotic group (-38.6% and -40.7%, early phase; -42.2% and -49.6%, late phase, respectively). For GRE images, statistically significant differences in mean hepatic relative SNR change were found between the cirrhotic group and noncirrhotic group (-14.2% and -44.5%, early phase; -28.5% and -56.4%, late phase on T2*-weighted GRE images (P <.001); 31.8% and 12.9%, early phase; 23.8% and 2.2%, late phase on T1-weighted GRE images (P <.05), respectively. CONCLUSION: Decreased overall phagocytic activity in cirrhotic liver is more likely due to Kupffer cell dysfunction than to Kupffer cell depletion, since magnetic susceptibility effects on T2*-weighted GRE images depend on intracellular SPIO cluster size.     

Malignant lesions of the liver identified on T1- but not T2-weighted MR images at 1.5 T

The authors reviewed their 21/2-year experience with a magnetic resonance (MR) imaging protocol for a 1.5-T MR imager that included T2-weighted fat-suppressed spin-echo, T1-weighted breath-hold gradient-echo, and serial dynamic gadolinium-enhanced T1-weighted gradient-echo imaging to identify histologic types of malignant liver lesions more apparent on T1- than on T2-weighted images. MR images of 212 consecutive patients with malignant liver lesions were reviewed. T2-weighted, T1-weighted, and dynamic contrast-enhanced T1-weighted images were examined separately in a blinded fashion. Seven patients demonstrated liver lesions (lymphoma [two patients] and carcinoid, hepatocellular carcinoma, colon adenocarcinoma, transitional cell carcinoma, and melanoma [one patient each]) on T1-weighted images that were inconspicuous on T2-weighted images. In all cases, the lesions were most conspicuous on T1-weighted images obtained immediately after administration of contrast agent. Histologic confirmation was present for all seven patients. The consistent feature among these lesions was that they were hypovascular, due either to a fibrous stroma or to dense monoclonal cellularity. These results suggest that in some patients with hypovascular primary neoplasms, the lesions may be identified only on T1-weighted images, and that immediate postcontrast T1-weighted images are of particular value in demonstrating lesions.     

Gadobenate dimeglumine (BOPTA) enhanced MR imaging: Patterns of enhancement in normal liver and cirrhosis

To determine whether gadobenate dimeglumine (BOPTA) will adequately enhance cirrhotic liver parenchyma, and to document the enhancement patterns in cirrhosis, 14 cirrhotic and 20 non-cirrhotic patients were evaluated before and 60–120 minutes after gadolinium-BOPTA. Proof of liver cirrhosis was biopsy (6), surgical resection (3), and clinical follow-up (5). Enhancement effects were compared quantitatively by determining the liver signal-to-noise ratio (SNR) and signal enhancement in both populations. Qualitatively assessment of the liver enhancement was performed and classified as homogeneous or heterogeneous. Quantitative analysis: cirrhotic liver parenchyma presented a higher increase in SNR values, relative to noncirrhotic liver parenchyma, on postcontrast images. Likewise the signal enhancement of cirrhotic liver parenchyma was superior to non-cirrhotic liver on T1-weighted SE images (P = .02) and in-phase GRE images (P < .001). There was no statistical difference on out-of-phase GRE images. Qualitative analysis: on T1-weighted SE postcontrast images, cirrhotic liver parenchyma showed a homogeneous enhancement in 7 patients and heterogeneous in 7. Whereas on GRE images, cirrhotic parenchyma showed heterogeneous enhancement in 9 patients and homogeneous in 5 patients. The heterogeneous enhancement was due to the presence of hypointense nodules in 7 patients and hyperintense nodules in 2 patients. In conclusion, our study has shown that the hepatobiliary contrast agent Gd-BOPTA is effective in the cirrhotic liver, demonstrating an increased liver enhancement compared with non-cirrhotic patients.     

Multifocal nodular fatty infiltration of the liver mimicking metastatic disease on CT: imaging findings and diagnosis using MR imaging

The aim of this study was to describe the MR appearance of multifocal nodular fatty infiltration of the liver (MNFIL) using T1-weighted in-phase (IP) and opposed-phase (OP) gradient-echo as well as T2-weighted turbo-spin-echo sequences with fat suppression (FSTSE) and without (HASTE). Magnetic resonance imaging examinations at 1.5 T using T1-weighted IP and OP-GRE with fast low angle shot (FLASH) technique, and T2-weighted FSTSE, T2-weighted HASTE of 137 patients undergoing evaluation for focal liver lesions were reviewed. Five patients were identified in whom CT indicated metastatic disease; however, no liver malignancy was finally proven. Diagnosis was confirmed by biopsy (n = 3), additional wedge resection (n = 1) or follow-up MRI 6–12 months later (n = 5). Regarding the identified five patients, the number of focal liver lesions was 2 (n = 2) and more than 20 (n = 3). The MR imaging characteristics were as follows: OP-image: markedly hypointense (n = 5); IP image: isointense (n = 2) or slightly hyperintense (n = 3); T2-weighted FSTSE-image: isointense (n = 5); T2-weighted HASTE image isointense (n = 1); slightly hyperintense (n = 4). On OP images all lesions were sharply demarcated and of almost spherical configuration (n = 5). Further evaluation by histology or follow-up MR imaging did not give evidence of malignancy in any case. Histology revealed fatty infiltration of the liver parenchyma in three patients. Magnetic resonance follow-up showed complete resolution in two patients and no change in three patients. Multifocal nodular fatty infiltration can simulate metastatic disease on both CT and MR imaging. The combination of in-phase (IP) and opposed-phase (OP) gradient-echo imaging can reliably differentiate MNFIL from metastatic disease. Received: 15 September 1999 Revised: 3 February 2000; Accepted: 7 February 2000     

Accuracy of liver fat quantification at MR imaging: comparison of out-of-phase gradient-echo and fat-saturated fast spin-echo techniques--initial experience

PURPOSE: To retrospectively determine the relative accuracy of liver fat quantification with out-of-phase gradient-echo magnetic resonance (MR) imaging and fat-saturated fast spin-echo MR imaging in patients with and without cirrhosis, with histologic analysis as the reference standard. MATERIALS AND METHODS: Committee on Human Research approval was obtained. Patient consent was not required. Data collection ended before HIPAA regulations were implemented, but patient anonymity was maintained. Twenty-seven patients, 16 with cirrhosis, were retrospectively identified who underwent MR imaging before histopathologic evaluation of liver fat at biopsy or surgery. The patient population consisted of 15 male and 12 female patients (mean age, 55 years; range, 16-75 years). One radiologist blinded to the histopathologic results recorded mean signal intensity derived from three regions of interest placed in the right and left lobes of the liver on three sections and signal intensity of the spleen from one region of interest within the same section. Liver fat was quantified with the relative loss of signal intensity on out-of-phase images compared with that on in-phase T1-weighted gradient-echo images and with relative loss of signal intensity on T2-weighted fast spin-echo MR images obtained with fat saturation compared with those obtained without fat saturation. Hotelling t test was used to compare correlation coefficients between relative signal intensity differences and histopathologically determined percentage of fat. RESULTS: In patients without cirrhosis, liver fat quantification with fat-saturated fast spin-echo MR imaging was significantly better than it was with out-of-phase gradient-echo MR imaging (r = 0.92 vs 0.69, P < .01). In patients with cirrhosis, liver fat quantification was correlated only with fat-saturated fast spin-echo MR imaging (r = 0.76, P < .01); the relative signal intensity loss on out-of-phase gradient-echo MR images was not correlated with histopathologically determined percentage of fat (r = 0.25, P = .36). CONCLUSION: Preliminary results suggest liver fat may be more accurately quantified with fat-saturated fast spin-echo MR imaging than with out-of-phase gradient-echo MR imaging, especially in patients with cirrhosis.     

Malignant lesions of the liver with high signal intensity on T1-Weighted MR images

Our purpose was to identify the histologic types of malignant liver lesions with high signal intensity (SI) on T1-weighted images and to describe the MR imaging features. Thirteen patients with malignant liver lesions high in SI on T1-weighted images were studied with a 1.5-T MR imager using pre- and serial postcontrast spoiled gradient-echo (SGE) sequences (all patients), T2-weighted fat-suppressed spin-echo sequences (all patients), precontrast T1-weighted fat-suppressed spin-echo sequences (five studies in five patients), and precontrast out-of-phase SGE sequences (seven studies in six patients). Images were reviewed retrospectively to determine number of lesions; lesion size; SI of lesions on T1-weighted, T2-weighted, and fat-attenuated T1-weighted images; distribution of high SI in lesions on T1-weighted images; and tumor enhancement pattern. Seven patients had multiple tumors high in SI on T1-weighted images and six patients had solitary tumors. Seventy-two lesions were less than 1.5 cm in diameter and 35 lesions were more than 1.5 cm in diameter. Nine patients had solid malignant lesions and four patients had cystic malignant lesions. All tumors more than 1.5 cm in diameter were heterogeneously high in SI on T1-weighted images, and all tumors less than 1.5 cm were completely homogeneous or homogeneous with a small central hypointense focus. All tumors were more conspicuous on T1-weighted fat-attenuated images, both on excitation spoiled fat-suppressed spin-echo or on out-of-phase SGE images with the exception of one fat-containing hepatocellular carcinoma (HCC). In one patient with melanoma metastases and one patient with multiple myeloma nodules, appreciably more lesions were detected on out-of-phase SGE images. Causes of hyperintensity were considered to be either fat, melanin, central hemorrhage, or high protein content, all of which may be seen in a variety of tumors. Fat-attenuation techniques are helpful in the detection of these lesions.     

Superparamagnetic Iron Oxide Hepatic MR Imaging: Efficacy and Safety Using Conventional and Fast Spin-Echo Pulse Sequences

The purpose of this study was to evaluate the technical efficacy and safety of iv ferumozldes (Feridex^a), a superparamagnetic iron oxide contrast agent for detection of hepatic lesions using conventional spin-echo and fast spin-echo MR images. Precontract and postcontrast MR studies were performed on 25 patients with suspected focal hepatic lesions. Conventional Tl-and T2-weighted MR images, as well as fast spin-echo and fat suppressed fast spin-echo MR images, were evaluated. Quantitative assessment of the contrast agent was performed obtaining region of interest measurements of the liver, spleen, and selected hepatic lesions. The pulse sequences were also evaluated subjectively for overall image quality and a subjective assessment of lesion detection. The use of a superparamagnetic iron oxide contrast agent led to a decrease in hepatic signal intensity on all pulse sequences. Lesion-to-liver contrast increased 41.1%, 36.5%, and 32.0% on the conventional T2, fast spin-echo, and fat suppressed fast spin echo pulse sequences, respectively. Lesion-to-liver contrast decreased on the T1-weighted postcontrast pulse sequence by 23.8%. Despite Improvement in lesion-to-liver contrast, radiologists subjectively preferred the precontract sequences because of overall better image quality. At a dose of 10 μmol/kg, fenunoxides favorably impacts lesion-to-liver contrast, and may be useful in hepatic imaging, more with conventional T2-weighted spin-echo pulse sequences than with T2-weighted fast spin-echo pulse sequences.     

Characterization of focal fatty change in the liver with a fat-enhanced inversion-recovery sequence

A T1-weighted inversion-recovery (IR) sequence was used to study 15 patients with possible fatty change in the liver. The inversion time (TI) was calculated for optimal suppression of normal liver signal (t-null). Conventional spin-echo (SE) and short TI IR (STIR) sequences were also performed. For seven documented benign focal fatty liver lesions, the T1-weighted IR (fatenhanced)sequence clearly enabled differentiation of normal from fat-infiltrated liver, whereas three of these lesions were isointense to normal liver with all other sequences. The livers of the other nine patients (two normal, one with diffuse fatty change, two with metastatic disease, one with hemangioma, one with focal nodular hyperplasia, one with simple cyst, and one with micronodular cirrhosis) showed homogeneous reduction of liver signal with the fat-enhanced IR sequence.     

Sclerosed liver hemangioma mimicking malignant tumor at MR imaging: Pathologic correlation

This case report illustrates atypical magnetic resonance (MR) imaging findings in a liver hemangioma mimicking a malignant lesion—lower signal intensity than cerebrospinal fluid on T2-weighted spin-echo images and lack of early enhancement on dynamic contrast material—enhanced gradient-echo images. Pathologic analysis demonstrated nearly total replacement of the vascular cavities by dense fibrous tissue. In this rare, sclerosed form, this lesion could not be defined as a hemangioma with MR imaging.     

Detection of malignant hepatic lesions before orthotopic liver transplantation: accuracy of ferumoxides-enhanced MR imaging

OBJECTIVE: The purpose of this study was to evaluate the diagnostic accuracy of ferumoxides-enhanced MR imaging for screening malignant hepatic lesions before orthotopic liver transplantation. MATERIALS AND METHODS: The study comprised 48 patients who underwent MR imaging within 6 months before transplantation. Imaging techniques included unenhanced and ferumoxides-enhanced T1-weighted gradient-echo and T2-weighted fast spin-echo sequences and ferumoxides-enhanced T2(*)-weighted gradient-echo sequences. Qualitative and quantitative analyses were performed; the gold standard was the histopathologic reports of explanted livers. RESULTS: Twenty patients had malignant hepatic lesions, and 24 hepatocellular carcinomas were histopathologically proven. The mean area under the receiver operating characteristic curve and the mean sensitivity were significantly greater for the image sets with ferumoxides-enhanced gradient-echo sequences than for those without these sequences. The mean sensitivity and specificity of all sequences were 85% and 74% on a per-patient basis, respectively. The mean contrast-to-noise ratio was significantly greater for the ferumoxides-enhanced T2(*)-weighted gradient-echo sequences than for any other sequences and for the ferumoxides-enhanced T1-weighted gradient-echo sequences than for unenhanced sequences and the ferumoxides-enhanced T2-weighted fast spin-echo sequences. CONCLUSION: Ferumoxides-enhanced gradient-echo sequences improved the diagnostic accuracy and the sensitivity for detecting malignant hepatic lesions in patients with end-stage cirrhosis of the liver. However, the specificity was not improved even after the administration of ferumoxides because of the false-positive lesions that were mainly the result of fibrotic changes.     

Regenerating nodules of liver cirrhosis: MR imaging with pathologic correlation

To determine the pathologic basis for low-intensity nodules seen on MR images of the liver in patients with cirrhosis, we obtained spin-echo and gradient-echo MR images in 20 cirrhotic patients in whom partial hepatectomies were subsequently performed for hepatocellular carcinoma. Low-intensity liver nodules were shown on gradient-echo and spin-echo images in eight patients. Pathologic study of the liver in these patients showed that these nodules were regenerating nodules containing hemosiderin. Low-intensity nodules were seen only on T2-weighted spin-echo images in four other patients. Microscopic examination of the liver in these patients showed regenerating nodules without hemosiderin deposits. Broad fibrous septa containing vascular spaces were present in two of these four patients. These results suggest that regenerating nodules containing hemosiderin or those that are surrounded by vascular fibrous septa are visible on MR images as low-intensity nodules and that gradient-echo images are useful in demonstrating nodules with hemosiderin.     

Focal nodular hyperplasia of the liver on ferumoxides-enhanced MR imaging: features on conventional spin-echo, fast spin-echo and gradient-echo pulse sequences

The aim of this study was to assess the efficacy of a superparamagnetic iron oxide, ferumoxides, in the detection and characterization of focal nodular hyperplasia (FNH) on MR conventional spin-echo (SE), fast spin-echo (FSE) and gradient-echo (GRE) images. Fourteen adults with 27 FNHs were evaluated at 1.5 T before and after injection of ferumoxides. T1-weighted and T2-weighted SE, T2-weighted FSE and T2^*-weighted GRE sequences were used and analysed qualitatively and quantitatively. One hundred percent of FNHs showed a significant postcontrast decrease in signal intensity on T2- and T2*-weighted images. Heavily T2-weighted SE images showed the maximum decrease in FNH signal-to-noise ratio (S/N). Postcontrast GRE T2^*-weighted images improved the detection of the central scar and the delineation of FNHs and demonstrated the best lesion-to-liver contrast-to-noise ratio (C/N). Postcontrast T1-weighted SE images showed the least lesion-to-liver C/N. Ferumoxides-enhanced MR imaging can help detect and characterize FNH. Conventional pre- and postcontrast T2-weighted SE images and postcontrast GRE T2*-weighted images should be used preferentially. Received: 30 November 1998; Revised: 5 April 1999; Accepted: 6 April 1999     

Contrast-enhanced MR imaging of diffuse and focal splenic disease with use of magnetic starch microspheres

The diagnostic value of magnetic starch microspheres (MSM), a new superparamagnetic contrast agent, was studied in experimental models of diffuse and focal splenic disease in rats by means of ex vivo relaxometry and in vivo magnetic resonance (MR) imaging. Owing to small differences in unenhanced T1 and T2 values between diffuse lymphoma and normal spleen, MR imaging failed to distinguish tumor-bearing animals from control animals by signal-to-noise ratios (SNRs) obtained with T1- and T2-weighted spin-echo sequences. One hour after injection of 20 μmol/kg MSM, lymphomatous spleen showed significantly (P <.001) reduced enhancement relative to normal splenic tissue. As a result, animals with diffuse lymphoma (SNR: 10.3 ± 1.7) could be easily differentiated from control animals (SNR: 5.5 ± 0.6) on T2-weighted (TR msec/TE msec = 2,000/45) images. In focal splenic disease, MSM produced normal enhancement of nontumorous splenic tissue, whereas relaxation times of tumors were not different before and after contrast agent injection. On T2-weighted images (2,000/45), the tumor-spleen contrast-to-noise ratio increased from (4.8 ± 1.6 to 21.8 ± 1.9 +354%), improving conspicuity of splenic tumors. The results show that MSM-enhanced MR imaging improves the detection of diffuse and focal splenic disease.     

Cartilage destruction in small joints by rheumatoid arthritis: assessment of fat-suppressed three-dimensional gradient-echo MR pulse sequences in vitro

Purpose. To assess the accuracy of different MR sequences for the detection of articular cartilage abnormalities in rheumatoid arthritis. Design and patients. Ten metacarpophalangeal joints and 10 metatarsophalangeal joints (specimens from arthritis patients undergoing ablative joint surgery) were examined with a fat-suppressed (FS) 3D FLASH, a FS 3D FISP, a FS 2D fast spin-echo T2-weighted, and a 2D FS spin-echo T1-weighted sequence. Each cartilage lesion and each cortical lesion was graded from 0 to 4 (modified Outerbridge staging system). Subsequently, the results of each sequence were compared with the macroscopic findings and statistically tested against each other. Results. The study shows that 3D gradient-echo sequences with fat suppression were best for imaging and grading of cartilage lesions in arthritis of the small joints of the hands and feet. Using 3D techniques, all grade 2, grade 3, and grade 4 lesions of cartilage or cortical bone were detected. Conclusion. FS 3D gradient-echo techniques were best for the detection and grading of hyaline cartilage and subchondral bone lesions in rheumatoid arthritis. MRI has a great potential as an objective method of evaluating cartilage damage and bone erosions in rheumatoid arthritis.     

Comparison of lesion enhancement on spin-echo and gradient-echo images.

PURPOSETo compare lesion enhancement after injection of gadopentetate dimeglumine on spin-echo and gradient-echo T1-weighted images.METHODSA total of 48 contrast-enhancing intracranial lesions were evaluated using a spin-echo and two gradient-echo T1-weighted pulse sequences. Percent contrast, contrast-to-noise, and signal-to-noise measurements were made on the spin-echo T1-weighted, three-dimensional gradient-echo, and multiplanar gradient-echo sequences.RESULTSThe measurements were somewhat different for the following categories of lesions: extraaxial, intraaxial with edema, and intraaxial without edema. The latter group provided the greatest diagnostic challenge: three of 19 such lesions 1 cm in size or smaller could not be identified on three-dimensional gradient-echo images, and one could not be identified on multi-planar gradient-echo images. The spin-echo T1-weighted sequence demonstrated significantly higher percent contrast (P < .05) and greater contrast to noise (P < .03) than either gradient-echo sequence for these small intraaxial lesions without edema. For extraaxial and intraaxial lesions with edema, percent C was similar for spin-echo T1-weighted and three-dimensional gradient-echo images, while contrast to noise was greater for spin-echo T1-weighted images. This reflected greater tissue noise with gradient-echo sequences.CONCLUSIONThe T1-weighted spin-echo sequence was preferred for detecting the full spectrum of contrast-enhancing lesions of the central nervous system.     

Dynamic gadolinium-enhanced echo-planar MR imaging of the liver: Effect of pulse sequence and dose on enhancement

To develop guidelines for clinical magnetic resonance imaging of the liver, the authors undertook an animal study to investigate the effect of dose and pulse sequence on liver signal intensity in gadopentetate dimeglumine—enhanced echo-planar imaging. Serial imaging of the liver was performed in anesthetized rats after intravenous administration of five different doses (0.01, 0.05, 0.1, 0.2, and 0.5 mmol/kg) of contrast agent, with six different pulse sequences. The results show that gadopentetate dimeglumine—enhanced echo-planar images obtained during the perfusion phase can yield either positive (due to increased T1 relaxation rates) or negative (due to susceptibility-induced increased T2 relaxation rates) liver enhancement depending on choice of pulse sequence and dose. At the current clinically recommended dose of 0.1 mmol/kg, maximal liver signal enhancement was seen with a T1-weighted inversion-recovery sequence, while maximal liver signal diminution was seen with a T2*-weighted gradient-echo sequence. The authors conclude that gadopentetate dimeglumine—enhanced echo-planar imaging can provide T1, T2, and T2* contrast that may be exploited for both lesion detection and lesion characterization.     

Rectal tumour staging: MR imaging using pelvic phased-array and endorectal coils vs endoscopic ultrasonography

The aim of this study was to compare MR imaging and endoscopic ultrasonography (EUS) for the local staging of rectal tumours. Forty-nine patients were examined on a 1.5-T MR unit using either a pelvic phased-array coil (n = 37) alone or combined with an endorectal coil (n = 12). Sagittal and axial sequences with T2-weighted fast spin-echo and axial T1-weighted spin-echo techniques were employed. The EUS technique was performed using a flexible endosonoscope. The results were compared with findings at histopathological sectioning of the specimen. The T-stage on MR correlated with histopathology in 32 of 49 patients and on EUS in 29 of 49 patients. The N-stage on MR correlated with histopathology in 22 of 49 patients and on EUS in 26 of 49 patients. Tumour penetration of the rectal wall was predicted by MR with 86 % sensitivity and 65 % specificity, and by EUS with 89 % sensitivity and 33 % specificity. Preoperative radiotherapy was administered to 40 of the patients after the examinations which may explain some of the overstaging by MR and EUS. Three patients with surgically and histopathologically confirmed invasion of neighbouring organs in the pelvis were detected preoperatively on MR but none on EUS. Tumour penetration of the rectal wall and local lymph node metastases cannot accurately be predicted with MR or EUS. Magnetic resonance, however, seems to be more useful for preoperative identification of clinically occult advanced disease. Received: 18 February 1999; Revised: 17 September 1999; Accepted: 20 September 1999     

Arterioportal shunts in cirrhotic patients: evaluation of the difference between tumorous and nontumorous arterioportal shunts on MR imaging with superparamagnetic iron oxide

OBJECTIVE: The study objective was to distinguish between the features of tumorous and nontumorous arterioportal shunts on superparamagnetic iron oxide-enhanced MR imaging in patients with cirrhosis. SUBJECTS AND METHODS: Ten arterioportal shunts in eight patients, including four tumorous and six nontumorous arterioportal shunts, were evaluated on T2-weighted turbo spin-echo and T2(*)-weighted gradient-echo sequences before and after administration of superparamagnetic iron oxide. Qualitatively, the relative signal intensity of the arterioportal shunt compared with that of the surrounding liver parenchyma was categorized into three grades: high, slightly high, and not detected. Quantitatively, signal-to-noise ratio, contrast-to-noise ratio, lesion-to-liver contrast, and percentage enhancement were calculated and compared between tumorous and nontumorous arterioportal shunts by a nonparametric statistical test (Mann-Whitney test). RESULTS: Qualitatively, all four tumorous arterioportal shunts appeared as areas of slightly high or high intensity without and with superparamagnetic iron oxide on T2-weighted turbo spin-echo images and changed from isointensity to high intensity after the administration of superparamagnetic iron oxide on T2(*)-weighted gradient-echo images. All nontumorous arterioportal shunts except one could not be recognized without or with superparamagnetic iron oxide on either sequence. Quantitatively, with superparamagnetic iron oxide the contrast-to-noise ratio and the lesion-to-liver contrast of the tumorous arterioportal shunts were significantly higher than those of the nontumorous arterioportal shunts. CONCLUSION: Tumorous arterioportal shunts are seen as areas of reduced signal loss, whereas most nontumorous arterioportal shunts are seen as areas of normal signal loss, like the normal liver parenchyma. The difference is more marked on T2(*)-weighted gradient-echo images than on T2-weighted turbo spin-echo images.