Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print January 22, 2004, 10.1148/radiol.2303020921
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2303020921v1
230/3/637    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hussain, H. K.
Right arrow Articles by Francis, I. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hussain, H. K.
Right arrow Articles by Francis, I. R.
(Radiology 2004;230:637-644.)
© RSNA, 2004

Gastrointestinal Imaging

T2-weighted MR Imaging in the Assessment of Cirrhotic Liver1

Hero K. Hussain, MD, Ibrahim Syed, MD2, Hanh V. Nghiem, MD, Timothy D. Johnson, PhD, Ruth C. Carlos, MD, MS, William J. Weadock, MD and Isaac R. Francis, MD

1 From the Departments of Radiology (H.K.H., I.S., H.V.N., R.C.C., W.J.W., I.R.F.) and Biostatistics (T.D.J.), University of Michigan Hospitals, 1500 E Medical Center Dr, MRI B2B311, Ann Arbor, MI 48109-0030. From the 2001 RSNA scientific assembly. Received July 26, 2002; revision requested September 10; final revision received July 3, 2003; accepted July 31. Address correspondence to H.K.H. (e-mail: hhussain@umich.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
PURPOSE: To assess if T2-weighted magnetic resonance (MR) imaging provides added diagnostic value in combination with dynamic gadolinium-enhanced MR imaging in the detection and characterization of nodular lesions in cirrhotic liver.

MATERIALS AND METHODS: Two readers retrospectively and independently analyzed 54 MR imaging studies in 52 patients with cirrhosis. In session 1, readers reviewed T1-weighted and dynamic gadolinium-enhanced images. In session 2, readers reviewed T1-weighted, dynamic gadolinium-enhanced, and respiratory-triggered T2-weighted fast spin-echo images. Readers identified and characterized all focal lesions by using a scale of 1–4 (1, definitely benign; 4, definitely malignant). Multireader correlated receiver operating characteristic (ROC) analysis was employed to assess radiologist performance in session 2 compared with session 1. The difference in the areas under the ROC curves for the two sessions was tested. In a third session, readers assessed conspicuity of biopsy-proved lesions on T2-weighted MR images by using a scale of 1–3 (1, not seen; 3, well seen) and identified causes of reduced conspicuity.

RESULTS: Two additional benign lesions were detected by each reader in session 2. Fifty-five lesions had pathologic verification, including 32 malignant, three high-grade dysplastic, and 20 benign nodules. There was no significant difference in the area under the ROC curves between the two sessions (P = .48). Thirty-two lesions were inconspicuous on T2-weighted MR images because of parenchymal heterogeneity, breathing artifacts (particularly in patients with ascites), and lesion isointensity with liver parenchyma. T2-weighted MR imaging was useful in the evaluation of cysts and lymph nodes.

CONCLUSION: T2-weighted MR imaging does not provide added diagnostic value in the detection and characterization of focal lesions in cirrhotic liver.

© RSNA, 2004

Index terms: Liver, cirrhosis, 761.288 • Liver neoplasms, 761.31, 761.323, 761.33 • Liver neoplasms, MR, 761.121411, 761.121412, 761.121413, 761.121415, 761.12143


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Prior to the introduction of contrast material–enhanced magnetic resonance (MR) imaging, T2-weighted imaging MR was the mainstay of hepatic MR imaging. Once dynamic gadolinium-enhanced imaging was introduced, it became an adjunct to the T2-weighted sequence for the detection and characterization of liver lesions, and several investigators reported the comparative accuracies of the two sequences (1–7). As the quality of fast sequences and contrast timing methods improved, dynamic gadolinium-enhanced MR imaging became more reliable, and the early and delayed enhancement patterns of hepatic lesions became important criteria for lesion characterization, in addition to signal intensity characteristics on T2-weighted MR images (1,5,6).

The main purpose of imaging in cirrhosis is to identify hepatocellular carcinoma (HCC) and distinguish it from other nonmalignant nodular lesions—namely, dysplastic and regenerating nodules that are prevalent in cirrhotic liver. Early reports suggested that on T2-weighted MR images, HCC always displayed high or equivalent signal intensity compared with liver parenchyma (8–10). Later studies showed that there is a major overlap in the signal intensity characteristics of regenerating, dysplastic, and HCC nodules on T2-weighted MR images (11–14), and the distinction between these lesions could not be made reliably by using the signal intensity characteristics alone. Furthermore, other benign processes in the cirrhotic liver, such as infarction and confluent hepatic fibrosis, can mimic HCC at T2-weighted MR imaging (15,16).

Arterial enhancement is the most common and important imaging characteristic of HCC (14,17). While considered a reliable feature, it can occasionally be seen in dysplastic and regenerating nodules, postbiopsy arteriovenous shunts, nontumorous arterioportal shunts, metastasis, and benign lesions such as cavernous hemangiomas and focal nodular hyperplasia (18–23). These benign lesions, however, are uncommon in cirrhotic liver (23).

The aim of our study was to determine if T2-weighted MR imaging provides added diagnostic value in combination with dynamic gadolinium-enhanced MR imaging in the detection and characterization of nodular lesions in cirrhotic liver.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Patient Selection
We retrospectively reviewed our institutional database for all hepatic MR examinations performed at our institution between October 1998 and September 2001 in patients with known cirrhosis. Only patients who underwent biopsy of focal lesions detected at MR imaging were included in this study. Fifty-four MR examinations performed in 52 patients, including 39 men and 13 women with a mean age of 54.7 years (range, 22–85 years), satisfied these criteria. Cirrhosis was caused by hepatitis C in 21 patients, hepatitis C and alcoholism in 12, alcoholism in four, hepatitis B in six, hepatitis B and C in two, steatosis in two, primary biliary cirrhosis in one, autoimmune hepatitis in one, and nodular regenerative hyperplasia of Budd-Chiari syndrome in one. The cause of cirrhosis was cryptogenic in two patients. Institutional review board approval was granted, and consent was waived because of the retrospective nature of the study.

MR Imaging
MR imaging was performed with a 1.5-T system (Signa Echospeed, LX version 8.3; GE Medical Systems, Milwaukee, Wis). The following sequences were used in all patients: transverse T1-weighted spoiled gradient-recalled-echo (SPGR) sequence with breath-hold and in-phase (repetition time msec/echo time msec, <180/1.8–2.4) and out-of-phase (<180/4.2–4.8) imaging, with (n = 8) or without (n = 46) use of a spin-echo (SE) sequence (<500/12); transverse T2-weighted fast SE sequence (<5,000/90) with respiratory triggering, echo train length of eight, and fat saturation; and dynamic gadolinium-enhanced breath-hold fat-suppressed T1-weighted SPGR two-dimensional (<200/1.2 with flip angle of 70°) or three-dimensional (3D) (<6/1.2 with flip angle of 12° and spectral fat suppression) sequence. The addition of a fat suppression pulse minimized the out-of-phase effect that resulted from using a minimum echo time of 1.2 msec.

Dynamic imaging was performed before and after intravenous injection of 20 mL of gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) or gadodiamide (Omniscan; Nycomed Amersham, Princton, NJ) in the arterial dominant, portal venous, and 2-minute delayed phases. Early in the study, dynamic imaging was performed by using a two-dimensional T1-weighted SPGR sequence (n = 15), and arterial phase images were acquired after a fixed delay of 15 seconds from the start of contrast material injection. Later dynamic examinations were performed by using a 3D T1-weighted SPGR sequence (n = 39), and arterial phase images were timed by using the automated contrast material–bolus detection technique (SmartPrep; GE Medical Systems).

Image Analysis
Lesion detection and characterization.—Two experienced radiologists (readers 1 and 2, with 4 and 9 years of experience in interpretation of liver MR images) retrospectively and independently reviewed all 54 MR studies on a commercially available workstation (Advantage Workstation; GE Medical Systems). Readers were blinded to the initial clinical MR interpretations and biopsy results.

Images were analyzed in two sessions 4 weeks apart. In session 1, readers reviewed T1-weighted SPGR in- and out-of-phase images (with or without SE) and dynamic (precontrast, arterial dominant, portal venous, and delayed phase) images. In session 2, readers reviewed T1-weighted SPGR in- and out-of-phase images (with or without SE), dynamic (precontrast, arterial dominant, portal venous, and delayed phase) images, and T2-weighted fast SE (respiratory triggered with fat saturation) images.

In each of these sessions, readers were asked to identify all focal hepatic lesions and characterize them by using a scale of 1–4 (1, definitely benign; 2, probably benign; 3, possibly malignant; and 4, definitely malignant). Lesions that fulfilled all or most imaging criteria for cysts, hemangiomas, and regenerative and nontransformed dysplastic nodules were assigned a score of 1 or 2.

Hemangiomas were characteristically well-defined lesions with low signal intensity on T1-weighted images and very high signal intensity on T2-weighted images compared with liver parenchyma. One of three enhancement patterns were present (24): (a) early uniform enhancement and delayed contrast material retention, (b) early peripheral nodular enhancement with centripetal progression to complete filling and delayed contrast material retention, or (c) early peripheral nodular enhancement with centripetal progression to incomplete filling and delayed contrast material retention with nonenhancing central scar.

Cysts were characteristically well-defined lesions with low signal intensity on T1-weighted images and very high signal intensity on T2-weighted images compared with liver parenchyma and no internal enhancement after contrast material administration. Signal intensity of regenerative nodules was lower than or equivalent to that of liver parenchyma on T1- and T2-weighted images, and enhancement was similar to that of liver parenchyma. Nontransformed dysplastic nodules had variable signal intensity on T1-weighted images and had equivalent or low signal intensity on T2-weighted images with no arterial enhancement or delayed contrast-enhanced hypointensity compared with liver parenchyma.

Lesions with some or all features suggestive of malignancy were assigned a score of 3 or 4. These features were high signal intensity on T2-weighted images, excluding cysts and hemangiomas, and/or arterial enhancement. Other features suggestive of malignancy included lesion hypointensity compared with liver parenchyma on delayed (2-minute) contrast-enhanced images, lesion heterogeneity, diameter more than 2 cm, and/or vascular invasion. Because of nonspecific imaging characteristics, identification of possible nontumorous arterioportal shunts was left to the readers’ discretion. These lesions manifest as peripheral wedge-shaped or round areas of arterial enhancement that become isointense or remain mildly hyperintense compared with liver parenchyma in subsequent phases and are isointense on T1-weighted images and iso- or mildly hyperintense on T2-weighted images (20,21). Typical examples of regenerative, dysplastic, and HCC nodules were shown to the readers prior to image interpretation.

Lesion conspicuity and signal intensity.— To help identify the biopsy-proved lesion correctly, the readers were provided with a representative image (other than the T2-weighted image) that showed the lesion. This image was chosen by one author (I.S.) who was not involved in the reading sessions and was usually selected from the arterial dominant phase images. The biopsy-proved lesion was marked with an arrow, and the marked image was saved separately on the workstation.

In a third interpretation session, readers 1 and 2 assessed the conspicuity of all biopsy-proved lesions on T2-weighted MR images by using a scale of 1–3 (1, not seen; 2, partially obscured; and 3, well seen). For lesions with a score of 1 or 2, the reason for reduced conspicuity or nonvisualization was recorded as being due to predominantly one or a combination of the following factors: artifact from ascites, artifact from breathing, parenchymal inhomogeneity, and lesion isointensity with liver parenchyma. In this session, the readers also recorded the signal intensity of all biopsy-proved lesions on the T1-weighted, T2-weighted, arterial dominant phase, and 2-minute delayed contrast-enhanced images as being hyperintense, hypointense, or isointense compared with liver parenchyma. Disagreement between readers in this session was resolved by means of consensus review.

Statistical Analysis
Multireader receiver operating characteristic (ROC) analysis for correlated data (25) was employed for analysis of the 55 lesions with proved pathologic findings. Multireader ROC analysis pools the results from the readers and generalizes the results to the population of all readers. The areas under the ROC curves from both sessions were compared. The area under the ROC curve represents the probability that a randomly chosen malignant lesion is ranked with greater suspicion than a randomly chosen benign lesion.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Lesion Detection
Fifty-two patients with known cirrhosis underwent 54 MR examinations. In session 1, reader 1 detected 107 lesions, and reader 2 detected 111 lesions. The four additional lesions detected by reader 2 were described as benign (one cyst, three regenerative nodules). In session 2, reader 1 detected 109 lesions, and reader 2 detected 113 lesions. Both readers detected the same two extra lesions after the addition of the T2-weighted MR sequence. These lesions were characterized by both readers as simple cysts and were approximately 1 cm in diameter.

Fifty-five lesions were sampled with core needle biopsy for proof of pathologic findings. One patient underwent biopsy of two lesions, and two other patients underwent biopsy of one lesion twice at 6-month intervals; these lesions were treated as independent lesions. The 55 lesions included 30 HCCs, two adenocarcinomas, three high-grade dysplastic nodules (small cell dysplasia), 19 regenerative nodules, and one case of focal nodular hyperplasia. Because of the inability to reliably differentiate high-grade dysplastic nodules from HCC at imaging and because of the likelihood that high-grade dysplastic nodules will transform into HCC, high-grade dysplastic nodules were treated in the analysis as malignant lesions. Lesions varied in diameter from 0.5 to 16.7 cm (mean, 3 cm). Both readers detected all 55 biopsy-proved lesions in each of the two sessions. The subsequent analyses are based on 55 biopsy-proved lesions in 52 patients.

Lesion Characterization
Reader rankings of benign and malignant lesions are shown in Table 1, and the changes in lesion characterization for both readers after the addition of the T2-weighted sequence are shown in Table 2. The estimated area under the ROC curves was 0.70 in session 1 and 0.73 in session 2. The 0.03 difference in area between the two sessions (standard error, 0.05) was not significant (P = .48; a P value of less than .05 was considered to indicate a statistically significant difference).


View this table:
[in this window]
[in a new window]

  		TABLE 1. Reader Confidence Rankings of Benign and Malignant Lesions

 

View this table:
[in this window]
[in a new window]

  		TABLE 2. Changes in Lesion Characterization after Addition of the T2-weighted MR Sequence

 
Considering the relatively small sample size of 55 lesions, it is important to determine if there is sufficient power to detect a difference when one really exists. By assuming that a difference of 0.10 in the area under the ROC curve is a relevant difference, there is 0.92 power to detect the observed difference of 0.03. With a relevant difference as small as 0.05, the power is still 0.66. The power curve for a relevant difference in the area under the ROC curve (area for session 2 minus area for session 1) from 0.03 to 0.2 is given in Figure 1.



View larger version (15K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1.  Power curve for correlated ROC analysis. Horizontal axis represents the difference in areas between the two ROC curves. Vertical axis is the power.

 
Both readers characterized seven lesions (six cysts, one hemangioma) as benign, with high confidence in both sessions. Reader 2 detected an additional simple cyst in session 1. Portal hilar and portocaval lymph nodes, a common finding in patients with cirrhosis, were more discernable on the T2-weighted MR images because of their slight hyperintensity compared with liver parenchyma.

Two patients returned for repeat imaging and repeat biopsy of a liver lesion suspected of being malignant. These were treated as separate lesions. The first patient had a 5-cm lesion in the hepatic dome that was scored as 4 (definitely malignant) by both readers in both sessions, but biopsy results were negative for malignancy. The patient underwent repeat imaging 6 months later. Imaging findings of a mass suspected of being HCC prompted repeat biopsy. The second biopsy was positive for malignancy. The second patient had a 5-cm lesion in the anterior segment of the right hepatic lobe with imaging features highly suggestive of malignancy. The lesion was scored as 4 by both readers in both sessions. Biopsy results were negative for malignancy. The patient underwent repeat imaging 6 months later, and the lesion had become larger in size and had portal vein invasion by tumor. This mass was scored as 4 by both readers in both sessions. The lesion was sampled for biopsy again, and results remained negative for malignancy; however, the lesion was clinically presumed to be malignant and was treated as such.

Lesion Conspicuity and Signal Intensity
On T2-weighted MR images, 23 of the 55 (42%) biopsy-proved lesions were graded as 3 (well seen), fifteen (27%) were graded as 2 (partially obscured), and 17 (31%) were graded as 1 (not seen). The predominant causes for reduced lesion conspicuity and nonvisualization were artifacts from ascites (n = 7) or breathing (n = 5), underlying liver parenchymal heterogeneity (n = 14) (Fig 2), and lesion isointensity with liver parenchyma (n = 6) (Fig 3). More than one cause was identified in 15 patients.



View larger version (110K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a.  MR images show a 4-cm HCC in the right lobe of the liver. (a) Transverse T1-weighted SPGR image (170/4.2 with 70° flip angle). Tumor (arrow) is hyperintense compared with liver parenchyma and has a central area of hypointensity, probably a central scar. (b) Transverse T2-weighted fast SE image (4,000/94 with respiratory triggering and fat suppression). Reduced lesion conspicuity (arrow) on this image is predominantly caused by heterogeneity of surrounding liver parenchyma and isointensity of the lesion compared with surrounding heterogeneous parenchyma. Note the hyperintense central scar. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (6/1.8 with 12° flip angle) shows marked enhancement of the lesion (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (6/1.8 with 12° flip angle). Lesion becomes hypointense compared with liver parenchyma. Note enhancing thin pseudocapsule (arrow) around the tumor. Central scar remains unenhanced.

 


View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b.  MR images show a 4-cm HCC in the right lobe of the liver. (a) Transverse T1-weighted SPGR image (170/4.2 with 70° flip angle). Tumor (arrow) is hyperintense compared with liver parenchyma and has a central area of hypointensity, probably a central scar. (b) Transverse T2-weighted fast SE image (4,000/94 with respiratory triggering and fat suppression). Reduced lesion conspicuity (arrow) on this image is predominantly caused by heterogeneity of surrounding liver parenchyma and isointensity of the lesion compared with surrounding heterogeneous parenchyma. Note the hyperintense central scar. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (6/1.8 with 12° flip angle) shows marked enhancement of the lesion (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (6/1.8 with 12° flip angle). Lesion becomes hypointense compared with liver parenchyma. Note enhancing thin pseudocapsule (arrow) around the tumor. Central scar remains unenhanced.

 


View larger version (129K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2c.  MR images show a 4-cm HCC in the right lobe of the liver. (a) Transverse T1-weighted SPGR image (170/4.2 with 70° flip angle). Tumor (arrow) is hyperintense compared with liver parenchyma and has a central area of hypointensity, probably a central scar. (b) Transverse T2-weighted fast SE image (4,000/94 with respiratory triggering and fat suppression). Reduced lesion conspicuity (arrow) on this image is predominantly caused by heterogeneity of surrounding liver parenchyma and isointensity of the lesion compared with surrounding heterogeneous parenchyma. Note the hyperintense central scar. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (6/1.8 with 12° flip angle) shows marked enhancement of the lesion (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (6/1.8 with 12° flip angle). Lesion becomes hypointense compared with liver parenchyma. Note enhancing thin pseudocapsule (arrow) around the tumor. Central scar remains unenhanced.

 


View larger version (135K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2d.  MR images show a 4-cm HCC in the right lobe of the liver. (a) Transverse T1-weighted SPGR image (170/4.2 with 70° flip angle). Tumor (arrow) is hyperintense compared with liver parenchyma and has a central area of hypointensity, probably a central scar. (b) Transverse T2-weighted fast SE image (4,000/94 with respiratory triggering and fat suppression). Reduced lesion conspicuity (arrow) on this image is predominantly caused by heterogeneity of surrounding liver parenchyma and isointensity of the lesion compared with surrounding heterogeneous parenchyma. Note the hyperintense central scar. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (6/1.8 with 12° flip angle) shows marked enhancement of the lesion (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (6/1.8 with 12° flip angle). Lesion becomes hypointense compared with liver parenchyma. Note enhancing thin pseudocapsule (arrow) around the tumor. Central scar remains unenhanced.

 


View larger version (112K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a.  MR images show 5-cm HCC in the liver dome. (a) Transverse T1-weighted SPGR image (185/4.2 with 70° flip angle). Lesion is isointense with liver parenchyma. (b) Transverse T2-weighted fast SE image (4,255/96 with respiratory triggering and fat suppression). No definite focal lesion is apparent. Reduced lesion conspicuity on this image is predominantly caused by isointensity of the lesion with heterogeneous liver parenchyma and artifact from ascites. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (7/1.7 with 12° flip angle) shows heterogeneous enhancement of the large dome lesion (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (7/1.7 with 12° flip angle). Lesion becomes hypointense compared with liver parenchyma. Note enhancing thin pseudocapsule (arrow) around the tumor.

 


View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b.  MR images show 5-cm HCC in the liver dome. (a) Transverse T1-weighted SPGR image (185/4.2 with 70° flip angle). Lesion is isointense with liver parenchyma. (b) Transverse T2-weighted fast SE image (4,255/96 with respiratory triggering and fat suppression). No definite focal lesion is apparent. Reduced lesion conspicuity on this image is predominantly caused by isointensity of the lesion with heterogeneous liver parenchyma and artifact from ascites. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (7/1.7 with 12° flip angle) shows heterogeneous enhancement of the large dome lesion (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (7/1.7 with 12° flip angle). Lesion becomes hypointense compared with liver parenchyma. Note enhancing thin pseudocapsule (arrow) around the tumor.

 


View larger version (151K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3c.  MR images show 5-cm HCC in the liver dome. (a) Transverse T1-weighted SPGR image (185/4.2 with 70° flip angle). Lesion is isointense with liver parenchyma. (b) Transverse T2-weighted fast SE image (4,255/96 with respiratory triggering and fat suppression). No definite focal lesion is apparent. Reduced lesion conspicuity on this image is predominantly caused by isointensity of the lesion with heterogeneous liver parenchyma and artifact from ascites. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (7/1.7 with 12° flip angle) shows heterogeneous enhancement of the large dome lesion (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (7/1.7 with 12° flip angle). Lesion becomes hypointense compared with liver parenchyma. Note enhancing thin pseudocapsule (arrow) around the tumor.

 


View larger version (150K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3d.  MR images show 5-cm HCC in the liver dome. (a) Transverse T1-weighted SPGR image (185/4.2 with 70° flip angle). Lesion is isointense with liver parenchyma. (b) Transverse T2-weighted fast SE image (4,255/96 with respiratory triggering and fat suppression). No definite focal lesion is apparent. Reduced lesion conspicuity on this image is predominantly caused by isointensity of the lesion with heterogeneous liver parenchyma and artifact from ascites. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (7/1.7 with 12° flip angle) shows heterogeneous enhancement of the large dome lesion (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (7/1.7 with 12° flip angle). Lesion becomes hypointense compared with liver parenchyma. Note enhancing thin pseudocapsule (arrow) around the tumor.

 
The signal intensity of all 55 biopsy-proved lesions is shown in Table 3. Fourteen of 30 (47%) HCCs were hyperintense on T2-weighted images, with 12 of 30 (40%) isointense and four of 30 (13%) hypointense compared with liver parenchyma. Interestingly, 10 of 19 (53%) regenerating nodules (Fig 4) and one of three (33%) dysplastic nodules were also hyperintense compared with liver parenchyma on T2-weighted images. Twenty-six of 30 (87%) HCCs showed arterial enhancement, and 27 of 30 (90%) became hypointense compared with liver parenchyma on the 2-minute delayed contrast-enhanced MR images.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Signal Intensity of 55 Biopsy-proved Lesions in Comparison to Liver Parenchyma

 


View larger version (116K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a.  MR images show 1.5-cm regenerating nodule in the right lobe of the liver. (a) Transverse T1-weighted SPGR image (180/4.2 with 70° flip angle). No lesion is apparent. Nodule is isointense with liver parenchyma. (b) Transverse T2-weighted fast SE image (4,750/94 with respiratory triggering and fat suppression). Lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (6.4/1.8 with 12° flip angle). Lesion enhances homogeneously (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (6.4/1.8 with 12° flip angle). Lesion is no longer apparent and has becomes isointense with liver parenchyma. This lesion disappeared in subsequent follow up MR studies at 3, 6, and 9 months.

 


View larger version (132K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b.  MR images show 1.5-cm regenerating nodule in the right lobe of the liver. (a) Transverse T1-weighted SPGR image (180/4.2 with 70° flip angle). No lesion is apparent. Nodule is isointense with liver parenchyma. (b) Transverse T2-weighted fast SE image (4,750/94 with respiratory triggering and fat suppression). Lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (6.4/1.8 with 12° flip angle). Lesion enhances homogeneously (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (6.4/1.8 with 12° flip angle). Lesion is no longer apparent and has becomes isointense with liver parenchyma. This lesion disappeared in subsequent follow up MR studies at 3, 6, and 9 months.

 


View larger version (145K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4c.  MR images show 1.5-cm regenerating nodule in the right lobe of the liver. (a) Transverse T1-weighted SPGR image (180/4.2 with 70° flip angle). No lesion is apparent. Nodule is isointense with liver parenchyma. (b) Transverse T2-weighted fast SE image (4,750/94 with respiratory triggering and fat suppression). Lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (6.4/1.8 with 12° flip angle). Lesion enhances homogeneously (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (6.4/1.8 with 12° flip angle). Lesion is no longer apparent and has becomes isointense with liver parenchyma. This lesion disappeared in subsequent follow up MR studies at 3, 6, and 9 months.

 


View larger version (144K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4d.  MR images show 1.5-cm regenerating nodule in the right lobe of the liver. (a) Transverse T1-weighted SPGR image (180/4.2 with 70° flip angle). No lesion is apparent. Nodule is isointense with liver parenchyma. (b) Transverse T2-weighted fast SE image (4,750/94 with respiratory triggering and fat suppression). Lesion (arrow) is hyperintense compared with surrounding liver parenchyma. (c) Transverse arterial dominant phase contrast-enhanced fat-suppressed 3D SPGR image (6.4/1.8 with 12° flip angle). Lesion enhances homogeneously (arrow). (d) Transverse 2-minute delayed contrast-enhanced fat-suppressed 3D SPGR image (6.4/1.8 with 12° flip angle). Lesion is no longer apparent and has becomes isointense with liver parenchyma. This lesion disappeared in subsequent follow up MR studies at 3, 6, and 9 months.

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
 
Investigators in several studies have compared the utility and accuracy of T2-weighted and dynamic gadolinium-enhanced MR sequences in cirrhotic and noncirrhotic liver, with varying results (1–7) favoring one over the other for focal lesion detection and characterization (2–7,26). A few reports have emerged that question the value of T2-weighted MR imaging of the liver (27–29). Lee et al (27) concluded that T2-weighted MR imaging provides added diagnostic information or confidence and that it was premature to omit it from liver imaging studies. Coulam et al (28), on the other hand, concluded that multiphase dynamic gadolinium-enhanced sequences alone depicted most of the clinically relevant focal liver lesions. These authors found that the addition of unenhanced T1- and T2-weighted sequences, while helpful for lesion characterization, increased lesion detection only minimally. Sheppard et al (29) suggested that T2-weighted imaging is useful for confirmation of the diagnosis of hemangioma but is not required otherwise.

Our study of 52 cirrhotic livers shows that lesion detection increased only minimally after the addition of the T2-weighted sequence. The two additional lesions detected on T2-weighted MR images were characterized as simple cysts. We also found that T2-weighted MR imaging in cirrhotic livers did not improve the reader’s ability to characterize nodular lesions in cirrhotic liver.

The T2-weighted MR sequence has been used as a key sequence for the detection and characterization of nodular lesions in cirrhotic liver (8–12). Some centers use the short inversion time inversion-recovery, or STIR, sequence instead of the T2-weighted fast SE sequence. STIR may be more sensitive for the depiction of lesions, but it introduces T1 tissue contrast, which may be confusing when characterizing lesions. In our study, we found that 32 of 55 (58%) solid lesions were not clearly discernable on T2-weighted images, including 17 lesions that were not visualized. This sequence is the longest in our liver imaging protocol, and it is the only non–breath-hold sequence, which makes it subject to breathing and motion artifacts. Despite the routine use of respiratory triggering in our study to reduce breathing artifacts (30), such artifacts—especially in patients with ascites—frequently degraded the image quality, resulting in reduced or no conspicuity of 12 lesions. Use of breath-hold T2-weighted MR sequences may help reduce breathing artifacts, but the value in improving lesion detection and characterization in cirrhotic liver has not been established (26). Sequences such as single-shot fast SE or half-Fourier acquisition turbo SE, while useful for the detection and characterization of cystic lesions, may obscure solid lesions because of the nature of the sequence (31).

The heterogeneity of the cirrhotic liver parenchyma as a result of nodular regeneration, fibrosis, and scarring resulted in reduced or no conspicuity of 14 lesions. Scar tissue has variable signal intensity on T2-weighted images, which is influenced by the chronicity of the process (ie, active fibrosis or mature fibrous tissue) (32), and confluent hepatic fibrosis can occasionally mimic HCC on T2-weighted MR images (16).

Nodular lesions in cirrhotic livers are separated into two broad categories: regenerative and dysplastic or neoplastic (33). Most regenerative nodules have signal intensity similar to that of liver parenchyma with all sequences (13,34–37). Dysplastic nodules have variable signal intensity on T1-weighted images and hypo- or isointensity on T2-weighted images compared with liver parenchyma (12,37–39). When siderotic, both regenerative and dysplastic nodules have low signal intensity on T1- and T2-weighted images (13,34–37). HCC nodules, on the other hand, display a wide range of signal intensities on T1- and T2-weighted images (8–12,14,22,23,36,39–46).

In our study, only 14 of 30 HCC nodules were hyperintense compared with liver parenchyma on T2-weighted images, and the remainder were iso- or hypointense. This isointensity resulted in reduced or no conspicuity of six lesions. It is possible that the heterogeneity and increased signal intensity of the cirrhotic liver parenchyma obscured the mildly hyperintense HCC nodules on T2-weighted images. The mean size of iso- and hypointense HCCs on T2-weighted images in our study was 4.5 cm (range, 1.8–16.7 cm); this observation is different from that of Kelekis et al (14), who found that small HCCs (<1.5 cm) are usually isointense with liver parenchyma on T1- and T2-weighted MR images. Unlike findings in other studies (10,12,13,23,34–39), 13 of 25 (52%) of the non-HCC nodules in our study were hyperintense on T2-weighted images, including 10 of 19 regenerative nodules, one of three dysplastic nodules, and two of three noncirrhotic nodules (one case of focal nodular hyperplasia, one adenocarcinoma). Hyperintensity of regenerative nodules on T2-weighted MR images is a rare finding when such nodules infarct (15) and in nodular regeneration of Budd-Chiari syndrome (47). Only one patient in our study had Budd-Chiari syndrome; therefore, the likely explanation for our results is that these were areas of active fibrosis that simulated focal lesions or may have been infarcted nodules.

While arterial enhancement is the most consistent feature of HCC (14,17,23,26,37,39,40), it is by no means a specific feature. Other benign and malignant lesions in cirrhotic liver may show arterial enhancement (18–21,39,47–52). The most troublesome are the arterial enhancing dysplastic and regenerative nodules (23). In our study, 16 of 25 (64%) non-HCC nodules enhanced in the arterial phase, including one of three dysplastic nodules, 12 of 19 regenerative nodules, and three of three noncirrhotic nodules. Baron et al (19) reported that approximately one-third of their false-positive diagnoses of HCC at computed tomography (CT) assigned during screening of a large transplantation patient population resulted from enhancing vascular lesions. Similar results were reported by Freeny et al (18), and in the study of Krinsky et al (39), four arterial enhancing high-grade dysplastic nodules were falsely diagnosed as HCC. Because of their malignant potential and the difficulty in differentiating them from HCC at imaging and biopsy, high-grade dysplastic nodules are managed similarly to HCC at our institution; therefore, in our statistical analysis, we regarded these lesions as malignant. Malignant transformation can occur in these nodules after 6 months or even as early as 91/2 weeks (53–56).

Many HCC nodules become hypointense compared with liver parenchyma on delayed (interstitial-phase) contrast-enhanced images. Some become isointense and rarely remain hyperintense (39,40). In our study, 27 of 30 (90%) HCCs became hypointense compared with liver parenchyma on the 2-minute delayed contrast-enhanced images. This percentage is higher than that reported in other series (39), but this is a known feature that has been documented at MR imaging and CT (17,57–59). Carlos et al (60) developed a predictive model of MR imaging correlates of malignancy in patients with cirrhosis and a liver mass, taking into account the clinical and demographic data of these patients. Their model demonstrated that only venous phase hypointensity, {alpha}-fetoprotein level, and number of lesions detected were significant predictors of malignancy (60).

High signal intensity on T1-weighted MR images is another feature that has been described in some dysplastic nodules and low-grade HCC and is attributed to the presence of copper binding protein, steatosis, or hemorrhage (43–46,61,62). Fifty percent of the HCCs in our study were hyperintense on T1-weighted images, and six HCCs (four hyperintense and two hypointense) were seen only on the T1-weighted images. Some studies have shown that T1-weighted MR sequences can be superior to the T2-weighted sequences in depiction of early HCC (43,44).

In accordance with other reports (12,14,18,19,39), we have shown that there is a major overlap in the signal intensity characteristics and enhancement patterns of nodular lesions in cirrhotic liver, and among all features, hyperintensity on T2-weighted images, once considered a major feature of HCC, is the least reliable. Therefore, the criteria for diagnosis of possible malignancy in a cirrhotic nodule should include other features, such as lesion hypointensity compared with liver parenchyma on delayed contrast-enhanced images. While it is possible to definitively characterize many nodular lesions in the cirrhotic liver—especially the larger lesions—occasionally, imaging findings do not allow a specific diagnosis to be assigned. In such cases, follow-up imaging should be performed after 3 to 6 months to assess rapid progression, or biopsy should be performed for pathologic confirmation (23). Correlation with {alpha}-fetoprotein levels can occasionally be useful, but not all tumors express {alpha}-fetoprotein (63). Furthermore, elevated {alpha}-fetoprotein levels are sometimes seen in patients with chronic liver disease and cirrhosis without HCC.

There are recognized limitations of our study. First, not every lesion detected with MR imaging was sampled for biopsy to confirm the diagnosis. Second, pathologic proof was obtained by means of needle biopsy and not surgical resection or explant correlation. Needle biopsy in cirrhotic liver has the major drawback of sampling error that results from sampling the wrong nodule or missing small foci of HCC in dysplastic nodules. Our practice reflects true clinical practice, however, in which surgical excision and hepatic transplantation are not always possible or feasible. Patient care in daily clinical practice is usually determined on the basis of the results of needle biopsy, except when there are gross discrepancies between biopsy results and clinical and/or imaging features. Third, because of the retrospective nature of our study, there is lack of uniformity of the MR imaging technique—particularly the timing method used for arterial phase imaging. The fixed delay method used in the earlier examinations may have resulted in incorrectly timed arterial phase images, which may potentially lead to lesion mischaracterization. This may have also led to underestimation of the true performance of dynamic gadolinium-enhanced MR imaging.

Our study showed that T2-weighted MR imaging has limited value in the detection and characterization of nodular lesions in cirrhotic liver because of reduced lesion conspicuity and the overlap in signal intensity characteristics of benign and malignant nodules. The main value of T2-weighted MR imaging may lie in its usefulness for the diagnosis of cysts, hemangiomas, and lymphadenopathy. It may be useful in lesion detection and characterization in patients unable to suspend respiration during dynamic imaging. For such patients, we also use non–motion-sensitive T1-weighted MR sequences, such as inversion-recovery gradient-recalled-echo imaging. Although it is of limited value in cirrhotic liver, we have not omitted the T2-weighted MR sequence from our liver protocol but have replaced the longer respiratory-triggered T2-weighted fast SE sequence with a breath-hold fast-recovery fast SE, or FRFSE, sequence.


   FOOTNOTES
 
2 Current address: Toledo Radiology Associates, Ohio. Back

Abbreviations: HCC = hepatocellular carcinoma, ROC = receiver operating characteristic, SE = spin echo, SPGR = spoiled gradient recalled echo, 3D = three-dimensional

Author contributions: Guarantor of integrity of entire study, H.K.H.; study concepts and design, H.K.H., I.S., R.C.C., I.R.F.; literature research, H.K.H., I.S.; clinical studies, H.K.H., H.V.N., R.C.C., W.J.W., I.R.F., I.S.; data acquisition, I.S., H.K.H., H.V.N.; data analysis/interpretation, H.K.H., I.S., T.D.J.; statistical analysis, T.D.J.; manuscript preparation, all authors; manuscript definition of intellectual content, H.K.H., I.S., T.D.J., R.C.C.; manuscript editing, all authors; manuscript revision/review and final version approval, H.K.H., T.D.J., I.R.F.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 

  1. Pawluk RS, Tummala S, Brown JJ, Borrello JA. A retrospective analysis of the accuracy of T2-weighted images and dynamic gadolinium-enhanced sequences in the detection and characterization of focal hepatic lesions. J Magn Reson Imaging 1999; 9:266-273.[CrossRef][Medline]
  2. Larson RE, Semelka RC, Bagley AS, Molina PL, Brown ED, Lee JK. Hypervascular malignant liver lesions: comparison of various MR imaging pulse sequences and dynamic CT. Radiology 1994; 192:393-399.[Abstract/Free Full Text]
  3. Semelka RC, Shoenut JP, Kroeker MA, et al. Focal liver disease: comparison of dynamic contrast-enhanced CT and T2-weighted fat-suppressed, FLASH, and dynamic gadolinium-enhanced MR imaging at 1.5 T. Radiology 1992; 184:687-694.[Abstract/Free Full Text]
  4. Hamm B, Mahfouz AE, Taupitz M, et al. Liver metastases: improved detection with dynamic gadolinium-enhanced MR imaging? Radiology 1997; 202:677-682.[Abstract/Free Full Text]
  5. Ito K, Choji T, Nakada T, Nakanishi T, Kurokawa F, Okita K. Multislice dynamic MRI of hepatic tumors. J Comput Assist Tomogr 1993; 17:390-396.[Medline]
  6. Low RN, Francis IR, Sigeti JS, Foo TK. Abdominal MR imaging: comparison of T2-weighted fast and conventional spin-echo, and contrast-enhanced fast multiplanar spoiled gradient-recalled imaging. Radiology 1993; 186:803-811.[Abstract/Free Full Text]
  7. Ward J, Baudouin CJ, Ridgway JP, Robinson PJ. Magnetic resonance imaging in the detection of focal liver lesions: comparison of dynamic contrast-enhanced TurboFLASH and T2 weighted spin echo images. Br J Radiol 1995; 68:463-470.[Abstract/Free Full Text]
  8. Ebara M, Ohto M, Watanabe Y, et al. Diagnosis of small HCC: correlation of MR imaging and tumor histologic studies. Radiology 1986; 159:371-377.[Abstract/Free Full Text]
  9. Itoh K, Nishimura K, Togashi K, et al. Hepatocellular carcinoma: MR imaging. Radiology 1987; 164:21-25.[Abstract/Free Full Text]
  10. Matsui O, Kadoya M, Kameyama T, et al. Adenomatous hyperplastic nodules in the cirrhotic liver: differentiation from HCC with MR imaging. Radiology 1989; 173:123-126.[Abstract/Free Full Text]
  11. Kadoya M, Matsui O, Takashima T, Nonomura A. Hepatocellular carcinoma: correlation of MR imaging and histopathologic findings. Radiology 1992; 183:819-825.[Abstract/Free Full Text]
  12. Earls JP, Theise ND, Weinreb JC, et al. Dysplastic nodules and HCC: thin-section MR imaging of explanted cirrhotic livers with pathologic correlation. Radiology 1996; 201:207-214.[Abstract/Free Full Text]
  13. Itai Y, Ohnishi S, Ohtomo K, et al. Regenerating nodules of liver cirrhosis: MR imaging. Radiology 1987; 165:419-423.[Abstract/Free Full Text]
  14. Kelekis NL, Semelka RC, Worawattanakul S, et al. Hepatocellular carcinoma in North America: a multiinstitutional study of appearance on T1-weighted, T2-weighted, and serial gadolinium-enhanced gradient-echo images. AJR Am J Roentgenol 1998; 170:1005-1013.[Abstract/Free Full Text]
  15. Kim T, Baron RL, Nalesnik MA. Infarcted regenerative nodules in cirrhosis: CT and MR imaging findings with pathologic correlation. AJR Am J Roentgenol 2000; 175:1121-1125.[Abstract/Free Full Text]
  16. Ohtomo K, Baron RL, Dodd GD, 3rd, Federle MP, Ohtomo Y, Confer SE. Confluent hepatic fibrosis in advanced cirrhosis: evaluation with MR imaging. Radiology 1993; 189:871-874.[Abstract/Free Full Text]
  17. Yamashita Y, Mitsuzaki K, Yi T, et al. Small HCC in patients with chronic liver damage: prospective comparison of detection with dynamic MR imaging and helical CT of the whole liver. Radiology 1996; 200:79-84.[Abstract/Free Full Text]
  18. Freeny PC, Kaakji K, Grossholz ME, Schmiedl UP, Nghiem HV. The significance of hypervascular hepatic nodules (abstr). Radiology 1999; 213(P):335.
  19. Baron RL, Marsh JW, Oliver JH, Confer S, Hunt LE, Peterson MS. Screening cirrhosis for hepatocellular carcinoma (HCC) with helical contrast CT: specificity (abstr). Radiology 1997; 205(P):143.
  20. Kim TK, Choi BI, Han JK, Chung JW, Park JH, Han MC. Nontumorous arterioportal shunt mimicking hypervascular tumor in cirrhotic liver: two-phase spiral CT findings. Radiology 1998; 208:597-603.[Abstract/Free Full Text]
  21. Yu JS, Kim KW, Jeong MG, Lee JT, Yoo HS. Nontumorous hepatic arterial-portal venous shunts: MR imaging findings. Radiology 2000; 217:750-756.[Abstract/Free Full Text]
  22. Fujita T, Honjo K, Ito K, Matsumoto T, Matsunaga N, Hamm B. High-resolution dynamic MR imaging of hepatocellular carcinoma with a phased-array body coil. RadioGraphics 1997; 17:315-331.[Abstract]
  23. Baron RL, Peterson MS. Screening the cirrhotic liver for hepatocellular carcinoma with CT and MR imaging: opportunities and pitfalls. RadioGraphics 2001; 21:S117-S132.[Abstract/Free Full Text]
  24. Semelka RC, Brown ED, Ascher SM, et al. Hepatic hemangiomas: a multi-institutional study of appearance on T2-weighted and serial gadolinium-enhanced gradient-echo MR images. Radiology 1994; 192:401-406.[Abstract/Free Full Text]
  25. Dorfman DD, Berbaum KS, Metz CE. Receiver operating characteristic rating analysis: generalization to the population of readers and patients with the jackknife method. Invest Radiol 1992; 27:723-731.[CrossRef][Medline]
  26. Fujita T, Ito K, Honjo K, Okazaki H, Matsumoto T, Matsunaga N. Detection of hepatocellular carcinoma: comparison of T2-weighted breath-hold fast spin-echo sequences and high-resolution dynamic MR imaging with a phased-array body coil. J Magn Reson Imaging 1999; 9:274-279.[CrossRef][Medline]
  27. Lee JH, Ros PR, Sica GT, Ji H, Zou KH. Are T2-weighted sequences really necessary for hepatic MR imaging? (abstr). Radiology 1999; 213(P):426.
  28. Coulam CH, Chan FP, Li KCP. Can a multiphasic contrast-enhanced three-dimensional fast spoiled gradient-recalled echo sequence be sufficient for liver MR imaging? AJR Am J Roentgenol 2002; 178:335-341.[Abstract/Free Full Text]
  29. Sheppard DG, Finn JP, Miller FH, Pereles FS. Liver MR imaging: is post-gadolinium imaging only required (abstr). Radiology 1999; 213(P):427.
  30. Mirowitz SA. T2-weighted fast spin-echo MR imaging of liver. AJR Am J Roentgenol 1998; 171:263-264.[Medline]
  31. Coates GG, Borrello JA, McFarland EG, Mirowitz SA, Brown JJ. Hepatic T2-weighted MRI: a prospective comparison of sequences, including breath-hold, half-Fourier turbo spin echo (HASTE). J Magn Reson Imaging 1998; 8:642-649.[Medline]
  32. Semelka RC, Chung JJ, Hussain SM, Marcos HB, Woosley JT. Chronic hepatitis: correlation of early patchy and late linear enhancement patterns on gadolinium-enhanced MR images with histopathology—initial experience. J Magn Reson Imaging 2001; 13:385-391.[CrossRef][Medline]
  33. International Working Party. Terminology of nodular hepatocellular lesions. Hepatology 1995; 22:983-993.[CrossRef][Medline]
  34. Murakami T, Kuroda C, Marukawa T, et al. Regenerating nodules in hepatic cirrhosis: MR findings with pathologic correlation. AJR Am J Roentgenol 1990; 155:1227-1231.[Abstract/Free Full Text]
  35. Ohtomo K, Itai Y, Ohtomo Y, Shiga J, Iio M. Regenerating nodules of liver cirrhosis: MR imaging with pathologic correlation. AJR Am J Roentgenol 1990; 154:505-507.[Abstract/Free Full Text]
  36. Taylor AJ, Carmody TJ, Quiroz FA, et al. Focal masses in cirrhotic liver: CT and MR imaging features. AJR Am J Roentgenol 1994; 163:857-862.[Abstract/Free Full Text]
  37. Mitchell DG. Focal manifestations of diffuse liver disease at MR imaging. Radiology 1992; 185:1-11.[Abstract/Free Full Text]
  38. Choi BI, Han JK, Hong SH, et al. Dysplastic nodules of the liver: imaging findings. Abdom Imaging 1999; 24:250-257.[CrossRef][Medline]
  39. Krinsky GA, Lee VS, Theise ND, et al. Hepatocellular carcinoma and dysplastic nodules in patients with cirrhosis: prospective diagnosis with MR imaging and explantation correlation. Radiology 2001; 219:445-454.[Abstract/Free Full Text]
  40. Yamashita Y, Fan ZM, Yamamoto H, et al. Spin-echo and dynamic gadolinium-enhanced FLASH MR imaging of hepatocellular carcinoma: correlation with histopathologic findings. J Magn Reson Imaging 1994; 4:83-90.[Medline]
  41. Muramatsu Y, Nawano S, Takayasu K, et al. Early hepatocellular carcinoma: MR imaging. Radiology 1991; 181:209-213.[Abstract/Free Full Text]
  42. Monzawa S, Omata K, Shimazu N, Yagawa A, Hosoda K, Araki T. Well-differentiated hepatocellular carcinoma: findings of US, CT, and MR imaging. Abdom Imaging 1999; 24:392-397.[CrossRef][Medline]
  43. Inoue E, Kuroda C, Fujita M, et al. MR features of various histological grades of small hepatocellular carcinoma. J Comput Assist Tomogr 1993; 17:75-79.[Medline]
  44. Winter TC, 3rd, Takayasu K, Muramatsu Y, et al. Early advanced hepatocellular carcinoma: evaluation of CT and MR appearance with pathologic correlation. Radiology 1994; 192:379-387.[Abstract/Free Full Text]
  45. Kelekis NL, Semelka RC, Woosley JT. Malignant lesions of the liver with high T1. J Magn Reson Imaging 1996; 6:291-294.[Medline]
  46. Mitchell DG, Palazzo J, Hann HW, Rifkin MD, Burk DL, Jr, Rubin R. Hepatocellular tumors with high signal on T1-weighted MR images: chemical shift MR imaging and histologic correlation. J Comput Assist Tomogr 1991; 15:762-769.[Medline]
  47. Vilgrain V, Lewin M, Vons C, et al. Hepatic nodules in Budd-Chiari syndrome: imaging features. Radiology 1999; 210:443-450.[Abstract/Free Full Text]
  48. Krinsky GA, Theise ND, Rofsky NM, Mizrachi H, Tepperman LW, Weinreb JC. Dysplastic nodules in cirrhotic liver: arterial phase enhancement at CT and MR imaging—a case report. Radiology 1998; 209:461-464.[Abstract/Free Full Text]
  49. Matsui O, Kadoya M, Kameyama T, et al. Benign and malignant nodules in cirrhotic livers: distinction based on blood supply. Radiology 1991; 178:493-497.[Abstract/Free Full Text]
  50. Lim JH, Cho JM, Kim EY, Park CK. Dysplastic nodules in liver cirrhosis: evaluation of hemodynamics with CT during arterial portography and CT hepatic arteriography. Radiology 2000; 214:869-874.[Abstract/Free Full Text]
  51. Park YN, Yang CP, Fernandez GJ, Cubukcu O, Thung SN, Theise ND. Neoangiogenesis and sinusoidal "capillarization" in dysplastic nodules of the liver. Am J Surg Pathol 1998; 22:656-662.[CrossRef][Medline]
  52. Hayashi M, Matsui O, Ueda K, et al. Correlation between the blood supply and grade of malignancy of hepatocellular nodules associated with liver cirrhosis: evaluation by CT during intraarterial injection of contrast medium. AJR Am J Roentgenol 1999; 172:969-976.[Abstract/Free Full Text]
  53. Sadek AG, Mitchell DG, Siegelman ES, Outwater EK, Matteucci T, Hann HW. Early hepatocellular carcinoma that develops within macroregenerative nodules: growth rate depicted at serial MR imaging. Radiology 1995; 195:753-756.[Abstract/Free Full Text]
  54. Takayama T, Makuuchi M, Hirohashi S, et al. Malignant transformation of adenomatous hyperplasia to hepatocellular carcinoma. Lancet 1990; 336:1150-1153.[CrossRef][Medline]
  55. Sakamoto M, Hirohashi S. Natural history and prognosis of adenomatous hyperplasia and early hepatocellular carcinoma: multi-institutional analysis of 53 nodules followed up for more than 6 months and 141 patients with single early hepatocellular carcinoma treated by surgical resection or percutaneous ethanol injection. Jpn J Clin Oncol 1998; 28:604-608.[Abstract/Free Full Text]
  56. Seki S, Sakaguchi H, Kitada T, et al. Outcomes of dysplastic nodules in human cirrhotic liver: a clinicopathological study. Clin Cancer Res 2000; 6:3469-3473.[Abstract/Free Full Text]
  57. Ohashi I, Hanafusa K, Yoshida T. Small hepatocellular carcinoma: two-phase dynamic incremental CT in detection and evaluation. Radiology 1993; 189:851-855.[Abstract/Free Full Text]
  58. Lim JH, Choi D, Kim SH, et al. Detection of hepatocellular carcinoma: value of adding delayed phase imaging to dual-phase helical CT. AJR Am J Roentgenol 2002; 179:67-73.[Abstract/Free Full Text]
  59. Loyer EM, Chin H, DuBrow RA, David CL, Eftekhari F, Charnsangavej C. Hepatocellular carcinoma and intrahepatic peripheral cholangiocarcinoma: enhancement patterns with quadruple phase helical CT—a comparative study. Radiology 1999; 212:866-875.[Abstract/Free Full Text]
  60. Carlos RC, Kim HM, Hussain HK, Nghiem HV, Francis IR, Fendrick AM. Predicting malignancy in cirrhosis using regression analysis. AJR Am J Roentgenol 2003; 180:893-900.[Abstract/Free Full Text]
  61. Ebara M, Watanabe S, Kita K, et al. MR imaging of small hepatocellular carcinoma: effect of intratumoral copper content on signal intensity. Radiology 1991; 180:617-621.[Abstract/Free Full Text]
  62. Ebara M, Fukuda H, Kojima Y, et al. Small hepatocellular carcinoma: relationship of signal intensity to histopathological findings and metal content of the tumor and surrounding hepatic parenchyma. Radiology 1999; 210:81-88.[Abstract/Free Full Text]
  63. Marrero JA, Fontana RJ, Su GL, Conjeevaram HS, Emick DM, Lok AS. NAFLD may be a common underlying liver disease in patients with hepatocellular carcinoma in the United States. Hepatology 2002; 36:1349-1354.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 Am. J. Roentgenol.Home page
G. Brancatelli, R. L. Baron, M. P. Federle, G. Sparacia, and K. Pealer
Focal Confluent Fibrosis in Cirrhotic Liver: Natural History Studied with Serial CT
Am. J. Roentgenol., May 1, 2009; 192(5): 1341 - 1347.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
J. M. Willatt, H. K. Hussain, S. Adusumilli, and J. A. Marrero
MR Imaging of Hepatocellular Carcinoma in the Cirrhotic Liver: Challenges and Controversies
Radiology, May 1, 2008; 247(2): 311 - 330.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
T. Parikh, S. J. Drew, V. S. Lee, S. Wong, E. M. Hecht, J. S. Babb, and B. Taouli
Focal Liver Lesion Detection and Characterization with Diffusion-weighted MR Imaging: Comparison with Standard Breath-hold T2-weighted Imaging
Radiology, March 1, 2008; 246(3): 812 - 822.
[Abstract] [Full Text] [PDF]

Home page
 Br. J. Radiol.Home page
K Coenegrachts, J Delanote, L Ter Beek, M Haspeslagh, S Bipat, J Stoker, F Van Kerkhove, L Steyaert, H Rigauts, and J W Casselman
Improved focal liver lesion detection: comparison of single-shot diffusion-weighted echoplanar and single-shot T2 weighted turbo spin echo techniques
Br. J. Radiol., July 1, 2007; 80(955): 524 - 531.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
D. A. Aguirre, C. A. Behling, E. Alpert, T. I. Hassanein, and C. B. Sirlin
Liver Fibrosis: Noninvasive Diagnosis with Double Contrast Material-enhanced MR Imaging.
Radiology, May 1, 2006; 239(2): 425 - 437.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
E. M. Hecht, A. E. Holland, G. M. Israel, W. Y. Hahn, D. C. Kim, A. B. West, J. S. Babb, B. Taouli, V. S. Lee, and G. A. Krinsky
Hepatocellular Carcinoma in the Cirrhotic Liver: Gadolinium-enhanced 3D T1-weighted MR Imaging as a Stand-alone Sequence for Diagnosis.
Radiology, May 1, 2006; 239(2): 438 - 447.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
A. E. Holland, E. M. Hecht, W. Y. Hahn, D. C. Kim, J. S. Babb, V. S. Lee, A. B. West, and G. A. Krinsky
Importance of Small (<=20-mm) Enhancing Lesions Seen Only during the Hepatic Arterial Phase at MR Imaging of the Cirrhotic Liver: Evaluation and Comparison with Whole Explanted Liver
Radiology, December 1, 2005; 237(3): 938 - 944.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
M. Kanematsu, S. Goshima, H. K. Hussain, T. D. Johnson, H. V. Nghiem, and I. R. Francis
Does T2-weighted MR Imaging Really Add No Value in Detection and Characterization of Focal Lesions in Cirrhotic Liver? * Dr Hussain and colleagues respond:
Radiology, February 1, 2005; 234(2): 638 - 640.
[Full Text]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2303020921v1
230/3/637    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Hussain, H. K.
Right arrow Articles by Francis, I. R.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Hussain, H. K.
Right arrow Articles by Francis, I. R.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE