| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, University of Pittsburgh Medical Center, 200 Lothrop
St., Pittsburgh, PA 15213.
2 Present address: Department of Radiology, University of Chicago, MC 2026, 5841
S. Maryland Ave., Chicago, IL 60637.
3 Present address: Radiology Ltd., 3170 E. Fort Lowell Rd., Tucson, AZ
85716.
4 Department of Surgery, University of Pittsburgh Medical Center, Pittsburgh, PA
15213.
Received July 30, 2002;
accepted after revision September 18, 2002.
Address correspondence to R. L. Baron.
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. Single-detector helical CT screening was undertaken in 1329 patients with cirrhosis who were referred for transplantation. The patients underwent one or more helical CT examinations over 30 months and were followed up for an additional 19 months or until transplantation. We predominantly used unenhanced and biphasic contrast-enhanced techniques with infusions of 2.5–5.0 mL/sec. Four hundred thirty patients underwent transplantation within this period. Liver specimens were sectioned at 1-cm intervals, with direct comparison of imaging and pathologic findings and histologic confirmations of all lesions. Prospective preoperative helical CT reports were used for the primary data analysis. A retrospective unblinded review was undertaken to determine characteristics of false-positive lesions diagnosed as hepatocellular carcinoma.
RESULTS. Thirty-five patients (8%) had false-positive diagnoses for hepatocellular carcinoma based on helical CT. Twenty of these patients (5%) showed hypoattenuating lesions seen during one of the three helical CT examination phases. Fifteen patients (3%) had hyperattenuating lesions seen during the arterial phase. Among the 15 hyperattenuating lesions, CT revealed the causes to be transient benign hepatic enhancement (n = 3), hemangiomas (n = 2), fibrosis (n = 2), peliosis (n = 1), volume averaging (n = 1), low-grade dysplastic nodule (n = 1), or undetermined (n = 5). Of the 20 hypoattenuating lesions, the causes were shown to be fibrosis (n = 8), focal fat (n = 4), infarcted regenerative nodules (n = 2), regenerative nodules (n = 1), fluid trapped at the dome of the liver (n = 1), hemangioma (n = 1), or undetermined (n = 3). Follow-up helical CT in 13 (72%) of 18 patients allowed a change in the diagnosis of hepatocellular carcinoma to a finding of no cancer present.
CONCLUSION. Helical CT screening for hepatocellular carcinoma in patients with cirrhosis has a substantial false-positive detection rate. Although most of lesions were hypoattenuating, a few hyperenhancing arterial phase lesions were proven not to be hepatocellular carcinoma. An awareness of imaging characteristics and follow-up imaging can help radiologists avoid a mistaken diagnosis in many patients.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Contrast-enhanced helical CT has become the most commonly used screening tool for depicting hepatocellular carcinoma in patients with cirrhosis, which is due in large part to the increased ability of arterial phase enhancement to visualize early, small lesions. Although the sensitivity of helical CT and contrast-enhanced MR imaging for depicting hepatocellular carcinoma in patients with cirrhosis has been well documented to be from 50% to 70% [1, 2], the specificity, to our knowledge, has never been documented. Controversy exists to some extent, in that one series reported that all enhancing lesions seen in patients with cirrhosis of the liver were hepatocellular carcinoma [3], whereas there have been anecdotal reports of many causes of enhancing as well as hypovascular lesions simulating hepatocellular carcinoma in patients with cirrhosis. Without knowledge of the actual specificity, one cannot put into perspective the utility of screening patients with cirrhosis for hepatocellular carcinoma with helical CT.
We undertook this study to determine the specificity of helical CT in screening cirrhotic patients for hepatocellular carcinoma and to provide information beyond that already existing in anecdotal collections of problem liver cases. Hopefully, our study will allow radiologists and other physicians to more accurately understand the role of helical CT in evaluating patients with cirrhosis.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Approval for this study was obtained from the institutional review board of our medical center. We performed a prospective correlation of the explanted livers with the preoperative CT findings in all 430 transplant recipients. The study included 248 males and 182 females (age range, 17–73 years; mean age, 54 years).
Imaging–Pathologic Correlation
Within 24 hr after transplantation, at the pathologic gross specimen
evaluation, the explanted livers were cut into 8- to 10-mm-thick sections that
were matched as closely as possible to the transverse orientation of the CT
images. The preoperative helical CT findings were prospectively and directly
correlated with the surgical pathologic findings. The number, size, and
location of all lesions seen on preoperative helical CT that were suspicious
for tumor were recorded prospectively. If a lesion seen on preoperative
helical CT was not found in the initial specimen sections, additional thinner
sections in additional planes were obtained to improve correlation between the
imaging and pathologic findings.
Specimen section maps were recorded to correlate the imaging and pathologic findings. Specimen pathology section numbers for the lesions were also recorded for subsequent direct histologic correlation with specific CT findings.
Helical CT Technique
All patients underwent an initial single-detector helical CT examination,
and many were followed up with additional examinations until transplantation.
Most patients underwent CT performed using a helical technique (HiSpeed
Advantage; General Electric Medical Systems, Milwaukee, WI).
Helical CT scans were obtained, when possible, with a triphasic technique, which included the acquisition of unenhanced images of the liver followed by the acquisition of images with IV contrast-enhanced biphasic helical CT during the hepatic arterial and portovenous phases of enhancement. Section collimation was 7 mm with a pitch of 1.5:1.0. Scanning was started after the initiation of contrast infusion at 20–28 sec for hepatic arterial phase imaging and 60–70 sec for portovenous phase imaging. The contrast-enhanced images were obtained with 150 mL of iodinated contrast material (injection of iothalamate meglumine 60% [Conray] or injection of ioversol 68% [Optiray 320]; Mallinckrodt, St. Louis, MO), which was administered IV with a mechanical power injector (Medrad, Pittsburgh, PA) at a rate of 2.5–5.0 mL/sec.
Helical CT Interpretation
The diagnoses of liver lesions based on helical CT were recorded in two
ways. First, the prospective original clinical reports on the helical CT scans
rendered by eight experienced faculty radiologists with subspecialty expertise
in abdominal imaging were retrospectively collected at the time of
transplantation and used for the primary data analysis for detecting tumor on
helical CT. All original reports describing masses as diagnostic or suspicious
for hepatocellular carcinoma were considered for our study as a prospective
diagnosis of hepatocellular carcinoma.
Hepatocellular carcinoma was typically defined as a well-or ill-defined lesion, single or multifocal, hypoattenuating on unenhanced scans, moderately and inhomogeneously hyperattenuating to the surrounding liver on the arterial phase of enhancement, and inhomogeneously hypoattenuating on the portovenous phase. The presence of a partial or complete capsule, foci of calcifications, fat, intralesional hemorrhage, or the invasion of an adjacent branch of the portal vein were further indications of hepatocellular carcinoma. Any lesion prospectively described as an enhancing lesion of uncertain cause with characteristics not typical of hemangioma were also considered in our study to be positive for hepatocellular carcinoma. Any hypovascular masses described explicitly as not characteristic of cyst or hemangioma and of uncertain cause were considered to be a potential finding of hepatocellular carcinoma and therefore a false-positive diagnosis of hepatocellular carcinoma. Conversely, if the interpreting radiologist reported a lesion as benign or the entire study did not report any lesion, our study was considered negative for hepatocellular carcinoma. Among the patients who underwent multiple CT examinations, the results were considered to be positive when one or more of the examinations were determined to have a positive finding.
Second, to prospectively correlate the imaging findings with the surgical pathologic findings, we retrospectively reviewed all cases of false-positive original findings of hepatocellular carcinoma at the time of diagnosis, based on the pathologic results, to determine the cause of the false-positive finding.
The attenuation of the tumor was judged relative to that of the surrounding liver and that of the blood pool on unenhanced images, as well as on images obtained in each phase of contrast medium enhancement (hepatic arterial, portal venous, and delayed phases).
To determine the ability of additional studies to clarify mistaken diagnoses of hepatocellular carcinoma, we evaluated subsequent follow-up CT studies. If the follow-up helical CT examination did not reveal the lesion or the report did not suggest malignancy, an assessment was determined as to the reason for clarification of the diagnosis or why the lesion was not visualized.
All false-positive studies were reviewed by two observers for consensus on the imaging characteristics of the lesions.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Demographic data for the patients with false-positive diagnoses included 18 men and 17 women who were 25–73 years old (mean age, 54 years). The causes of cirrhosis in these 35 patients were alcohol (n = 11), cryptogenetic (n = 7), autoimmune (n = 6), primary biliary cirrhosis (n = 3), primary sclerosing cholangitis (n = 3), hepatitis C (n = 3), hepatitis B (n = 1), and hepatitis B and C (n = 1).
Lesion Enhancement Characteristics
Of the 35 patients, 16 underwent a complete triphasic helical CT
examination, 17 underwent unenhanced and portovenous phase imaging only, and
two underwent unenhanced imaging only.
Review of lesion characteristics confirmed 15 lesions to be homogenously hyperattenuating after contrast infusion, whereas 20 were hypoattenuating, either after contrast administration or if no contrast had been administered on unenhanced images.
Among the 15 hyperattenuating lesions, pathology proved hemangiomas (n = 2) (Fig. 1A, 1B), focal confluent fibrosis (n = 2), peliosis (n = 1) (Fig. 2A, 2B), and benign regenerative nodule with low-grade dysplastic changes (n = 1) (Fig. 3A, 3B, 3C). In nine patients in whom nothing was detected at pathology to account for the false-positive imaging findings, a retrospective review of the images determined a cause for the false-positive findings in four: a wedge shaped hyperattenuating area that was interpreted retrospectively as a transient hepatic attenuation difference in three patients or volume averaging adjacent to normal structures that was interpreted as the reason for the false-positive finding in one patient. In five of the patients in whom no apparent cause could be seen at pathology (Fig. 4A, 4B), no retrospective explanation for the CT findings could be determined when reviewing the helical CT images.
|
|
|
|
|
|
|
|
|
Among the 20 hypoattenuating cases, pathology revealed the causes to be focal fibrosis (n = 8) (Fig. 5A, 5B), focal fat (n = 4), infarcted benign regenerative nodules (n = 2), benign regenerative nodule (n = 1), and fibrosed hemangioma (n = 1). In four patients, nothing was detected at pathology to account for the false-positive findings. A review of the images in these four patients revealed extrahepatic fluid trapped at the dome of the liver that simulated a peripheral mass in one patient (Fig. 6A, 6B) and no apparent cause seen at pathology or in reviewing the images in the three remaining patients.
|
|
|
|
In 23 patients, the false-positive lesions were reported on the initial CT examination. In 12 patients, the false-positive lesions were first identified on follow-up helical CT images, although they had not been present on the initial helical CT examination. Of the 35 patients with false-positive diagnoses of hepatocellular carcinoma, eight had one helical CT examination, 14 had two examinations, six had three examinations, and seven had four examinations. The interval between serial helical CT scans ranged from 1 to 36 months, with an average of 8 months. Eighteen patients underwent a follow-up helical CT scan after the finding of mistaken hepatocellular carcinoma was rendered.
In these 18 patients, the lesion either disappeared or was not visualized on subsequent helical CT scans in 10 patients (55%), was unchanged in five patients (28%) and was again suggested as hepatocellular carcinoma, and was characterized as benign in three patients (17%).
Of the three patients in whom the lesion was characterized appropriately as benign on follow-up helical CT examinations, in one patient the findings were due to a bulging liver contour adjacent to scarring and retraction that remained stable on follow-up helical CT and was characterized as such when the adjacent retraction was recognized as the reason for the bulge in the unaffected liver. Focal fibrosis was the underlying cause in the other two patients. In one patient, the fibrosis showed focal enhancement that was erroneously interpreted as evidence of malignancy, whereas in the other patient, fibrosis was present at the dome of the liver with a vertical orientation and therefore not easily recognized because the lesion appeared round on axial images. The lesions in both patients were correctly interpreted as fibrosis on subsequent helical CT examinations. The stability of the lesions in these patients over time and the recognition of associated findings suggested the correct diagnosis to the interpreting radiologist.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Although many studies have attempted to assess the sensitivity of helical CT in the depiction of hepatocellular carcinoma in patients with cirrhosis, no study, to our knowledge, has specifically investigated the specificity of helical CT. Other lesions on helical CT anecdotally reported in patients with cirrhosis include regenerative nodules [9, 10], focal confluent fibrosis [11], cysts and hemangiomas [12], and rare enhancing dysplastic nodules [13].
Although prior studies have determined that other findings can simulate hepatocellular carcinoma, our series puts into perspective the expected false-positive rate for hepatocellular carcinoma in a large screening population. Our study confirms that a substantial false-positive diagnosis rate (8%) of hepatocellular carcinoma occurs in an experienced liver transplantation center, with 5% of the lesions appearing hypoattenuating to the liver and 3% showing substantial enhancement during arterial phase contrast-enhanced imaging. This false-positive rate of 8% must be put into perspective with the reported sensitivity rate of 11–14% for hepatocellular carcinoma in similar large imaging–transplantation studies [1, 2, 14]. Thus, during the screening of patients with cirrhosis using helical CT, a false-positive finding might be expected almost as often as a true-positive finding. This statistic is important because a diagnosis of hepatocellular carcinoma will substantially impact treatment strategies in patients with cirrhosis. The data from our study suggest that the depiction of small lesions on helical CT during the screening of patients with cirrhosis should not be assumed to be hepatocellular carcinoma and should not be used to alter patient treatment until the lesions are confirmed.
An awareness of key imaging characteristics of benign lesions can be used to avoid false-positive diagnoses, particularly in focal confluent fibrosis [11], transient hepatic attenuation difference [15], and flash-filling hemangiomas [16]. Follow-up helical CT afforded a second CT observer the opportunity to recognize key characteristics for these benign diagnoses and also allowed the development of progressive changes, particularly with focal fibrosis, that make these diagnoses possible. Although focal confluent fibrosis has been well documented [11], the evolutionary process leading to a characteristic imaging finding has not been emphasized. Several patients in our series showed irregular arterial phase enhancement within the lesions. It is important to recognize that such enhancement can be seen in fibrosis before the development of characteristic findings such as volume loss with capsular retraction and use follow-up helical CT to confirm the benign diagnosis (Fig. 7A, 7B, 7C, 7D).
|
|
|
|
Similarly, recognition of enhancement patterns of flash-filling hemangiomas [16], arteriovenous malformations [17], and peliosis during all phases of imaging is critical to avoid a mistaken diagnosis of hepatocellular carcinoma. These small lesions do not show the typical nodular enhancement seen in larger hemangiomas but appear to fill in instantly the entire lesion during the arterial phase of contrast-enhanced imaging. Although it has become routine to compare liver lesion attenuation with surrounding liver attenuation, lesion attenuation should be related to that of blood pool attenuation for an accurate diagnosis. These small blood pool lesions will have attenuation similar to blood pool on all phases of imaging, including unenhanced helical CT, whereas solid lesions such as hepatocellular carcinomas may have similar attenuation on one phase but not on all three phases (Fig. 8A, 8B, 8C). Particularly helpful can be determining in small vascular lesions that persistent enhancement occurs in the portovenous phase of imaging, whereas hepatocellular carcinoma will have a tumor washout, typical of vascular solid tumors.
|
|
|
An important determination in our study was that of the 18 patients with false-positive findings and a follow-up examination; the subsequent CT examination in 13 (72%) allowed clarification and withdrawal of the diagnosis of hepatocellular carcinoma.
The most problematic lesion in our study was the large benign regenerative nodule that showed marked arterial phase contrast enhancement with portal venous phase washout, simulating that of hepatocellular carcinoma. Histologic examination, even on intense retrospective review, revealed only mild dysplastic changes, a common finding in cirrhotic regenerative nodules in our experience. Other authors have reported rare substantial arterial phase enhancement in benign regenerative nodules on helical CT and MR imaging [2, 13]. The progression from benign regenerative nodule to dysplastic nodule to hepatocellular carcinoma has been well studied pathologically [18], and correlations with imaging are possible [19] that show decreasing portal flow and increasing arterial blood flow, as evidenced by contrast enhancement as nodules progress through this transition. Although these imaging changes correlate with pathology, there is a spectrum of correlation, and early or delayed changes in imaging can occur when compared with the progression. In addition, much of the histologic diagnosis is based on subjective criteria [18].
As in the study by Krinsky et al. [2], who compared MR imaging with explanted livers, our study involved a substantial incidence of small enhancing nodules for which no underlying cause could be established on extensive evaluation of the pathologic specimens or the imaging studies. Three such cases were seen as hypoattenuating on helical CT, and five such cases were enhancing and hyperattenuating homogeneously during arterial phase helical CT. In these cases, during the time of correlative imaging and gross specimen review, additional sections were obtained both axially and vertically through the specimen to ensure that no lesion was visible to explain the lesions seen on the image. Several possibilities exist for these lesions. One possibility (similar to the one case of a large enhancing nodule that was histologically confirmed) is that these represent enhancing benign nodules and a false-positive diagnosis of hepatocellular carcinoma. This classification was required by our research design and incorporated into our results. Unlike the one case of an enhancing nodule that was clearly larger than surrounding nodules and easily identifiable in the gross specimen, these remaining false-positive lesions were all smaller, approximately 0.75–1.5 cm in diameter. At sectioning of the liver, the involved region of the liver contained numerous similar-sized nodules that all appeared of approximately the same size and appearance. Realistically, although histologic samples were taken from the area where the nodule was thought to be, we had no way of ensuring that the exact nodule that enhanced was sampled. In fact, because all the nodules had an appearance similar to the examining eye, the exact nodule was not sampled. Although we have categorized these lesions as false-positive, they could be true-positive lesions in which visual changes had not yet occurred to discolor the nodule to be recognized at gross examination.
Our study has several limitations. To determine the diagnosis based on helical CT, we used the original reports from our abdominal imaging faculty, representing eight radiologists who interpreted the findings without specific controlled diagnostic criteria from our daily practice. However, this study design may more accurately reflect the true clinical setting in which patients are being encountered. Further, early in our study, we obtained scans at a lower rate of contrast injection (2.5–3.0 mL/sec) than we used in the later part of the study because we became aware that faster rates of injection had better enhancement characteristics for lesion detection during the arterial contrast phase. Had the faster rate been used throughout the study, an even higher rate of false-positive lesions could have been encountered. Lastly, not all patients could undergo either triphasic or even contrast-enhanced imaging because of poor venous access, contrast material allergies, or significant renal failure. Had arterial phase images been obtained in all patients, there could have been a slightly higher rate of enhancing false-positive lesions.
In summary, despite its limitations, helical CT remains one of the best screening tools available for detecting hepatocellular carcinoma in patients with cirrhosis. However, it is important to be aware of the numerous benign lesions that can be encountered in patients with cirrhosis that simulate hepatocellular carcinoma, even with experienced radiologists. Because the detection of hepatocellular carcinoma is so difficult in its early stages and with the knowledge that a substantial false-positive detection rate occurs with helical CT screening, one cannot assume that all lesions encountered on helical CT are malignant. We recommend confirming all such cases before either withholding or initiating treatment for presumed hepatocellular carcinoma. The use of follow-up helical CT in such cases can be helpful to confirm the benign nature of many lesions encountered or to show progression typical of hepatocellular carcinoma. Although the fact that lesions can simulate hepatocellular carcinoma on helical CT is not new, one needs to be aware that the incidence of false-positive lesions may be as common as that of true-positive carcinoma lesions.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  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]
|
|
|
|
 |
|
 D. Marin, M. Di Martino, A. Guerrisi, G. De Filippis, M. Rossi, S. Ginanni Corradini, R. Masciangelo, C. Catalano, and R. Passariello Hepatocellular Carcinoma in Patients with Cirrhosis: Qualitative Comparison of Gadobenate Dimeglumine-enhanced MR Imaging and Multiphasic 64-Section CT Radiology, April 1, 2009; 251(1): 85 - 95. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Marin, C. Catalano, G. De Filippis, M. Di Martino, A. Guerrisi, M. Rossi, and R. Passariello Detection of Hepatocellular Carcinoma in Patients with Cirrhosis: Added Value of Coronal Reformations from Isotropic Voxels with 64-MDCT Am. J. Roentgenol., January 1, 2009; 192(1): 180 - 187. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. W. Anderson, J. B. Kruskal, and R. A. Kane Benign Hepatic Tumors and Iatrogenic Pseudotumors1 RadioGraphics, January 1, 2009; 29(1): 211 - 229. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
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]
|
|
|
|
|
|
A. Ronzoni, D. Artioli, R. Scardina, L. Battistig, E. Minola, S. Sironi, and A. Vanzulli Role of MDCT in the Diagnosis of Hepatocellular Carcinoma in Patients with Cirrhosis Undergoing Orthotopic Liver Transplantation Am. J. Roentgenol., October 1, 2007; 189(4): 792 - 798. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. H. Kim, J. M. Lee, Y. J. Kim, J. Y. Choi, G. H. Kim, H. Y. Lee, and B. I. Choi Detection of Hepatocellular Carcinoma on CT in Liver Transplant Candidates: Comparison of PACS Tile and Multisynchronized Stack Modes Am. J. Roentgenol., May 1, 2007; 188(5): 1337 - 1342. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Iannaccone, A. Laghi, C. Catalano, P. Rossi, F. Mangiapane, T. Murakami, M. Hori, F. Piacentini, I. Nofroni, and R. Passariello Hepatocellular Carcinoma: Role of Unenhanced and Delayed Phase Multi-Detector Row Helical CT in Patients with Cirrhosis Radiology, February 1, 2005; 234(2): 460 - 467. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Giorgio, G. Ferraioli, L. Tarantino, G. de Stefano, V. Scala, F. Scarano, C. Coppola, and L. Del Viscovo Contrast-Enhanced Sonographic Appearance of Hepatocellular Carcinoma in Patients with Cirrhosis: Comparison with Contrast-Enhanced Helical CT Appearance Am. J. Roentgenol., November 1, 2004; 183(5): 1319 - 1326. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J Scholmerich and D Schacherer Diagnostic biopsy for hepatocellular carcinoma in cirrhosis: useful, necessary, dangerous, or academic sport? Gut, September 1, 2004; 53(9): 1224 - 1226. [Full Text] [PDF]
|
|
|
|
 |
|
 J A Guthrie Cirrhosis and focal liver lesions: MRI findings Imaging, September 1, 2004; 16(4): 351 - 363. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
