Improving the prediction of hepatocellular carcinoma in cirrhotic patients with an arterially-enhancing liver mass
2005; Lippincott Williams & Wilkins; Volume: 11; Issue: 3 Linguagem: Inglês
10.1002/lt.20357
ISSN1527-6473
AutoresJorge A. Marrero, Hero K. Hussain, HAHN V. NGHIEM, Ramsey K. Umar, Robert J. Fontana, Anna S. Lok,
Tópico(s)Liver Disease and Transplantation
ResumoLiver TransplantationVolume 11, Issue 3 p. 281-289 Original ArticlesFree Access Improving the prediction of hepatocellular carcinoma in cirrhotic patients with an arterially-enhancing liver mass Jorge A. Marrero, Corresponding Author Jorge A. Marrero jmarrero@umich.edu Division of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, MI Telephone: 734-615-4628; FAX: 734-936-7392 J.A.M. and H.K.H. contributed equally to this manuscript.Division of Gastroenterology, University of Michigan, 3912 Taubman Center, Ann Arbor, MI 48109-0362Search for more papers by this authorHero K. Hussain, Hero K. Hussain Department of Radiology, University of Michigan, Ann Arbor, MI J.A.M. and H.K.H. contributed equally to this manuscript.Search for more papers by this authorHahn V. Nghiem, Hahn V. Nghiem Department of Radiology, University of Michigan, Ann Arbor, MISearch for more papers by this authorRamsey Umar, Ramsey Umar Division of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, MISearch for more papers by this authorRobert J. Fontana, Robert J. Fontana Division of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, MISearch for more papers by this authorAnna S. Lok, Anna S. Lok Division of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, MISearch for more papers by this author Jorge A. Marrero, Corresponding Author Jorge A. Marrero jmarrero@umich.edu Division of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, MI Telephone: 734-615-4628; FAX: 734-936-7392 J.A.M. and H.K.H. contributed equally to this manuscript.Division of Gastroenterology, University of Michigan, 3912 Taubman Center, Ann Arbor, MI 48109-0362Search for more papers by this authorHero K. Hussain, Hero K. Hussain Department of Radiology, University of Michigan, Ann Arbor, MI J.A.M. and H.K.H. contributed equally to this manuscript.Search for more papers by this authorHahn V. Nghiem, Hahn V. Nghiem Department of Radiology, University of Michigan, Ann Arbor, MISearch for more papers by this authorRamsey Umar, Ramsey Umar Division of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, MISearch for more papers by this authorRobert J. Fontana, Robert J. Fontana Division of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, MISearch for more papers by this authorAnna S. Lok, Anna S. Lok Division of Gastroenterology, Department of Internal Medicine, University of Michigan, Ann Arbor, MISearch for more papers by this author First published: 17 February 2005 https://doi.org/10.1002/lt.20357Citations: 180 Telephone: 734-615-4628; FAX: 734-936-7392 AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract In the United States, cirrhotic patients with known or suspected hepatocellular carcinoma (HCC) are prioritized for liver transplantation. Noninvasive criteria for the diagnosis of HCC rely on arterial enhancement of a mass. The aim of this study was to determine whether clinical, laboratory, and / or radiologic data can improve the prediction of HCC in cirrhotic patients with an arterially-enhancing mass. Between May 2002 and June 2003, dynamic gadolinium-enhanced magnetic resonance imaging (MRI) of consecutive patients with liver cirrhosis and a solid mass were reviewed by 2 radiologists blinded to the clinical diagnosis. Clinical, laboratory, and radiologic data were recorded for all patients. A total of 94 patients with cirrhosis and an arterially-enhancing liver mass were studied, 66 (70%) of whom had HCC. Alpha-fetoprotein (AFP) >20 ng/mL (P = .029), tumor size >2 cm (P = .0018), and delayed hypointensity (P = .0001) were independent predictors of HCC. Delayed hypointensity of an arterially-enhancing mass had a sensitivity of 89% and a specificity of 96% for HCC. The presence of delayed hypointensity was the only independent predictor of HCC among patients with arterially-enhancing lesions <2 cm (odds ratio, 6.3; 95% confidence interval [CI], 1.8-13), with a sensitivity of 80% and a specificity of 95%. In conclusion, delayed hypointensity of an arterially-enhancing mass was the strongest independent predictor of HCC, regardless of the size of the lesion. If additional studies confirm our results, the noninvasive criteria utilized to make a diagnosis of HCC should be revised. (Liver Transpl 2005;11:281–289.) The incidence of hepatocellular carcinoma (HCC) in western countries is rising and is expected to further increase over the next 10–15 years. Furthermore, HCC has become a leading indication for liver transplantation in the United States due to granting of additional model for end-stage liver disease points to patients with known or suspected HCC.1 To date, cytopathologic analysis remains the gold standard for a definitive diagnosis of HCC. However, liver biopsy carries a risk of bleeding and tumor seeding,2, 3 and is not always possible due to inaccessible location of the mass, ascites, and / or coagulopathy. In addition, the tissue sample may be insufficient for a definitive diagnosis. A study by Torzilli et al.4 indicated that the preoperative diagnosis of HCC based on clinical, laboratory, and imaging data had an accuracy of 99%, suggesting that the use of needle biopsy for a diagnosis of HCC can be drastically reduced. The European Association for the Study of Liver Disease HCC Conference has provided nonhistologic criteria for a diagnosis of HCC,5 and the United Network for Organ Sharing (UNOS) policy for the transplantation of patients with HCC, does not require histologic confirmation of the tumor (http://www.unos.org/PoliciesandBylaws from July 2004). The noninvasive criteria for the diagnosis of HCC proposed by UNOS rely heavily on imaging characteristics, in particular arterial enhancement. Using the UNOS criteria at a single center in the United States, 33% of patients transplanted for HCC did not have tumor after the explant was examined and 63% of the misdiagnosed tumors had arterially-enhancing lesions ≤2 cm in diameter.6 These data suggest that arterial enhancement is a consistent but nonspecific feature of HCC and is less useful in cirrhotic patients with small (i.e., <2 cm) masses. Magnetic resonance imaging (MRI) has been proposed as a sensitive and specific imaging modality for the evaluation of liver masses in patients with cirrhosis. A total of 3 studies of MRI with explant correlation showed a sensitivity of 55, 76, and 77%, and a specificity of 57, 75, and 86%, respectively, for the detection of HCC in patients with cirrhosis.7-9 These studies indicated that the sensitivity of MRI for a diagnosis of HCC decreases significantly with lesions <2 cm in diameter. In the present study, 94 consecutive patients with known cirrhosis and an enhancing liver mass were prospectively evaluated with dynamic gadolinium-enhanced MRI. The aim of this study was to determine whether the combination of clinical, laboratory, and / or radiologic data can improve the prediction of HCC, especially among patients with enhancing masses 20 ng/mL) alpha fetoprotein (AFP) level; suggestion of a mass on ultrasound or a computed tomography scan without arterial phase; or unexplained symptoms (such as abdominal pain, increased ascites, jaundice, or weight loss). The etiology of liver disease was determined as previously described.11 Patient demographics, cause of cirrhosis, presence of ascites or hepatic encephalopathy, serum levels of aspartate aminotransferase, alanine aminotransferase, total bilirubin, alkaline phosphatase, albumin, creatinine, AFP, international normalized ratio, model for endstage liver disease score, Child-Turcotte-Pugh score, white blood cell count, and platelet count were obtained. MRI Technique All studies were performed on a 1.5-Tesla scanner (Signa; General Electric Medical Systems, Milwaukee, WI), and included the following sequences: axial longitudinal relaxation time–weighted, dual-echo gradient recalled-echo; axial transverse relaxation time–weighted fat-suppressed fast-recovery fast spin-echo; and dynamic gadolinium-enhanced imaging with a 3-dimensional spoiled gradient recalled-echo sequence.12 This sequence was acquired precontrast, and following the injection of 20-cc gadopentetate dimeglumine (Magnevist; Berlex Laboratories, Wayne, NJ) via a power injector in the arterial-dominant, portal-venous, and equilibrium (2-minute delayed) phases of enhancement, followed by a delayed acquisition at 5 minutes postgadolinium. MRI Data The magnetic resonance images were reviewed by 2 radiologists (H.K.H. and H.V.N.) with expertise in hepatic imaging, prior to the actual knowledge of the etiology of the liver mass. The radiologists reviewed the images independently in the 1st 1 of 3 of the cases, and simultaneously with the final opinion rendered by consensus in the remaining 2 of 3 of the cases. The purpose of the independent review was to obtain a measure of agreement between readers. While aware of the diagnosis of cirrhosis but blinded to the remainder of the patient's clinical data, the 2 radiologists were asked to evaluate the imaging studies for the presence of an arterially-enhancing mass, including homogeneously-, heterogeneously-, and ring-enhancing lesions. The number, size, location, signal characteristics, lesion hypointensity relative to surrounding liver in the portal-venous, 2- or 5-minute postgadolinium delayed-phases, and the presence of vascular invasion was recorded for each case. Finally, the radiologists were asked to provide a global consensus on the probability that the lesion is HCC by assigning a high or low probability based on the above characteristics and their overall impression. The radiologists used standard MRI features to characterize focal arterially-enhancing lesions in patients with cirrhosis.7, 13-16 Arterially-enhancing nodules were classified as 1 of 4 lesions: HCC included all arterially enhancing lesions >2 cm regardless of their other imaging features, and all arterially-enhancing lesions with transverse relaxation time–hyperintensity and / or delayed hypointensity regardless of their size; dysplastic nodule included ≤2-cm arterially-enhancing lesions with hyperintense longitudinal relaxation time signal, no transverse relaxation time hyperintensity, and no delayed hypointensity compared to the rest of the liver parenchyma; nonspecific enhancing nodules (NSEN) included ≤2-cm arterially-enhancing lesions with no corresponding signal changes on any of the other imaging sequences; and hemangiomas, based on a well-defined lesion with low longitudinal relaxation time and very high transverse relaxation time signal intensity compared to liver parenchyma, and one of 3 enhancement patterns: early uniform enhancement and delayed contrast retention; early peripheral nodular enhancement with centripetal progression to complete filling and delayed contrast retention; and early peripheral nodular enhancement with centripetal progression to incomplete filling and delayed contrast retention with nonenhancing central scar. Verification The final diagnosis of HCC and regenerative and dysplastic nodules was determined by histologic examination of the lesion of interest seen on MRI that led to the interpretation rendered by the radiologists. Hemangioma was verified by typical imaging features and NSEN by follow-up imaging. Patients with HCC were staged according to the UNOS-modified Tumor Node Metastasis staging system. Patients considered to have NSEN had a minimum of 3 (mean, 4 ± 2; median, 4; range, 3-9) MRI examinations over a mean follow-up of 26 ± 12 months (median, 25 months; range, 13-34) with no change in the number, size, and imaging features of the lesions, or the AFP value. Statistical Analysis Log transformation was used for AFP to account for skewness. Kappa statistics was used to determine the level of agreement between the 2 radiologists for the cases read independently (with regard to tumor size, number of lesions, presence of arterial enhancement, signal characteristics, and delayed hypointensity). We performed a per-patient analysis in which the final diagnosis of the main lesion seen on MRI, and later verified by pathology or imaging follow-up, was the main diagnosis. Pearson correlation was used to correlate the consensus diagnosis of the radiologists to the actual diagnosis of the liver mass. A 1-way analysis of variance was used to determine differences in continuous variables among the various diagnoses of liver masses. The Wilcoxon test was used for model for end-stage liver disease score and tumor size. Fisher's exact test was used to determine differences among categoric variables. Univariate analysis was performed to identify demographic, laboratory, clinical, and radiologic correlates of HCC. The laboratory and clinical criteria included etiology of liver disease, albumin, creatinine, total bilirubin, AFP, international normalized ratio, model for end-stage liver disease score, Child-Turcotte-Pugh score, white blood cell count, and platelet count. The radiologic criteria examined were the presence of arterial enhancement, delayed hypointensity, number of lesions, largest diameter of the main lesion, and portal vein thrombosis. Variables with P values <.10 in the univariate analysis were then subjected to multivariate analysis by forward logistic regression to identify independent factors associated with HCC. The adjusted odds ratio and its confidence interval were obtained from the final model. A 2-tailed P value of <.05 was used to determine statistical significance. All analyses were performed using SAS 8.1 (SAS Institute, Cary, NC). Results Patients During the study period, 106 patients with cirrhosis and a suspected liver mass underwent MRI. A total of 12 patients were excluded from the analysis; 8 had no visible mass on MRI and 4 had simple hepatic cysts. The 8 patients in whom no mass was identified have been followed for a mean of 18.3 months (range, 14-29 months) and have undergone a mean of 2.1 ± .7 MRI examinations with no evidence of an enhancing mass. The remaining 94 patients form the basis of this study. Demographics, etiology of underlying liver disease, and laboratory values at presentation for the patients who had solid arterially-enhancing liver mass(es) on MRI are listed in Table 1. Table 1. Comparison of Cirrhotic Patients With HCC Vs. Benign Liver Masses** Staging is the UNOS TNM system. HCC (n = 65) NSEN (n = 20) DN (n = 3) RN (n = 1) Hemangioma (n = 5) All (n = 94) Age (years) 58 ± 10†† P = .005 HCC vs. DN, NSEN, Hemangioma. 53 ± 8 45 ± 7 53 52 ± 9 56 ± 11 Gender (M:F) 45 : 20 13 : 7 2 : 1 1 : 0 4 : 1 65 : 29 Ethnicity (NHW:AA:As:H) 46 : 9 : 5 : 5 16 : 3 : 0 : 1 2 : 0 : 0 : 1 1 : 0 : 0 : 0 3 : 1 : 0 : 1 68 : 13 : 5 : 8 Etiology n (%) HCV 46 (71) 10 (50) 2 (67) 1 (100) 2 (40) 61 (65) HBV 3 (5) 4 (19) 1 (33) 0 2 (40) 10 (11) Cryptogenic 12 (18) 3 (15) 0 0 0 15 (16) Alcohol 4 (6) 2 (12) 0 0 1 (20) 7 (7) PBC 0 1 (4) 0 0 0 1 (1) Indication n (%) AFP > 20 ng/mL 21 (31) 7 (33) 1 (33) 0 0 29 (31) Abnormal US/CT 41 (64) 13 (67) 2 (67) 1 (100) 5 (100) 62 (66) Symptoms 3 (5) 0 0 0 0 3 (3) AFP (median) ng/mL n (%) 34†† P = .005 HCC vs. DN, NSEN, Hemangioma. 13.3 98.8 56 2.3 21.2 200 18 (28) 0 0 0 0 18 (19) MELD score 11 ± 4‡‡ P = .006 HCC vs. DN, NSEN, Hemangioma. 9 ± 2 8 ± 2 9 7 ± 0.2 9.4 ± 3.7 CTP score 7.1 ± 1 6.9 ± 1 7.2 ± 1.4 8 6.3 ± 0.7 7 ± 2 Total bilirubin (mg/dL) 2.4 ± 2 1.1 ± .6 1.2 ± .6 1.8 .8 ± .3 1.95 ± 1.3 Platelet (k/mm3) 112 ± 52 114 ± 61 106 ± 48 113 118 ± 46 115 ± 41 Listed for OLT n (%) 21 (31) 0 1 (33) 1 (100) 0 23 (24) Staging 5/ 16/ 27/ 18 NA NA NA NA NA Abbreviations: HCC, hepatocellular carcinoma; NSEN, non-specific enhancing nodule; DN, dysplastic nodule; RN, regenerative nodule; NHW, non-Hispanic white; AA, African American; As, Asian; H, Hispanic; HCV, hepatitis C; HBV, hepatitis B; PBC, primary biliary cirrhosis; US, ultrasound; CT, computed tomography; MELD, model for endstage liver disease; CTP, Child-Turcotte-Pugh score; OLT, orthotopic liver transplant. * Staging is the UNOS TNM system. † P = .005 HCC vs. DN, NSEN, Hemangioma. ‡ P = .006 HCC vs. DN, NSEN, Hemangioma. Liver Masses A total of 65 (69%) patients had a diagnosis of HCC, 1 (1%) had a regenerative nodule, and 3 (3%) had dysplastic nodules based on histologic analysis. A total of 20 (21%) patients were diagnosed with nonspecific enhancing nodules for which histology was not possible due to the small size of the nodules; all had follow-up imaging showing no change in number or size of the lesions over a period of 13–34 months. A total of 4 of the 65 patients with HCC were initially considered to have nonspecific enhancing nodules, but interval growth of the nodules on repeat MRI after 6 months led to biopsies that revealed HCC. Delayed-phase hypointensity was not present on the initial MRI, but became apparent during follow-up imaging when HCC was diagnosed. A total of 5 patients (4%) had hemangiomas based on typical MRI features. Figure 1 shows the algorithm of how the main liver mass diagnosis was achieved. Figure 1Open in figure viewerPowerPoint Algorithm indicating how patients with cirrhosis and a liver mass were evaluated. Radiologic Characteristics of the Liver Masses An average of 2.6 liver masses per patient (range, 1-6) and a mean maximal diameter of 3.6 cm (range, .5-15) were found. There were 26 (28%) patients with lesions <2 cm. A total of 71 (76%) patients had unilobar masses. A total of 60 (64%) had delayed hypointensity of the arterially-enhancing mass, and 5 (5%) had portal vein thrombosis. The imaging characteristics of the masses according to the diagnoses are shown in Table 2. The 2 radiologists had excellent agreement with regards to the probability of HCC and radiologic characteristics, with a kappa value of .837 (95% confidence interval [CI], .77-.90). Furthermore, there was excellent correlation between the radiologists' consensus diagnosis and the final pathologic diagnosis of HCC (correlation coefficient, .64; 95% CI, .78-.89; P < .001). Examples of a patient with a HCC and a nonspecific enhancing nodule are shown in Figures 2 and 3, respectively. Table 2. Radiological Characteristics of the Liver Masses** Data presented as mean ± SD unless indicated otherwise. HCC (n = 65) NSEN (n = 20) RN (n = 1) DN (n = 3) Hemangioma (n = 5) All (n = 94) No. of mass 2.5 ± 1.4 2.6 ± 1.8 1 1.8 ± 1.3 2 ± 2.6 2.6 ± 1.6 n (%) 1 21 (33) 11 (55) 3 (100) 2 (40) 36 (38) n (%) 2 15 (23) 4 (20) 0 1 (20) 20 (18) n (%) ≥3 29 (44) 5 (25) 0 2 (40) 36 (38) Size (cm) 4.3 ± 3†† P < .001 HCC vs. DN, NSEN. 1.1 ± 0.2 2.0 2.2 ± 1.3 4.8 ± 7 3.6 ± 3 n (%) <2 5 (8) 20 (100) 0 1 (33) 0 26 (28) n (%) ≥2 60 (92) 0 1 (100) 2 (6) 5 (60) 68 (72) Location (R:L:B) 39 : 10 : 16 11 : 4 : 5 1 : 0 : 0 3 : 0 : 0 2 : 1 : 2 56 : 15 : 23 Arterial enhancement n (%) 65 (100) 20 (100) 1 (100) 3 (100) 5 (100) 94 (100) Delayed hypointensity n (%) 59 (89)‡‡ P < .001 HCC vs. DN, NSEN, and Hemangioma. 1 (5) 0 0 0 60 (64) Portal vein thrombosis n (%) 5 (8)‡‡ P < .001 HCC vs. DN, NSEN, and Hemangioma. 0 0 0 0 5 (5) Abbreviations: HCC, hepatocellular carcinoma; NSEN, non-specific enhancing nodule; RN, regenerative nodule; DN, dysplastic nodule; R, right; L, left; B, bilobar. * Data presented as mean ± SD unless indicated otherwise. † P < .001 HCC vs. DN, NSEN. ‡ P < .001 HCC vs. DN, NSEN, and Hemangioma. Figure 2Open in figure viewerPowerPoint MR imaging of HCC in a cirrhotic patient. Axial dynamic magnetic resonance images through the liver in the arterial-phase (A), and at 2 minutes following gadolinium injection (B). There is a 2-cm arterial-enhancing lesion (arrow) in the right lobe, which becomes hypointense (arrow) to the liver in the delayed phase. Biopsy showed the lesion to be HCC. Note the pseudocapsule (arrowheads) around the lesion on the 2-minute delayed postgadolinium image. Figure 3Open in figure viewerPowerPoint A 1.5 × 1.6 cm arterially-enhancing nodule (arrows) remains stable between August 2001 and February 2004. The lesion does not show hypointensity (arrows) on delayed postgadolinium imaging. Note that the lesion is better seen on the delayed image of 2004. The patient underwent 6 scans between the dates mentioned above. Due to its stability, this lesion was labeled as a NSEN. A total of 23 patients were placed on the liver transplant waiting list. A total of 7 patients had suspected HCC (1 had an AFP value >200 with an arterially-enhancing lesion >2 cm; 3 had an arterially-enhancing lesion >2 cm; 3 had a suspicious lesion that was treated with radiofrequency ablation prior to transplant), and 16 had a histology-confirmed HCC. MRI detected 39 nodules (maximal diameter, 2.6 ± .5) in these patients, while 34 (maximal diameter, 2.9 ± 1.4) nodules were detected at the time of explant examination (P = .405 for difference in number of nodules; P = .32 for difference in diameter). Of the 34 nodules identified by explant examination, 1 was a high-grade dysplastic nodule (1 patient), 1 was a regenerative nodule (1 patient), and 32 (21 patients) were HCC on explant examination. The patient with a regenerative nodule had an enlarging mass from 8 mm (initially classified as a nonspecific enhancing nodule) to 2.0 cm by MRI over a 12-month period, and the lesion had homogeneous arterial enhancement without delayed hypointensity. Of the 37 nodules seen on MRI in the HCC patients, 33 had delayed hypointensity (21 patients transplanted for HCC had arterial enhancement in the main nodule and 20 had delayed hypointensity). Predictors of HCC In the univariate analysis, age >50 years, AFP >20 ng/mL, model for end-stage liver disease (MELD) score >10, size >2 cm, and the presence of delayed hypointensity of an arterially-enhancing mass were predictors of HCC. In the logistic regression model, AFP >20 ng/mL (P = .029), tumor size >2 cm (P = .0018), and delayed hypointensity of an arterially-enhancing mass (P = .0001) were independent predictors of HCC (Table 3). The presence of delayed hypointensity of an arterially-enhancing mass had a sensitivity of 89% (59 / 65 patients with HCC) and a specificity of 96% (1 − [1 / 29] without HCC) in predicting a diagnosis of HCC. Table 3. Independent Predictors of HCC Variable OR (95% CI) P Value AFP > 20 ng/mL 11.7 (2.3–30.7) .02 Size > 2 cm 27.9 (3.5–36) .001 Delayed-hypointensity** Of an arterially enhancing mass. 61 (3.8–73) .0001 * Of an arterially enhancing mass. Table 4 compares the clinical and radiologic characteristics of the patients with liver masses <2 or ≥2 cm. A total of 5 patients with HCC had lesions <2 cm in diameter. All 5 HCC patients with arterially-enhancing lesions 50 years, AFP > 20 ng/mL, and delayed hypointensity were associated with HCC lesions <2 cm. However, the presence of delayed hypointensity was the only independent predictor of HCC among patients with HCC and lesions <2 cm (odds ratio, 6.3; 95% CI, 1.8-13). For arterially-enhancing lesions <2 cm, the presence of delayed hypointensity had a sensitivity of 80% (4 / 5 in patients with HCC) and a specificity of 95% (1 − [1 / 21] in patients without HCC) for the diagnosis of HCC. Table 4. Comparison of Patients With Lesions < and > 2 cm Variable >2 cm (n = 67) <2 cm (n = 27) P Value Age 58 ± 10 52 ± 8 .005 Gender (M:F) 47 : 20 18 : 9 .496 Race (NHW:AA:As:H) 48 : 9 : 4 : 6 20 : 4 : 1 : 2 .629 Etiology (HCV:HBV:Crypto:Alc) 44 : 7 : 10 : 5 17 : 3 : 5 : 2 .643 AST 94 ± 22 88 ± 27 .452 ALT 73 ± 18 67 ± 25 .395 Bilirubin (ng/mL) 1.9 ± 2 1.3 ± 0.7 .104 MELD 10.2 ± 4 9.4 ± 2 .108 CTP 7.2 ± 1.4 6.8 ± 1.3 .243 Platelet 113 ± 21 111 ± 34 .121 AFP (ng/mL) n (%) 2941 ± 12859 51.6 ± 139 .006 20 ng/mL 39 (57) 9 (35) Lesion number 2.5 ± 1.4 2.3 ± 1.8 .916 Arterial enhancement n (%) 66 (97) 25 (96) .229 Delayed hypointensity n (%) 54 (81) 6 (23) <.0001 Portal vein thrombosis n (%) 5 (8) 0 0.02 Liver mass diagnosis <0.001 HCC 60 5 DN 2 1 RN 1 0 NSEN 0 20 Hemangioma 5 0 Abbreviations: NHW, non-Hispanic white; AA, African American; As, Asian; H, Hispanic; HCV, hepatitis C; HBV, hepatitis B; Crypto, cryptogenic; Alc, alcohol; AST, aspartate aminotransferase; ALT, alanine aminotransferase; MELD, model of endstage liver disease; CTP, Child-Turcotte-Pugh score; AFP, alpha-fetoprotein; HCC, hepatocellular carcinoma; DN, dysplastic nodule; RN, regenerative nodule; NSEN, non-specific enhancing nodule. Discussion In this prospective study of 94 cirrhotic patients, we found that delayed hypointensity of an arterially enhancing lesion was the most important independent predictor for a diagnosis of HCC regardless of the tumor size. Arterial enhancement (vascularity) is considered an essential characteristic of HCC,17, 18 and is used as the only radiologic feature for noninvasive diagnosis of HCC by UNOS. However, arterial enhancement is a nonspecific feature, and may be seen in other benign lesions such as hemangiomas (type 1), focal nodular hyperplasia, hepatic adenoma, dysplastic nodules, and, rarely, regenerative nodules.16 Furthermore, vascular abnormalities commonly seen in cirrhotic livers, such as nontumorous arterioportal shunts, also enhance in the arterial phase and may mimic HCC.15 Longitudinal relaxation time and transverse relaxation time signal characteristics may help distinguish some of these lesions, but are not always helpful.19 A review by UNOS of 666 patients with HCC in whom the explant pathology report was available showed that 146 (22%) patients had no tumor on explant examination, and 68 (10%) had no nodule or evidence of HCC.20 The reliance on arterial enhancement alone led to a significant number of patients receiving higher priority for transplant than was necessary. In our study, arterial enhancement was present in all patients with HCC and cirrhotic patients with other liver masses. By contrast, delayed hypointensity of the arterially enhancing mass was present in 89% of the patients with HCC and only 5% of patients with other arterially-enhancing liver masses. The exact reason why HCC lesions become hypointense compared to surrounding liver parenchyma on delayed postgadolinium imaging is unknown. The total number of intranodular arteries (preexisting hepatic arteries and neovascularized arteries) is often greater in HCC nodules than it is in the surrounding nonneoplastic hepatic parenchyma.21, 22 It is possible, therefore, that early venous drainage (washout) via neovascularity is the cause for delayed hypointensity. It is also possible that there is no early venous drainage but lesions appear relatively hypointense compared to surrounding fibrotic parenchyma, which retains contrast and appears hyperintense on delayed imaging, or that the lesion does not have portal venous supply and appears hypointense relative to the surrounding liver.23 We avoided the term "washout" and used delayed hypointensity instead, since gadolinium washout may not be the only cause for lesion hypointensity. Moreover, we d
Referência(s)