Hepatic Adenoma and Focal Nodular Hyperplasia: Diagnosis and Criteria for Treatment

Luciano De Carlis,* Vincenzo Pirotta,* GianFranco Rondinara,* Cosimo V. Sansalone,* Giovanni Colella,* Giuseppe Maione,* Abdallah O. Slim,* Antonio Rampoldi, Alberto Cazzulani, Luca Belli, and Domenico Forti*
Focal nodular hyperplasia (FNH) and adenoma are rare benign hepatic tumors, and the standards for diagnosis and treatment still remain controversial. Usually adenoma is an indication for resection, due to its tendency to bleed and to degenerate; FNH, on the contrary, may be treated conservatively. Preoperation differential diagnosis is, however, difficult, often impossible. Materials and methods. Thirty-eight patients with presumed hepatic adenoma and/or FNH were studied at our department from 1984 to 1996. Preoperative assessment included clinical evaluation and symptoms, laboratory tests, liver biopsy, ultrasound scan, computed tomography scan, magnetic resonance imaging, scintigraphy, and angiography. Thirteen patients had a presumed diagnosis of FNH, 16 of adenoma, and 9 of undetermined benign lesions; 27 had hepatic resections (3 with laparoscopic technique), and 11 were not operated on and are actually under a strict follow-up observation. Results. The nal diagnosis was 19 FNH and 19 adenomas (2 of which contained areas of hepatocarcinoma). Presumed diagnosis was conrmed in 71% of cases. Use of oral contraceptives, abdominal symptoms, and pathologic liver test results were more frequent in patients with adenomas. There were no deaths after surgery. All resected patients were tumor free during the follow-up, and in 10 of the 11 nonoperated cases, the size of the nodules remained unchanged. We conclude that precise diagnosis of these benign liver tumors remains difficult and sometimes impossible, despite new imaging techniques. Hepatic resections can be performed under very safe conditions; laparoscopic surgery may play a role in selected cases. Adenomas and uncertain cases are clear indications for surgery. Only when a diagnosis of FNH can be rmly conrmed in asymptomatic patients is strict observation without surgery recommended. Copyright r 1997 by the American Association for the Study of Liver Diseases

n contrast with hemangioma,1 focal nodular hyperplasia (FNH) and hepatic adenoma are very uncommon benign lesions affecting the liver, and their diagnosis and differentiation may be difficult. Moreover, their natural history is not well defined. Because of all these considerations, their surgical indication and treatment remain controversial. As a consequence of the widespread use of improved imaging modalities, these tumors are now recognized more frequently, and more information is available on their behavior.2 In particular, a

From the *Department of Surgery and Abdominal Transplantation, the Department of Radiology, and the Department of Hepatology, Niguarda Hospital, Milan, Italy. Address reprints request to Luciano De Carlis, MD, Divisione di Chirurgia Generale e dei Trapianti Addominali, Pizzamiglio 27, Ospedale Niguarda, 20162 Milano, Italy. Copyright r 1997 by the American Association for the Study of Liver Diseases 1074-3022/97/0302-0009$3.00/0

strict correlation exists between these tumors and the use of oral contraceptives.3,4 Hepatic adenomas have the tendency to grow to conspicuous sizes, and spontaneous ruptures or bleeding are relatively frequent. Malignant degeneration has been reported in some cases, and resection is therefore advisable.5 On the contrary, FNH is often an incidental finding, and to date there is no convincing report showing that these tumors can bleed or degenerate. Because of this, resection may be avoided when the diagnostic assessment evidences FNH.2 In clinical practice, computed tomography (CT), magnetic resonance imaging (MRI), ultrasound (US), and angiography are used in an attempt to determine the nature of the solitary masses of the liver, but accurate distinction between adenoma and FNH before surgery is often difficult.6-10 Furthermore, percutaneous needle biopsy cannot differentiate these tumors with accuracy.11,12

160

Liver Transplantation and Surgery, Vol 3, No 2 (March), 1997: pp 160-165

Hepatic Adenoma and FNH

161

The authors report herein their experience in the treatment of these benign lesions of the liver. Preoperative findings were matched with definite diagnoses and with the results of surgery; when surgery was unadvisable, the clinical courses of these patients were closely followed up through time. The aim of this study was both to define the diagnostic criteria and establish in which cases surgical treatment of these tumors is indicated.

Materials and Methods

From January 1984 to May 1996, 38 patients with either hepatic adenoma or FNH were observed in our surgical department. Nine patients were observed in the first 6 years, whereas the remaining 29 were referred to us between 1990 and 1996. The patient population included 37 women and 1 man, ranging in age from 21 to 57 years (average age, 32.6). No chronic liver diseases nor abnormalities in serum alphafetoprotein levels were detected in any patients. Thirty women (78.9%) had a history of oral contraceptive consumption for an average time of 5.8 years (range, 7 months to 12 years) before diagnosis. Ten (26.3%) patients were completely asymptomatic, and the lesions were discovered during periodic routine examinations (8 cases) or laparotomies (2 cases) performed for different medical reasons; 23 (60.5%) complained of abdominal pain, which was acute in 9 (23.6%); 11 (28.9%) had a palpable mass, and 15 (39.4%) suffered from vague digestive troubles with fatigue and sense of heaviness in the right abdomen. All patients were evaluated with routine laboratory analyses, including liver tests. Only alkaline phosphatase, gammaglutamyl transpeptidase, and red blood cell count showed some abnormalities in 14 cases (36.8%). US scan, liver scintigraphy, CT, and selective hepatic angiography were performed in all cases. MRI, available to us since 1990, was employed in the last 28 patients (Figs. 1, 2). The diagnosis was made by adopting predefined criteria, slightly modified by the authors. (Table 1).10,11,12 Percutaneous fine-needle liver biopsies were performed in all except 3 patients, where fresh frozen section specimens were obtained during laparotomy. FNH was preoperatively diagnosed in 13 cases and adenoma in 16. In 9 patients, a differential diagnosis could not be obtained. Two symptomatic patients with diagnosis of FNH, 16 with diagnosis of adenoma, and 9 with uncertain diagnosis underwent liver resection. Eleven patients were not operated on because preoperative study, including histology, showed the typical features of FNH; in 8 of these cases no clinical symptoms were evident, and in 7, moreover, the lesions were not easily resectable because of their central location in the liver parenchyma. All 11 of these lesions were the only lesion in each patient, with an average size at CT of 4.2 cm (range 2.5 to 5.5). Right hepatectomy was performed in 3 cases, left hepatectomy in 2, left lateral lobectomy in 4, and segmentectomy or enucleation in 18. An intraoperative US scan was used routinely

Figure 1. Typical CT appearance of an adenoma of the left liver lobe. (A) A hypodense area on the left lateral hepatic segments is present before contrast administration. (B) A typical marked contrast enhancement is evident in the early arterial phase.
to determine the location of the tumor and its relationship with the vascular system. Three superficial nodules, two located in the third and one in the sixth liver segment were excised by laparoscopic technique. Nodules were solitary in 35 of 38 patients (92.1%), whereas 3 patients had multiple tumors: Two had two FNH and 1 had three adenomas. The size of the different nodules ranged from 2.5 to 22 cm (mean, 8.7 cm). Intratumor hemorrhage was noted in five nodules. All the lesions were submitted to extensive evaluation by a trained pathologist. Follow-up was completed in 100% of cases and ranged from 2 months to 12 years (average, 46 months). Patients underwent an annual check-up with clinical examination, US scan, and biochemical

162

De Carlis et al

Table 1. Diagnostic Criteria for Adenoma and FNH Adenoma Enlarging nodules At CT, ipodensity followed by a marked contrast enhancement, calcications, capsule, and fat inltration At angiography, vascular supply from the periphery to the center of the node At biopsy, sheets of normal hepatocytes without bile ducts and Kupffer cells FNH At CT, isodensity with an iperdense central scar (50% of cases) At angiography, a central feeding artery with rapid visualization of the suprahepatic vein At MRI, isointense lesion on T1T2 with hyperintense central scar on T2 At scintigrams, normal or increased uptake At biopsy, normal hepatocytes separated by brous septa, proliferating vessels, bile ducts, and inammatory cells

Figure 2. Focal nodular hyperplasia of the left liver lobe. (A) MRI shows an isointense mass with a little hyperintense central scar on T2-weighted images. (B) At selective angiography a marked hypervascular lesion appears; the feeding artery is evident with rapid contrast lling from the center to the periphery of the node.

results appear in Table 2. In 2 patients, an accurate pathological examination revealed areas of welldifferentiated hepatocarcinoma within the adenomatous nodules. Two presumed adenomas were determined to be FNH, whereas in the nine undetermined cases, four were diagnosed as FNH and five as adenomas. All presumed FNH were confirmed both by pathological examinations and by follow-up data. The 11 patients with unresected FNH are regularly followed in our outpatient clinic, as mentioned earlier (average follow-up, 23.7 months; range, 5 to 39), and the clinical courses were uneventful except for 1. All 11 patients presented typical CT, MRI, angiographic, and/or histological features of FNH. In one case the nodule size increased from 4.5 to 5.5 cm and is now under strict evaluation. Preoperative biopsy and postop-

data; CT and/or MRI were performed only when indicated. Statistical analysis was based on the Students t-test, assuming significance when P , .05. The therapeutic protocol was approved by the ethical committee of the hospital, and an informed consent was obtained from each patient included in the study.

Table 2. Comparison Between Presumed and Denitive Diagnosis in the 38 Considered Patients FNH Presumed diagnosis Final diagnosis FNH Adenoma 13 13* 0 Adenoma 16 2 14 Uncertain 9 4 5

Results
The final diagnosis in the resected cases was FNH in 8 patients and hepatic adenoma in 19. These

*Including the 11 nonresected patients. Including the 2 cases with areas of HCC.

Hepatic Adenoma and FNH

163

erative surgical pathologic evaluation or follow-up data (in the nonoperated cases) were in agreement, thus allowing a definite diagnosis in 19 of 35 cases (54.2%). All preoperative studies showed a diagnostic accuracy of 71% (27/38 patients). Oral contraceptive use was more frequent in the patients with adenoma (17/19 or 89.4%) than in those with FNH (13/19 or 68.4%; P 5 ns). Acute pain (possibly related to intranodular bleeding) and pathologic liver test results were significantly more frequently associated with the presence of an adenoma or an hepatocarcinoma (P , .05). All hemorrhagic nodules were adenomas. Other clinical features of our patient population are shown in Table 3. No perioperative deaths occurred in the patients who underwent liver resection. Three patients had subdiaphragmatic fluid collections: One was reoperated and a small biliary fistula was sealed; the other 2 maintained percutaneous drainages for a few days. Minor complications occurred in 5 other patients including pleural effusion in 2, pneumonia in 1, and wound suppuration in 2: All were treated conservatively. The average hospitalization time was 10.9 days (range, 625). The 3 patients operated with the laparoscopic technique showed no postoperative problems and were discharged from hospital on the 4th postoperative day. During the follow-up, one patient died in a traffic accident 3 years after the resection. All the others are alive

with no evidence of tumor recurrence. All patients had discontinued oral contraceptive use.

Discussion
Our experience seems to confirm that FNH, when correctly diagnosed, may be managed conservatively and monitored with repeated US scans.13,14,15 Problems may exist in obtaining a certain differential diagnosis between FNH and adenoma and, in some cases, between benign and malignant tumors. From our data, only 15/19 (78.9%) of FNH had a correct preoperative diagnosis with accurate imaging techniques. US scan is nonspecific in the differentiation of these lesions but has a great value as a noninvasive method in the follow-up of both resected and nonresected patients. CT permits diagnosis in typical cases when a central scar within the nodule or a feeding vessel can be observed, but these characteristic pictures are present only in 50% of patients. Moreover, fibrolamellar carcinoma may present an important fibrotic component, similar to the central scar described as typical for FNH. MRI has an accuracy comparable to CT, and when used together, they may add 10% to 15% to specificity. Selective hepatic angiography was performed routinely in this series of patients, giving excellent diagnostic confirmation without any related complications. Concern exists about its extensive utilization for benign hepatic lesions because it has the disadvantage of being an invasive procedure. The preference for angiography results from our extensive experience in the treatment of portal hypertension, in which it proved to be extremely safe and exhaustive. Furthermore, the importance of angiography is incomparable for technical reasons when planning a liver resection. The procedure has diagnostic value for FNH when a feeding artery to the mass is demonstrable: This was the case in 11 of 19 (57.8%) of our patients with FNH, and in all, this diagnosis was confirmed either by postoperative pathologic evaluation or by follow-up data. In our experience, in the typical cases, a suprahepatic vein selectively draining the mass was usually rapidly seen along with the feeding artery (Fig. 3). To our knowledge, this observation is not reported in the literature and seems to be a pathognomonic picture of FNH; no patient with adenoma or other hepatic masses evidenced such angiographic features. Scintigraphy shows normal uptake in all cases of FNH due to the presence of Kupffer cells, but recent data

Table 3. Clinical Features of the Patients FNH (n 5 19) Age (mean 1 range) 33.4 (23-57) Oral contraceptive use 13 (68.4%) Symptoms None 8 (42.1%) Abdominal pain 8 (42.1%) Acute pain 0 (0%) Palpable mass 5 (26.3%) Vague 5 (26.3%) Biochemical alterations 2 (10.5%) Single lesion 19 (100%) Size (mean) 9.2 Adenoma (n 5 19*) 31.8 (21-43) 17 (89.4%) 2 (10.5%) 15 (78.9%) 9 (47.3%) 6 (31.5%) 10 (52.6%) 12 (63.4%) 16 (84.2%) 7.9

P Value
ns ns ns ns ,.05 ns ns ,.05 ns ns

*Including two cases with areas of HCC.

164

De Carlis et al

Adenoma was correctly diagnosed in 12 of 19 cases (63.1%); patients had symptoms present in a higher percentage of cases, especially when bleeding or sudden growth occurred13,14 (Fig. 4). Laboratory tests in most cases show alterations in stasis indexes. In our series, these signs had statistical significance in the differentiation between adenoma and FNH. CT and MRI frequently demonstrate the presence of necrosis or hemorrhage within the nodules (five cases in our series; Fig. 4); these findings, however, may be encountered also in malignant lesions, such as large hepatomas.7,9,10 Either an enlarging lesion or anemia on subsequent controls may indicate the presence of an adenoma. On scintigram a reduced uptake is usually evident, but not always. Percutaneous liver biopsy alone is reported to be of little value in the diagnosis of these benign tumors due to the frequent lack of specific features in a small specimen; moreover, the material is often inadequate, and typical signs were present in only 54.2% of our cases.11,12 Other problems are related to the fact that biopsy may be contraindicated in hemorrhagic lesions, and the distinction between adenoma and well-differentiated hepatocellular carcinoma remains difficult.11 Our study confirms the strict correlation between adenoma and the use of oral contraceptives; less evident is the correlation in cases of FNH, but it undoubtedly seems that the incidence in patients using sex hormones who manifested FNH is higher than the percentage of

Figure 3. Typical angiographic images of focal nodular hyperplasia of the right liver lobe. (A) Early arterial phase showing a hypervascular mass with a central feeding artery: The contrast dye rapidly lls the node from the center to the periphery. (B) In late phases the node is completely opacied, and a suprahepatic vein, selectively draining the node, is well evidenced.

demonstrate normal uptake also in 25% of cases of adenoma; four of our cases confirmed this finding.6,8,13 Laboratory tests and symptoms are not diagnostic. In case of FNH, however, there is a tendency to observe asymptomatic masses, incidentally seen, without any biochemical abnormalities.

Figure 4. A 27-year-old female with double hepatic adenoma. CT scan showed an enormous hemorrhagic and necrotic mass arising from the left liver lobe and occupying the whole left hypocondrium. Another nonhemorrhagic, contrastenhanced lesion is evident on the fourth liver segment.

Hepatic Adenoma and FNH

165

women in the general population in Italy using these drugs (68.4% v. 30%). This fact may reflect a selection bias in the study but is, in our opinion, an interesting finding. In the last 10 years, hepatic surgeons have largely improved their results, and hepatic resections are now performed safely, with low morbidity and very low mortality rates. In specialized surgical units these operations are done without any need of transfusions and with reduced hospitalization time.5,13-16 Laparoscopic surgery may, in very selected cases (superficial plongeant lesions), be an operative option.17 The prolonged duration of laparoscopic procedures compared with laparotomic techniques is still a major concern; nevertheless the duration of the operation usually does not affect the recovery of patients in overall good condition, and hospitalization time is reduced (4 v. 10.9 days in our series). An important area on which to focus when studying these lesions is the risk of not identifying malignant tumors. Two patients in our series had adenomas containing degenerated areas of hepatocarcinoma. Malignant transformation of adenoma is a rare event, but recent reports point out this possibility in an increasing percentage of cases.18 In conclusion, all the diagnostic preoperative studies in our series led to the right diagnosis in 71% of cases, with more than one quarter being misdiagnosed. It is our opinion that this is the actual limit in the treatment of these tumors. Our philosophy, therefore, is to resect all lesions preoperatively classified as adenoma, independent of their location and size. In cases of undetermined diagnosis, we usually resect all easily resectable lesions and keep under close observation those in which the risks of resection seem high; any increase in size or in imaging characteristics should be signal for excision. Asymptomatic patients who have a diagnosis of FNH based on the aforementioned typical signs are also under repeated clinical and ecographic controls.19 The extensive use of hepatic resection in such cases can be justified by offering patients the guarantee of a higher recovery rate, as well as fewer complications, which specialized liver centers offer today.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

References
1. Belli L, De Carlis L, Beati C, Rondinara GF, Sansalone CV, Brambilla G. Surgical treatment of symptomatic

18. 19.

giant hemangiomas of the liver. Surg Gynecol Obstet 1992;174:474-478. Kerlin P, Davis GL, McGill DB, Weiland LH, Adson MA, Sheedy PF. Hepatic adenoma and focal nodular hyperplasia: Clinical, pathologic and radiologic features. Gastroenterology 1983;84:994-1002. Klastin G. Hepatic tumors: Possible relationship to use of oral contraceptives. Gastroenterology 1977;73:386394. Edmonson HA, Henderson B, Benton B. Liver-cell adenomas associated with use of oral contraceptives. N Engl J Med 1976;294:470-472. Leese T, Farges O, Bismuth H. Liver cells adenomas: A 12 year surgical experience from a specialist hepatobiliary unit. Ann Surg 1988;203:558-564. Welch TJ, Sheedy PF, Johnson TM, Stephens DH, Charboneau JW, Brown ML, et al. Focal nodular hyperplasia and hepatic adenoma: Comparison of angiography, CT, US and scintigraphy. Radiology 1985;156:593595. Mathieu D, Bruneton JN, Drouillard J, Caron-Pontreau C, Vasile N. Hepatic adenomas and focal nodular hyperplasia: Dynamic CT study. Radiology 1986;292: 1355-1357. Vilgrain V, Flejou JF, Arrive L, Belghiti J, Najmark D, Meny Y, et al. Focal nodular hyperplasia of the liver: MR imaging and pathologic correlation in 37 patients. Radiology 1992;184:699-703. Coombs RJ, Woldenberg LS, Skeel RT, Bishara HM, Merrick HW. Magnetic resonance imaging of hepatic adenoma. Clin Imaging 1990;14:44-47. Bennet WF, Bova JG. Review of hepatic imaging and a problem oriented approach to liver masses. Hepatology 1990;12:761-775. Anthony PP. Tumors and tumor like lesions of the liver and biliary tract. In: MacSween RNM, Anthony PP, Scheuer PJ (eds). Pathology of the liver (ed 2). Edinburgh, Churchill Livingstone, 1987:574-645. Casarella WJ, Knowles DM, Wolf M, Johnson PM. FNH and liver cell adenoma: Radiologic and pathologic differentiation. Am J Roentgenol 1978;131:393-402. Belghiti J, Pateron D, Panis Y, Vilgrain V, Flejou JF, Benhamou JP, Fekete F. Resection of presumed benign liver tumours. Br J Surg 1993;80:380-383. Iwatsuki S, Todo S, Starzl TE. Excisional therapy for benign hepatic lesions. Surg Gynecol Obstet 1990;171: 240-246. Pain JA, Gimson AES, Williams R, Howard ER. Focal nodular hyperplasia of the liver: Results of treatment and options in management. Gut 1991;32:524-527. Habib NA, Koh MK, Zografos G, Awad RW, Bottino G. Elective hepatic resection for benign and malignant liver disease: Early results. Br J Surg 1993;80:1039-1041. Reich H, McGlynn F, De Caprio J, Budin R. Laparoscopic excision of benign liver lesions. Obstet Gynecol 1991;78:956-958. Foster JH, Berman MM. The malignant transformation of liver cell adenomas. Arch Surg 1994;129:712-717. Reddy KR, Shiff ER. Approach to a liver mass. Semin Liver Dis 1993;13:423-435.