Anatomic Pathology / COMPARATIVE GENOMIC HYBRIDIZATION ON FOCAL NODULAR HYPERPLASIA Comparative Genomic Hybridization Synchronous Occurrence of Focal Nodular Hyperplasia and Hepatocellular Carcinoma in the Same Liver Is Not Based on Common Chromosomal Aberrations

Occasionally hepatocellular carcinomas (HCCs) occur synchronously with or within focal nodular hyperplasias (FNHs), raising the question of a putative causal relationship. In the present study, we used comparative genomic hybridization to investigate the occurrence of genomic aberrations in FNHs, which might lead to hepatocarcinogenesis. Tissue samples from FNHs and nonlesional liver tissue were obtained from 7 women. None of the patients had a chronic diffuse liver disease. A synchronous HCC not spatially related to FNH was present in 1 patient. Two patients had received oral contraceptives. Genomic aberrations were found in only 1 FNH. No aberration was found in the FNH occurring synchronously with HCC, but the HCC included gains at chromosomes 1q, 5, 12, and 19q and losses at 4p, 7q22-q35, 9p, 17p, 21q, and 22q. No aberrations were found in nonneoplastic liver tissues. Our findings support the notion that FNH is not a preneoplastic lesion for the occurrence of HCC in humans and that the synchronous occurrence of FNH and HCC is coincidental in our case. Hepatocellular carcinomas (HCCs) might evolve from precancerous lesions, such as dysplastic nodules, and, occasionally, from liver cell adenomas.1,2 Study of these lesions might facilitate understanding of hepatocarcinogenesis. Identification of candidate oncogenes and tumor suppressor genes for hepatocarcinogenesis is a major challenge and might permit identification of patients at risk and an evaluation of targets for therapeutic treatments. It generally is accepted that cancer proceeds through accumulation of mutations in genes that govern cell proliferation and death. HCCs display genomic alterations, including DNA rearrangements, loss of heterozygosity, chromosomal amplifications, loss of imprinting, and mutations. Different genes have been implicated in the pathogenesis of HCC and may be divided into 4 major groups: (1) genes regulating DNA damage response (P53 pathway: p53), (2) genes involved in cell cycle control (retinoblastoma [RB1] pathway: RB1, p16INK4A, cyclin D), (3) genes involved in growth inhibition and apoptosis (transforming growth factor beta pathway: mannose-6-phosphate/insulin-like growth factor 2 receptor [M6P/IGF2R], SMAD2, SMAD4), and (4) genes responsible for cell-cell interaction and signal transduction (adenomatous polyposis coli [APC]/beta-catenin pathway: APC, beta-catenin, E-cadherin) (for a review see Öztürk3). Hepatocarcinogenesis could be mediated by loss of heterozygosity (RB1, M6P/IGF2R gene, E-cadherin gene, BRCA2 [breast cancer 2]), somatic mutation (p53, RB1, p16INK4A, M6P/IGF2R, SMAD2, SMAD4, beta-catenin, APC, BRCA2), de novo methylation (p16INK4A, E-cadherin gene), and/or functional inactivation (p53). The pattern of genomic alterations shows great variability. As yet there is no evidence for an ordered sequence of Am J Clin Pathol 2003;119:265-271 265 265 DOI: 10.1309/EF69VNDLVPWVE4QV 265 © American Society for Clinical Pathology Kellner et al / COMPARATIVE GENOMIC HYBRIDIZATION ON FOCAL NODULAR HYPERPLASIA genomic events leading to hepatocarcinogenesis, and the absence of an obviously inherited predisposition to liver cancer has hampered the identification of critical genes in hepatocarcinogenesis, making synchronous liver cell tumors in a noncirrhotic liver particularly interesting for the study of tumorigenesis. The aim of the present study was to prove by comparative genomic hybridization (CGH) analyses whether synchronous FNH and HCC in a patient without evidence of a chronic diffuse liver disease might have been related to a common genomic aberration. Data obtained in this particular case were compared with data for 6 additional FNHs obtained from patients without synchronous HCC. Materials and Methods Seven FNHs and 1 HCC from 7 patients were used in the present study. All patients underwent operation between February 1998 and February 2001. Patients’ ages ranged from 22 to 54 years (mean, 43 years; median, 48 years). The diameter of the resected tumors ranged from 1.8 to 16.0 cm (mean, 7.0 cm; median, 8.9 cm) ❚Table 1❚. One patient had both an FNH and an HCC (case 4, Table 1). Tissue samples (FNHs, HCC, and nonlesional liver tissue) for CGH were obtained immediately after surgery, snap frozen in liquid nitrogen, and stored at –80°C until further use. Histologic Specimen Processing The remaining tissue was used for routine histologic examination. Tissue samples from every tumor and nonneoplastic liver were fixed in 10% neutralized formalin and embedded in paraffin. Deparaffinized serial sections were stained using H&E, periodic acid–Schiff reagent with and without diastase pretreatment, Masson trichrome stain, reticulin stain, and iron stain. Immunostaining was performed with polyclonal antibodies directed against alpha fetoprotein (DAKO, Glostrup, Denmark) as described elsewhere.4 DNA Preparation and CGH For DNA preparation, a small frozen tissue piece (average, approximately 2 mm3) was dissected using a scalpel and dissolved in lysis buffer of the QIAamp DNA Mini kit (Qiagen, Hilden, Germany). Further DNA preparation, including RNA digestion, was performed according to the manufacturer’s instructions. CGH was performed as described by Arnold et al,5 with few modifications. In brief, genomic DNA was labeled with digoxigenin11–deoxyuridine triphosphate and biotin-16–deoxyuridine triphosphate (Roche Diagnostics, Penzberg, Germany) for reference and test DNA. Reaction conditions were optimized to yield labeled fragment sizes ranging from 200 to 2,000 base pairs. Two hundred to 300 ng of biotinylated tumor DNA was ethanol coprecipitated with the same amount of labeled control DNA and 10 to 15 μg of unlabeled human Cot-1 DNA (Gibco BRL, Life Technologies, Eggenstern, Germany). The probe mixture was dissolved in 10 μL of hybridization buffer (50% formamide, 10% dextran sulfate, 2× standard saline citrate [SSC]), denatured, and, after preannealing (30 minutes), hybridized onto normal metaphase chromosomes (Vysis, Bergisch Gladbach, Germany) for 2 to 3 days at 37°C in a moist chamber. Posthybridization washes were performed without formamide as follows: 1× phosphate-buffered detergent (PBD) (Appligene Oncor, Illkirchen, France) at room temperature for 5 minutes, 2× SSC at 70°C for 5 minutes and again 1× PBD at room temperature for 5 minutes. The DNA was detected using anti–digoxigeninrhodamine (red) and streptavidin–fluorescein isothiocyanate conjugates (green). Slides were counterstained with 4',6 diamidino-2-phenylindole-dihydrochloride at a concentration of 0.1 μg/mL in an antifade solution (Appligene Oncor). 266 Am J Clin Pathol 2003;119:265-271 266 DOI: 10.1309/EF69VNDLVPWVE4QV © American Society for Clinical Pathology ❚Table 1❚ Clinical Characteristics and Comparative Genomic Hybridization (CGH) Results Case No./ Sex/Age (y) Diagnosis Size (cm) Site (Segment) Hormone Use CGH 1/F/22 FNH 5.6 S 7 None No aberrations 2/F/40 FNH 2.0 S 7 None No aberrations 3/F/43 FNH 9.0 Right lobe None No aberrations 4/F/47 FNH 3.5 S 7 and 8 None No aberrations HCC 16.0 S 5 and 6 None +1q, –4p, +5, –7q22-q35, +12, –17p, +19q, –21q, –22q, possible –3 and –9 5/F/48 FNH 10.2 Left lobe Oral contraceptives +1p33-p36.3, –1p22-p31, –5q14-q23, –6q11.1-q23, +9q34, –12q21, +12q24.1-q24.3, –14q14-q31, +16, +17p, +17q25, +19, +20q, +22, possible –4q 6/F/49 FNH 7.5 S 6 None No aberrations 7/F/54 FNH 1.8 S 3 Climanest No aberrations FNH, focal nodular hyperplasia; HCC, hepatocellular carcinoma; S, segment.