423

Neuroblastoma is the most common extracranial solid tumor in children and the most common tumor in infants.1 The etiol-ogy of neuroblastomas is not clearly understood.1 They arise in tissues which have sympathetic nerves, such as the adrenal glands, and paramedian sympathetic chain. A neuroblastoma arising in solid organs other than the sympathetic nervous system is ex-tremely rare, and only a few cases of primary pancreatic neuro-blastoma have been reported.2 Neuroblastomas in the liver are reportedly metastatic lesions, particularly as a stage 4S adrenal gland neuroblastoma and a primary hepatic neuroblastoma has not been reported yet. A solitary hepatic neuroblastoma has been reported in one literature but it was a late hepatic metas-tasis of an adrenal neuroblastoma.3 Here, we report a case of a bona fide primary hepatic neuroblastoma in a patient who had neither history of neuroblastoma nor evidence of other neoplas-tic lesions in other organs.

CASE REPORT

A 29-year-old woman was admitted with a liver mass detect-ed by computed tomography (CT) at another hospital. On the initial visit to the other hospital, she complained of foamy urine. She had been healthy previously. A physical examination re-

vealed a palpable liver of two finger breadths and normal blood pressure. Initial laboratory tests, such as a complete blood count and liver function tests, revealed no remarkable findings. Serum and urine catecholamine levels were not measured because of the normal blood pressure and the absence of clinical suspicion. On an abdominal CT scan, a huge (16×12 cm) mass with cen-tral necrosis (Fig. 1) was found in the medial segment of the left lobe and the anterior segment of the right lobe. The mass show-ed both early and delayed contrast enhancement that proceeded from the periphery to the center. A liver needle biopsy had al-ready been performed at the other hospital before admission, and that biopsy was diagnosed histologically as a neuroendo-crine carcinoma. Because a primary hepatic neuroendocrine car-cinoma is extremely rare, a systemic survey was conducted to identify possible primary tumor deposits. Chest and abdomino-pelvic CT scans revealed nothing but the hepatic mass, and a whole body positron emission tomography-CT scan revealed no significant extrahepatic hypermetabolic foci. Because metastatic neuroendocrine carcinoma of unknown origin was the most probable clinical possibility at that time, she has received five cycles of chemotherapy (vincristine, adriamycin, and cyclophos-phamide #2/etoposide and ifosfamide #1/vinblastine, ifosfamide, and cisplatin #2) as an initial treatment. However, the tumor size did not change on a follow-up abdominal CT scan, despite

Primary Hepatic Neuroblastoma

- A Case Report -

Geunyoung Jung · Jihun Kim

Department of Pathology, University of Ulsan College of Medicine, Asan Medical Center, Seoul, Korea

Neuroblastoma is a malignant tumor of primordial neural crest origin. It usually develops along the sympathetic nervous system, such as the adrenal glands or paramedian sympathetic chain and metastasizes to the liver most frequently. However, a primary hepatic neuroblastoma has not been reported yet. Here, we report a case of 29-year-old woman who presented with a solitary hepatic mass. Grossly, the mass was large, creamy, rubbery firm, and showed focal hemorrhage and central cavitation. Microscopically, the tumor cells were arranged in small nests of spindle to ovoid cells with abundant neuropil. The neuroblastic nature of the tumor was confirmed by im-munohistochemistry and electron microscopy. No extrahepatic mass was found, despite a thor-ough systemic survey such as chest and abdominopelvic computed tomography (CT) scans and a whole body positron emission tomography-CT study. To the best of our knowledge, this is the first report of a bona fide primary hepatic neuroblastoma.

Key Words: Neuroblastoma; Liver neoplasms; Microscopy, electron; Adult

Received: November 24, 2010Accepted: December 14, 2010

Corresponding AuthorJihun Kim, M.D.Department of Pathology, University of Ulsan College of Medicine, Asan Medical Center, Pungnap 2-dong, Songpa-gu, Seoul 138-736, KoreaTel: +82-2-3010-4556Fax: +82-2-472-7898E-mail: jihunkim@amc.seoul.kr

The Korean Journal of Pathology 2011; 45: 423-427http://dx.doi.org/10.4132/KoreanJPathol.2011.45.4.423

� Geunyoung�Jung·Jihun�Kim424

the combination chemotherapy. Subsequently, she underwent a trisegmentectomy (segments 4, 5, and 8) to remove the residual hepatic mass. Neither enlarged lymph nodes nor peritoneal dis-semination was detected intraoperatively.

The resected liver specimen showed a huge (15×8.5×7.3 cm), solid mass with central cavitation. The cut surface was creamy white, rubbery firm, and focally hemorrhagic (Fig. 2A). The mass abutted the capsule and resection margin. The non-neo-plastic hepatic parenchyma adjacent to the mass was not cir-rhotic. Microscopically, the tumor cells formed nests of variable sizes that dissected the hepatic sinusoids and connective tissue (Fig. 2B). The tumor replacing hepatic parenchyma left the mul-tifocally entrapped hepatic parenchyma intact, and tissue de-struction was unremarkable. Individual tumor cells were small and relatively monotonous in their appearance and had hyper-chromatic nuclei and indistinct nucleoli (Fig. 2C). Notably, the tumor cells had abundant fibrillary cytoplasm, which was inter-mingled with neuropil-like stroma (Fig. 2D). Although the neu-ronal differentiation of this tumor was reflected by the abundant neuropil, no typical ganglion cells were found. Additionally, no typical tumor cell rosette was found. Immunohistoche mical staining was performed to confirm the neuroblastic nature of the tumor. Tumor cells were strongly reactive to synaptophysin (Fig. 2E), CD56, and neurofilament protein (Fig. 2F), support-ing the notion of neuronal differentiation. CD99 was stained weakly in fibrillary material. Cytokeratin expression was not detected. A transmission electron microscopic examination re-vealed several dense core granules and relatively well-defined microtubules, further supporting the neuronal differentiation (Fig. 3). Considering tumor morphology and the results of an-cillary studies together, we diagnosed this tumor as a neuroblas-

toma. Her initial liver biopsy slide was reviewed, and the histo-logical features of the biopsy specimen were identical to those of the resected specimen. Therefore, we corrected the diagnosis of the biopsy specimen. Because a primitive neuroectodermal tumor (PNET) can show histological features similar to the pre-sent case and has been reported in the liver,4 we perform ed a re-verse transcription-polymerase chain reaction (PCR) study to detect EWS-FLI1 and EWS-ERG1 fusion transcripts, but the results were negative (data not shown). In addition, MYCN gene amplification was not found in the PCR-based study (data not shown).

The post-operative follow-up period was uneventful and no recurrent tumor was detected on a follow-up CT examination taken 4 months after surgery. Serum and urine catecholamine levels were in the normal range. Because the diagnosis of neuro-blastoma was not made before resection, it is unknown whether the preoperative serum or urine catecholamine levels were ele-vated or not.

DISCUSSION

In the present report, we described a hepatic neuroblastoma that was presumably of primary liver origin. This case is ex-traordinary and interesting because the tumor was present only in the liver and because it developed in a previously healthy pa-tient of relatively old age. To the best of our knowledge, our case is the first primary hepatic neuroblastoma. Although spon-taneous regression or delayed recurrence of a neuroblastoma has been reported as not uncommon,5 we could not find any evi-dence of neoplastic lesions outside of the liver or any previous history of neuroblastoma.

A neuroblastoma usually occurs in organs with sympathetic nervous tissue.1 The adrenal gland is the most common organ, followed by the neck, mediastinum, and pelvis. Neuroblastoma of the liver is detected as metastatic disease along with a prima-ry adrenal tumor almost exclusively. A case of solitary hepatic neuroblastoma has been reported, but that case was a delayed metastasis after complete remission of an adrenal neuroblasto-ma.3 However, we could not detect any evidence of neuroblas-toma throughout the body in the present case.

An alternative explanation is that the solitary hepatic neuro-blastoma formed by spontaneous regression of a primary neuro-blastoma, which occurred in other organs with subsequent late hepatic metastasis. Indeed, neuroblastomas sometimes regress spontaneously. In a large-scale study investigating the natural

Fig. 1. A dynamic arterial-phase abdominal computed tomography scan image. A huge necrotic mass is located in segments 4, 5, and 8 of the liver with weak peripheral enhancement.

425Primary�Hepatic�Neuroblastoma

C D

B

E F

Fig. 2. (A) Gross appearance of the resected liver specimen. A well demarcated creamy white solid mass is replacing the resected liver. In the center is a large cavity that is filled with yellow serous fluid on arrival (removed during cutting). Some hemorrhagic spots and entrapped non-neoplastic hepatic parenchyma are evident within the tumor. (B) Microscopically, the tumor consists of nests of small blue round cells and neuropil-like stroma (H&E,×40). (C) The tumor cells are relatively monotonous and have a round, mildly hyperchromatic nuclei and a moder-ate amo unt of clear or fibrillary cytoplasm (H&E,×100). (D) Notably, the tumor cells have abundant fibrillary cytoplasm that is intermingled with neuropil-like stroma (H&E,×200). (E) Immunohistochemical staining for synaptophysin is strongly positive in the tumor cells (Synaptophysin, ×200). (F) The fibrillary matrix is stained with neurofilament protein immunostaining (Neurofilament protein,×200).

A

course of a subset of patients with a neuroblastoma (stage I or II, <5 cm in diameter) without any treatment, 13 out of 22

cases (55%) showed spontaneous regression.6 The mechanism of spontaneous regression is reportedly a type of programmed cell

� Geunyoung�Jung·Jihun�Kim426

Fig. 3. Transmission electron microscopic image. Several dense core granules (arrows) and microtubules (arrowheads) are clearly seen (scale bar=200 nm).

death associated with the accumulation of autophagic vacuoles due to lysosomal-associated protein multispanning transmem-brane 5 mediated lysosomal destabilization.7 Spontaneous re-gression is more common in infants than children and is more frequent in MYCN-non-amplified stage 4S neuroblastoma.8 However, this “spontaneous regression and late hepatic metas-tasis” hypothesis cannot be demonstrated in the present case because there was no evidence of a previous neuroblastoma.

The pathological diagnosis of the present case was relatively straightforward. A possible differential diagnosis was PNET, not only because a case of primary hepatic PNET has been re-ported but also because a PNET can show similar histological features to the present case. However, the absence of the EWS-FLI1 and EWS-ERG1 fusion transcripts along with the typical neuroblastic nature of the present case excluded the possibility of PNET.

The histogenesis of a primary hepatic neuroblastoma is un-known. However, it is plausible that a primary hepatic neuro-blastoma may arise from intrahepatic sympathetic nervous tis-sue, because sympathetic nervous system innervates the liver and plays important roles such as tissue repair and glucose me-tabolism.9,10 In agreement with this idea, a few cases of primary hepatic paraganglioma have been reported.11-14 Because neuro-blastomas and paragangliomas share a common histogenesis and there are several examples of hybrid or transitional tumors between the two,15 the present case may represent one end of the “sympathetic nervous tumor” spectrum that shows predo-minant neuroblastic features.

In this report, we described the first primary hepatic neuro-blastoma with detailed clinicopathological features. Morpho-

logically evident neuronal differentiation and the absence of MYCN gene amplification in our case suggested a less aggres-sive clinical course but the exact biological behavior of the pres-ent case remains to be determined. We suggest that neuroblas-toma should be considered in the differential diagnosis when a solitary hepatic tumor histologically resembles a neuroendo-crine tumor. Furthermore, cases similar to the present one should be collected and followed up further to define the nature of this intriguing tumor.

REFERENCES

1.EsiashviliN,AndersonC,KatzensteinHM.Neuroblastoma.CurrProblCancer2009;33:333-60.

2.ChauhanJC,KumarP,DuttaAK,BasuS,KumarA.AssessmentofdietarycompliancetoglutenfreedietandpsychosocialproblemsinIndianchildrenwithceliacdisease.IndianJPediatr2010;77:649-54.

3.KatoK,IshikawaK,ToyodaY,et al.Laterecurrenceofneuroblasto-mastage4Swithunusualclinicopathologicfindings.JPediatrSurg2001;36:953-5.

4.ManiS,DuttaD,DeBK.Primitiveneuroectodermaltumoroftheliver:acasereport.JpnJClinOncol2010;40:258-62.

5.YonedaA,OueT,ImuraK,et al.Observationofuntreatedpatientswithneuroblastomadetectedbymassscreening:a“waitandsee”pilotstudy.MedPediatrOncol2001;36:160-2.

6.OueT,InoueM,YonedaA,et al.Profileofneuroblastomadetectedbymassscreening,resectedafterobservationwithouttreatment:resultsoftheWaitandSeepilotstudy.JPediatrSurg2005;40:359-63.

7.InoueJ,MisawaA,TanakaY,et al.Lysosomal-associatedproteinmultispanningtransmembrane5gene(LAPTM5)isassociatedwithspontaneousregressionofneuroblastomas.PLoSOne2009;4:e7099.

8.BénardJ,RaguénezG,KauffmannA,et al.MYCN-non-amplifiedmetastaticneuroblastomawithgoodprognosisandspontaneousregression:amolecularportraitofstage4S.MolOncol2008;2:261-71.

9.DicostanzoCA,DardevetDP,NealDW,et al.Roleofthehepaticsympatheticnervesintheregulationofnethepaticglucoseuptakeandthemediationoftheportalglucosesignal.AmJPhysiolEndo-crinolMetab2006;290:E9-E16.

10.ObenJA,RoskamsT,YangS,et al.Sympatheticnervoussystemin-hibitionincreaseshepaticprogenitorsandreducesliverinjury.He-patology2003;38:664-73.

11.RimmelinA,HartheiserM,GangiA,et al.Primaryhepaticpheo-

427Primary�Hepatic�Neuroblastoma

chromocytoma.EurRadiol1996;6:82-5.12.RodríguezDíezR,AñóGarcíaMdelC,SuárezFernandezA,Mu-ñozRodríguezM.Primaryhepaticpheochromocytoma.MedClin(Barc)1999;113:795.

13.CortiB,D’ErricoA,PierangeliF,FiorentinoM,AltimariA,GrigioniWF.Primaryparagangliomastrictlyconfinedtotheliverandmim-ickinghepatocellularcarcinoma:animmunohistochemicalandin

situhybridizationstudy.AmJSurgPathol2002;26:945-9.14.ChangH,XuL,MuQ.Primaryfunctioninghepaticparaganglio-ma:acasereport.AdvTher2006;23:817-20.

15.KhanAN,SolomonSS,ChildressRD.Compositepheochromocy-toma-ganglioneuroma:arareexperimentofnature.EndocrPract2010;16:291-9.