A 72-year-old man with a history of chronic alcoholism and cirrhosisChild score A
was referred to the oncology department of a tertiary hospital in Cali, Colombia,
for assessment of a growing mass in his oral cavity. Additionally, a mass located on
the left adrenal gland was detected during the surveillance cirrhosis controls. On
examination, an exophytic lesion of approximately 10 cm on the left mandible was
noted. He was hospitalized so laboratory tests and procedures could be performed to
establish the primary cancer diagnosis and treatment.Serologic tests were negative for hepatitis B and C panels. Serum alpha-fetoprotein
(AFP) and carcinoembryonic antigen were not significantly altered. Abdomen magnetic
resonance imaging (MRI) and computed tomography (CT) scan showed changes consistent
with cirrhosis; no hypervascular changes or signs of hepatocellular carcinoma (HCC)
were detected (Fig 1). Positron emission
tomography–CT detected abnormal hypermetabolic activity in the mandible
corresponding to the mass, pelvis bones, and adrenal mass, suggesting neoplastic
lesions, but no significant activity in the liver (Fig
2). Biopsies of adrenal and mandible lesions showed morphology and
immunohistochemistry consistent with hepatoid differentiation in carcinoma. These
findings did not, however, correlate with the imaging evidence.
Fig 1
Axial sequence volumetric interpolated breath-hold examination T1-weighted
images. (A) Early arterial, (B) arterial, (C) portal venous, and (D)
equilibrium phases. Altered hepatic segmentation and contours compatible
with chronic liver disease; diffuse changes are shown in the enhancement
pattern of focal lesions with no determination of hypervascular
behavior.
Fig 2
[18F]Fluorodeoxyglucose positron emission tomography
(PET)–computed tomography (CT) scan. (A) coronal and (B) saggital
whole-body images showed (C) expansile lytic hypermetabolic lesion on the
left side of mandible associated with (D) left adrenal hypermetabolic mass
and (E) lytic hypermetabolic lesions on the right iliac bone and sacrum.
Heterogeneous hepatic distribution of radiotracer without visualization of
focal lesions on PET and CT images.
Axial sequence volumetric interpolated breath-hold examination T1-weighted
images. (A) Early arterial, (B) arterial, (C) portal venous, and (D)
equilibrium phases. Altered hepatic segmentation and contours compatible
with chronic liver disease; diffuse changes are shown in the enhancement
pattern of focal lesions with no determination of hypervascular
behavior.[18F]Fluorodeoxyglucose positron emission tomography
(PET)–computed tomography (CT) scan. (A) coronal and (B) saggital
whole-body images showed (C) expansile lytic hypermetabolic lesion on the
left side of mandible associated with (D) left adrenal hypermetabolic mass
and (E) lytic hypermetabolic lesions on the right iliac bone and sacrum.
Heterogeneous hepatic distribution of radiotracer without visualization of
focal lesions on PET and CT images.The history of cirrhosis along with the hepatoid characteristics of the adrenal and
mandibular tumors suggested metastatic HCC. However, the absence of a liver tumor on
imaging raised the possibility of an adrenocortical carcinoma with hepatoid
differentiation, a rare tumor with an even rarer presentation.[1,2]The patient experienced progression 1 year later despite two treatment lines. At this
time, a decision to perform an exploratory laparoscopy with liver biopsy was made.
At laparoscopy, the liver was cirrhotic and diffusely nodular without a dominant
mass. Numerous representative biopsies were taken from different areas. They
revealed histologic and immunophenotypic findings of infiltrative HCC.Tumor markers were taken serially; initial AFP was 9.4 ng/mL (normal values, <
10.0 ng/mL), and during follow-up, AFP values were as follows: 7.3, 3.17, and 4.8
ng/mL. Carcinoembryonic antigen values were < 0.5 ng/mL (normal values,
< 3.0 ng/mL). Both tumor markers were considered nonsignificant.The adrenal biopsy reported that no normal adrenal gland tissue was seen. Regarding
the immunohistochemical markers, the tumor was focally positive for pancytokeratin,
Hep-Par-1, and arginase-1 (Arg-1), consistent with hepatoid differentiation. The
tumor cells were negative for AFP, glypican-3, and thyroglobulin. All adrenal
markers, including inhibit, melon-A, calretinin, and podoplanin, were negative. The
proliferation index was high (Ki-67), and CD10 highlighted a canalicular pattern
within the tumor (Fig 3). The mandibular biopsy
showed morphology and immunoprofile results similar to those of the adrenal tumor
(Fig 3). The report concluded with a
diagnosis of oncocytic carcinoma with hepatoid features, and clinical and radiologic
correlation was recommended.
Fig 3
(A) Mandible biopsy (hematoxylin and eosin [HE] ×20). The biopsy was
composed entirely of large polygonal oncocytic cells arranged in trabeculae
and clusters. (B) Left adrenal core biopsy (Hep-Par-1 immunohistochemical
stain). (C) Left adrenal core biopsy (arginase-1 [Arg-1] immunohistochemical
stain). The tumor was positive for Hep-Par-1 and Arg-1, consistent with
adrenal gland with hepatoid differentiation. (D) Liver biopsy (HE).
Histologic sections show loss of tissue architecture, with hepatocytes
arranged in trabeculae or acini. (E) Liver biopsy (cytokeratin AE1/AE3).
Diffusely and weakly positive immunostain. (F) Liver biopsy. (Hep-Par-1
immunohistochemical stain). Granular cytoplasmic staining pattern.
(A) Mandible biopsy (hematoxylin and eosin [HE] ×20). The biopsy was
composed entirely of large polygonal oncocytic cells arranged in trabeculae
and clusters. (B) Left adrenal core biopsy (Hep-Par-1 immunohistochemical
stain). (C) Left adrenal core biopsy (arginase-1 [Arg-1] immunohistochemical
stain). The tumor was positive for Hep-Par-1 and Arg-1, consistent with
adrenal gland with hepatoid differentiation. (D) Liver biopsy (HE).
Histologic sections show loss of tissue architecture, with hepatocytes
arranged in trabeculae or acini. (E) Liver biopsy (cytokeratin AE1/AE3).
Diffusely and weakly positive immunostain. (F) Liver biopsy. (Hep-Par-1
immunohistochemical stain). Granular cytoplasmic staining pattern.RNA in situ hybridization for albumin was performed using RNAview (Affymetrix,
Cambridge, MA) on the adrenal core biopsy and was positive, confirming hepatoid
differentiation and suggesting that the tumor was either from the liver or an
unusual hepatoid variant of adrenal cortical carcinoma.The liver biopsy showed a neoplastic lesion composed of hepatocytic cells arranged as
tubules and rosettes. Immunohistochemistry analysis showed positive results for
cytokeratin AE1/AE3 and HePar-1. The markers C7-C20 and AFP were negative. The
proliferation index (Ki-67) was 30%. The report concluded that the histology and
immunochemistry markers were consistent with a diagnosis of HCC (Fig 3).
DISCUSSION
Here we discuss a case of HCC presenting as metastatic tumor in the mandible and
adrenal gland, without evidence of a dominant hepatic primary lesion on imaging. The
patient had as first clinical manifestation a mandibular mass and was subsequently
found to have a left adrenal tumor, both with hepatoid differentiation. Multiple
imaging tests (CT scan, MRI, and positron emission tomography–CT) failed to
detect a definitive liver tumor, and serum biomarkers for HCC remained negative.
From this point on, some hypotheses were proposed, the first one being metastatic
HCC. Although < 1% of cases of HCC involve oral metastases[3] and approximately 5% of HCCs may
initially present as extrahepatic metastases,[4] it is important to note that, in the group of patients with
HCC metastasis to the oral cavity, in approximately 66% of cases, a metastatic oral
tumor is noted before the primary hepatic lesion.[5]The second hypothesis regarding a possible diagnosis was a hepatoid carcinoma of the
adrenal gland. This type of cancer is extremely rare, with histopathologic features
that mimic those of HCC. It is aggressive and tends to raise serum markers such as
AFP[1,3]; however, serum markers and immunohistochemistry
analysis were negative in our patient.The current guidelines[6,7] state that the diagnosis of HCC in
cirrhotic patients should be confirmed based on imaging or, less frequently, on
biopsy analyses. Imaging criteria consist of detection by at least three-phase
contrast-enhanced CT or MRI of the HCC radiologic hallmark,[6] characterized by intense arterial
uptake or enhancement followed by contrast washout or hypointensity in the delayed
venous phase.[6] Liver lesions
< 1 cm should be evaluated by at least three-phase contrast-enhanced CT or
MRI every 3 to 6 months. Liver lesions > 1 cm should first be evaluated by
one of the two imaging techniques mentioned. A finding of two classic enhancements
is considered to be diagnostic of HCC.[6] The radiologic pattern identifies HCCs with limited levels of
sensitivity but up to 100% specificity, depending on the size of the liver nodules
detected.[8] Biopsy analyses
are recommended for focal hepatic lesions with atypical imaging features or
ambiguous findings on CT or MRI, for lesions > 1 cm if only one or nonclassic
enhancement patterns are present on the results of the two imaging techniques, or
for lesions detected in the absence of cirrhosis.[6,9] The biggest
difficulty in our patient case was the lack of findings suggestive of HCC on
imaging, which is the main diagnostic criterion. According to a recent meta-analysis
and systematic review,[10] in which
the test performance of imaging techniques for the detection of HCC was measured, in
nonsurveillance settings, CT had a sensitivity of 83% and a specificity of 91% for
HCC diagnosis. MRI had a sensitivity and specificity of 86% and 89%,
respectively.Liver biopsy can enable diagnosis of HCC in 70% to 90% of cases,[11]and immunohistochemistry analysis
should be used as an ancillary tool in the diagnosis of HCC. The triad of Arg-1,
Hep-1, and glypican-3 has been recommended as the most effective method of
determining metastatic carcinoma HCC, with Arg-1 being the most specific
marker.[12] Mandible and
adrenal gland samples were positive for Hep-Par-1 and Arg-1, supporting the hepatoid
nature of these lesions.Genetic analysis consisted of detecting albumin in the adrenal gland biopsy via RNA
in situ hybridization, which confirmed the hepatoid characterization of the sample.
This test is considered highly sensitive in poorly differentiated HCC versus
immunohistochemistry, and the combination of Arg-1 and genetic analysis
significantly improves diagnostic accuracy.[13]Finally, several liver biopsies were taken, and their pathologic analysis revealed a
morphologic and immunohistochemical pattern compatible with infiltrative metastatic
HCC. This subtype is characterized by a diffuse and ill-defined phenotype that
corresponds to 7% to 20% of cases of HCC. It is considered a diagnostic challenge
because of the difficulty of distinguishing, in the imaging, cancerous cells from
background changes in cirrhosis. Whereas for nodular HCC the radiologic criteria are
defined, for infiltrative HCC they are unclear, because imaging techniques fail to
detect this subtype in approximately 40% of cases.[4,14,15] The macroscopic pattern is
extrapolated to images, with a permeative and ill-defined appearance, which present
an inconsistent uptake in the arterial phase, reported as minimal or as patches that
can be visualized as iso- or hypointense. A diagnostic key for this subtype of HCC
is the presence of portal vein thrombosis, which is found to be present in 68% of
cases.[4,14,15] However,
none of these imaging signs were present on our patient’s images.To conclude, this is a rare case of HCC presenting with oral and adrenal masses but
with no definitive liver tumor on imaging and serologic testing. This case
highlights the importance of early pathologic diagnosis of HCC in cases where there
is a clinical suggestion of HHC and images are inconclusive. The clinical team must
take into account the different morphologic subtypes of HCC, not only the nodular
subtype that is widely described by current literature. The infiltrative variant of
HCC is diagnostically challenging and may lead to a late diagnosis.
Authors: Aram Demirjian; Peter Peng; Jean-Francois H Geschwind; David Cosgrove; Jacob Schutz; Ihab R Kamel; Timothy M Pawlik Journal: J Gastrointest Surg Date: 2011-07-02 Impact factor: 3.452
Authors: Peter J Kneuertz; Aram Demirjian; Amin Firoozmand; Celia Corona-Villalobos; Nikhil Bhagat; Joseph Herman; Andrew Cameron; Ahmet Gurakar; David Cosgrove; Michael A Choti; Jean-Francois H Geschwind; Ihab R Kamel; Timothy M Pawlik Journal: Ann Surg Oncol Date: 2012-04-03 Impact factor: 5.344
Authors: Mohammad Shahid; Aysha Mubeen; Julie Tse; Sanjay Kakar; Adrian C Bateman; Darrell Borger; Miguel N Rivera; David T Ting; Vikram Deshpande Journal: Am J Surg Pathol Date: 2015-01 Impact factor: 6.394
Authors: Roger Chou; Carlos Cuevas; Rongwei Fu; Beth Devine; Ngoc Wasson; Alexander Ginsburg; Bernadette Zakher; Miranda Pappas; Elaine Graham; Sean D Sullivan Journal: Ann Intern Med Date: 2015-05-19 Impact factor: 25.391
Authors: Arich R Reynolds; Alessandro Furlan; David T Fetzer; Eizaburo Sasatomi; Amir A Borhani; Matthew T Heller; Mitchell E Tublin Journal: Radiographics Date: 2015 Mar-Apr Impact factor: 5.333