Abstract
We describe the first case of primary thymic adenocarcinoma with signet ring cell features. The patient was a 39-year-old Filipino male who presented with a huge anterior mediastinal mass extending to the left supraclavicular fossa, which underwent an incisional biopsy. Extensive clinicoradiological work-up showed no evidence of any primary tumor in other organs, and radiological imaging confirmed a primary tumor location in the thymic area. He later developed bilateral pleural and pericardial effusions and eventually died of his tumor. The biopsy contained a neoplastic growth of solid nests made of cells with prominent signet ring features in sclerotic stroma. Immunohistochemically, the tumor cells were reactive for cytokeratin 7, carcinoembryonic antigen, and CD5 and negative for cytokeratin 20, TTF1, napsin A, α-fetoprotein, PAX-8, CD-117, CA19-9, CA-125, CDX2, p63, and CD99. No genetic alterations of ALK, RET, and ROS1 were found, nor was any ALK or ROS1 immunostaining detected, as known to occur in a fraction of primary pulmonary adenocarcinomas. Morphologically, this thymic tumor resembled signet ring cell adenocarcinomas of other locations, including, stomach, pancreas, and lung, but CD5 immunoreactivity strongly supported the clinical and radiological evidence of a primary thymic origin. In the English literature, only 58 cases of primary thymic adenocarcinoma are on record and this is the first report of a signet ring cell variant, which further broadens the morphological spectrum of thymic adenocarcinoma subtypes.
Introduction
Among primary thymic malignant tumors, adenocarcinoma is extremely rare and was first reported in 1989, 1 though it was not until 1997 when it was recognized as a distinctive histologic subtype. 2 To date, a total of 58 cases of primary thymic adenocarcinoma have been reported in the medical literature.3,4 It accounts for approximately 1.6% to 3.9% of thymic carcinomas.3,5
No specific clinical presentations are reported for thymic adenocarcinomas. The diagnostic algorithm usually includes radiographic examination such as computed tomography (CT), magnetic resonance imaging, positron-emission tomography, and biopsy of the mediastinal lesion. Morphologically, adenocarcinomas of the thymus are heterogeneous and do not differ much from the respective counterparts in other nearby organs. Therefore, immunohistochemical (IHC) analyses are of great help to attempt to define an origin. In general, malignancies of thymic origin show immunoreactivity for carcinoembryonic antigen (CEA), CA19-9, CD117, p63, and above all for CD5.6-10 Therefore, immunopanels including CD5, cytokeratin (CK) 5/6, TTF-1, calretinin, estrogen receptor, progesterone receptor, gross cystic disease fluid protein-15 (GCDFP-15), and CD57 were suggested to take primary mediastinal tumors apart from metastatic deposits of lung, thyroid, pleura, or breast derivation.11-13
According to the literature, thymic adenocarcinomas follow a heterogeneous course, largely depending on tumor stage and on histologic type, with papillary adenocarcinoma being more indolent than adenocarcinomas not otherwise specified (NOS) that showed a 5-year survival of approximately 50%. 3
The latest version of the World Health Organization classification of thoracic tumors 14 described several subtypes of thymic adenocarcinomas, including papillary, mucinous, adenoid cystic, and NOS carcinoma types. No mention is made to the signet ring cell type, which is a well-known feature of some gastric but also pulmonary, breast, or intestinal adenocarcinomas. Indeed, in reported cases of thymic adenocarcinomas, the presence of focal “signet ring-like” features have been recorded in 2 cases by Kwon et al 3 and 1 case by Ra et al, 15 although these cases were defined as adenocarcinoma NOS and as mucinous carcinoma in these 2 studies, respectively.
We here describe a case of primary thymic signet ring cell adenocarcinoma in a 39-year-old man. To the best of our knowledge, this entity has so far never been reported in the literature and broadens the morphological spectrum of thymic adenocarcinoma subtypes.
Case Report
A 39-year-old Filipino male was admitted to the Quirino Memorial Medical Center in Philippines in September 2014 due to persistent cough and dyspnea. The patient reported to be a smoker (24-pack year) and occasional alcohol drinker. Laboratory tests discovered anemia with neutrophilic leukocytosis, while lung opacities and bilateral pleural effusions were found at chest X-rays. CT scan revealed the presence of a lobulated mass in the anterior mediastinum measuring 7.4 × 8.1 × 13.4 cm that extended to the left hilar regions and the left supraclavicular area (Figure 1.1). The tumor had a heterogeneous enhancing pattern. Infiltration of the surrounding fat was evident, with encasement of the tracheal carina and the great vessels (aorta, pulmonary artery, and veins), but no evidence of endoluminal vascular embolization (Figure 1.2). Bilateral pleural effusions were identified. Neither pulmonary nor extrathoracic tumor masses in liver, pancreas, spleen, intestine, and adrenal glands were found. Thoracentesis and mediastinal mass biopsy (with access from the left supraclavicular infiltrated area) were performed and submitted to cytological and histopathological examination.

Panel displaying the thoracic computed tomography (CT) and the histopathology findings. (1.1) CT revealed the presence of a lobulated mass in the anterior mediastinum measuring 7.4 × 8.1 × 13.4 cm that extended to the left hilar regions and the left supraclavicular area. (1.2) The tumor had a heterogeneous enhancing pattern. Infiltration of the surrounding fat was evident, with encasement of the tracheal carina and the great vessels (aorta, pulmonary artery, and veins), but no evidence of endoluminal vascular embolization. (1.3 to 1.6) Solid nests can be observed within the tumor with predominant signet ring cells immersed in a sclerotic stroma. Tumor cells demonstrate oval-shaped, irregular, hyperchromatic nuclei with prominent nucleoli, and an abundant vacuolated cytoplasm (1.3, hematoxylin-eosin [H&E], 40×; 1.4, H&E, 100×; 1.5, H&E, 200×; and 1.6, H&E, 400×).
Cytological examination of pleural fluid showed reactive mesothelial cells with no evidence of malignant cells. The mediastinal mass biopsy consisted of a 1.5 × 1 cm tissue sample. Microscopically, a malignant tumor made of solid nests with predominant signet ring cells immersed in a sclerotic stroma was observed. Tumor cells had oval-shaped, irregular, hyperchromatic nuclei with prominent nucleoli, and an abundant vacuolated cytoplasm (Figure 1.3-1.6). No thymic tissue was identified in this biopsy that for practical purposes was taken from the tumor periphery protruding into the supraclavicular space. No necrosis was observed, and the mitotic count was relatively low, as also confirmed by a Ki67 index of 20%. Mucin stain was positive in intracellular vacuoles, and no mucin lakes were present in the tumor. By IHC, the tumor cells were strongly positive for CK7, CEA, and CD5, and negative for CK20, CDX2, CD99, PAX8, CD117, p63, TTF1, napsin A, α-fetoprotein, CA19-9, CA125, ALK, and ROS1 (Figure 2). In agreement with the IHC findings, no ALK gene rearrangement was identified by fluorescence in situ hybridization (FISH), nor did the NanoString technology using the ALK/ROS1/RET gene panel reveal any rearrangements.

Immunohistochemical findings. Mucin stain is positive in intracellular vacuoles, and no mucin lakes are present within the tumor (50×, 100×, 200×). The tumor shows strong immunoreactivity for CD5 (10×, 100×, 200×). No reactivity for TTF1 and napsin A is observed (200×).
Few days after admission, the patient developed severe shortness of breath and pericardial effusion. Cardiac tamponade was suspected and an emergency pericardiostomy was performed, but the patient died during the surgical procedure. An autopsy was not performed, as rejected by the family.
Material and Methods
Histology and Immunohistochemistry
Cytological and histological samples were anonymized by a pathology staff member not involved in this study, and only coded and de-identified materials were used throughout and the research was performed in compliance with the ethical guidelines of the Declaration of Helsinki. Pleural effusions were fixed in ethanol and processed for smears and cell block preparations, while the supraclavicular mass biopsy was fixed in formalin and processed for conventional histopathological stains, namely, hematoxylin-eosin (H&E) and mucin stains (periodic acid-Schiff after diastase treatment). Additional serial sections were obtained for FISH, Nanostring (see below), and IHC, which were performed with a standard procedure in an automated Ventana BenchMark immunostainer (Ventana Medical Systems, Tucson, AZ). The following primary antibodies were used: CK7, CK20, CEA, TTF1, napsin A, α-fetoprotein, Ki67, and ALK (from Ventana Roche, Tucson, AZ), CD5, CDX2, CD99, p63, and CA125 (from Novocastra-Leica Biosystems, Newcastle, England), CD117, CA19-9 (both from DakoAgilent, Denmark), PAX8 (Biocare Medical System Inc, Concord, CA), and ROS1 (from Cell Signaling Technology, Leiden, Holland). Positive and negative controls were included for each IHC run.
Fluorescence In Situ Hybridization for ALK
DNA FISH was performed in 4 µm-thick paraffin sections using the Vysis ALK Break Apart FISH Probe Kit (Abbott Laboratories, Abbott Park, IL), according to the manufacturer’s instructions. Briefly, the paraffin sections were deparaffinized, dehydrated, and incubated in pretreatment solution at 98°C for 15 minutes. Then the slides underwent protease digestion with pepsin at 37°C for 60 minutes (Life Technologies, Carlsbad, CA). After dehydration, probe mixture was applied to each sample. The slides were coverslipped and sealed with rubber cement. Slides and probes were co-denatured at 73°C for 3 minutes and hybridized at 37°C overnight in the dark, using the Vysis HYBrite Denaturation/Hybridization Unit. The slides were dehydrated and mounted in antifade with DAPI (4′,6-diamidino-2-phenylindole) and stored in the dark before signal evaluation. The slides were then automatically acquired at 40× magnification with the motorized Metafer scanning system (Carl Zeiss MetaSystems GmbH, Jena, Germany) and Axio Imager epifluorescence microscope.
NanoString Technology for Fusion Transcripts of ALK, ROS1, and RET Genes
NanoString uses a combination of capture and reporter probes to uniquely capture and tag a particular RNA molecule; only when the 2 probes hybridize close in the same transcript a positive signal will result. The RealQuant Lung Fusion Genes kit (Diatech Pharmacogenetics, Jesi, Italy) was employed for the detection of the fusion transcripts of the genes ALK, ROS1, and RET through NanoString technology (NanoString Technologies, Seattle, WA) that allows to detect known and unknown fusion variants of ALK, ROS1, and RET using 45 molecular probes in a single multiplex reaction. In the same multiplex reaction are included further 27 assays and related probes: 21 for reference genes necessary for data normalization and 6 for genes differentially expressed in normal and tumor lung tissue, useful for the evaluation of RNA quality and quantity and for neoplastic cell content. Briefly, after assessing tumor cellularity on a H&E-stained slide, four 10 µm-thick tissue sections were collected in an Eppendorf tube. RNA extraction was performed using a Roche FFPET RNA Isolation Kit (Roche, Mannheim, Germany). RNA concentration was assessed using NanoDrop 2000 Spectrophotometer (Thermo Fisher Scientific, Grand Island, NY). Total RNA hybridization and nCounter analyses were performed according to the RealQuant kit instructions. Data analysis was automatically performed by iGENETICS RealQuant software (Diatech Pharmacogenetics).
Discussion
We report the first case of primary thymic adenocarcinoma with extensive signet ring cell features. Among thymic carcinomas, only 58 adenocarcinomas are on record in the English literature, 3 and focal signet ring-like features have been mentioned in only 2 studies.3,15
Although the patient could not undergo radical resection due to the huge tumor size (up to 13 cm) and the extensive infiltrative growth in the mediastinal organs, this diagnosis is supported by radiological data that prove an anterior mediastinal location with an extensive infiltration of mediastinal vessels and the trachea and also partially encasing the esophagus. In addition, the immunophenotypic profile revealed a remarkable expression of CD5, together with the less specific staining for CK7 and CEA, and in the complete absence of reactivity for other commonly used markers of pulmonary or intestinal origin. These findings will be shortly commented.
Histologically, the current case appeared to contain signet ring cells throughout the tumor specimen, in the absence of mucin lakes or enteric-type tubular formations. This tumor shared the known morphological features of signet ring cell adenocarcinomas of other organs, including a glandular or more commonly a solid nest growth of vacuolated atypical cells, positive for mucin stains. Mitotic figures were common and the Ki67 proliferative index was elevated, consistent with the tumor size and the extensive infiltrative growth observed at diagnosis. The possibility of a metastatic spread of a signet ring cell carcinoma to the thymus and mediastinum was immediately considered. Searching the English literature, the first mention of signet ring cell-like features in a tumor confined to the anterior mediastinum dated back to 2007. 15 Histologically, this tumor had scattered signet ring cell in abundant extracellular mucin lakes and expressed focally CD5 and strongly CK7, which lead to final diagnosis of primary thymic mucinous adenocarcinoma. 15 Such reported diagnosis is supported by the current World Health Organization classification of thymic tumors, since signet ring cell-like features coupled with extracellular pool of mucin lakes feature enteric-type mucinous carcinomas. 16
Conversely, our case cannot exactly be classified as a mucinous (colloid) adenocarcinoma in the absence of the definitional criterion of presence of extracellular mucin lakes, nor as an adenocarcinoma NOS, in the absence of an enteric immunophenotyped. 16
A large panel of markers has been investigated to characterize the nature of this extremely uncommon growth pattern for a malignant neoplasm in the thymic region. Regarding thymic-specific markers, CD5 expression reliably indicates thymic derivation of malignant epithelial tumor, mainly of the squamous type. Nevertheless, even the rare adenocarcinomas of primary thymic origin are reported to express, more or less extensively, such biomarker.14,17 On the other hand, CD117 is reported to label squamous carcinomas but not adenocarcinomas of thymic origin. 14
For the purpose of differential diagnosis, specific products of pulmonary adenocarcinoma were primarily tested. TTF1 is reported to stain up to 85% of signet ring cell carcinomas of the lung and napsin A stains a large fraction of lung adenocarcinomas. 14 The absence of reactivity in both markers allowed us to presume an origin other than pulmonary. Other sources of signet ring cell carcinomas are the stomach, biliary tract, and the intestine, which are generally reported to express a special type of CK20 (100% of colorectal, 62% of pancreatic, 50% of gastric adenocarcinomas, and 43% of cholangiocarcinoma) and also CDX2 (in 90% of gastric and of 89% colon cancers). 18 Both markers were negative in the currently described adenocarcinoma; thus, an intestinal origin of the tumor was disregarded. Finally, a limited set of molecular markers was investigated, targeting the most common ones associated to signet ring cell features in lung adenocarcinoma. To this purpose, ALK, together with RET and ROS1 rearrangements were tested using IHC, FISH, and gene expression procedures. None of the above-mentioned genes resulted in alteration in the current case, thus failing to support the potential pulmonary origin of the tumor.
A relevant differential diagnosis is with benign or low-grade thymomas since this tumor may occasionally display signet ring cell features. As a rule, overt invasive growth and cytological atypia, as observed in the current case, favors thymic carcinoma over thymoma. 19 In addition, the IHC profile helps confirm the malignant thymic tumor phenotype with expression of CK7 and of CD5 (absent in thymoma), 8 and no immunoreactivity for p63 nor CD99 (both positive in thymoma),16,19 a profile also documented in the current case.
The role of radiologists in diagnostic process regarding thymic neoplasias is of immense importance. 20 Thymic adenocarcinomas often show invasion of surrounding structures, usually accompanied by lymph node and distant metastasis. Frequently, the malignant disease is biopsy-proven; therefore, radiological imagining is fundamental in order to provide correct diagnosis. 14 In this case an extensive tumor growth was documented up to the left supraclavicular fossa were indeed biopsy sample was taken.
In conclusion, the absence of any extra-mediastinal location together with the specific thymic carcinoma immunoprofile (in the absence of lung adenocarcinoma biomarkers) supported the thymic origin of this tumor. This extremely rare adenocarcinoma variant further expands the spectrum of reported subtypes of primary adenocarcinoma of the thymus and includes this organ among the possible sites of origin of signet ring cell carcinomas.
Footnotes
Authors’ Note
This study was presented in part at the Hong Kong International Academy of Pathology meeting (November 2015) and the ITMIG meeting in Turin (September 2017).
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Ethical Approval
Not applicable, because this article does not contain any studies with human or animal subjects.
Informed Consent
Not applicable, because this article does not contain any studies with human or animal subjects.
Trial Registration
Not applicable, because this article does not contain any clinical trials.
