Signet-ring cells in pleural and peritoneal effusions identified on Wright stains – A diagnostic pitfall

INTRODUCTION

Signet-ring cell (SRC) carcinoma is one of the most subtle malignancies causing serous effusions. These cells typically present as single cells, with cytoplasmic vacuoles pushing the nucleus to the periphery of the cell. The nuclear-to-cytoplasmic ratio is low, and the nuclei can be bland with subtle atypia. These malignancies can be difficult to diagnose even when correlating with surgical specimens on histologic sections. SRCs are associated with many malignancies, most commonly gastric, breast, and ovarian carcinomas. SRCs have also been reported in many other carcinomas, including colonic adenocarcinoma, thyroid, squamous cell, prostatic adenocarcinomas, pancreatic colloid carcinoma, mucin-producing salivary gland adenocarcinomas, extramammary Paget’s, and lung adenocarcinoma, as well as tumors of alternate histogenesis including lymphoma, gastrointestinal stromal tumor, and melanoma.^[1-13] Diagnosis of malignant SRC in serous effusions can be very challenging, and ancillary studies are frequently utilized to confirm the diagnosis.

SRC change is not only limited to mucin-producing tumors but has also been described in benign reactive processes.^[14-16] Cytoplasmic vacuolization can occur as a degenerative change in histiocytes and mesothelial cells, two ubiquitous components of serous effusions. These degenerative vacuoles can compress the nucleus and push it to the periphery of the cell, resulting in a perfect mimicker of malignant SRC. SRC change has also been observed as an artifact of cell shrinkage or procedural mechanical trauma to tissues, especially in Wright-stained slides.^[17,18] While mucin stains can be utilized to distinguish the secretory mucin of malignant SRC from the degenerative vacuoles, they are not always positive in SRC adenocarcinoma and can be difficult to interpret. Importantly, mucin stains are not available for the Wright-stained cytospins, which are already known to be problematic for specific diagnosis.

Malignancy involving serous cavity is an unfavorable sign. Prognosis is poor, and most patients survive less than several months after presentation. Treatment choices are limited to palliative care for the majority of patients once diagnosis is confirmed. Thus, cytological diagnosis of SRC carcinoma is crucial for management decisions. Misinterpretation as SRC carcinoma without correlation may have serious consequences including premature overtreatment or inappropriate hospice referral.

The purpose of this study is to examine morphologic features of benign SRC mimickers in serous effusion, summarize features that distinguish benign from malignant SRCs, and understand the clinical scenarios in which benign SRCs are seen. We further investigate the appearance of benign SRCs in preparations commonly utilized by cytologic methods (Papanicolaou stain) versus methods typical of a hematology laboratory (Wright stain). Morphologic characteristics between the two methods are also compared.

MATERIAL AND METHODS

The Cleveland Clinic Institutional Review Board granted permission for this study. Consecutive cases (n = 400) of Wright-stained body fluid samples received in the hematology laboratory from October 2013 to January 2014 were reviewed. Serous effusions with SRC-like cells were retrieved. Corresponding cytology specimens that were collected at the same time or within 1 week of the hematology specimen were retrieved and reviewed. For patients with multiple specimens available, only the specimens submitted simultaneously to both hematology and cytology laboratories were selected for examination.

Body fluid received in the hematology laboratory was prepared by Cytospin (Thermo Scientific, Waltham, Massachusetts) method, and slides were Wright stained. Specimens received in the cytology laboratory were prepared using ThinPrep (Hologic, Marlborough, Massachusetts) method, and slides were Papanicolaou stained.

To compare the morphology of benign SRC-like cells and malignant SRC, serous effusion specimens with a cytology diagnosis of “positive for malignancy, adenocarcinoma with signet-ring cell features” were retrieved from the Anatomic Pathology CoPathPlus (Cerner Corporation, Kansan City, Missouri) database at Cleveland Clinic from April 2012 to April 2013. A total of six cases were identified. Both Wright-stained slides from hematology laboratory and the corresponding Papanicolaou-stained slides from cytology laboratory were reviewed. The clinical data were obtained from the institution’s electronic medical record system (Epic Systems Corporation, Verona, Wisconsin).

RESULTS

Morphology of SRC-like cells and malignant SRC

On a Wright stain, SRC-like cells typically present as single cells within a chronic inflammatory cell background. The SRC-like cells have space-occupying cytoplasmic vacuoles that push the nucleus to a peripheral location. The nucleus is small, curved, and indented by the vacuole(s) [Figures 1 and 2]. Nucleoli are typically not seen. The majority of SRC-like cells have one large cytoplasmic vacuole; however, small numerous vacuoles can also be seen. In some cells, the vacuole overlies the partially folded nucleus.

Close

Figure 1:

(a) Specimen from a patient with liver cirrhosis prepared by cytospin and Wright-stained. Many SRC-like cells are seen. (b) Same specimen prepared by ThinPrep and Papanicolaou-stained showed no signet-ring cells (×4).

Export to PPT

Close

Figure 2:

Representative images of benign signet-ring cells on Wright-stained slides (×40).

Export to PPT

Compared with SRC-like cells, malignant SRCs have larger, hyperchromatic nuclei with irregularly condensed chromatin. The nuclei are pushed to the periphery of the cells like their benign counterparts but one side of the nuclei is typically flat and not indented by the cytoplasmic vacuoles. The cytoplasmic vacuoles are usually small and numerous [Figures 3 and 4]. On Papanicolaou-stained slides, these cells show similar morphology, with large irregular nuclei and nuclear eccentricity, low nuclear-tocytoplasmic ratio, abundant vacuoles in the cytoplasm, and numerous nucleoli. Some cases can be very challenging with subtle atypia and can be mistaken for mesothelial cells or histiocytes with degenerative changes. Knowledge of a history of malignancy and comparison to an existing surgical pathology specimen, if present, can greatly help to make the correct diagnosis.

Close

Figure 3:

Representative images of malignant signet-ring cells on Wright-stained slides (×40).

Export to PPT

Close

Figure 4:

Malignant SRCs on Papanicolaou-stained slides (right). These single cells may be subtle and can be missed on routine cytologic staining (×20). Wright stain (left) makes nuclear atypia more apparent (×20).

Export to PPT

DISCUSSION

The aim of this study was to examine pleural and peritoneal effusions with benign signet-ring cells and compare morphology to the malignant counterpart. The morphology of serous fluids across several different preparations has been well-described in the literature as well as recent reference books.^[19] Of the 400 Wright-stained body fluids, 18 cases (4.5%) harbored prominent signet-ring cells. Of these cases, 13 cases (72.2%) had liver cirrhosis, 3 cases (16.7%) had end-stage renal disease (ESRD), one case had pancreatic adenocarcinoma (5.5%), and one case had uterine adenocarcinoma. Signet-ring cells were commonly found in patients with hepatic cirrhosis on examination of Wright-stained fluids [Figure 1]. An inflammatory cell background (lymphocytes and neutrophils) was noted in all of the benign cases. Hence, the finding of signet-ring cells can be the result of cellular degeneration secondary to inflammation in patients with chronic illnesses and repeated effusions, as demonstrated in this study in clinical settings of liver cirrhosis or ESRD. These cells were only identified on Wright-stained slides and Papanicolaou-stained smears and they were not identified on ThinPrep slides, indicating that those cells were likely disrupted by the ThinPrep method.

The typical morphology of SRC-like cells includes prominent cytoplasmic vacuolization and small, curved nuclei indented by the vacuole(s) [Table 1]. Nucleoli are typically not seen. Most cells have one large vacuole, but small numerous vacuoles may also be seen. In contrast, the nuclei of malignant SRCs are larger with higher nuclear-to-cytoplasmic ratio, the cytoplasm more often contains multiple vacuoles, and the cells are not distended. Overall, the malignant cells have a more rigid appearance.

At Cleveland Clinic, body fluid prepared in the hematology laboratory is signed out by medical technologists and hematopathologists. Unless there are questions about the diagnosis, these slides are typically not reviewed by cytopathologists. As SRC-like cells are not uncommon findings in the hematology laboratory, both hematopathologists and laboratory technicians become immune to their sighting in daily practice. When there is morphologic suspicion for malignancy, correlation with the cytology preparation and cell block findings is warranted. This is especially important in cases with SRC-like morphology, which can have a bland cytologic appearance in cytology preparations and, paradoxically, an alarming cytologic appearance in the Wright-stained slides. Misclassification of these cells may lead to inappropriate treatment such as extensive workup to detect a metastatic primary, excessive treatment, or premature end-of-life decisions.

As many body fluid specimens are prepared by ThinPrep in the cytology laboratory, benign SRCs are not seen with Papanicolaou staining. Malignant SRCs present as a second population in serous effusions [Figure 3]. Still, some cases can be very challenging as they have only slightly larger nuclei with no or subtle atypia compared with mesothelial cells. Usually, these cells are noticed and dotted, yet a benign diagnosis is rendered by cytotechnologists. In addition, malignant signet-ring cells are not commonly present in serous effusions. Cleveland Clinic receives about 1000 serous effusion cytology samples every year, with a malignancy rate of 12.5%. In all of 2013, there were only six cases signed out as “adenocarcinoma with signet-ring cell features.” Of the six cases, five were high-grade adenocarcinomas with obviously malignant cells and diagnosis was not difficult. However, the remaining case was very challenging. The patient had a history of pancreatic adenocarcinoma with peritoneal carcinomatosis. The concurrent mesenteric biopsy showed infiltrating tumor cells forming glands. The tumor cells showed similar low nuclear-to-cytoplasmic ratio with abundant cytoplasmic mucin (peritoneal fluid) [Figure 3 and 4]. Without knowledge of the concurrent peritoneal biopsy, this case may have been signed out as atypical cells. These cells may have been considered to be reactive mesothelial cells/histiocytes.

Signet-ring cell adenocarcinoma, lobular breast adenocarcinoma, small cell carcinoma, and mesothelioma represent the most subtle malignancies in serous effusion. Malignant serous effusions predict poor outcomes and poor survival rates in patients. Hence, the distinction of benign and malignant signet-ring cells is crucial in the diagnosis of such cases. As mentioned earlier, signet-ring cells have been identified in a number of malignancies, including gastric, breast, ovarian, lymphoma, gastrointestinal stromal tumor, prostatic adenocarcinomas, pancreatic, salivary gland adenocarcinomas, lung adenocarcinoma, and other malignancies.^[1-13] However, signet-ring cell change can occur as a degenerative change in histiocytes and mesothelial cells, common components of serous effusions. These represent benign mimickers of signet-ring cell adenocarcinoma and represent a diagnostic challenge to cytopathologists.

Signet-ring-like cells have been reported by many authors in benign and reactive conditions in the gastrointestinal tract. Artifactual signet-ring-like cells have been identified in gastric lymphomas and can be of either lymphoma cell or epithelial origin.^[17,20] Benign signet-ring cell aggregates have also been identified in Peutz-Jeghers polyps, in benign colonic and gallbladder mucosa, and in cases of pseudomembranous colitis.^[15,21,22] As a result of degeneration, histiocytes may undergo signet-ring cell change and may be mistaken for signet-ring carcinoma. Guerrero-Medrano et al. described signet-ring sinus histiocytosis as a reactive disorder incidentally identified in histologic sections of pelvic and axillary lymph nodes of patients with carcinoma of the prostate and breast, respectively. The histiocytic signet-ring cell change mimics metastatic adenocarcinoma, and the distinction could be resolved by special stains for mucin and immunohistochemistry.^[14] Wu et al. described signet-ring cells produced as a result of procedural trauma in a duodenal stricture biopsy from elderly male patient. The signet-ring cells represented degenerated epithelial cells detached from normal glands as a result of trauma.^[18]

Signet-ring-like cells are a frequent finding in cytology specimens, particularly serous effusions. Wu et al. studied the cytopathologic significance of signet-ring cells in a wide range of cytology specimens including fine-needle aspiration specimens, washings, and fluids. There were 83 cases of signet-ring cell diagnosis over a 16-year period, and 13 of them (16%) were benign. Reactive mesothelial cells were the most common benign finding (10 of 13 cases; 77%). The other benign cases were due to reactive epithelium (two cases) and chronic bronchitis with goblet cells (one case). The specimens in which benign signet-ring cells were identified included paracentesis, brushing, effusions, and washes. However, the clinical presentation of the benign cases is not known, and the cytologic preparation used to diagnose those benign cases was not given.

The interpretation of benign signet-ring cells and their distinction from malignant signet-ring cells is crucial to avoid false-positive diagnosis of signet-ring carcinoma. Morphology can be useful in distinguishing benign from malignant SRCs. However, correlation with the ThinPrep specimen and knowledge of the clinical setting should be performed before reporting of SRCs on Wright-stained slides. According to this study, those cells are lost on ThinPrep preparation and are most commonly identified in the setting of liver cirrhosis. Knowledge of the morphologic difference between benign and malignant SRCs, as well as correlation with the ThinPrep specimen, is crucial to avoid false positive carcinoma diagnosis, with a low threshold for interlaboratory comparison of different specimen preparations.

CONCLUSION

Mesothelial cells and histiocytes can undergo signet-ring cell change and mimic signet-ring adenocarcinoma cells, especially on cytospin prepared, Wright-stained slides. Their presence on Wright-stained slides may represent a diagnostic pitfall. This study has shown that these benign mimickers are more frequently seen in patients with liver cirrhosis and ESRD, probably as a result of the chronic inflammatory state. They typically are not present in ThinPrep prepared and Papanicolaou-stained slides. Careful integration of all available clinical data and pathologic material related to a patient procedure is necessary in this critical diagnostic situation.