Diagnostic Pitfall of Gastric Signet-Ring Cells: How to Diagnose a Benign Signet-Ring Cell From a Malignant One

Patient 1

A 61-year-old man, with a long history of Helicobacter pylori infection, presented with new-onset permanent acid reflux (Table 1). He was started on proton pump inhibitor (PPI) therapy. Endoscopy revealed a normal mucosal pattern, with no visible lesion. Random gastric biopsies were performed and showed large cells with a vacuole occupying most of the cytoplasm and pushing the nucleus to the periphery, scattered in the superficial lamina propria (Figure 1A). Normocrine crypts, well-differentiated glandular structures with discrete chronic inflammation were also observed. Positive keratin AE1/AE3 staining confirmed the presence of signet-ring epithelial cells (Figure 1B). Other staining showed wild-type p53 expression, low Ki67 index, strong membranous staining for E-cadherin (Figure 1C-E) and undisrupted reticulin staining (Figure 1F) pleading for the absence of signet-ring cell adenocarcinoma infiltrating into the lamina propria. Helicobacter pylori was not detected by immunohistochemistry. Further investigations, including contrast-enhanced thoracoabdominal tomography, endoscopic ultrasound, and a PET-FDG, showed no evidence of local or widespread gastric malignancy.

OPEN IN VIEWER

Figure 1.

Superficial aggregates of signet-ring cell change in the gastric mucosa at high magnification (A). Signet-ring cell change express keratin AE1/AE3 on immunohistochemistry (B), show a strong membranous expression of E-cadherin (C), a wild-type p53 expression (D) and a low Ki67 labeling index (E). Reticulin is intact around the signet-ring cell change (F).

OPEN IN VIEWER

Table 1.

A Summary Table Listing Clinical Symptoms, Endoscopic Findings, Histologic Findings (Inflammation Setting, Immunohistochemistry, Special Stain), and Follow-Up Treatment.

Patient	Age	Symptoms	Endoscopic Findings	Inflammation Setting	Immunohistochemistry	Special Stain	Treatment
1	61	Helicobacter pylori infection Permanent acid reflux.	Normal mucosal pattern, with no visible lesion	Discrete chronic inflammation	Wild-type p53 Low KI67 Strong positivity of Ecadherin	Undisrupted reticulin	Follow-up with biopsies PPI therapy
2	73	New-onset microcytic anemia Eight years prior : gastric biopsies with the same lesion	Nodular aspect of the mucosae	Regenerative gastric mucosae	Wild-type p53 Low KI67 Moderate positivity of Ecadherin	Undisrupted reticulin	Mucosectomy Follow-up with negative biopsies
3	56	Dyspepsia abdominal pain	Normal with no visible lesions	No inflammation No intestinal metaplasia	Wild-type p53 Low KI67 Moderate positivity of Ecadherin	Undisrupted reticulin	4 years follow-up
4	61	Gastrointestinal bleeding	Nodular aspect of the mucosae	Regenerative gastric mucosa without ulceration	Wild-type p53 Low KI67 Strong positivity of Ecadherin	Undisrupted reticulin	Mucosectomy Follow-up with negative biopsies

Abbreviation: PPI, proton pump inhibitor.

Repeat endoscopy was performed to look for signet-ring cell adenocarcinoma. New gastric biopsies were taken which showed persistence of signet-ring cells within the crypts and stained for keratin AE1 to AE3 without infiltration of the lamina propria. Three months later, the control showed persisting mild inflammation and intestinal metaplasia without any signet-ring cells. The diagnosis of benign signet-ring cell change was finally accepted. Ten months after the initial diagnosis, the patient remained healthy with stable symptoms on PPI therapy. Close follow-ups remain scheduled.

Patient 2

A 73-year-old man underwent upper GI endoscopy for the evaluation of new-onset microcytic anemia. Gastric antral biopsies showed the presence of superficial signet ring cell aggregates in the setting of a regenerative gastric mucosae. This led to the initial diagnosis of signet ring cell adenocarcinoma. Previous biopsies taken 8 years prior to presentation were retrieved and showed similar signet ring cells in all histological sections in the setting of mucosal exulceration. However, at that time, the patient refused surgery and decided to return to Turquey. He returned to the gastroenterologist after 8 years with abdominal pain and no follow-up. New biopsies were performed. The same signet ring cells were observed in the new biopsies as described above. Endoscopic mucosectomy was chosen as a staging procedure and en bloc resection was performed. Histology confirmed the presence of focal signet ring cells in the gastric mucosa, positive for keratin AE1/AE3. Strong membranous staining for E-cadherin, a wild-type p53, low Ki67, and reticulin staining confirmed an intact mucosal architecture (Figure 2A-F). The diagnosis of signet ring cell change was considered the most likely diagnosis given the lack of disease progression over the 8 years. Follow-up was recommended, but unfortunately the patient died of a myocardial infarction one year later.

OPEN IN VIEWER

Figure 2.

Superficial aggregates of signet-ring cell change in the gastric mucosa at high magnification (A). Signet-ring cell change express keratin AE1/AE3 on immunohistochemistry (B), show a strong membranous expression of E-cadherin (C), a wild-type p53 expression (D) and a low Ki67 labeling index (E). Reticulin is intact around the signet-ring cell change (F).

Patient 3

A 56-year-old man underwent gastric endoscopy for dyspepsia and abdominal pain. Endoscopic examination was normal with no visible lesions. On histological evaluation, antral biopsies showed aggregates of vacuolated and atypical cells in the mucosa positive for keratin AE1/AE3, with low Ki67, a wild-type p53 and no loss of E-cadherin (Figure 3A-E). Reticulin staining highlighted the absence of alterations in the tissue architecture (Figure 3F). There was no evidence of inflammation or intestinal metaplasia or Helicobacter pylori infection. There was no evidence of ischemia. This presentation raised the differential diagnosis between signet-ring cell change and signet-ring cell adenocarcinoma. Close endoscopic follow-up was recommended and control biopsies performed 2 months later were completely normal. No atypical vacuolated cells were found. To date (4 years later), no further biopsies have been obtained for this patient.

OPEN IN VIEWER

Figure 3.

Superficial aggregates of signet-ring cell change in the gastric mucosa at high magnification (A). Signet-ring cell change express keratin AE1/AE3 on immunohistochemistry (B), show a strong membranous expression of E-cadherin (C), a wild-type p53 expression (D) and a low Ki67 labeling index (E). Reticulin is intact around the signet-ring cell change (F).

Patient 4

A 61-year-old man underwent gastric biopsies for GI bleeding. Histological examination revealed regenerative gastric mucosa without ulceration and superficial aggregates of vacuolated cells with discrete atypia. This prompted a differential diagnosis between signet-ring cell change and signet-ring cell adenocarcinoma and endoscopic submucosal dissection (ESD) was recommended. The surgical specimen was obtained after 2 months and still containing the same superficial aggregates of signet-ring cells described in the previous biopsies, positive for keratin AE1/AE3 (Figure 4A and B). Immunohistochemistry results showed low Ki67, a wild-type p53, and negative staining for CD68 (macrophage marker) and endocrine markers. E-cadherin was strongly positive and the reticulin staining showed a preserved tissue architecture (Figure 4C-F). Discrete chronic inflammation has also been seen. Although both signet-ring cell change and signet-ring cell adenocarcinoma were considered, signet-ring cell change was the preferred diagnosis given the reassuring immunostaining of E-cadherin, Ki67, p53, and reticulin). Endoscopy and control biopsies at 4 months were reassuring and compatible with scar tissue. Endoscopic surveillance at 6 months and then annually was recommended. No repeat biopsies have been performed to date.

OPEN IN VIEWER

Figure 4.

Superficial aggregates of signet-ring cell change in the gastric mucosa at high magnification (A). Signet-ring cell change express keratin AE1/AE3 on immunohistochemistry (B), show a strong membranous expression of E-cadherin (C), a wild-type p53 expression (D) and a low Ki67 labeling index (E). Reticulin is intact around the signet-ring cell change (F).

Physiopathology

Although the exact pathogenesis of signet-ring cell change is unknown, several authors have proposed a similar hypothesis. Gastric epithelial cells undergo signet-ring cell change, a benign degenerative differentiation process in response to mucosal injury.^9–11 Boncher et al theorized that “benign epithelial cells subjected to damage detach from the basement membrane and adjacent cells. In doing so, their columnar shape is lost, the cells round up and their mucin droplet pushes the nucleus to the periphery resulting in hallmark appearance of the traditional signet-ring cell.” ⁹ However, the development of these cells remains unexplained and difficult to study. In the first and third patients, as well as in the follow-up biopsies of Mitsuhashi et al, the disappearance of the signet-ring cells can be attributed to irreversible degenerative changes that eventually lead to apoptosis. In the case of (partial) mucosal healing, the absence of new degenerated cells would be expected.

Diagnostic Tools

The incidental discovery of signet-ring cells in a biopsy specimen should lead clinicians to consider signet-ring cell adenocarcinoma first. Clinicians rely on endoscopic examination to rule out tissue invasion and biopsy specimens without malignant features. Genetic testing is recommended in an appropriate clinical setting. The diagnosis of signet-ring cell change is suggested on the basis of cell morphology, their superficial presence in the lamina propria and immunohistochemistry results. The diagnosis is subsequently confirmed with continued negative findings on follow-up endoscopic examinations. The persistence of signet-ring cell change in the stomach for years without a pejorative evolution may occur, as described in patient 2 which is, to our knowledge, the first to illustrate the natural history of signet-ring cell change and to show that signet-ring cell change can persist for a very long time.

Narrow-band imaging or chromoendoscopy in combination with magnifying endoscopy can be used to analyze the gastric mucosa in detail. ¹² Endoscopic ultrasound plays an important role in ruling out tissue invasion. However, no endoscopic tool is accurate enough to properly assess the depth of lesions within a 500-μm threshold, and none can distinguish signet-ring cell change from very early gastric cancer.^12,13

The cornerstone of signet-ring cell change diagnosis remains accurate histological analysis. Microscopically, signet-ring cells in both signet-ring cell change and signet-ring cell adenocarcinoma show strong cytoplasmic positivity for mucins (periodic acid–Schiff staining), reflecting the normal cellular mucin content of the cell.^5,6,8,9,11 Signet-ring cells in signet-ring cell change are mostly located superficially in the lamina propria and lack malignant cytomorphological features of cellular and nuclear atypia. Especially in small biopsy specimens, morphological differentiation from signet-ring cell adenocarcinoma based on histological and immunohistochemical profiles seems to allow better discrimination.^6,8,9,11

Cytokeratin and reticulin staining are useful in highlighting the tissular architecture and showing the confinement of signet-ring cells in signet-ring cell change within the basement membrane, thus excluding cellular invasion into the lamina propria. Tissue morphology is related to the underlying disease. ⁶

Signet-ring cells in signet-ring cell change show strong or moderate membranous staining for E-cadherin, a wild-type p53 staining, and generally have a low Ki67 labeling index.^5,6,8,9,11 Only Boncher et al ⁹ identified a focally high Ki67 index (up to 20%) in less than 2% of cells in their histological sample, with no other malignant atypia.

In contrast, signet-ring cell adenocarcinoma exhibits the hallmarks of malignant cellular atypia, an infiltrative pattern and changes in the adjacent gastric mucosa. Many signet-ring cell adenocarcinoma show reduced or absent E-cadherin staining, some show increased nuclear p53 positivity or complete loss of expression, and usually have a high proliferation index as reported in various studies.^8,9

Several other mimics of signet-ring cell adenocarcinoma deserve brief discussion. Firstly, macrophages are often present in association with gastric injury. They are usually easily distinguished from diffuse-type gastric carcinoma but can be a diagnostic challenge when they contain abundant cytoplasmic mucin (foamy macrophages). Immunohistochemistry allows identification of macrophages by negative keratin staining and strong diffuse staining for CD68. Secondly, cytoplasmic vacuolization of oxyntic glands may occur in patients receiving PPIs. ¹⁴ This finding may simulate the appearance of signet-ring cell adenocarcinoma in situ. Finally, glandular clear cell degeneration has also been described in areas of ischemia. ¹⁰ These cells appear within glands with a degenerative lacy pattern that can be confused with dysplasia. They show nuclear atypia, such as hyperchromatic, enlarged, slightly irregular nuclei, and a vacuolated cytoplasm. ¹⁰

Distinguishing between very early signet-ring cell adenocarcinoma and signet-ring cell change remains a challenging task as biopsies are limited by their localized nature. In early hereditary diffuse gastric cancer (HDGC), the most common histological abnormalities are small clusters of typical signet-ring cells confined to the superficial lamina propria without infiltration below the muscularis mucosae.^15,16 However, there are 2 specific scenarios associated exclusively with HDGC that require testing for the CDH1 gene, followed by CTNNA1 if the first test is negative: when signet-ring cells with hyperchromatic and depolarized nuclei are identified within the basal membrane of a gland, effectively replacing the normal glandular cells, and when there is a pagetoid extension of a row of signet-ring cells beneath the preserved epithelium of glands and foveolae and within the basal membrane.^15,16 Therefore, genetic testing should be considered in patients where specific histological findings or historical evidence strongly suggest the presence of HDGC. ¹⁷

Although imperfect, repeat biopsies are essential for a correct diagnosis. If clinical doubt remains, ESD can be performed to rule out very early signet-ring cell adenocarcinoma, with a focus on achieving an R0 resection margin. ¹⁸

Recently, attention has been focused on the molecular profiles of gastric cancers and clinically relevant biomarkers for early disease detection.^12,19–21 Although promising in the field of gastric oncology, emerging conclusions seem premature and have not yet been validated.