PERSPECTIVES IN CLINICAL GASTROENTEROLOGY AND HEPATOLOGY

Histopathologic Features and Diagnostic Criteria

Histologically, fundic gland polyps consist of one or more dilated oxyntic glands, lined by flattened parietal and mucous cells (Figure 1). Fundic gland polyps are among the most characteristic lesions of the stomach: the recognition of the dilated oxyntic glands with flattened parietal and mucous cells in slides stained with H&E is immediate and unequivocal (Figure 1B and C). One caveat is that when the surface of a polyp is eroded the regenerative appearance may be misinterpreted as dysplasia. True dysplasia, particularly high grade, is exceedingly rare and is virtually limited to fundic gland polyps found in patients with polyposis syndromes. No special stains or molecular studies are warranted.

Clinical Approach

The finding of multiple characteristic polyps in the oxyntic portion of the stomach in a patient taking PPIs is essentially diagnostic of fundic gland polyps. Generally, when first encountered, one or more representative polyps should undergo a biopsy examination to confirm the diagnosis. Large polyps (>1 cm in diameter) should be removed entirely to confirm the diagnosis because fundic gland polyps rarely exceed this size.⁶ A biopsy specimen from the polyp in these cases is not adequate because if the polyp is not of the fundic gland type, biopsy sampling may not include crucial areas of possible dysplasia or neoplasia. In addition, a thorough visual inspection of the remaining polyps should be made; any lesion that appears significantly different from the others should be undergo a biopsy examination, or, if possible, be removed. Specifically, size larger than 1 cm, ulceration, and unusual location such as the antrum should prompt a more aggressive approach. When fundic gland polyps are found in a young patient, especially if numerous (eg, ≥20), the possibility of a polyposis syndrome should be considered. Individuals with familial polyposis syndromes typically are younger than the average patient with fundic gland polyps (mean age, 40 y),^7–9 and occasionally have polyps in the antrum.¹⁰ When gastric polyps are associated with duodenal adenomas a familial polyposis syndrome strongly should be considered and colonoscopy should be recommended.

Fundic gland polyps rarely are found in stomachs affected by H pylori infection and, therefore, in the absence of a familial polyposis syndrome, concerns about gastric cancer are moot.¹¹ Nonetheless, when polyps are innumerable or large (>1 cm) there may be cause for concern regarding eventual outcome. Although no guidelines exist, we suggest that when either more than 20 polyps are present or their size is larger than 1 cm one should consider reducing or preferably stopping the medication to assess whether this will result in regression of the polyps.¹² If regression occurs, it is unknown whether PPIs can be reinstituted. Practically, if surgical therapy is not an option, one might consider a different PPI and at the minimally effective dose. Although there does not seem to be a correlation between serum gastrin levels and the presence of fundic gland polyps, it may be worthwhile to measure gastrin levels in these patients.¹³ A high level (>400 pg/mL) suggests profound acid suppression. If the patient is not an intrinsic hypersecretor and does not have a gastrinoma or Zollinger–Ellison syndrome, PPIs should be withdrawn, and, if gastroesophageal reflux disease symptoms persist, replaced with an H₂-receptor antagonist. Recent reports of gastric carcinoids associated with profound PPI-induced acid suppression^{14, 15} and the increasing awareness of other potential adverse effects (eg, interference with the absorption of a variety of other medications, possible hip fracture, and Clostridium difficile infection) have provided further impetus to using PPIs less often and at the minimal effective dose.^15–22

Histopathologic Features and Diagnostic Criteria

The typical features of hyperplastic polyps include elongated, grossly distorted, branching, and dilated hyperplastic foveolae lying in an edematous stroma rich in vasculature, and small haphazardly distributed smooth muscle bundles with varying degrees of chronic and active inflammation (Figure 2). Between 1% and 20% of hyperplastic polyps have been reported to harbor foci of dysplasia (Figure 3). This wide range is more likely a reflection of the different criteria used in the assessment of dysplasia than because of real geographic or biological variability. Mutations of the p53 gene, chromosomal aberrations, and microsatellite instability all have been detected in these polyps.^30–32 The overall prevalence of carcinoma in hyperplastic polyps is less than 2%, and it is more frequent in polyps larger than 2 cm.^33–35

When foci of dysplasia or carcinoma are diagnosed within a hyperplastic polyp, clinicians often inquire whether the lesion was in fact an adenoma or a carcinoma erroneously identified as a hyperplastic polyp. This is almost never the case. Adenomas (as detailed later) consist entirely of dysplastic epithelium; in contrast, hyperplastic polyps with dysplastic foci are inflammatory lesions formed by long distorted branching, and dilated hyperplastic foveolae lying in a vascular edematous stroma. Finding foci of dysplastic (ie, neoplastic) epithelium within these lesions does not call for a reconsideration of their original pathogenesis or classification, although it does change their management.

The histopathologic diagnosis rests on the detection of the earlier-mentioned features in traditional H&E-stained slides. Because erosions, ulcerations, and inflammation, even extensive, are expected features of these polyps, pathologists are encouraged to avoid adding the word “inflammatory” to the diagnostic line. This invariably results in calls from puzzled clinicians who want to know whether hyperplastic and inflammatory gastric polyps are different or, as one put it, one and the same. They are one and the same.

When inflammatory polyps measure less than 1 cm, the routine examination of representative sections from each polyp is considered adequate. Larger polyps (>1 cm) may harbor dysplasia and even carcinoma. Therefore, they should be sectioned in 2- to 3-mm slices, and sections should be prepared from each slice. This practice will allow the detection of otherwise easily missed dysplastic or neoplastic lesions (Figure 3). Outside the research arena, no special stains or molecular studies are necessary. If no specimens from other parts of the stomach are included, an immunohistochemical stain for H pylori can be helpful.

Clinical Approach

In view of the potential cancer risk, all hyperplastic polyps larger than 1 cm should be excised completely. If dysplasia or intramucosal carcinoma is found, but the stalk is not affected, the lesion can be considered completely removed and most likely cured.³⁶ The excision of the polypoid lesion always should be accompanied by additional sampling of the unaffected mucosa to obtain reliable information about the topography and severity of the background gastritis and atrophy.

When hyperplastic polyps arise in a background of chronic atrophic gastritis (a precursor lesion for gastric adenocarcinoma) the severity and extent of the atrophic gastritis should be evaluated. Risk stratification for gastric cancer can be assessed histologically using the Operative Link for Gastritis Assessment (OLGA) or the Operative Link on Gastritis/Intestinal Metaplasia Assessment staging systems.^{37, 38} Both require histologic grading of adequate samples from the antrum and corpus; grades then are combined to provide a risk stratification category. One can use the 5-biopsy specimen protocol recommended by the updated Sydney system.³⁹ We prefer an extended 7-biopsy protocol consisting of 3 specimens from the antrum (which can be submitted in a single formalin container), 2 from the lesser curvature of the corpus, and 2 from the greater curvature of the corpus, with each set in a separate container.⁴⁰ The topographic separation of specimens from the antrum and corpus is crucial because pseudopyloric metaplasia (a feature of corpus atrophy) mimics the antral mucosa histologically. Although it has not been established whether the additional information obtained by this extended sampling protocol substantially improves risk assessment, we prefer it because it allows a more detailed evaluation of the extent and distribution of atrophy.

If present, H pylori should be eradicated and an endoscopic follow-up evaluation should be scheduled between 3 and 6 months after therapy to confirm successful eradication. Alternatively, a noninvasive test such as the urea breath test may be used.⁴¹ In many instances any remaining small hyperplastic polyps will have regressed or disappeared.^{42, 43} Patients with OLGA stages III and IV (moderate diffuse atrophy or severe atrophy in either the corpus or antrum, usually accompanied by extensive intestinal metaplasia) should be considered for longterm endoscopic surveillance. The best intervals for such surveillance are unclear (eg, annual, bi-annual, or some other interval) as is the number of years for which it should be continued. No evidence-based guidelines exist; therefore, local recommendations should be followed when available.

Clinical Approach

Gastric adenomas frequently arise in a background of chronic atrophic gastritis. Because this is a precursor lesion for gastric adenocarcinoma, in addition to completely excising all adenomas, the severity and extent of the atrophic gastritis should be evaluated. The same biopsy protocols^{39, 40} and the use of the OLGA or Operative Link on Gastritis/Intestinal Metaplasia Assessment system suggested in the section “Hyperplastic Polyps” should be followed.^{37, 38} It must be emphasized that adenomas are neoplastic lesions (ie, past the stage of preneoplastic) and, therefore, all patients with a diagnosis of gastric adenoma need to be placed in a surveillance program irrespective of their atrophy stage. Eradication of H pylori followed by confirmation of the cure by biopsy examination or urea breath test is necessary in these patients.

Histopathologic Diagnostic Tips: Special Stains, Immunohistochemistry, and Molecular Studies

Gastric adenomas are neoplastic lesions with malignant potential. Therefore, multiple sections from each lesion must be examined to exclude invasion. Outside the research arena, neither special stains nor molecular studies are necessary.

Clinical Approach

The vast majority of GISTs that measure less than 1 cm are asymptomatic and are detected incidentally during an endoscopic examination scheduled for other indications. As GISTs grow, they may cause erosions or ulcerations in the overlaying mucosa or compress adjacent structures; consequently, the 2 more common manifestations are bleeding (occult or overt) and pain.

All GISTs must be considered as having malignant potential: up to 50% of patients with larger GISTs (>2 cm) have metastatic disease at presentation, usually to the liver.⁵² In practice, there is good correlation between size, mitotic activity, and clinical behavior of GISTs. Surgical resection is recommended for lesions greater than 2 cm; endoscopic enucleation followed by surveillance is an option for smaller GISTs. Endoscopic removal is controversial, however, because of reports of positive resection margins and tumor spillage.⁵⁵ Tyrosine kinase inhibitors are used as targeted therapy in cases of metastasis and surgically unresectable GISTs.⁵⁹ Neoadjuvant therapy with use of tyrosine kinase inhibitors after surgical resection of high-risk GISTs deters recurrence, but the optimal duration of therapy has not been determined.⁶⁰

Histopathologic Diagnostic Tips: Special Stains, Immunohistochemistry, and Molecular Studies

Immunohistochemical stains are central to the diagnosis of GISTs. The c-kit proto-oncogene mutation (most often occurring in exon 11) is the key molecular event in gastric GISTs and can be detected by an immunohistochemical stain directed against the KIT protein. The antibody used for the staining, designated as CD117, stains approximately 95% of GISTs (Figure 5C); the remaining 5% (which typically have a more epithelioid morphology) stain with antibodies to the DOG-1 or platelet-derived growth factor α.^{56, 57, 61} If none of these 3 stains is positive on a tumor with morphologic characteristics suggestive of GIST, tumors of smooth muscle (leiomyomas) or neural (neuromas, schwannomas) origin should be considered and immunostaining for actin and S-100 should be performed.

Predictors of behavior that pathologists must evaluate in a GIST include tumor size and mitotic counts. In general, it can be stated that the larger a tumor the greater the likelihood that it has metastasized. Figures most often cited indicate that 15% of GISTs smaller than 2.5 cm metastasize, in contrast to more than 80% of GISTs larger than 6 cm.⁵⁷ These data, reproduced in many subsequent studies, emphasize the malignant potential of even small GISTs. Most studies have found that higher mitotic counts are associated with decreased survival; however, the way mitotic counts traditionally are reported (per high-power field) are imprecise, not standardized, and subject to many technical and interpretational variables. Only unequivocally high mitotic counts (eg, >5 mitoses/50 high-power field) obtained under rigorously controlled conditions (details of the methods used to find and count 50 separate fields) and precise information on the exact area of a high-power field (which varies 3-fold in different types of commonly used microscopes) can be used to make a prognostic assessment.⁶¹ Therefore, when clinicians receive a report indicating a high mitotic count in a GIST, they should interpret it cautiously and question the pathologist about the methods used to reach that conclusion.

Clinical Approach

Most inflammatory fibroid polyps are asymptomatic, but larger polyps have been reported to cause abdominal pain, early satiety, anemia, and gastric outlet obstruction.⁶⁶ The endoscopic ultrasound appearance, characterized by an indistinct margin, a hypoechoic homogeneous lesion, and location within the second or third layer with an intact fourth layer, may be helpful in establishing the diagnosis.^67–69

Histopathologic Diagnostic Tips: Special Stains, Immunohistochemistry, and Molecular Studies

Because fibroepithelial polyps are extremely rare, general pathologists may spend their entire career without ever seeing (or at least recognizing) one. However, once familiar with their unique morphology they can be diagnosed instantly (Figure 6). Immunohistochemical staining for CD31 (an endothelial marker) is strongly positive, but it is almost never necessary for the diagnosis and should be used only as a teaching tool.

Clinical Approach

Type I and II tumors often can be removed endoscopically. In select patient with numerous and recurrent type I neuroendocrine tumors, antral resection could be a reasonable option. Antrectomy works by reducing the gastrin-producing cell mass in the stomach, thus removing the stimulus (ie, hypergastrinemia) for ECL cell proliferation.⁷⁹ Newer treatments such as the gastrin receptor antagonist netazepide are being investigated and could represent an alternative new medical treatment for type I gastric carcinoids.⁸⁰ Patients with sporadic carcinoids (type III) may present with anemia, epigastric pain, or signs and symptoms caused by metastases, including the rare carcinoid syndrome (characterized by cutaneous flushing, diarrhea, bronchospasm, and cardiac valvular lesions). Surgery followed with chemotherapy is the treatment of choice.

Histopathologic Diagnostic Tips: Special Stains, Immunohistochemistry, and Molecular Studies

Neuroendocrine tumors are best diagnosed with the help of an immunohistochemical stain (synaptophysin, chromogranin A, or CD56). These 3 stains are essentially equivalent in their ability to highlight neuroendocrine markers: their specific use is mostly a matter of preference for each pathologist. In addition, a Ki-67 stain should be performed to count the percentage of proliferating cells and to determine the grade of the tumor (Figure 8).

Follow-up Evaluation

There is a dearth of data on both the short- and the long-term follow-up evaluation of gastric polyps; therefore, no evidence-based guidelines exist.⁸² A surveillance endoscopy on nonfundic gland polyps within 1 year is a reasonable approach to evaluate the site for recurrence and to assess for new polyps. Follow-up evaluation after resection of polyps with high-grade dysplasia or early cancer should be individualized, but (at least for the first 2–3 years) short intervals (eg, 6 mo) would seem desirable.⁸³ Gastric carcinoids managed endoscopically (usually type 1) should be followed up with endoscopy every 1 to 2 years.