, Christopher VandenBussche2 and Syed A. Hoda3
CBL Path, Rye Brook, NY, USA
Department of Pathology, Johns Hopkins University Department of Pathology, Baltimore, MD, USA
New York Presbyterian Hospital Weill Cornell Medical College, New York, NY, USA
The gastrointestinal (GI) tract is comprised of an upper and lower tract and associated accessory organs. Cytology of the GI tract is performed in order to detect neoplastic and non-neoplastic lesions and often includes concurrent procurement of needle care biopsy (NCB). Sites amenable to cytological sampling include the esophagus, stomach, duodenum, pancreatobiliary and liver from the upper GI tract, rectum and anus from the lower GI tract, and regional lymph nodes. See Chap. 4, “Gastrointestinal Tract Exfoliative Cytology,” for introductory details, PSC reporting guidelines, and ancillary tests for GI specimens.
Indication, Collection, and Laboratory Processing of Cytological Samples
The examination of GI cytology is an efficient and cost-effective method for diagnosing GI tract lesions. It allows for rapid interpretation of FNA specimens using rapid on-site evaluation (ROSE); this also allows for proper specimen triage for various ancillary tests. An accurate diagnosis relies on procuring an adequate sample, optimal specimen preparation in order to enhance cytomorphology, and expertise in the interpretation of these challenging entities.
Endoscopic Ultrasound-Guided FNA (EUS-FNA)
Endoscopic ultrasound-guided fine needle aspiration (EUS-FNA) is a well-established, reliable, accurate, and safe diagnostic technique for assessing lesions of the GI tract and adjacent organs. EUS-FNA can also be used to sample mediastinal, splenic, and adrenal gland lesions. The sensitivity and specificity of EUS-FNA are at least 80 and 100 %, respectively.
EUS-FNAs are usually performed for masses in the esophagus, stomach, esophageal/gastric submucosal lesions, pancreatobiliary tract, and regional lymph nodes including periesophageal/gastric, peri-pancreatobiliary, and abdominal lymph nodes. It is the procedure of choice for establishing a pancreatic malignancy and for sampling pancreatic cysts.
EUS-FNA of pancreatic masses is performed with linear endosonographic instruments. The aspiration needles are usually 22- or 25-gauge and are selected based on the vascularity of the lesion or viscosity of cyst fluid. EUS-FNA may be performed in conjunction with NCB (Tru-Cut 22-gauge Shark Core needles) for mucosal, submucosal, fibrotic, and stromal-rich lesions. ROSE of cytology is usually performed for FNAs of both solid masses and cystic lesions, as it reduces rate of false-negative FNAs. After the cyst fluid is analyzed cytologically, cyst fluid is triaged for biochemical testing [carcinoembryonic antigen (CEA) and amylase] and/or molecular testing for mutation analysis. Additional passes are performed if a residual solid component (mural nodule) is identified.
EUS is better than CT or MRI in determining vascular invasion and detecting small tumors and lymph node metastases.
Laboratory Processing of GI Specimens
For FNA, the specimen is processed as conventional smears, LBP, and/or a cell block. An accurate cytological diagnosis requires an adequate and well-prepared sample, as well as expertise in the interpretation of these preparations. See Chap. 1 for details of LBP preparations.
Advantages of FNA Specimens for GI Tract over NCB
Ability to assess solid pancreatic lesions; subepithelial, submucosal, and mural nodules; and regional lymph nodes via EUS-FNA, assessment of cystic lesions, collection of cyst fluid for chemical and molecular analyses, and shorter turnaround time.
EUS-FNA of Submucosal Gastric and Esophageal Lesions
EUS-FNA is an effective method for tissue diagnosis of gastrointestinal submucosal tumors (SMTs). The standard endoscopic NCB is usually nondiagnostic because the overlying mucosa is usually thick and prevents procurement of tissue from the lesion. The SMTs include gastrointestinal stromal tumors (GISTs). The diagnosis of GIST is based on cytomorphology and immunohistochemical analysis. The latter is crucial for differentiating these lesions from benign smooth muscle lesions such as a leiomyoma or schwannoma. Therefore, it is necessary to obtain adequate tissue for the immunohistochemical analysis in the case of GI SMTs. Some surprises may be encountered such as esophageal duplication cyst.
EUS-FNA of Regional Lymph Nodes
EUS-FNA is an important modality for sampling regional deep-seated lymph nodes, including mediastinal and abdominal, for diagnosis of infections, metastatic malignancy, or primary hematopoietic tumors. Adequate specimen should be procured for ancillary tests including immunohistochemistry, flow cytometry, and molecular analyses, such as next-generation sequencing (when pertinent), to increase diagnostic yield. Lymphoid lesions may be successfully subclassified on EUS-FNA, particularly, if the lymphoma is recurrent. Cytopathologists should work in collaboration with hematopathologists for work-up of lymphoid lesions.
The differential diagnosis of solid and cystic pancreatic lesions include:
Solid pseudopapillary tumor
Acinar cell carcinoma
Intraductal papillary mucinous neoplasms (IPMNs)
Mucinous cystic neoplasm (MCN)
Solid Pancreatic Lesions
Any of the above-noted solid pancreatic lesions may be encountered on EUS-FNA. The clinician performing the procedure may have a suspicion for a particular entity based on the clinical history and presentation, preceding CT and MRI findings and appearance on EUS. Fortunately, each entity has specific cytological features and distinct immunoprofile and occasionally a distinct location. Common diagnostic pitfalls include chronic pancreatitis that may mimic well-differentiated pancreatic adenocarcinoma and vice versa and benign acinar cells that may mimic neuroendocrine (NE) tumor and vice versa.
Premium Wordpress Themes by UFO Themes
Cystic pancreatic lesions are detected by imaging techniques and are sampled by EUS-FNA, when clinically and radiologically indicated. Clinically, it is important to determine if the pancreatic cyst is mucinous or non-mucinous and benign or malignant. Inflammatory pseudocysts can be managed medically in most patients, and serous cysts which are benign neoplasms are not surgically resected unless the patient is symptomatic or the cyst is large (>4 cm). Neoplastic mucinous cyst can be IPMN or MCN. Although both are mucinous epithelium-lined cystic tumors, they differ in age, gender, location in the pancreas, relationship to pancreatic ducts, and management. All main pancreatic duct IPMNs are resected. Management guidelines for suspected branch duct-type IMPNs are based on clinical and imaging findings. Small cysts (<3 cm) without high-risk clinical or imaging findings are conservatively managed. Patients with high-risk imaging features or suspicious/malignant cytology are surgically managed. Patients with worrisome imaging features will be evaluated by EUS-FNA for cytological assessment. All MCNs are resected for assessment of high-grade dysplasia. Aspiration of pancreatic cysts is also valuable for obtaining material for cytological, biochemical, and molecular analyses.
Diffuse signet ring-type gastric adenocarcinoma. (a) EUS-FNA yielded this material from a gastric lesion. This pattern is classic for signet ring gastric cancer, in which some cells take on the appearance of a signet ring (see the cell at 2 o’clock) – the “hole” of the ring is a mucin vacuole, which appears mostly white and the “ring” is the thin rim of cytoplasm. The malignant nucleus is large, is eccentrically placed, and appears to bulge out of the cell, completing the picture of a signet ring. These carcinomas often diffusely involve the stomach as single cells, and thus FNA often produces numerous discohesive single cells rather than tissue fragments. Note the nuclear size differences, the coarse chromatin, and the irregular nuclear borders seen in these cells. These cells may resemble histiocytes, and a high degree of suspicion may be required for their detection. In this direct smear (DS), inflammatory cells partially obscure cell detail (Pap-stained DS). (b) Isolated signet ring cells with the crescentic-shaped nuclei compressed toward the periphery by large intracytoplasmic mucinous vacuole, characteristic of the tumor. Mucinous cytoplasmic condensation is evident in the cell on the left. Background is clean (TP). If cell block material is present, keratin markers or special stain for mucicarmine can differentiate the cells from histiocytes, which express CD68 or KP-1 
EUS-FNA of gastrointestinal stromal tumor (GIST). (a) This aspirate shows numerous uniform spindle cells with monotonous, slender, well-aliened nuclei with pale chromatin, no nucleoli, and delicate wispy basophilic cytoplasm. The cytoplasmic borders are indistinct, giving a syncytial appearance. The fragment is too cellular to be a fragment of normal connective tissue, which is sometimes seen when a lesion is not properly sampled during the procedure. Note the myxoid stromal quality. Mitoses were not evident (TP). (b) GIST on a conventional smear shows similar nuclear morphology. The myxoid stroma is more diffuse (Pap-stained DS). (c) Same case of GIST as above processed as SP specimen. Main differences are that the stroma appears denser and cells are more compact than TP (SP). (d) Positive immunostain for c-kit (a cell-surface transmembrane tyrosine kinase also known as CD117). Immunostain for DOG1 was also positive (CB). (e) The resection shows fascicles of uniform, bland spindle cells with pale, eosinophilic cytoplasm and myxoid stroma (H&E). Radiologically, this lesion was submucosal, distal, well marginated, heterogeneous, and enhancing. GISTs are the most common mesenchymal neoplasms of the gastrointestinal tract and should be differentiated from other mesenchymal tumors. They harbor specific activating mutations in the KIT or platelet-derived growth factor receptor-α (PDGFRA) which makes them responsive to pharmacologic inhibitors, such as imatinib mesylate, and sunitinib malate. Overall, it is difficult to completely distinguish GIST from other spindle-shaped neoplasms in this region, and immunohistochemical studies are often critical for a definitive diagnosis. Approximately 90 % of GISTs are positive for c-kit and 95 % are positive for DOG-1 . S-100 positivity is suggestive of a nerve sheath tumor, such as schwannoma, whereas smooth muscle actin (SMA) reactivity is seen in smooth muscle neoplasms, such as leiomyoma. Clinical suspicion for GIST may arise from imaging studies showing a submucosal location of the tumor. (f and g) Gastric duplication cyst; a second lesion was identified as a well-defined, 3-cm hypoechoic lesion with a heterogeneous internal structure, along the greater curvature of the stomach/peripancreatic by a previously performed computed tomography (CT) scan. It was suspected to be a lymph node suspicious for non-Hodgkin lymphoma. The cystic nature of the lesion was not diagnosed due to the solid-tissue appearance of the lesion, presumably, because of the mucinous nature of the fluid (DQ). The images demonstrate thick mucin with embedded ciliated columnar epithelial cells. Mucin is more diffuse and stains magenta on DQ and clumped and pale basophilic on TP (f, DQ-stained DS; g, TP). Histologically, the cysts were lined with pseudostratified ciliated columnar as seen on subsequent resection
Benign gastric and duodenal mucosa. (a) Benign gastric foveolar cells showing a honeycomb arrangement with an apical mucin cup (TP). (b) Benign gastric mucosa with branching mucus glands with vacuolated cytoplasm (H&E). (c) Benign duodenal mucosa appears as flat, two-dimensional sheet with an organized honeycomb configuration. Cells are polarized and uniform with low nucleus-to-cytoplasmic ratios. The nuclei are monotonous with bland chromatin. Most importantly, duodenal mucosa shows goblet cells, which appear as evenly distributed “eosinophilic holes” in the fragments which are not seen in pancreatic mucinous neoplasms (TP). EUS-FNA of the pancreas must transverse the normal GI tract mucosa and may be seen as contaminants in pancreatic FNA specimens. EUS-FNA for pancreatic head and body/tail lesions are performed via the duodenum and the stomach, respectively. Because the GI contaminants are also mucin-secreting epithelium, one must carefully distinguish them from cystic mucinous tumors
WordPress theme by UFO themes
%d bloggers like this: