Endoscopic ultrasound-guided fine-needle aspiration cytology in diagnosing intra-abdominal lesions

INTRODUCTION

In the early 1980s, attachment of ultrasonography (USG) probe to endoscope led to better visualization of gastrointestinal (GI) and abdominal organs.^[1] In the early 1990s, Vilmann et al.^[2] used endoscopic ultrasound (EUS) fine-needle aspiration cytology (FNAC) for the pancreatic lesion. Since then EUS-guided FNAC has been adapted to acquire tissue from mucosal and submucosal lesions, as well as peri-intestinal structures including lymph nodes, pancreas, adrenal gland, gallbladder, bile duct, liver, kidney, and lung.^[3-16] Among them, the pancreas and lymph node are the most common targeted sites.^[4] As the anatomical configuration of the pancreas is complex, it is difficult to obtain cytology samples without exploratory laparotomy. USG-guided FNAC is employed to detect the pancreatic lesions by various studies.^[17] However, EUS-guided FNAC has better accuracy over other modalities like computed tomography (CT) or magnetic resonance imaging in detecting pancreatic lesion, as it can detect lesion <3 cm in size along with feasibility to take cytopathology samples.^[4,5,13]

In the present study, we are evaluating the role of EUS FNA in different intra-abdominal lesions.

MATERIAL AND METHODS

EUS of the lesion was carried out using Echotip ultra 22 and color Doppler. The puncture technique was the fanning one, guiding the needle in the various planes. The needle size was 22 gauges for larger lesions and 24 gauge for smaller ones. [Figure 1] shows the details of the instrument. Two to three passes were made for every mass lesion. FNAC was done transgastric (body and tail lesions) or transduodenal (head and uncinate process) depending on the location of the lesion. Similarly, transduodenal FNAC was done for the periportal lymph nodes [Figure 2]. Onsite cytologist evaluation was not available in these cases.

Close

Figure 1:

Echotip ultra 22 needle used for performing endoscopic ultrasound-guided fine-needle aspiration and its various components.

Export to PPT

Close

Figure 2:

(a) Periportal lymph node as seen in endoscopic ultrasonography, (b) Endoscopic ultrasound (EUS) elastography showing predominantly green color suggestive of soft lesion, and (c) EUS fine-needle aspiration being performed.

Export to PPT

The aspirated slides were air-dried for May Grunwald Giemsa stain as well as quickly fixed in 95% ethanol for Papanicolaou stain. The remaining sample or separate aspirate was taken for cell block in 10% formalin, if required, based on clinical suspicion.

RESULTS

During 19 months, EUS-guided FNA was performed in 71 patients, for different intra-abdominal lesions. There were 36 (50.7%) male and 35 (49.2%) female patients with a mean age of 50 ± 15 years, the age ranging from 19 to 80 years. [Table 1] shows the FNAC site of intra-abdominal lesions. Four patients underwent FNAC from different lesions simultaneously. Since the details history of ten cases was unknown, so they were already excluded from the study. In addition, three cases yielded non-diagnostic aspirate which was also excluded from the study.

Pancreas was the most common site of EUS-guided FNAC comprising 60% (n = 45) of cases followed by intra-abdominal lymph nodes 17.3% (n = 13) and stomach 9.3% (n = 7). In total of 38 cases of neoplastic lesion, there were 50.7% intra-abdominal lesions, out of which 68.4% (n = 26) of cases had pancreatic neoplasm. Two cases proved to be tuberculosis [Table 2]. Among the rare neoplasm, metastatic melanoma was confirmed in peripancreatic lymph node FNAC whereas schwannoma was confirmed on immunocytochemistry of cellblock prepared from esophageal FNAC. Cellblock was prepared in nine cases with neoplastic findings [Table 3].

Among pancreatic lesion, adenocarcinoma comprised major proportion of 24.4% (n = 11) followed by neuroendocrine tumor (NET) 15.6% (n = 7). Solid pseudopapillary neoplasm (SPN) (n = 2) comprised 4.4% of cases whereas there was one case where the differential diagnosis were SPN and NET as it cannot be confirmed even on immunocytochemistry. NETs, on the other hand, showed the discrete population as well as acinar arrangement of tumor cells with characteristic salt and pepper chromatin [Figure 3a and b]. Adenocarcinoma was characterized by cohesive clusters of moderately pleomorphic tumor cells with coarse chromatin, prominent nucleoli, and moderate vacuolated cytoplasm [Figure 3c and d]. SPN of pancreas showed characteristic thin fibrovascular cores with arborizing papillae giving “Chinese-lettering” configuration and monomorphic tumor cells arranged in a palisaded manner around these clusters with characteristic nuclear grooving and absence of mitosis [Figure 4a and b]. Many discrete cells and foamy macrophages are also noted [Figures 4c and d]. Immunocytochemistry was used for confirmation on the cellblock, where these tumors showed membranous positivity for CD 10 and aberrant nucleocytoplasmic expression of beta-catenin. However, the tumor was negative for chromogranin and synaptophysin excluding the possibility of NET. While in one case, the tumor was positive for CD10; however, beta-catenin showed cytoplasmic positivity and was negative for chromogranin and synaptophysin, creating a diagnostic dilemma between SPEN and NET.

Close

Figure 3:

(a) Fine-needle aspiration (FNA) smear shows cellular smear with clusters and singly scattered tumor cells (×120 May Grunwald Giemsa Stain), (b) Higher magnification to appreciate the cellular morphology with salt and pepper chromatin morphologically consistent with neuroendocrine tumor (×240 May Grunwald Giemsa Stain). (c) FNA smears from head of pancreas lesion show compact 3D clusters and scattered population of tumor cells (×120 May Grunwald Giemsa Stain), (d) Higher magnification shows moderate nuclear pleomorphism with irregular nuclei, coarse chromatin, and inconspicuous nucleoli, morphologically consistent with adenocarcinoma. (×420 May Grunwald Giemsa Stain).

Export to PPT

Close

Figure 4:

(a) Fine-needle aspiration smear shows branching papillary fragments with central thin delicate capillaries (×220 May Grunwald Giemsa Stain), (b) Higher magnification shows relatively uniform oval to round eccentrically placed nuclei, few showing characteristic nuclear grooves morphologically consistent with solid pseudopaillary neoplasm. (×420 Hematoxylin and Eosin), (c) Discrete cells with moderate to abundant cytoplasm (×220) May Grunwald Giemsa Stain), (d) Occasional foamy macrophages present. (×420 May Grunwald Giemsa Stain).

Export to PPT

Two cases each were reported as mucinous neoplasm and acinar cell carcinoma, respectively. One case was suspicious for malignancy. One case previously diagnosed as GI stromal tumor (GIST) turned out to be schwannoma on immunocytochemistry as the tumor showed diffuse positivity for S-100 and was negative for c-kit as well as DOG1.

One case of a peripancreatic lymph node showed features of non-Hodgkin lymphoma [Figure 5a and b]. The cases with tuberculous lymph node show necrosis and positive for acid-fast bacilli in Ziehl Neelsen staining [Figure 5c and d].

Close

Figure 5:

(a) Non-Hodgkins lymphoma in peripancreatic lymph node shows discrete population of atypical lymphoid cells (×220 May Grunwald Giemsa Stain), (b) Oil immersion view of atypical lymphoid cells with prominent nucleoli and vacuolated cytoplasm (×1200 May Grunwald Giemsa Stain), (c) Necrotic aspirate from tubercular pre-aortic lymph node (×220 May Grunwald Giemsa Stain), (b) Acid fast bacilli highlighted on Ziehl Neelsen stain (×1200) (d).

Export to PPT

Histopathology correlation was also available in few cases. Two cases of adenocarcinoma on cytology were inadequate for an opinion on biopsy. Two cases of NET on cytology were confirmed even on resected Whipple’s specimen. Three cases that were negative for malignancy on cytology confirmed to be benign pathology even on histopathological examination. However, there was one case that was reported as acinar cell neoplasm on cytology was proved to be pancreatitis on histopathological examination. Among the neoplastic lesion diagnosed on cytology in the intra-abdominal lesion, adenocarcinoma constituted the major proportion followed by the NET.

DISCUSSION

The present study showed the usefulness of EUS-guided FNA in diagnosing different intra-abdominal lesions. It is a rapid, minimally invasive, and accurate method in the assessment of neoplastic as well as non-neoplastic lesion involving the GI tract and other adjoining structures. With a diagnosis such as tuberculosis, lymphoma, and inflammatory lesion, immediate intervention can be done based on the cytological diagnosis. Most of the time, EUS can easily delineate the origin of the lesion whether the lesion is intrinsic or is extrinsic to the wall of the lumen. However, a tissue sample for EUS FNA is required to establish a definitive diagnosis.^[4,10-16] The overall sensitivity, specificity, negative predictive value, positive predictive value, and accuracy of EUS FNA when compared to EUS alone are 87.3 versus 96.8%, 100 versus 89.1%, 85.2 versus 95.3%, 100 versus 92.4%, and 92.7 versus 93.6, respectively.^[18] However, overall sensitivity and specificity can be increased by on-site cytological evaluation.^[19] Zargar et al. have reported that the diagnostic accuracy of endoscopic mucosal forceps biopsy was lower than endoscopic FNA, that is, 87% versus 94%, respectively.^[20]

As in congruence with the previous institutional studies, our study also found pancreatic lesions to be the most common site in EUS-guided FNA followed by intra-abdominal lymph nodes.^[4] EUS FNA has a very high specificity, sensitivity, and positive predictive value in diagnosing adenocarcinoma.^[4,10,21] While we have a very small number of cases which underwent histopathological evaluation as well, it was noted that cytological diagnosis was equally reliable in case of NET.

GIST is one of the most challenging lesions encountered in EUS FNA of GI lesions which can be assessed by the fact that one case previously diagnosed as GIST turned out to be schwannoma on immunocytochemistry. Hence, cell block from EUS FNAC provides a potential tool to carry out immunocytochemistry or molecular analysis in case of diagnostic confusion and achieving higher diagnostic efficacy. EUS FNA can accurately diagnose SPN on cytology based on its typical architecture, morphology, and clinical profile. Sometimes, it is difficult to differentiate it from NET because of overlapping morphology and immunocytochemistry between the two entities.

The most important role of EUS FNA is in the staging of pancreatic adenocarcinoma especially in advanced pancreatic carcinoma where surgical intervention is futile.^[4,5] To improve the diagnostic efficacy, some investigator advocates the use molecular methods in EUS FNA samples. Takahashi et al. noted that K-ras point mutations were found in 74% of pancreatic cancers, while it is absent focal pancreatitis lesions.^[22] Similarly, Hosoda et al. suggested that achieving a diagnosis from EUS-FNA specimens of invasive ductal carcinomas, endocrine tumors, and acinar cell tumors, using immunostaining for CK7, CDX2, chromogranin, and synaptophysin with K-ras mutation analysis could serve the purpose.^[23]

CONCLUSION

EUS FNA is a valuable, accurate, and reliable tool in diagnosing neoplastic and non-neoplastic lesions of the GI tract, especially in pancreatic lesion because of its complex anatomy. An additional ancillary technique like immunocytochemistry can add more diagnostic value especially in the case of GIST, SPN, and NET.

COMPETING INTEREST STATEMENT BY ALL AUTHORS

The authors do not have any competing interest.

AUTHORSHIP STATEMENT BY ALL AUTHORS

Dr. Pankaj Kumar, Surinder Singh Rana, Reetu Kundu, Parikshaa Gupta, Manish Rohilla, Nalini Gupta, Radhika Srinivasan, Pranab Dey.

Pankaj Kumar: Data collection and manuscript preparation.

Surinder Singh Rana: Performing the USG-guided fine-needle aspiration of all these cases.

Reetu Kundu, Parikshaa Gupta, Manish Rohilla, Nalini Gupta, Radhika Srinivasan: Data collection.

Pranab Dey: Designing the study, analyzing data, and manuscript preparation.

ETHICS STATEMENT BY ALL AUTHORS

This retrospective study was carried out in the department of Cytopathology, PGIMER, Chandigarh, India. Patient consent was taken during the EUS guided fine needle aspiration cytology. Approval of the study was taken by the Departmental Committee and the conditions of the Helsinki Accords were followed in this study.