Key points

Introduction

Tissue acquisition is of paramount importance to confirm diagnosis and guide treatment in a wide variety of thoracic and abdominal neoplasms. In the past decade, endoscopic and minimally invasive techniques have become the procedures of choice to sample deep structures that could only be biopsied through open techniques in the past. The introduction of endoscopic ultrasound (EUS) has revolutionized the management of patients presenting with gastrointestinal (GI) malignancies, reaching the status of standard of care in most industrialized nations. Tumors that in the past required surgical biopsies with prolonged convalescence are now routinely accessed endoscopically, allowing expedited recovery and accelerated initiation of definitive therapies. A high sensitivity and specificity coupled with an excellent safety profile has turned EUS–fine-needle aspiration (FNA) into the preferred approach for staging mediastinal lymph node involvement in lung cancer, biopsy of pancreatic and periampullary tumors, diagnosis of submucosal tumors of the GI tract (particularly GI stromal tumors [GISTs]), and biopsy of deep-seated lymphomas. Growing experience with pancreatic and gastric tumors has allowed expansion of the indications of this approach to now include other conditions such as esophageal cancers, rectal tumors, and lung diseases. To date, EUS-guided FNA procedures offer a diagnostic accuracy of 70% to 98% depending on the location of the target lesion and experience of the operator.

Despite current widespread availability of EUS-FNA, the technique is associated with limitations related to accessibility and interpretation of cytology samples. Among the limitations of this technique, is that it only provides a cytologic specimen often with scant cellularity and, by definition, devoid of histologic architecture. EUS-FNA requires multiple needle passes and an on-site cytopathologist. Disruption of the tissue architecture during sampling of malignancies necessitating complete tissue analysis fordiagnosis and grade differentiation, such as sarcomas or lymphomas, is the most notable limitation. In addition, patients with inflammatory processes that mimic cancer pose challenges to the endoscopist and cytopathologist interpreting the results. Furthermore, in the era of molecular profiling and personalized oncologic therapies, the need for complete histologic samples has become of paramount importance. Because of these restraints, growing interest in the use of larger caliber needles has prompted trials comparing FNA with core biopsy techniques or a combination of both. Several studies have shown the efficacy and safety profile of EUS-guided core biopsies in a variety of different sites. This article discusses the importance of core tissue acquisition in GI oncology, specifically focusing on upper GI and hepatopancreatobiliary conditions.

Introduction

Tissue acquisition is of paramount importance to confirm diagnosis and guide treatment in a wide variety of thoracic and abdominal neoplasms. In the past decade, endoscopic and minimally invasive techniques have become the procedures of choice to sample deep structures that could only be biopsied through open techniques in the past. The introduction of endoscopic ultrasound (EUS) has revolutionized the management of patients presenting with gastrointestinal (GI) malignancies, reaching the status of standard of care in most industrialized nations. Tumors that in the past required surgical biopsies with prolonged convalescence are now routinely accessed endoscopically, allowing expedited recovery and accelerated initiation of definitive therapies. A high sensitivity and specificity coupled with an excellent safety profile has turned EUS–fine-needle aspiration (FNA) into the preferred approach for staging mediastinal lymph node involvement in lung cancer, biopsy of pancreatic and periampullary tumors, diagnosis of submucosal tumors of the GI tract (particularly GI stromal tumors [GISTs]), and biopsy of deep-seated lymphomas. Growing experience with pancreatic and gastric tumors has allowed expansion of the indications of this approach to now include other conditions such as esophageal cancers, rectal tumors, and lung diseases. To date, EUS-guided FNA procedures offer a diagnostic accuracy of 70% to 98% depending on the location of the target lesion and experience of the operator.

Despite current widespread availability of EUS-FNA, the technique is associated with limitations related to accessibility and interpretation of cytology samples. Among the limitations of this technique, is that it only provides a cytologic specimen often with scant cellularity and, by definition, devoid of histologic architecture. EUS-FNA requires multiple needle passes and an on-site cytopathologist. Disruption of the tissue architecture during sampling of malignancies necessitating complete tissue analysis fordiagnosis and grade differentiation, such as sarcomas or lymphomas, is the most notable limitation. In addition, patients with inflammatory processes that mimic cancer pose challenges to the endoscopist and cytopathologist interpreting the results. Furthermore, in the era of molecular profiling and personalized oncologic therapies, the need for complete histologic samples has become of paramount importance. Because of these restraints, growing interest in the use of larger caliber needles has prompted trials comparing FNA with core biopsy techniques or a combination of both. Several studies have shown the efficacy and safety profile of EUS-guided core biopsies in a variety of different sites. This article discusses the importance of core tissue acquisition in GI oncology, specifically focusing on upper GI and hepatopancreatobiliary conditions.

Diagnosis of pancreatic and periampullary tumors

The diagnostic yield for EUS-FNA of solid pancreatic tumors ranges from 75% to 98%, with rare false-positives and a false-negative rate up to 15% in the setting of chronic pancreatitis. EUS-FNA has also been proved to be helpful in the evaluation of periampullary masses that cannot be well visualized on computed tomography scan. EUS core needle biopsy (CNB) seems to be a useful adjunct in those cases in which lymphoma or histology other than ductal adenocarcinoma are suspected. By providing a histologic specimen, a better microscopic examination of the tissue may be performed while providing additional tissue for immunohistochemical characterization.

Early studies have investigated the accuracy of EUS-CNB with no clear advantage compared with FNA. In a pilot study of 18 patients, 3 of whom had pancreatic masses, Varadarajulu and colleagues determined the specimen adequacy and diagnostic accuracy of both techniques and concluded that there were no significant differences between EUS-CNB and EUS-FNA in diagnostic accuracy (78% vs 89%). Wittmann and colleagues subsequently published their experience in 83 pancreatic patients who underwent EUS-FNA alone (lesions <2 cm) or the combination of both sampling modalities (lesions ≥2 cm). In this series, adequate samples were obtained by FNA in 94% and by CNB in 81%, compared with 87% and 92% from nonpancreatic sites (n = 76), respectively. In this study, the combination of both techniques resulted in more adequate samples from nonpancreatic cases than EUS-FNA alone ( P = .044). In pancreatic cases, the investigators reported an accuracy of EUS-FNA alone of 77%; for EUS-CNB alone, 56%; and for EUS-FNA/CNB, 83%. The complication rate of this combined approach was minimal. A more recent study from the United Kingdom of 113 patients with pancreatic masses showed similar complication rates to EUS-FNA techniques.

Diagnosis and characterization of GIST and retroperitoneal soft tissue sarcomas

EUS techniques have been increasingly used in the armamentarium of diagnostic methodologies of submucosal tumors of the GI tract. Guidelines for determination of malignancy are based on tumor size, heterogeneity of the lesion, irregularity of the borders, presence of enlarged lymph nodes, and invasion of vascular structures. When FNA is added to the procedure, the sensitivity of cytologic sampling has been reported to range from 88% to 100%. Nevertheless, in cases of suspected GIST, it may not possible to obtain features that could potentially dictate the therapeutic course of action, such as mitotic counts and immunohistochemical stains, from the cytologic sample. Because of these limitations of FNA, some investigators have explored the possibility of a 19-gauge spring-loaded EUS-CNB needle in the diagnosis of suspected GIST.

DeWitt and colleagues enrolled 38 consecutive patients in a prospective single-center study of patients with lesions greater than 2 cm to undergo a EUS-CNB once the on-site FNA was deemed suboptimal. In this study, EUS-CNB provided diagnostic histology and positive immunochemistry for c-kit for 79% and 97% of the patients, respectively. Based on these encouraging results, the investigators concluded that, for the initial biopsy of GIST, EUS-CNB might be considered an acceptable alternative to EUS-FNA. In contrast, in a separate European randomized crossover study, Fernández-Esparrach and colleagues found that, when inadequate samples were obtained, the overall diagnostic accuracy of EUS-FNA was similar to that of EUS-CNB. Among the samples that were adequate, immunohistochemistry could be performed in 74% of EUS-FNA samples and in 91% of EUS-CNB samples ( P = .025). Based on this, the investigators concluded that EUS-CNB is not superior to EUS-FNA in GISTs because of the high rate of technical failure of Tru-Cut biopsy. However, when an adequate sampling was obtained with EUS-CNB, immunohistochemical phenotyping was almost always possible. Most investigators agree that EUS-CNB is useful when immunohistochemistry is necessary and previous FNA attempts have failed to provide enough tissue acquisition. The accuracy of dual sampling (EUS-CNB plus EUS-FNA) is superior to either technique alone; however, for nontransduodenal routes, the failure rate of EUS-CNB is low and the accuracy for the detection of malignancy is similar to that of EUS-FNA.

With regard to the use of EUS-CNB techniques for retroperitoneal masses such as sarcomas, the data are spare and limited to case reports, probably because, given their size, most tumors are correctly diagnosed by percutaneous core biopsy methods. The use of EUS-CNB for retroperitoneal masses could potential be applied in the future more frequently to determine tumor grade or molecular features of the mass once advances in genetic profiling become more clinically relevant.

Diagnosis of deep-seated lymphomas

EUS-FNA is routinely used, with excellent sensitivity and specificity, to sample peri-intestinal lymph nodes when nodal metastases are suspected. Despite its accuracy, diagnosis of deep-seated lymph nodes, and specifically lymphomas, can be challenging, particularly when flow cytometry is performed on the aspirated atypical lymphoid cells. In addition to tissue architecture, immunophenotype and genetic abnormalities should be assessed during diagnosis and classification of lymphoproliferative disorders. Despite the limitations inherent in EUS-FNA, investigators from Japan recently reported 240 patients with suspected lymphoma who were diagnosed by EUS-FNA. Ninety-six percent of the patients were accurately diagnosed by FNA with flow cytometry showing unusual cell populations in a significant number of patients, deeming EUS-CNB unnecessary. In those patients in whom diagnostic uncertainty is present, EUS-CNB can potentially provide additional information to aid in the final diagnosis of patients with enlarged generalized lymphadenopathy.