Endoscopic ultrasound in the staging of exocrine pancreatic cancer

INTRODUCTION — Exocrine pancreatic cancer is a common cause of cancer-related death in the United States and is second only to colorectal cancer as a cause of digestive cancer-related death [1,2]. Surgical resection offers the only potential cure, but because late presentation of the disease is common, only 10 to 15 percent of patients are candidates for surgical resection, and five-year survival rates remain limited. The main obstacles to potentially curative resection are the presence of distant metastases and/or invasion of nearby major vascular structures. (See "Epidemiology and nonfamilial risk factors for exocrine pancreatic cancer" and "Overview of surgery in the treatment of exocrine pancreatic cancer and prognosis".)

This topic will review the role of endoscopic ultrasound in the staging of pancreatic cancer. The general approach to the diagnosis and staging of pancreatic cancer is discussed separately. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer".)

TECHNIQUE — The technical aspects of performing endoscopic ultrasound (EUS) and EUS-guided fine needle aspiration or biopsy are presented elsewhere:

●(See "Endoscopic ultrasound: Examination of the upper gastrointestinal tract".)

●(See "Endoscopic ultrasound-guided fine needle aspiration in the gastrointestinal tract".)

●(See "Endoscopic ultrasound-guided fine needle biopsy in the gastrointestinal tract".)

STAGING — Staging patients with pancreatic cancer includes determining the tumor-node-metastasis (TNM) stage (table 1) as well as evaluating for vascular invasion [3]. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Imaging studies'.)

Multiple imaging modalities, including endoscopic ultrasound (EUS), are available for staging pancreatic cancer. EUS is effective for obtaining a tissue diagnosis, even if the tumor is poorly visualized by other imaging modalities. EUS is also used to evaluate patients with nonmetastatic disease that appears resectable on initial imaging (eg, computed tomography [CT] or magnetic resonance imaging [MRI]) [4].

Goal of staging — The goal of preoperative staging is to identify patients who might benefit from surgery while avoiding surgery in patients with locally advanced or metastatic disease. However, even with efforts to establish resectability preoperatively, many patients have unresectable tumors at surgery. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer".)

TNM staging — Pancreatic cancer is staged using the TNM system, which is outlined in the table (table 1). EUS is useful for tumor and node staging, while alternative imaging modalities, such as CT or MRI, are required for metastasis staging. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Staging system and the staging workup'.)

In general, EUS is more accurate than CT for T-staging of smaller tumors, whereas CT is more accurate than EUS for larger tumors [5]. EUS appears to be as accurate as multiphase, contrast-enhanced, multidetector CT (CE-MDCT) for N-staging [6].

Evaluation of vascular invasion — The presence or absence of vascular invasion is a major determinant of tumor resectability. However, vascular invasion does not automatically preclude resection since some vessels involved with the tumor can be resected and/or reconstructed. Thus, it is important to know whether vascular invasion is present and, if it is, to determine the specific anatomic location and the degree of vascular invasion.

Vascular invasion is probable if any of the following is present [7]:

●Peripancreatic venous collaterals in an area of a mass that obliterates the normal anatomic location of a major portal confluence vessel

●Tumor within the vessel lumen

●Abnormal vessel contour or irregular wall with loss of the vessel-parenchymal sonographic interface

ACCURACY OF EUS — The reported accuracy of endoscopic ultrasound (EUS) for tumor-node-metastasis (TNM) staging and for determining vascular invasion is highly variable among studies [8]. Some studies show superiority to other imaging modalities, while other studies show equivalent or even inferior results for EUS [9]. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Imaging studies'.)

Several factors, such as the experience of the endosonographer and the endosonographer's knowledge of prior diagnostic imaging studies [10], may influence the staging accuracy of EUS.

The type of instrument used does not appear to influence staging accuracy. Both radial and linear array EUS instruments perform equally well, although the latter has the advantage of permitting fine needle aspiration (FNA) or fine needle biopsy (FNB) to be obtained [11]. (See 'EUS with tissue sampling' below.)

TNM stage — The reported accuracy of EUS for T-staging is highly variable, ranging from 63 to 94 percent, with reported accuracies for N-staging ranging from 44 to 82 percent [4,10,12-20]. Comparing studies for T-staging is difficult because of changes in T-staging criteria over time. For example, two studies showed an accuracy of 63 and 67 percent for T-staging and 67 and 44 percent for N-staging [19,20]; however, these studies were based on an earlier version of the TNM staging classification. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Staging system and the staging workup'.)

Vascular invasion — The reported sensitivity of EUS for detecting vascular invasion has varied widely, with estimates ranging from 50 to 100 percent [21]. A meta-analysis of 29 studies showed that EUS had high specificity (90 percent) but only moderate sensitivity (73 percent) for detecting vascular invasion [21].

EUS may not be as accurate in excluding vascular invasion in patients with small pancreatic tumors [21]. However, EUS is superior to other imaging modalities, such as CT and transabdominal ultrasound, for detecting vascular invasion for tumors that are smaller than 3 cm (image 1) [22,23].

The accuracy of EUS also varies with the vessel being examined. The accuracy for detecting invasion into the superior mesenteric artery and vein is lower than that for detecting portal or splenic vein invasion [10,24,25].

COMPARISON OF EUS WITH OTHER IMAGING TECHNIQUES — Studies have compared endoscopic ultrasound (EUS) with helical CT, multidetector CT (MDCT), and MRI for initial diagnosis and/or staging of pancreatic cancer [9,26]. However, direct comparison of the data can be difficult because of differences in inclusion criteria and reference standards. The imaging modalities chosen to stage patients with pancreatic cancer should take into account the studies' accuracy as well as local expertise.

Helical CT — Multiple studies have compared the accuracy of EUS with helical CT for the evaluation of pancreatic cancer resectability [6,19,20,27-33]. However, the studies are heterogeneous, and most had methodologic flaws.

These limitations were highlighted in a systematic review that examined 11 studies with 678 patients [6]. Only three studies met all but one of the quality criteria. The most important and frequent study limitations were lack of a consecutive series of patients and biased patient selection for surgery that potentially affected validity. The older studies included in that review suggested that EUS was superior to CT for staging and determining vascular invasion. However, subsequent studies found that helical CT was more accurate than EUS for T-staging (73 versus 63 percent) and was equivalent for N-staging (62 to 63 percent versus 65 to 69 percent). EUS and CT were also equivalent with regard to detecting vascular invasion. Additional studies are needed before definitive conclusions can be made.

Contrast-enhanced multidetector CT — CT scanners with multiple rows of detectors permit imaging of larger volumes of tissue while acquiring both arterial and venous phases in shorter periods of time [19,34]. Contrast-enhanced multidetector CT (CE-MDCT) has improved the evaluation of the main pancreatic duct and the detection of small tumors [35].

Only a few studies comparing EUS with multiphase CE-MDCT for detecting and staging pancreatic cancer are available. In a report of 80 patients who underwent EUS for pancreatic cancer, EUS was superior to CE-MDCT for tumor detection (sensitivity 98 versus 86 percent) and primary T-staging (accuracy 67 versus 41 percent), but similar for nodal staging (accuracy 44 versus 47 percent) [20]. Fifty-three patients underwent surgery and 25 were found to have resectable tumors. EUS and CE-MDCT correctly identified 88 and 92 percent of the resectable tumors. Of the 28 surgical patients with unresectable tumors, EUS and CE-MDCT correctly identified the tumors as being unresectable in 68 and 64 percent, respectively.

Magnetic resonance imaging — Studies comparing EUS with MRI suggest that EUS may be more sensitive for detecting small tumors while providing complementary information regarding resectability [16,19,36,37].

A retrospective study focused on 63 patients with pancreatic cancer who underwent both examinations, using surgical results as the gold standard [36]. The tumor was resectable in 36 patients (57 percent). The sensitivities of EUS and MRI for determining resectability were 61 and 73 percent, respectively. EUS and MRI both predicted resectability in 18 patients, of whom 16 (89 percent) were found to be resectable at the time of surgery. Both studies predicted unresectability in 17 patients, of whom 13 (76 percent) were unresectable at the time of surgery.

In a prospective study of 48 patients who were staged with EUS and MRI prior to surgery, the correlation between EUS and MRI was fair (kappa 0.42), with agreement in the patient's staging seen in 74 percent [38]. EUS understaged 13 patients (27 percent), whereas MRI understaged 12 (25 percent). Only one patient was overstaged by EUS, and none were overstaged by MRI. MRI was more likely to report metastatic disease or arterial involvement.

PET scanning — The role of positron emission tomography (PET) scanning for the staging of pancreatic cancer is discussed separately. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer", section on 'Positron emission tomography scanning'.)

EUS WITH TISSUE SAMPLING

General principles — Histologic confirmation is used establish a diagnosis of pancreatic cancer. Endoscopic ultrasound (EUS) with tissue sampling may facilitate the diagnosis and staging of pancreatic cancer.

EUS-guided fine needle biopsy (FNB) is used for sampling solid pancreatic lesions located in the pancreatic head, neck, body, and tail. EUS-FNB may be particularly helpful when EUS-guided fine needle aspiration (FNA) specimens have been nondiagnostic or additional tissue studies (eg, molecular testing) are needed [39,40]. The technique and accuracy of EUS-FNB for solid pancreatic lesions and other gastrointestinal lesions is discussed separately. (See "Endoscopic ultrasound-guided fine needle biopsy in the gastrointestinal tract".)

EUS with tissue sampling can be used to:

●Biopsy lesions that are too small to be characterized by computed tomography (CT) or magnetic resonance imaging (MRI) [41]

●Biopsy lesions that are too encased by surrounding vascular structures to permit safe percutaneous biopsy [41]

●Demonstrate malignant invasion of lymph nodes located in the celiac, para-aortic, retroduodenopancreatic, or superior mesenteric regions [42]

●Biopsy suspected small metastases in the left lobe of the liver [43]

●Diagnose carcinomatosis through sampling of ascitic fluid [44]

Indications for EUS-FNA and EUS-FNB include [4,27,45,46]:

●Ruling out other types of malignancy involving the pancreas that can mimic adenocarcinoma (eg, lymphoma and small cell carcinoma, metastatic disease, and gastrointestinal neuroendocrine tumors), as well as nonmalignant processes such as autoimmune or chronic pancreatitis.

●Assisting with surgical planning (eg, a more limited resection may be possible in patients with neuroendocrine tumors).

●Confirming the diagnosis prior to surgery. However, as noted in the next section, a negative result does not exclude malignancy.

Accuracy — EUS-guided tissue sampling of solid pancreatic lesions is associated with high rates of technical success and diagnostic accuracy:

●EUS-FNB – Data suggest that diagnostic yield of EUS-FNB of solid pancreatic lesions is high (approximately 99 percent) [39,40]. The technical aspects, accuracy, and adverse events associated with EUS-FNB are discussed separately. (See "Endoscopic ultrasound-guided fine needle biopsy in the gastrointestinal tract".)

●EUS-FNA – EUS-FNA is technically successful in 90 to 95 percent of procedures, with high sensitivity and specificity [4]. A meta-analysis of 15 studies with 1860 patients found that overall, the sensitivity of EUS-FNA for pancreatic cancer was 92 percent, and the specificity was 96 percent [47]. In a second meta-analysis that included 41 studies with 4766 patients, the pooled sensitivity of EUS-FNA for diagnosing the correct etiology for a solid pancreatic mass was 89 percent, with a specificity of 96 percent [48].

The diagnostic accuracy appears to be lower in the setting of chronic pancreatitis (74 versus 91 percent in one report) [49], particularly in patients with obstructive jaundice [50].

The benefits of EUS-FNA are illustrated by a study of 559 patients undergoing EUS-FNA for evaluation of pancreatic masses [51]. In that study, when using strict cytologic criteria, the sensitivity for EUS-FNA diagnosing pancreatic adenocarcinoma was 77 percent, with a specificity of 99 percent. When patients with atypical or suspicious cytologies were reclassified as positive for malignancy, the sensitivity increased to 93 percent, without a change in the specificity.

The accuracy of EUS-FNA can improve with additional FNA passes and onsite histologic interpretation [52,53]. In one report, the presence of a cytopathologist in attendance for all aspiration procedures was associated with a high degree of accuracy (95 percent) in diagnosing pancreatic cancer [52]. The absence of a cytopathologist during the EUS-FNA procedure required an average of at least five to six passes from the pancreatic mass to ensure adequate cellularity. In another study, there was a high yield from only two FNA passes when the samples were examined both by histology and cytology [54]. In that study, cytopathologists were not onsite during the procedure.

The size of the needle used for FNA may also be important. In a meta-analysis of eight studies with 1292 patients, the sensitivity of FNA using a 25-gauge needle was higher than that for a 22-gauge needle (1 versus 0.85) with similar specificity (0.97 versus 0.93) [55].

Molecular genetic analysis — The addition of molecular genetic analysis (eg, assay for K-ras or p53 gene mutations by reverse transcription-polymerase chain reaction or restriction length polymorphisms, promoter methylation of tumor suppressor genes, differential protein expression [proteomics]) to cytologic examination may improve the sensitivity of EUS with tissue sampling for diagnosing pancreatic cancer, especially in patients with small primary tumors [56-63].

In a study with 394 patients with pancreatic masses, K-ras mutations were present in 266 of 307 patients (87 percent) with pancreatic ductal adenocarcinoma and in 3 of 87 patients (3 percent) with non-pancreatic ductal adenocarcinoma [62]. When the results of the K-ras mutation analysis were combined with cytohistopathologic results, the sensitivity for detecting pancreatic ductal adenocarcinoma was 93 percent, and the specificity was 100 percent. However, others report a lower sensitivity (38 percent) of K-ras mutations for the detection of pancreatic cancer [61]. At present, molecular analysis of pancreatic aspirate is not a routine component of the diagnostic evaluation for pancreatic masses.

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Pancreatic cancer".)

SUMMARY AND RECOMMENDATIONS

●Background – Staging patients with pancreatic cancer includes evaluating for tumor-node-metastasis (TNM) stage (table 1) and determining if vascular invasion is present. (See "Clinical manifestations, diagnosis, and staging of exocrine pancreatic cancer".)

A variety of studies, including endoscopic ultrasound (EUS) and computed tomography, are available for preoperative diagnosis and staging of pancreatic cancer. The imaging modalities chosen should take into account the studies' accuracy as well as local expertise.

●Staging – The goals of preoperative staging are to identify patients who might benefit from surgery and to avoid surgery in patients who have locally advanced or metastatic disease. The reported accuracy of EUS for TNM staging and for determining vascular invasion is highly variable. (See 'Staging' above.)

●Tissue sampling – Histologic confirmation is used to establish a diagnosis of pancreatic cancer. EUS with tissue sampling may facilitate the diagnosis and staging of pancreatic cancer. (See 'EUS with tissue sampling' above.)

EUS-guided fine needle biopsy (FNB) is used for sampling solid pancreatic lesions located in the pancreatic head, neck, body, and tail. EUS-FNB may be particularly helpful when EUS-guided fine needle aspiration (FNA) specimens have been nondiagnostic or additional tissue studies (eg, molecular testing) are needed. The technique and accuracy of EUS-FNB for solid pancreatic lesions and other gastrointestinal lesions is discussed separately. (See "Endoscopic ultrasound-guided fine needle biopsy in the gastrointestinal tract".)