Pancreatic Cancer

Diagnostic Imaging Continued

•   Ultrasonography

  • Preoperative ultrasonography may be useful in assessing pancreatic tumors.⁹

    • The important endpoints having to do both with tumor characteristics and ultimate resectability decisions may be addressed using this technology.

    • A comparison of spiral CT vs. color-Doppler ultrasound was considered in prediction of portal-mesenteric trunk involvement in pancreatic cancer.

      • In that study 95 patients with pancreatic cancer were assessed with respect to portal-mesenteric involvement (PMT) using both methods. Vascular involvement was evaluated based on five stages. The preoperative findings were compared with intraoperative ultrasound. 82 of the 95 patients ultimately underwent surgery. For the CT technique (spiral CT) the sensitivity was 98% with a specificity of 79% and overall accuracy of 80.2%. The positive predictive value was 87.5%; whereas, the negative predictive value was 96%. Referencing the same variables, the results from color Doppler ultrasound were 92.3%, 72.7%, 72.8%, 79.5%, and 88.8%, respectively. Results are summarized in the table below. These authors concluded that the gold standard was spiral CT for assessment ofportal-mesenteric trunk involvement in pancreatic cancer with no advantage gained by using color Doppler ultrasound. Furthermore, these workers concluded that both techniques are useful in deciding whether the pancreatic tumor might be resectable.²⁸  

  • Comparison of the utility of spiral CT and color-Doppler ultrasonography in predicting portal-mesenteric trunk involvement in pancreatic cancer²⁸

    Imaging Method	Computed Tomography (CT)	Ultrasonography
    Sensitivity	98%	92.3%
    Specificity	79%	72.7%
    Overall Accuracy	80.2%	72.8%
    Positive Predictive Value	87.5%	79.5%
    Negative Predictive Value	96%	88.8%

  • Another study compared ultrasonography with both helical CT and computed tomographic angiography with respect to their ability to detect unresectable periampullary cancer. (Periampullary cancer refers to cancer forming near the ampulla of Vater (An enlargement of the ducts from the pancreas and liver where they join and enter the small intestine.)

    • Periampullary cancer is often unresectable because of either local vascular involvement or the presence of distant metastatic disease. Color Doppler ultrasound imaging as well as grayscale imaging increase the likelihood of detecting vascular involvement. In this study ²⁹,23 consecutive patients presenting with periampullary cancer were evaluated in terms of prospective disease staging by comparing helical CT scanning and CT angiography with color Doppler/grayscale abdominal ultrasound. Absence of vascular involvement was graded as 0; whereas, total vessel occlusion was designated grade 4. Vascular structures assessed included superior mesenteric artery, splenic vein, superior mesenteric vein, and portal vein.

    • Statistical analysis described in detail in the published manuscript supported the conclusion of  agreement between CT angiography and ultrasonography with respect to detection of vascular involvement in all vessels. Both methods identified unresectable disease.

    • In addition, both methods, CT angiography and ultrasonography, were equally poor in preoperative identification of both metastases and lymphadenopathy. As a premise, there must be adequate ultrasonographic visualization of the pancreatic head in order for this technique to be comparable in accuracy to that observed with CT angiography for identification of unresectable periampullary disease. CT angiography may be helpful in assessing vascular involvement beyond which had been available through conventional CT scanning.

    • More recently, however,  dynamic contrast infusion with helical (spiral) CT methods may be the more commonly used approach in resolution of this question.⁴  

  • Endoscopic ultrasonography which will be considered in the next section is an important method for detection of pancreatic lesions.

    • One study (30) considered this technique and 50 patients most of whom had pancreatic cancer (42) with the remaining 8 exhibiting nodular fibrosis as a consequence of chronic pancreatitis. Using endoscopic ultrasound lesions were readily identified in all patients even if the tumor size was less than 20 mm in diameter. Pancreatic lesions were typically hypoechoic (dark) masses.

    • Differentiation of malignant from benign pancreatic lesions was more likely in tumors larger than 30 mm in size but challenging in those lesions less than 20 mm in size.

    • Furthermore, vascular involvement of disease might be facilitated using endoscopic ultrasound.

     

  • Pancreatic Cancer and Associated Anatomy as Visualized by Utrasonography³¹  

    Attribution:  Harrison's: Principles of Internal Medicine

    Sonogram above: the pancreatic carcinoma is designated by "P". Dilation of intrahepatic bile ducts (d), portal vein (pv) and inferior vena cava ((IVC) is also evident.31

  •  

    Ultrasonographic examination of pancreatic carcinoma localized in the head of the pancreas⁴  

    Attribution: Abeloff's Clinical Oncology

    In this example, the pancreatic mass exhibits low signal intensity (hypoechoic). pancreatic tissue is seen around the tumor.4

  • Endoscopic ultrasonography

    • Endoscopic ultrasonography (EUS) may be useful if the pancreatic tumor cannot be visualized on CT imaging. Furthermore, information obtained by EUS can be useful in some special circumstances. These circumstances include not only diagnosis and staging generally but also if tissue diagnosis must be obtained prior to neoadjuvant therapy (discussed later), before a patient enters a clinical trial, and/or prior to chemotherapy for in advance disease management.⁹

    • EUS involves imaging through the stomach wall or duodenal wall; routinely, specific diagnosis as to tumor type might not be necessary prior to resection, since any "suspicious" lesion detected by imaging would usually be managed by surgery.⁹

    •  However, endoscopic ultrasound, in some centers, may be increasingly used for both diagnosis and staging.4 EUS is effective in analyzing tumor size, mesenteric and portal vein involvement and in assessing regional lymph node involvement. EUS can also be used for a tissue sampling both from pancreatic lesions and from lymph nodes located near the stomach and duodenum. Analysis of these nodes may reveal locoregional disease. An advantage of endoscopic ultrasound is that it may be performed under only mild sedation and outside the operating room setting.⁴

    • In patients who receive EUS guided fine-needle aspiration can typically expect reliable, accurate results, thereby avoiding exploratory surgery for pancreatic cancer diagnosis. The accuracy of this approach was assessed in a study involving 233 patients. Disease classification was established using contrast-enhanced multislice CT scanning thus identifying patients with resectable, locally advanced, or metastatic disease. Using EUS-fine needle aspiration (FNA), cancer was established in 216 of the 233 patients (93%), with 15 individuals ( 6%) exhibiting benign disease and with unknown diagnosis in 2 patients (1%). The sensitivity of EUS-FNA for a pancreatic malignancy diagnosis was 91%; whereas, specificity was 100% and accuracy was 92%. When just considering the 216 patients proven to have cancer, EUS-FNA was diagnostic in 197 or 91% of the patients. 90% of the 107 patients with resectable disease was identified and 97% of patients with locally advanced disease was identified, compared with 87% of those with metastatic disease.³²

      • A related technique, intraductal ultrasound (IDUS) can be used during ERCP (endoscopic retrograde cholangiopancreatography, discussed next). Intraductal ultrasound can be helpful in evaluation of biliary and pancreatic disorders given the smaller diameter, flexibility, and image quality offered by IDUS imaging systems. Pancreatic disorders that may be most effectively analyzed using this technology include undefined strictures, ampullary neoplasms, and luminal  filling defects.³³