Pancreatic Cancer Screening: State of the Art

Christian Gemmel; Axel Eickhoff; Lars Helmstädter; Jürgen F Riemann


Expert Rev Gastroenterol Hepatol. 2009;3(1):89-96. 

In This Article

Screening Techniques

There is increasing interest regarding the screening of individuals at risk with different imaging modalities and serologic markers to detect premalignant and early malignant lesions of the pancreas. Until now, no ideal single screening test exists for pancreatic cancer.

A variety of tumor markers have been tested for the diagnosis of pancreatic cancer. The serum marker carbohydrate antigen 19-9 is relatively insensitive and nonspecific for early lesions and is best used as a marker for therapeutic response.[57]

The most common screening techniques used for the detection of pancreatic cancer are EUS, CT scanning, MRI and ERCP.

Brentnall et al. screened 14 patients with a family history of pancreatic cancer for pancreatic dysplasia. In total, seven of the 14 patients demonstrated dysplasia on the basis of clinical history and abnormalities on EUS and ERCP, and were referred for subsequent pancreatectomy. All seven patients showed evidence of dysplasia in their pancreatectomy specimens. Spiral CT scanning and serum tumor markers demonstrated a low sensitivity with regards to the detection of pancreatic dysplasia. Therefore, the authors concluded that clinical data combined with imaging studies (i.e., EUS and ERCP) should be used to identify high-risk patients with pancreatic cell dysplasia.[58]

Endoscopic ultrasonography - as an invasive method - entails a relatively low risk of complications, is able to detect pancreatic masses as small as 5 mm in size and can detect vascular invasion. The major drawbacks of this technique are operator dependence and a limited field of visualization for detecting lesions located in the body-tail as well detecting distant metastases.[59]

Endoscopic ultrasound has been described as a highly sensitive method; however, results for accuracy, especially in the staging of pancreatic cancer, differ. Initial studies showed excellent accuracy up to 94%, but the initial euphoria subsided as the results of later publications differed.[60,61] Accuracy seems to be approximately 60-70%. With regards to sensitivity, the determination of tumor size and extent, lymph-node involvement, and vascular infiltration, EUS was equal or superior to other imaging modalities such as CT imaging and MRI in most studies.[61] Kimmey and colleagues used EUS over CT scanning, MRI and PET in the evaluation of a family with a high risk for pancreatic cancer.[62]

Rapid advances in technology led to better CT scanning and MRI results. Soriano et al. compared EUS with helical CT, MRI and angiography in a prospective study with histopathological or surgical confirmation of their clinical results.[61] When each imaging technique was examined alone, helical CT achieved the highest accuracy in terms of assessing the extent of the primary tumor, locoregional extension, vascular invasion, distant metastasis, tumor "Tumor, Node, Metastasis" stage and tumor resectability, whereas EUS achieved the highest accuracy with regards to assessing tumor size and lymph node involvement.[61,63,64] Regarding the question of irresectability, a recent study concluded that CT and EUS should be used together to achieve the highest accuracy.[65]

Canto et al. screened for early pancreatic neoplasia with an EUS-based[66] and an EUS- and CT-based[67] protocol. In the latter protocol, pancreatic abnormalities were compared among high-risk individuals and control subjects. If EUS was abnormal, EUS/FNA and ERCP were performed. The EUS- and CT-based approach found eight patients out of 78 high-risk individuals with pancreatic neoplasms confirmed by surgery or FNA, and no pancreatic neoplasia among the 149 controls. These included six patients with IPMNs and one patient with PanIN, who were treated surgically, and one patient with an IPMN who chose not to be treated. This latter patient later developed an invasive cancer. This study demonstrates that if a precursor lesion is found, curable treatment is possible before progression to invasive cancer occurs.

Abnormalities suggestive of chronic pancreatitis were more common among high-risk individuals. Differentiation to focal pancreatitis is one of the major problems in diagnosing pancreatic cancer. Contrast-enhanced EUS using perfusion characteristics seems to be to a promising technique to discriminate between focal inflammation and pancreatic carcinoma.[68]

Jansen et al. investigated the feasibility of EUS elastography and the assessment of elastographic patterns of the normal pancreas and the pancreas affected by inflammatory or focal disease in 73 patients. They concluded that EUS elastography may be suitable for the early diagnosis of chronic pancreatitis but is not able to distinguish chronic pancreatitis from various types of neoplastic and, in particular, malignant tumors, probably owing to the similarities in their mechanical structure.[69]

Fine-needle aspiration is the best biopsy tool for the evaluation of pancreatic cancer, according to Harewood and Wiersema, who demonstrated the presence of pancreatic masses in 185 patients when prior biopsies performed by CT guidance or ERCP produced negative results. EUS-guided FNA had a sensitivity of 94% and an accuracy of 92% for detecting pancreatic cancer in patients with a negative ERCP tissue diagnosis, and 90% sensitivity and 84% accuracy in patients with negative CT-guided biopsies.[70]

A normal EUS examination appears to have a high negative-predictive value. In two studies, including 76 and 135 patients and a mean follow-up period of 23.9 and 25 months, respectively, none of the patients with clinically indeterminate pancreatic cancer and a normal EUS developed pancreatic cancer.[71,72] By contrast, Soriano et al. found negative-predictive values for locoregional extension, lymph node involvement and vascular invasion to be only 44, 65 and 74%, as confirmed by intraoperative or histopathological findings.[61]

The ability of MRI to diagnose pancreatic carcinoma has improved with improvements in technology and its application. Both gadolinium- and mangafodipir trisodium-enhanced MR images are useful for the evaluation of local tumor extension and vascular involvement of pancreatic carcinoma. Enhanced MRI has equal accuracy to helical CT for the determination of local tumor extent and vascular involvement, except for duodenal invasion and portal venous system involvement.[73,74]

Endoscopic retrograde cholangiopancreatography as an invasive method is best for the evaluation of periampullary and ductal malignancies. Brush cytology and pancreatic-fluid sampling can further help to establish the final diagnosis of pancreatic cancer. However, owing to a number of risks involved with the use of ERCP, such as pancreatitis, bleeding, perforation, sepsis and cholangitis, its indications have to be carefully considered.[75,76]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: