Pancreatic cancer is the fifth leading cause of cancer mortality in the United States and has the lowest survival of any cancer. Roughly 15% to 20% of patients with pancreatic cancer present with disease localized to the pancreas. In these patients, surgical resection offers the best option for prolonging life and the only option for long-term survival. Although once associated with high operative morbidity and mortality, pancreatic resection can now be performed safely at many centers. This review will focus on the preoperative assessment, perioperative management, and outcome in the subset of patients with potentially resectable pancreatic cancer.
Spiral Computed Tomography
Spiral computed tomography (CT) scanning is the primary imaging study for patients with suspected pancreatic cancer. This technique offers both a sensitive means of diagnosing the disease as well as a cost-effective, noninvasive means of staging and determining resectability. Dual-phase spiral CT scanning is preferred, with an arterial phase 20 to 25 seconds after intravenous contrast injection using 3- to 4-mm collimation followed by venous phase 60 to 70 seconds after contrast injection using 5- to 7-mm collimation. CT evidence of a pancreatic mass, the local extent of the tumor, and the presence of metastatic disease are all important in determining whether a patient with suspected pancreatic cancer is a candidate for surgery (Figure 1).
Pancreatic cancer usually appears as a hypodense mass on spiral CT. Overall, the sensitivity of dual-phase spiral CT in detecting pancreatic cancer ranges from 85% to 95%. The technique is less sensitive for small lesions (< 15 mm), but sensitivity approaches 100% for larger lesions (> 15 mm).
Spiral CT is also accurate in predicting resectability based on the proximity of the primary tumor to major vascular structures, and CT evidence of vascular involvement correlates with overall survival in patients with pancreatic cancer.[5-7] Preservation of the fat planes around major vessels suggests lack of tumor invasion and is consistent with resectability.[5,6]
Lu et al examined 48 major vessels in 25 patients with pancreatic cancer using both spiral CT and operative dissection to determine resectability. Tumor contiguity to major vessels (portal vein, superior mesenteric vein or artery, hepatic artery, and celiac axis)ie, invasion of less than 25% of the vessel circumferencewas associated with resectability in all cases. Circumferential contiguity exceeding 50% precluded resection in over 95% of cases. The presence of periportal collaterals or dilated small peripancreatic veins suggests portal vein occlusion and is also a reliable sign of unresectability. Spiral CT angiography uses axial images to generate three-dimensional (3D) vascular images similar to those produced with traditional angiography.
Spiral CT is not sensitive in detecting small hepatic metastases, and lesions smaller than 1 cm are commonly missed. In addition, the presence of enlarged peripancreatic lymph nodes on CT scans does not correlate with the presence of metastatic cancer or survival. Enlarged lymph nodes are often benign, and metastatic cancer is commonly present in normal-sized lymph nodes. Therefore, the presence of an enlarged lymph node should not discourage surgical referral. Small peritoneal metastases are also commonly missed by CT scans in the absence of ascites.
Other Imaging Studies
Magnetic Resonance Imaging
Magnetic resonance imaging (MRI) with contrast administration provides comparable information on tumor extent and vascular or hepatic involvement to that obtained with spiral CT. Cholangiopancreatography with MRI produces images of the biliary tract and pancreatic duct of similar diagnostic quality to endoscopic cholangiopancreatography. MRI is useful in patients with significant allergies to contrast or for cases in which the CT scan does not demonstrate a mass in a patient with suspected pancreatic cancer.
Endoscopic ultrasound is as sensitive as dual-phase spiral CT in detecting pancreatic masses and can be used to accurately assess vascular involvement. Endoscopic ultrasound-guided fine-needle aspiration is also a safe method of obtaining a tissue diagnosis with less theoretical risk of tumor cell implantation than that associated with percutaneous biopsy. However, routine use of this invasive diagnostic test is unwarranted; its use should be limited to patients with equivocal CT scan findings or for obtaining a tissue diagnosis in patients with unresectable tumors.
Positron-emission tomography is a newer modality that uses the increased metabolism of labeled glucose by pancreatic cancer cells to form images. This technique may provide a sensitive means of detecting hepatic, nodal, or peritoneal metastases, and may also be able to differentiate benign from malignant pancreatic masses.
Endoscopic retrograde cholangiopancreatography is very sensitive in diagnosing ductal adenocarcinoma of the pancreas. The finding of a long irregular stricture in an otherwise normal pancreatic duct is virtually pathognomonic in the appropriate clinical setting. However, given the diagnostic accuracy of dual-phase spiral CT, this study is rarely necessary and should be reserved for patients in whom the diagnosis of pancreatic cancer is not straightforward.
Staging laparoscopy and laparoscopic ultrasound have been used to compensate for the low sensitivity of CT in detecting small peritoneal and hepatic metastases. Appropriate laparoscopy may avoid a nontherapeutic laparotomy in patients with limited survival due to unresectable pancreatic cancer. Several studies have examined the role of laparoscopic staging in patients with periampullary malignancies thought to be resectable after conventional imaging studies.[10-12]
Callery et al evaluated 50 patients with hepatobiliary and pancreatic malignancies using laparoscopy and laparoscopic ultrasound. Of these patients, 44% had either metastases or vascular invasionmissed by dynamic CT scanningthat precluded curative resection. In a larger series of 203 patients with periampullary cancer, Nieveen van Dijkum et al identified metastatic disease in only 15% of patients using laparoscopic staging.
As CT technology improves, the yield of laparoscopic staging may be decreasing. Using helical CT scans and 3D CT angiography, Saldinger et al graded vascular involvement in 52 patients with pancreatic cancer. Of 35 patients with minimal or no vascular involvement, 94% were resectable. The incidence of metastases or vascular involvement precluding resection increased dramatically with greater vascular involvement seen on CT.
Laparoscopic ultrasound appears to provide little additional information to that obtained with high-quality dual-phase spiral CT. Thus, laparoscopy should be used selectively in patients at higher risk of having peritoneal or small hepatic metastases (lesions in body and tail of pancreas, grade 2 or 3 vascular involvement, ascites, larger tumors, or other findings suspicious for unresectability on CT).[3,5] Laparoscopy is not warranted in patients who would benefit from palliative surgery if unresectable.