Endoscopic Ultrasound in the Diagnosis and Staging of Pancreatic Cancer

The article by Drs. Levy and Wiersema is an excellentoverview of the indications, technical nuances, and efficacy of endoscopicultrasound in the diagnosis and staging of pancreatic neoplasms. Endoscopicultrasonography was introduced into the diagnostic armamentarium forgastroenterology approximately 15 years ago. Although the literature suggests ageneral increase in the utility and experience with endoscopic ultrasound, thetechnique remains most effective in the hands of experienced experts like Drs.Levy and Wiersema. Their article is a complete and thorough review of theindications and expected accuracy of the technique when evaluating a variety ofdifferent pancreatic lesions.

The authors, however, understate the technical expertise required to achievethe level of results summarized in their overview. There is no question thatendoscopic ultrasound has become a common procedure used in centers specializingin pancreatic diseases. It must be clearly understood, however, that this is avery operator-dependent technology. The ability to accurately characterizepancreatic masses as inflammatory or neoplastic is very much dependent upon thetechnical skills of the person performing the endoscopic ultrasound. Similarly,its utility for obtaining tissue biopsy using endoscopic ultrasound fine-needleaspiration (FNA) techniques is directly related to the aggressiveness,experience, and technical competency of the physician performing the study.

Advantages and Limitations

In their review, the authors carefully and appropriately point out thestrengths and weaknesses of this technique. One limitation that deserves moreattention is the impact that operator inexperience can have on the utility ofthe procedure when interpreting what often may be ambiguous or equivocal imagesof complicated pancreatic lesions. This is the most obvious drawback of theprocedure, since it needs to be done in centers of excellence by physicians withsignificant experience and considerable diagnostic skills.

The authors thoroughly review endoscopic ultrasound and its utility in thediagnosis and staging of pancreatic solid mass and cystic lesions. The techniqueof endoscopic ultrasound has become essential in the characterization andlocalization of pancreatic islet tumors. As they pointed out, the endoscopicultrasound’s ability to successfully localize small islet cell tumors imbeddedin the substance of the pancreatic parenchyma has diminished the utilization ofinvasive technology such as venous sampling and angiography.

Cost-Effectiveness

Both of the latter techniques are considerably more expensive than endoscopicultrasound, and are associated with potential morbidity not found in endoscopicultrasound. Use of endoscopic ultrasound technology has resulted in significantcost savings in terms of localizing neuroendocrine tumors and assessingresectability of the pancreas.

Other important considerations in the management of islet cell tumors are thepotential presence of multifocal lesions, lymph node involvement, and extensionof the tumor into important vascular structures. As noted by the authors,endoscopic ultrasound addresses all of these issues in a very cost-efficient andminimally invasive fashion.

Indications

Several recent studies of cystic neoplasms of the pancreas have indicatedthat the incidence of these lesions is rising as they are detected incidentallyon routine computed tomograpy (CT) scans of the abdomen for unrelated causes.Thus, the pancreatic surgeon now is seeing an increased number of patientsreferred for diagnostic evaluation and possible pancreatic resection ofotherwise asymptomatic cystic lesions. Levy and Wiersema review the differentialdiagnosis for cystic lesions and indicate that the role of endoscopic ultrasoundmay be crucial in distinguishing between an inflammatory pseudocyst and aneoplastic cystic tumor of the pancreas.

In the case of an inflammatory pseudocyst, the treatment decision depends onthe size and symptoms of the pseudocyst and whether it needs to be drained witha surgical drainage procedure. On the other hand, a mucinous cystic neoplasm ofthe pancreas has the potential to become malignant if not already franklyinvasive at the time of diagnosis. Thus, a thorough characterization andassessment of the lesion may determine whether the lesion is a pseudocyst thatshould perhaps be treated with a surgical drainage procedure or a cysticneoplasm that may require a more formal pancreatic resection.

The ability to obtain cystic fluid using endoscopic ultrasound-guidedaspiration gives the surgeon and gastroenterologist additional information inmaking treatment decisions. An asymptomatic lesion that has the radiographiccharacteristics of a benign serous lesion, in which the fluid is nonviscouswithout atypical cells, may be safely observed in a high-risk surgical patient.On the other hand, the presence of a thick mucous viscous type of fluid andatypical columnar cells would remove any ambiguity or equivocation regarding thenecessity for resection of this lesion.

Most of the article appropriately addresses the most common neoplasticprocess of the pancreas—the presence of a typical pancreatic adenocarcinoma inthe head of the gland. For patients with these pancreatic neoplasms, the surgeontries to obtain as much information as possible concerning the extent of thetumor and the presence of local or distant tumor spread. The level of testingnecessary to determine these findings remains controversial in the surgicalliterature.

Accuracy of Endoscopic Ultrasound

Advocates of limited preoperative evaluation followed by surgical explorationin all patients with nonmetastatic pancreatic tumors believe that intraoperativeevaluation is the most sensitive method of determining resectability. If thetumor is found to be unresectable, subsequent operative palliation is the bestway to provide sustained relief of symptomatic jaundice or gastric outletobstruction.

Others argue that direct intraoperative assessment of the extent ofretroperitoneal tumor growth in relation to the superior mesenteric arteryorigin is not completed until the final step, after gastric and pancreatictransection, when the surgeon is committed to resection even if the entire tumorcannot be safely removed. Recent studies suggest that for most patients withunresectable disease, laparotomy for palliation may be avoided because of recentadvances in endoscopic, percutaneous, and laparoscopic methods of biliarydecompression.

Creation of Standards of Care

The algorithm of care outlined by the authors in the final pages of theiroverview is now approaching the standard of care at major centers treating largenumbers of patients with pancreatic carcinoma. The staging evaluation ofpatients with suspected or known pancreatic carcinoma begins with a multiphasethin-section CT scan. The value of beginning with the CT scan is the clearsuperiority of this technology over endoscopic ultrasound for evaluating thepresence of hepatic, peritoneal, or omental metastases.

As technology advances, the role of CT scan and magnetic resonance imagingwill continue to have notable strengths in assessing locally advanced extensionof the tumor into the peripancreatic, retroperitoneal soft tissues, thrombosisof the portal-superior mesenteric vein confluence, and encasement of theceliac trunk or superior mesenteric artery. Indeed, three-dimensionalreconstruction of the relationship of the primary tumor to the above-mentionedstructures has markedly reduced the number of patients undergoing exploratorysurgery who are found to have unresectable locally advanced tumors at laparotomy.

The thin-section contrast-enhanced CT scan currently is the procedure ofchoice in the preoperative evaluation for pancreatic duodenectomy, for thereasons mentioned above. However, it does have inadequacies: The ability toidentify small pancreatic lesions surrounded by chronic fibrosis or edema in thehead of the pancreas is known to be limited, as the authors point out.Endoscopic ultrasound can be particularly helpful in identifying such ambiguousor vague mass lesions. In addition, endoscopic ultrasound has particular valuein assessing the presence of portal vein invasion or encasement of the celiacaxis.

Perhaps one point needing clarification is the complementary role oftriphasic thin-section CT with endoscopic ultrasound. The information obtainedfrom each of these tests should be considered complementary rather thancompetitive. Using these two modalities, the sensitivity, specificity, andaccuracy of assessing mass lesions in the head of the pancreas should exceed90%. As noted by Levy and Wiersema, perhaps one of the evolving roles ofendoscopic ultrasound will be the ability to obtain a fine-needle aspiration andbiopsy when considering patients for novel preoperative therapy trials.

In these clinical trials, it is imperative to obtain accurate histologicconfirmation of a diagnosis of pancreatic adenocarcinoma. In addition,endoscopic ultrasound FNA can be performed on any suspicious regional lymphnodes to optimize the preoperative staging and stratification of patientsenrolled in these trials. This has become the most important role ofpreoperative endoscopic ultrasound in our current practice at NorthwesternUniversity Medical School. Also noted by the authors is the potential fortherapeutic endoscopic ultrasound with celiac plexus neurolysis in nonoperablepatients.

Conclusions

Drs. Levy and Wiersema have written a comprehensive review ofstate-of-the-art endoscopic ultrasound technology. As noted, this technology isvery much operator-dependent, but in excellent hands it serves as an extremelyuseful and complementary tool to current radiologic imaging with CT scan

References:

1. Brenin DR, Talamonti MS, Yang YET, et al: Cystic neoplasms of thepancreas. Arch Surg 130:1048-1054, 1995.

2. Williams DB, Sahai AV, Aabakken L, et al: Endoscopic ultrasound guided-fine-needleaspiration biopsy: A large single center experience. Gut 44:720-726, 1999.

3. Suits J, Frazee R, Erickson RA: Endoscopic ultrasound and fine-needleaspiration for the evaluation of pancreatic masses. Arch Surg 134(6):639-643,1999.