Surgical Management of Pancreatic Cancer

June 1, 2002

Drs. Ahrendt and Pitt should be congratulated on a comprehensive and well-presented review of the surgical management of pancreatic cancer. Unfortunately, pancreatic cancer continues to be a major cause of cancer-related death. The majority (80%) of patients still present with unresectable locally advanced or metastatic disease.

Drs. Ahrendt and Pitt should be congratulated on acomprehensive and well-presented review of the surgical management of pancreaticcancer. Unfortunately, pancreatic cancer continues to be a major cause ofcancer-related death. The majority (80%) of patients still present withunresectable locally advanced or metastatic disease.

Preoperative Imaging Modalities and Laparoscopy

In the 1980s, several surgical series reported dismal resectability rates of25%. Due to poor imaging techniques, most patients were being stagedintraoperatively by manual palpation of the plane between the tumor and themesenteric vessels. Patients found to be unresectable had then undergone a majorlaparotomy with little benefit, and postoperative recovery delayed theirtreatment options.

Dual-phase contrast-enhanced spiral computed tomography (CT) hasrevolutionized both the detection and staging of pancreatic cancer. The accuracywith which spiral CT predicts resectability ranges from 75% to 90%.[1] CTcriteria for surgical resectability include (1) the absence of extrapancreaticdisease, (2) a patent superior mesenteric vein-portal vein confluence, and (3)no direct tumor extension into the celiac axis or superior mesenteric artery.

Investigators at the Massachusetts General Hospital reported that 24% ofpatients thought to have resectable pancreatic cancers on CT scan had occultmetastatic disease found on diagnostic laparoscopy.[2] Patients found to haveoccult disease were spared an unnecessary laparotomy. However, with improvedstate-of-the-art spiral CT, routine use of staging laparoscopy may not be easilyjustified.[1]

Ultrasound and PET

More recently, endoscopic ultrasound has been shown to be helpful indetecting small pancreatic cancers. In our practice, the technique is used inpatients with a clinical suspicion of pancreatic cancer and an equivocal ornegative CT scan. Although the procedure is very user-dependent, it offers theadditional benefit of image-guided tissue diagnosis. More experience withwhole-body positron-emission tomography may help stage patients more accuratelyand also may be able to differentiate benign from malignant pancreatic tumors.

Surgical Results and Pathologic Analysis

Despite improved resectability rates and decreased perioperative mortality athigh-volume surgical centers, long-term survival followingpancreaticoduodenectomy remains poor. Median survival in most series ranges from18 to 20 months, and 5-year actuarial survival rates range from 7% to 25%.Patterns of failure after curative resection for pancreatic carcinoma involveboth local recurrence (60%) and distant hepatic metastases (60%).[3] Even amongpatients thought to be resectable for cure by preoperative CT and intraoperativeexploration, 50% will have either gross or microscopic involvement of thesurgical margins.[4] The most commonly involved is the retroperitoneal margin,which corresponds to the tissue along the proximal 3 to 4 cm of the superiormesenteric artery wall.

Several studies have demonstrated that patients with grossly ormicroscopically positive surgical margins have a median survival of only 8 to 10months, similar to the survival of patients with unresectable locally advancedtumors. From a technical aspect, failure to mobilize the superior mesenteric-portalvein may result in a positive margin due to incomplete removal of the uncinateprocess and the mesenteric soft tissue adjacent to the superior mesentericartery. New treatment strategies to maximize margin-negative resections inpatients with pancreatic cancer will hopefully be one component of improvedlocoregional tumor control.

Both the evaluation of future innovative treatment strategies and thedevelopment of reproducible prognostic predictors of treatment failure depend onan accurate standardized pathologic assessment of the surgical specimen.[5]Retrospective pathologic analysis of archival material does not allow foraccurate assessment of resection margins or the number of lymph nodes retrieved.A prospective pathologic evaluation must be initiated by the surgeon’sorientation of the specimen with the pathologist, and identifying the bile duct,pancreatic, and retroperitoneal transection margins.

The final pathologic evaluation of permanent sections should include adescription of tumor histology and differentiation, a gross and microscopicevaluation of the tissue of origin (pancreas, bile duct, ampulla of Vater, orduodenum), an assessment of the maximal transverse tumor diameter, and a reportof lymph node status. A standardized pathologic analysis will allow for anaccurate comparison of novel treatments of pancreatic cancer.

Adjuvant and Neoadjuvant Treatment Strategies

In 1985, the Gastrointestinal Study Group (GITSG) reported the first clinicaltrial evaluating the use of adjuvant fluorouracil (5-FU)-based chemoradiationfor resectable pancreatic cancer. This study demonstrated an improvement in bothmedian and 2-year overall survival. With recent data showing a survival benefitfor gemcitabine (Gemzar) over 5-FU in patients with advanced disease, studiesare currently examining the role of gemcitabine in the adjuvant setting.

The phase III Radiation Therapy Oncology Group (RTOG) trial 9704 is comparingpostoperative adjuvant infusional 5-FU vs gemcitabine, both followed by 5-FU-basedchemoradiation. Unfortunately, the morbidity and often prolonged recovery timeassociated with pancreaticoduodenectomy prevent the timely delivery ofpostoperative chemoradiation in at least 25% to 30% of eligible patients.[6]

This risk of delaying postoperative adjuvant chemoradiation has promptedinvestigators to assess the safety and efficacy of administering preoperativechemoradiation to patients with resectable adenocarcinoma of the pancreas. Theuse of preoperative chemoradiation is supported by the following considerations:

  • Radiation therapy is more effective on well-oxygenated cells that havenot been devascularized by surgery.

  • Peritoneal tumor implantation due to the manipulation of surgery may beprevented by preoperative chemoradiation.

  • The high frequency of margin-positive resections recently reportedsupports the concern that the retroperitoneal margin of excision, even whennegative, may be only a few millimeters wide.

  • Patients with disseminated disease evident on restaging studies afterchemoradiation will not be subjected to laparotomy.

  • Because chemoradiation will be administered first, delayed postoperativerecovery will have no effect on the delivery of multimodality therapy.

A prospective trial of preoperative 5-FU chemoradiation demonstrated thatthis treatment was well tolerated, and all patients received treatment.[7] Thelocal recurrence rate decreased dramatically to 11%, compared to 60% inhistorical controls treated with surgery alone. This improved locoregional tumorcontrol rate was associated with a decrease in the incidence of positivesurgical margins (18%).

Future Directions

Because preoperative chemoradiation is a locoregional therapy, this strategyis not expected to significantly improve distant metastatic disease rates.Future treatment strategies should incorporate preoperative chemoradiation forlocoregional disease control and novel systemic treatment to control possiblemicrometastatic disease (and to possibly downstage locally advanced tumors).Major improvements in overall survival will await these novel systemictherapies, but the importance of locoregional tumor control should not beunderestimated.

At Emory, we have incorporated a preoperative strategy with full-doseinduction chemotherapy followed by chemoradiation for improved locoregionaldisease control. This phase I/II investigation examined the safety of induction5-FU, cisplatin, and gemcitabine followed by hyperfractionated chemoradiation in29 patients. Remarkably, we have seen a 40% objective partial and completeresponse rate with the induction chemotherapy. We are now investigating thistreatment strategy in a phase II trial in borderline and resectable patients.


1. Pister PW, Lee JE, Vauthey JN, et al: Laparoscopy in the staging ofpancreatic cancer. Br J Surg 88:325-337, 2001.

2. Jimenez RE, Warshaw AL, Rattner DW, et al: Impact of laparoscopic stagingin the treatment of pancreatic cancer. Arch Surg 135:409-415, 2000.

3. Griffin JF, Smalley SR, Jewell W, et al: Patterns of failure aftercurative resection of pancreatic carcinoma. Cancer 66:56-61, 1990.

4. Willett CG, Lewandrowski K, Warshaw AL, et al: Resection margins incarcinoma of the head of the pancreas. Ann Surg 217:144-148, 1993.

5. Staley C, Cleary K, Abbruzzese J, et al: Need for standardized pathologicstaging of pancreaticoduodenectomy specimens. Pancreas 12:373-380, 1996.

6. Evans DB, Wolff R, Crane C: Neoadjuvant strategies for pancreatic cancer.Oncology 15:727-744, 2001.

7. Staley CA, Lee JE, Cleary KR, et al: Preoperative chemoradiation,pancreaticduodenectomy, and intraoperative radiation therapy for adenocarcinomaof the pancreatic head. Am J Surg 171:118-125, 1996.