Inoperable pancreatic adenocarcinoma is a dilemma that oncologists frequently encounter. Only 15% to 20% of patients are diagnosed when cancer of the pancreas is still surgically resectable. However, pancreaticoduodenectomy is the only curative option for this disease and should be offered to all patients who meet resection criteria and do not have significant comorbidities. For inoperable pancreatic cancer, the goals of treatment are to palliate symptoms and prolong life. Improved survival in locally advanced disease has been demonstrated with chemoradiation plus fluorouracil or with gemcitabine (Gemzar) alone. In metastatic disease, single-agent gemcitabine has been associated with improvement in symptoms and survival. Trials combining various chemotherapeutic agents with gemcitabine have not had a significant impact on overall survival, although meta-analyses suggest a small benefit. The targeted agent erlotinib (Tarceva) has shown a modest improvement in overall survival in combination with gemcitabine. This combination is another option for first-line therapy in patients with locally advanced or metastatic disease. Despite these recent advances, survival for patients with inoperable pancreatic cancer continues to be poor. Future investigations need to focus on understanding the molecular nature of this malignancy, with the goal of developing interventions based on this knowledge.
In this issue of ONCOLOGY, Dr. Wisinski and her colleagues at the Robert H. Lurie Comprehensive Cancer Center provide a valuable review of the therapeutic approaches to unresectable pancreatic cancer. They are thorough in their approach, and their review outlines the use of chemoradiation as well as the studies of chemotherapy, from fluorouracil to gemcitabine (Gemzar) and finally to combinations of chemotherapy. In doing so, they provide a valuable catalog of the work performed to date. They might have emphasized more, however, the conceptual challenges to the definition of unresectable pancreatic cancer itself.
What defines a pancreatic cancer as "unresectable"? It is a difficult question, as the answer is slowly evolving with advances in imaging technology, refinements in chemotherapy and radiation, and novel strategies toward the disease. Oncologists now realize that not all "unresectable" cancers are created equal, and the increasing use of neoadjuvant therapy reflects this understanding.
What we define as unresectable should reflect an important reality. That is, only patients who undergo a margin-negative (R0) resection are likely to derive benefit from surgery. The literature consistently demonstrates median survival times of 8 to 12 months among patients with positive surgical margins. Survival times improve to as much as 26 to 28 months after R0 resections. Thus, the definition of unresectability should reflect the likelihood that our current treatments for pancreatic cancer could achieve the goal of negative margins. Encasement of major vessels such as the superior mesenteric artery (SMA) or celiac axis, portal vein occlusion, and the presence of celiac or para-aortic lymphadenopathy are universally acknowledged as findings that preclude R0 surgery. We label these as "locally advanced" cancers.
One of the weaknesses frequently encountered in the pancreatic cancer literature is the lack of clear distinction between cancers that are truly locally advanced and those that might be made resectable with neoadjuvant therapy. In fairness, these distinctions are evolving as surgical technique and neoadjuvant therapies improve. However many of the trials cited by Wisinski were conducted during an era of intraoperative staging, inferior imaging, and at a time when the subtle differences between truly locally advanced cancers and borderline resectable disease would not have been appreciated.
Indeed, many studies spanning the 1980s (Eastern Cooperative Oncology Group [ECOG], Gastrointestinal Tumor Study Group [GITSG]) through the 1990s to the present time (Burris, Moore) grouped together patients with either locally advanced or metastatic disease. The inclusion of variable numbers of locally advanced with metastatic cases complicates comparison across trials and, more importantly, precludes the identification of patients for whom neoadjuvant therapy might make an R0 resection possible.
Improvements in preoperative imaging now allow for more refined staging of patients. The dramatic improvements in computed tomography (CT) imaging techniques and specific imaging protocols to visualize the relationship between the pancreas, its surrounding vasculature, and the liver can reduce the number of patients incorrectly staged as having resectable disease. One study by Vargas and others at Stanford found that, using multidetector CT, 20 of 23 patients identified on imaging as resectable were indeed able to undergo margin-negative resections. These authors note that while earlier technology incorrectly staged 21% to 47% of patients as resectable, their rate was 13% with multidetector CT scanning. Only one patient had a positive margin, which was actually anticipated by the CT scan. Conversely, in a study by Bao and colleagues, venous involvement of > 180° seen on CT was universally predictive of an R1 resection.
Thus, with increasing resolution of preoperative CT scanning, with more narrowly spaced images, and with optimal timing of contrast, many tumors that would previously have been labeled "locally advanced" might be appreciated as being more amenable to neoadjuvant treatment. Likewise, improved imaging will identify more subtle findings that may suggest unresectable features or metastatic disease, reducing the number of futile surgical procedures.
At M.D. Anderson, we have defined a category of "borderline resectable" cancers that, when recognized and treated with appropriate preoperative therapy, might be made resectable. As indicated by Wisinski et al, these include cancers that abut the SMA but not more than 180° in circumference, cancers that involve short segments of major vessels amenable to reconstruction, and cancers that encase a short segment of the hepatic artery. Our practice has been to treat such patients with neoadjuvant chemotherapy, followed by chemoradiation, in an effort to downstage these tumors.
Dr Wisinski correctly points out that novel therapies are needed. The inherent difficulty is that promising agents appear slowly, and targeted agents such as bevacizumab (Avastin), cetuximab (Erbitux), and erlotinib (Tarceva) have shown little if any improvement in outcomes. In that sense, the large phase III trials conducted in recent years have not yielded significant treatment advances. One strategy to more quickly identify effective agents might be to test candidate drugs in smaller randomized phase II trials before proceeding with larger trials.
Certainly, there will be future improvements in chemotherapy. However, improvements in trial design incorporating newer imaging technology, strategic thinking, and patient selection can make a positive impact for our patients today.
-David R. Fogelman, MD
-James L. Abbruzzese, MD
1. Klassen DJ, MacIntyre JM, Catton GE, et al: Treatment of locally advanced pancreatic cancer of the stomach and pancreas: A randomized comparison of 5-fluorouracil alone with radiation plus concurrent and maintenance 5-fluorouracil-An Eastern Cooperative Oncology Group study. J Clin Oncol 3:373-378, 1985.
2. Kalser MH, Ellenberg SS: Pancreatic cancer. Adjuvant combined radiation and chemotherapy following curative resection. Arch Surg 120:899-903, 1985.
3. Burris HA, Moore MJ, Andersen J, et al: Improvements in survival and clinical benefit with gemcitabine as first line therapy for patients with advanced pancreas cancer: A randomized trial. J Clin Oncol 15:2403-2413, 1997.
4. Moore MJ, Goldstein D, Hamm J, et al: Erlotinib plus gemcitabine compared with gemcitabine alone in patients with advanced pancreatic cancer: A phase III trial of the National Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 25:1960-1966, 2007.
5. Vargas R, Nino-Murcia M, Trueblood W, et al: MDCT in pancreatic adenocarcinoma: Prediction of vascular invasion and resectability using a multiphasic technique with curved planar reformations. AJR 182:419-425, 2004.
6. Bao PQ, Johnson JC, Lindsey EH, et al: Endoscopic ultrasound and computed tomography predictors of pancreatic cancer resectability. J Gastrointest Surg Oct 3, 2007 (epub ahead of print).
7. Varadhachary G, Tamm EP, Abbruzzese JL, et al: Borderline resectable pancreatic cancer: Definitions, management, and role of preoperative therapy. Ann Surg Oncol 13:1035-1046, 2006.