Pancreatic cancer remains the least curable of all cancers. Over 37,000 people in the United States were diagnosed with pancreatic cancer in 2005; only 2% to 3% can expect to live 5 years or more using current treatment techniques.[1] Complete surgical removal of the cancer is considered a sine qua non for a cure to pancreatic cancer, although prolonged disease-free survival can very rarely be achieved in patients with localized pancreatic cancer who receive chemoradiation without surgery.

In this issue of ONCOLOGY, Drs. Keedy and Berlin provide a capable review of key clinical trials shaping current thinking surrounding the adjuvant therapy of pancreatic cancer. They conclude (appropriately, in my view) that adjuvant chemotherapy represents a standard of care, and that the role of adjuvant chemoradiation remains uncertain. However, their article—entitled "Adjuvant Therapy for Pancreatic Cancer: To Treat or Not to Treat?"—might be more aptly entitled "Adjuvant Therapy for Pancreatic Cancer: Can We Treat It Better?" as cure rates remain only 15% to 20% when applied to this select patient population.

Patient Selection

How might we improve curative therapy for pancreatic cancer? To begin with, patient selection has been highly variable among the world's best adjuvant pancreatic cancer studies. There are many reasons for this. First, studies to date have not required state-of-the-art imaging (triphasic, highresolution, three-dimensional reconstructed computed tomography [CT], endoscopic ultrasound). Second, the decision as to who is resectable has generally been left to the discretion of individual surgeons as opposed to prospective rules. Third, a lack of independent pathologic review may lead to the inclusion of cancers other than adenocarcinoma of the pancreas (eg, other periampullary cancers) into clinical trials.

Additionally, pathologic criteria for study entry into adjuvant trials have varied. For example, the Gastrointestinal Tumor Study Group (GITSG) trial cited by the authors excluded node-positive patients, the European Organisation for Research and Treatment of Cancer (EORTC) trial excluded patients with T3 or T4 primary lesions, and the Charité Onkologie (CONKO)-001 trial excluded patients with postoperative tumor marker values (CA 19.9 or carcinoembryonic antigen [CEA]) greater than 2.5 times normal.

Many studies (eg, the European Study Group for Pancreatic Cancer [ESPAC]-1 trial) did not require mandatory CT restaging prior to initiation of adjuvant treatment. Finally, selection biases likely exist as to who is considered postoperatively for protocol entry. As mentioned by the authors, 20% to 30% of patients even at high-volume centers never receive adjuvant therapy for a variety of reasons (postoperative complications, comorbid conditions, patient/physician attitudes, etc). In community practice, this number may be as high as 60%.[2] Thus, patient populations between major trials may vary considerably, hindering comparative analysis.

Surgery

Although surgery provides the foundation for cure in pancreatic cancer, its efficacy is limited. The goal of surgery in localized pancreatic cancer is to achieve an R0 resection (no residual macroscopic or microscopic disease). Patients who have undergone an R1 resection (residual microscopic disease) experience similar overall survival to those with de novo presentation of locally advanced disease (ie, 9- to 12-month median survival, and a 5-year survival rate close to 0%).[3] Typically, 20% to 30% of patients in adjuvant clinical trials for pancreatic cancer have had an R1 resection. Local recurrence rates folowing pancreaticoduodenectomy only for pancreatic cancer range from 50% to 85%. Systemic recurrence rates are as high as 90%, usually occurring within months from surgery.

Five-year disease-free survival of resected pancreatic cancer without additional therapy is generally 10% or less (note the median disease-free interval of 6.9 months and 5-year overall survival of 5.5% for the observation arm of the CONKO-001 trial). Given the inverse relationship between surgical morbidity/mortality and surgical volume for pancreaticoduodenectomy,[4] patients who undergo this procedure at low-volume centers may actually have a higher risk of death than cure if surgery is the only therapeutic modality employed.

Chemotherapy