Survival from pancreatic cancer: it’s not just about the surgical mortality

The article by Speer and colleagues in this issue of the Journal reporting outcomes in patients undergoing pancreatectomy for pancreatic cancer in Victoria in 2002–2003 indicates the poor outcome for this condition on a population basis.1 Consistent with international series are the observations that only a small percentage (10%–15%) of patients have resectable disease at diagnosis; that survival is significantly improved by obtaining a clear margin; that the median survival is relatively short after resection; and that long-term cure is rare, with even some of the small number of 5-year survivors relapsing after this milestone.

Speer et al’s report considers adenocarcinoma in all pancreatic locations, but here I consider their findings regarding cancer of the head and neck of the pancreas, treated with pancreaticoduodenectomy. What is striking in their study is the large number of surgeons undertaking this procedure: 31 surgeons performed 75 resections over the 2-year period — about one per surgeon per year. The rate of positive margins was nearly 40%, and 90-day mortality was 9.3%. Of 133 patients explored with intent to resect, 58 did not proceed. While benchmarking indicators is an inexact science, these figures would suggest room for improvement. There is considerable evidence that concentrating complex procedures such as pancreatectomy in specialised centres leads to significant outcome improvement.2

Western Australia has responded to the evidence of inferior outcomes with widely distributed procedures by limiting pancreatectomy to three hospitals. It has also been mandated in WA that all cases are to be discussed at a multidisciplinary meeting and that the surgeons require appropriate credentialling. These requirements were based on the accreditation criteria for training units in hepatic, pancreatic and biliary surgery as prescribed by the Australian and New Zealand Hepatic, Pancreatic and Biliary Association (ANZHPBA). Evidence of improved outcomes with this approach is awaited (Professor David Fletcher, Winthrop Professor of Surgery, Fremantle Hospital, WA, personal communication).

The patient population reported in the Victorian series probably did not routinely receive adjuvant chemotherapy. Randomised trials reported since 2004 have demonstrated an approximate doubling of long-term survival after pancreaticoduodenectomy with adjuvant chemotherapy,3-5 and it is now the standard of care. What is less certain is the benefit of neoadjuvant chemoradiotherapy. Neoadjuvant strategies are popular in the United States but have not been tested in randomised trials. There has been considerable interest in using neoadjuvant chemoradiotherapy in patients with borderline resectable disease, to increase both rates of margin-negative resection and long-term survival.6,7 What is evident from reported series is that systemic relapse is common and that micrometastases occur early. Increasing focus on the local site without a strategy for the systemic nature of the disease will be unlikely to improve outcomes. In recognition of this, and to redress the lack of randomised studies evaluating neoadjuvant chemoradiotherapy, the European Study Group for Pancreatic Cancer (ESPAC) is proposing a three-armed, randomised Phase II study (ESPAC-5). While all patients will receive postoperative chemotherapy, they will be divided into the three trial arms of surgery, chemotherapy before resection, and chemoradiotherapy before resection.8

There has been significant recent interest in defining goals and research priorities for pancreatic cancer.8,9 A common theme emerging is the necessity for translational research and the identification of molecular markers or targets for therapeutic agents. Identifying molecular mechanisms or different genetic mutations in pancreatic cancer may help explain differences in outcome and responsiveness to current chemotherapeutic agents. Considerable progress has been made with mapping of the pancreatic cancer genome through the PanScan project, which has confirmed significant heterogeneity.10 Recent publications have identified genetic polymorphisms affecting pancreatic cancer susceptibility,11 expression of vascular endothelial growth factor as a negative prognostic factor,12 and serum levels of carbohydrate antigen 19-9 (CA19-9) as having a potential therapeutic predictive and prognostic role.13

Identification of markers or targets may improve selection of patients for different treatment strategies and realise the potential for personalised therapy.14,15 Benefits could include avoidance of major surgery in patients who will relapse early, better selection of either neoadjuvant or adjuvant agents, and stratification of palliative treatments by identifying better prognosis subgroups among those with unresectable disease.

Returning to the concept of specialised centres, the obvious primary objective of forming such centres is to concentrate surgical experience and improve surgical and non-surgical outcomes. Further, treatment of the subpopulation of patients with borderline resectable disease with a view to progressing to margin-negative resection is a considerable clinical and resource challenge6 and is not the domain of an occasional pancreatic surgeon. Finally, there is the challenge of improving knowledge and joining collaborative research networks, and enrolling patients in clinical trials with a translational component. It is only through exploring the translational aspects and developing more specific personalised treatments14,15 that outcomes will improve and patient suffering from the pursuit of futile treatments will be minimised.

Considering all these factors, the moment has arrived to grasp the nettle of specialised centres in Australasia. It is not just about surgical mortality, but providing better care to patients with pancreatic cancer and contributing to the research effort and knowledge base for this lethal condition. The quality of pancreatic cancer surgery will be a major discussion item at this year’s meeting of the ANZHPBA in November.