Whipple resection: The need for specialization, standardization and centralization.

Surgery for pancreatic head adenocarcinoma and other cancers of the peri-ampullary region has evolved over time. Although the natural biology of the disease is aggressive and long-term survival remains dismal, the technical refinements over the last two decades, coupled with improved chemo-radiotherapy that have improved outcomes more than ever before. Widely regarded as the most difficult of abdominal operations, sustained good outcomes of Whipple resection entail a long and steep learning curve, uncompromising attention to standardization and the urgent need to specialize in hepato-pancreato-biliary surgery. The clamor for centralization, especially in regions of low incidence of peri-ampullary and pancreatic head cancer, is expected to grow in the coming years.

The paper by Binziad et al. in this issue of SAJC[1] is another attempt to address the issue of perioperative and long-term outcomes after a pancreatoduodenectomy. The novelty of the paper lies in the fact that it originates from a region of low incidence and as a consequence highlights a number of areas known to influence outcomes after pancreatoduodenectomy.

Major factors dictating perioperative outcomes are texture of the pancreatic gland and duct size, and the type of pancreatic and digestive reconstruction. Other factors that play a crucial role are the level of surgical training of the operating surgeon, impact of high volumes (both hospital and surgeon) and concept of standardization of the procedure in dedicated teams. Dedicated teams include not just surgeons, but intensive care specialists, interventional radiologists and therapeutic endoscopists among others.

Tumor location is directly related to biliary and pancreatic duct dilatation. A meta-analysis[2] highlighted that the incidence of post-operative pancreatic fistula (POPF) was the highest in lower common bile duct tumors (non-dilated main pancreatic duct and softer pancreas) and the lowest in the pancreatic head cancers (dilated common bile duct and main pancreatic duct). Soon after, Hamanaka et al.[3] demonstrated that harder the pancreas and greater the duct dilatation, lesser was the pancreatic secretion with subsequent low-risk of POPF after pancreatic reconstruction. Modern imaging makes it possible to assess ductal size and assess the likely pancreatic consistency pre-operatively. This often but not always, can provide a pre-operative road-map to the pancreatic surgeon and enable him to predict and differentiate between a “high-risk versus low-risk pancreatic anastomosis.”

Pancreatic anastomotic leak remains the Achilles heel of pancreatoduodenectomy and is clearly the single most important factor influencing morbidity and mortality after pancreatoduodenectomy. Innumerable studies have focused on the technique of pancreatic anastomosis (pancreaticojejunostomy and pancreaticogastrostomy). Meta-analysis, as pointed out by the authors, concluded that there is no superiority of one procedure over the other.[4] Thus, it is not the choice of the procedure, but how well it is carried out, that matters. Any duct to mucosa anastomosis performed without tension, with fine sutures (4-0/5-0 PDS), with gentle handling of the pancreas and ensuring good vascularity without distal obstruction, will likely yield excellent results.[5] If a standardized approach is followed, excellent results, even outside of centers of excellence, can be attained.[6]

Gaining experience in performing an extremely challenging and demanding procedure such as pancreatoduodenectomy is dependent on a combination of factors. Surgical volumes (both hospital and surgeon) and the “center effect” play a key role in surgical training and evolution. Birkmeyer et al. in 2002[7] provided compelling evidence highlighting superior results in high volume centers across United States. A surgeon performing at least 16 pancreatoduodenectomies per year was likely to have a low mortality rate of 3.8% versus a surgeon performing less than 16 procedures per year with the rate increasing alarmingly. Closely related to this aspect is the issue of negotiating the learning curve. At least 60 resections are necessary to attain a level of proficiency and achieve acceptable results in a procedure where the learning continues throughout the life of a surgeon. Such a number is possible only in tertiary centers of excellence and surgeons aspiring to be dedicated pancreatic surgeons should be aware of this aspect.[8]

Developing trained teams in the developing world, especially in areas where incidence of pancreatic and periampullary cancers is much lower than the Western world and far east, can therefore be a challenging and demanding task. Service configuration with gradually increasing experience can go a long way in improving outcomes even in areas of low incidence.[9] The results of the authors indicate that they need to incorporate concepts of super-specialization, standardization, service reconfiguration, and centralization, following which a marked improvement in their perioperative outcomes can be expected.