My preliminary review of literature has not yielded much information to answer this question. There are techniques described to prevent complications following PG, however all the articles i have so far come across regarding management of pancreatic fistulas were following PJ anastomosis.

I am also wondering about the reasons why after coming in vogue for some time during the 1990s, PGs lost favor among surgeons for a while. Could it be that the leaks following PG are more difficult to manage and lead to more serious complications because of activation of pancreatic enzymes.

Raza,

initial reports suggested that PG has lower leak rates than PJ. however RCT's by Yeo, Bassi, and Duffas have not shown that to be true. Only one RCT by Fernandez Cruz reported lower PF with a modified PG reconstruction as compared to PJ.

coming to your actual query. it is believed that PF after PG has lower complication than that after PJ. reason being that PJ leak is always associated with activated pancreatic enzymes due to mixing with jejunal contents. on the contrary pancreatic enzymes are inactivated by acid in a PG reconstruction...hence less inflammation and damage.this is also supported by a study by Aranha who showed morbidity, mortality and reexploration rates were higher after PJ leak than PG leak. Furthermore since the HJ is close to the PJ a little amount of bile leak may also ensue so more damage more inflammation.

Many surgeons still use PG reconstruction.

Thank you very much Dr Wani for a very detailed explanation on this issue, especially the pathophysiologic basis of the difference in leak rates.. I think the debate will continue which of the two procedures is better in terms of morbidity.. A recent study in lancet oncology has tried to answer this question and has shown PG to be superior to PJ, however, concerns regarding the methodology hamper a wider implication of these results..

My impression is that PJ is routinely constructed proximal to the biliary anastomosis with the intent of preventing the two mixing together to activate the pancreatic enzymes. I am sure almost all of us have manged many pancreatic fistulas following PJ that settle down only with drainage since the enzymes are inactivated and the fistulas eventually close spontaneously in absence of sepsis..

I cannot think the same about PG, and my thought process is (and i stand to be corrected) that in addition to the pancreatic anastomosis in PG, there is a gastrojejunostomy also fashioned which can bathe the pancreatic anastomosis with a constant supply of bile and intestinal secretions.. in event of a PG leak, therefore, I would anticipate the pancreatic secretions to be activated, causing widespread inflammation and thus leading to a higher incidence of hemorrhagic complications.

Since most of our experience of managing pancreatic leaks following pancreatoduodenctomies are following PJ as can be seen from the available literature, I raise this question whether the leaks following PG should be managed more agressively compared to PJ leaks even in the absence of sepsis... I particularly wanted to know if there should be a lower threshold for interventions like re-exploration and irrigation/drainage, percutaneous biliary diversion, and more liberal use of CT angiography/ angiographic embolization compared to PJ leaks.

I understand that it is difficult to answer this question from centres with low or medium volumes, however maybe surgeons practicing PG reconstruction in high volume centre have a better idea if these leaks fare better, worse or the same as PJ leaks..

regards, raza

Dear Raza,

i worked in a high volume pancreatic centre where we did more than 50 whipples a year. as per one of our ongoing studies we were comparing PG vs PJ reconstruction. although the study is not yet complete i can confirm that the leak rates do not seem to be significantly different , the morbidity especially after leaks when they do occur does seem to be higher with PJ. the PJ group had higher grade fistulas(isgpf) and required more interventions, including angio for pf related bleeds. no patient in the pg group developed a bleed. intra abdominal septic complications were also less in the pg group..

furthermore i disagree with your proposed justification.... i agree with dr nazir. pancreatic enzymes are not activated by mixing with bile or juice but once they come in contact with enterokinase in the jejunal mucosal brush border...so i dont see how more activated enzymes can leak after pg

anyway i do think more and more pancreatic surgeons are doing pg now and soon substantial literature will be available to answer your questions

Thank you Dr Wani for your explanation...

i guess the bottom line is that since the site of reconstruction is different, the way leaks behave may be different as well... and since the literature on management on leaks mainly comes from experiences with PJ, I guess we will have to wait till more experience with PG is transformed into publications..

regards, raza

All available PRT didn't show any difference btw PJ and PG.

Personally I think that PJ should be considered the standard after PD, but of course, considering the results reported in literature, everyone can be liberal to decide the preferred technique. When fistula occur in PG, not only the abdominal drain should be on place, but also a naso-gastric tube.

Good discussion,our practice has been both procedures.I will not say one is better than other in low and medium volume centres.However personnel pref is for PJ as we are used to it.

Bett NJ. Regulation of enterokinase synthesis in animal and human small intestine by luminal signals: its implication in upper gastrointestinal surgery. Br J Surg 1979;66:708-711.

Bett NJ reported that there was only minimal mucosal enterokinase activity in the remaining proximal jejunum in 5 patients given a proximal PD for a periampullary tumor in an endoscopic follow-up 6 months later.

This means that enterokinase may not play the major role in activating the trypsinogen with the resection of the proximal jejunum after the pancreaticoduodenectomy. If the trypsinogen is not activated, which enzyme sholud have the role to cause delayed hemorrhage after a PJ leakage?

Dear Raza:

It is my thereotic hypothesis that lipase (no need to be activated), not trypsinogen (need to be activated by enterokinase which is nearly ablated after PD), works with bile in an alkaline condition is the cause for the erosion of arterial wall and eventually leads to delayed hemorrhage and mortality after PD.

In terms of fistula rate,  PJ or PG should not be recognized as an universal technique for comparing. There are lots of different PJs and PGs.

More PGs use whole thickness pancreas suture, hence it may cause less fistula than the conventional PJs which conventionally don't use whole thickness pancreatic suture and easier to cause pancreatic tissue laceration while make a tie.

If fistual does occure, then two factors, mixing with bile and in an alkaline condition, will determine the severity of the complication. In such case PG is thereotically safer than PJ since it is in a acidic condition, especially if the bile is diversed from the stomach, such as double loop reconstruction PG.

I prefer a looped-biliary-diversion PJ reconstructed by whole thickness pancratic sutures .