Dr.Chang,

Thank for for the most interesting question!

It is the trypsin indeed that causes the blood vessels damage followed wiith hemorrhagic imbibition of tissues. These alteratins one can see in cases both of the haemorrhagic necrosis of the pancreas and from venomous snakebites (viper, cobra, etc.).

Under the normal conditions. when trypsinogen enters the small intestine from the pancreas, enterokinase (a serine peptidase) secretions from the duodenal and jejunal mucosa cleaves the lysine 15 - isoleucine 16 peptide bond of the zymogen. As a result, the zymogen trypsinogen breaks down into trypsin. However, once the trypsin is also a serine protease, and thus, if a small amount of trypsin is generated, it participates in cleavage of its own zymogen, generating even more trypsin. The process of trypsin activation is autocatalytic, and does not demand much of enterokinase to run.

Besides, there may ectopic production of the small quantities of the enterokinase in neighbouring organ, like the stomach, liver, and even in the pancreas by ectopic islets as a consequence of adaptaion, congenital anomaly, or so.

However, I do not recollect any quantitive or qualitve data about the enterokinase production in the early period after the Wipple procedure. Moreover, to the best of my knowledge, I do not know any about the stomach condition and its secretion pattern after the WP. Just to consider, what a metabolic chaos should begin after the removement of duodeno-pancreatic complex! Seemingly, this has been scarcesly investigated till now.

Major hemorrhage following PD is due to arterial erosions in most of instances. Though there is no firm experimental evidence, it is presumed that enzyme Elastase present in pancreatic juice digests Elastin in the arterial walls leading to erosions.

Dr.  Borodach and Dr. Negi:

Thank you for your expert-comments. The common explanation in the literature about this question is trypsinogen and sometimes elastase. Of cause elastase needs to be activated by the enterkinase and trypsin as well.

But, when I analyzed biliary diversion and non-biliary diversion PD reconstruction I found that PJ leakage, delayed hemorrhage, and mortality due to delayed hemorrhage rates were18.2% (12/66), 0%, 0% and13.1% (272/2079), 20.2% (55/272), 50.9% respectively.

Therefore, I use a Q shape biliary diversion for PD reconstruction, I had one PJ lekage patient with a initial whitish dischage (without mixing the bile) found from the drain at POD-7 and the patient did well for 24 days (POD-31) without sentinel hemorrhage. He had two weeks of severe constipation without being noticed. Then, bile discharge (yellowish stained) began to appear from the drain at POD-31. Eight days later (POD-39), sentinel hemorrhage appeared and progressed to massive bleeding with impending shock at POD-54. At the waiting period (3 hours) for the operation, I try normal saline instillation from the fistula above the PJ to neutralize the patient’s pH condition (bile and lipase work in alkaline condition). Unexpectedly, within 30 minutes, the patient’s bleeding gradually slowed sufficiently to justify cancelling the scheduled reoperation and finally stopped 3 days later.

The above evidence, although not strong enough, may indicate that the bile have involved in the process of delayed hemorrhage after PJ leakage.

Can you comment on the role of the bile?

Dr Chang

Thanks very much for the interesting question. Pancreatic anastomosis and its complications is still an area of debate in the surgical field, however,  the main concern for surgeons in almost all cases  is the anastomotic leakage as expected. Very few articles in the literature address the post operative physiology of pancreas and the nature, amounts and stimuli responsible for its secretions and enzymes. Is pancreatico jejunostomy more physiologic? Is pancreatico gastrostomy more safe ( as enzymes would be inactive)? I think theories will remain theories untill there is solid evidence for understanding what's really going on.

Thanks again for your sharing the interesting case of saline instillation,  sure I will keep it in mind.

Current PD reconstructions, if not changed, remains a PJ leakage rate around 15%. It seems that no improvement can be further expected. Therefore, my first consideration is how to prevent the occurrence of delayed hemorrhage and the consequent mortality. And gradually I found out that biliary-diversion PD may be the answer. And the case I mentioned above gave me the hint that my theory may be right. In that case, bile can work with lipase in the pancreatic secretions and cause hemorrhage. If there is any research evidence to support or deny the bile/lipase theory will be most welcome.

Now I prefer whole thickness (pancreatic and jejunal) mattress suture for the end-to-side PJ anastomosis to prevent the formation of PJ leakage.

Dr.Chang,

I would like to know about your way to perform PJ-recontruction (a drawing would the best if possible) and what namely did you find to be the source of that hemorrage, please. I hope that is not too difficult, is it?

There is a lot of reason for bleeding to occur both in the eerly and remote periods of the pancreato-duodenal resection depending on the main disease the patient was operated upon, and depending not. For example, my colleagues from the Novosibirsk Regional Hospital found arterial pseudoaneurisms development in the chronic pancreatitis that may become the source of post-operative bleeding after the Beger procedure [Blagitko E.M, Dobrov S.D, Polyakevich AS., ,,, all were in Russian]. Actually, I may take the email address of Dr.Polyakevich over to you if needed to communicate with.

It should be noticed, that the pancreato-duodenal complex is the very heart of the digestive system and, therefore, of the whole homeostasis. So, the consequences of its removement are difficult to account in all those diversity and complexity. What will be (and how homogenous in various persons) the stomach secretion under conditions of the acute abolition of the secretin stimulation during digestion and prostprandial? How much will be the intestinal microflora changes, and what degree of the SIBO syndrome will develop, if any? And so on, and so on... The consequences of the PDR demand the most complex. long, and multidisciplinary approach to perform: from its endocrynology to microbiology, from immunology to dietology.

We prefer usually (wherever possible) the PG-reconstruction to perform. These patients go better, and we for ourselves explain this on grounds of very low, almost sterile stomach content comparing to that of the jejunum. However, we have not investigated or even confirmed this idea. Perhaps, somewhen, somehow in Future...

Sorry for being boring and long. Just one thing to finish.

T.Terada et al, 1990. found that app.5% of the human liver contain the exocrine pancreatic tissue [Terada T, Nakanuma Y, Kakita A. / Pathologic observations of intrahepatic peribiliary glands in 1,000 consecutive autopsy livers: Hetero-

topic pancreas in the liver // Gastroenterology 1990; 98: 1333-1337]. Besides, the presence of both the exocrine and endocrine islets is characteristic for the teleost liver. So, one can raise a question, why bile acids/salts do not activate the lypase in this case? Why these livers do no explode like bombs? Still, there is no answer.

Did you study the patient bile? Perhaps, it contained something like the enterokinase itself? 

Dear Prof. Borodach,

Thank you for your information of liver origin pancreatic enzymes.

Instead of trypsinogen and elastase, I assume that lipase may play a pivotal role in the mechanism of delayed hemorrhage (DH) or aneurysm formation. In my case of PJ leakage, "DH" happened 8 days later when bile mixed with pancreatic secretions.

I don't study the bile, but I am very curious about the mechanism of "DH".

I practice my whole thickness mattress liver suture on liver resection. (Fig. 2 - Chang YC, Nagasue N, Lin XZ, Chen CS. Easier hepatic resections with a straight needle. Am J Surg. 2001;182:260-4.)Therefore, I try to apply it on PJ. When you make a knot (mattress suture) including whole pancreatic tissue and whole layer of jejunum below  (post. jejunal wall) and above (ant. jejunal wall), this knot will not tear the pancreatic tissue at all, no matter its texture is soft or hard. This whole thickness mattress suture is applied on the PG anastomosis as well. 

Article Easier hepatic resections with a straight needle

Dear Dr.Chang,

Thank you for explanations of the pancreato-jejunal anatomosis technique! However, I am not sure that I understood it right due to my very superficious knowledge of English. Do you not include the duct of Wirsung into the suture line at all while performing the PJ anastomosis?

Dear Prof. Borodach:

 A short stent tube is inserted into Wirsung duct before starting whole thickness mattress suture to prevent iatrogenic suture-ligation of the pancreatic duct. 1. To set two mattress sutures (from ant-post. pancras to seroa-mucosa of post. jejunum; then, mucosa-seroa of post. jejunum to post-ant. pancreas) without tying, to the pancreas, cephalic to the stent. 2. another three same sutures without tying to the caudal side of the stent, These 5 longitudinal mattress sutures are enough to cover the pancreatic stump. 2. A second line of mattress sutures may be added 5 mm distal to this line - this is optional. 3. A continued suture to the margine of the whole layer of post. jejunum-to dorsal margin of the pancreatic stump (not whole layer), then to the margine of the ant. jejunum-to ventral margin of the pancreatic stump and make the first knot for the PJ anastomosis. 4. the pre-hold first line whole thickness suturs now start to suture the whole layer of the ant. jejunum and make individula knots. (one line of whole thickness suture has 5 konts, if two lines it will be 10 knots). 5. I will add another whole thickness suture on the cephalic and caudal edges of the PJ, respectively.