Sunday, February 21, 2016

Why Did This Happen?


It was a Whipple. Not Mr. Whipple of Charmin fame, but the operation which carried the name of Whipple. The proper name is Pancreatoduodenectomy and it is a procedure which most surgeons decline to undertake.
Named for Allen Whipple, the operation consists of removal of the gallbladder, a portion of the common bile duct, the head of the pancreas and the duodenum with the most proximal jejunum. In its original form it was performed in 2 stages, first the resection and then, the following day, the reconstruction. It is an operation for tumors in the head of the pancreas, distal common bile duct and duodenum. Occasionally it is utilized for benign strictures of the distal bile duct, often due to chronic pancreatitis.
To me the operation presents a study in anatomy and an exercise of careful dissection. The portal triad, consisting of bile duct, hepatic artery and portal vein must be teased apart, vessels to the liver preserved and the pancreas gently lifted off the superior mesenteric and portal vein.
Mark needed such an operation.
Mark was 52 and came to the hospital because he had turned yellow. He had noticed the color change for about two weeks. He also felt weak, had vague abdominal discomfort and had lost about fifteen pounds. In the Emergency Room he was found to have gallstones, dilated bile ducts and “focal” pancreatitis on CT Scan of the abdomen.
His bilirubin level was 12, which is why he looked yellow. Normal bilirubin level is under 1.0.
Fortune put me on call for the ER that day and I was consulted by Mark’s admitting physician. His presentation was not typical of gallstone pancreatitis. His pain was mild. The bilirubin was too high, weight loss did not go along with acute pancreatitis.
And, when I reviewed his CT Scan, the finding of “focal” pancreatitis looked like a tumor to me and to another Radiologist colleague with whom I was reviewing the images.
Further work up with MRCP revealed the stricture in the bile duct and a definite mass in the pancreas.
Mark needed a Whipple.
He was otherwise healthy. The surgery was scheduled for two days later.
At 9:08 am on a Tuesday I made my chevron incision in Mark’s upper abdomen, dealt with the usual bleeders in the abdominal wall and entered his peritoneal cavity. There was no fluid, usually a good sign, the liver was discolored from his jaundice, but free of any masses. Palpation of the head of the pancreas revealed the expected hard mass, a sign of either inflammation or tumor.
Next came the first decision. Should I biopsy the pancreas? Would such a biopsy change anything? Suppose the biopsy revealed cancer. The treatment would be to resect the tumor, that is, to continue with the Whipple procedure. But, suppose the biopsy only revealed inflammation, or fibrosis, without any tumor? The answer is that it wouldn’t make a difference. Mark still needed the Whipple. Pancreatic cancer is a funny beast. It causes an intense fibrotic reaction. The tumor often is obscured by this fibrosis and diagnosis, particularly on intraoperative frozen section is often very difficult. Mark’s presentation was textbook cancer of the pancreas. Even if his jaundice turned out to be benign disease, which was unlikely, the proper treatment was still the Whipple.
The first decision was, therefore, simple; no biopsy.
Next I removed the gallbladder. It contained a bunch of stones, but otherwise wasn’t inflamed. It was in the bucket in about ten minutes.
Time to get going.
I Kocherized the duodenum, which meant I divided the peritoneal attachments of the duodenum which allowed me to lift the pancreas and duodenum off any underlying retroperitoneal structures. The main structure of concern was the Inferior Vena Cava. I was now able to hold the entire head of the pancreas in my hand. The mass felt hard, while the neck felt soft and spongy, the way normal pancreas was supposed to feel. I mobilized as much as I could, until I could feel the pulse of aorta adjacent to the pancreas.
Time to tackle the porta hepatis.
This part of the operation is the time to be careful and methodical. The common hepatic artery arises from the celiac axis, a major branch from the aorta. This common hepatic artery branches into the proper hepatic artery which is one of the major structures of the porta hepatis and supplies arterial blood to the liver, and the gastroduodenal which supplies the duodenum and the pancreas, but also gives rise to the right gastric artery which is important to preserve in the so called pylorus preserving Whipple which is what I planned for Mark.
Carefully I exposed the arteries. It turned out to be one of the easier dissections of these structures I’d ever performed. There was considerable space between the common bile duct and the arteries, making separation of these structures very straightforward. All the branches were dissected and encircled with silk ties, but none of these arteries were divided yet.
Next the bile duct was identified, dissected free and also encircled. It was dilated to about 1.6 cm, about twice normal size. Behind the bile duct was the portal vein, usually the make or break structure in Whipples. This large vein is formed where the superior mesenteric vein and splenic vein meet. It abuts the head of the pancreas and tumors often grow into and around it, which renders these tumors unresectable, at least for cure. The anterior surface of the vein is dissected free so that the neck of the pancreas can be lifted off the vein.
Mark’s portal vein easily separated from the pancreas.
Once all this dissection being done it was time for the real operation to start.
The gastroduodenal artery was ligated and divided, while the proper hepatic and right gastric arteries were preserved. The bile duct was also transected, an act which always fills me with an oxymoronic combination of fear and delight. This unusual feeling is due to the repeated drilling during residency to “know where the bile duct is, do not injure the bile duct,” every time a gallbladder was removed.
Next the neck of the pancreas is lifted off the portal vein and divided, vessels running with this organ are sutured as there is always some bleeding. Great care is taken to not suture the pancreatic duct.
The portal vein is now separated from the pancreas. There are always a number of small branches running from this vein into the head of the pancreas. Each is dissected free, clipped and divided.
The jejunum, just beyond the ligament of Treitz is divided and the vessels going to the distal duodenum are divided, mostly using my trusty “Ligasure” device.
Finally, the uncinate process of the pancreas is freed from the its connections posterior to the portal vein. Mark had an unusually large collateral vessel in this area which required special attention to control. The vessel was feeding into the pancreatic mass and could not be preserved. I always worry when I find, and sacrifice, an unexpected and unusually large artery; worry that it might be important. In Mark’s case I my worries were well founded.
These final maneuvers completed the resection and the specimen: head of the pancreas, duodenum and portion of bile duct was handed off to the circulator who passed it on to Pathology for immediate examination.
“Looks like adenocarcinoma of the Pancreas,” the friendly neighborhood Pathologist reported, “margins are free.”
“Thank you,” I replied and I set about the task of reconstructing the damage I had created.
At this point in the operation what I have is the divided pancreas, bile duct and duodenum. These organs are designed to meet at the Ampulla of Vater where bile and pancreatic juice are dumped into the duodenum to join and digest ingested food. The task at hand is to reconnect each structure to the intestine to restore the normal digestive function. Connecting the Pancreas to the bowel is considered the Achilles heel of the Whipple procedure.
Back to work. The divided pancreas was dissected free a bit more which allowed me to put the about three centimeters of the divided pancreas inside the small bowel. Sutures of 3-0 Prolene were used to suture the thin, weak capsule of the pancreas to the bowel. Sometimes the pancreas is more fibrotic and will hold suture fairly well. Mark’s pancreas was more normal and the sutures held, but I knew I was going to worry for a few days.
The bile duct was next, an anastomosis using 4-0 Vicryl in a single layer, straightforward and uncomplicated.
Finally, the duodenum, just beyond the stomach, was anastamosed in 2 layers to the small bowel, using 3-0 Vicryl and 3-0 silk. Drains were placed, the belly closed and he was brought to the recovery room.
Mark was very stable for the first 24 hours. There was slight tachycardia, heart rate around 105, but nothing else unusual. The two drains put out serosanguinous fluid as expected and the volume of fluid put out was low.
Labs on the first day revealed a bilirubin of seven, down from thirteen preop. His white blood cell count was high at 21k, but this is not uncommon with the stress of such a major operation. His hemoglobin level actually was higher than preop.
Hemoconcentration, very common after Whipples.
I increased his fluids and went on my way. That evening I made my usual call to the ICU to check on Mark and my other ICU patient.
“He’s more tachycardic now, heart rate 130,” his nurse reported. “Urine output is OK and oxygen saturation is 100%. He looks comfortable.
“Give him a fluid bolus, a liter of lactated ringer’s,” I ordered and I went to bed.
I didn’t hear anything more that night. I saw Mark first thing the following morning. His heart rate still was hovering around 130. His WBC was 20k, he had a mild metabolic acidosis and his renal function was a little worse.
I ordered a stat CT of the abdomen and gave him more fluid and antibiotics.
What’s wrong with Mark?
His drains weren’t putting out very much and the fluid was the expected serosanguinous. The NG tube was bile, his abdominal exam was unremarkable.
Everything looks OK, except for him. Maybe the CT will help.
The CT was finished later in the day: Post op changes, no fluid collections, nothing that would explain Mark’s persistent tachycardia and elevated WBC.
Maybe he’s just one of those patients who becomes tachycardic under the stress of surgery.
Over the years I’ve had a handful of patients whose heart rates go up for a few days after surgery, even after a relatively minor procedures. For example, one young woman had excision of a retroperitoneal node to diagnose lymphoma. She maintained a heart rate in the 120’s for five days afterwards. The resident staff, including me, scratched our collective heads and searched for an etiology. None was ever found and her heart rate eventually became normal and she went home.
Unfortunately, this was not the case for Mark.
The next day I noticed that one of his abdominal drains now was putting out bile. Even worse, Mark was tachypneic, respiratory of 36.
He’s septic. I need to take him back to surgery.
I expected that he had a leak from the pancreatic anastomosis. What I found was very unexpected.
There was bilious fluid in the abdomen, as I anticipated, but the some of the bowel was dead. Specifically, the loop of small bowel which I had used to reconstruct the GI tract was necrotic, about forty centimeters. The small bowel beyond the duodeno-jejunostomy was pink and viable. In addition the right colon didn’t look right.
I began by taking down each anastomosis. The pancreas, bile duct, stomach and duodenum all looked OK, well perfused and essentially healthy. I resected the gangrenous bowel and redid each anastomosis. The pancreas was even more difficult to handle, due to the surrounding inflammation and edema.
Maybe I should ligate the pancreatic duct and not do a pancreatic anstamosis? Or, maybe remove the rest of the pancreas?
There were plus and minuses on each side. Ligating the pancreatic duct likely would preserve the pancreatic endocrine function and prevent Mark from becoming a very brittle diabetic, but would also likely lead to pancreatitis which could be severe. I wasn’t sure Mark would survive a bout of rip roaring pancreatitis.
Removing the pancreas eliminated the risk of pancreatitis, but would leave Mark with pancreatic endocrine insufficiency, which meant diabetes which could be difficult to manage as well as the need to take supplemental pancreatic enzymes for the rest of his life.
In the end I redid the pancreatic anastomosis, wrapped it with omentum and put drains all around it. The bile duct and duodenal anastomosis were straightforward.
I looked at the right colon again.
It looks dead.
Fifteen minutes later this part of the colon was sitting in a bucket and I was creating an ileostomy.
Finally finished, Mark was deposited in the ICU where his overall condition improved immediately. His heart rate came down, his bilirubin came down, he maintained his blood pressure, renal function remained stable and the drains were only putting out serous fluid.
Still, a question nagged at me.
Why did Mark develop this devastating complication?
I wish I knew. I’ve postulated that the unusually large vessel which was supplying the pancreas may have also been the major arterial supply to the proximal small bowel. The colon gangrene I chalked up to low flow and vasoconstriction, although it is possible that sacrificing that vessel also affected the colon.
Perhaps the middle colic artery, which is the major blood supply to the right colon took an anomalous course adjacent to the pancreas. Maybe the mesentery twisted causing the gangrene, although it didn’t look this way at the second operation.
I never found out for sure.
I had planned to return Mark to surgery 48 hours later, to examine the bowel for further ischemia and necrosis, but his overall condition so dramatically improved that I cancelled this procedure.
I also wish I could say that Mark made a 100% complete and uncomplicated recovery, but this was not the case. He improved for about ten days. But, on the eleventh day he started putting out more fluid from one of the drains adjacent to the pancreatic anastomosis.
He required prolonged support on a ventilator, dialysis for a time and TPN. He had this fistula for about three months before it finally healed and he was able to get on with his life.
His pathology revealed adenocarcinoma of the pancreas, no spread to any lymph nodes.
I hope he is cured.