How Death Becomes Life

by Joshua Mezrich

  While in general it has been hard to show a significant benefit to pancreas transplant in terms of survival (unlike with a kidney transplant, which clearly prolongs life), there are no more thankful patients than the recipients of pancreas transplants. For the first time in their lives, these patients are not defined by their blood sugar. They can eat what they want, they can sleep through the night without worrying they won’t wake up, and they don’t have to call themselves diabetic anymore.

  I recently evaluated Mary J. for a pancreas transplant. A pleasant forty-five-year-old woman, she came with her four kids in tow, ages two to fourteen. She looked totally healthy, and for a minute I wasn’t sure what she was doing there. In general, we don’t perform pancreas transplants for type 1 diabetes until patients have developed kidney failure. That’s because the antirejection medications have so many side effects that the benefit of getting off insulin may be outweighed by the disadvantages of taking immunosuppressive drugs. Still, we will consider doing a pancreas transplant alone for patients who have such brittle diabetes (a severe form that involves unpredictable swings in blood sugar level and difficulty getting sugars under control) that they develop hypoglycemic unawareness (a complication in which the patient is unaware of dangerous drops in her blood sugar level, which can lead to seizures or death).

  Mary is one of those patients. She has episodes, pretty much every other day, in which she will suddenly black out. She could be at work, at the grocery store, or asleep. Once, she blacked out while driving her kids in the car. She has had to give up her job and needs help taking care of her children. Her diabetes is so hard to control that her blood sugar level either flies off the charts or bottoms out, becoming low enough to cause seizures. But unlike most diabetics, Mary can’t tell when her blood sugar is getting low. There is no warning, no sweating or light-headedness. Her glucose level rises and drops so precipitously that an insulin pump hasn’t helped, and her life has become unlivable.

  Yet all that will be fixed with a pancreas transplant. (Perhaps there is one other option for her: A recent patient of ours had a dog that could detect when her blood glucose level got too low based on her smell and taste when he licked her skin. If her level got out of whack while she was asleep, he would wake her up so she could take some sugar or insulin. We gave her a transplant, which worked great, but the dog lost his job.)

  IN THE EARLY 1900s, there was no successful treatment for type 1 diabetes. Children who developed this disease were placed on starvation diets until sugar could no longer be detected in their urine, with the hope that someone would figure out how to treat them before they died of starvation. By the turn of the twentieth century the medical community recognized that the hormone responsible for handling sugar in the blood (eventually to be called insulin) came from the pancreas, though numerous researchers failed in their attempts to isolate this protein. Finally, in January 1922, insulin was isolated and injected into a patient by the most unlikely of investigators, Frederick Banting, a surgeon with virtually no surgical practice or research background. His persistence, or stubbornness (likely compounded by a case of PTSD from his service in World War I), together with the help of a couple of real scientists, J. J. R. Macleod and James Bertram Collip, and a tireless medical student, Charles Best, at the University of Toronto, led to the isolation of insulin and its use as a life-saving treatment for thousands of children diagnosed with this disease. In the 1920s and ’30s, there was hope that insulin would return diabetic patients to normal lives (other than having to inject insulin multiple times a day). By the mid-twentieth century, patients with type 1 diabetes were living into their forties and fifties, but were now facing the horrible complications just mentioned, including kidney failure. Although the discovery of insulin prolonged the lives of these patients, the difficulty in measuring blood sugar and knowing when and how much insulin to use and the inability to tightly regulate sugars throughout the day and night were still major problems. It seemed possible that transplanting the whole pancreas could solve all this.

  There were some attempts in the 1950s to transplant pancreata in dogs, but the vast majority of transplants led to devastating complications from the leaking of highly destructive pancreatic enzymes. Various techniques were employed to prevent this, ranging from tying off the pancreatic duct (which led to scarring that ruined the organ) to bringing the duct out through the skin and collecting the enzymes in a bag. Some investigators tried radiating the organ to destroy the parenchyma, the tissue responsible for creating the digestive juices (hoping it wouldn’t also destroy the insulin-secreting islet cells), but that failed as well. Others included a cuff of duodenum, the bowel that collects the drainage of pancreatic juices as they exit the pancreatic duct, in the anastomosis rather than just using the pancreatic duct. Initially, the bowel would be attached to an adjacent loop of small bowel in the recipient, allowing these juices to flow through the intestine (which is where they go normally).

  By the late 1960s, some experiments in dogs were showing a survival rate of half a year or more, good enough for a few investigators to attempt the procedure in humans. On December 17, 1966, a team at the University of Minnesota that included Drs. William Kelly and Richard Lillehei (the younger brother of C. Walt Lillehei) performed the first pancreas transplant in a human, a twenty-eight-year-old woman, along with a kidney transplant to cure the renal disease caused by her diabetes. In this operation they opted not to include the donor duodenum but, rather, ligated the pancreatic duct. The blood vessels to the pancreas were anastomosed to the iliac vessels, a technique similar to that used for kidney transplantation. The pancreas worked right away, allowing the patient to come off insulin and have normal blood sugar levels regardless of what she ate. Sadly, she died after two months, having suffered rejection of the organ and, ultimately, infection.

  Heartened by the mild success of this transplant, the team performed thirteen more pancreas transplants over the next seven years. Nine of them included a kidney as well, and the majority was transplanted with a cuff of the duodenum sewn onto adjacent small bowel. Although one graft survived for a year, the rest were lost in a much shorter time, with the patients dying secondary to infection or renal failure. It is fair to say pancreatic transplantation was not accepted as a viable option in those days, given the success of insulin and the reasonable-ish outcomes of transplanting kidneys alone in diabetic patients. A few researchers soldiered on, though, led primarily by the Minnesota team. By 1977, fifty-seven pancreas transplants had been performed with only one long-term survivor. Many of these grafts were lost because of leaking at the site where the duodenum was sewn to the bowel, causing surgeons to remove this cuff of duodenum and transplant the pancreas alone. Outcomes for pancreas transplant didn’t really start to improve until the advent of cyclosporine in the 1980s, which dramatically reduced organ rejection.

  Pancreatic transplant is unique among organ transplants. Unlike the heart, liver, and kidney, all of which are necessary for the patient’s survival, the pancreas primarily provides a quality-of-life benefit. The patients could otherwise survive on insulin, albeit as slaves to their own blood sugar. After witnessing a few horrible complications after pancreas transplant (requiring multiple returns to the OR, leaking organs and drains, and open wounds), I wondered if it was worth it. Wouldn’t it be better to use an insulin pump and get a kidney if needed?

  I thought this, that is, until I got to know some pancreas recipients and heard their stories. While there are many successful ones, which I could share, sometimes the most complicated cases are the most revealing.

  El Diablo (or Disconnected

  Pancreatic Duct Syndrome)

  I had my first encounter with El Diablo about eight years ago. I was on my way to my lab when my phone rang. On the screen was the name of one of my partners, Janet.

  “Hey, I’m getting killed today,” she said. “Think you can do a quick ex lap? It’s some previous transplant with free air, probably a colon perf or duodenal ulcer.”
>
  I pushed aside the multiple excuses that entered my brain, ranging from the lab work I wanted to do to a dread illness I was developing. “Sure, no problem.”

  Twelve hours later, I was still in the OR. Quick, my ass. The patient was many years out from a combined kidney-pancreas transplant. He had recently been treated for rejection, and had recovered from this, with the pancreas still functioning, although not as well as before. His blood sugars were borderline, and he was close to needing insulin. His kidney wasn’t working that well, either—he still wasn’t on dialysis, but he was not far away.

  Earlier in the week “JB” was at home working on his roof when he fell off the ladder and hit his belly. A few days passed, and the pain that had started out mild had increased in severity until he finally made his way to the hospital. A CT showed free air, indicating a perforation in a portion of his bowel, usually a clear indication of the need for emergency surgery.

  When we got in, we found adhesions everywhere—that is, scar tissue most likely formed from his previous surgery that had caused all of JB’s bowel to stick together. We spent hours dissecting out the various tissues, taking care not to injure any of the bowel as we went along. After many hours of this, we deduced that there was no hole in the stomach, duodenum, or colon. When we got to the transplanted pancreas, however, we found a hole right where the duodenum had been sewn onto the small bowel during his years-old kidney-pancreas transplant. The hole was leaking pancreatic and small bowel contents (succus—we generally use the term succus, or succus entericus, to describe small bowel contents) everywhere. Given the amount of spillage and how sick the patient was, I knew I had to take out JB’s pancreas. It wasn’t working that well anyway, and clearly if I left it in, it was going to kill him. I would need to resect the portion of bowel onto which the duodenal cuff had been sewn and then hook that bowel back together. That should be easy. I would also need to find the artery and vein that supplied the organ, which had been sewn onto the iliac vessels going down to his right leg. Finally, I would need to scoop out the body and tail of the pancreas.

  Digging out the blood vessels proved to be nearly impossible. Every thing was so stuck that it was like carving through a piece of hard marble. That’s the thing about scar tissue. You never know how thick it will be, how difficult to take down, until you get in there. Usually transplant patients have less scar tissue than other patients who have undergone surgery—the immunosuppression actually suppresses its development. But not always. Eventually I carved my way across the vessels and divided them.

  After reconnecting his bowel, I tried to peel out the pancreas. It was so stuck and surrounded by so much inflammation that, to be honest, I couldn’t tell what was pancreas and what was surrounding tissue that needed to stay in there. Once I had removed a handful of yellow tissue, I placed a drain and closed up. After talking to JB’s family, I made my way home at around midnight.

  When I spoke to JB the next day, he hadn’t been told yet that I had removed his pancreas. He had been too sedated after surgery, so I figured I’d let him know when I saw him on my rounds. He was devastated. I told him we’d had no choice, that we’d had to take the organ out to save his life.

  “Well, how long until I can get another one?” he asked.

  I told him it was too early to talk about that. “Let’s just focus on getting through this for now.” Before I left, I looked at the drain I’d put in him. The bag had pinkish fluid in it, and I noticed that the output was pretty high—almost half a liter. Hmm. Should dry up . . .

  The next morning at around 6:00, I flipped through JB’s chart. His labs looked okay, but I again noticed that he had drained almost another liter of fluid overnight. Could I have left some of that pancreas in there? I texted my fellow, John, and asked him to have the fluid tested for the presence of amylase, an enzyme made by the pancreas. By 10:00 a.m. I saw the result: well over 1,000 (way too high; the level in normal body fluid should be under 100). There clearly was some pancreas still in there that was leaking amylase. I figured the leak would dry up; after all, the damn organ didn’t have any blood supply, it had to die. But that zombie pancreas just wouldn’t. The next few days, JB kept pouring out pancreatic juice. I thought about taking him back to the OR, and even talked to him about it.

  “I knew you shouldn’t have taken my pancreas out,” he said.

  While I was still agonizing over what to do, I got a text from John, my fellow. “You’ll never believe what’s coming out of JB’s drain now. Succus.” JB was leaking from where I’d sewn the bowel back together. And why wouldn’t he be? His anastomosis had been bathed in pancreatic juices for three days.

  This trip to the OR went only marginally better than the last one. It seemed as if a bomb had gone off in JB’s body since the last time I was in there. I carved my way to the small bowel and was able to take down that anastomosis and put it back together. I then spent at least an hour trying to find and remove any other tissue that looked like pancreas. I called in a partner who had years of experience operating on this beast of an organ, and together we peeled out even more tissue. When I couldn’t imagine anything possibly being left, I closed JB up, leaving a drain.

  The next morning I looked at the drain—and it looked back at me, laughing. There was the same fluid, in the same quantity. I sent JB down to get a CT scan, and sure enough, I spotted the culprit: a little nubbin of pancreas, smaller than a golf ball, had been left behind. When I showed the scan to my partners, Dr. Sollinger quietly whispered, “El Diablo.”

  The name had been introduced to Sollinger at a conference he attended in South America, to describe scenarios in which the pancreatic duct gets fully disrupted and disconnects a portion of the pancreas from the rest of the organ, causing a remnant of pancreas to continuously leak pancreatic juices. Sollinger told me I would be back in the OR.

  I had decided to wait it out for a few more days when I got another text from John. “Guess what’s in the drain now,” he said.

  I could only imagine.

  “Shit,” he said. “There’s shit in the drain.”

  Back to the OR, take three. This time El Diablo had spewed its venomous juices through the wall of the colon. We removed a large piece of colon and brought up an ostomy. JB would wear a bag now. We carved out more tissue that looked like pancreas, and I again had a partner come in and help. Sure that it was all gone now, I left a drain. When we went to close, we noted the poor condition of JB’s fascia, the strength layer around the muscles that keeps surgical wounds intact; from all the operations and the leaking stool and toxic juices, it had been eaten away. He was definitely going to develop a wound problem.

  Next day: bedside of JB, drain, El Diablo. I went back to my office and put on the Rolling Stones’ “Sympathy for the Devil.” I listened to the words, feeling more and more helpless.

  Pleased to meet you,

  Hope you guess my name.

  But what’s puzzling you

  Is the nature of my game.

  How the hell was I going to solve this?

  A few days later: back in the OR with JB and more leaking bodily fluids. This time, I couldn’t even get into his belly. Everything was stuck together. Eventually, we cored a small tunnel into the space where I thought the remaining pancreas might be. I was nearly crazed, pulling out tissue to send to Pathology and thinking, Die, you fucking zombie, die! By the third sample, the tissue finally came back positive for pancreas. We continued to send tissue to Pathology until it came back negative again. It just had to be gone now. And it was.

  The next day the drain showed no output, which was good news. But his kidney had failed, and he was back on dialysis. He had bedsores and couldn’t walk. Still, slowly, over months, he healed. He finally made his way to a nursing home. While there, he developed a nonhealing ulcer on one of his feet, which led to a below-the-knee amputation. Yet, somehow, he got through all this and made it home. Although his kidney function was marginal, he was weaned off dialysis. Six months later, I got word
that he wanted to see me in the clinic. When he showed up, he was barely recognizable. He was as thin as a rail, looked about ninety years old (he was sixty), and was missing a leg, but he was smiling. He thanked me for saving his life.

  Then he asked me, “When can I get another pancreas?”

  That devil had nearly killed him. How could he entertain another? I told him he really wasn’t a candidate (he was too old and had been through too much), and no one would want to operate on him.

  He told me he wasn’t surprised, but then said, “I wish you’d left it in. I don’t want to be diabetic again. I would rather have died.” And he did, about six months later.

  You will never meet a more grateful patient than one who has received a new pancreas and now no longer has type 1 diabetes. No matter what complications that damn devil may cause, patients almost always want another one. Sure, the kidney is the real lifesaver; it gets patients off dialysis, a living hell. And of course, kidney patients are grateful, too. But there is something special about the pancreas. It truly is a bedeviling organ.

  The Future of Pancreas Transplant

  Outcomes after pancreas transplant have improved dramatically, thanks to better immunosuppression and the commitment and persistence of a handful of investigators. Both the University of Minnesota and my own program at the University of Wisconsin have performed more than a thousand such transplants each. Graft survival at one-year rivals that for kidneys, and we have many patients who are twenty years post-op and doing great.

  That said, this is one area where technology undoubtedly will surpass the challenges and benefits of organ transplant. The first-ever closed-loop pump system, which monitors a patient’s blood sugar every five minutes and then responds to low levels by lowering the basal rate of insulin, was recently approved. While this device’s sensor is subcutaneous (that is, under the skin), I’ll bet that in the near future there will be devices that can monitor glucose continuously through transdermal sensors (or some other noninvasive way, such as through contact lenses, a strategy that is being developed). Eventually, these devices will be able to constantly adjust someone’s insulin dose to maintain near-perfect glucose levels throughout the day. I presume they will be controlled by smartphones, so perhaps if a patient wants to eat a piece of cake, he can take a picture of the cake with his phone, which will then calculate the needed bolus of insulin and deliver it as the cake is being eaten. Whether that will prevent the long-term complications associated with diabetes will take years to determine.