Science / Medicine

Dr. Michael Harrison is co-director of the 10-year old Fetal Treatment Program at the University of California, San Francisco. With his colleagues at UCSF, Harrison, a pediatric surgeon, pioneered techniques that allow surgeons to operate on unborn babies, to correct potentially lethal problems too serious to wait until birth. Since 1982, close to a dozen patients have undergone open fetal surgery, all at UCSF. In an interview last month with Marcia Barinaga, Harrison, 44, reflected on the developments that made such surgery possible, and the ways in which it can help otherwise doomed fetuses.

The thrust of most prenatal diagnosis in the past, before 10 years ago, was basically a search-and-destroy mission, because you were looking for bad things in order to counsel the family. If they wanted, they would choose to terminate the pregnancy. That was the only real option.

Now we’re really getting to where we can do a service for families.

The fetus used to be just a hidden little rascal in there. No one knew anything about him really, except that you could listen to him a little bit. Now he’s a patient.

There were two thrusts that came together and brought this to fruition. The technology was just developing 10 years ago so they could really for the first time see problems in the fetus.

Technical things were coming together: The sonographers were seeing the fetus; the genetics folks could make some genetic diagnoses, and the whole area of prenatal diagnosis opened up. That’s one-half of the equation.

The other half came from folks like me and my colleagues, who really just looked at the babies after birth. We were seeing babies who had some very frustrating problems, and the frustration was that they would often die. And when you began looking at what was happening to these little babies, it was clear that we were there too late in the process. It was like the cancer that’s already metastasized, the heart attack that’s already gone on and can’t be fixed. You say, ‘Oh my God, if I could have been there just a little bit earlier, we could have saved this patient.’

The classic ones, the first ones that we saw, were rather simple-minded ones--like diaphragmatic hernia. Most babies we see as surgeons after birth with diaphragmatic hernia die. Originally, we thought they die because they have a hole in their diaphragm and all their intestines are in their chest. But that’s just not the case, because we learned to fix that in newborns long ago. The babies die because their lungs haven’t grown. And that issue was settled before they were born.

It became relatively clear quickly that the only real way to save the bad diaphragmatic hernias would be to take this lump of intestine out of the chest before birth. Then, if the lung would grow, by the time he comes out and takes his first breath, he’ll be all right.

The problem was, we didn’t know whether it was true. Obviously you couldn’t answer that in a human being. We had to go to animals.

But diaphragmatic hernia is only one of a spectrum of diseases that we began to recognize in newborns that might make sense to fix before birth. The second one was obstructed urinary tracts. The fetus can’t urinate, and so he doesn’t fill up the fluid around him. When the fluid goes away, his lungs are small, just like a diaphragmatic hernia, and, in addition, it backs up the urine into his kidneys and wrecks his kidneys.

One of the things that we’ve studied over all this period of time is the effect of decreased amniotic fluid on lung growth, and the effect of fetal breathing movements on lung growth.

Now we were beginning to see the light at the end of the tunnel. There are sick kids that may need to be fixed, and we had better figure out which things truly make sense to fix and which don’t. We showed in that disease that blocking the urinary tract makes small lungs and makes wrong kidneys, and releasing the obstruction early enough ameliorates those bad effects on the kidneys and lungs. The thinking was rather simple, and basically correct.

Another aspect of the story is the human side. Here we were trying to sort out which things to go after, and at the same time we were starting to see many human fetuses with these problems who were referred here from all over the United States.

Women would have an ultrasound for one reason or another, and by serendipity really, something abnormal would be found in the fetus. With diaphragmatic hernia, you can see the stomach bubble on the sonogram. It’s not in the right place--it’s inside the chest, when it’s supposed to be in the belly.

So now we were seeing the diseases in human fetuses, and we could ask the question: What happens if you leave them alone? You would never think about intervening even if you’ve done the experimental work unless you knew the natural history of the disease, if left alone. So here was a whole bunch of years of work--documenting the natural history of fetal diseases.

We’re not the only ones doing this. People all over the world are doing it, and we’re constantly comparing notes and getting together, in an international group once a year. Hundreds of fetuses have been observed, without intervention, with all these different problems.

Now we have two parts of the picture in place--we have knowledge of the pathophysiology, from experiments in animals, and we have knowledge of the natural history of the human disease. There were a couple more pieces that had to fall in place before we could think about doing anything.

The next piece is the terrible dilemma of the mother. The things you might consider doing to help her fetus usually involve some risk to the mother. It’s hard to justify putting her at risk when she herself doesn’t have the problem.

Before we could consider any intervention, we had to study the risk to the mom. Obviously we shouldn’t do it by just trying it on a human and seeing what happens.

That involved another huge block of work. We’ve now done well over 100 fetal operations on monkeys. There are many aspects to the technical business of opening the uterus, removing the fetus part way, monitoring him, anaesthesia for the mother, anaesthesia for the baby, how to give him blood, how to do all the things you need to do, to do surgery on somebody. It’s not quite as easy when he’s inside someone else.

One thing that really is worrisome is that by making an incision in the uterus and working on it that the uterus can go into labor and you lose the game because the kid is delivered early. It’s a constant struggle. There’s not a magic drug that stops labor.

Another purely surgical aspect that was very challenging--and it wasn’t clear when we started that we could do it--was just the simple business of opening a mid-gestation pregnant uterus without having the mom bleed to death. This is the bloodiest organ you’ll ever see. It’s quite incredible. By the time you deliver and have a Caesarian section, your uterus is thinned out, and its ready to give up anyway. But in mid-gestation, it’s just a big, thick, extremely vascular muscle. Those are the things you need to work out in the animal model.

It is absolutely crystal clear that none of this would be possible without the studies in sheep and rabbits and monkeys. There were many, many technical details that got worked out in this series of experiments. And we did eventually show that we could do what we intended to do.

That was the final piece, and when that fell into place, we were ready to offer this to moms with very unusual problems--they’re all rare. And so we began.

‘So now we were seeing the diseases in human fetuses, and we could ask the question: What happens if you leave them alone?’

The crucial issue for all this fetal treatment is selection of the right fetus. The diseases all come in a spectrum. Take the urinary tract example. There are kids who are very badly affected, and even if you could relieve the obstruction, you wouldn’t reverse all the bad effects, and they’ll be bad kids with small lungs, and die. Or perhaps even worse, they will have bad kidneys and have a miserable life. You clearly don’t want to do anything to those; that’s taking a risk for no gain. Probably the worst thing that many people fear for their children is that they’ve converted a situation where the baby would have died to a situation where he’s alive but has a miserable life.

On the other end of the spectrum you have the kids who are very mildly affected, who will do very well with standard treatment after birth. Now those kids you don’t want to intervene on, because they don’t need it. And that’s a lot of kids.

It’s only the kid in the middle, who is so badly damaged that he wouldn’t live without intervention in utero, but not so badly damaged that he can’t recover, if you do your procedure. That’s the kid you want to go after.

Probably our biggest contribution has been in selection of fetuses. We can do the tests and counsel the families much more accurately than we ever could in the past about what to expect. Is your child going to live or die? If he lives, is he going to be a renal cripple? And that’s been a very big contribution to a lot of families--I think even a bigger contribution than the treatment, because the treatment’s only for a very rare individual.

The people who are candidates for any kind of intervention are a small subgroup of those people who are biologically right. One of them will be just as happy to abort and start over, the next one will be unwilling to consider that kind of risk for this pregnancy. Another may not want a child anyway. There are lots of things that enter into it.

Our total experience with open fetal surgery is getting close to a dozen now, and it’s all good from the mom’s point of view. From the kid’s point of view it all depends on his disease, of course. The urinary tract obstruction cases have been the best stories. Diaphragmatic hernia is about an order of magnitude more difficult technically and physiologically for the fetus. We’re just in the stage of embarking on that.

The first cases were all the urinary tract ones, because it’s the thing that fell in place earliest. The first thing was to do a very simple maneuver where you insert a needle through the mother’s abdomen to place a little catheter into the fetal bladder and essentially bypass the obstruction, and let the urine drain into the amniotic fluid. That’s the one thing that’s been replicated pretty much around the world. Probably hundreds have been done with catheters.

The open fetal surgery story is much easier to talk about, because we’re the only ones that I’m aware of who do this. We’ve done five for urinary tract obstruction, and we’re very happy with it. You make an opening into the bladder, that comes right out the lower abdomen, below the belly button. It’s frequently done after birth too, that’s where we got it. I think it’s really the only way to save severely affected kids who you find early in gestation, because the catheters aren’t good enough to last for three to four months. They get pulled out, pulled in, plugged. They’re just little tubes. The catheter is sort of temporary, and this way it is permanent--until birth, and then you fix it.

The outcome for the moms has been probably the happiest part of the whole story--the moms have done well. The first few have gone on and had other children.

It certainly is a wonderful feeling of accomplishment, to be able now, with regularity--a couple in the last few weeks--to take the kid out, do something to him, put him back in, and have him be happy. And have the mom be happy. It’s just the start, but it’s what makes the future.

There will be further development of open fetal surgical techniques. There will be some nifty things we can do, like take out some tumors and things like that.

There is a great big tumor coming out of the tailbone, called a sacrococcygeal teratoma. It is very clear now that we can pick out the babies with this tumor who are going to die in mid-gestation. They die of heart failure. The tumor is about as big as the kid, and it takes up so much blood that the kid has to pump so much blood that his heart gives out and he dies. It would seem that if you just cut off the blood supply to the tumor, you could bail those kids out. That remains to be proven, but I think that it’s a very good candidate.

There are probably many things that can be done, but only if you can actually expose the fetus and do the work. Now you can do that so you can develop all kinds of nifty things, ways to help fetuses. It’s very much like developing a heart-lung machine for heart surgery. It’s the enabling step.