New Hi-Tech Images Guide Surgeons' Hands
There are some worries surgeons don't share with patients before an operation. That they are going into the body blind, while carrying a sharp knife, is one of them. "It's kind of like a labyrinth. You can only see right in front of you, but not around the next bend," says Alexandra Golby, a neurosurgeon at Brigham and Women's Hospital in Boston. Adds Christopher Moir, a pediatric surgeon at the Mayo Clinic in Rochester, Minn., "You hope the structures look like what you've seen before, but you really don't know."
How can this be, in an era where high-tech scans with fancy names like computed tomography and magnetic resonance imaging give unprecedented views of the inner human body? The big reason, surgeons say, is that most scans omit crucial information. MRI and CT images that show the shape of body parts, for instance, don't capture unhealthy electrical activity or blood flow there; the section of a heart causing a dangerous rhythm, for instance, can look just like normal heart muscle. And because most scans are done before surgery, they don't provide the real-time detail that would allow Golby, say, to shift a cut left or right by a millimeter or two–a move that can make the difference between safely removing a brain tumor or stealing a patient's speech.
Now, however, imaging is getting a whole new look. New and powerful computer programs can meld detailed images of anatomical structure with vibrant views of the structures in action, making the invisible visible–and treatable. "Matching these images to one another accurately is the key. It's making a lot of impossible things possible," says Richard Robb, a pioneer in the field who runs Mayo's biomedical imaging lab. Robb has developed scans that reveal abnormal spots of electrical activity causing epileptic seizures in the brain, allowing surgeons to zero in on those points and remove them from patients for whom, previously, surgery would have posed too great a risk. In experiments, combining multiple scans into one easy-to-read image has pinpointed deadly, rapid heart rhythms, too–and may turn a dangerous six-hour repair procedure, much of it spent poking around looking for the abnormal activity, into a relatively simple one- to two-hour job.
Doctors have also moved MRI from pre-op to the surgical table, where real time scanning gives ongoing views as the operation progresses. Already, the technique is making brain surgery and prostate cancer surgery more precise and leading to better treatment. "This really is the dawn of a new surgical era," says Michael Schulder, a neurosurgeon at New Jersey Medical School University Hospital in Newark who uses MRI during tumor surgery. "We're taking a lot of the guesswork away."
An Epileptic's Story
Michael Hutton certainly hoped for better than guesswork when surgeons cut into his brain in 2004 to battle his disabling epileptic seizures. "I told the doctors that I didn't have a lot of extra brain," says the 45-year-old from Chippewa Falls, Wis. "So I couldn't afford for them to take out the wrong spot."
He couldn't afford to skip surgery, either. Hutton, who installs insulation in houses, had taken a bad tumble off a roof during a seizure and smacked his head. "I just kind of zoned out," he says. Nothing was broken, that time. But he was having 70 seizures a year, sometimes four in a day, and epilepsy medications weren't helping.
But finding the exact spot to cut out was a problem. Bursts of odd electrical activity can be hard to locate using conventional techniques like electrodes on the scalp, especially if the seizure is deeper in the brain as Hutton's was. A scalp electroencephalogram often isn't powerful enough to reveal deep misfires, says Elson So, a neurologist and epilepsy specialist at Mayo, where Hutton was a patient. Surgeons can cut off the top of the skull and lay electrodes all over the brain surface, but exposing so much brain often leads to excess bleeding or a serious infection.
Instead, Hutton benefited from an imaging technique pioneered by So and Robb. "We injected him with a kind of 'radioactive ink' that tags the area of the seizure," says So. The ink, a molecule about as radioactive as a chest X-ray, gloms on to flowing blood. Seizure regions are very active, attracting a lot of blood and thus ink. A scan sensitive to this ink, called SPECT for "single photon emission computed tomography," showed a big clump in Hutton's right temporal lobe. (Positron emission tomography, a related type of scan often used by doctors, is a bit too slow to capture seizure location.)
Next, a computer program lays this image over an MRI, which is better at showing brain anatomy. "Once you get the two images to line up, you have this hot spot located on a map of the brain that a surgeon can use," says Robb. Mayo surgeons opened a small hole in Hutton's skull instead of a large one, and removed an area "smaller than the tip of my thumb," Hutton recalls.
The treatment worked. Today, Hutton is back at his job, has regained the driver's license he'd surrendered, is no longer a dangerous driver, and is seizure-free on a low dose of just one medication. "I even got back into my favorite hobby," he says. "Sky-diving."
The Mayo system, called SISCOM and now used at many hospitals across the country, can help 70 percent of patients like Hutton with temporal lobe epilepsy, estimates So. "It makes surgery a safe option for a lot more people. Without it, and if they don't respond to meds, only about 5 percent of them will ever be able to drive again," he says. Plus, adds Bruce Fisch, director of the epilepsy center at Louisiana State University Health Sciences Center in New Orleans, the pinpoint accuracy "allows you to stay away from important functional areas of the brain" when deciding where to cut.
Carving Out Cancers
One of the major challenges in brain surgery is that the good bits can look a lot like the bad parts. When a tumor has insinuated itself around nerve fibers, removing it without damaging healthy tissue becomes very difficult. "You can't really see bundles of nerve fibers. The surface of the brain all kind of looks like cooked egg whites," says Brigham's Golby. Even an MRI scan often doesn't distinguish a low-grade tumor from its surroundings–which may control speech and vision.
So Golby and her colleagues have taught MRI a new way of seeing. The trick is to combine three different types of MRI scan into one image. First, they use a technique known as "functional MRI" to identify brain regions linked to vital abilities; fMRI maps blood flow in the brain as patients perform such tasks as pointing to an object or reading. Areas that show increased blood flow are the ones responsible for those abilities.
Next, Golby marries the fMRI image to another MRI scan, tweaked by powerful computer programs to reveal the nerve fibers. Water molecules move in particular patterns along such fibers; the computer can pick out those patterns in the scan and turn them into an image outlining the fibers. Finally, a detailed MRI showing brain anatomy is fused with the other two. "We can visualize functional areas, along with the nerve tracts and the cancer, see them all in one image," Golby says–and then plan a safe way to remove the tumor before making a single cut.
Once an operation begins, neurosurgeons have to cope with a second big hurdle: "brain shift." The brain moves slightly in the skull during surgery, so landmarks identifying healthy brain tissue in even the most detailed pre-op scan end up a centimeter or so away from where they started. At Brigham, surgeons have responded by moving an MRI scanner into the operating room. Unlike the long tube dreaded by the typical MRI candidate, the OR version features two large, circular magnets, one at the head of the surgical table and the other surrounding its middle; Brigham doctors call it "the double donut." They stand between the magnets to operate. Scans made as the operation progresses capture any changes and project them on a monitor, Golby says. Recent advances have produced a smaller, more portable version of the scanner that's now used in other hospitals.
With MRI capability in the operating room, surgeons also can check to see that they got all of the cancer before closing up the patient, reducing the chances of recurrence and the need for more surgery. "You really don't want to go back in," says radiologist Ferenc Jolesz, director of the National Center for Image Guided Therapy, based at Brigham. In a 2003 study, this kind of MRI found residues of brain cancer in 36 percent of patients while they were still on the table, which surgeons were then able to remove. In another 31 percent, imaging revealed the cancer had been fully removed, avoiding the need for addition probing and cutting of the brain.
The Brigham MRI is being used to guide cancer therapy at the other end of the body, too. For many patients with early-stage, localized prostate tumors, brachytherapy–implanting 50 to 100 or so tiny radioactive seeds in the prostate, using needles–is very effective. Usually, however, the location of these implants is guided by ultrasound images, which can give a muddy view of the prostate. That's a weakness: In addition to targeting and curing the cancer, radiologists want to avoid damaging nerves and other tissues, which can cause impotence and incontinence. By taking a finely detailed image as each needle is inserted, urologists and radiologists can make sure it is going into the right place.
That approach was very appealing to Charlie Kireker, 56, who had seeds implanted at Brigham last June. "If it's good enough for brain surgery, it's good enough for me," says the 56-year-old investment fund manager, who lives near Middlebury, Vt. Kireker got needled 23 times, with five seeds planted per stick. By January of this year, his score on the prostate-specific antigen blood test for cancer, which had shot up from less than 1 to 2.75 in the months before the procedure–speedy climbs like that are often signs that the cancer is growing–was back down to a negligible 0.63. "And I've had no physical problems," he says. "Everything works."
At this point, it's hard to say that using MRI in this way leads to more cures, cautions Ashutosh Tewari, a urologist and prostate cancer specialist at Weill Cornell Medical Center in New York City. "It's more precise than ultrasound, but that doesn't necessarily translate to better cancer care. I haven't seen any data on that," he says. (Brigham and Women's is the only institution currently using the technique.) Clare Tempany, a radiologist at Brigham, agrees, saying that because prostate cancer grows so slowly, doctors have to wait at least 10 years for real outcome data, and their first MRI patient won't reach the 10-year mark until this fall. The data on incontinence and impotence, however, are already becoming clear. When compared with ultrasound patients in a 2004 study, MRI patients fared much better in terms of side effects.
Other hospitals are not rushing to install the massive–and massively expensive–double donut for prostate surgery, because of reservations like Tewari's and because the cramped space limits a doctor's ability to move around for other operations. But for brain surgery, a few dozen, including the Cleveland Clinic, Children's National Medical Center in Washington, D.C., and New Jersey Medical School University Hospital in Newark, have bought the smaller, portable MRI, which brackets a patient's head but allows doctors more space. The $1 million to $2 million price tag is "a big upfront cost," says New Jersey's Schulder. "But we did a cost analysis and found there were significant savings because the patients had shorter hospital stays."
Beating Bad Heart Rhythms
At Mayo, Robb is now working toward shorter stays on the operating table for heart patients. Pushing beyond single-frame snapshots taken during surgery, he's experimenting with movies of the heart that combine several different types of scans to show detailed anatomical shapes as well as bursts of electricity while a doctor operates. "This is huge for treating bad heart rhythms," says John Haller, program director for image-guided interventions at the National Institute of Biomedical Imaging and Bioengineering, part of the National Institutes of Health. "These operations now take six or eight hours and someone could die on the table. Robb's technique could cut it down to just one or two hours."
That would come as good news to many of the country's 2 million people plagued by bad rhythm atrial fibrillation, which causes the upper chambers of the heart to flutter 300 or 400 times per minute. (A normal heart rate is between 60 and 100.) Those rapid flutters produce blood clots and strokes. To fix them, doctors snake several tiny catheters through blood vessels into the heart. Some of the catheters have an electrode at the end to pick up errant electrical signals. Currently, doctors track the electrodes with fluoroscopy, a kind of moving X-ray. When they suspect an area of being bad, they zap it with a tiny energy burst, killing the abnormal muscle sections. Or killing normal sections. If the heart rhythms don't return to normal, the electrodes were in the wrong spot. Doctors move them and try again.
"It's trial and error, trial and error," says Douglas Packer, a cardiac electrophysiologist at Mayo. "I'm standing there moving these catheters with my hands, and I'm looking at one screen that shows their position on a murky two-dimensional fluoroscopy image, and another that shows some details of the heart" using ultrasound. He's also building up a map, on another screen, of points that represent different aspects of the heart's electrical activity. "I have to combine all of this information in my head and hold it there. So it's hard to find your way around. I'm pretty good at it, but it takes hours to find the trouble spots and treat them."
Working with Packer, Robb has devised a system to show all of this complicated stuff on one computer screen. Three-dimensional ultrasound captures the movement of the heart while fluoroscopy or another technology tracks the electrodes, which, along with a standard electrocardiogram, provide an ongoing stream of data about electrical activity. This information is all mapped onto a detailed anatomic image, taken by computed tomography or CT, that changes as the heart beats–and can be manipulated on the fly to show different parts of the heart from different angles.
"The result is like actually being in the heart," Packer says. "I see things as they happen, and where they happen." In animal tests, and computerized re-creations of procedures using actual patient data, the technique has cut errors and hours way down. Later this year, Robb and Packer intend to start a trial on actual heart patients. The moment will be a long way from the one about 30 years ago, Robb recalls, "when a surgeon said to me, 'If I can see it, I can fix it.'" These new views eliminate a lot of blind spots, and for patients that means more and better repairs.