July/August 2008

A Messy Art

Continued from page 6

By Katrina S. Firlik

For example, a relatively recent advance, popularized within the past 10 years or so, is the programmable valve. In years past, shunts came in three basic flavors: low, medium, and high pressure. If a surgeon decided after placing a shunt that it needed to drain either more or less cerebrospinal fluid, the only option was to surgically remove the old valve from underneath the scalp and insert a different one. This is accomplished by cutting the tubing at both ends of the old valve, inserting little metal connectors into the tubing left behind, and patching in the new valve by cinching the new tubing onto the connectors with suture. Not so elegant. Given the prospect of sending an elderly patient back to the operating room, most surgeons have a fairly high threshold for going forward with a valve change.

Programmable valves largely did away with those return trips to the OR: the pressure setting of the valve can be changed in the office, noninvasively and painlessly, with a magnetic device. But that process introduces a fiddle factor of its own. For one thing, the setting can be tweaked almost endlessly, in 10-millimeter increments. Deciding when, how often, and how much to change a shunt setting is a messy art. Overdrainage can make fluid pressure drop too low, causing headaches; underdrainage can leave the original symptoms poorly controlled. Sometimes the surgeon never quite finds a patient's sweet spot. Some patients return to the office over and over again, hoping to find relief for every symptom, even unrelated ones. And sometimes the family disagrees with the patient; then the surgeon has to pick which party to please.

That's not the only problem with programmable valves. In at least one popular brand of shunt, the powerful magnetic force of an MRI scan can change the valve's pressure setting inadvertently. (This was not a problem with the traditional, nonprogrammable shunts.) And these days, MRI scans are ordered at the drop of a hat. Let's say that a patient with a programmable shunt develops a hip problem, and her orthopedic surgeon orders an MRI. There are a couple of potential pitfalls here. One is that the patient (especially an NPH patient with memory problems) may forget to tell her neurosurgeon about the scan. Furthermore, the radiologist may not realize that the shunt is programmable or that an MRI can change the setting. I have seen patients whose settings had been off-kilter for more than a year following an MRI scan.

Ideally, when a patient with a programmable shunt susceptible to this problem undergoes an MRI scan, imaging of the shunt valve is ordered for the same day, so that the valve's setting can be confirmed. These images then have to be read by a radiologist or neurosurgeon who is familiar with that particular shunt. If the setting is off, then the neurosurgeon needs to reset the valve and perhaps even send the patient back for repeat imaging.

A more advanced programmer uses a built-in ultrasound to confirm a valve's setting without requiring separate imaging. But this introduces two new problems. The first is that some patients complain about getting ultrasound gel on their heads and hair. Second, the programmer is so sensitive and temperamental that it may not work in rooms with either too much noise or too much electrical equipment. This pretty much describes most doctors' offices. In my first experience with the new programmer, I tried close to a dozen times to adjust the valve setting before I gave up, used the old programmer, and sent the patient off for fluoroscopic imaging.

Meanwhile, a competing manufacturer has designed an altogether different programmable shunt that is advertised as MRI compatible. Not only that, but its programmer is almost pocket-sized, whereas the programmer for the old shunt is housed in a heavy, unwieldy, briefcase-like container. When I heard about this new shunt, I jumped at the opportunity to try it. It seemed almost too good to be true: no need to worry that MRI scans would change the settings, no need to bother with fluoroscopy, and no need to lug a heavy programmer. My first few cases with the new shunt went fine from the surgical standpoint. But it turned out that despite what I had been led to believe, MRI compatibility cannot be guaranteed. As I learned from a representative of the company that made the competing shunt I'd just forsaken, the fine print reveals that follow-up imaging of the new valve after an MRI is still officially recommended. Conclusion? Again, I can't win.