The American FDA has a track record of sitting back and letting Europe guinea-pig medical innovations for us, especially durable goods. Many a web search has resulted in the familiar shutdown: “not available in America, although in common use in Europe.” I myself have participated in medical tourism for a durable implantable device that’s been in use in Europe since the mid-1980s, but still isn’t available stateside, for reasons beyond my ken.
One such device, which has settled into widespread use in Europe, but remains under provisional FDA oversight instead of outright approval, is the PillCam. You may have heard about the device a few years back, but it’s still hung up in the world of “off-label” use. While the FDA has provisionally approved the PillCam for use, they only allow it after a patient has experienced an unsuccessful endoscopy or colonoscopy.
Here’s how it works. There are three versions of the device: the PillCam ESO, for the esophagus, the PillCam SB for the stomach, and the PillCam Colon for the lower GI tract. All versions of the PillCam contain a little chemical battery that’ll give it eight to ten hours of recording time, plus a light source and two tiny CCD cameras. A patient who’s about to use it will strap on a belt containing a receiver and an SD card or equivalent storage medium. Then they swallow the tiny pill.
The camera proceeds through the GI tract, takes between 2,600 and 57,000 pictures while it’s on its way through, and “produces no pain or even sensation as it moves through the colon” over a period of a few hours. After it’s taken all the close-ups and beauty shots it can, the thankfully disposable PillCam winds up flushed down the toilet within about 72 hours. But the photos are on the SD card, ready for medical examination.
Given the generally extremely unpleasant experience of having a colonoscopy or endoscopy, it’s a major perk that the patient feels nothing at all during this procedure. According to a study published in the New England Journal of Medicine, there were no serious adverse events among 320 subjects given the PillCam Colon. The study concluded that use of the PillCam was “a safe method of visualizing the colonic mucosa through colon fluids without the need for sedation or insufflation.” Later feedback from clinics using the PillCam pointed out that some patients insisted on having X-rays to prove that they’d actually passed the camera, because they never felt a thing.
This is exactly what doesn’t happen.
There are still a few drawbacks. The PillCam isn’t as good at diagnosing polyps as a traditional colonoscopy, but it has the advantage of being able to reach the 15-20 feet of small intestine that colonoscopy just can’t see. It also tends to get hung up if a patient already has a bowel obstruction or another obstructing GI condition like polyps or diverticulosis. It can’t replace the colonoscopy, but it sure can offer an adjunct therapy that isn’t nearly as uncomfortable.
Likewise, the PillCam ESO doesn’t image the stomach, so it can’t really catch signs of peptic ulcers. For that, an endoscope is still the right approach, because it allows the doctor administering the endoscopy to also take a biopsy. Having the PillCam catch a picture of a possible ulcer would just mean the person would have to go through an endoscopy later anyway. Bummer. (This is part of why the FDA hasn’t approved it as a first-line diagnostic method.)
But the PillCam ESO’s ability to thoroughly image the upper GI tract without patient discomfort is unparalleled. It can catch problems in the esophagus, like signs of the chronic acid reflux damage that can contribute to GERD.
And this capture-all-the-data approach lends itself well to computer-aided diagnostics. Machine learning is making great strides in identifying subtle but important conditions in skin cancer, for example — and skin is made of epithelial tissue, just like the membranes lining the GI tract. This suggests an area of diagnostics that’s ready for some AI to sink its digital teeth into the problem. Hey, Watson — doing anything this weekend?