Monday, September 22, 2008

Virtual Reality




















Last week, The New England Journal of Medicine published a study evaluating the efficacy of CT colonography (virtual colonoscopy) compared to standard optical colonoscopy. 2600 patients received CT colonography with a follow up standard colonoscopy on the same day. The results show that virtual colonoscopy identified 90% of adenomas/cancers greater than 1cm in size.

As KevinMD suggests, this could instigate a turf war between radiologists and gastroenterologists as both specialties fight to tout their procedure as the surveillance technique of choice of colorectal cancer. Because virtual colonoscopy is completely non-invasive (no long black tube up the rear, elimination of the risk of perforation) one can certainly see the appeal to the general population as the preferred screening method, but I'm not necessarily convinced that that will be the case because, ultimately, this may be a turf war the radiologists don't want to win.

The current theory of colorectal cancer development is one of step wise progression from polyp to dysplastic polyp to frank invasive cancer. The reason polyps are removed at colonoscopy is that polyps are thought to be pre-cancerous lesions; therefore take out the polyp before it has a chance to transform into a cancer, the thinking goes.

However, there's a kink in the theory. This year in JAMA, a study was published by the folks at Palo Alto called "Prevalence of nonpolypoid (flat and depressed) colorectal neoplasms in asymptomatic and symptomatic adults." It's a cross sectional study of over 1800 patients who underwent colonoscopy at a California VA. What they found was a 9.35% prevalence of non-polypoid lesions. Furthermore, after biopsies were performed, these lesions were found to have a higher incidence of harboring invasive cancer regardless of lesion size. Now, flat polypoid lesions are difficult to spot on normal optical colonoscopy. You have to proceed much more slowly and make sure you thoroughly inspect the entire mucosa of the colon. Even so, these lesions can be easily missed. With virtual colonoscopy, there's no chance of spotting them.

So even though the public may demand more CT Colonographies, all it's going to take is a couple of missed flat lesions or even peduncular polyps (10% miss rate in the NEJM article!) that turn into a stage III cancer and the turf war will come to a very quick ceasefire. I don't think GI guys have too much to worry about yet.....

4 comments:

Bongi said...

i do a lot of colonoscopies and recently my colleague and i had exactly this discussion (he had read the quoted article about flat lesions. i had not). i suppose you could say there is nothing like the real thing?

OHN said...

I would imagine (as a lay person) that the prep would be the same for either the virtual or scope. If that is the case, with a healthy dose of loopy drugs it seems to make sense to go with the standard route so removal of anything suspicious could be taken care of in one visit vs coming back for removal if something is found on virtual. (As a side note, my mother became septic and died after a perforation during a colonoscopy done by a resident so I know that can be a concern but I still think that "hands on" so to speak makes more sense).

doc said...

Vurtual colonoscopy will also cut the referral strings between general surgeons and GI docs. GS's will begin to get referrals straight from PCP's and radiologists which is less power for GI docs. I think the GI docs should be worried. The technology involved will only improve with time.

Frank Drackman said...

So what if you have Polyps? You still need someone there in real time. And X-rays aren't totally benign either.