LONDON — In a French study of 545 patients, wireless capsule colonoscopy achieved an overall sensitivity of only 39% and a positive predictive value of just 46%, Dr. Jean Paul Galmiche reported at the 13th World Congress of Gastroenterology.
The 16-center study was designed to discover the reliability of the colon capsule (Pillcam, Given Imaging Ltd.)—which travels through the GI tract equipped with a tiny camera—for finding colon polyps of at least 6 mm. Two cohorts were studied: asymptomatic patients aged 50-74 years needing routine screening (30% of subjects), and patients at increased risk for colorectal cancer because of a personal or family history of colon polyps or cancer (70% of subjects). The patients' mean age was 60 years.
The colon capsule results were compared with results from conventional colonoscopy performed the following day by colonoscopists who were blinded to the capsule findings. Images obtained during the two procedures underwent subsequent examination by an outside panel of experts.
The interim analysis by Dr. Galmiche's group at the University Hospital Nantes, France, showed that the wireless capsule colonoscopy (WCC) results achieved an overall sensitivity of 39% and a positive predictive value of 46% in terms of the total number of polyps correctly identified.
This was no match for conventional colonoscopy, which had a positive predictive value of 57% and a sensitivity of about 85% for all polyps regardless of size.
“At this stage of the development of the technology, WCC cannot replace colonoscopy for screening and surveillance of patients at risk for colon cancer,” Dr. Galmiche said in an interview.
“Future research should focus on better preparation and improved technology, that is to say, a new capsule generation,” he added.
The primary problem is that, despite rigorous attempts at preprocedure cleansing, colonic debris frequently obscured the wireless capsule's capacity to transmit reliable images. Only 52% of the 377 WCC patients were judged to have sufficiently “good to excellent” colon cleansing to allow full visualization from the technology.
In contrast, more than 85% of the 183 colonoscopy patients in the study achieved the same “good to excellent” bowel prep results.
The study began with patients on a 3-day low-residue diet, with 3 L of polyethylene glycol solution taken the night before the procedure, and another liter taken the following morning. Domperidone and a laxative booster were used to ensure prompt expulsion of the video capsule—which occurred within 10 hours in 93% of the patients.
About one-quarter of the way through the study, the 3-day low-residue diet was replaced with an all-liquid diet for the remaining patients, to improve visualization.
“Somewhat better results have been achieved with this newer approach to preparation,” Dr. Galmiche said.
In addition to the drawback of the high cost of WCC, the noninvasive procedure was associated with 20 adverse events, chiefly excessive abdominal pain and vomiting. However, these effects were all considered related to the bowel prep used before swallowing the video capsule, rather than mishaps with the capsule itself.
Dr. Galmiche noted that the WCC procedure had a negative predictive value of 88% for identifying polyps larger than 10 mm.
But he concluded by saying its sensitivity was “not optimal” and that a new iteration of the technology would need to have dramatic improvements before it could contend with colonoscopy as a gold standard screening tool for colorectal cancer.
The Nantes group's presentation was formally cited as one of the two Best Abstracts at the meeting.
Disclosures: Dr. Galmiche disclosed that he is on the advisory board of Given Imaging but has no financial links. The study was sponsored by a National Research Grant.