WASHINGTON –
“The benefit appears limited to proximal polyps,” Heiko Pohl, MD, said at the annual Digestive Disease Week®. In that prespecified subgroup, which included two-thirds of enrolled patients, placement of hemostatic clips on defects left after removing polyps 20 mm in diameter or larger cut the rate of delayed, severe bleeding by two-thirds, compared with patients with large defects not treated with clips. This result represented a number needed to treat with clips of 15 patients with large proximal polyps to prevent one episode of delayed severe bleeding, said Dr. Pohl, a gastroenterologist at the VA Medical Center in White River Junction, Vt.
Mitchel L. Zoler/MDedge News
Dr. Heiko Pohl
Although the results that Dr. Pohl reported came from a trial that originally had been designed to generate data for Food and Drug Administration approval for using the clips to close defects following large polyp removal, the clips received approval for this indication from the agency in 2016 while the study was still in progress.
But Dr. Pohl maintained that the new evidence for efficacy that he reported will provide further impetus for gastroenterologists to use clips when they remove larger polyps in proximal locations. “I think this study will help standardize treatment of mucosal resections and change clip use,” he said in an interview.
“This was a terrific study, and one that needed to be done,” commented John R. Saltzman, MD, professor of medicine at Harvard Medical School and director of endoscopy at Brigham and Women’s Hospital in Boston. But Dr. Saltzman, who spoke from the floor during discussion of Dr. Pohl’s report, added that data on the average number of clips required to close defects were needed to assess the cost-effectiveness of the treatment, data that Dr. Pohl said were available but still being analyzed.
“We have to know how many clips to use and how to close the polyp,” Dr. Saltzman said. Dr. Pohl estimated that roughly four or five clips had been used per defect, but he cautioned that this estimate was preliminary pending his complete analysis of the data.
The CLIP (Clip Closure After Endoscopic Resection of Large Polyps) study enrolled patients with at least one nonpedunculated colonic polyp that was at least 20 mm in diameter at 16 U.S. centers, as well as one center in Montreal and one in Barcelona. The patients averaged 65 years of age, and 6%-7% of patients had more than one large polyp removed during their procedure. Randomization produced one important imbalance in assignment: 25% of the 454 patients in the clipped arm were on an antithrombotic drug (either an anticoagulant or antiplatelet drug) at the time of their endoscopy, compared with 33% of the 464 patients in the control arm.
The study’s primary endpoint was the incidence of “severe” bleeding within 30 days after the procedure. The study defined severe bleeding as an event that required hospitalization, need for repeat endoscopy, need for a blood transfusion, or need for any other major intervention, explained Dr. Pohl, who is also on the staff of Dartmouth-Hitchcock Medical Center in Lebanon, N.H.
Such events occurred in 3.5% of the patients who underwent clipping and in 7.3% of control patients who received no clipping, a statistically significant difference (P = .01). Among patients with proximal polyps, the bleeding rates were 3.3% among clipped patients and 9.9% among controls, also a statistically significant difference. Among patients with distal polyps the bleeding rates were 4.0% among clipped patients and 1.4% among controls, a difference that was not statistically significant.
Dr. Pohl and his associates ran three other prespecified, secondary analyses that divided the enrolled patients into subgroups. These analyses showed no significant effect on outcome by polyp size when comparing 20-39 mm polyps with polyps 40 mm or larger, treatment with an antithrombotic drug, or method of cauterization. The median time to severe bleeding was 1 day among the controls and 7 days among the clipped patients.
Aside from the difference in rates of delayed bleeding, the two study arms showed no significant differences in the incidence of any other serious postprocedure events. The rates of these nonbleeding events were 1.3% among clipped patients and 2.4% among the controls.
The researchers ran all these analyses based on the intention-to-treat assignment of patients. However, during the study, 9% of patients assigned to the control arm crossed over ended up receiving clips during their procedure after all, a rate that Dr. Pohl called “surprisingly high,” whereas 14% of patients assigned to the clip arm never received clips. A per-protocol analysis that censored patients who did not receive their assigned treatment showed that, among the remaining patients who underwent their assigned treatment, the rate of delayed, severe bleeds was 2.3% among the 390 patients actually treated with clips and 7.2% among the 419 controls who never received clips, a statistically significant difference, he reported.
Dr. Pohl also noted that it was “somewhat surprising” that clipping appeared to result in complete closure in “only” 68% of patients who underwent clipping and that it produced partial closure in an additional 20% of patients, with the remaining patients having mucosal defects that were not considered closed by clipping.
The study was funded by Boston Scientific, the company that markets the hemostatic clip (Resolution 360) tested in the study. Dr. Pohl had no additional disclosures. Dr. Saltzman had no disclosures.
SOURCE: Pohl H et al. Digestive Disease Week, Presentation 886.