Study: Compounded Avastin not linked to increased risk of endophthalmitis

Issue: October 25, 2015

Intravitreal injections of repackaged Avastin did not pose a greater risk of endophthalmitis than Lucentis, according to a study.

Brian L. VanderBeek, MD, MPH, and colleagues at the Scheie Eye Institute, Philadelphia, reported their results in JAMA Ophthalmology.

“To me, the most significant finding is the fact that, for over 7 years of data in a national database, the rate of endophthalmitis in Avastin was no different than the rate for Lucentis,” VanderBeek told Ocular Surgery News.

In response to reports of an endophthalmitis outbreak linked to repackaged bevacizumab in 2011, the FDA issued a preliminary draft guidance allowing the use of compounded bevacizumab with a beyond-use date that limits use to within 5 days of repackaging.

“The big concern was whether the repackaging of Avastin induced extra risk above and beyond the baseline risk. That appears to have not happened,” VanderBeek said. “I’m hopeful that the FDA takes this study into consideration before making a final judgment on its draft guidance.”

Analysis of medical claims data

The authors retrospectively analyzed medical claims data on 383,810 intravitreal injections of Avastin (bevacizumab, Genentech) and Lucentis (ranibizumab, Genentech) given to 58,612 patients between 2005 and 2012.

Specifically, 296,565 injections of bevacizumab were given to 51,116 patients and 87,245 injections of ranibizumab were given to 7,496 patients.

Primary diagnoses were age-related macular degeneration, diabetes and retinal vein occlusion.

Mean time to diagnosis of endophthalmitis was 4.75 days after injection.

The authors wrote that, to the best of their knowledge, the study used the single largest data source to evaluate the risk of endophthalmitis after bevacizumab or ranibizumab injection.

“I used what’s called an administrative claims database. Essentially, a single insurance company puts together the data it has on all of its beneficiaries over a certain amount of time. It de-identifies that data so we’re not able to re-identify people, making it safe for research use. It then turns around and sells it to researchers who use it for purposes such as this study,” VanderBeek said.

Results

There were 71 cases of endophthalmitis: 49 in the bevacizumab group and 22 in the ranibizumab group. The overall rate of endophthalmitis was 0.018%. Separately, the rates were 0.017% in the bevacizumab group and 0.025% in the ranibizumab group. The between-group difference was not statistically significant.

An analysis with a broader definition of endophthalmitis showed 97 cases of endophthalmitis: 69 in the bevacizumab group and 28 in the ranibizumab group. However, that analysis did not show a stronger correlation between bevacizumab and endophthalmitis, the authors wrote.

Overall, there was no correlation between intravitreal bevacizumab and the incidence of endophthalmitis, VanderBeek said.

“Anytime you do a procedure on any part of the body, one of the risks is infection. Without exception, you cannot 100% completely eliminate the risk of infection. So, even in the Lucentis group, you’re going to see some level of infection,” he said. – by Matt Hasson

Reference:
VanderBeek BL, et al. JAMA Ophthalmol. 2015;doi:10.1001/jamaophthalmol.2015.2556.

For more information:
Brian L. VanderBeek, MD, MPH, can be reached at Scheie Eye Institute, Department of Ophthalmology, University of Pennsylvania Perelman School of Medicine, 51 N. 39th St., Philadelphia, PA 19104; email: brian.vanderbeek@uphs.upenn.edu.

Disclosure: VanderBeek reports no relevant financial disclosures.