Surgery and RVO Risk

Download PDF

Does cataract surgery reduce the risk of retinal vein occlusion (RVO)? A study of 4 million patients from the IRIS Registry yielded unexpected findings: First, cataract extraction did not appear to be protec­tive against development of RVOs, despite its effect on intraocular pressure (IOP) reduction. Second, the presence of diabetic retinopathy (DR) emerged as the strongest predictor of RVO development.¹

Power of big data. Both findings highlight the power of using large databases to explore questions that are impractical for randomized controlled trials. “Big data allows for even small effects to be teased out, so the finding that cataract surgery does not reduce the risk of RVOs despite lowering IOP was surprising,” said Andrew Chen, MD, at the University of Washington, Seattle, a senior coauthor of the study.

Study design. To determine the risk of developing RVO, the researchers emulated randomized controlled trials with a machine learning model. Patients were classified as belonging to the treatment or control groups based on known risk factors for cataract de­velopment. This allowed the two groups of patients to be selected according to the same set of rules. Factors included age, sex, primary insurance type, and history of DR, glaucoma, and narrow angles.

Study findings. Of the 4 million patients, there were a total of 2,062 central RVO events within one year of undergoing uncomplicated cataract surgery—or, for 1:1 matched controls, one year from the baseline visit. Of these, 1,141 occurred in the surgery group, and 921 oc­curred in controls. In addition, there were 3,488 branch RVO events, with 1,942 in the surgery group and 1,547 in controls.

The bottom line: Although surgery did not prevent RVO development, the number of RVOs in both groups was relatively small, and the proportion of eyes that did not develop either type of RVO was greater than 99.8%.

DR risk. DR was the strongest predictor associated with developing central RVO (hazard ratio [HR] 2.79; p < .001) and branch RVO (HR 1.97; p < .001) after cata­ract surgery. “The magnitude of the increased risk asso­ciated with DR was not expected given the results from prior epidemiologic studies, such as the Blue Mountain Eye study,” Dr. Chen said.

In discussing this discrepancy, the researchers noted that they relied on DR codes rather than systemic dia­betes mellitus codes in their analysis. “Thus, we would have only considered diabetes cases severe enough to have ocular manifestations, unlike previous studies,” they wrote.

Up next. Dr. Chen cautioned that relationships in a retrospective study do not imply causation. He called for future research that delves into the pathophysiol­ogy of the disease, noting that future versions of the IRIS Registry will include more variables that have been implicated as risk factors for RVOs, such as axial length. “We still do not fully understand the reason for why RVOs occur.”

—Miriam Karmel

___________________________

1 Bagdasarova Y et al. Ophthalmology Science. Published online July 13, 2021.

___________________________

Relevant financial disclosures—Dr. Chen: None.

For full disclosures and the disclosure key, see below.

• Graft Rejection Detected in Action

• Refining Dx of Pentosan Polysulfate Maculopathy

• New Consensus on RRD Repair Risks