Journal Review

by Jennifer Bisquera, MD

The STOP-ROP Multicenter Study Group. Supplemental Therapeutic Oxygen for Prethreshold Retinopathy of Prematurity (STOP-ROP), A Randomized, Controlled Trial. I: Primary Outcomes. Pediatrics. 2000;105(2):295-310

Abstract

Introduction: The STOP-ROP Multicenter Study Group attempted to determine the efficacy and safety of supplemental oxygen to reduce the progression of ROP and the need for peripheral ablation therapy in infants with prethreshold ROP.

Methods: In this multicenter randomized controlled trial, infants in 30 centers were screened for ROP every 2 weeks. Those with confirmed prethreshold ROP and median pulse oximetry <94% were randomized, either to a conventional oxygen arm with pulse oximetry targets at 89–94% saturation or to a supplemental arm with pulse oximetry targets at 96–99%, for at least two weeks or until both eyes were at study endpoints. An adverse endpoint was reached when threshold criteria was met for laser or cryo therapy in at least one study eye. A favorable endpoint was either regression to Zone III for at least two consecutive weeks or full retinal vascularization. At 3 months’ corrected age, ophthalmic findings, pulmonary status, growth, and interim illnesses were recorded.

Results: Of the 649 infants enrolled over 5 years, 92% reached ophthalmic endpoints and 92% completed the three-month assessment. The rate of progression to threshold was 48% in the conventional arm and 41% in the supplemental arm. The authors concluded that supplemental oxygen at pulse oximetry saturations of 96–99% did not significantly reduce the number of infants requiring peripheral ablative surgery but also did not cause additional progression of prethreshold ROP. However, increased levels of oxygenation increased the risk of adverse pulmonary events (8.5% conventional vs 13.2% supplemental).

Analysis

Strengths: This is a well-designed trial. It had a large number of premature infants who were highly comparable in both treatment arms and a low rate of non-completion or drop-out. A strict standard of significance was used, allowing cautious interpretation of results. Other relevant pediatric data was gathered at 3 months’ corrected age: higher incidence of CLD exacerbation/pneumonia and need for diuretics, oxygen, and hospitalization shown in the high oxygen arm of the study. Analysis of the number needed-to-treat demonstrated that one episode of pneumonia/CLD occurred for each case of peripheral ablative surgery prevented.

Weaknesses: The study did not clearly rule out the potential for oxygen supplementation to reduce the progression of ROP to threshold, as was previously reported in two smaller clinical studies. This study lacked the appropriate power expected for two reasons: 1) the adverse ophthalmic outcome rate in the conventional group was higher than expected (expected was 30%; actual 48% in the conventional group); and 2) the target sample size (880) was reduced (to 649) with a resultant power of about 80% instead of the 90% planned.

Discussion

Several possible reasons exist to explain why this study had different results from the two previous clinical studies showing benefit: 1) STOP-ROP infants had lower birth weights and, therefore, may have had more severe ROP compared to subjects in the other studies; 2) lower oxygen saturation levels in the STOP-ROP infants; 3) delay in initiation of treatment (screening exams every two weeks vs. weekly in the other studies), and; 4) use of historical controls in one of the two studies. In addition, all eligible infants were enrolled in the two other studies, regardless of baseline median pulse oximetry readings. All three studies did not show any harm to the eyes with oxygen supplementation. The STOP-ROP study is the only one that showed potentially harmful effect to the lungs.

This study has significant clinical implications. Supplemental oxygen does not appear to be harmful to eyes that already have ROP. So, attempts may not be necessary to tightly control oxygenation in infants with ROP for fear of making the ROP worse, especially in settings that require supplemental oxygen for cardiopulmonary reasons. However, contrary to previously published work, supplemental oxygen was shown to be potentially harmful for some infants with more severe CLD.