Laser treatment is the standard of care for threshold or highrisk prethreshold ROP. The ETROP study included 361 eyes with type 1 ROP treated mostly by laser and rest of them by cryotherapy ⁽³⁾. Retreatment was conducted in 13.9% of the infants ⁽³⁾. After 6-year follow-up, 8.9% of the treated children had unfavorable structural results ⁽⁴⁾. In the current study, there was no one requiring laser retreatment and surgical intervention in 26 patients undergoing primary laser ablative therapy. The fair outcome is possibly caused from small sample size, less proportion of zone I disease, and using nearly confluent laser pattern. Small sample size in the study can lead to statistical bias. Infants with zone I disease often had a poor outcome. Unfavorable structural outcomes following laser therapy were found in 22.2% of the high-risk prethreshold eyes with zone I disease at age 2 years in the ETROP trial ⁽³). In a prior study, 64% of patients having eyes with anterior zone I disease needed retinal detachment surgery following laser therapy ⁽²⁶⁾, and 100% of posterior zone I eyes requiring surgical repair reported in another article ⁽²⁷⁾. Different mechanisms of zones I and II ROP have been speculated. Zone I disease was thought to be correlated with aberrant vasculogenesis, which was less dependent on VEGF-mediated mechanisms, thus explaining its insensitivity to laser therapy, while zone II disease was considered angiogenesis-dependent and associated with VEGF; thus, laser ablation was effective ⁽²⁾. Superior outcomes in this study was possibly because zone I disease was found in 15.4% of the laser group in our case series, less than 41.3% of the patient with zone I disease in the ETROP study ⁽³⁾. Conventional laser treatment regimens suggested a burn density spaced one to one and a half burn widths apart ^(28,29). The prior study advocated that dense laser spots with more complete destruction of the avascular retina resulted in better anatomical outcome than sparse laser application ⁽²⁵⁾. We had fair clinical results in laser group by applying nearly confluent laser photocoagulation with half- to one-burn widths apart to avoid under-treatment, or ‘skip areas’ leading to persistent abnormal vascularity with progression to retinal detachment. 

In recent years, intravitreal anti-VEGF has become another option for treating advanced ROP. Bevacizumab injection initially serves as a supplemental therapeutic agent for severe laser-refractory ROP ^(1,20). IVB can also treat ROP when media opacities precluding laser photocoagulation such as vitreous hemorrhage, hyphema, and undilated pupils with tunica vasculosa lentis ^(2,20). Because the procedure of intravitreal injection was quick, bevacizumab therapy is better for the patient too sick for lengthy laser ⁽¹⁹⁾. Bevacizumab monotherapy has been a trend for treatment of ROP currently. The BEAT-ROP study collected 140 eyes with zone I or zone II posterior stage 3 plus disease receiving IVB management ⁽¹¹⁾. Before 54 weeks’ PMA, 4% of the treated eyes needed repeated bevacizumab injection, and 4.3% of them having unfavorable anatomical outcome ⁽¹¹⁾. A multicenter study recruited 162 eyes with prethreshold ROP treated by IVB ⁽¹²⁾. After mean 13.7-month follow-up, 1% of the treated children needed repeated bevacizumab injection due to persistence of plus disease, 9% of them requiring laser salvage therapy because of lack of response to bevacizumab, and 2% of them having unfavorable anatomical outcome ⁽¹²⁾. In this article, 11.7% of bevacizumab group received repetitive bevacizumab or laser salvage, which was comparable to the results in previous studies (from 4% to 10%) ^(11,12). All the infants of bevacizumab group had favorable outcome either after single injection of bevacizumab or/and additional laser treatment. The fair clinical result is possibly caused from small sample size, less proportion of zone I disease, and using laser salvage treatment. Zone I disease was found in 23.5% of the bevacizumab group in our case series, less than 44% of the patient with zone I disease in the BEAT-ROP study ⁽¹¹⁾. A prior article described 17 eyes with recurrence of ROP after initial treatment with IVB monotherapy ⁽¹⁷⁾. Five eyes progress to retinal detachment without laser salvage, but the other 12 eyes had favorable outcome following adjuvant laser applied ⁽¹⁷⁾. Laser salvage can further inhibit VEGF level in the eye by destroying the avascular retina that is responsible for continued VEGF expression. Complete regression of ROP was found after performing laser salvage in two patients with ROP reactivation in our study.

There were several reports comparing the anatomical results of laser and IVB treatment. A prospective study randomly conducted laser in one eye and IVB in the fellow eye in 12 infants with zone I type 1 ROP ⁽¹⁸⁾. The authors found all eyes treated with a bevacizumab were noted to have abnormalities at the periphery (large avascular area, abnormal branching, shunt) or the posterior pole (hyperfluorescent lesion, absence of foveal avascular zone). But these lesions were not observed in the majority of the lasered eyes after examination of fundus fluorescein angiography at age of 9 months ⁽¹⁸⁾. Higher chance of ROP reactivation was reported after bevacizumab monotherapy than after laser monotherapy in several prior studies. A prospective study performed IVB in one eye and laser therapy in contralateral eye of 14 patients with stage 3 posterior ROP ⁽¹³⁾. Three of 14 eyes (21.42%) had recurrence of ROP after single bevacizumab treatment, higher than one of 14 eyes (7.14%) with recurrence receiving conventional laser therapy ⁽¹³⁾. Another preceding retrospective comparative study demonstrated that more ROP recurred in 22 bevacizumabtreated eyes (14%) than 32 laser-treated eyes (3%) ⁽²¹⁾. We also discovered our patients receiving bevacizumab treatment had higher rate of ROP reactivation than those treated by laser. The different response between laser and bevacizumab for ROP reactivation can be hypothetically explained as follows: bevacizumab can temporarily inhibit intraocular VEGF, ROP may reactivate while vitreous VEGF levels elevate as anti-VEGF effect vanishing. Laser treatment destroys the avascular retinal tissue, which can permanently lower the intraocular VEGF ⁽³⁰⁾. Close followup and possible repeated IVB or/and laser salvage should be addressed before bevacizumab management for type 1 ROP. Favorable structural outcome can be achieved in most of the patients with proliferative ROP treated either bevacizumab monotherapy or conventional laser according to prior reports ^(13,21,22). Huang CK and coauthors reported only one eye developed retinal detachment among 32 laser-treated eyes with type 1 ROP, while no unfavorable results found in 22 bevacizumab-treated eyes ⁽²¹⁾. Favorable outcomes were discovered in all eyes with neovascular ROP after laser or IVB in two prior reports ^(13,22). We had similar findings in this study, in which neither laser-treated eyes nor bevacizumab-treated eyes had unfavorable outcome.

In this case series, laser was performed in the earlier years in patients with ROP (from 2002 to 2009), whereas intraocular bevacizumab injected in the recent years (from 2009 to 2015). The fact can explain the earlier cases undergoing laser had longer follow-up period than cases receiving bevacizumab treatment lately. Because recent advances of pediatric care for prematurity, cases with smaller GA and lower BBW can survive. The rate of zone I disease, that is, vasculopathy occurring close to the optic nerve, because the eyes were extremely immature and associated younger age and lighter body weight at birth, was on the rise ⁽²⁾. The bevacizumab group in current years tends to have significantly less GA and BBW and more zone I disease than the laser group in the early years in the study. There were some limitations to our study. Some bias may occur because this was not a prospective or multicenter study and all subjects coming from the same institution. Our study composed of small case number that may underestimate or overestimate the results. However, we uniquely compared clinical effects of laser and IVB in Chinese infants with type 1 ROP after average follow-up more than 4 years.

In summary, laser therapy and IVB were both effective treatments for type 1 ROP. Single session of laser ablation on peripheral retinal avascular area in nearly confluent pattern was sufficient for complete regression of ROP. Single IVB was appropriate for managing most of cases with ROP in bevacizumab group, but a small proportion (nearly one tenth) of them had reactivation requiring repeated injections or/and laser salvage.