Evaluation of regression, reactivation, retreatment and retinal detachment after injections of three different anti-VEGF drugs for retinopathy of prematurity Authors: Melih Akıdan, Ayşe Cengiz Ünal, Muhammet Kazım Erol Published: 2026 ABSTRACT AbstractThis study aimed to compare the regression, reactivation, retreatment, and retinal detachment (RD) outcomes of eyes with type 1 retinopathy of prematurity (ROP) and aggressive ROP (A-ROP) treated with different anti-VEGF drugs: bevacizumab (IVB), ranibizumab (IVR), and aflibercept (IVA). This single-center retrospective study included infants treated with anti-VEGF therapy for type 1 ROP or A-ROP between 2014 and 2020. Ultra-widefield (UWF) retinal images were captured with Optos confocal laser ophthalmoscopy. Data on initial and retreatment injection rates, zone findings, time intervals between treatments, additional laser therapy rates and durations, and RD cases were analyzed in detail. The study included 550 eyes of 284 infants, 52.5% (149) of whom were female and 47.5% (135) of whom were male. The gestational age (GA) ranged from 22 to 33 weeks and BW ranged from 450 to 1650 g. The first injection rate across all cases was 99.8% (549) and these injections occurred at between 33 and 58 weeks. The retreatment injection rate was 9.6% (53) and these injections occurred at between 35 and 60 weeks. The time interval between injections ranged from 1 to 26 weeks. Post-injection, laser therapy was administered to 15.8% (87) of eyes. RD was observed in 1.8% (10) of eyes between 38 and 120 weeks. Initial injections for RD cases occurred significantly earlier. Anti-VEGF therapy has recently been used as an effective treatment for retinopathy of prematurity. The retreatment rate (9.6%) and RD rate (1.8%) are quite low. Long-term follow-up remains crucial for managing late reactivation and complications.Keywords: Type 1 ROP, Aggressive ROP, Anti-VEGF, Retinal detachmentSubject terms: Diseases, Health care, Medical research Introduction Retinopathy of prematurity (ROP) is the leading cause of preventable vision loss among premature infants1. While ROP resolves spontaneously without requiring treatment in most cases, timely retinal examinations, follow-up, and contemporary treatment approaches are crucial in preventing permanent blindness in severe cases2. Studies in the existing literature have extensively examined various aspects of ROP management, including comparisons of monotherapy with laser, combined laser and anti-VEGF treatments, and different anti-VEGF drugs, in terms of efficacy, retreatment requirements, and ocular or systemic complications. Landmark studies such as ETROP, BEAT-ROP (Bevacizumab Eliminates the Angiogenic Threat for Retinopathy of Prematurity), RAINBOW (Ranibizumab Compared with Laser Therapy for ROP), and FIREFLEYE (Aflibercept for ROP-IVT Injection versus Laser Therapy) have significantly shaped our understanding of these interventions3–5. Additionally, meta-analyses and observational studies, including those by Zhou et al., Chen et al., Ortiz-Seller et al., and Pfeil et al., have provided the most recent insights into different anti-VEGF treatments6–9. Due to the need for sedation or general anesthesia, the potential for permanent retinal damage, visual field loss, and high refractive errors such as myopia associated with laser therapy, anti-VEGF drugs are increasingly considered as the first choice in ROP treatment today10,11. Although anti-VEGF drugs provide rapid and dramatic results with less anatomical damage, concerns regarding ocular and systemic safety, as well as long-term outcomes, prevent them from being considered the gold-standard treatment12. The most debated aspects are reactivation, the decision for a second intravitreal injection, injection timing, and laser photocoagulation decisions. Although the opinion that reactivations are more frequent and retreatments occur later with anti-VEGF compared to laser therapy holds slightly more weight, further studies are needed on anti-VEGF drugs and laser therapy applications7,9,13. In this study, eyes with type 1 ROP and A-ROP were treated with different anti-VEGF drugs including bevacizumab (IVB), ranibizumab (IVR), and aflibercept (IVA).Post-injection assessments were conducted to compare regression, reactivation, and retreatment injection rates between the agents. Additionally, among patients who developed RD following anti-VEGF injections, systemic and ocular risk factors were analyzed in detail to elucidate the progression of the condition. Material and methods Approval for this study was obtained from the local ethics committee of the University of Health Sciences Antalya Training and Research Hospital. The study was conducted in accordance with the ethical principles of the Declaration of Helsinki. Informed consent was obtained from the parents or guardians of all infants before ROP examination and treatment. This single-center retrospective study included patients who underwent anti-VEGF injection for type 1 ROP or A-ROP between January 2014 and December 2020 and were followed for at least 36 months (Exception: Patient 2, right eye, diagnosed with stage 5a disease and directly treated with surgery). Patients who did not complete anti-VEGF or other treatments, had other ocular diseases, were withdrawn from follow-up by their parents/guardians, or died during follow-up were excluded from the study. The study included infants identified by a neonatologist as having gestational age (GA) of < 32 weeks and body weight (BW) of < 1500 g or high risk for ROP. GA was estimated based on fetal ultrasound. Initial ROP examinations were conducted when the infants reached GA of 32 weeks or at a chronological age of 4 weeks. Prior to the initial dilated fundus examination, GA, BW, sex, and prenatal and postnatal comorbid risk factors were recorded using standardized forms. Dilated fundus examinations were performed using indirect ophthalmoscopy, and Optos device (Optos PLC, Dunfermline, UK) ultra-widefield (UWF) images were obtained by two ophthalmologists (Dr. M.A. and Dr. M.K.E). The ROP examination assessed pupillary dilation, vitreous opacity, the presence of plus disease (defined as tortuous and dilated vessels or iris neovascularization with/without vitreous opacities), and the stage, location, and extent of ROP disease. ICROP-2 and ETROP were used as ROP terminology since ICROP-3 had not yet been published between 2014 and 2020, when the patient data were collected3,14. In this article, the data were retrospectively examined and recategorized according to ICROP-315. In addition to the treatment and follow-up parameters for eyes with type 1 ROP and A-ROP, the study retrospectively reviewed the development of refractive errors, corneal opacity, cataracts, preretinal or intravitreal hemorrhage, and RD. All patients were administered one of three anti-VEGF medications among IVB (0,625 mg/0,025 mL), IVR (0,25 mg/0,025 mL), and IVA (0.25 mg/0.025 mL) as monotherapy within the first 72 h following diagnosis (or within 24 h for A-ROP)3,16–18. IVB, the most cost-effective anti-VEGF option, is primarily used intravenously for cancer treatment but is widely used off-label for ocular administration, with demonstrated efficacy in ROP treatment. IVR and IVA, FDA-approved for ocular use, are effective alternatives with single-use dosing options. Parents or guardians were thoroughly informed about the anti-VEGF drugs and their potential ocular and systemic side effects, and the choice of drug was left to their discretion. Baseline characteristics were largely similar between the treatment groups. No corrections were made between the treatment groups or during statistical analyses that might have influenced the outcomes. In accordance with the 2018 recommendations of the American Academy of Pediatrics, follow-up examinations were conducted on the first day and first week post-injection, followed by weekly or biweekly evaluations until ROP had completely regressed and retinal vascularization reached zone 319. Regression was defined as reduction of at least one stage and the resolution of plus disease. UWF retinal images and notes from follow-up visits were recorded in patient files and compared with results from previous weeks. Retreatment was performed with a second injection in cases of reactivation (typically observed at < 72 weeks). Laser treatment was not used as the first-line option due to its potentially weaker efficacy and risks of adverse structural retinal outcomes, visual field loss, and high refractive errors, particularly in zone 1. However, the laser treatment was administered in combination with the first injection in cases of A-ROP where severe zone 1 findings (with hemorrhages, massive proliferation) were present. This treatment was administered in the absence of regression after repeated treatment injections, or as monotherapy in late reactivation (> 72 weeks) cases within 72 h. Reactivation was defined as the recurrence of plus disease, extraretinal fibrovascular proliferation, neovascularization, the formation of a new ridge after initial regression, or disease progression despite prior treatment. Patients who developed persistent avascular retina (PAR) and underwent laser treatment for PAR were excluded from this study. After the necessary sterilization, the intravitreal injection was performed with a 30 G needle 1.5 mm away from the limbus, and topical antibiotic drops were started. Laser photocoagulation was performed using an indirect laser at a wavelength of 810 nm (diod lazer), photocoagulation(average power 150–200 mW, average duration 200 msn, average number of shots 5000–7000) was applied to the entire avascular retina. As post-laser treatment, topical antibiotics (0.3% moxifloxacin) and steroids (0.10% betamethasone) were administered four times daily for one week. While 4 patients (4 eyes) had stage 4a, 5 patients (5 eyes) had stage 4b. In surgeries for 4a and 4b ROP, lens sparing vitrectomy was performed using 23 gauge 4 port chandelier light releasing the fibrous tissue dissection and traction. For stage 5a ROP, in addition to vitrectomy with extensive fibrovascular membrane dissection and lensectomy, scleral buckling surgery was performed. In patient 2 (one eye) with stage 5A, we applied a scleral buckle prior to pars plana vitrectomy membrane dissection to facilitate easier peeling of the peripheral membranes. In addition, in two patients with stage 4A disease, due to the presence of an iatrogenic tear in an area with no traction or minimal traction during vitrectomy, intensive laser treatment was performed around the tear. Intraocular gas injection was performed in these patients. Silicone was not used in any of the patients. All injections, laser treatments, and vitrectomy surgeries were carried out by the same surgeon (M.K.E.) following similar steps for all patients. Statistical analysis Statistical analysis was performed using IBM SPSS Statistics 25 (Copyright SPSS Inc., 1989, 2017). Categorical variables were presented as frequency (n) and percentage (%), while continuous variables were expressed as mean ± standard deviation (SD) and minimum–maximum values. Non-parametric tests were applied regardless of normal distribution when the sample size (n) of the group was less than 30. Accordingly, the Mann–Whitney U test was used for independent two-group analyses and the Kruskal–Wallis H test was employed for comparisons involving more than two independent groups. In cases where the data did not show normal distribution, the Mann–Whitney U test was used for comparisons of two independent groups. For the analysis of independent categorical variables, the Pearson chi-square test, Fisher exact test, Fisher Freeman-Halton exact test, and post hoc Bonferroni correction were utilized. A p-value of 0.05 was considered as the threshold for statistical significance. Results This study included 550 eyes of 284 patients, 52.5% (149) of whom were female and 47.5% (135) of whom were male. Among the eyes, 277 (50.4%) were right eyes and 273 (49.6%) were left eyes. The patients’ GA ranged from 22 to 33 weeks (mean: 27.59 ± 2.17 weeks) and BW ranged from 450 to 1550 g (mean: 922.64 ± 199.83 g) (Table 1). Statistically significant relationships were observed between GA, BW, and anti-VEGF injection parameters. As BW and GA decreased, the first injection was performed at earlier weeks (positive correlation, p < 0.001 for both) and retreatment injections increased (negative correlation, p = 0.001 and p = 0.016) (Table 2). Similar statistical results were obtained between GA, BW, and the medications used for injections (Table 3). Table 1Baseline characteristics all infants.All CasesPatients/Eye284 / 550Gender, M/F (%)47,5% (135) / 52,5% (149)GA (weeks) mean ± SD (range)27,59 ± 2,17 (22–33)BW (g) mean ± SD (range)922,64 ± 199,83 (450–1650)Type 1 ROP and Zone 1/2 (%)96,9% (533) / 5,4% (29) / 94,6% (504)Stage 1/2/3 (%)0,7% (4) / 6,6% (35) / 92,7% (494)Plus disease (%)95,6% (526)A-ROP (%)2.9% (16)Laser (% and week)15,8% (87) / 49,53 ± 13,63 (35–112)Laser (Type 1 ROP and A-ROP, %)13% (71) / 100% (16)RD (% and week)1,8% (10) / 57,83 ± 30,87 (38–120)RD (Type 1 ROP and A-ROP, %)44,4% (4) / 55,6% (5)M/F: Male/Female, GA: Gestational age, SD: Standard deviation, BW: Birth weight, ROP: Retinopathy of prematurity, AROP: Aggressive ROP, RD: Retinal detachment. Baseline characteristics all infants. M/F: Male/Female, GA: Gestational age, SD: Standard deviation, BW: Birth weight, ROP: Retinopathy of prematurity, AROP: Aggressive ROP, RD: Retinal detachment. Table 2Correlation analysis GA and BW with anti-VEGF (all cases) injection parameters.All casesGABWFirst injection weekr0,661**0,578**p < 0,001 < 0,001Second injection weekr0,0830,315p0,6630,09Total number of injectionsr-0,196**-0,143*p0,0010,016Time interval between injectionsr-0,1040,187p0,5830,323GA: Gestational age, BW: Birth weight ṠSpearman Rho Correlation analysis *The correlation is significant at the 0.05 level. **The correlation is significant at the 0.01 level. Correlation analysis GA and BW with anti-VEGF (all cases) injection parameters. GA: Gestational age, BW: Birth weight ṠSpearman Rho Correlation analysis *The correlation is significant at the 0.05 level. **The correlation is significant at the 0.01 level. Table 3Evaluation of anti-VEGF drugs (IVB, IVR, IVA) associations with GA and BW.n1/n2/n3IVBIVRIVApǩFirst injection133/87/63GA (weeks)28 (26–30)28 (26–29)28 (26–29)0,872BW (g)900 (770–1070)910 (805–1070)930 (805–1040)0,888Second injection11/9/10GA (weeks)26 (23–28)27,5 (27–28)26 (26–26)0,215BW (g)800 (520–1000)850 (720–1130)905 (740–914)0,738GA: Gestational age, BW: Birth weight, IVB: Intravitreal bevacizumab, IVR: Intravitreal ranibizumab, IVA: Intravitreal aflibercept. ǩKruskal-Wallis H Test, Med(IQR). Evaluation of anti-VEGF drugs (IVB, IVR, IVA) associations with GA and BW. GA: Gestational age, BW: Birth weight, IVB: Intravitreal bevacizumab, IVR: Intravitreal ranibizumab, IVA: Intravitreal aflibercept. ǩKruskal-Wallis H Test, Med(IQR). The overall rate of first injections was 99.8% (549). Only one eye (1.8%) was directly diagnosed with stage 5a ROP and underwent surgery (Table 4). First injections were administered at 33–58 weeks (mean: 39.98 ± 5.54 weeks). The medications used for first injections were IVB (47.5%, 261), IVR (21.9%, 120), and IVA (30.6%, 168). The retreatment injection rate was 9.6% (53), administered between 35 and 60 weeks (mean: 44.91 ± 7.05 weeks). The interval between injections ranged from 1 to 26 weeks (mean: 8.2 ± 5.75 weeks). For retreatment injections, IVB was used for 43.4% (23), IVR for 34% (18), and IVA for 22.6% (12) of cases. The retreatment rates for each drug were 8.8% (23) for IVB, 15% (18) for IVR, and 7.1% (12) for IVA. Statistical similarities were found between the drugs in terms of retreatment rates and injection intervals (p = 0.069 and p = 0.0943) (Table 4). Table 4Evaluation of anti-VEGF injection parameters in eyes with all cases, type 1 ROP (and Zone findings), A-ROP and retinal detachment.All casesType 1 ROPZone 1 ROPZone 2 ROPA-ROPRDpFirst injection, (n, %)549 (99,8%)533 (100%)29 (5,4%)504 (94,6%)16 (100%)9 (90%)First injection (week)39,98 ± 5,54 (33–58)39 (36–43)µ35 (34–39)µ39 (36–43)34 (34–35)µ35 (34–35)µ < 0,001µFirst injection (drug, %)IVB261 (47,5%)253 (47,5%)1 (3,4%)a252 (50%)b8 (50%)4 (44,4%)0,716IVR120 (21,9%)118 (22,1%)14 (48,3%)a104 (20,6%)b2 (12,5%)2 (22,2%)IVA168 (30,6%)162 (30,4%)14 (48,3%)a148 (29,4%)b6 (37,5%)3 (33,4%)Second injection, (n, %)53 (9,6%)43 (8,1%)*11 (37,9%)*32 (6,3%)10 (62,5%)*7 (70%)* < 0,001*Second injection (week)44,91 ± 7,05 (35–60)46 (40,5–48)µ41 (37–47)46 (42–48)40 (36–42)µ42 (36–47)*0,038µ < 0,001*Second injection (drug)IVB23 (43,4%)17 (39,5%)a1 (9,1%)a16 (50%)b6 (60%)a3 (43%)0,017*IVR18 (34%)18 (41,9%)b*8 (72,7%)a10 (31,3%)b0 (0)b*2 (28,5%)IVA12 (22,6%)8 (18,6%)a2 (18,2%)a6 (18,8%)a4 (40%)a2 (28,5%)Retreatment interval (week)8,2 ± 5,75 (1–26)8 (5–12)6 (1–11)9 (5–12)5,5 (1–7)7 (1–12)0,189Total injections1,09 ± 0,30 (0–2)1 (1–1)µ1 (1–2)µ1 (1–1)2 (1–2)µ2 (1,5–2)µ < 0,001µROP: Retinopathy of prematurity, A-ROP: Aggressive ROP, RD: Retinal detachment, IVB: Intravitreal bevacizumab, IVR: Intravitreal ranibizumab, IVA: Intravitreal aflibercept. *Pearson Chi-Square Test, (Fisher’s Exact Test, Fisher’s Freeman Halton Exact Test, Post-Hoc Bonferroni Correction for drug parameters at 2nd injection), n(%). µMann Whitney U test, Med(IQR). Each superscript letter (a,b) denotes subgroups that do not differ significantly from each other at the 0.05 significance level. Evaluation of anti-VEGF injection parameters in eyes with all cases, type 1 ROP (and Zone findings), A-ROP and retinal detachment. 0,038µ < 0,001* ROP: Retinopathy of prematurity, A-ROP: Aggressive ROP, RD: Retinal detachment, IVB: Intravitreal bevacizumab, IVR: Intravitreal ranibizumab, IVA: Intravitreal aflibercept. *Pearson Chi-Square Test, (Fisher’s Exact Test, Fisher’s Freeman Halton Exact Test, Post-Hoc Bonferroni Correction for drug parameters at 2nd injection), n(%). µMann Whitney U test, Med(IQR). Each superscript letter (a,b) denotes subgroups that do not differ significantly from each other at the 0.05 significance level. Type 1 ROP accounted for 96.9% (533) of cases, with 5.4% (29) of these cases located in zone 1 and 94.6% (504) in zone 2. Among these cases, 0.7% (4) were stage 1, 6.6% (35) were stage 2, and 92.7% (494) were stage 3. Plus disease was observed in 95.6% (526) (Table 1). First injections were performed in all type 1 ROP cases (100%, 533) between 36 and 43 weeks (median:39 weeks). Statistically, injections were administered later in zone 2 and more frequently in zone 1 (p < 0.001) (Table 4). Reactivation occurred in 8.1% (43) eyes between 40.5 and 48 weeks and retreatment injections were performed during this period. Retreatment injections were less frequent and occurred at later weeks in cases of type 1 ROP compared to A-ROP (p < 0.001 and p = 0.038). Post hoc Bonferroni correction showed that the IVR retreatment rate was significantly higher in cases of type 1 ROP compared to A-ROP, while IVB and IVA rates were similar between the groups (p = 0.017). Additionally, the total number of injections was significantly lower in cases of type 1 ROP than A-ROP (p < 0.001) (Table 4). A-ROP accounted for 2.9% (16) of cases (Table 1). First injections were administered in all A-ROP cases (100%, 16) at 34 or 35 weeks (median: 34 weeks). These injections were performed earlier compared to type 1 ROP (p < 0.001) (Table 4). Due to Severe Zone 1 findings (with hemorrhages, massive proliferation), the laser treatment was combined with the first injection for 37.5% (6) of the eyes. Reactivation was observed in 62.5% (10) eyes and retreatment injections were administered at 36–42 weeks (median: 40 weeks). Reactivation and retreatment injections occurred earlier in A-ROP than in type 1 ROP (p < 0.001). The interval between injections was statistically similar (p = 0.189). The IVR retreatment rate was significantly lower in cases of A-ROP compared to type 1 ROP, while IVB and IVA rates were similar (p = 0.017). The total number of injections was significantly higher in A-ROP compared to type 1 ROP (p < 0.001) (Table 4). Laser treatment was performed in 15.8% (87) of the cases, with treatment occurring between 35 and 112 weeks (mean: 49.53 ± 13.63 weeks) (Table 1). No statistical differences were observed between the drugs regarding laser treatment rates or timing (p = 0.966 and p = 0.628). For eyes with RD (80%, 8), both injections and laser treatment were performed. Laser treatment was applied in 13% (71) of type 1 ROP cases and 100% (16) of A-ROP cases (Table 1). In this study, RD was observed in 1.8% (10) of the cases. RD occurred between 38 and 120 weeks (mean: 57 ± 34.05 weeks). Among the eyes that developed RD, 44.4% (4) were Type 1 ROP and 55.6% (5) were A-ROP (Table 1). One eye (9%) was directly diagnosed with stage 5a disease. Excluding that case, RD cases had statistically earlier initial injections compared to eyes without RD (p = 0.11) (Table 4). Retreatment was performed on the eyes except for the two cases. Retreatment injection rates were significantly higher in RD cases (70%, 7) compared to non-RD cases (p < 0.001). No statistically significant differences were found among anti-VEGF drugs in terms of the number of injections, intervals between injections, or time to RD development (Table 4). The systemic risk factors of the 9 patients who developed RD are listed in Table 5 and the clinical characteristics of these eyes are detailed in Table 6. Table 5Evaluation of risk factors in 9 patients with retinal detachment.Patient123456789SexMFFMMFMFFGA272828273124252224BW7801020100012701100670930450730MG--- + ----Preeclamsia- + - + - + --CPR--- + + - + --RDS + + + + - + + + + PDA + ----- + --BPD + + + -- + + + + PT + + - + + + + + -Ext + ---- + ---MV + (30 days) + (10 days) + (40 days) + (50 days)- + (23 days) + (30 days) + (35 days) + (37 days)CBAP- + (5 days)- + (3 days)- + (2 days)---Apnea--- + -----Pnomonea + + + --- + --Sepsis + + + --- + + -Hipoglycemia--------Hiperglycemia--------Cafein + + + + + + + + + NEC- + + + -- + --IVH--- + - + ---CS + - + -- + - + + Vitamin A---------PN + (16 days) + (30 days) + (50 days) + (8 days) + (5 days) + (20 days) + (30 days) + (30 days) + (26 days)GA; gestastional age, BW; birth weight, MP; multiple gestation, CPR; cardiopulmonary resuscitation, RDS; respiratory distress syndrome, PDA; patent ductus arteriosus, BPD; bronchopulmonary dysplasia, PT; phototheraphy, EXT;* , MV; mechanical ventilation (and duration), CBAP; *, NEC; necrotizing enterocolitis, IVH; intraventricular hemorrhage, CS; corticosteroid (postnatal), PN; parenteral nutrition (and duration). Evaluation of risk factors in 9 patients with retinal detachment. GA; gestastional age, BW; birth weight, MP; multiple gestation, CPR; cardiopulmonary resuscitation, RDS; respiratory distress syndrome, PDA; patent ductus arteriosus, BPD; bronchopulmonary dysplasia, PT; phototheraphy, EXT;* , MV; mechanical ventilation (and duration), CBAP; *, NEC; necrotizing enterocolitis, IVH; intraventricular hemorrhage, CS; corticosteroid (postnatal), PN; parenteral nutrition (and duration). Table 6Evaluation of 9 patients (18 eyes) with retinal detachment.PatientEyeType 1 ROPA-ROP1. Injweek1Inj2. Injweek2InjCataractLPCLPC weekRD week and stageSurgeryPatient features1R + , Z2, S3,P + -45IVA--- + 112120, S4aLSVLate reactivL + , Z2, S3,P + -45IVA--- + 112--Late reactiv2R---------49, S5aL + V + SBEarthquakeL + , Z2, S3,P + -49IVB--- + 5260, S4bLSVEarthquake3R + , Z2, S3,P + -35IVB47IVB---104, S4bLSVLate reactivL + , Z1, S3,P + -35IVB41IVB- + 104--Late reactiv4R + , Z1, S3,P + -34IVR47IVR- + 47---L + , Z1, S3,P + -34IVR47IVR- + 4752, S4aLSVCrunch Synd5R- + 35IVR36IVR- + 3840, S4bLSVA-ROPL------------6R- + 34IVB35IVB + + 3538, S4bLSVIVI complL- + 34IVB35IVB- + 35---7R + , Z2, S3,P + -34IVA60IVA------L- + 34IVA60IVA- + 6064, S4aLSVA-ROP8R- + 35IVB42IVB + + 4660, S4aLSVCataractL- + 35IVB42IVB- + 46---9R- + 36IVA44IVA- + 5466, S4bLSVA-ROPL------------ROP; retinopathy of prematurity, A-ROP; aggressive ROP, Inj; injection, LPC; laser photocoagulation, RD; retinal detachment, R; right, L; left, Z; zone, S; stage, P; plus disease, IVA; intravitreal aflibercept, IVB; intravitreal bevacizumab, IVR; intravitreal ranibizumab, LSV; lens sparing vitrectomy, L; lensectomy, V; vitrectomy, SB; scleral buckling, Reactiv; reactivation, Synd; syndrome, IVI compl; intravitreal injection complication. Evaluation of 9 patients (18 eyes) with retinal detachment. ROP; retinopathy of prematurity, A-ROP; aggressive ROP, Inj; injection, LPC; laser photocoagulation, RD; retinal detachment, R; right, L; left, Z; zone, S; stage, P; plus disease, IVA; intravitreal aflibercept, IVB; intravitreal bevacizumab, IVR; intravitreal ranibizumab, LSV; lens sparing vitrectomy, L; lensectomy, V; vitrectomy, SB; scleral buckling, Reactiv; reactivation, Synd; syndrome, IVI compl; intravitreal injection complication. Discussion In this study, it was observed that initial injections were administered later and more frequently in eyes with type 1 ROP compared to eyes with A-ROP. The rate of retreatment injections was lower in cases of A-ROP and these injections occurred in later weeks. The retreatment injection rate was found to be higher with IVR. The retreatment injection rates for IVB and IVA were similar. In eyes with A-ROP, the first and second injections were administered earlier compared to eyes with type 1 ROP. The type 1 ROP and A-ROP groups had similar intervals between injections regardless of the drug used. Eyes that developed RD received their first anti-VEGF injections earlier and required more second injections compared to other cases. However, there was no significant difference in the number of injections, intervals between injections, or the time to RD development in eyes with RD. Reactivation, retreatment interval and rate In infants treated with anti-VEGF drugs, peripheral retinal vascularization progresses very slowly and may even remain incomplete in some eyes20,21. Lepore et al. reported that, 4 years after IVB injections for zone 1 ROP, 75% of cases had extensively non-vascularized peripheral retinas; similarly, Karkhaneh et al. noted that 45% of cases still had avascular areas at 90 weeks of GA after IVB injections for zone 2 ROP22. Following anti-VEGF treatments, reactivations and increases in retreatment frequency are often observed after a silent period following full regression, likely due to the “anti-VEGF wear-off effect” or the presence of peripheral avascular retina9. Retreatment for reactivations is mostly required in cases of severe posterior diseases such as plus disease and/or stage 3 ROP and A-ROP, being especially common in A-ROP as it can develop as long as retinal vascularization remains incomplete2,23. Lundgren et al. linked high recurrence rates following anti-VEGF treatment to a higher rate of A-ROP (5.5%) in their study24. In cases involving retreatment, there is a general tendency to continue with the same treatment modality (anti-VEGF or laser therapy) used initially. Although laser therapy may be considered during retreatment to reduce the risk of systemic side effects associated with repeated IVB injections, treatment generally continues with the same anti-VEGF drug22. In our study, the majority of ROP cases were located in zone 2 and these patients presented with stage 3 and plus disease. Although the number of eyes with A-ROP (2.3%, 16) was relatively low, consistent with the literature, these patients had higher rates of retreatment injections, with the first and second injections administered earlier compared to type 1 ROP eyes. In our study, no switching to a different anti-VEGF drug occurred during retreatment. Regardless of the anti-VEGF drug used, laser therapy was not directly applied after retreatment injections unless late reactivation occurred. Reports on reactivation based on GA generally occur between 37 and 60 week25. Reactivations following IVB have been reported at up to 65 weeks, suggesting that longer follow-up may be required compared to IVR2. Fleck et al., in their RAINBOW-based study, reported retreatment after ranibizumab administration of 0.1 mg at a maximum of 50 weeks (mean: 54.5 days) and after administration of 0.2 mg at a maximum of 52 weeks (mean: 53.5 days)26. Mirghorbani et al. found that most reactivations (97%) following IVB occurred within 150 days after treatment, with the risk gradually declining and being significantly reduced after 3 years. They also observed a higher rate of reactivation in cases of zone 1 ROP compared to zone 2 (7.6% vs. 3%) 21. In our study, although late reactivations were observed at 104 and 112 weeks [Patients 1 and 3, Table 6], retreatment injections were administered at a maximum of 60 weeks (44.91 ± 7.05 weeks). We also observed that post-injection reactivations were more common in cases of zone 1 ROP and that injections were performed later in cases of zone 2 ROP (37.9% vs. 6.3%). While the literature primarily focuses on comparisons between anti-VEGF drugs and laser therapy or between single or dual anti-VEGF drugs, there are relatively few studies evaluating reactivation, retreatment rates, and intervals among all anti-VEGF drugs. In the BEAT-ROP study with IVB, the retreatment rate was 8%, while it was 31% in the RAINBOW study with 0.2 mg IVR25,27. Pfeil et al. compared their retreatment rates (Germany, 16.2%) with those of Taiwan (12.9%), the UK (13.1%), and Sweden (30%)9,13,28,29. They reported retreatment rates for IVR (24.5%) and IVB (12%)9. The SWEDROP and Chinese cohort studies reported high retreatment rates for IVR (68.4% and 64%, respectively), while the G-ROP 1 and 2 studies reported rates of 16.5% for both IVR and IVB13,30,31. Smaller case series also evaluated retreatment rates, reporting ranges of 6.8%-14.4% for IVB and 18.75%-64% for IVR 2. For IVA, Chen et al. and Sukgen et al. reported retreatment rates of 11.8% and 13.9%, respectively 32,33. In the recent study by Lee et al., similar to ours, the retreatment rate across three anti-VEGF agents was 9.9%, with rates of 6% for IVB, 22.2% for IVR, and 13.9% for IVA. They highlighted that IVR administration was an independent risk factor for reactivation34. Süren et al. reported retreatment rates of 25.9% for IVB, 37.6% for IVR, and 23.2% for IVA, showing significantly higher retreatment rates and shorter intervals for IVR35. In the meta-analysis by Chen et al., which compared IVA, IVR, IVB, conbercept (IVC), and laser treatment, IVR was found to have higher retreatment rates than other agents. Regarding intervals between treatments, IVR had a significantly shorter interval than IVA but was similar to the others7. Chang et al., in their meta-analysis, emphasized that IVB had lower retreatment rates than IVR and IVA with shorter intervals for IVR (IVA: 12.96 ± 0.47 weeks, IVB: 11.36 ± 0.54 weeks, IVR: 9.29 ± 0.43 weeks)36. In general, retreatment intervals are reported as 12–31 weeks for IVB, 4–16 weeks for IVR27,32. Yang et al. and Jin et al. reported intervals 8.01 ± 3.86 (4–12) weeks and 14.2 ± 1.03 (13–16)weeks for IVA37. The literature suggests that IVR is associated with more frequent retreatments and shorter intervals, likely due to its shorter half-life and faster clearance from the vitreous. Differences in disease severity (e.g., A-ROP with a high risk of reactivation) and chance effects may also contribute to these findings32,38,39. In our study, the overall retreatment rate was 9.6% (53) (Table 7). The interval between injections ranged from 1 to 26 weeks (8.2 ± 5.75 weeks). Retreatment injection rates by medication were 8.8% for IVB, 15% for IVR, and 7.1% for IVA. The retreatment interval by medication was 6 weeks (6–12 weeks) for IVB, 8 weeks (5–12 weeks) for IVR, and 7 weeks (1–17 weeks) for IVA. Statistically, there was no significant difference between the medications in terms of retreatment injection rates or intervals. However, the IVR retreatment rate was significantly higher for type 1 ROP eyes compared to other eyes. Our study, which included 550 eyes, demonstrated a relatively low overall retreatment rate compared to the literature. Also, our retreatment injection rates by medication were lower but retreatment intervals were similar. Factors such as relatively higher GA and BW, a lower number of zone 1 ROP and A-ROP cases, cessation of risk factors in most patients except those who developed RD, and the later timing of the first injection (39.98 ± 5.54 weeks) may have contributed to these outcomes. In alignment with the literature, the higher IVR retreatment rate for type 1 ROP eyes may be due to drug half-life and dosage. Table 7Retreatment rate and interval results of the included studies.StudyYearRetreatment rateRetreatment intervalWong et al. 422015 %83 (IVR) 5.9 weeksMintz-Hittner et al.272016 8% ( IVB) 51.2 weeksRAINBOW262019 15% (0.2mg IVR) 17% (0.1mg IVR) 53.5 days (0.2mg IVR) 54.5 days (0.1mg IVR) Sukgen et al.322019 48.1% (IVR) 13.9 % (IVA) 8.2 ± 0.92 weeks (IVR) 14.2 ± 1.03 weeks (IVA)G-ROP 1 and 2312020 16.5% (IVB)16.5% (IVR)Chen et al.332020 11.8% (IVA)Ling et al.282020 12.9% 10% (IVB) 20.8% (IVR) 8.8 ± 3.9 weeks (IVB) 8.3 ± 1.6 week (IVR) SWEDROP132020 66.7% 50% (IVB) 68.4 (IVR) 11.5/12.8 weeks (IVB) 8.3/9.0 weeks (IVR) FIREFLEYE52022 4.8% (IVA) Süren et al.352022 25.9% (IVB) 37.6% (IVR) 23.2% (IVA) 13±2.37 weeks (IVB) 8±1.7 weeks (IVR) 12±2.47 weeks (IVA) Pfeil et al.92024 15.6% 12% (IVB) 24.5% (IVR) 57 days (IVB) 60 days (IVR) Lee et al.342024 9.9% 6% (IVB) 22.2% (IVR) 13.9% (IVA) 84 ± 45 days Yang et al.37202414.9% (IVR) 15.8% (IVA)Our study20269.6%8.8% (IVB)7.1% (IVA)15% (IVR)8.2±5.75 weeks IVB: Intravitreal bevacizumab, IVR: Intravitreal ranibizumab, IVA: Intravitreal aflibercept. Retreatment rate and interval results of the included studies. 15% (0.2mg IVR) 17% (0.1mg IVR) 53.5 days (0.2mg IVR) 54.5 days (0.1mg IVR) 48.1% (IVR) 13.9 % (IVA) 8.2 ± 0.92 weeks (IVR) 14.2 ± 1.03 weeks (IVA) 16.5% (IVB) 16.5% (IVR) 12.9% 10% (IVB) 20.8% (IVR) 66.7% 50% (IVB) 68.4 (IVR) 11.5/12.8 weeks (IVB) 8.3/9.0 weeks (IVR) 25.9% (IVB) 37.6% (IVR) 23.2% (IVA) 13±2.37 weeks (IVB) 8±1.7 weeks (IVR) 12±2.47 weeks (IVA) 15.6% 12% (IVB) 24.5% (IVR) 57 days (IVB) 60 days (IVR) 9.9% 6% (IVB) 22.2% (IVR) 13.9% (IVA) 14.9% (IVR) 15.8% (IVA) 9.6% 8.8% (IVB) 7.1% (IVA) 15% (IVR) IVB: Intravitreal bevacizumab, IVR: Intravitreal ranibizumab, IVA: Intravitreal aflibercept. GA, BW and comorbid risk factors Although it is crucial for clinicians to understand the predisposing factors for reactivation and the interval of reactivation to ensure proper patient follow-up, these aspects remain incompletely understood. Bas et al. highlighted in their TR-ROP study that low GA and BW are independent risk factors for severe ROP40. As BW and GA decrease, the rates of reactivation and retreatment tend to rise in parallel with the incidence of ROP and treatment-requiring ROP, and they may even correlate with the reactivation interval23,41,42. Wong et al., in their study involving infants with low GA (mean: 23.48 weeks) and low BW (mean: 620 g), reported a retreatment rate of 83% and a reactivation interval of 5.9 weeks42. In our study, the mean GA was 27.59 ± 2.17 weeks and the mean BW was 922.64 ± 199 g. Our retreatment rate was 9.6% (53), with an average interval of 8.2 ± 5.75 weeks between injections. Statistically, we observed that as GA and BW decreased, the first injection was administered earlier and the retreatment injection rate increased. However, no significant relationship was found between the reactivation interval and GA or BW. Additionally, no statistically significant relationship was found between GA, BW, and the anti-VEGF drugs used in our study. Numerous comorbid risk factors are known to be associated with ROP. Persistent or newly added risk factors may lead to reactivation by increasing VEGF levels from avascular retinal areas23,41,43. In our study, as shown in Table 5, all of the patients had comorbid risk factors, and particularly among patients with RD, different ocular or systemic risk factors potentially contributing to persistence were noted. Anti-VEGF, laser therapy and retinal detachment Post-reactivation retreatment with anti-VEGF injections or laser therapy can result in early or late RD. Barry et al. compared anti-VEGF and laser treatments for type 1 ROP before 36 weeks of GA and found RD rates of 0% with anti-VEGF and 7.9% with laser treatment within the first 8 weeks, with no significant difference after 36 weeks. They highlighted the rapid efficacy of anti-VEGF in preventing early progression31. Shah et al. reported similar findings in A-ROP cases44. This difference may be attributed to the extended VEGF presence in the vitreous after laser therapy, while anti-VEGF drugs act immediately upon administration, sustained by their half-life31. Gunzenhauser et al. reported reactivation requiring laser therapy 9 weeks after an IVR injection for type 1 ROP, with RD developing 1 week later. They attributed early and late RD following intravitreal injections to vitreous traction, potential pars plana entry, or excessive laser application45. Similar to our study, laser therapy was preferred following the second injection or for late reactivations to mitigate systemic side effects, persistent avascular retina, and fibrosis-related RD risks31,46,47. Laser therapy was applied in 15.8% (87) of cases in our study at 35–112 weeks, avoiding excessive and high-power applications. RD occurred in 1.8% (10) of the cases between 38 and 120 weeks, including two cases of late-onset RD. Among these, 44.4% (4) were type 1 ROP eyes and 55.6% (5) were A-ROP eyes. RD was statistically associated with earlier initial anti-VEGF administration and higher retreatment injection rates. This may have stemmed from the predominance of A-ROP eyes, patient-specific characteristics, or other contributing factors. Clinical characteristics of the 9 patients with retinal detachment When examining the patients sequentially, it was observed that Patient 1 developed Stage 4a ROP following a single IVA dose despite late reactivation and laser treatment. Even though the patient’s risk factors were no longer present, the long half-life and administered dose of IVA, progression due to remaining peripheral avascular areas following laser treatment, and the direct vitreotractive effects of anti-VEGF or laser treatments may have contributed to late reactivation and subsequent RD development. Patient 2 was diagnosed with stage 5a disease in the right eye upon presentation at our clinic, while stage 4b ROP developed in the left eye despite IVB injection and laser treatment. The patient’s records revealed no signs of ROP requiring treatment prior to the earthquake, during which the patient was trapped under debris, subsequently developing crush syndrome. This syndrome involves systemic and metabolic effects, such as dysrhythmias, rhabdomyolysis, acute renal failure, and sepsis, stemming from muscle and end-organ damage48. Although reports on the ocular effects of crushing injuries include retinal hemorrhages, retinal folds, and retinoschisis, the literature on the ocular impacts of crush syndrome remains limited49,50. Crush syndrome might have directly contributed to persistent progression due to metabolic dysregulation or caused vascular contractional effects by elevating the anti-VEGF drug concentration well beyond therapeutic levels. Patient 3 developed stage 4b ROP in the right eye with late reactivation after two doses of IVB injections to both eyes, and laser treatment was applied to the left eye. Similar mechanisms as in the case of Patient 1 might have been responsible. Patient 4 developed crunch syndrome in the left eye during the fifth week following a second IVR dose and laser treatment to the same eye [Fig. 1]. Crunch syndrome, characterized by progressive tractional retinal detachment following anti-VEGF injections, is associated with proliferative diabetic retinopathy, ROP, and retinal vein occlusion51,52. It typically develops 1 to 6 weeks after injections and is most frequently linked with IVB51. High-dose anti-VEGF use, an increased number of injections, the severity of retinopathy, preexisting fibrosis, large ischemic areas, and significant fibrovascular proliferation are known risk factors51–53. In this case, no fibrovascular proliferation or traction was observed during the examination prior to the second injection. Factors such as increased injection frequency, unseen vitreoretinal adhesion contractions caused by the injection, and the reinforcing effect of laser treatment on anti-VEGF may have contributed to the development of crunch syndrome. To our knowledge, this is the first reported case of crunch syndrome following IVR injection. Patient 5 presented with A-ROP in the right eye while showing no ROP findings in the left eye. After two doses of IVR and laser treatment, stage 4b ROP developed in the right eye. ROP usually progresses symmetrically between the eyes in terms of presence and severity. Lin et al., in their G-ROP studies, noted a very low risk of type 1 ROP developing in the contralateral eye within the first 4 weeks if one eye had type 1 ROP. They emphasized the need for further studies to explain such asymmetry54. Kumawat et al. reported 3.9% asymmetry in ROP among 410 patients, attributing it to local ocular factors and possible mechanisms such as regional variations in retinal neurovascular development and retinal light dose55. The lack of any ROP findings in the other eye of our patient with unilateral A-ROP strongly supports the role of local dynamics in ROP development. Patient 6 received two IVB doses in both eyes due to A-ROP. During the second injection in the right eye, lens capsule perforation occurred as a complication, leading to cataract development. Laser treatment was subsequently added for both eyes, with stage 4b ROP developing in the right eye [Fig. 2]. Cataracts can develop due to mechanical lens injury or as a result of the anti-VEGF drug’s molecular properties without direct contact56,57. Following posterior capsule rupture, inflammatory mechanisms may initiate cataract progression and, upon dislocation of lens fragments into the vitreous cavity, pathologies such as glaucoma, chronic uveitis, retinal detachment, and chronic cystoid macular edema may occur58,59. The inflammation hypothesis is well supported in the literature regarding ROP progression60. We believe that in our patient, the inflammatory mechanism triggered by capsule perforation during injection accelerated ROP progression alongside cataract formation in the right eye. Patient 7 underwent two IVA injections for type 1 ROP in the right eye and A-ROP in the left eye, achieving regression. However, stage 4a ROP developed in the left eye following laser treatment at 60 weeks. Similar factors as in the cases of Patients 1 and 3 could have contributed. Patient 8, with A-ROP in both eyes and mild cataract findings in the right eye, received two doses of IVB and laser treatments for both eyes, leading to stage 4a ROP in the right eye. Preterm infants face a risk of cataract development due to systemic factors, often bilaterally but commonly asymmetrical in some cases57. Similar mechanisms as in the case of Patient 5 might have contributed to both cataract formation and ROP progression. Patient 9 developed stage 4b ROP in the right eye despite two IVA doses and laser treatment for A-ROP, with no ROP findings in the left eye. Similar factors to those seen in the case of Patient 5 could explain this progression. Ultra-widefield (UWF) retinal images with Optos device (Patient 4, left eye). A Type 1 ROP and B reactivation after first intravitreal ranibizumab (IVR) injection. C At the 5th week, tractional retinal detachment (crunch syndrome) following second injection and laser treatment. D Reattachment after lens sparing vitrectomy (LSV) surgery. Ultra-widefield (UWF) retinal images with Optos device (Patient 6, right and left eye). A, B Aggressive retinopathy of prematurity (A-ROP). C Right eye: lens capsule perforation and cataract development during retreatment intravitreal bevacizumab (IVB) injection. D Left eye: second IVB injection. E Right eye: stage 4b ROP and F regression of ROP findings after injections and laser treatment in the left eye. Limitations and conclusion Our study has some important limitations. First, anti-VEGF drugs other than IVB, IVR, and IVA were not included in our study. Moreover, the choice of anti-VEGF drug was made by the patients’ parents or guardians, resulting in a lack of randomization and equal distribution. Further studies with larger sample sizes and longer follow-up periods are needed to better evaluate the outcomes of anti-VEGF injections and to compare all anti-VEGF agents. Additionally, the detailed assessments of our patients who developed RD reflect the need for additional studies to support some of the hypotheses we have proposed with further evidence in the literature. In conclusion this study, consistent with the literature, has demonstrated based on comparisons of anti-VEGF drugs for type 1 ROP that IVR retreatment injections were performed at a significantly higher rate. Retreatment rates were comparable for IVB (8.8%) and IVA (7.1%), but significantly higher for IVR (15%). Injections were administered earlier in cases of zone 1 disease and in later weeks for zone 2. As BW and GA decreased, the initial injection was administered in earlier weeks and the need for retreatment injections increased. Among eyes that developed RD, injections were performed in earlier weeks and the rate of retreatment injections was high. Each patient who developed RD was thoroughly evaluated in our study, and some of our findings, to the best of our knowledge, were discussed for the first time in the literature. While anti-VEGF drugs are frequently the first choice in the treatment of ROP, it is essential to closely and continuously monitor patients over the long term for potential complications that may arise following injections. Author contributions MA, ACU and MKE performed the literature search. MA and ACU were responsible for the study concept and design. MA and ACU collected the data. MA and MKE performed the statistical analysis. MA, ACU and MKE were responsible for the interpretation of the data. MA and MKE drafted the manuscript. All authors critically revised the manuscript for important intellectual content. All authors reviewed and approved the final manuscript. Data availability The data that support the findings of this study are available from the corresponding author, MA, upon reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The data that support the findings of this study are available from the corresponding author, MA, upon reasonable request.