Table 5.
Observational studies comparing neurodevelopmental outcomes between intravitreal anti-VEGF and LPC treatment modalities for TR-ROP
| Authors/first author | Year | Participants | Treatment modalities being compared | Outcomes | Risk of bias assessment with ROBINS-I tool9 |
|---|---|---|---|---|---|
| Arima et al. [53] | 2020 | 53 infants with type 1 ROP |
IVB (0.625 mg/0.025 ml), LPC (type unspecified) |
At 18 months CA, IVB group had: Lower mean score in the Language-Social domain DQ (P = 0.010), even after the adjustment for GA and birth weight (P = 0.030) No differences between groups in KSPD scores at 18 months CA, for: Postural-Movement domain DQ (P = 0.100) Cognitive-Adaptive domain DQ (P = 0.170) Overall domain DQ (P = 0.100) |
Moderate overall risk of bias |
| Raghuram et al. [15] | 2019 | 60 eyes of 34 infants with TR-ROP |
IVB (0.625 mg/0.025 ml), Diode LPC |
At 18–24 months CA, no differences between groups in rates of: Moderate to severe NDI (P = 0.380) Severe NDI (P = 0.120) Cerebral palsy (P = 0.500) Hearing loss requiring amplification (P = 0.690) Motor Bayley scores (P = 0.580) Cognitive Bayley scores (P = 0.830) Language Bayley scores (P = 0.450) |
Moderate overall risk of bias |
| Therani et al. [17] | 2019 | 111 eyes of 64 infants with TR-ROP |
IVB (dose unspecified), LPC (type unspecified) |
At 18–24 months CA, no differences between groups in: NDI scores (OR = 1.63; 95% CI: 0.54–4.87, adjusted OR = 1.77; 95% CI: 0.46–6.73) sNDI scores (OR = 2.19; 95% CI: 0.80–5.98, adjusted OR = 2.31; 95% CI: 0.75–7.14) |
Serious overall risk of bias due to missing information in three domains of bias assessment |
| Kang et al. [36] | 2019 | 314 eyes of 165 infants with type 1 ROP |
IVR (0.25 mg/0.025 ml), Diode LPC |
Mean follow-up: 36.3 ± 31.9 months. No differences between groups in: Rate of adverse neurodevelopmental outcomes (P = 0.73) |
Moderate overall risk of bias |
| Zhang et al. [44] | 2019 | 283 infants with TR-ROP |
IVB (dose unspecified), LPC (type unspecified) |
After 2 years of follow-up, no differences between groups in rates of: CP (P = 0.060, adjusting for IVH) Motor delay (P = 0.540) Cognitive delay (P = 0.680) Language delay (P = 0.060) |
Serious overall risk of bias due to missing information in two domains of bias assessment |
| Chen et al. [54] | 2018 | 49 eyes of 25 patients with TR-ROP |
IVB (0.625 mg/0.025 ml), Diode LPC |
At an average of 20 months CA, no differences between groups in: Rates of neurodevelopmental delay at 20 months CA (adjusted OR = 0.87; 95% CI, 0.08–9.46) |
Moderate overall risk of bias |
| Peyton et al. [47] | 2018 | 22 infants with type 1 ROP |
IVB (dose unspecified), LPC (type unspecified) |
LPC group had: Higher rate of CP (P = 0.005) No differences between groups in Bayley-III standard scores for: Cognitive outcomes (P > 0.050) Language outcomes (P > 0.050) Motor outcomes (P > 0.050) Rate of neurodevelopmental delay (P = 0.090) Rate of hearing loss (P = 0.770) Rate of bilateral vision loss (P = 0.100) |
Serious overall risk of bias due to baseline confounding |
| Lien et al. [52] | 2016 | 62 infants with type 1 ROP |
IVB (0.625 mg/0.025 ml), Diode LPC, Combination of IVB (0.625 mg/0.025 ml) and diode LPC |
24 months after treatment diode LPC group had: Higher mean MDI score (P = 0.030) Higher mean PDI score (P = 0.002) than the IVB + LPC group No differences between three groups in: MDI and PDI scores measured at 6 months, 12 months, and 18 months after treatment (P = 0.380 and 0.830, 0.790 and 0.056, 0.100 and 0.697, respectively) |
Moderate overall risk of bias |
| Morin et al. [55] | 2016 | 125 infants with TR-ROP |
IVB (dose unspecified), LPC (type unspecified) |
At 18 months CA, LPC group had: Higher motor composite score (P = 0.020) Lower rate of neurodevelopmental disabilities (OR = 3.1; 95% CI: 1.2–8.4) No differences between groups in: Language composite scores (P > 0.050) Cognitive scores (P > 0.050) |
Moderate overall risk of bias |
| Kong et al. [22] | 2015 | 80 eyes of 42 patients with type 1 ROP || |
IVB (0.625 mg/0.025 ml), LPC (type unspecified) |
At 1 year CA, no differences between groups in: Gross motor DQ (P = 0.780) Visual motor DQ (P = 0.840) Language DQ (P = 0.840) Cognitive DQ (P = 0.830) |
Serious overall risk of bias due to time-varying confounding |
anti-VEGF anti-vascular endothelial growth factor, LPC laser photocoagulation, TR-ROP treatment-requiring retinopathy of prematurity, ROBINS-I risk of bias in non-randomized studies of interventions, IVB intravitreal bevacizumab, CA corrected age, NDI neurodevelopmental impairment, OR odds ratio, CI confidence interval, sNDI significant neurodevelopmental impairment, ROP retinopathy of prematurity, IVR intravitreal ranibizumab, CP cerebral palsy, IVH intraventricular hemorrhage, MDI mental developmental index, PDI psychomotor developmental index, DQ developmental quotient, GA gestational age, KSPD Kyoto scale of psychological development