Systemic outcomes of anesthesia method in anti-vascular endothelial growth factor therapy for retinopathy of prematurity: General vs topical

Abstract

BACKGROUND

The type of anesthesia needed while performing an anti-vascular endothelial growth factor (VEGF) injection in preterm infants with retinopathy of prematurity is not well standardized. In Oman, general anesthesia (GA) remained the first choice by ophthalmologists treating retinopathy of prematurity (ROP) infants at the time of starting the study. Short-term systemic outcome in GA vs topical anesthesia (TA) is the primary outcome of this study.

AIM

To analyze the short-term systemic outcomes of GA in infants undergoing anti-VEGF injections vs bedside injections under TA.

METHODS

This is an observational study with historical comparison. Estimated sample size is 36 (18 in each group), based on the anticipated rate of difference in the outcome between the two cohorts of 50%, type I error (α) probability of 5%, power of 80%, and the allocation ratio of 1:1. The statistical test considered was Fisher’s exact test. The calculation was done using the G*Power version 3.1.9.2. Infants were from two main tertiary hospitals in Oman.

RESULTS

A total of 36 infants with retinopathy of prematurity were enrolled in the study: 18 in the retrospective (GA) cohort and 18 in the prospective (TA) cohort. The Primary outcome-the need for oxygen and/or respiratory support during the 48 hours post-procedure-occurred in 27.8% of the GA cohort, as compared to none of TA cohort (P value = 0.045). Cost and duration of the procedure, including anesthesia, were noted to be higher in the GA group. Surgeons’ preference revealed that 70% prefer topical vs 30% general in terms of safety, 75% topical vs 25% general in terms of logistics, and 100% prefer topical in terms of cost and staff efforts.

CONCLUSION

TA is associated with a reduced need for oxygen and/or respiratory support within 48 hours post-procedure, and better performance in duration, cost, and surgeon preference compared to GA.

Key Words: Retinopathy of prematurity; General anesthesia; Topical anesthesia; Anti- vascular endothelial growth factor injection

Core Tip: To date, no studies have compared the short-term systemic effects of general anesthesia (GA) in infants undergoing Anti- vascular endothelial growth factor injections vs those treated at bedside with topical anesthesia. This study addressed this gap by retrospectively analyzing outcomes under GA and comparing them to prospective data from bedside procedures under topical anesthesia. Topical anesthesia as compared to GA is associated with a significantly reduced need for oxygen and/or respiratory support within 48 hours post-procedure. In addition, topical anesthesia showed better performance in duration, cost, and surgeon preference.

INTRODUCTION

The type of anesthesia used during anti-vascular endothelial growth factor (VEGF) injections in preterm infants with retinopathy of prematurity (ROP) varies, depending on institutional protocols. Options include general anesthesia (GA), bedside procedures with sedation, or using only topical anesthetic drops. In Oman, GA has typically been the preferred choice, often requiring the transfer of unstable preterm infants to tertiary care centers for safety. However, a group from Harvard Medical School recommended bedside injections to maintain normal feeding schedules and avoid the systemic risks associated with GA[1].

Research has also highlighted the pain associated with ROP screening and treatment. A study by Mattar et al[2] evaluated pain in infants treated under topical anesthesia (TA) using the premature infant pain profile and concluded that TA, combined with soothing measures like pacifiers, is a safe and effective method for treating ROP. Another study found that while tetracaine was effective as a topical anesthetic, the pain was primarily caused by the use of the eyelid speculum, not the injection itself[3].

Sedation without intubation has been proposed as a less invasive alternative to GA, though it remains understudied[4]. Tokgöz et al[5] suggested sevoflurane inhalation as a suitable alternative for anti-VEGF injections without major complications. GA, while effective, has been linked to short-term complications like apnea and cardiorespiratory instability and long-term effects on neurodevelopment. There is a 30% chance of post-operative apnea in preterm infants under 44 weeks of gestational age, and even a single exposure to anesthesia has been associated with cognitive and language development delays[6].

Animal studies have established the neurotoxic effects of anesthesia on developing brains, but findings in humans remain inconsistent. While some studies link early anesthesia exposure to increased risks of cognitive impairment, others show no significant association[7-10].

To date, no studies have compared the short-term systemic effects of GA in infants undergoing anti-VEGF injections vs those treated at the bedside with TA. Our study will address this gap by retrospectively analyzing outcomes under GA and comparing them to prospective data from bedside procedures under TA.

MATERIALS AND METHODS

Design and settings

A prospective cohort study with historical comparison. The study was conducted at two tertiary care medical institutions in Oman.

Study population

The sample size was estimated at 36 (18 in each group). The estimation was based on the anticipated rate of difference in the outcome between the two cohorts of 50%, type I error (α) probability of 5%, power of 80%, and the allocation ratio of 1:1. The statistical test considered was Fisher’s exact test. The calculation was done using the G*Power version 3.1.9.2. Inclusion criteria were infants with ROP undergoing anti-VEGF injections under GA and infants with ROP indicated for anti-VEGF injections under topical anesthesia. Exclusion criteria were suspected sepsis/illness in the previous 48 hours before the procedure, congenital anomalies/syndromic and patient receiving narcotics. A standardized information sheet was used for both groups to elicit the clinical parameters 48 hours post-procedure.

An additional customized survey was used to assess surgeons’ preferences for GA vs TA developed by the research team based on four criteria: Safety, logistics, cost, and procedure location. The survey was distributed to 15 ophthalmology surgeons involved in injecting ROP infants.

Ethical considerations

Ethical approval was obtained from both ethical committees in the allocating sites. No direct harm was suspected to the participants. Consent formed was used as per institutional protocol for any surgical intervention.

Statistical analysis

The data was analyzed with IBM SPSS Statistics, Version 26.0. Armonk, NY, United States: IBM Corp. Independent samples t-test and χ² test were used to compare the association between variables. A P value < 0.05 is considered statistically significant. The risk ratio and cost-effectiveness ratio were also assessed.

RESULTS

Demographic data

Total of thirty-six infants with ROP were included, divided equally between the two cohorts of GA and TA, out of whom 19 (52.8%) were males and 17 (47.2%) were females. A total of 12 (33.3%) infants were treated in Al-Nahdha hospital either directly or referred from other Ministry of Health institutions 16 (44.4%), and Sultan Qaboos university hospital 8 (22.2%). The mean gestational age for GA and TA cohorts was 27.18 and 25.13 weeks respectively (P = 0.020). A significant slight difference in mean birth weight was observed (0.99 kg vs 0.7 kg, P = 0.004) (Table 1).

Table 1 Demographic data of study population, n (%).

	Total (n = 36)	GA (n = 18)	TA (n = 18)	P value
Females	17 (47.2)	12 (66.7)	5 (27.8)	0.044
Males	19 (52.8)	6 (33.3)	13 (72.2)	0.044
ANH	12 (33.3)	9 (50)	3 (16.7)	0.151
SQUH	8 (22.2)	5 (27.8)	3 (16.7)	0.151
Others	16 (44.4)	4 (22.2)	12 (66.7)	0.151
Gestational age (weeks), mean (SD)	Not applicable	27.18 (2.4)	25.13 (1.6)	0.020
Birth weight (kg), mean (SD)	Not applicable	0.99 (0.25)	0.71 (0.14)	0.004

GA: General anesthesia; TA: Topical anesthesia; ANH: Al-Nahdha hospital; SQUH: Sultan Qaboos university hospital.

Primary outcomes

There was occasional apnea reported in both cohorts, but more commonly in the GA cohort (16.6% vs 5.6%, P = 0.6). However, the difference was statistically insignificant. The need for oxygen within 48 hours post-procedure was seen in 5 (27.8%) infants of GA cohort as compared to 1 (5.5%) infant of the TA cohort. However, the need for respiratory support within 48 hours post-procedure was seen in 5 (27.8%) infants of the GA cohort as compared to 0 (0%) in TA cohort. There were no significant differences observed in clinical parameters, including respiratory rate, heart rate, blood pressure, and oxygen saturation 48 hours post-procedure (Table 2).

Table 2 Primary outcome and clinical parameters 48 hours post-procedure, n (%).

	GA	TA	P value
Episodes of apnea	3 (16.6)	1(5.6)	0.601
Need for oxygen 48 hours post-procedure	5 (27.8)	1 (5.6)	0.177
Need for respiratory support 48 hours post procedure	5 (27.8)	0 (0.0)	0.045
RR (breath/minute), mean (SD)	42.5 (10.8)	47 (6.7)	0.187
HR (beat/minute), mean (SD)	144.8 (9.0)	150 (10.1)	0.771
Syst. p(B) (mmHg), mean (SD)	80.3 (7.5)	77.9 (13.6)	0.046
Dia. P(B) (mmHg), mean (SD)	41.3 (5.9)	50 (17.3)	0.001
Oxygen saturation (%), mean (SD)	98.4 (2.1)	96.8 (2.5)	0.443

HR: Heart rate; GA: General anesthesia; RR: Respiratory rate; TA: Topical anesthesia; p(B): Blood pressure.

Risk and risk difference analysis

The risk ratio for adverse events-defined as oxygen need and respiratory support 48 hours post procedure- was calculated and is presented in (Table 3). Among infants with ROP who were exposed to TA, the risk of having an adverse event was 72% less than that among infants with ROP who were exposed to GA.

Table 3 Risk ratio between two groups.

	Topical anesthesia	General anesthesia	Total
Adverse outcome (+)	1	5	6
Adverse outcome (-)	17	13	30
Total	18	18	36
Risk in TA (R1)	1/18 = 0.05
Risk in GA (R0)	5/18 = 0.28
Risk difference	0.05-0.28 = -0.23
Risk ratio	0.05/0.28 = 0.18

GA: General anesthesia; TA: Topical anesthesia.

Secondary outcomes

Duration of the procedure: The Average duration of the procedure was 15 minutes in TA cohort compared to 55 minutes in GA cohort with a difference of 40 minutes (Figure 1).

Figure 1 Calculated duration of the procedure general anesthesia vs topical anestheisa. GA: General anesthesia.

Cost of the procedure:Table 4 Illustrates the differences in the list of items considered for the cost calculation in each type of procedure, specifically in terms of medications, items, and human resources. The details of the average cost of each item are in the GA method are presented in Table 5, with a total cost of 2722 $. All values are average prices in the private and government sectors. The human resources needed to perform the injection in TA required three personnel, which is contrary to 7-8 persons while performing it under GA (Tables 4 and 5).

Table 4 Summary of list of items considered in the cost calculation of each type of procedures.

	General anesthesia	Topical anesthesia
Pre-op	Medication (Lucentis)	Medication (Lucentis)
	Anesthesia consultation +/- any subspeciality asked by anesthesia
Intra-op	Anesthesia doctor + anesthesia nurse + medications + instruments	NICU staff
	OR nurse/ophthalmic	NICU nurse
	Surgeon	Surgeon
	Shifting patient from the ward to OR
	OR time 30-60 minutes
Post-op	Post-op recovery	NICU nurse observation
	Shifting from OR to ward

NICU: Neonatal intensive care unit; OR: Operation theater.

Table 5 Summary of the cost of each stage of the procedure in the general anesthesia group.

	GA	Cost
Pre-op	Anesthesia consultation +/- any subspeciality asked by anesthesia	130 $
Intra-op	Anesthesia doctor + medications + instruments	390 $
	OR time 30-60 minutes	2150 $
Post-op	Post-op recovery	52 $
Total	2722 $

GA: General anesthesia.

Surgeon preference

A total of 12/15 (80%) responded to the procedure preference survey; 25% peadiatric ophthalmologists, and 75% are retina specialists. In terms of the safety of the procedure and systemic condition of preterm infants, 70% preferred TA vs 30% preferred GA. In terms of logistics of the hospital, including booking and preparation of the infants, 75% preferred TA vs 25% preferred GA. In terms of cost and staff effort (nurses, peadiatrician, anesthesia), 100% preferred TA and none preferred GA. Finally, regarding preference for location of the procedure, 63.6% preferred to do it under TA in minor operation room (OR)/SCBU/NICU, and 36.4% preferred to do it in the main OR under GA (Figure 2).

Figure 2 Surgeons’ preference survey results between general anesthesia vs topical anesthesia in injecting ant-vascular endothelial growth factor. GA: General anesthesia.

DISCUSSION

The type of anesthesia used for the different treatment modalities for ROP and associated conditions systemic outcomes were studied in the literature. The use of TA alone for cryotherapy can result in more severe cardiorespiratory complications and systemic instability[11]. Another study showed that TA for laser treatment is not safe in comparison to systemic GA/sedation[12]. Although TA was found to be a good alternative for pain relief in anti-VEGF injections, there was no well-established data regarding systemic outcomes, as it is less painful than cryotherapy and laser treatment.

Systemic well-being during and post-procedure is one of the main concerns in infants with ROP. This study objectively analyzed systemic well-being by examining the need for oxygen and respiratory support within 48 hours post-procedure. This was found to be significantly increased in the GA cohort (27.8%) compared to only 5.5% in the TA cohort. Some studies evaluated the faster return to normal baseline in terms of vitals; however, we did not find it statistically significant between the two groups in our study. For example, Barry et al[13] found that infants who received the injection under bedside sedation returned to their baseline faster than those who were treated with laser under GA. However, it is well established that the two modalities of treatment are not equal in terms of the amount of stress and pain caused. An observation found in our data is that the TA cohort had lower birth weight compared to GA, which can be used to explain that those infants will be less likely to tolerate GA in comparison to bigger infants. After calculating the risk ratio among the infants who were exposed to TA, the risk of having an adverse systemic outcome was 72% less than infants who were exposed to GA (Table 3).

In addition to establishing systemic safety, it is important to know the cost-effectiveness of both methods. Our study showed that performing the anti-VEGF injection under TA lasted for a significantly shorter duration (15 minutes vs 55 minutes). The cost was estimated to be much lower in the TA cohort, by simply avoiding all the costs needed to perform the procedure under GA. Theoretically, if there are 100 anti-VEGF injections planned in future, it will save 272200 $ which is the cost of GA alone (Table 5). However, we believe that the cost is underestimated because it was calculated based on the local government and private prices of each step in the procedure and with the help of well-established operation theatre room cost per minute in the literature[14]. Another consideration is the human resources needed. Around 8 personnel are required to do the injection under GA, compared to only three personnel for the topical procedure.

In the surgeon’s preference survey performed in this study, 60% preferred TA to GA in the 4 assessed domains (Figure 2). However, 36.4% of them preferred to do the injection in the main operation theatre/room compared to the bedside. Interestingly, 100% of them preferred TA in terms of cost and staff efforts. Overall, the study has shown a preference for using TA while injecting anti-VEGF compared to GA in most studied aspects. The calculated cost-effectiveness ratio overall was 544.2 $/avoided outcome. Based on the available data from 36 infants with ROP in this study, the anti-VEGF injection under TA is safer systemically than GA, and it is more cost-effective. Although the side effects of the procedure itself under TA were not studied, it is important to mention that there were no reported ocular adverse effects during or after the injection under TA in the 18 eyes injected.

The strength of this study lies in its main objective of comparing two different anesthesia interventions for anti-VEGF in preterm infants with ROP.

This study has some limitations that are worth discussing. The sample size was small, which makes it a limitation, as a larger sample size is likely to show more significant differences in outcomes between the two cohorts. The study mixed design with one cohort being prospective, and the GA is retrospective, with the lack of randomization of the interventions carrying the risk of bias. However, by the end of the study, the national practice had already moved to injecting the ROP infants under TA and avoiding GA since surgeons noted its feasibility. Moreover, this will prevent the need to transfer unstable infants to a tertiary hospital for the injection when the procedure can be done at the bedside under TA by a trained surgeon.

In future, further randomized studies are recommended based on the data we found to expand on the return to normal baseline vitals, long-term effects and ocular outcomes of the injections themselves.

CONCLUSION

Anti-VEGF injections for infants with ROP are safe when performed under GA and TA. However, infants showed systemically better outcomes in the topical cohort in terms of the need of oxygen or respiratory support within 48 hours post-procedure. In addition, TA showed better performance in procedure duration, cost, and surgeon preference. Future randomized studies are recommended to further confirm and expand these findings.