Ahmed glaucoma valve implant for refractory glaucoma in children: A systematic review and meta-analysis

Abstract

Purpose

The aim of this study was to evaluate the efficacy and safety of the Ahmed glaucoma valve in pediatric patients with refractory glaucoma.

Methods

A comprehensive literature search was conducted across multiple major databases, including PubMed, Embase, the Cochrane Library of Systematic Reviews, Science Direct, China's National Knowledge Infrastructure, and the Wanfang database. We retrieved studies published before December 2022 that met the inclusion criteria, including clinical controlled trials (randomized controlled trials) and clinical noncontrolled trials (non-randomized controlled trials) on the use of Ahmed glaucoma valve in pediatric patients with refractory glaucoma. We performed a meta-analysis and systematic review. The efficacy measures included intraocular pressure, number of anti-glaucoma medications, visual acuity, and success rate. The safety measures were complications. Statistical analysis was performed using RevMan 5.0 software.

Results

We identified 46 eligible studies: Compared with geographic location and study type, the Ahmed glaucoma valve showed a decrease in postoperative intraocular pressure and number of anti-glaucoma medications compared to preoperative levels in children with refractory glaucoma (P < 0.001). Compared with etiological, the Ahmed glaucoma valve showed a decrease in intraocular pressure after surgery compared to preoperative levels in children with refractory glaucoma (SMD: 14.57, 95% CI: 14.05–15.08, P < 0.00 1), and a decrease in postoperative number of anti-glaucoma medications compared to preoperative number of anti-glaucoma medications (SMD: 1.45, 95% CI: 1.37–1.54, P < 0.001). Compared with trabeculectomy revision surgery, there was no significant difference in the complete success rate between the two groups (SMD: 0.86, 95% CI: 0.52–1.39; P = 0.37).

Overall, the postoperative intraocular pressure at the time of Ahmed glaucoma valve implantation was lower than that at the time of trabeculectomy revision surgery (SMD: 1.01, 95% CI: 0.71–1.31, I²= 99%, P < 0.001). Subgroup analyses based on whether mitomycin C was use d or not. There was a statistically significant difference in intraocular pressure between Ahmed's glaucoma valve surgery and preoperative (SMD: 14.13, 95% CI: 13.47–14.80, P = 0.007). Comparison of cumulative complete success rates of Ahmed S2, S3, and Ahmed FP7, FP8 in Ahmed glaucoma valve surgery (SMD: 0.74, 95% CI: 0.38–1.45, I²= 85%, P = 0.38). There is no statistical difference between the two groups. Choroidal effusion and anterior chamber hemorrhage are the two most common adverse events after Ahmed's glaucoma valve surgery.

Conclusions

The Ahmed glaucoma valve implantation has some effectiveness in reducing intraocular pressure in children with refractory glaucoma, but there are still many complications. Valve model may not be the key factor affecting the postoperative effectiveness and adverse reactions of refractory glaucoma in children.

Introduction

Primary open-angle glaucoma is still one of the main causes of irreversible blindness worldwide. ⁴⁵ When local anti-glaucoma medications fail to control intraocular pressure (IOP) and conventional surgery has a poor prognosis, it is classified as refractory glaucoma (RG). ⁴⁶ The success rate of most modern filtering surgeries for glaucoma ranges from 70% to 90%, but for refractory glaucoma, the success rate is only 11% to 52%. Treatment is more challenging, with a poorer prognosis and severe optic nerve damage in advanced stages, posing a major challenge for ophthalmologists. ⁴⁷

The causes of pediatric glaucoma are complex and diverse, and the cooperation and compliance of children in eye examination and treatment are relatively poor, which increases the difficulty of treatment. This increase may lead to adverse events such as structural damage to child's eyes, optic nerve disease, corneal edema, and amblyopia.^48,49 Although some children can be treated with medication, there are few types of eye pressure-lowering drugs suitable for children, and they have corresponding side effects. In addition, the overall success rate of drug treatment for pediatric glaucoma is poor, so surgery often becomes the preferred option. ²

In the past two decades, several types of glaucoma drainage devices (GDDs) have been developed. Research has shown that GDD implantation is an effective method for treating refractory glaucoma, and in recent years, the number of people using this surgery to treat refractory glaucoma in children has been increasing. Compared to adults, children have smaller eyes and more elastic scleral tissue, which makes GDD implantation more challenging. ⁵⁰ The most common implanted aqueous humor drainage device is the Ahmed glaucoma valve (AGV) implant. It is a one-way flow valve device with a reservoir in a plate that allows flow at a specific IOP value. This mechanism prevents high filtration and early hypotony. The S2/S3 type is made of polypropylene, which is hard and thick, while the FP7/FP8 type is made of silicone, which is soft and thin, with a smooth surface and a gradually narrowing valve disc.

Currently, the evaluation of AGV surgery in the treatment of refractory pediatric glaucoma has been inconsistent. Some specialists report success rates as high as 93% for refractory congenital/infantile glaucoma. The implantation of AGV seems to be a feasible option for treating refractory pediatric glaucoma, yet others report suboptimal outcomes. Tung et al. ⁵¹ found that in some children with refractory pediatric glaucoma, a single Ahmed glaucoma pediatric glaucoma surgery cannot control IOP. The effectiveness of GDD reimplantation for the second time is not ideal, especially with a longer implantation time. After AGV surgery, children may experience abnormal eye movements and strabismus. ⁵²

With the development of randomized controlled trials (RCTs) and evidence-base d medicine, the efficacy of AGV implantation in adult patients has been fully evaluated, and its long-term effects on IOP, visual acuity, and visual field outcomes have been examined. Compared to those of adult glaucoma, the incidence of pediatric glaucoma is lower, and existing studies are limited by small sample sizes and high attrition rates, which restrict the ability to conduct large-scale intervention studies on the efficacy of AGV implantation. However, there are not many domestic reports on the application of AGV implantation surgery in children.^53–57 In this study, we collected existing research data on the treatment of refractory glaucoma in children with AGV implantation and conducted a quantitative analysis, hoping to fill this data gap and provide a reference for future relevant studies.

Methods:

Search strategy and selection of studies

We followed the reporting guidelines and criteria of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.^35,43 We searched PubMed, EMBASE, and the Cochrane Library of Systematic Reviews (up to December 2022) using the following search terms: “Ahmed,” “refractory glaucoma,” “high eye pressure,” “intraocular pressure,” “IOP” “Ahmed glaucoma drainage device,” “glaucoma implants,” “aqueous humor drainage device,” “aqueous humor shunt surgery,” “children,” “adolescents,” “infants,” and other keywords (for the complete search strategy, see Supplemental Table 1).

Inclusion and exclusion criteria

Studies that meet the following criteria are considered to meet our meta-analysis: the inclusion criteria include the following: ① the literature must be published in official journals; ② the sample size > 5 eyes; ③ the study design involves RCT and non-randomized clinical trials (N-RCT) (including prospective studies, cross-sectional studies, and retrospective studies); and ④ eyes diagnosed with glaucoma undergoes AGV surgery; ⑤ at least one of the following reported results: pre and postoperative IOP, pre and postoperative NOAM, visual acuity, success rate, and adverse events; ⑥ no perception of light loss. Exclusion criteria: ① follow-up time < 3 months; ② case reports, reviews, meta-analyses, animal experiments, letters, meeting and full-text abstracts, and duplicate publications; ③ literature with poor quality or incomplete data that cannot extract relevant indicators; ④ non-English literature; ⑤ literature that cannot be accessed in full.

Data extraction

For the data in the literature, two different researchers (GlLi and JHe) independently extracted and cross-checked the data to reach a consensus agreement. If the final results are inconsistent, they will be evaluated and decided by a third-party researcher (XJ Fu). Any differences have been resolved through discussion. The collected information includes the first author, publication year and location, study design (study type), children (quantity, age, sex, race, and etiology of glaucoma), follow-up time, and model of AGV drainage device implantation. The primary outcome measures of this study are the IOP, number of anti-glaucoma medications (NOAM), and visual acuity (VA). Safety assessments include complications (adverse events) and a complete success rate of surgery.

Statistical analysis

Using Rev-Man 5.4 software (Information Management Systems Group, Oxford, United Kingdom), a meta-analysis was conducted to summarize the included literature. The heterogeneity was assessed using Cochran's Q test. If I² > 50% and P ≤ 0.1, it indicated the presence of heterogeneity among the included studies, and the random-effects model was used. Conversely, if the studies showed homogeneity, a fixed-effects model was employed. The relative risk (RR) or standardized mean difference (SMD) was calculated by weighing the individual risk ratio or SMD against its reciprocal variance. IOP and NOAM were continuous variables, while the success rate was a binary variable. The RR was used as the effect size measure for determining the success of surgery, with a significance of P < 0.05 indicating a statistically significant difference.

To assess the robustness of the results, a sensitivity analysis was conducted by systematically excluding individual studies from the meta-analysis to reflect the influence of each study on the summary estimate. The random-effects estimates were generally similar before and after the removal of any individual study, indicating a high stability of the meta-analysis results. Funnel plot analysis demonstrated a symmetrical distribution of the results, indicating the absence of publication bias (see Supplemental Figures 1 to 11).

The etiology of glaucoma is classified into three categories according to the definition of the Childhood Glaucoma Research Network (CGRN) ⁴⁴ : ① primary glaucoma (including primary congenital glaucoma, PCG, and juvenile open-angle glaucoma); ② glaucoma following cataract surgery (GFCS); and ③ all other secondary glaucoma excluding GFCS.

The definition of success varies among different studies published so far.^5,32,33,83 Our definition of success rate in this study, based on a comprehensive review of domestic and foreign literature, covers the range of success rates included in all included literature. The standard for defining surgical success is summarized as follows: postoperative IOP is controlled within 4–23 mm Hg, including treatment with or without glaucoma drugs, and there are no signs of glaucoma progression.

Results

Results of the literature search

The meta-analysis identified 1076 studies through a search strategy, of which 1076 were from PubMed (N = 440), Embase (N = 393), Science Direct (Elsevier) (N = 243), China's National Knowledge Infrastructure (N = 0), and Wanfang database (N = 0). Of these, 337 studies were excluded due to duplicate content. After reading the titles and abstracts, 595 studies were excluded. Abstract sections of potential full-text studies were carefully reviewed, and studies were assessed for compliance (n = 144). Final scrutiny of comparisons excluded studies that failed to download the entire text in its entirety (N = 15); animal and cell studies (N = 6); case reports (N = 3); letters and replies (N = 1); reviews and meta-analyses (N = 27); and meeting abstracts (N = 33) and outcome metrics that did not have at least one of these measurements (pre- and postoperative IOP, pre- and postoperative number of medications administered, visual acuity, success rate, and complications) of the studies (N = 13). After eligibility screening, 46 studies were included for qualitative synthesis (see Figure 1). Table 1 summarizes the characteristics of the included studies.

Figure 1.

PRISMA flow diagram delineating the search, screening, and eligibility assessment process.

Table 1.

The characteristics of the studies included in the meta-analysis.

First author		Year	Study design	Number of eyes	Age	Gender	Ahmed devices studied	Etiologis studied	Success definition	Follow-up time	MMC
1. Balekudaru et al. 1	India	2014	Retrospective	71	6.8 (4.9)	40：31	FP7, S2	Primary, GFCS, secondary	IOP between 6 mm Hg and 18 mm Hg without medications, no loss of light perception, and no loss of vision due to endopthalmitis, choroidal hemorrhage or phthisis during the follow-up period.	＞ 3 m	NR
2. Chen et al. 2	America	2015	Retrospective	119	6.8 (5.7)	61：58	FP7, S2	Primary, secondary	IOP 5–21, 20% reduction from baseline IOP with or without medications	6.1 （3.3）y	NR
3. Das et al. 3	India	2005	Retrospective	122	< 12	NR	S2	NR	IOP of 22 mm Hg or less and 5 mm Hg or more and at least a 30% reduction in IOP without visually devastating complications or additional glaucoma surgery	12.51 ± 8.37 （3–24） m	NR
4. Dave et al. 4	India	2015	Retrospective	11	8	4:4	FP8	Primary	IOP > 5 and ≤ 18 mm Hg during examination under anesthesia with or without medications and without serious complications or additional glaucoma surgery	17.9 ± 9.3 (6.2–35.4) m	NR
5. Djodeyre et al. 5	Spain	2001	Retrospective	35	4.4 (4.7)	18:17	S2, S3	Primary, GFCS, secondary	IOP < 22, no other signs of glaucomatous progression (increased cornea, anteroposterior eye diameters, and cup-to-disc ratio), no additional glaucoma surgeries or medications or devastating visual complications	12.6 ± 10.8 （0–37.9） m	Capsulectomy was performed, in two cases with MMC application, and the IOP decreased to normal limits.
6. Eksioglu et al. 6	Canada	2017	Retrospective	16	14.2 (3.3)	8:3	FP7, S2	Secondary	IOP 6–21 with (qualified success) or without (complete success) antiglaucoma medications and without the need for further glaucoma or tube extraction surgery	64.46 ± 33.56 (12–118) m	NR
7. El Sayed and Awadein—silicone 7	Egypt	2013	Prospective	25	2.9 (5.4)	NR	FP7, FP8	Primary, GFCS, secondary	IOP ≤ 22, no medications, no signs of glaucoma progression	2 y	NR
8. El Sayed and Awadein—polypropylene 7	Egypt	2013	Prospective	25	2.8 (5.6)	NR	S2, S3	Primary, GFCS, secondary	IOP ≤ 21, no medications, no signs of glaucoma progression	2 y	NR
9. Esfandiari et al.—Ahmed 8	America	2019	Retrospective	28	4.4 (2.1)	9:7	FP7	GFCS	IOP 5–21 and ≥ 20% reduction from baseline IOP, without additional glaucoma surgery	4 y	NR
10. Geyer et al. 9	Israel	2021	Retrospective	41	7.9 (7.3)	NR	FP7	GFCS	IOP ≤ 22 (with or without glaucoma medications), without glaucoma reoperation, and without significant visual complications during the follow-up period	61.1 ± 46.5 m	NR
11. Helmy et al. 10	Egypt	2016	RCT	33	1.28 (0.5)	17:16	FP8	Primary	IOP 6–21, without additional medical or surgical treatment, stable corneal diameter, decreased corneal edema, improved corneal clarity, and stable or reversed cup disc/ratio	3.5 y	NR
12. Khan et al.—polypropylene 11	Saudi Arabia	2009	Retrospective	31	0.9 (0.5)	17:14	S1, S2	Primary, GFCS, secondary	IOP ≤ 22 with or without antiglaucoma medications and without visually significant/threatening complications	2 y	Some surgeons routinely used MMC 0.2%–0.4% for 2–4 minutes topretreat the scleral bed and overlying conjunctiva in the area where the valve was to be implanted
13. Khan An and Almobarak—silicone 11	Saudi Arabia	2009	Retrospective	11	1.2 (0.5)	5:6	FP7	Secondary, primary, and GFCS	IOP ≤ 22, with or without antiglaucoma medications and without visually significant/threatening complications	2 y	Some surgeons used MMC 0.2%–0.4% for 2–4 minutes topretreat the scleral bed and overlying conjunctiva in the area where the valve was to be implanted
14. Kirwan et al. 12	Ireland	2005	Retrospective	19	8	7:6	S2	GFCS	IOP ≤ 15, without medications	32 （3–84） m	In 10 eyes 0.5 mg/mL MMC was applied to the sclera for 3 minutes
15. Lee et al. 13	Korea	2015	Retrospective	11	7.1 (7.2)	7:4	FP8, S3	Primary	IOP 5–21, no decrease in vision due to complications or changes in IOP and no need for an additional glaucoma operation	94.6 (55–120) m	NR
16. Mahdy—Ahmed+MMC 14	Egypt	2011	RCT	20	4 (5)	6:8	S3	Primary	IOP 10–21, without antiglaucoma medications or additional glaucoma surgery and without visually devastating complications as endophthalmitis or phthisis bulbi and no loss of light perception	12 m	Preoperative application of MMC (0.4 mg/mL), 3 minutes
17. Mahdy—Ahmed 14	Egypt	2011	RCT	20	5 (4)	10:8	S3	Primary	IOP 10–21, without antiglaucoma medications or additional glaucoma surgery and without visually devastating complications as endophthalmitis or phthisis bulbi and no loss of light perception	12 m	NR
18. Mofti et al. 15	Saudi Arabia	2020	Retrospective	178	5.8 (5.5)	89:89	F7, F8, S2	Primary, Secondary	IOP 5–21, without the need for subsequent glaucoma surgery	4.6 ± 3.1 y	NR
19. Mokbel et al. 16	Egypt	2019	Retrospective	14	NR	NR	Unspecified	Primary	IOP 5–21 with (qualified) or without (complete) medication, with stability of ocular biometric measurement	6 m	NR
20. Morad et al. 17	Canada	2003	Retrospective	60	6 (4.9)	25:19	Unspecified	Primary, GFCS, secondary	IOP 5–21, without medications	24.3 ± 16 (3–60) m	NR
21. Novak-Lauš et al. 18	Croatia	2016	Retrospective	10	< 11	6:4	S3	Primary, GFCS, secondary	IOP 6–21, without medications and no loss of vision due to complications such as choroid hemorrhage or phthisis during the follow up period		NR
22. O’malley Schotthoefer et al. 19	America	2008	Retrospective	38	0.75 (0.2–15)	NR	FP7, S2	Primary, GFCS	IOP ≤ 21, with or without medications onlast follow-up and without severe complication or further glau-coma surgery	5.5 (0.5–10.5) y	NR
23. Ou et al. 20	America	2009	Retrospective	30	1.9 (2.6)	12:7	FP7, S2	Primary	IOP 5–23 and at least a 15% reduction from the preoperative IOP , without serious complications, additional glaucoma surgery, or loss of light perception	57.6 (48.0) m	NR
24. Pakravan et al. 21	Iran	2007	RCT	15	10.9 (5.1)	12:3	Unspeci fied	GFCS	IOP 5–21, with or without medications	13.1 ± 9.7 m	After dissection of conjunctiva and tenon up to the limbus, MMC 0.02% was applied with multiple pieces of soaked sponges over the sclera and under the conjunctiva for 2 minutes
25. Promelle and Lyons 22	Canada	2021	Retrospective	81	6.4 (5.1)	35:28	FP7, S2	Primary, GFCS	IOP 5–21, with no glaucoma medication, no subsequent glaucoma surgery, and no severe complication	10 y	The sponges were soaked with six to seven drops of MMC 0.4 mg/mL and reinserted into the subtenon space, with sponges facing toward tenon’s
26. Razeghinejad et al. 23	Iran	2014	Retrospective	33	2.7 (3.1)	17:5	FP7, FP8 S2, S3	Secondary and primary	IOP 6–21, with or without medication, and no serious complications, additional glaucoma surgery, or loss of light perception	32.6 ± 18.3 m	NR
27. Senthil et al. 24	India	2018	Retrospective	24	2 (1.5–3)	NR	FP7, FP8	Primary	IOP 5–21, with or without topical antiglaucoma medications.	27 （15–39） m	NR
28. Senthil et al.—secondary ac 24	India	2018	Retrospective	41	NR	NR	FP7, FP8	Secondary	IOP 5–21, with or without topical antiglaucoma medications.	27 （15–39） m	NR
29. Soyugelen Demirok et al. 25	Turkey	2019	Retrospective	6	9 (4.6)	3:0	FP7, S2	Secondary	IOP 5–22, with or without glaucoma medication.	19.16 ± 8.8 (12–36) months	NR
30. Speiss and Peralta Calvo 26	Spain	2021	Retrospective	29	NR	10:13	FP7, FP8, S2	GFCS	IOP ≤ 21, with and without adjunctive topical treatment	85.41 ± 56.24 m	NR
31. Al-Torbak 27	Saudi Arabia	2004	Retrospective	20	NR	7:10	NR	Primary	IOP 5–21, with or without antiglaucoma medi-cations and without loss of light perception, further glaucoma surgery, or devastating complications	30.9 ± 11.1 (18–60) m	NR
32. Khan et al. 28	Middle East	2021	Retrospective	70	8.97 ± 5.20	35:35	NR	Primary, GFCS, secondary	IOP 6–21, ≥ 20% reduction in IOP from baseline	25.33 ± 11.03 m	NR
33. Kirwan	Ireland	2005	Retrospective	19	96 m （9–18）	7:6	S2	GFCS			NR
34. Elhefney et al. 29	Egypt	2016	Prospective	22	16.3 ± 9.7 m	10:2	FP8	Primary, GFCS, secondary	IOP < 18, without adjuvant glaucoma medication	24.1 ± 4.3 m	NR
35. Hamush et al. 30	America	1999	Retrospective	11	NR	NR	NR	Secondary	IOP < 21, no additional glaucoma surgery, no expulsive choroidal hemorrhage, and no retinal detachment	910.5 ± 574.1 days	NR
36. Kaushik et al. 31	India	2019	Retrospective	24	7.91 ± 5.02	16∶8	FP8, FP7	Secondary	IOP 5–21, without additional surgical maneuver, and operative complications	2.12 ± 0.87 y	NR
37.Coleman et al. 32	America	1997	Prospective	24	6.61 ± 5.67	11∶10	NR	Primary, GFCS, secondary	IOP < 22 mm Hg for the last 2 follow-ups in eyes with preoperative IOP > 22 mm Hg, or an IOP that was lowered by at least 20% from preoperative values in eyes with preoperative IOP < 22 mm Hg, and no additional glaucoma surgeries or visually devastating complications	16.3 ±11.2 m	NR
38. Englert et al. 33	America	1999	Retrospective	27	6.44 ± 5.88	10∶13	S2	Primary, GFCS secondary	IOP ≤ 21, without visually devastating complications or additional glaucoma surgery (exclusive of tube revision)	3–31 (12.6–8.2 m	NR
39. Yang and Park 34	Korea	2008	Retrospective	34	5.5 ± 4.2	18∶11	S2, S3	Primary, GFCS, secondary	IOP ≤ 21, with or without glaucoma medication	29.1 (3–31) m	NR
40. Al-Omairi et al. 36	Saudi Arabia	2017	Retrospective	44	6.7 ± 3.1	23∶21	FP7	Primary, GFCS, secondary	IOP ≤ 21, with or without medications	8.1± 5.1 m	MMC soaked in Weck-Cel sponges (0.2 mg/mL) was placed in the sub-Tenon space for 3–4 minutes in some of the cases
41. Kafkala et al. 37	America	2005	Retrospective	7	9–13 y (mean, 11 y)	2∶4	S2	Primary, GFCS, secondary	IOP 4–22, with or without glaucoma medications, with no additional glaucoma surgery or visually devastating complications	36.8 (6–60) m	NR
42. Englert et al.—Pakravan-2 38	Iran	2007	Retrospective	30	9.1 ± 4.1 y	T+MMC: 6/7	–	Secondary	IOP 5–21, with or without medications	14.8 ± 11 m	After dissection of conjunctiva and tenon up to the limbus, MMC 0.02% was applied with multiple pieces of soaked sponges over the sclera and under the conjunctiva for 2 minutes
				T+MMC:15	T+MMC:10.9 ± 5.1 y	AGI+MMC:13/2	–		IOP 5–21, with or without medications	13.1 ± 9.7 m	MMC 0.02% was applied by a large piece of sponge over the sclera for 2 minutes
				AGI+MMC:15	AGI+MMC:14.8 ± 11 y		NR				NR
43. Huang et al. 39	China	2015	Retrospective	14	4.29 ± 1.88 y	7:4	FP-8, FP-7	Primary	IOP 6–21, and IOP reduction of at least 30% relative to preoperative values; without the signs of the surgical failure	18.29 ± 10.96 m	NR
44. Elwehidy et al. 40	Egypt	2018	Prospective	19	AGV: 11/8	25:16	–	Primary, GFCS, secondary	IOP < 21 mm Hg, without any sight-threatening complications	24.9 ± 4.8 m	AGV revision consisted of deroofing of the capsule around the plate of the AGV with application of MMC (0.2 mg/mL for 2 minutes) to the sub-Tenon space before closure of the Tenon and conjunctiva in two separate layers
				22	VT: 14/8				IOP < 21 mm Hg, without any sight-threatening complications	25.2 ± 5.8 m	NR
45. Al-Mobarak and Khan An 41	Egypt	2009	Retrospective	31	11.1 ± 5.4 m	16:11	S1, S2, FP7	Primary, GFCS, secondary	IOP ≤ 22, without glaucoma medication and no visually significant/threatening complications		Use MMC 0.2%–0.4% for 2–4 minutes to pretreat the scleral bed and overlying Tenon tissue/conjunctiva in the area where the valve was to be implanted in patients of this age group
46. Beck et al. 42	America	2003	Retrospective	46	7.0 ± 5.1 m	NR	S2, S3	Primary, GFCS, secondary	IOP < 23 mm Hg, no further glaucoma surgery performed or recommended, no devastating complication, and stable ocular dimensions	31.5 ± 22.6 m	MMC in aconcentration of 0.25–0.5 mg/mL was soaked in a 7.5 mm corneal light shield or 4 mm of Weck cell sponge and applied for 3–5 minutes intraoperatively

Abbreviations: SD: standard deviation; IOP: intraocular pressure; RCT: randomized controlled trial; GFCS: glaucoma following cataract surgery; MMC: mitomycin C; NR: not reported; m: months; y: years.

Note: Etiologies classified according to the International Consensus Classification: “primary glaucoma” includes primary congenital glaucoma and juvenile open-angle glaucoma. Glaucoma following cataract surgery (GFCS) was classified separately. All secondary glaucomas other than glaucoma following cataract surgery were classified as “secondary glaucoma.”^44,49

Characteristics of the included studies

A summary of the characteristics of the 46 included studies can be found in Table 1. These studies were published between 1997 and 2021, and the statistical analysis included children from 16 countries. A total of 1614 eyes (1492 children) were included, and 55% were male. The sample sizes ranged from 6 to 178 patients, and the ages ranged between 0.55 and 14.8 years. Based on the study methods, there were six prospective studies, 38 retrospective studies, and four RCT studies. Based on the geographical population s of the study subjects, there were 12 studies on North American populations, six studies on European populations, 20 studies on Asian populations, and nine studies on African populations. Based on the follow-up time, there were one study with a 3-month follow-up, one study with a 6-month follow-up, nine studies with a 12-month follow-up, five studies with an 18-month follow-up, 12 studies with a 24-month follow-up, seven studies with a 36-month follow-up, one study with a 48-month follow-up, and 12 studies with a follow-up of more than 48 months. Classified by whether the patients had undergone preoperative glaucoma surgery, 32 studies included patients with a history of glaucoma surgery, while 14 studies included patients without such a history. Classified by the presence or absence of pseudophakic or aphakic eyes, 23 studies involved patients with pseudophakic or aphakic eyes, and 23 studies involved patients without pseudophakic or aphakic eyes (please refer to Supplemental Table 2 for detailed information).

Intraocular pressure

Changes in IOP before and after surgery (subgroup by geographical location)

A total of 43 studies were included in this analysis, and this study was divided into three ethnic subgroups, Europe, North America, and Asia, according to the geographic location of the study population. Four studies from Europe, 12 studies from North America, and 27 studies from Asia were included. The total study sample consisted of 1498 eyes that underwent preoperative IOP testing for AGV surgery and 1435 eyes that underwent postoperative IOP testing for AGV surgery, as shown in Figure 2.

Figure 2.

Forest plot of comparing IOP before and after AGV surgery (subgroup by geographical location).

IOP: intraocular pressure; AGV: Ahmed glaucoma valve.

After performing a heterogeneity test with an I² > 50% and solidly combining the data using a random-effects model, the IOP was significantly lower after surgical implantation of the AGV surgery (SMD: 14.61, 95% CI: 14.25–14.98, P < 0.001), with a statistically significant difference between the preoperative and post-operative IOP.

The geographic locations of the study populations were Europe (SMD: 11.61, 95% CI: 10.63–12.58, I² = 93%, P < 0.001), North America (SMD: 13.68, 95% CI: 13.06–14.30, I² = 80%, P < 0.001), and Asia (SMD: 14.18, 95% CI: 13.57–14. 79, I² = 88%, P < 0.001), which had lower postoperative IOP than preoperative IOP in AGV surgery.

Changes in IOP before and after surgery (subgroup by study type)

A total of 43 studies were included in this analysis, and these studies were divided into three subgroups, prospective studies, retrospective studies, and RCTs, according to the geographical location of the study population. Of these, five studies were prospective studies, 36 studies were retrospective studies, and two studies were RCTs. The total study sample consisted of 1379 eyes that underwent preoperative IOP testing for the AGV surgery procedure and 1317 eyes that underwent postoperative IOP testing for the AGV surgery procedure, as shown in Figure 3.

Figure 3.

Forest plot comparing IOP before and after AGV surgery (subgroup by study type).

IOP: intraocular pressure; AGV: Ahmed glaucoma valve.

After performing a heterogeneity test with an I² > 50% and solidly combining the data using the random-effects model, IOP was significantly lower after AGV implantation than before (SMD: 14.43, 95% CI: 13.97–14.89, P < 0.001).

The study populations included prospective studies (SMD: 13.06, 95% CI: 11.89–14. 23, I² = 82%, P < 0.001), retrospective studies (SMD: 14.72, 95% CI: 14.20–15.24, I² = 83%, P < 0.001), and RCT studies (SMD: 14. 12, 95% CI: 12.33–15.92, I² = 5 2%, P < 0.001) on the IOP changes before and after AGV surgery, with postoperative IOP being lower than preoperative levels.

Changes in IOP before and after surgery (subgroup by etiological)

A total of 38 studies were included in this analysis, and these studies were divided into four subgroups according to etiology. After performing a heterogeneity test with an I² > 50% and solidly combining the data using the random-effects model, IOP was significantly lower after AGV implantation than before (SMD: 14.57, 95% CI: 14.05–15.08, I² = 92.2%, P < 0.001) (see Figure 4).

1. Among them, the IOP changes before and after glaucoma valve surgery in Ahmed patients with primary + GFCS + secondary disease (SMD: 13.35, 95% CI:12.61–14.08, I² = 79%, P < 0.001) were statistically significant between the two groups.

2. Among them, the IOP changes before and after Primary glaucoma valve surgery in Ahmed (SMD: 15.17, 95% CI: 14.14–16.20, I² = 81%, P < 0.001) were statistically significant between the two groups.

3. Among them, the IOP changes before and after glaucoma valve surgery in Ahmed due to secondary disease (SMD: 18.31, 95% CI: 16.88–19.74, I² = 87%, P < 0.001) were statistically significant between the two groups.

4. Among them, the IOP changes before and after glaucoma valve surgery in Ahmed with GFCS (SMD: 14.32, 95% CI: 12.86–15.78, I²= 81%, P < 0.001) we restatistically significant between the two groups.

Figure 4.

Forest plot of comparing IOP before and after AGV surgery (subgroup by etiological).

IOP: intraocular pressure; AGV: Ahmed glaucoma valve.

Glaucoma medications

Changes in NOAM before and after surgery (subgroup by geographic location)

A total of 36 studies were included in this analysis, and this study was divided into three ethnic subgroups, Europe, North America, and Asia, according to the geographic location of the study population. Two studies were from Europe, 11 were from North America, and 23 were from Asia. The total study sample consisted of 1255 eyes that underwent preoperative NOAM testing for AGV surgery procedures and 1267 eyes that underwent postoperative NOAM testing for AGV surgery procedures, as shown in Figure 5.

Figure 5.

Forest plot of comparing NOAM before and after AGV surgery (subgroup by geographic location).

NOAM: number of antiglaucoma medications; AGV: Ahmed glaucoma valve.

After performing a heterogeneity test with an I²> 50% and solidly combining the data using the random-effects model, the NOAM was significantly lower after surgical implantation of the AGV (SMD: 1.20, 95% CI: 1.12–1.27; I²= 98%, P < 0.001).

Among the study populations geographically located in Asia (SMD: 1.11, 95% CI: 1.02–1.20, I²= 99%, P < 0.001), North America (SMD: 1.48, 95% CI: 1. 30–1.65, I²= 71%, P < 0.001), and Europe (SMD: 0.00, 95% CI: −0.66–0.66, Z = 0, P = 1.000) all had a statistically significant reduction in the NOAM after AGV surgery compared with the NOAM before surgery.

Changes in the NOAM before and after surgery (subgroup by study type)

A total of 36 studies were included in this analysis, and this study was divided into three subgroups according to the geographical location of the study population: prospective studies, retrospective studies, and RCT studies. Of these, 30 studies were prospective studies, four studies were retrospective studies, and two studies were RCTs. The total study sample consisted of 1245 eyes that underwent AGV surgery preoperative IOP testing and 1254 eyes that underwent AGV surgery postoperative NOAM testing, as shown in Figure 6.

Figure 6.

Forest plot comparing NOAM before and after AGV surgery (subgroup by study type).

NOAM: number of antiglaucoma medications; AGV: Ahmed glaucoma valve.

After performing a heterogeneity test with an I²> 50% and solidly combining the data using the random-effects model, the NOAM was significantly lower after surgical implantation of the AGV (SMD: 1.21, 95% CI: 1.14–1.28, P < 0.001), and there was a statistically significant difference between the preoperative and postoperative NOAM.

The study populations included prospective studies (SMD: 1.45, 95% CI: 1.37–1.5 4, I² = 85%, P < 0.001), retrospective studies (SMD: 1.13, 95% CI: 0.86–1.40, I² = 4 0%, P < 0.001), and RCT studies (SMD: 0.47, 95% CI: 0.32–0.63, I² = 98%, P < 0.00 1) on post-AGV surgery NOAM lower than post-surgery NOAM.

Changes in the NOAM before and after surgery (subgroup by etiological)

A total of 32 studies were included in this analysis, and these studies were divided into four subgroups according to etiology. After performing a heterogeneity test with an I²> 50% and solidly combining the data using the random-effects model, NOAM was significantly lower after AGV implantation than before (SMD: 1.4 5, 95% CI: 1.37–1.54, I² = 89%, P < 0.001) (see Figure 7).

1. Among them, the NOAM changes before and after glaucoma valve surgery in Ahmed patients with primary + GFCS + secondary disease (SMD: 1.37, 95% CI: 1.24–1.51, I²= 80%, P < 0.001) were statistically different between the two groups.

2. Among them, the NOAM changes before and after primary glaucoma valve surgery in Ahmed (SMD: 1.05, 95% CI: 0.86–1.24, I² = 47%, P < 0.001) were statistically significant between the two groups.

3. Among them, the NOAM changes before and after glaucoma valve surgery in Ahmed due to secondary disease (SMD: 1.84, 95% CI: 1.68–1.99, I² = 95%, P < 0.001) were statistically different between the two groups.

4. Among them, the NOAM changes before and after glaucoma valve surgery in Ahmed with GFCS (SMD: 1.35, 95% CI: 1.08–1.62, I² = 70%, P < 0.001) were statistically different between the two groups.

Figure 7.

Forest plot of comparing NOAM before and after AGV surgery (subgroup by etiological).

NOAM: number of antiglaucoma medications; AGV: Ahmed glaucoma valve.

AGV surgery versus TB revision surgery

Meta-analysis of IOP during the follow-up period after AGV surgery versus TB revision surgery

A total of eight studies were included in this analysis, and these studies were divided into five subgroups of 1, 6, 12, 24, and > 24 months for analysis according to the duration of follow-up. One study was included in the 1-month follow-up period, one study in the 6-month follow-up period, three studies in the 12-month follow-up period, two studies in the 24-month follow-up period, and one study in the >24-month follow-up period. The study sample consisted of 256 eyes that underwent AGV surgery and 257 eyes that underwent TB surgery.

After performing the heterogeneity test, I² > 50%, using the random-effects model to combine the data, the overall postoperative follow-up IOP after AGV surgery was lower than the postoperative follow-up IOP after TB revision surgery (SMD：1.01, 95% CI: 0.71–1.31, I²= 99%， P ＜ 0.001), there was a statistical difference between the two groups (see Figure 8).

1. At the one-month postoperative follow-up period (SMD： –0.50, 95% CI: –0. 93–0.07， Z = 2.29, P = 0.02), the IOP after AGV surgery was lower than that after TB revision surgery.

2. At the postoperative 6-month (SMD: 6.9, 95% CI: 6.20–7.60, Z = 19.46, P ＜ 0.001), the IOP after AGV implantation was lower than the IOP after TB revision surgery.

3. At the 12-month follow-up (SMD：–0.47, 95% CI: –1.12–0.18, I²= 86%, P = 0. 16), there was no significant difference between the two groups.

4. At the 24-month follow-up (SMD：–0.11, 95% CI: –1.02–0.81, I²= 0%, P = 0.82), there was no significant difference between the two groups.

5. At the more than 24 months follow-up (SMD: 10. 00, 95% CI: 6.88–13. 12, Z = 6.29, P < 0.001), the IOP after TB revision surgery was lower than that after AGV surgery more than 24 months, and there was a statistical difference between the two groups.

Figure 8.

Forest plot of comparing changes in IOP at postoperative follow-up AGV surgery with TB revision surgery (subgroup by follow-up times).

IOP: intraocular pressure; AGV: Ahmed glaucoma valve; TB: trabeculectomy.

Forest comparison of postoperative complete success rate between AGV surgery and TB revision surgery (stratified by follow-up time)

These included a 12-month follow-up period (RR: 0.45, 95% CI: 0.18–1.12, I² = 0%, P = 0.09), a 24-month follow-up period (RR: 0.84, 95% CI: 0.27–2.61, I² = 0%, P = 0.76), a 36-month follow-up period (RR: 1.44, 95% CI: 0.44–4.73, P = 0.55), and a 48-month follow-up period (RR: 1.14, 95% CI: 0.50–2.64, I² = 0%, P = 0.75). There was no significant difference between the two groups (RR: 0.8 6.14, 95% CI: 0.52–1.39, I² = 3.7%, P = 0.37) (see Figure 9).

Figure 9.

Forest comparison of postoperative complete success rate between AGV surgery and TB revision surgery (subgroup by follow-up time).

TB: trabeculectomy; AGV: Ahmed glaucoma valve.

Comparison of surgical VA before and after surgery

Five studies were included in this analysis, and the total study sample consisted of 103 eyes that underwent preoperative VA testing for AGV surgery and 124 eyes that underwent postoperative testing for AGV surgery. After performing a heterogeneity test with an I² > 50%, the random-effects model was used to combine the data (see Figure 10). Overall, there was no significant difference in the incidence of VA after AGV surgery (SMD: 0.08, 95% CI: −0.04–0.19, I² = 93%, P = 0.21) between the two groups.

Figure 10.

Forest plot of comparing surgical VA before and after AGV surgery.

VA: visual acuity; AGV: Ahmed glaucoma valve.

Complications

Choroidal effusion and hyphema were the two most common adverse events after AGV surgery. Due to the many adverse reactions and the insufficient number of indicators for each outcome to be used in the meta-analysis, meta-analysis for this aspect was not conducted in this study. Adverse events in the Ahmed group are shown in Supplemental Table 3.

Comparison of AGV surgery with preoperative and postoperative IOP (subgroup by MMC)

A total of 17 studies were included in this analysis, and the total study sample consisted of 537 eyes that underwent preoperative IOP testing for AGV surgery and 185 eyes that underwent postoperative IOP testing for AGV surgery, with a total of 12 studies without MMC application and five studies with MMC application. After performing a heterogeneity test with an I² > 50% and solidly combining the data using the random-effects model, IOP was significantly lower after AGV implantation than before (SMD: 14.13, 95% CI: 13.47–14.80, I² = 86.4%, P = 0.007) (Figure 11).

1. Among them, a statistically significant difference between preoperative and postoperative IOP comparisons for AGV surgery without MMC application (SMD: 14.64, 95% CI: 13.88–15.40, I² = 46%, P < 0.001).

2. Among them, a statistically significant difference between preoperative and postoperative IOP comparisons for AGV surgery with MMC application (SMD: 12.45, 95% CI: 11.07–13.84, I² = 84%, P < 0.001).

Figure 11.

Forest plot of comparing IOP before and after AGV surgery (subgroup by MMC).

IOP: intraocular pressure; AGV: Ahmed glaucoma valve; MMC: mitomycin-C.

Comparison of the complete success rates of Ahmed S2, S3 and Ahmed FP7, FP8 in AGV surgery

This analysis included a total of three studies with a combined study sample of 112 eyes undergoing AGV surgery in the S2, S3, and 106 eyes undergoing AGV surgery in the FP7 and FP8 models. After conducting a test for heterogeneity, with an I² value > 50%, the random effects model was used to combine the data (see Figure 12). There was no statistically significant difference in the postoperative complete success rate bet ween these two different material models (RR: 0.74, 95% CI: 0.38–1.45, I² = 85%, P = 0. 38).

Figure 12.

Forest plot of comparing complete success rates of S2, S3 models and FP7, FP8 models in Ahmed glaucoma valve (AGV) surgery.

Discussion

IOP and medication

Ou et al. ²⁰ reported that the success of glaucoma surgery was lower in Hispanic and female populations. Malik et al. ⁵⁸ also observed differences in the racial distribution of glaucoma patients, stating that Middle Eastern children have invasive scars above the eyelid margin, which can affect the success of surgery. Experts have long suggested that race is a significant risk factor for glaucomatous damage and that glaucomatous disease is population-specific and sensitive.^59,60 North America has a rich variety of ethnic groups, and the fundamental reason for this is that it is difficult to classify the proportion of ethnic groups and people. The literature included in this study did not classify the North American patients included in the statistics by race. Therefore, we are unable to determine the specific proportion of ethnic groups included in the study. This is the main reason why this study conducted subgroup analysis based on the geographical location of the study population, rather than distinguishing by race. Some of the published articles on evidence-based medicine have also indicated that including literature in systematic reviews and predominantly retrospective meta-analyses will undoubtedly have the potential to cause recall bias, and the type of study included in the literature is also necessary. Considering the essential effects of geographic location and type of study, in this study, after evaluating the efficacy of AGV implantation surgery before and after surgery (in terms of changes in IOP and NOAM) and after explicitly excluding the heterogeneity of geographic location and type of study, it was found that the implantation of the AGV surgery was effective in decreasing the IOP and NOAM for children on several continents, such as Asia, Europe, and North America, which is consistent with the findings of some previous studies.^61–63

Currently, there is controversy regarding the impact of different models of Ahmed glaucoma drainage valves. Some relevant studies have suggested that the reduction in IOP after glaucoma valve implantation is related to the patency of the drainage tube, the surf ace area of the filtering bleb, and the permeability of the filtering bleb capsule. Previous experimental results have shown that, under certain conditions with other factors constant, the hypotensive effect is proportional to the surface area of the drainage plate. Some studies have suggested that the drainage plate of Ahmed FP7/FP8 glaucoma valves is mad e of silicone material, which has a smoother surface and softer texture. This results in a higher eye pressure control rate and a lower incidence of fibrous encapsulation. Previous studies have shown that differences in these two models of biomaterials are related to the hypotensive effect and success rate of AGV surgery.^7,11 Wang et al. ⁶⁴ reported that the effectiveness of Ahmed FP7/FP8 in the treatment of refractory pediatric glaucoma is superior to that of the S2/S3 models. In our study, when we included models of valve s made of a single material in a subgroup, there were statistically significant differences between the preoperative and postoperative IOP for each material model. However, the fin al combined summary results showed no differences. Additionally, we compared the success rates of polypropylene and silicone valve models in the literature and found no differences. This finding suggests that different valve material types may not be the main factors influencing the hypotensive effect and success rate.

Etiological subgroups

Adult glaucoma is mainly characterized by chronic open-angle glaucoma, while primary congenital glaucoma is the main type of glaucoma in children. In the past, differences in the type of glaucoma between children and adults can directly affect the surgical outcome of GDD implantation. Cataract ultrasonic emulsification and extracapsular cataract extraction can reduce IOP, especially for patients with shallow anterior chambers.^65,66 The effectiveness of these procedures in combination with glaucoma implant devices is uncertain. Additional surgeries may increase the risk of adverse events and increase the heterogeneity of outcomes. The methods and frequency of preoperative surgeries we included in this study also varied. So it is not possible to divide them into different subgroups to discuss whether the previous number of surgeries had an impact on the surgical outcome. So in order to eliminate the interference of heterogeneity and better identify differences in results based on etiology, we divided them into subgroups according to etiology. We analyzed the changes in preoperative and postoperative IOP and medication dosage under the etiological grouping, and the results showed statistical differences. It indicates that the etiology is a factor that affects the IOP and medication efficacy of AGV surgery, and indirectly proves the reliability of our etiology grouping data. Unfortunately, due to the insufficient number of indicators and literature for each outcome, this study did not conduct a meta-analysis on its correlation with success rate.

Comparison of the results with those of TB revision surgery

Currently, there is limited research on the comparison of AGV surgery and TB revision surgery in children with refractory glaucoma. Some experts have found that AGV drainage is more suitable for children with preexisting conjunctival scarring and thin sclera. The filtration bleb is far away from the limbus, theoretically reducing the risk of severe complications such as infection compared to traditional TB surgery. Additionally, scholars have pointed out that due to the smaller size of children's eyes and frequent damage to anatomical structures during surgery, traditional TB surgery in children presents specific technical challenges and can easily lead to serious complications such as iris and ciliary body entrapment and vitre. For children with refractory glaucoma, although the use of traditional TB surgery or TB surgery combined with antimetabolite drugs has a higher initial success rate in primary childhood glaucoma, their efficacy is somewhat limited in secondary glaucoma, and the filtration area is prone to scar formation, resulting in a low long-term success rate.^71–73 Oguz et al. found in their comparison of traditional TB surgery and AGV implantation that the success rate was higher after AGV surgery. However, another study found that success rates after both surgeries were similar. ⁷⁴ Our study also showed no differences in success rates between the two surgical methods. In our comparison of postoperative IOP follow-up between AGV surgery and TB revision surgery, we observed differences in the postoperative follow-up of IOP at 1 month, 6 months, and more than 24 months after surgery. However, due to the limited number of included literature, heterogeneity interference cannot be completely ruled out.

Fewer studies have compared the results of AGV surgery with those of TB surgery in children with refractory glaucoma. Some experts have shown that the AGV is more suitable for children with conjunctival scarring and thin sclera, where the filter bubble is farther away from the corneoscleral limbus. Theoretically, the probability of complications such as severe infections will be lower than that of TB surgery. Moreover, scholars have stated that because children have smaller eye s than adults and anatomical structures are often damaged during surgery, TB surgery has particular technical difficulties during childhood surgery and quickly leads to severe complications such as iris and ciliary body embedding and vitreous loss.^67–70 For refractory glaucoma in children, while TB surgery alone or TB surgery combined with antimetabolic medications has high initial success rates in primary pediatric glaucoma, these treatments have somewhat limited efficacy in secondary glaucoma. They are prone to scarring of the filtration zone and low long-term success rates.^71–73 In their results comparing TB surgery and AGV implantation, Oguz et al. reported that AGV implantation had a greater success rate after surgery.

Another study reported similar success rates after both surgeries. ⁷⁴ There are m any types of revision surgeries for TB nowadays, including repair surgery after previous failed TB, viscotrabeculoomy, etc. These surgeries are a renovation of traditional TB. We did not include traditional TB in this study, but rather as a postoperative salvage and repair surgery for TB. We collectively refer to this type of surgery as reconstructive surgery on top of TB. Our study also revealed no difference between the success rates of the two surgical modalities. We found a difference in postoperative follow-up IOP between the AGV surgery group and the TB surgery group at 1, 6, and > 24 months, postoperatively. The heterogeneous interference issue could not be completely ruled out due to the limited literature.

Complications

According to reports, the incidence of postoperative complications in children with glaucoma implanted with GDD is 50%, ⁷⁵ and the range of surgical interventions for postoperative complications is 9%–50%.^27,33,73,76 Most reports indicate that complications related to AGV implantation, such as larger wound healing reactions and higher IOP differences in children after surgery, are more common in childhood than in adulthood.^77–79 Compared to nonvalve glaucoma drainage implants, AGV exhibits fewer complication s in the early postoperative period, with the most common complications being high IOP or hypotonia.^47,80,81 Later-stage complications often include wound leakage, choroidal hemorrhage, macular cystoid edema, iris prolapse, and adhesions.^67,69,82 There is a significant risk of late-stage complications such as fibrous encapsulation, corneal decompensation, and tube misplacement, especially in younger children.

Some research has reported that complications such as choroidal effusion and anterior chamber hemorrhage in children after surgery can resolve on their own. ³⁰ However, the most common reasons and risk factors for increased IOP and surgical failure are the growth of fibrovascular tissue inward and tube exposure. ² Among the 46 included studies, 13 mentioned the occurrence of severe postoperative adverse events, such as endophthalmitis. One study mentioned fibrous encapsulation in the context of complications, nine studies mentioned choroidal effusion, eight studies mentioned tube exposure, 10 studies mentioned retinal detachment, nine studies mentioned choroidal detachment, 10 studies mentioned hyphema, seven studies mentioned shallow anterior chamber, five studies mentioned hypotony, one study mentioned high IOP, one study mentioned cyst formation, and one study mentioned corneal edema. Our combined results of complications in children with refractory glaucoma implanted with AGV indicate a higher probability of complications after surgery, leading to a higher risk of surgical failure and secondary surgery.

Success rate

The definitions of success after surgery in different studies differ among published studies.^5,32,33,83 For our definition of success, we have compiled and considered domestic and foreign literature encompassing the range of success rates included in this study. We have categorized the criteria for defining surgical success as postoperative IOP control between 4 and 23 mm Hg, including both treatment and untreated with glaucoma medication, and no signs of glaucoma progression.

Previous studies have reported that the one-year success rate of GDD in pediatric glaucoma patients is 57%–86%.^84,85 In the literature included in this study, the one-year success rate in pediatric refractory glaucoma patients was 46%–100%, the two-year success rate was 43%–93%, the three-year success rate was 49.2%–71%, the four-year success rate was 17%–83%, the five-year success rate was 33%–70%, and the 10-year success rate was 36.8%–42%. In recent years, with a better understanding of the disease, early diagnosis, and improvements in surgical techniques, the success rate of PCG surgery has increased dramatically. This rate is higher than the success rate of secondary glaucoma surgery. Secondary glaucoma has a higher failure rate and usually requires additional surgical intervention. ⁷⁶ In the studies included in our present systematic review, the one-year success rate for children with PCG was 64%–97%, the two-year success rate was 43%–93%, and the four-year success rate was 17%–83%. The one-year success rate for children with secondary glaucoma was 56.3%–91%, and the four-year success rate was 42.2%–75%. The one-year success rate for children with GFCS was 31.58%–95. 1%, the two-year success rate was 63.9%–89.8%, and the four-year success rate was 54%–71%. In our present pooled comparison of success rates for subgroups of glaucoma patients, we also found that the complete success rate of PCG was greater than the surgical success rate of other types of childhood glaucoma.

The overall success rate of our current study is generally consistent with previous findings: the postoperative success rate of the use of the AGV as a surgical treatment option for childhood glaucoma after refractory lens resection is moderate.^26,58 Similar to other GDD implantation surgeries, the success rate decreased with increasing follow-up time, especially for some specific pediatric glaucoma types, such as uveitic glaucoma, for which the success rate was relatively lower.

Mitomycin C

The impact of MMC use on the efficacy and adverse effects in pediatric patients with refractory glaucoma is not fully understood. ⁵⁸ In classic TB surgery, the use of anti-scarring agents to reduce fibrosis improves surgical success. Nevertheless, the same advantage has yet to be established in AGV surgery, and there are conflicting results in the published literature.^12,14,41 Thirteen studies in our current study used MMC, and 34 (nearly 2/3) did not. We compared the preoperative and postoperative intraocular pressures of AGV surgery with and without MMC and found a statistically significant difference (P < 0.001) in the changes in IOP before and after surgery with and without MMC. The final convergence result showed statistical differences (P = 0.007). We infer that MMC may be a key factor affecting the postoperative efficacy of refractory glaucoma in children.

Advantages and disadvantages

Our study's main strengths are that it is the first systematic review and meta-analysis of AGV implantation outcomes in a refractory pediatric population. Second, for the first time, we compared the postoperative follow-up IOP and success rates of AGV surgery and conventional TB surgery. Third, we compared the effects of silicone and polypropylene materials in the AGV on IOP and postoperative success rates and whether the combined use of MMC affects the postoperative effect. These open questions raised in previous studies were answered for the first time.

There are several limitations to our study. First, the small sample size and incomplete population baseline data in the included literature weakened the validity of this test. Second, the degree of surgical success and complication criteria for the included studies are not currently standardized in the industry. ⁸⁶ Third, not all of the children included in the selected studies were first implanted with AGVs, and studies of children's eyes with prior AGV implantation were not excluded. This is because conjunctival scarring from the last ocular surgery may limit the success of repeat glaucoma surgery in children. Fourth, the scarcity of postoperative follow-up IOP data for AGV surgery and TB revision surgery has increased the heterogeneity of the results.

Conclusion

In summary, AGV implantation surgery is a reliable method for reducing IOP and inhibiting use in children with refractory glaucoma, but the postoperative complication rate remains high, with a decreasing long-term success rate. The use of MMC during AGV implantation surgery in children may be the main factor affecting postoperative efficacy and safety, but the valve model is not. To determine the effectiveness and safety of AGV implantation in treating glaucoma, longer follow-up periods and larger sample sizes of RCTs (especially in European countries) are needed.