EFFICACY AND SAFETY OF AMNIOTIC MEMBRANE TRANSPLANTATION IN VITRECTOMY FOR EXTRA-LARGE OR LONG-STANDING MACULAR HOLES

Supplemental Digital Content is Available in the Text.

The authors use amniotic membrane transplantation to treat extra-large or long-standing macular holes (MH) and evaluate its efficacy and safety. Eight eyes from eight patients underwent vitrectomy with amniotic membrane transplantation over the MH. All patients' MHs are closed and macular function are improved. This method maybe an useful treatment.

Abstract

Purpose:

To evaluate the efficacy and safety of amniotic membrane transplantation during vitrectomy for extra-large or long-standing macular holes (MH).

Methods:

A single-center, noncomparative prospective observational study included eight eyes from eight patients with extra-large or long-standing MHs (six recurrent, two primary). All patients underwent vitrectomy with amniotic membrane transplantation over the MH. Baseline assessments included best-corrected visual acuity, optical coherence tomography, multifocal electroretinography, macula microperimetry, and adaptive optics imaging to evaluate photoreceptor cells on the graft surface. Outcome measures were recorded during follow-up.

Results:

All eight eyes achieved MH closure during the follow-up period, with 1 case developing amniotic membrane graft displacement that did not cause MH recurrence. After the follow-up ended, at 18 months postoperatively, this case experienced a recurrence of the MH due to rhegmatogenous retinal detachment(mean follow-up: 14.3 ± 8.1 months). Baseline MH diameter was 825.2 ± 265.9 µm. Best-corrected visual acuity improved in four eyes (50%) and stabilized in others. Multifocal electroretinography showed increased amplitude and shortened peak time at 1 month postoperatively, peaking at 3 months and stabilizing by 6 months. Microperimetry revealed improved retinal sensitivity in five eyes (62.5%), with enhanced 2° and 4° fixation rates. Absolute scotomas resolved posttreatment. Adaptive optics imaging confirmed photoreceptor cell attachment on the graft surface. No adverse events related to the amniotic membrane were observed.

Conclusion:

Amniotic membrane transplantation facilitates extra-large or long-standing MH closure, improves macular function, and demonstrates favorable safety. This approach may enhance photoreceptor recovery, supporting its clinical utility.

Idiopathic macular hole (MH) is a common macular disorder characterized by a full-thickness break in the foveal neurosensory retina, accompanied by significant visual acuity loss, metamorphopsia, and other symptoms. Epidemiological studies report an incidence of idiopathic MH ranging from 0.02% to 0.7%.¹ However, surgical closure fails or recurs in some cases due to large hole size, prolonged duration, or other factors. These challenging cases with low surgical success rates are termed extra-large or long-standing MHs.

Current research focuses on improving surgical techniques to enhance closure rates for extra-large or long-standing MH. Previous studies have demonstrated successful closure using approaches such as expanded internal limiting membrane (ILM) peeling, ILM flap inversion/insertion, autologous lens capsule transplantation, autologous serum application, or neurosensory retinal transplantation.^2–7 Nevertheless, these methods face limitations in reoperation scenarios due to difficulties in harvesting autologous tissues or procedural complexity.

Amniotic membrane exhibits unique bioactive and prohealing properties, including secretion of antiinflammatory factors, suppression of proinflammatory cytokines, antiscarring and antiadhesion effects, and provision of extracellular matrices for cell migration and proliferation.^8–13 It has been successfully applied in treating persistent corneal epithelial defects, neurotrophic keratitis, bullous keratopathy, corneal band keratopathy, and corneal perforations.^14–18 In addition, it has shown efficacy in oral mucosal repair, cartilage regeneration, chronic wound/ulcer management, urethral reconstruction, and pericardial patching.^19–25 In this prospective observational study, we evaluate the efficacy and safety of biological amniotic membrane transplantation for extra-large or long-standing MH closure.

METHODS

This was a single-center, noncomparative, prospective observational case study. The study protocol was approved by the Medical Ethics Committee of Tianjin Medical University General Hospital (IRB2022-YX-034-01). All participants provided written informed consent before enrolment.

Participants

Patients diagnosed with extra-large or long-standing MHs at the Department of Ophthalmology, Tianjin Medical University General Hospital between October 2020 and October 2023 were included. Based on inclusion and exclusion criteria, eight eyes from eight patients were ultimately enrolled. We define “long-standing MHs” as MHs that recur after initial surgical closure and remain open ≥3 months A MH with a diameter of ≥800 µm is defined as “extra-large”.

Inclusion criteria: (1) Age 18 years and older, regardless of sex; (2) diagnosis of primary extra-large or long-standing MH with a minimum diameter ≥800 µm; (3) history of vitrectomy with ILM peeling, but with recurrent MH postoperatively (minimum diameter ≥200 µm). Provision of written informed consent. For patients with both eyes meeting inclusion criteria, the study eye was selected by the investigator.

Exclusion criteria: (1) Age younger than 18 years; (2) primary MH <800 µm or recurrent MH <200 µm; (3) secondary MHs due to other vitreoretinal diseases; (4) pathologic myopia-associated MHs (defined as axial length ≥26.0 mm, corneal refractive power >6.00 D, posterior staphyloma on B-scan, or posterior retinal/choroidal atrophy with scleral exposure on fundus imaging); (5) pregnancy or lactation; (6) history of psychiatric disorders; (7) known allergy to amniotic tissue.

Ocular exclusion criteria: (1) Active ocular infection in either eye; (2) severe media opacity (e.g., corneal scarring) in the study eye affecting imaging; (3) history of non–MH-related macular diseases in the study eye; (4) history of glaucoma, ocular hypertension, or glaucoma surgery in the study eye; (5) history of uveitis in the study eye; (6) congenital structural abnormalities in the study eye.

Patients with systemic conditions deemed unsuitable by investigators were excluded.

Assessment

All eligible clinical cases underwent detailed preoperative ophthalmic examinations, including: best-corrected visual acuity (BCVA), noncontact tonometry (NCT), slit-lamp biomicroscopy, dilated fundus photography or ultra-widefield fundus imaging, optical coherence tomography (OCT), multifocal electroretinography (mfERG), and macular microperimetry.

Postoperative outcome measures: (1) Integrity and position of the amniotic membrane graft; (2) presence or absence of hypopyon or inflammatory exudate in the anterior chamber.

Follow-up protocol

Patients were followed for 12 months postoperatively. Evaluations at 1, 3, 6, and 12 months included BCVA, NCT, OCT (ZEISS Cirrus HD-OCT 5000 device), macular microperimetry (Nidek, Gamagori, Japan/MP-1), and mfERG (ROLAND RETI-Port/Scan 21). Adaptive optics (AO) imaging was used to assess photoreceptor cell distribution over the amniotic membrane graft (Robotrak Technologies Co., Ltd.  Nanjing, China, AOSLO System).

Interventions

All enrolled cases underwent surgery performed by the same surgeon. Patients aged 50 years and older with lens opacity at study enrolment received combined phacoemulsification with intraocular lens implantation. Patients younger than 50 years with clear lenses underwent vitrectomy for MH repair without cataract surgery.

Vitreoretinal surgery

Vitrectomy was performed with triamcinolone acetonide injection for vitreous staining. For primary MHs, indocyanine green (ICG) was used to stain and peel the ILM within the retinal vascular arcades. For recurrent MHs, repeat staining ensured sufficient ILM peeling.

Amniotic membrane transplantation

ICG staining identified the epithelial side of the AM. An AM implant (35 × 35 mm; Ruitai Biological Technology Co., Ltd., Guangzhou, China) was prepared according to the MH size. The epithelial side was aligned with the injector and implanted into the MH during fluid–air exchange. Postoperatively, the vitreous cavity was filled with 14% C₃F₈ gas, and proper positioning of the AM implant was confirmed (Figure 8).

Fig. 8.

Amniotic membrane transplantation procedure: (A) Inject ICG-stained ILM into the MH area, peel off the ILM around the MH, and perform air–fluid exchange. B. Prepare the amniotic membrane graft, and stain the epithelial surface of the amniotic membrane with ICG to help determine its epithelial surface after implantation into the MH. C. Cut the amniotic membrane graft to a diameter equal to that of the MH. D. Place the trimmed amniotic membrane graft into the injector (red arrow). E. Insert the injector into the eye and push the amniotic membrane graft out at the MH to tamponade (not cover) the MH to prevent graft movement (red arrow). F. Confirm the correct position of the amniotic membrane implantation using intraoperative OCT and ICG staining of the amniotic membrane (see Figure 8 for the surgical step diagram and Video 1 and 2, Supplemental Digital Content 1 and 2, https://links.lww.com/IAE/C978 and https://links.lww.com/IAE/C979). No PFO was used during the surgery.

Outcomes Measure

Regular ophthalmic examinations were performed within 2 weeks postoperatively. Examinations included: (1) NCT; (2) evaluation of anterior chamber status under slit-lamp microscopy; (3) indirect ophthalmoscopy under mydriasis to assess the amniotic membrane graft in the macular area.

Follow-ups were conducted at 1 month, 3 months, 6 months, and 12 months postoperatively. Outcomes included: (1) BCVA; (2) NCT; (3) OCT; (4) (mfERG); (5) macular microperimetry; (6) AO imaging; (7) adverse events, such as elevated intraocular pressure, hypotony, retinal hemorrhage or exudates, and vascular occlusion. Lens opacity in phakic patients after vitrectomy was not considered an adverse event.

Sample Size

Epidemiological studies have shown that the incidence of idiopathic MHs ranges from 0.02% to 0.7%.¹ However, in real-world clinical practice, the incidence of extra-large or long-standing MHs may be lower than the rates reported in these epidemiological studies. To date, no specific epidemiological research has focused on extra-large or long-standing MHs. Consequently, the actual sample size collected in this study was limited, with 8 cases included. For a noncomparative prospective observational study, this sample size is deemed feasible.

Statistical Methods

Data statistics were performed using GraphPad Prism 8 and IBM SPSS Statistics 23. Images and tables were edited with PS and GraphPad Prism 8.

Results

Participants Characteristics

This study included eight eyes from eight patients with extra-large or long-standing MHs who underwent vitrectomy combined with amniotic membrane transplantation. As summarized in Table 1, the mean age of patients was 64.6 ± 12.2 years, all female. Among them, 4 cases (50%) had a history of hypertension, 1 case (12.5%) had Type 2 diabetes mellitus, and 1 case (12.5%) had coronary atherosclerotic heart disease. Three cases (37.5%) involved the left eye and 5 cases (62.5%) the right eye. Six cases (75%) were recurrent MHs after previous surgery, while 2 cases (25%) were primary MHs. Seven eyes (87.5%) were phakic and one eye (12.5%) was pseudophakic. The mean disease duration was 30 months (range 3–84 months). The minimum MH diameter averaged 825.2 ± 265.9 µm. Postoperative follow-up averaged 14.3 ± 8.1 months, with the longest observation period reaching 22.4 months.

Table 1.

Baseline Demographic Characteristics of Patients With Extra-large or Long-standing Holes

No.	Gender	Age	Eye	Medical History	Primary/Recurrent	BCVA	IOP (mmHg)	Diameter of MH (μm)
1	F	66	Left	HBP	Recurrent	0.2 (Snellen 20/100)	17	830
2	F	66	Right	HBP	Primary	0.1 (Snellen 20/200)	13	811
3	F	70	Left	HBP, DM	Primary	0.12 (Snellen 20/160)	19	903
4	F	36	Right	None	Recurrent	FC	13	932
5	F	72	Right	HBP, CAHD	Recurrent	0.12 (Snellen 20/160)	13	635
6	F	62	Left	None	Recurrent	0.1 (Snellen 20/200)	13	873
7	F	72	Right	None	Recurrent	0.04 (Snellen 20/500)	13	341
8	F	73	Right	None	Recurrent	0.1 (Snellen 20/200)	16	1,277

CAHD, coronary atherosclerotic heart disease; DM, diabetes mellitus; F, female; HBP, high blood pressure; IOP, intraocular pressure.

Case 1

Female, 70 years old, extra-large or long-standing MH (Primary), hole diameter 903 µm (Figure 1).

Fig. 1.

Follow-up outcomes of the patient preoperatively and at 1, 3, 6, and 12 months postoperatively included visual acuity, OCT, macular microperimetry, and mfERG. Postoperative visual acuity remained stable, and OCT at 12 months postoperatively revealed closure of the MH with the amniotic membrane graft intact (white arrow). Macular microperimetry at 1 and 3 months postoperatively demonstrated improved central fixation (red arrow), and distinct potential wave peaks emerged on mfERG at 6 months postoperatively (yellow arrow).

Case 2

Female, 36 years old, extra-large or long-standing MH (Recurrent), hole diameter 932 µm (Figure 2, A and B).

Fig. 2.

A. Patient's visual acuity demonstrated significant improvement at 12 months postoperatively. OCT revealed closure of the MH (white arrow). Microperimetry at 12 months postoperatively showed marked improvement in central fixation compared with preoperative measurements (red arrow). mfERG revealed significant electrical potential peaks at both 1 and 12 months postoperatively, with a more pronounced peak observed at 12 months (yellow arrow). B. Postoperative fundus photographs showed displacement of the amniotic membrane graft (white arrow in fundus photograph of the right sequence). OCT demonstrated that the amniotic membrane graft remained subretinal below the MH area throughout the follow-up period (white arrow in OCT images of the right sequence). The MH remained closed during the entire follow-up period (OCT images in left sequence and OCT in A).

Case 3

Female, 72 years old, extra-large or long-standing MH (Recurrent), hole diameter 635 µm (Figure 3).

Fig. 3.

This patient experienced a mild decrease in visual acuity at 1 and 3 months postoperatively, but vision returned to preoperative levels by 6 months and remained stable at 12 months postoperatively. OCT at 12 months demonstrated MH closure with the amniotic membrane graft in place (white arrow). Macular microperimetry revealed significant improvement in retinal sensitivity at 6 months compared to earlier measurements, which remained stable through 12 months postoperatively (red arrow). The most prominent improvement in waveform amplitudes on mfERG was observed at 1 month postoperatively (yellow arrow).

Changes in visual acuity and Complications

In this study, 7 cases (87.5%) of extra-large or long-standing MHs demonstrated retained amniotic membrane and hole closure during follow-up, while 1 case experienced amniotic membrane graft displacement. However, the patient did not experience MH recurrence after the displacement, and the MH remained closed throughout the follow-up period. After completing the follow-up for this study, at 18 months postoperatively, the MH recurred due to rhegmatogenous retinal detachment involving the macula. As summarized in Table 2, the preoperative BCVA (LogMAR) was 1.14 (Snellen 20/276). Visual acuity gradually improved with extended follow-up, reaching 0.94 (Snellen 20/174) at the 6-month postoperative follow-up. Among the cases, four showed improved postoperative BCVA compared with preoperative baseline measurements, while four maintained stable BCVA postoperatively. Intraocular pressure remained within normal ranges for all patients preoperatively and postoperatively. During follow-up, 2 cases developed cataracts requiring phacoemulsification with intraocular lens implantation, both procedures being uneventful. From postoperative day 1 through the end of follow-up, no complications—including hypopyon, inflammatory exudates in the anterior chamber, purulent vitreous cavity, retinal vascular occlusion, or retinal hemorrhage/edema—were observed in any study participants.

Table 2.

Postoperative Follow-up of Visual Acuity and Complications

	Preop	1 month post	3 months post	6 months post
BCVA (LogMAR)	1.14	1.13	1.02	0.94
Snellen visual acuity	20/276	20/270	20/209	20/174
Complications
MH-related retinal detachment	1
Cataract	2

Changes in OCT

The postoperative healing patterns of MHs are classified as follows: Type I healing—Reconstruction of the foveal laminar retinal structure within the outer limiting membrane (OLM) without hyperreflective or hyporeflective masses perpendicular to the retinal structure (Ia: continuous OLM; Ib: discontinuous OLM); Type II healing—Presence of hyperreflective masses perpendicular to the foveal structure without intraretinal cysts in the foveal area (IIa: partial continuity with foveal laminar retina; IIb: complete discontinuity of all foveal laminar retina); Type III healing—Presence of intraretinal cysts within the foveal structure showing discontinuity with all foveal laminar retinal layers.²⁶

In our study cohort (Figure 4), postoperative healing patterns demonstrated: At 1 month postoperatively – 1 case of Type Ib, 2 cases of Type IIa, 1 case of Type IIb, and 4 cases of Type III; At 3 months postoperatively – 2 cases of Type Ib, 2 cases of Type IIa, 1 case of Type IIb, and 3 cases of Type III; At 6 months postoperatively – 2 cases of Type Ib, 2 cases of Type IIa, 1 case of Type IIb, and 3 cases of Type III (P = 0.069).

Fig. 4.

A. Patient OCT demonstrates Type Ib healing, preoperative (upper) and 6-month postoperative (lower). B. Patient OCT shows Type IIa healing, preoperative (upper) and 6-month postoperative (lower). C. Patient OCT reveals Type IIb healing, preoperative (upper) and 6-month postoperative (lower). D. Patient OCT displays Type III healing, preoperative (upper) and 6-month postoperative (lower). E. Follow-up and changes in MH healing patterns.

Changes in mfERG

To evaluate changes in macular electrophysiological activity after biological amniotic membrane treatment, we used mfERG to assess photoreceptor cell potential levels in patients after MH closure.

The study results (Table 3) demonstrated that the amplitude of the R1 wave increased at 1 month postoperatively, peaked at 3 months postoperatively, and stabilized by 6 months postoperatively. The peak time of the R1 wave (P1) occurred earlier compared with preoperative measurements. The amplitude of the R2 wave gradually increased postoperatively, reaching its peak at 3 months and stabilizing by 6 months, although its peak time showed no significant shortening compared with preoperative values. For the R3-R5 waves, no notable amplitude elevation was observed postoperatively, with stability maintained at both 3 and 6 months, and no significant shortening in peak time compared with preoperative measurements.

Table 3.

Changes in mfERG and P1

	R1	R2	R3	R4	R5
Response densities (nV/deg2)
Pre	26.34 ± 13.86	21.79 ± 11.60	19.87 ± 6.84	13.63 ± 5.02	7.54 ± 2.61
Post-1 month	36.62 ± 22.48	22.39 ± 10.46	17.72 ± 5.84	12.61 ± 4.15	8.82 ± 3.12
Post-3 months	41.07 ± 20.40	27.84 ± 9.41	22.15 ± 8.20	14.04 ± 5.55	9.18 ± 4.31
Post-6 months	33.58 ± 13.91	25.97 ± 16.39	18.06 ± 7.97	10.24 ± 3.50	5.66 ± 2.07
P (one-way ANOVA)	0.673	0.4145	0.2338	0.387	0.7875
P1 IT m seconds
Pre	47.38 ± 6.71	41.85 ± 2.96	42.37 ± 0.71	42.33 ± 1.12	41.52 ± 1.18
Post-1 month	37.51 ± 9.19	42.76 ± 5.50	40.65 ± 4.40	40.10 ± 3.82	40.43 ± 3.33
Post-3 months	39.76 ± 5.23	41.44 ± 2.71	41.49 ± 3.20	41.49 ± 3.25	41.19 ± 3.35
Post-6 months	42.47 ± 8.00	44.73 ± 2.83	44.10 ± 1.73	44.47 ± 2.47	44.80 ± 2.46
P (one-way ANOVA)	0.1755	0.4688	0.3843	0.123	0.1723

Figure 5 demonstrates a case of extra-large or long-standing MH successfully treated with amniotic membrane transplantation, showing macular mfERG findings before and after treatment. At 1 month post-treatment, the macular potential peak amplitude exhibited an increase, while at 3 months post-treatment, the potential peak amplitude demonstrated stabilization in the macular region.

Fig. 5.

mfERG of patients with extra-large or long-standing MH before surgery (A), 1 month after surgery (B), and 3 months after surgery (C).

Changes in Macular Microperimetry

To further evaluate functional changes in the macula after amniotic membrane treatment, we assessed macular sensitivity and fixation stability using the macular integrity assessment (Table 4 and Figure 6). Observations revealed that the mean macular sensitivity increased from 14.29 ± 9.01 dB preoperatively to 16.35 ± 7.12 dB at the 6-month postoperative follow-up. No significant increase was observed in the number of fixation points within the 1° macular area. The fixation points within the 2° area increased from 75 ± 15 preoperatively to 81 ± 15 postoperatively; however, neither difference reached statistical significance (P > 0.05).

Table 4.

Statistical Results of Macular Microperimetry before and after Vitrectomy for MH Treatment

Parameters	Pre	Post-1 month	Post-3 months	Post-6 months	P (one-way ANOVA)
MMS (dB)	14.29 ± 9.01	14.53± 6.69	16.10 ± 8.81	16.35 ± 7.12	0.875
P1 (%)	43 ± 23	42 ± 2	49 ± 22	37 ± 14	0.4382
P2 (%)	75 ± 16	78 ± 17	82 ± 18	81 ± 15	0.5808
63% BCEA	8.59 ± 4.20	8.49 ± 5.20	7.99 ± 7.05	7.50 ± 4.12	0.7561
95% BCEA	25.70 ± 12.53	25.41 ± 15.55	23.94 ± 21.13	22.48 ± 12.4	0.7579

P1 and P2 represent the number of fixating points which was in circles of 1° and 2° radius, respectively.

BCEA, bivariate contour ellipse area; MMS, mean macular sensitivity.

Fig. 6.

Macular integrity assessment and OCT results of MH before and after treatment.

Changes in Photoreceptor Cells in the Macular Area

This study used AO technology to observe whether photoreceptor cells were attached to the surface of the amniotic membrane graft. We compared differences in photoreceptor cell density within the macular region between surgical eyes with extra-large or long-standing MHs and their contralateral eyes. Figure 7 demonstrates that the photoreceptor cell density in the MH area was significantly reduced compared with the macular region of the contralateral eye. Despite the marked reduction in photoreceptor cell density, cellular migration and attachment to the surface of the amniotic membrane graft were observed.

Fig. 7.

Adaptive optics imaging of photoreceptor cells: Highlights correspond to photoreceptor cells. Figures A–C show AO images after amniotic membrane transplantation for extra-large or long-standing MH. White arrows indicate photoreceptor cells attached to the amniotic membrane graft. Figures D–F present AO images of the contralateral eye, with red arrows highlighting normal photoreceptor cells in the macular region.

Discussion

The ILM coverage technique has emerged as a popular surgical approach for MHs. Several comparative clinical studies have evaluated ILM peeling alone versus ILM coverage for large MHs, demonstrating that the coverage technique significantly improves postoperative MH closure rates and visual acuity compared with ILM peeling alone.^3,27 However, other studies report that while ILM coverage effectively achieves anatomical closure and visual improvement in large MHs, the visual outcomes do not differ statistically from those of ILM peeling alone.^28–30 Subsequent modifications, such as radial or temporal ILM coverage, have been explored, but no correlation has been found between coverage patterns and final visual outcomes.^31,32 Intriguingly, some clinical observations suggest that ILM coverage may even result in inferior visual recovery compared with ILM peeling.³³ Notably, Bleidißel³¹ and Ramtohul³⁰ highlighted the association between the integrity of the outer retinal layers and BCVA, suggesting that restoration of outer macular structures is critical for functional recovery. Although ILM coverage aids MH closure, its role in enhancing local macular function remains debated.

In addition, the ILM flap technique exhibits certain limitations in managing recurrent MHs. For cases where the ILM was adequately removed during the initial surgery, reoperation faces challenges in obtaining sufficient ILM material. ILM flaps harvested outside the vascular arcade, lacking a pedicle connected to the MH edges, are prone to dislocation.

Based on these findings, we propose that materials used for MH repair should not only promote hole closure but also support retinal repair in the macular region. The amniotic membrane, owing to its favorable biological activity and reparative properties, has recently gained attention in vitreoretinal surgery for its therapeutic potential.^34–36 Therefore, in this study, we used amniotic membrane treatment for a series of extra-large or long-standing MH cases. We evaluated postoperative improvements in macular structure and function, while also assessing the safety profile of amniotic membrane in MH surgery.

In our study, all eight cases of extra-large or long-standing holes achieved complete closure. One case experienced amniotic membrane graft displacement. No amniotic membrane-related adverse events, such as hypopyon or empyema in the anterior chamber or vitreous cavity, retinal hemorrhage, hard exudates, or retinal vascular occlusion, were observed during the entire follow-up period. These findings align with previously reported outcomes,^37–39 suggesting the safety of amniotic membrane application in vitreoretinal surgery. Although some studies have identified graft dislocation as a primary complication of amniotic membrane use for MH repair,^37–39 our experience indicates that performing amniotic membrane transplantation after fluid–air exchange effectively mitigates dislocation caused by intraoperative fluid turbulence. In this study, a patient with extra-large or long-standing MH developed amniotic membrane graft displacement postoperatively. In this case, there was significant atrophy of the retinal pigment epithelial (RPE) in the macular area, and RPE abnormalities may have adverse effects on the stability of the amniotic membrane graft. Owing to the limited number of cases included in this study, the above hypotheses and analysis require further validation through more subsequent case observations. Furthermore, graft size is a critical factor for ensuring stable positioning. We used a new device specifically designed for amniotic membrane transplantation to assist in completing the surgical procedure (Figure 9). By tailoring the amniotic membrane to the MH diameter measured through preoperative OCT and embedding the graft into the MH area (Figure 8), stable fixation of the membrane can be reliably achieved.

Fig. 9.

Picture of 23-G amniotic membrane injector; the red arrow indicates the injection handle. A. Injector in the noninjection state. B. Injector after injection, with a silicone tip visible at the head end (white arrow).

During surgical procedures, we observed that the amniotic membrane graft, when wetted by irrigation fluid, becomes challenging to flatten smoothly due to the surface tension of the fluid. In addition, repeated grasping with surgical microforceps risks damaging the epithelial cells of the amniotic membrane graft. To address these challenges, we designed and developed a specialized injector for amniotic membrane transplantation (Figure 9), which effectively resolves the aforementioned issues while simplifying the procedural steps.

Regarding BCVA at the final follow-up of all observed cases, more than 50% of patients exhibited improved final visual acuity compared with baseline levels. The highest proportion of vision improvement (≥0.1 logMAR) was observed at 3 months postoperatively, reaching 87.5%, followed by a gradual stabilization phase. A similar trend was noted in mfERG results: the peak improvement in the photoreceptor-derived potential at the fovea reached its maximum at 3 months postoperatively and subsequently stabilized. An intriguing observation was that macular microperimetry results demonstrated enhanced central fixation stability and retinal sensitivity at the final follow-up. However, these parameters initially declined during the first postoperative month before gradually recovering and surpassing baseline levels. Notably, one patient with a 22-month follow-up exhibited a slow but continuous improvement in BCVA even beyond the 12-month postoperative period. Collectively, these findings indicate that visual function improvements in patients with extra-large or long-standing MH follow a prolonged and gradual course. This aligns with observations from a study by Zhongshan Ophthalmic Center,²⁶ which tracked 97 patients with extra-large MHs (>900 µm) treated by ILM coverage. In that 15-month study, BCVA improvements were still detectable at 13 months for holes measuring 900–1,000 µm, while patients with holes >1,000 µm showed progressive BCVA gains starting at 6 months postoperatively.

Another notable observation was the presence of well-defined photoreceptor cell attachments on the amniotic membrane graft surface in AO images, providing microscopic evidence supporting functional macular recovery. The biological amniotic membrane inherently serves as an excellent cellular scaffold material by offering extracellular matrix components for cell adhesion.¹¹ Studies have demonstrated that amniotic membrane as a culture substrate promotes the expression of differentiated characteristics in RPE cells.⁴⁰ These cultured cells exhibit characteristic ultrastructural features of epithelial cells, including distinct apical microvilli and intercellular junctions, with experimental observations confirming rapid RPE cell adhesion to the epithelial basement membrane side of amniotic membrane within 24 hours.⁴⁰ Furthermore, RPE cells cultured on a human amniotic membrane showed upregulated expression of RPE65, cellular retinaldehyde binding protein, and tyrosinase-related protein-2, along with significantly increased protein production of vascular endothelial growth factor, thrombospondin-1, and pigment epithelium-derived factor.⁴⁰ These biochemical changes establish a homeostatic microenvironment conducive to both local retinal tissue and RPE cell functionality, where RPE cell integrity critically influences regional retinal structure and function. Collectively, these findings suggest that amniotic membrane transplantation in MH regions may facilitate functional restoration of local photoreceptors or RPE cells, thereby improving long-term visual outcomes.

In summary, the use of amniotic membrane in surgery for extra-large or long-standing MHs is safe and effective. Postoperatively, patients achieved MH closure with improvements in visual acuity and function. If applied in clinical practice, this approach may benefit complex cases such as large-diameter MHs, recurrent MHs, traumatic MHs, or laser-induced MHs.

This clinical trial has several limitations. In terms of study design, double-blinding was impractical due to the significant morphological differences between amniotic membrane and ILM tissue. In addition, the sample size was relatively small.

Supplementary Material

Download video file ^{(13.4MB, mp4)}

Download video file ^{(10.4MB, mp4)}