Anatomical and Functional Outcomes of Relaxing Parafoveal Nasal Retinotomy for Refractory Macular Hole Repair

Abstract

Purpose:

This work aimed to assess postoperative outcomes associated with relaxing parafoveal nasal retinotomy for refractory macular hole repair.

Methods:

This was a retrospective interventional study of patients with persistent or recurrent macular holes following 1 or more standard repair procedures with pars plana vitrectomy and internal limiting membrane peeling. Patients received an additional pars plana vitrectomy and relaxing parafoveal nasal retinotomy, followed by fluid-air and air-gas exchange. Key postoperative outcomes included the achievement of macular hole closure and changes in visual acuity from baseline.

Results:

Thirteen patients with refractory macular holes were included, with a median age of 65 years (range, 49-90 years). The aperture diameter of the 13 macular holes ranged from 180 to 799 µm (median, 538 µm). Vitrectomy and relaxing parafoveal nasal retinotomy were performed in all 13 eyes, and after a median follow-up of 12 months (range, 3-34 months), anatomical closure was achieved in 12 of 13 eyes (92.3%). Overall, visual acuity (mean ± SE) improved significantly from 1.20 ± 0.15 logMAR (approximate Snellen equivalent, 20/320) at baseline to 0.84 ± 0.11 logMAR (Snellen, ∼ 20/125) during postoperative follow-up (P < .05). Central and paracentral scotomas were observed in 8 of 11 eyes with postoperative Humphrey visual field 10-2 and/or 24-2 data available.

Conclusions:

Relaxing parafoveal nasal retinotomy may be an effective method to promote anatomical closure and improve vision outcomes in patients with recalcitrant macular holes.

Introduction

Idiopathic full-thickness macular holes (FTMHs) are anatomical defects of the neurosensory retina and a common cause of metamorphopsia and central vision loss among adults aged 55 years and older. ¹ The first FTMH to be described in the literature was a traumatic case in 1869 ² ; however, our current understanding of idiopathic macular holes (MHs) is largely founded on clinical observations published by Gass more than 100 years later. ³ Although the pathogenesis of idiopathic MH remains unclear, imaging modalities including optical coherence tomography (OCT) have greatly enhanced our ability to visualize, diagnose, classify, and monitor MH in recent years. ^4,5

FTMH were long considered an incurable condition; however, treatment options and closure rates have evolved alongside our understanding of their pathology. In 1991, Kelly and Wendel first performed pars plana vitrectomy (PPV) with vitreous cortex detachment and fluid-gas exchange in 52 patients with idiopathic MHs, and they achieved closure in 58% of cases. ⁶ With ongoing refinement and the addition of internal limiting membrane (ILM) peeling, vital dyes, and alternative postoperative positioning, closure rates have exceeded 85% in more recent studies. ^{7 -9} Based on available evidence, vitrectomy is an effective method to improve anatomical and visual outcomes in patients with idiopathic MHs, and it represents the standard of care in current clinical practice. ^1,10

Despite current advances in primary repair procedures, postoperative outcomes for large MHs (aperture diameter > 400 µm) are suboptimal compared with smaller defects, ¹¹ and achieving long-term closure in persistent or recurrent MHs remains a surgical challenge. Several procedures involving inverted ILM flap techniques, autologous ILM transplant, lens capsular flap transplant, MH hydrodissection, autologous neurosensory retinal transplant, and human amniotic membrane plugs have been developed to promote the closure of large or refractory MHs ^{12 -17} ; however, a consensus surgical approach has not yet been established.

Newer techniques for large and recalcitrant MHs are typically more complex and carry inherent risks, which may limit their use to surgeons who have appropriate comfort and experience in these procedures. ¹² We describe a novel, simpler, and safer technique that may be readily adopted in routine clinical practice. The aim of this retrospective case series was to evaluate the anatomical and functional outcomes associated with relaxing parafoveal nasal retinotomy for refractory MH repair.

Methods

Study Design

This was a retrospective, multicenter, interventional study of relaxing parafoveal nasal retinotomy in eyes with recalcitrant MHs. Eligible patients presented to Carle Foundation Hospital (Champaign, Illinois) or University Retina and Macula Associates (Oak Forest, Illinois) between June 2016 and January 2019 and had persistent or recurrent MHs following 1 or more standard repair procedures with PPV and ILM peeling. Patients with significant macular pathologies (eg, advanced age-related macular degeneration, significant diabetic retinopathy, or retinal vein occlusions) or those who were unable to adhere to postoperative face-down positioning were ineligible to participate. Prior to surgery, risks and benefits associated with the procedure were discussed with patients deemed eligible by the investigators.

Patient Evaluation

Baseline characteristics collected from preoperative clinic notes included patients’ age and sex; concomitant ocular pathologies; the laterality, diameter, mechanism, and duration of the MH; and previous interventions for MH repair. The preoperative and postoperative size of the MH was assessed by spectral-domain OCT, with aperture diameter defined as the smallest distance at any level of the macula and basal diameter taken to be the widest distance at the level of the retinal pigment epithelium (RPE). Other ocular assessments included visual acuity (VA), lens status, postoperative peripapillary retinal nerve fiber layer (RNFL) thickness measured by OCT, and in some cases, postoperative Humphrey visual field (HVF) 10-2 and/or 24-2. Snellen VA was based on habitual correction or pinhole, whichever was best available at preoperative and postoperative visits.

Surgical Procedure

All patients underwent 25-gauge transconjunctival PPV performed by 1 of 3 surgeons (M.T., R.J., or V.S.). Residual ILM was stained with intravitreal indocyanine green and a wider, circumferential ILM peel was performed if present. Gentle endodiathermy with a 25-gauge tapered tip was used to create a small retinotomy at the midpoint between the fovea and optic disc, visualized by a small area of whitening on the retina. This was followed by gentle suction at the retinotomy site using a soft-tipped cannula, then fluid-air exchange and further suction to allow complete drainage of the MH (Figure 1). Intravitreal gas tamponade with isoexpansile or slightly hyperexpansile sulfur hexafluoride or perfluoropropane (C₃F₈) was applied at the end of the procedure, and patients were advised to maintain a face-down position for 7 to 10 days after surgery.

Figure 1.

Schematic diagrams showing (A) the approximate placement of parafoveal nasal retinotomy for MH repair, and (B) the proposed mechanism by which parafoveal nasal retinotomy and gentle suction promote MH closure.

Outcome Measures

Postoperative outcomes were assessed when the gas tamponade agent was completely reabsorbed. The primary end point was the achievement of MH closure determined by OCT; for those that closed, the restoration of the subfoveal ellipsoid zone was additionally assessed. Secondary outcomes included postoperative RNFL thickness, HVF 10-2 and/or 24-2 (if available), and VA change from baseline at the postoperative follow-up visit.

Statistical Analysis

Baseline characteristics and postoperative outcomes are presented using descriptive statistics, including mean, SE, median, range, and interquartile range (IQR). Snellen VA was converted to the logarithm of the minimal angle of resolution (logMAR) units for quantitative analysis, and a paired t test was used to detect a significant difference between preoperative and postoperative mean VA. Statistical significance was defined as P less than .05.

Results

Patient Population

This retrospective case series included 13 eyes from 13 patients with refractory MHs. Preoperative patient and MH characteristics are summarized in Table 1. The cohort comprised 5 men and 8 women, with a median age of 65 years (range, 49-90 years).

Table 1.

Preoperative Patient and Macular Hole Characteristics.

Pt	Age, y	Sex	Concomitant ocular pathologies	Macular hole characteristics
				Laterality	Aperture diameter, μma	Basal diameter, μmb	Mechanism/History	Chronicityc	Prior interventions
1	49	M	Glaucoma suspect (both eyes)	OD	180	1327	Traumatic; OGI with IOFB	Chronic	PPV + ILMP + MP ×2
2	68	F	AMD	OD	410	1234	Idiopathic; ocriplasmin use with dramatic hole enlargement	Chronic	Ocriplasmin; PPV + ILMP + MP
3	90	F	Moderate-stage glaucoma	OS	422	1043	Idiopathic	Nonchronic	PPV + ILMP + MP
4	76	M	None	OS	665	1206	Idiopathic	Nonchronic	PPV + ILMP + MP ×2
5	64	F	None	OD	253	810	Idiopathic; VMT related	Nonchronic	PPV + ILMP + MP
6	65	M	None	OS	276	327	Post-RD repair FTMH; unknown if iatrogenic or ERM related	Chronic	RD repair; MP ×2; PPV + ILMP + MP + SO
7	68	F	Mild NPDR	OS	704	1280	Idiopathic	Nonchronic	PPV + ILMP + MP + 20% SF6; PPV + ILMP + ILM flap overlay + 14% C3F8; PPV + SO
8	85	M	Intermediate dry AMD	OD	432	1156	Idiopathic; ERM-related FTMH	Nonchronic	PPV + ILMP + 14% C3F8
9	64	M	None	OS	700	1306	Idiopathic	Chronic	PPV + ILMP + 14% C3F8
10	71	F	Severe-stage POAG	OS	601	1156	Idiopathic	Nonchronic	PPV + ILMP + ILM flap overlay + 14% C3F8
11	65	F	None	OS	538	1126	Idiopathic	Nonchronic	PPV + ILMP + temporal retinotomy + 14% C3F8
12	57	F	DME	OD	799	1833	Idiopathic	Nonchronic	PPV + EL + ILMP + ILM flap overlay + 14% C3F8
13	60	F	None	OS	680	1226	Idiopathic	Chronic	PPV + ILMP + 25% SF6

Abbreviations: AMD, age-related macular degeneration; C₃F₈, perfluoropropane; DME, diabetic macular edema; EL, endolaser photocoagulation; ERM, epiretinal membrane; F, female; FTMH, full-thickness macular hole; ILM, internal limiting membrane; ILMP, internal limiting membrane peeling; IOFB, intraocular foreign body; M, male; MP, membrane peeling; NPDR, nonproliferative diabetic retinopathy; OD, right eye; OGI, open-globe injury; OS, left eye; POAG, primary open-angle glaucoma; PPV, pars plana vitrectomy; Pt, patient; RD, retinal detachment; SF₆, sulfur hexafluoride; SO, silicone oil; VMT, vitreomacular traction.

^a Aperture diameter of the MH was assessed by spectral-domain optical coherence tomography (OCT) and defined as the smallest distance measured at any level of the macula.

^b Basal diameter of the MH was assessed by spectral-domain OCT and defined as the widest distance measured at the level of the retinal pigment epithelium.

^c Patients with chronic MH were those with symptom duration of more than 6 months or diagnosis prior to initial vitrectomy.

The aperture diameter of the 13 MHs ranged from 180 to 799 µm (median, 538 µm; IQR, 343-690 µm). Eleven eyes had idiopathic MHs, 1 eye had a traumatic MH (patient 1), and 1 eye developed a MH following retinal detachment repair (patient 6). Five of 13 eyes had chronic MHs that had persisted for 6 months or more. All 13 patients had refractory MHs following 1 or more PPV procedures with ILM peeling, with or without ILM flap overlay. 1 eye had an idiopathic MH that had failed to close following a previous temporal retinotomy procedure (patient 11).

Postoperative Outcomes

Vitrectomy and relaxing parafoveal nasal retinotomy were performed on all 13 patients, as described in Table 2. Residual ILM was detected in 11 of 13 eyes, and a wider, circumferential ILM peel was additionally performed in these cases. All patients maintained a face-down position for at least 7 days after surgery, and postoperative follow-up ranged between 3 and 34 months (median, 12 months; IQR, 7-18.5 months).

Table 2.

Relaxing Parafoveal Nasal Retinotomy Procedure and Postoperative Outcomes.

Pt	Surgical procedure	Face-down, d	Follow-up, mo	Macular hole closurea	Lens status		VA, Snellen equivalent (logMAR)
					Preoperative	Postoperative	Preoperative	Postoperative
1	PPV + NR + 18% C3F8	7	34	Type 1; EZ band restoration	1 + NS	PCIOL	CF 2 ft (2)	20/80 (0.602)
2	PPV + ILMP + NR + 18% C3F8	7	24	Type 1; no EZ band restoration	PCIOL	PCIOL	CF 1 ft (2.301)	20/400 (1.301)
3	PPV + ILMP + NR+ 25% SF6	7	12	Type 1; EZ band restoration	PCIOL	PCIOL	20/100 (0.699)	20/80 (0.602)
4	PPV + ILMP + NR + 25% SF6	7	13	Type 1; partial EZ band restoration	PCIOL	PCIOL	CF 3 ft (1.824)	20/600 (1.477)
5	PPV + ILMP + NR + 25% SF6	7	12	Type 1; partial EZ band restoration	2 + PSC; 1 + NS	PCIOL	20/200 (1)	20/40 (0.301)
6	PPV + SOR + NR + 14% C3F8	14	11	Type 2	PCIOL	PCIOL; PCO	20/200 (1)	20/120 (0.778)
7	PPV + ILMP + EL + NR + 16% C3F8	14	13	Type 1; no EZ band restoration	PCIOL	PCIOL	20/200 (1)	20/100 –1 (0.699)
8	PPV + ILMP + ERMP + NR + 16% C3F8	14	12	Type 1; EZ band restoration	PCIOL	PCIOL	20/120 –1 (0.778)	20/120 –2 (0.778)
9	PPV + ILMP + NR + 14% C3F8	10	7	Type 1; EZ band restoration	PCIOL	PCIOL; 2 + PCO	20/160 –1 (0.903)	20/80 (0.602)
10	PPV + ILMP + NR + 14% C3F8	10	4	Type 1; no EZ band restoration	PCIOL	PCIOL	20/120 (0.778)	20/160 –2 (0.903)
11	PPV + ILMP + NR + 14% C3F8	7	3	Type 1; no EZ band restoration	2 + NS; 2 + CC	2 + NS; 2 + CC	20/400 +1 (1.301)	CF 4 ft (1.699)
12	PPV + ILMP + EL + NR + 14% C3F8	7	7	Type 1; partial EZ band restoration	PCIOL	PCIOL	20/100 –2 (0.699)	20/100 +1b (0.699)
13	PPV + ILMP + NR + 25% SF6	7	24	Type 1; partial EZ band restoration	2 + NS	PCIOL	20/400 (1.301)	20/60 (0.477)

Abbreviations: C₃F₈, perfluoropropane; CC, cortical cataract; CF, counting fingers; EL, endolaser photocoagulation; ERMP, epiretinal membrane peeling; EZ, ellipsoid zone; ft, feet; ILMP, internal limiting membrane peeling; NR, nasal retinotomy; NS, nuclear sclerotic cataract; PCIOL, posterior chamber intraocular lens; PCO, posterior capsule opacification; PPV, pars plana vitrectomy; PSC, posterior subcapsular cataract; Pt, patient; SF₆, sulfur hexafluoride; SOR, silicone oil removal; VA, visual acuity.

^a Type 1 closure indicates complete restoration of the foveal defect and attachment to the underlying retinal pigment epithelium (RPE); type 2 closure indicates that while the edges of foveal defect had thinned and attached to the underlying RPE, the foveal defect persisted postoperatively.

^b Postoperative VA measured at 2 months’ follow-up.

MH closure was achieved in 12 of 13 eyes (closure rate, 92.3%; Table 2, Figure 2). Of those that achieved closure, 8 eyes showed complete or partial restoration of the subfoveal ellipsoid zone. Some eyes with MH closure exhibited retinal thinning on postoperative OCT (eg, patients 7, 11, and 13), likely owing to multiple surgical procedures, ILM peeling, and tissue manipulation. Nasal retinotomy was unsuccessful in 1 chronic MH that developed following retinal detachment repair and had failed 3 previous attempts to achieve closure (patient 6). The retinotomy site remained open in 5 of 8 eyes with available OCT data; in these cases, postoperative enlargement of the open retinotomy or accumulation of subretinal fluid at the site was not observed.

Figure 2.

Postoperative OCT, visual field, and RNFL outcomes in 13 eyes that underwent relaxing parafoveal nasal retinotomy for refractory MH repair. HVF indicates Humphrey visual field; OCT, optical coherence tomography; RNFL, retinal nerve fiber layer.

Nine of 13 eyes (69.2%) achieved VA improvement during postoperative follow-up, including the eye that failed to achieve MH closure (see Table 2). Three of 4 eyes that were phakic at baseline underwent phacoemulsification and intraocular lens placement following MH repair, and all 3 achieved VA improvement postoperatively. Overall, mean VA (±SE) improved significantly from 1.20 ± 0.15 logMAR (approximate Snellen equivalent, 20/320) at baseline to 0.84 ± 0.11 logMAR (Snellen, ∼20/125) during postoperative follow-up (P < .05; Figure 3).

Figure 3.

Preoperative and postoperative VA (A, mean ± SE; B, patient-level results) in 13 eyes that underwent relaxing parafoveal nasal retinotomy for refractory MH repair. VA indicates visual acuity.

Postoperative OCT indicated normal RNFL thickness in 11 of 13 eyes (see Figure 2). Significant RNFL thinning was observed in patient 10 due to severe glaucoma in both eyes, and symmetrical nasal RNFL thinning in patient 9 was likely an anatomical variant. Postoperative HVF 24-2 and/or 10-2 showed no detectable defect in 3 of 11 eyes with available HVF data (patients 2, 10, and 13), while an enlarged central scotoma and small paracentral scotomas were observed in the remaining 8 eyes. Patient 4 suffered a cerebrovascular accident during the postoperative period and developed a homonymous hemianopia that subsequently confounded his VA and visual field outcomes. Overall, paracentral scotomas were well tolerated, and patients reported better functionality with postoperative paracentral scotoma than with central scotoma due to the MH.

Expanded Case Study

Patient 1 was a 49-year-old man with a traumatic MH in his right eye following an open-globe injury with an intraocular foreign body. The MH was chronic (duration > 1 year), 180 µm in aperture diameter (1327 µm in basal diameter), and had failed to close after 2 previous PPV and ILM peeling procedures at a different institution (1 with cataract extraction; both with good patient compliance). We performed a third vitrectomy with parafoveal nasal retinotomy, followed by gas tamponade with 18% C₃F₈ and 7 days of face-down positioning. No residual ILM was found during the procedure, and intraoperative OCT showed that retinotomy immediately allowed the edges of the hole to become mobile (Supplemental Video). During postoperative follow-up, complete MH closure was achieved and Snellen VA improved from counting fingers at 2 feet (2 logMAR) to 20/80 (0.602 logMAR). Minimal defects were detected on postoperative HVF and none on RNFL OCT (Figure 4).

Figure 4.

(A) Preoperative optical coherence tomography image of a chronic, traumatic MH treated with relaxing parafoveal nasal retinotomy after 2 unsuccessful vitrectomy procedures. At 1 year postoperatively, (B) complete MH closure (with an open retinotomy) was achieved, (C) minor central and paracentral visual field defects were observed, and (D) no retinal nerve fiber layer loss was detected.

Conclusions

Although PPV with ILM peeling, gas tamponade, and postoperative face-down positioning is considered the standard surgical approach for primary MH repair, there is no consensus strategy for the treatment of large or refractory MHs. Several methods have been developed, including inverted ILM flap techniques that have achieved closure rates of 88% to 98% in eyes with large MHs. ^{12 -14} However, in cases in which the ILM has previously been peeled, treatment options may be limited to reoperation and tamponade with long-acting gas or silicone oil, which requires removal and has been associated with poorer postoperative outcomes. ¹⁸

Several studies have assessed the feasibility of additional vitrectomy procedures for persistent or recurrent MHs. For example, D’Souza et al performed reoperation on 30 patients with MHs that had failed to close or reopened following primary vitrectomy and ILM peeling. ¹⁹ Reoperation consisted of PPV, enlargement of the ILM rhexis, gas tamponade with nonexpansile C₃F₈, and 1 week of face-down positioning. Similarly, Hillenkamp et al reported postoperative outcomes for 28 patients who received an additional vitrectomy for the treatment of persistent MHs. ²⁰ In most cases, autologous platelet concentrate was applied to the MH prior to tamponade with 16% sulfur hexafluoride or silicone oil. In these studies, anatomical closure was reported in 47% and 68% of cases, respectively. ^19,20

Our case series follows several others that have sought to improve postoperative outcomes with retinotomy and related techniques. In 2012, Reis et al performed radial retinal incision on 7 patients with FTMH that persisted after vitrectomy and ILM peeling. ²¹ Five perifoveal incisions were made 1 macular-hole diameter away from its margins and extended centripetally toward the defect. Following gas tamponade and face-down positioning, anatomical closure was achieved in all patients, and mean VA improved by 5.6 lines. However, it should be noted that this is a challenging technique with high inherent risks of hemorrhage and/or RPE damage, and the authors did not describe the methods used to create and judge the depth of their retinal incisions.

One year later, Charles et al described the use of temporal arcuate retinotomy in 6 eyes with large MHs that had failed to close following primary repair. ²² Successful closure and improved vision were achieved in 83% and 50% of eyes, respectively; however, retinal thinning and RPE defects at the retinotomy site were observed in some patients. In this study, vertical scissors were used to create the retinotomy, which likely damaged the underlying RPE through an inability to judge the depth of penetration under the retina.

In a simpler and safer procedure, Tsipursky first used endocautery and a soft-tipped cannula to create a paracentral retinotomy in 6 eyes with large MHs that had failed at least 1 prior repair procedure. ²³ With this technique, anatomical closure was achieved in 83% of cases, and mean VA improved from 1.28 logMAR (Snellen, ∼20/400) at baseline to 0.89 logMAR (Snellen, ∼20/160) during postoperative follow-up. In a patient who developed a large MH following PPV for proliferative diabetic retinopathy, Karacorlu et al performed superior and inferior arcuate relaxing retinotomies after previous repair attempts had failed. ²⁴ This procedure resulted in successful MH closure, improved VA, and no intraoperative or postoperative complications.

Anatomical and functional outcomes associated with relaxing nasal retinotomy for MH repair were recently described in a case report by Knight et al. ²⁵ In this study, a 69-year-old woman with juxtafoveal telangiectasia presented with Snellen VA 20/200 and a stage 1 MH that measured 647 µm wide at its base. The defect reopened 1 month after primary vitrectomy and ILM peeling, measuring 425 µm at its narrowest width. During a second vitrectomy and superonasal retinotomy, intraoperative OCT showed that the defect narrowed to 219 µm immediately after retinotomy. After 6 months, Snellen VA was 20/100 and both the MH and retinotomy site had closed completely.

The present study expands on the work of Tsipursky ²³ and Knight et al ²⁵ to demonstrate the utility of relaxing nasal retinotomy in a larger cohort of patients with recalcitrant MHs and over a longer period of follow-up. In our case series, anatomical closure was achieved in 12 of 13 eyes (92%); of those that closed, 10 eyes had large MHs (aperture diameter > 400 µm), and 8 eyes achieved complete or partial restoration of the subfoveal ellipsoid zone. Our technique failed to resolve 1 chronic MH that had failed 3 previous repair procedures; however, postoperative VA was found to be improved in this patient. Interestingly, 4 of 5 eyes with chronic MHs achieved anatomical closure with nasal retinotomy, and mean VA (±SE) in these eyes improved significantly from 1.63 ± 0.32 logMAR (Snellen, ∼20/800) to 0.75 ± 0.19 logMAR (Snellen, ∼20/100). These results suggest that the chronicity of a MH should not prevent a surgeon from attempting closure with nasal retinotomy. Overall, 9 of 13 eyes achieved vision gains during postoperative follow-up, and mean VA improved significantly from 1.20 ± 0.15 logMAR (Snellen, ∼20/320) to 0.84 ± 0.11 logMAR (Snellen, ∼20/125).

The proposed mechanism by which relaxing retinotomy promotes MH closure is by reducing tangential traction on the defect, ^12,22,24,25 although increased glial cell proliferation may also play a role. ²¹ In our study, retinotomy was followed by suction at the retinotomy site using a soft-tipped cannula, which served to drain the MH. Intraoperatively, we observed that suction immediately allowed the edges of the MH to come together and narrowed the width of the defect, suggesting that this technique may additionally promote MH closure by further relaxing the adjacent retinal tissue (see Figure 1). Alternatively, fluid aspiration and tissue manipulation may assist by lysing adhesions between the edge of the hole and underlying RPE that otherwise prevent MH closure; however, further studies are needed to test the effectiveness of these techniques in the absence of nasal retinotomy.

Advantages of relaxing retinotomy for MH repair include its application of commonly used surgical techniques, a rapid anatomical response confirmable on intraoperative OCT, the preservation of central VA, and its amenability in eyes that have received prior vitrectomy. The procedure carries a risk of inducing paracentral scotoma; however, we found that the functional VA improvements achieved by most patients during postoperative follow-up generally outweighed the development of small temporal and nasal scotomas. Patients with significant macular pathologies (eg, advanced age-related macular degeneration, significant diabetic retinopathy, or retinal vein occlusions) were excluded from this study, and we suggest that relaxing retinotomy should be contraindicated for these patients in clinical practice. Owing to the potential for postoperative RNFL thinning in patients with severe glaucoma, we additionally recommend that these patients be excluded from receiving relaxing retinotomy for persistent or recurrent MHs.

Limitations of our case series included its small sample size of 13 patients; however, our positive findings warrant confirmation in larger comparative studies. Although a circumferential peel of residual ILM was additionally performed in 11 of 13 eyes, we believe that relaxing retinotomy facilitated MH closure, given that all eyes in this study had previously failed 1 or more standard repair procedures with ILM peeling. In eyes for which data from previous procedures were available, most MHs decreased in size after initial repair, suggesting that these holes were more likely refractory than having failed to close because of other factors, such as noncompliance with postoperative face-down positioning.

We acknowledge that relaxing retinotomy is a “destructive” surgical technique that sacrifices tissue to promote MH closure, and that the retinotomy site remained open for most of our cases. Although we observed that open retinotomy sites and RNFL thickness remained stable during postoperative follow-up (median duration of 12 months in this study), further studies are needed to assess the longer-term safety of this procedure. Furthermore, ganglion cell layer thickness was not assessed, microperimetry was not performed, and most patients did not have preoperative HVF and RNFL thickness data available for comparison. Our study did not include a control or comparator group, so we did not compare the effectiveness of relaxing retinotomy against alternative techniques for refractory MHs. Although we advocate the use of an ILM flap or scaffolding technique for the primary repair of large MHs, the results of our study suggest that relaxing retinotomy may facilitate long-term anatomical closure and functional improvement in patients with persistent or recurrent MHs.

Future studies may seek to refine our technique and further improve postoperative outcomes. For example, injecting a subretinal balanced salt solution prior to retinotomy may create a localized serous retinal detachment that promotes MH closure, ²⁶ while the use of an excimer laser for retinotomy may increase surgical precision and minimize the risk of RPE damage. We performed nasal retinotomy based on early observations that the nasal macula was often the most foreshortened area among eyes with recalcitrant MHs and the belief that this procedure may carry less risk of retinal detachment than temporal retinotomy. Further studies may alternatively perform temporal retinotomy to avoid the maculopapillary bundle; however, 1 patient in our study had a refractory MH after previously receiving this procedure (patient 11).

In conclusion, this retrospective case series found that most patients who received relaxing parafoveal nasal retinotomy for refractory MH repair achieved anatomical closure and vision gains during postoperative follow-up. Larger comparative studies are needed to confirm our findings; nevertheless, our data suggest that this technique may be an effective and readily adopted treatment strategy for patients with persistent or recurrent MHs.