Surgical Macular Hole Closure Without a Gas Tamponade

Abstract

Purpose:

This work aims to evaluate the outcomes of a series of macular hole (MH) surgical procedures in patients who had pars plana vitrectomy (PPV) with internal limiting membrane (ILM) peel and without gas tamponade.

Methods:

Patients from a retina specialty clinic who had MHs were identified for this interventional case series. Patients with small MHs were offered inclusion into the trial. Patients with larger MHs were excluded. They underwent standard 3-port PPV and ILM peel without gas or air to treat small MHs. The main outcomes that we measured were closure of MH and visual results.

Results:

Small MHs in 5 patients were managed with PPV and ILM peel alone. The average preoperative hole size at its narrowest width was 227 µm (range, 173-294 µm). Four of 5 patients (80%) had successful hole closure without a gas tamponade and improved vision; 1 patient did not have hole closure and was treated with an in-office gas bubble to close the hole. The average preoperative vision at 3 months was 20/80– (54 Early Treatment Diabetic Retinopathy Study [ETDRS] letters) and improved to 20/30–2 (73 ETDRS letters) in the 4 patients whose MHs were closed with surgery without a gas bubble. This was statistically significant (P = .003). The hole that did not close initially without gas tamponade was the largest in the series.

Conclusions:

Patients with small MHs can be successfully treated with a vitrectomy and ILM peel alone without a gas tamponade.

Introduction

In 1991, a landmark paper was published by Kelley and Wendel that showed that with removal of the vitreous and placement of a gas bubble and face-down positioning, visual improvement was seen in a significant number of patients with full-thickness macular holes (MHs). ¹ The intent of the authors was to reduce the localized detachment of the retina around the hole. However, it soon became clear that the MHs closed with these surgical maneuvers, and the initial success rate was 58% in 52 eyes, which improved to 73% 2 years later in a larger series of 170 eyes. ²

Internal limiting membrane (ILM) removal improved hole-closure rates even further. Closure rates increased to 92% to 100% in retrospective studies. ^3,4 Strict face-down positioning was recommended initially for 2 or more weeks. The face-down position requirement was later reduced, except for larger holes. Many surgeons first used perfluoropropane (C₃F₈), but then transitioned to sulfur hexafluoride (SF₆). Some physicians used air to close the MH. ⁵ The next step in the alteration of this surgical procedure was the recognition that strict face-down positioning was not critical and similar closure rates could be achieved without looking down. ^6,7

Nonsurgical closure of holes was also investigated using pharmacological vitreolysis. Ocriplasmin (Jetrea; ThromboGenics), which was used for enzymatic cleavage of the vitreous and allowed 30% of all MHs to close. In small MHs, the closure rate increased to 47.8%. ⁸

We set out to conduct a small pilot study of patients with small MHs and without posterior vitreous detachments to see whether these holes could close with surgical management and without any gas placement.

Methods

Patients who had MHs were identified in a single retina center from 2017 to 2019. Patients were included if they had a small MH (less than 300 µm, measured at the narrowest width of the hole on spectral-domain optical coherence tomography [OCT]). Patients with larger MHs (greater than 400 µm, measured at the narrowest width of the hole on spectral-domain OCT) were excluded. The patients were given the option of having surgery without a gas bubble, after it was explained that this was experimental. Five patients consented to the surgical procedure.

The patients were treated with a 23-gauge, standard 3-port pars plana vitrectomy. In all 5 cases, a posterior vitreous detachment was created using the vitrector or a soft-tip cannula. One patient (patient 2) had an epiretinal membrane, which was peeled first. All patients had a dilute solution of indocyanine green (0.1%) injected into the eye for less than 1 minute and then removed. A faint staining of the ILM was achieved. At the surgeon’s discretion, either a Tano scraper or Finesse Loop was chosen to begin the ILM rhexis. The ILM was removed 2 to 3 disc diameters around the central fovea. Then the peripheral retina was carefully examined 360° using scleral depression to ensure there were no breaks. The cannulas were removed, the sclerotomies were sutured closed if the wounds were not self-sealing, and no patients had either a gas or air tamponade.

Results

Five patients with small MHs were treated with pars plana vitrectomy and ILM peel alone. The average age of the patients was 68.2 years (range, 61-80 years). Four of the 5 patients were female. Three of the 5 patients were pseudophakic at presentation, and 2 patients were phakic. During the follow-up period after vitrectomy surgery, 1 of those 2 patients had cataract surgery.

The average preoperative hole size at its narrowest width was 227 µm (range, 173-294 µm), and the average greatest basal width was 497 µm (range, 259-767 µm) (Table 1). One patient had an epiretinal membrane. Four of the 5 patients had vitreomacular traction, subretinal fluid, and an elevated rim of the MH. Four of the 5 patients also had intraretinal cystic changes surrounding the MH (Figures 1 –5).

Table 1.

Patient Demographics, Clinical Data, and Outcomes.

Patient	Sex	Narrowest width of hole, μm	Base of hole, μm	Preoperative vision	Hole closed without gas	VA at 3 mo postoperatively	BCVA
1	M	201	584	20/100-	Yes	20/40–	20/40–
2	F	163	376	20/70	Yes (POD10)	20/30 – 2	20/25 + 2
3	F	239	259	20/60-2	Yes	20/30	20/30
4	F	173	377	20/70-2	Yes	20/30 – 2	20/25
5	F	294	767	20/100+	No	20/70–	20/40 + 2

Abbreviations: BCVA, best-corrected visual acuity; F, female; M, male; POD, postoperative day; VA, visual acuity.

Figure 1.

Patient 1. (A) Preoperative vision is 20/80. (B) Postoperative day 1, vision is 20/200 with closed macular hole. (C) Postoperative week 1, vision is 20/60. (D) Postoperative month 3, vision is 20/40.

Figure 2.

Patient 2. (A) Preoperative vision is 20/70. (B) Postoperative day 1, vision is 20/80 with a smaller gap, but the macular hole is still open. (C) Postoperative day 4, vision is 20/60 with a decreasing gap between the edges of the macular hole. (D) Postoperative day 7, the hole is almost closed, with a very small gap. Vision is 20/50. (E) On postoperative day 10, the hole is closed, and vision is 20/50. (F) At postoperative month 3, the hole is closed, and vision is 20/30.

Figure 3.

Patient 3. (A) Preoperative vision is 20/60. (B) Postoperative day 1, vision is 20/60 with closed hole. (C) Postoperative week 1, vision is 20/40. (D) Postoperative month 3, vision is 20/30.

Figure 4.

Patient 4. (A) Preoperative vision is 20/70. (B) Postoperative day 1, vision is 20/200 with a closed hole. (C) Postoperative week 1, vision is 20/40. (D) Postoperative month 3, vision is 20/30.

Figure 5.

Patient 5. (A) Preoperative vision is 20/100. (B) Postoperative day 1, vision is 20/400 with an open macular hole. (C) Postoperative day 3, vision is 20/200 with an open hole and slightly larger gap between the edges of the hole. (D) On postoperative day 4, the decision is made to inject 0.3 mL of 100% perfluoropropane (C₃F₈) with face-down positioning. On postoperative day 5, which is the first day after in-office gas placement, the hole is smaller, with vision at 20/200. (E) On postoperative day 8, the third day after in-office gas placement, the hole continues to get smaller, and vision is 20/200. (F) On postoperative day 14, the ninth day after in-office gas placement, the hole is closed, and vision is 20/100. (G) At postoperative month 3, vision is 20/50.

Four of 5 patients (80%) had successful MH closure without a gas tamponade and with improved vision. Of the 4 holes that closed without gas, 75% (3 of 4) closed on postoperative day (POD) 1 (see Figures 1, 3, and 4), whereas 1 hole (of patient 2) did not. At each visit, patient 2’s MH became smaller, until the hole closed on POD10 (see Figure 2). For the other patient (patient 5) who did not have hole closure on POD1, the hole became smaller, but on POD3 the hole enlarged. On POD4, the patient had an in-office gas bubble placed using 0.3 mL of 100% C₃F₈ with face-down positioning. Closure of the hole was achieved afterward (see Figure 5).

The average preoperative vision was 20/80– (54 Early Treatment Diabetic Retinopathy Study [ETDRS] letters) and improved at 3 months to 20/30–2 (72.8 ETDRS letters) in eyes that had closure with surgery and without a gas bubble. This was statistically significant (P = .003). The final best-corrected visual acuity (BCVA) in this group increased to 74.5 ETDRS letters (20/30) (P < .001). At 3 months all holes were closed, with increased vision averaging 69.6 ETDRS letters (20/40) (P = .006); all eyes that had initial closure without a gas bubble had 3 lines or greater of VA increase, or an average increase of 18 ETDRS letters (range, 15-24 ETDRS letters). At final BCVA, all patients had greater than 3 lines of VA increase (range, 15-24 ETDRS letters with average increase of 20 ETDRS letters). The hole that did not close initially without gas tamponade was the largest in the series (see Table 1).

Conclusions

MH surgery is highly successful with great visual outcomes that have continued to improve since it was originally described. ¹ MHs closed surgically have significant improvements in vision, with the average VA increasing to 20/32 to 20/40. ^9,10 In this series, the average BCVA increased to the 20/30 level by 3 months.

MH surgery has evolved over time. ^1,2 The addition of ILM peeling has increased the rate of successful hole closure. ^3,4 The procedure further improved for the patient by using shorter-acting gas bubbles and then air for some small holes, ⁵ with decreased positioning and for some no positioning, ^6,7 while still maintaining great MH closure rates and visual improvement. Because pharmacological vitreolysis achieved hole closure, ⁸ albeit at inferior levels compared with surgery, we theorized that MHs could be closed with vitrectomy and removal of the ILM alone, without a gas or air bubble.

What is the purpose of the bubble in MH surgery? It is possible that the bubble allows the environment to be dry and smooth, which may promote cellular proliferation. Also, the gas bubble allows the retinal pigment epithelium pump to remove the cuff of subretinal fluid surrounding the MH, possibly bringing the edges of the hole closer to apposition. There are also theories that the gas bubbles prevent fluid currents from washing away the glial cells that plug the hole. ¹¹ When examined histopathologically, MHs were sealed with Müller cell proliferation, and a defect of varying sizes (16-250 μm) remained in the photoreceptor layer. ¹²

On indirect fundoscopic examination, when looking through gas filled eyes, most holes appear closed on POD1. OCT has been performed through a gas bubble or silicone oil, and it has been noted in many cases that holes appeared closed within 24 hours of surgery. ⁵ There has been good evidence seen on OCT angiography that the foveal avascular zone is smaller after MH surgery with the removal of the ILM. ^13,14 The closure of holes on POD1 suggests that cellular proliferation may not be the only mechanism for hole closure since it would seem unlikely for proliferation to occur that quickly. If there is a reduction in size of the hole with just vitrectomy and ILM removal, could MH surgery be performed without a bubble with reasonable closure rates? In this series of MHs, 60% closed on POD1 (3 of 5 patients), and another MH closed later without gas tamponade. The closure rate with vitrectomy surgery and ILM peeling without gas in this series was 80%.

Given the rapid closure rate at 24 hours in our study and the recognized reduction in the size of the foveal avascular zone, there appeared to be a clear-cut mechanical shrinkage of the hole induced by the surgery. Previously we reported on a patient in whom closure had failed in the primary MH surgery, and in the second surgery, we created a retinotomy to close the hole. We imaged and measured the hole using intraoperative OCT. Immediately after creating the retinotomy, the intraoperative OCT showed the size of the hole had significantly reduced in size from 425 µm to 225 μm. ¹⁵ These findings suggested that there are at least 2 mechanisms for hole closure. First there is a mechanical shrinkage of the hole and second, in larger holes, there is a cellular proliferative response. Our present results suggest that the mechanical shrinkage caused by removing the ILM and reducing the tangential traction on the hole is adequate to close holes sized 250 µm or less, while more glial proliferation may be required to close larger holes.

We felt that we could introduce a bubble in the office setting postoperatively if there was inadequate closure, hopefully avoiding additional surgery. Fortunately, in the one patient whose hole did not close and started becoming larger, an in-office gas bubble was successful in closing the hole.

Several other studies have examined surgical closure of MHs without a gas tamponade. ^16,17 Stopa et al described a different technique to close MHs using viscoelastic to assist with creating a temporal ILM flap. They reported successful hole closure in 11 of 12 patients. Visual results were similar to our case series at 3 months, with improved vision to the 20/40 level. ¹⁶ Lally and Kasetty described a series of 20 small MHs. Seventeen of 20 MHs (85%) closed with 1 surgery. The MHs in that series, on average, closed much later than in our series but had a similar closure rate. The 3 holes that did not close required a subsequent vitrectomy surgery. The visual results were recorded at 1 month, with improved vision to the 20/50 level. ¹⁷

The advantages of not placing a bubble are modest but not inconsequential. The more rapid development of cataract would likely be reduced without gas, and patients could see better immediately. There are no position requirements and there are no travel restrictions due to a gas bubble. Also, examining the retina’s response to surgery without a gas bubble helped further elucidate the mechanism of closure in small MHs.

Patients with small MHs may be successfully managed with a vitrectomy and ILM peel alone without a gas tamponade. The patients in this study all had improved vision that was statistically significant. The 1 patient who did not have successful closure of the MH after surgery had an in-office gas injection with hole closure 10 days later, and her final vision was similar to the other patients. This is a small pilot study that does not prove the value of a no-gas technique. Larger studies need to be conducted on small MHs to see whether the closure rates without gas reach those with gas and determine the maximum size limit of the MHs (potentially < 250μm) that may be amenable to this approach.