From Manchester Royal Eye Hospital (MMM, IA, GST, PES); and University of Manchester (MMM, PES), Manchester, United Kingdom.
Supported by the Manchester Academic Health Sciences Centre and the NIHR Manchester Biomedical Research Centre.
The authors have no financial or proprietary interest in the materials presented herein.
Address correspondence to Paulo E. Stanga, MD, Manchester Royal Eye Hospital, Oxford Road, Manchester, M139WH, United Kingdom. E-mail: firstname.lastname@example.org
The healing responses after full-thickness macular hole surgery are well recognized.1 Recently, outcomes of surgery without postoperative face-down positioning have been reported with good results.2–4 However, the strong light reflexes emanating from gas–macular interfaces may prevent detailed in vivo visualization of the extent of gas–macula tamponade.5
Face-down positioning has been prescribed to patients after full-thickness macular hole surgery to ensure gas–inner macula tamponade and maximize anatomical hole closure and healing. Fourier-domain optical coherence tomography (FD-OCT) is routinely performed with the head in the upright position at the chin-rest and the patient seated. We hypothesized that this may simulate the upright head positions that occur during normal daily activities for patients at home or outdoors. In this scenario, the ability to noninvasively demonstrate optimal gas–inner macular tamponade in the upright position may avoid requiring face-down positioning for patients postoperatively.
Imaging of gas–inner retinal tamponade using FD-OCT has yet to be reported. We attempted to address the extent of gas tamponade at the inner retinal interface after macular hole surgery in the presence of all commonly used gas tamponade agents.
Design and Methods
A consecutive series of patients with full-thickness macular hole were prospectively studied between July 2008 and March 2009. All patients underwent standard 20-gauge pars plana vitrectomy and Membrane-Blue–assisted (Dorc International, Rotterdam, The Netherlands) internal limiting membrane peel followed by injection of perfluoroethane (C2F6), perfluoropropane (C3F8), or sulfur hexafluoride (SF6) gas.
Patients were instructed to maintain face-down positioning overnight, with uncomplicated surgery in all cases. On the first postoperative day, we performed FD-OCT (3D OCT-1000; Topcon, Inc., Paramus, NJ) using 6 × 6 mm three-dimensional scans, high-definition overlapping line scans, and raster scans to assess the extent of gas tamponade at the inner retinal interface. The sequential FD-OCT scans were image-tracked using vessel and disc landmarks. The anatomical fovea was localized using three-dimensional cropping and vascular landmarks, and the foveal zone was further analyzed using the x-, y-, and z-planes on the Topcon software analysis program. The decision for the postoperative face-down positioning regimen was based on the biomicroscopy findings and substantiated with FD-OCT imaging. Patients were subsequently examined at 2 and 12 weeks following surgery to evaluate anatomical hole closure.
We studied 10 patients who underwent uncomplicated surgery. On day 1 postoperatively, the lower meniscus of the gas bubble was inferior to the lower vascular arcade with greater than 75% gas fill of the vitreous cavity in all eyes (Fig. 1). FD-OCT demonstrated satisfactory tamponade between the central macula and the bubble meniscus when the patient was in the upright position (Fig. 2). Interoferometric signal anomalies prevented a clear image of the fovea, but the inner and outer retina landmarks could be accurately identified along the curvature of the eye (Fig. 3).
Figure 1. Slit-Lamp Photography of a Patient on the First Postoperative Day Demonstrating the Optimal Location of Gas Bubble Within Vitreous Cavity. The Gas Bubble Is Seen Within the Vitreous Cavity and the Lower Meniscus Visible Above the Inferior Pupil Margin. Examination Revealed the Lower Meniscus of the Gas Bubble Was Inferior to the Lower Vascular Arcade with a 75% Gas Fill of the Vitreous Cavity.
Figure 2. Three-Dimensional Fourier-Domain Optical Coherence Tomography (FD-OCT) on the First Postoperative Day. (A) FD-OCT En-Face Scan Demonstrates Sectors of Hyperreflectivity that Correlate with Areas of Gas–Inner Retina Interface. Curvature of the Posterior Pole Prevents Complete Tamponade Hyperreflectivity from Being Visualized in a Single FD-OCT Scan. The Black Marker and Arrow Are Positioned at the Anatomical Fovea. (B) Color Fundus Photograph Shows the Flash Reflection from the Posterior Surface of the Gas Bubble at the Macula. The Black Arrow Localizes the Anatomical Fovea. The Edges of the Flash Area Correlate with the Areas of Hyperreflectivity. (C) Horizontal (FD-OCT) Scan Demonstrates Intense Hyperreflectivity that Demarcates the Apex of Tamponade Between the Fovea and Posterior Surface of the Gas Bubble. The Vertical White Marker Passes Through the Foveal Center. The Inner Surface of the Retina Appears to Be Continuous Across the Fovea.
Figure 3. Three-Dimensional Fourier-Domain Optical Coherence Tomography (FD-OCT) on the First Postoperative Day. (A) FD-OCT Scan Demonstrates the Posterior Curvature of the Macula. The Vertical White Marker Designates the Anatomical Center of the Fovea. The Inner Surface of the Foveal Center Is Well Demarcated by the Hyperreflectivity (thick Arrow), Induced by the Interface Between the Edge of the Gas Bubble and the Surface of the Inner Retina After Reduction of the Gain or Noise Levels. The Outer Nuclear and Outer Plexiform Layers Are Formed (dotted Arrow). The Junction Between the Inner and Outer Segments of the Photoreceptors (JI/OSP) and Retinal Pigment Epithelium Are Visualized as Moderately Hyperreflective Layers (dashed Arrow). (B) Three-Dimensional FD-OCT Scan with Subtraction of the Inner Retina and Retinal Pigment Epithelium Layers to Show the JI/OSP. The Vertical White Marker and Black Arrow Designate the Anatomical Center of the Fovea. The JI/OSP Surface Lacks Reflectivity Because the Intense Hyperreflectivity that Occurs at the Gas–Inner Retina Interface Blocks Penetration of the Interferometric Signals from the JI/OSP Posterior to the Gas Bubble Surface, Thus Creating an Artefact that Corresponds with the Imprint of the Gas Bubble on the Inner Surface of the Retina as Visualized from Above. The Area of Gas Tamponade at the Macula Is a 3-Disc Diameter Circular Zone Centered at the Fovea.
FD-OCT en-face imaging demonstrated sectors of hyperreflectivity that correlated with areas of gas tamponade at the inner retinal interface. Horizontal FD-OCT scanning demonstrated a localized point of intense hyperreflectivity that accurately demarcated the apex of tamponade between the fovea and posterior surface of the gas bubble. The inner retinal surface was well demarcated by the hyperreflectivity induced by the interface between the posterior edge of the gas bubble and the surface of the inner retina. Evaluation of individuals’ FD-OCT 6-mm scans demonstrated that coverage of gas bubble tamponade extended down to the lower vascular arcade.
FD-OCT was able to show gas tamponade at the inner retinal interface in patients who had inadequate biomicroscopic visualization of the fovea postoperatively due to overlapping of the lower bubble meniscus across the fovea. We found that a gas fill of 50% did not achieve satisfactory gas tamponade at the inner retinal interface at 2 weeks postoperatively (Fig. 4).
Figure 4. Three-Dimensional Fourier-Domain Optical Coherence Tomography (FD-OCT) of a Patient 2 Weeks Following Surgery with a 50% Vitreous Cavity Fill of SF6 Gas. (Upper Left) Arrow Indicates the Plane of Scanning. The Fovea Is not Visualized on Either Biomicroscopy or Fundus Photography Due to Masking by the Lower Edge of the Gas Bubble. (Upper Right) Three-Dimensional Scan of 50% Gas Effects on Interferometry Image. Lack of Intraretinal FD-OCT Signal Due to Intravitreal Gas Bubble Above the Fovea (right Side of Scan), with a Much Higher Retinal Signal Below the Bubble Anterior to the Macula (left Side of Scan). (Lower) Horizontal FD-OCT Demonstrates Anatomical Hole Closure and There Are No Reflectivity Changes Associated with Gas–Inner Retinal Tamponade.
At day 1 postoperatively, there was 80% (8 eyes) complete hole closure (Fig. 5), with partial closure in two eyes (Fig. 6). After 2 weeks, surgery resulted in anatomical closure of full-thickness macular hole in 90% of eyes. A single myopic patient with full-thickness macular hole developed a rhegmatogenous retinal detachment within 1 week of surgery, with associated unsuccessful full-thickness macular hole closure.
Figure 5. (A) Three-Dimensional Fourier-Domain Optical Coherence Tomography (FD-OCT) of a Patient with a Stage 4 Macular Hole with Cystoid Spaces Within the Inner Plexiform Layer at the Edges of the Hole. On Day 1 Following Surgery Using C3F8, a Gas–Inner Retinal Interface of Tamponade Is Seen (thick Arrow) on the (B) x-Plane and (C) y-Plane. The Outer Retinal Layers Have Conjoined with Complete Closure of the Macular Hole and Restoration of the Foveal Contour. (D) After 2 Weeks, the Foveal Contour Is Present with Atrophy of Intraretinal Tissue Layers. The Junction Between the Inner and Outer Segments of the Photoreceptors and Retinal Pigment Epithelium Layers Is Thickened, and Continuity of These Outer Retinal Layers at the Base of the Fovea Confirms Hole Closure (dotted Arrow).
Figure 6. Three-Dimensional Fourier-Domain Optical Coherence Tomography (FD-OCT) of a Patient with a Stage 2B Macular Hole and Foveal Detachment in the (A) x-Plane and (B) y-Plane. On Day 1 Following Surgery Using C2F6, a Gas–Inner Retinal Interface of Tamponade Is Seen (thick Arrow, C and D) with Partial Hole Closure. The Inner and Middle Foveal Layers Are Conjoined and Have Normal Reflectivity Below the Surface (dashed Arrow). The Subretinal Fluid Has Reduced and the Junction Between the Inner and Outer Segments of the Photoreceptors and Retinal Pigment Epithelium Layers Is Intact at the Base of the Fovea (dotted Arrow). At 2 Weeks, the Foveal Contour (thick Arrow) Is Present with a Small Pocket of Intraretinal Fluid, and There Is Closure of the Macular Hole in the (E) x-Plane and (F) y-Plane.
The landmarks for gas-filled eyes based on indirect ophthalmoscopy have been well documented by Lincoff et al.6 However, visualization of gas–inner retinal interfaces, and more importantly the gas–inner foveal interface, are not well described. Postoperative gas-filled eyes have poor foveal fixation patterns, but mapping software on the Topcon FD-OCT system allowed localization of the anatomical fovea in our patients.
Biomicroscopy is used to determine the correlation between the percentage of gas fill and foveal tamponade in the immediate postoperative period. The SF6 gas bubble contracts over the initial 2 weeks, and visualization of the foveal zone may be difficult in the presence of an overlapping inferior bubble meniscus. FD-OCT enabled accurate foveal imaging below the bubble meniscus. At 2 weeks postoperatively, there was no longer any gas tamponade at the inner retinal interface in eyes with a gas fill of 50% when patients were in the upright position.
We have demonstrated FD-OCT imaging of the gas tamponade at the inner retinal interface that may extend across a 3-disc diameter circular zone centered at the fovea. The posterior surface of the gas bubble maintained gas–inner retinal tamponade across the whole area of coverage (Fig. 3). Three-dimensional FD-OCT with en-face subtraction may produce a detailed imprint of the area of gas tamponade coverage, designated by a circular zone of hyperreflectivity, as early as the first postoperative day. Due to the curvature of the posterior pole, evaluation of multiple en-face subtraction scans may be required to localize the plane of gas tamponade (Fig. 2).
The inner retinal surface is well demarcated by hyperreflectivity induced by the interface between edges of the gas bubble and surface of the inner retina after reduction of the gain or noise levels. High-definition overlapping and choroidal FD-OCT scan acquisition further improved visualization of gas–retinal interfaces. On the first postoperative day, we did not observe any differences between C2F6, C3F8, and SF6 at different percentage concentrations with regard to reflectivity signals or tamponade effects at inner retinal interfaces. In the immediate postoperative period, the tamponade effects of possible suboptimal gas fill may be assessed using FD-OCT.
The main aim of this small study was to investigate whether a 75% gas fill can achieve gas–inner retinal tamponade in the upright position. We defined hole closure as no interruption of the internal line of the inner retina in any macular area on FD-OCT scanning. We have analyzed sequential overlapping high-definition, choroidal, raster, and three-dimensional horizontal FD-OCT scans in detail for this study, and it is unlikely that an unclosed macular hole was overlooked by our method. We found that FD-OCT may noninvasively validate this tamponade effect across the fovea to facilitate macular hole closure in the upright position, and this may allow patients to avoid postoperative face-down positioning after surgery.
A limitation of our study is the lack of sequential scans at closer intervals after surgery that may have provided useful information on healing responses. These healing responses after surgery are recognized to occur within the first week after surgery. There are only a few reports of OCT examinations performed in the early postoperative period following macular hole surgery, and these studies have reported healing responses and hole closure within the first week after surgery. Hasler and Prünte have reported 93% hole closure at day 3 postoperatively using air tamponade.7 Kasuga et al. acquired good OCT images through the gas bubble in 4 of 7 eyes with SF6 tamponade on the first postoperative day and all 4 eyes showed a closed hole.5 Eckardt et al. used a customized OCT technique and reported 54.5% hole closure at 24 hours, 75.7% hole closure at 48 hours, and 12.1% open holes at 24 hours following surgery using air tamponade.8 In our study, on day 1 after surgery, 80% had complete macular hole closure and two eyes had partial closure (Figs. 5 and 6). Because there were no open holes on day 1 postoperatively, we did not perform sequential OCT scans until the 2-week visit. Of the two partially closed eyes, one myopic patient with macular hole developed a retinal detachment at 1 week and one patient developed complete hole closure at 2 weeks.
It would be of interest to vitreoretinal surgeons to note exactly when different gases stopped tamponading the foveal center in the upright position. In our study, we observed an absence of gas–inner retinal tamponade with a 50% gas fill. Lincoff et al. extensively studied the properties of gas, including half-life and disappearance times, within the vitreous cavity in rabbit and human eyes.9 According to them, the approximate end points for a 50% gas fill were SF6 at 6 days, C2F6 at 10 days, and C3F8 at 35 days.9
This is the first report of high-resolution FD-OCT imaging to evaluate gas tamponade effects following full-thickness macular hole surgery. During the initial 2 weeks postoperatively, we demonstrated that all types of intraocular gas may achieve optimal gas tamponade at the inner retinal interface in the upright position. These FD-OCT observations may help clinicians decide whether to use face-down positioning following surgery.
- Madreperla SA, Geiger GL, Funata M, de la Cruz Z, Green WR. Clinicopathologic correlation of a macula hole treated with cortical vitreous peeling and gas tamponade. Ophthalmology. 1994;101:682–686.
- Simcock PR, Scalia S. Phacovitrectomy without prone posture for full thickness macular holes. Br J Ophthalmol. 2001;85:1316–1319. doi:10.1136/bjo.85.11.1316 [CrossRef]
- Tranos PG, Peter NM, Nath R, et al. Macular hole surgery without prone positioning. Eye. 2007;21:802–806. doi:10.1038/sj.eye.6702339 [CrossRef]
- Dhawahir-Scala FE, Maino A, Saha K, Mokashi AA, McLauchlan R, Charles S. To posture or not to posture after macular hole surgery. Retina. 2008;28:60–65. doi:10.1097/IAE.0b013e31813c68a2 [CrossRef]
- Kasuga Y, Arai J, Akimoto M, Yoshimura N. Optical coherence tomography to confirm early closure of macular holes. Am J Ophthalmol. 2000;130:675–676. doi:10.1016/S0002-9394(00)00587-0 [CrossRef]
- Lincoff HA, Kreissig I, Smith RJ, Serag YR, Laroche D. Clinical measurement of intraocular gas [article in German]. Klin Monatsbl Augenheilkd. 1996;208:100–106. doi:10.1055/s-2008-1035178 [CrossRef]
- Hasler PW, Prünte C. Early foveal recovery after macular hole surgery. Br J Ophthalmol. 2008;92:645–649. doi:10.1136/bjo.2007.130971 [CrossRef]
- Eckardt C, Eckert T, Eckardt U, Porkert U, Gesser C. Macular hole surgery with air tamponade and optical coherence tomography-based duration of face-down positioning. Retina. 2008;28:1087–1096. doi:10.1097/IAE.0b013e318185fb5f [CrossRef]
- Lincoff H, Coleman J, Kreissig I, Richard G, Chang S, Wilcox LM. The perfluorocarbon gases in the treatment of retinal detachment. Ophthalmology. 1983;90:546–551.