Sir,

Macular holes in proliferative diabetic retinopathy (PDR) may present an unusual surgical challenge when associated with a macular tractional retinal detachment (TRD).1 We successfully closed such a hole, presumed secondary to TRD, using standard pars plana vitrectomy and membrane peeling, without further intervention specifically addressing the macular hole.

Case report

A 45-year-old insulin-dependent diabetic man was referred with a recent drop in vision of the left eye; best-corrected visual acuity (BCVA) was 6/6 OD and 6/36 OS. Anterior segment examination was unremarkable except for bilateral pseudophakia. Fundus examination revealed postscatter-photocoagulation PDR OU and an extramacular TRD OS. Both eyes underwent additional panretinal photocoagulation. After 2 months, the new vessels regressed; but increasing vitreoretinal traction caused cystoid macular oedema OS. Vitrectomy was advised; but the patient returned only after 4 months, when BCVA had decreased to 2/60 due to macular TRD with a full-thickness macular hole (Figure 1(a)). Optical coherence tomography confirmed the macular hole, and revealed a vitreous membrane bridging the hole (Figure 1(b)). Pars plana vitrectomy, removal of fibrous proliferation, air–fluid exchange, and silicone oil injection were performed without complication. No attempt was made to peel the internal limiting membrane (ILM). After 1 month, left eye had an attached retina and closed macular hole (Figure 1(c)), confirmed by OCT (Figure 1(d)). Silicone oil was removed 6 months later. At 1 year postvitrectomy, significant posterior capsular opacification was noted. After YAG capsulotomy, BCVA improved to 6/36.

Figure 1
figure 1

(a) Fundus photograph of the left eye demonstrating a tractional macular detachment and a vertically-oriented oval full-thickness macular hole. The arrow shows the direction of optical coherence tomography (OCT) scan. (b) OCT demonstrating full-thickness macular hole with vitreous bridging the thickened edges of the hole. Only the superior edge (arrow) is everted. The base of the hole is not seen due to retinal elevation in excess of 2 mm. (c) Fundus photograph of the left eye 1 month after vitrectomy demonstrating an attached macula, closure of the macular hole, folds in the internal limiting membrane, residual glial tissue along the superior arcade, and light reflections from the silicon oil interface. (d) OCT of the left fovea 1 month after vitrectomy demonstrating a closed macular hole, minimal central (228 μm) and nasal macular thickening and subfoveal pigment epithelial hypertrophy.

Full size image

Comment

Macular holes in PDR are caused by fibrovascular traction at the edges of a fovea weakened by oedema and ischaemia. Particularly when eccentric, they may also result from a ruptured cyst of macular oedema, or from sudden dissection of the posterior hyaloid by a premacular haemorrhage. Idiopathic macular holes also occur coincidently with PDR.1, 2, 3 Holes in a TRD typically change its profile from concave to convex.4 The concave profile of TRD and the vertically-oriented macular hole in this patient pointed to fibrovascular traction, possibly exacerbated by preceding cystic changes. The preoperative OCT demonstrated a tractional membrane bridging the hole, everting its superior edge, and flattening the inferior edge. This was in contrast to the ‘pregnant drawbridge’ OCT configuration characteristic of idiopathic macular holes.5 Although we could not rule out inadvertent removal of ILM during membrane removal, it was not specifically attempted.

Though macular hole-closure rates in PDR are comparable to those in nondiabetic eyes, visual outcomes are compromised by coexisting macular oedema, ischaemia, and TRD.2, 4 The visual recovery in this case could also have been influenced by the surgical delay, and toxicity of silicone oil to the retinal pigment pithelium/photoreceptors.6 We report a good surgical and visual outcome in a rare case of tractional macular hole in diabetic TRD.