Long-term natural history of spontaneous closure of a full-thickness macular hole in retinitis pigmentosa

Summary

We report 18 years of follow-up in a middle-aged Caucasian woman with retinitis pigmentosa (RP) and a non-tractional full-thickness macular hole (FTMH). The case was initially reported after 3 years of observation; this update provides a further 15 years of follow-up and, to our knowledge, represents the longest documented natural history of spontaneous FTMH closure in RP.

At presentation, spectral-domain optical coherence tomography (SD-OCT) demonstrated a right-eye FTMH without vitreomacular traction. The patient elected conservative management with serial monitoring. The FTMH subsequently closed spontaneously, leaving a stable lamellar configuration that has remained anatomically unchanged without recurrence for almost two decades. Visual acuity gradually declined, likely due to progressive retinal degeneration and cataract progression, while OCT confirmed long-term foveal stability.

This case demonstrates that spontaneous closure of a non-tractional FTMH in RP, although rare, can remain stable in the long term, supporting careful observation in selected patients.

Background

Retinitis pigmentosa (RP) is a progressive inherited retinal dystrophy characterised by photoreceptor degeneration and retinal pigment epithelium dysfunction. It typically causes nyctalopia, progressive peripheral visual-field constriction and eventually, central visual loss. Cystoid macular oedema (CMO) and epiretinal membrane are among the most common macular complications and have been reported in up to half of patients. In contrast, the development of full-thickness macular hole (FTMH) in RP is uncommon, with reported prevalence estimates of 0.5%–4.5%.¹

Unlike the aetiology of idiopathic FTMH, which is usually tractional, the pathogenesis of FTMH in RP remains incompletely understood. Proposed mechanisms include chronic cystoid macular degeneration, Müller-cell dysfunction and progressive retinal tissue fragility with vitreomacular traction or epiretinal membrane formation.

Pars plana vitrectomy (PPV) with internal limiting membrane (ILM) peeling has been used to treat RP-associated FTMH, and several reports describe anatomical macular hole closure.^{2 3} However, functional improvement may be limited by irreversible photoreceptor loss and residual underlying retinal degeneration. Surgery may also carry particular risk in eyes with RP, including iatrogenic retinal tears, retinal detachment, cataract progression and further macular atrophy. OCT has transformed the assessment of these cases by enabling detailed evaluation of macular morphology and the vitreoretinal interface.⁴

Spontaneous closure of idiopathic macular holes is well recognised, but reports in RP remain rare. To our knowledge, the original report of this patient remains the only published case of spontaneous FTMH closure in RP.⁵ The present report provides a unique and valuable extended 18-year follow-up of FTMH in RP and documents the long-term anatomical and functional course after conservative management.

Case presentation

A middle-aged Caucasian woman with longstanding RP was referred 18 years ago for assessment of an FTMH in the right eye. She reported symptoms of life-long nyctalopia and progressive peripheral vision loss. She had a strong family history of RP affecting her father and brother, suggesting an autosomal dominant pattern of inheritance. Genetic testing was unavailable at presentation and has not subsequently been performed. Her medical history included well-controlled hypertension. She had no RP-associated syndromic features, such as deafness or cardiac abnormalities.

At presentation, best-corrected visual acuity (BCVA) was LogMAR 0.9 (Logarithm of the Minimal Angle of Resolution 0.9 equivalent to Snellen 6/48) in the right eye and LogMAR 0.3 (6/12) in the left eye. Anterior segment examination showed early bilateral nuclear sclerotic cataracts. Intraocular pressures were within normal limits. Fundus examination showed typical features of RP, including bilateral mid-peripheral bone-spicule pigmentation, attenuated retinal vessels and waxy optic disc pallor. Spectral-domain optical coherence tomography (SD-OCT) confirmed an FTMH in the right eye (figure 1A), with no evidence of epiretinal membrane, vitreomacular traction or lamellar hole-associated epiretinal (LHEP) proliferation. A posterior vitreous detachment (PVD) was already present.

Figure 1. Baseline and 3-year right foveal SD-OCT B-scans and colour fundus photographs. Top row: Baseline SD-OCT (A) showing a right FTMH (arrow) and the corresponding colour fundus photograph (B) showing the location of the FTMH (arrow). Bottom row: 3-year SD-OCT (C) showing spontaneous FTMH closure with a residual lamellar defect and hyperreflective tissue (arrow), with the corresponding colour fundus photograph (D) showing the location of the residual lamellar defect (arrow). Panels are reproduced from the original report.⁵ FTMH, full-thickness macular hole; SD-OCT, spectral-domain optical coherence tomography.

Given the absence of tractional pathology and the patient’s preference to avoid surgery, conservative management was chosen. Serial OCT showed progressive reduction in hole size and spontaneous closure by 12 months, leaving a lamellar defect with partial foveal reconstitution, remaining evident at the 3-year visit (figure 1C). BCVA improved slightly to LogMAR 0.8 (Snellen 6/36). During 18 years of follow-up, right BCVA gradually declined to LogMAR 0.9 (Snellen 6/48) at 10 years and LogMAR 1.0 (Snellen 6/60) by 18 years. Serial OCT demonstrated long-term anatomical stability of the right fovea without FTMH recurrence (figures 2 and 4). The left eye showed persistent cystoid macular changes and subsequently developed a lamellar macular hole (figures 3–5).

Figure 2. Right foveal SD-OCT at 10 years (E) and 18 years (F), showing a stable lamellar defect with cystoid macular changes and lamellar hole-associated epiretinal proliferation (arrows). Compared with figure 1, the grey-scale Spectralis images more clearly demonstrate residual outer retinal tissue. Asterisks indicate residual tissue columns bridging the foveal and parafoveal inner and middle retina, considered to represent Müller-cell tissue. SD-OCT, spectral-domain optical coherence tomography.

Figure 3. Left foveal SD-OCT at 10 years (G) and 18 years (H) showing cystoid macular changes and a lamellar hole. The arrow indicates subfoveal preservation of the four outer retinal bands—the external limiting membrane, ellipsoid zone, interdigitation zone and retinal pigment epithelium-Bruch’s membrane complex—using IN.OCT nomenclature. Asterisks indicate loss of these in the parafoveal, perifoveal and peripheral macula. SD-OCT, spectral-domain optical coherence tomography.

Figure 5. Left-eye ultra-widefield pseudo-colour fundus photograph and fundus autofluorescence at 18 years, showing similar RP-associated peripheral pigmentary change, vascular attenuation and parafoveal autofluorescence abnormalities. RP, retinitis pigmentosa.

Figure 4. Right-eye ultra-widefield pseudo-colour fundus photograph and fundus autofluorescence at 18 years. The colour image shows widespread mid-peripheral bone-spicule pigmentation, vascular attenuation and optic disc pallor, consistent with advanced RP. Fundus autofluorescence shows parafoveal hyperautofluorescence surrounded by hypoautofluorescence, corresponding to retinal pigment epithelium and photoreceptor loss. RP, retinitis pigmentosa.

The patient underwent uncomplicated left cataract surgery during follow-up. A 3-month trial of topical dorzolamide three times daily to both eyes for CMO was ineffective and subsequently discontinued. During later follow-up, she underwent treatment for jaw cancer. She continues to receive low-vision support and uses magnifiers for reading.

Investigations

Baseline and 3-year follow-up assessments used SD-OCT (Topcon 3D OCT-1000 Mark I; Topcon Corporation, Tokyo, Japan) and colour fundus photography (NIDEK AFC-300). Subsequent assessments used SD-OCT (Heidelberg Spectralis) and ultra-widefield pseudo-colour fundus photography with fundus autoflorescence (Optos California).

Baseline SD-OCT of the right eye demonstrated a FTMH without vitreomacular traction. Serial imaging documented spontaneous closure with formation of a stable lamellar configuration maintained at 3, 10 and 18 years.

Differential diagnosis

Multimodal imaging showed no evidence of vitreomacular traction, epiretinal membrane, LHEP, choroidal neovascularisation or retinal trauma as alternative differential diagnoses for the FTMH.

Treatment

Observation: Serial observation of the right FTMH.

Medical treatment: A 3-month trial of topical dorzolamide for CMO, with no clinical benefit.

Surgery: Uncomplicated left cataract surgery.

Rehabilitation: Ongoing low-vision service support.

Outcome and follow-up

At 18 years, the right FTMH remained closed, with a stable lamellar defect on OCT. Right-eye BCVA had declined to LogMAR 1.0 (Snellen 6/60). Left-eye BCVA remained LogMAR 0.3 (Snellen 6/12) after cataract surgery, despite persistent CMO.

Discussion

FTMH is an uncommon but recognised complication of RP. The pathogenesis of RP-associated FTMH is likely multifactorial, involving cystoid macular degeneration, Müller-cell dysfunction and progressive foveal fragility. In this case, the absence of vitreomacular traction suggests that progressive de-roofing of cystoid spaces and loss of retinal tissue loss produced a predominantly degenerative or atrophic, rather than tractional, FTMH.

Macular holes were traditionally staged using the Gass system, which describes progression from foveal elevation without a retinal break (stage I), through small full-thickness defect (stage II), to an established FTMH without PVD (stage III) and finally to FTMH with complete PVD (stage IV).^{6 7} More recently, the International Vitreomacular Traction Study Group introduced an OCT-based classification according to minimum linear diameter, vitreomacular interface status and aetiology, where FTMH is categorised as small (<250 µm), medium (250–400 µm) or large (>400 µm).⁸

These classifications are clinically useful because hole size and tractional status influence both the likelihood of spontaneous closure and surgical outcomes. However, FTMH in RP may represent a distinct degenerative phenotype driven by chronic cystoid change and retinal tissue fragility rather than classic anteroposterior vitreomacular traction. Standard idiopathic macular hole classifications should therefore be applied cautiously when planning management in RP.

Current evidence to guide the management of RP-associated FTMH remains limited. PPV with ILM peeling has achieved anatomical closure in some series^{2 3 9 10} but functional gains are often modest. More complex techniques, including inverted ILM flaps and autologous lens capsule transplantation, have also been described,^{11 12} but may entail greater procedural complexity and risk. A reported recurrence of macular hole 2 years after apparent successful surgery further illustrates the variable course in this population.¹³

This case suggests that observation may be a reasonable initial strategy in selected patients with non-tractional FTMH, particularly when underlying retinal degeneration limits the expected visual benefit of surgery. The 18-year follow-up shows that spontaneous closure can remain anatomically stable while avoiding surgical risk. Management should be individualised in FTMH should be individualised and guided by multimodal imaging, vitreomacular interface morphology, underlying retinal disease and expected functional prognosis.

Limitations include the absence of genetic confirmation and the inherent constraints of a single observational case of a rare condition. However, the extended follow-up provides rare and valuable longitudinal evidence on the natural history and conservative management of RP-associated FTMH.

Learning points.

• Spontaneous closure of a non-tractional full-thickness macular hole in retinitis pigmentosa is possible and may remain anatomically stable for almost two decades.

• Careful observation may be an appropriate management option in selected patients, particularly when traction is absent and the anticipated functional benefit of surgery is limited.

• Subsequent visual decline may reflect progression of the underlying retinal dystrophy rather than recurrence of the macular hole.