Intracorneal hypopyon within a DSAEK graft

Abstract

Introduction:

Intracorneal hypopyon is a rare condition involving the accumulation of inflammatory cells and debris within the structures of the cornea. We present an unusual case of an intracorneal hypopyon within a Descemet-stripping automated endothelial keratoplasty (DSAEK) graft.

Patient and Clinical Findings:

A 54-year-old woman status post DSAEK for bullous keratopathy secondary to prior trabeculectomy and cataract surgery developed a corneal ulcer. Corneal cultures grew Serratia marcescens, and topical antibiotics were given. Despite clinical improvement, a dense, deep corneal opacity with overlying epithelial defect persisted.

Diagnosis, Intervention, and Outcomes:

Anterior segment optical coherence tomography (AS-OCT) identified an intracorneal hypopyon between the Descemet membrane and corneal stroma of the DSAEK graft. Antibiotics were continued, while topical steroids, initially reduced in response to infection, were increased modestly. The patient's condition and hypopyon gradually improved.

Conclusions:

Ophthalmologists should consider the possibility of an intracorneal hypopyon masquerading as a persistent stromal infiltrate in the setting of infectious keratitis. Unlike unresponsive corneal infiltrates, intracorneal hypopyons may represent sterile inflammatory debris after resolved infection. AS-OCT was critical to distinguish the intracorneal hypopyon from a persistent infiltrate, which may have prevented unnecessary and/or invasive interventions.

A hypopyon represents an aggregate of inflammatory cells within the anterior chamber in the setting of intraocular inflammation. An intracorneal hypopyon, also known as “pseudohypopyon” or “double hypopyon” (the latter in the presence of an anterior chamber hypopyon), is a rare clinical entity, wherein a hypopyon forms within a potential space in the cornea, typically described as either a cavitating abscess in the corneal stroma or a bullous separation of the Descemet membrane (DM) and endothelium from the overlying corneal stroma.^1,2 Intracorneal hypopyons typically form spontaneously in the setting of microbial keratitis but may occur in the absence of infection (ie, sterile hypopyon).^1–7 To the authors' knowledge, this is the first described case of a corneal hypopyon within a Descemet-stripping automated endothelial keratoplasty (DSAEK) graft, between the stroma and the DM of the graft.

Patient Consent Statement

Written informed consent to publish deidentified protected health information and images was obtained from the patient. The Mass General Brigham Institutional Review Boards waived the need for ethics approval. This study was conducted in accordance with HIPAA regulations and the tenets of the Declaration of Helsinki.

CASE REPORT

A 54-year-old woman with a history of chronic angle-closure glaucoma and pseudophakic bullous keratopathy in both eyes, after cataract extraction over 10 years prior, established care at our institution for ongoing management. She has advanced glaucoma and a history of trabeculectomy in each eye. She underwent penetrating keratoplasty in the right eye and DSAEK in the left eye for visual rehabilitation. The DSAEK graft failed within 2 years, and she underwent a second DSAEK surgery which also failed.

Two years after her second DSAEK, she presented for a routine follow-up with increased pain and blurriness in her left eye. Her visual acuity was hand motions (from a baseline of 20/150 corrected distance visual acuity), and her intraocular pressure was 9 mm Hg. Slitlamp examination was notable for a large epithelial defect (approximately 8.0 × 4.0 mm) with a deep, dense, white infiltrate (approximately 3.5 mm in diameter). The corneal infiltrate was cultured, and topical fortified vancomycin and tobramycin were started at 1 drop hourly to the left eye. Her long-term topical prednisolone—which had been maintained at 4 times per day because of graft failure—was reduced to twice daily.

The corneal cultures grew Serratia marcescens, with a broad range of sensitivities, including tobramycin, levofloxacin, and ciprofloxacin. She was maintained on topical tobramycin and moxifloxacin for 1 week, followed by moxifloxacin alone, at gradually decreasing frequency. Despite overall clinical improvement, a persistent area of dense opacity was noted centrally with an overlying persistent epithelial defect (Figure 1).

Figure 1.

A: Slitlamp photograph of the left eye demonstrating dense corneal opacity. B: AS-OCT of the left eye showing, from posterior to anterior: the Descemet membrane of the DSAEK graft (outlined by blue dashes), the intracorneal hypopyon (red asterisk), the posterior stroma of the DSAEK graft (outlined by yellow dashes), and the anterior border of the DSAEK graft (outlined by red dashes).

An anterior segment optical coherence tomography (AS-OCT) scan revealed that the area of persistent opacity could be attributed to a collection of material in the interface between the DSAEK graft stroma and its DM—that is, an intracorneal hypopyon. There was no anterior chamber hypopyon, minimal anterior chamber inflammation, and there was no detachment of the DSAEK graft. Topical steroids were returned to her baseline of prednisolone 4 times daily, and the patient experienced symptomatic improvement. Topical antibiotics were reduced to prophylactic dosing (twice daily moxifloxacin), and over the ensuing months, the intracorneal hypopyon improved (Figure 2).

Figure 2.

AS-OCT depicting improvement of intracorneal hypopyon in the left eye with the use of topical steroids and antibiotics, but with thinning of the overlying host corneal stroma. From posterior to anterior: the Descemet membrane of the DSAEK graft (blue dashes), the hypopyon (red asterisk), the posterior stroma of the DSAEK graft (yellow dashes), and the anterior border of the DSAEK graft (red dashes).

As of her latest follow-up, 6 months after initial presentation, her epithelium has healed under a bandage contact lens. Given severe stromal thinning, once the inflammation resolves, the patient will undergo penetrating keratoplasty for visual rehabilitation.

DISCUSSION

In this report, we have described a case of a hypopyon forming between the donor corneal stroma and the donor DM of a DSAEK graft. The clinical appearance as a nonhealing deep corneal stromal infiltrate and the absence of an anterior chamber hypopyon made this a diagnostic challenge.

The presence of an intracorneal hypopyon, especially in the absence of an anterior chamber hypopyon, can be difficult to identify on clinical examination. In many of the published reports describing intracorneal hypopyon, a clear fluid–fluid level of pus from clear fluid is visible within the cavity between the host corneal stroma and DM.^1–3,5,6 If there is an absence of a fluid–fluid level, as in this reported case, the intracorneal hypopyon may appear as a deep corneal stromal infiltrate. AS-OCT was therefore critical for accurate diagnosis in our case. In our case, a careful review of the barriers between corneal structures (ie, DSAEK–host interface and DSAEK stroma–DM interface) was necessary. Owing to the presence of a small amount of fluid surrounding the hypopyon, edges of the hyperreflective focus posterior to the DSAEK stroma could be identified and used to confirm the diagnosis (Figure 1). As the hypopyon resolved, the relative amount of fluid within the Descemet detachment increased, making the delineation more readily apparent (Figure 2). Misdiagnosis or delayed diagnosis of a corneal hypopyon may have prompted repeated cultures or even more invasive diagnostic and/or therapeutic interventions.

In addition to the failing graft, the presence of the intracorneal hypopyon may have contributed to the delay in healing observed in this patient. Hypopyon consists of decaying leukocytes, fibrin, and necrotic debris.⁸ We hypothesize that cytokines and chemokines released by decaying leukocytes in a confined space just posterior to the corneal stroma served to prolong inflammation and delay wound healing.

Intracorneal hypopyon may be associated with severe corneal infections, and outcomes in reported cases associated with infection are poor. Several reported cases ultimately required evisceration.^1,5 In another case, the hypopyon resolved spontaneously along with the resolution of the corneal infection.⁴ Surgery to drain the intracorneal hypopyon may be considered, although risks of this procedure include intraocular spread of infection and chronic Descemet detachment.^9,10 In our case, as the intracorneal hypopyon improved with topical medical therapies, surgical intervention was avoided, thereby minimizing the risk for further complications. Furthermore, surgery was avoided in our case, as the hypopyon was presumed sterile given the improvement while on steroids and reduction in antibiotics to prophylactic dosing.

In summary, we have presented an unusual case of an intracorneal hypopyon forming between the donor corneal stroma and donor DM of a DSAEK graft in the setting of Serratia marcescens keratitis, without associated anterior chamber hypopyon. Intracorneal hypopyon may mimic or accompany a nonhealing corneal ulcer, and AS-OCT may be important in accurately diagnosing this condition.

WHAT WAS KNOWN

• Intracorneal hypopyons typically form in the setting of microbial keratitis with an associated anterior chamber hypopyon.

• Misdiagnosis of intracorneal hypopyons may lead to unnecessary diagnostic and/or therapeutic interventions.

WHAT THIS PAPER ADDS

• Ophthalmologists should be aware of the possibility of intracorneal hypopyons within the donor corneal stroma and Descemet membrane layers of a DSAEK graft.

• Without a fluid level intracorneal hypopyon may evade clinical diagnosis and can masquerade as an unresolving corneal infiltrate. AS-OCT is useful in successfully diagnosing this condition.