Article Text

Ultrastructural alterations in the stroma adjacent to non-inflammatory corneal perforations associated with long standing rheumatoid arthritis
  1. Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kyoto, Japan
  2. Department of Ophthalmology, Shinkawabashi General Hospital, Tokyo, Japan
  1. Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kyoto, Japan
  2. Department of Ophthalmology, Shinkawabashi General Hospital, Tokyo, Japan
  1. Kohji Nishida, MD, PhD, Department of Ophthalmology, Kyoto Prefectural University of Medicine, Kawaramachi Hirokoji, Kyoto 602, Japan.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Editor,—Patients with rheumatoid arthritis commonly have two types of corneal ulceration. One is a peripheral corneal ulceration thought to be caused by complex mediated hypersensitivity.1 The other is a paracentral corneal ulceration that tends to perforate the cornea rapidly. Its pathogenesis remains unknown, though a key feature is the lack of inflammation, leading to the suggestion that it is the result of surface disease (drying).


We examined two individuals, a 38 year old woman and a 76 year old woman, with unilateral corneal epithelial defects and aqueous tear deficiencies. Both patients had been treated for rheumatoid arthritis for several years; however, neither had been on systemic corticosteroids. The 76 year old had had successful cataract surgery 1 week previously and was using 0.1% betamethasone eye drops. At the time of presentation both women had severe dry eye (possibly secondary Sjögren’s syndrome, though no biopsies were taken to confirm this) as assessed by detailed slit lamp examination with fluorescein and rose bengal staining, and by the Schirmer test. The 38 year old was started on 0.1% fluorometholone and 0.3% norfloxacin and dibekacin eye drops four times a day, and the 76 year old was switched to the topical application of 0.1% betamethasone and 0.3% ofloxacin ointments four times a day. A few weeks later paracentral corneal perforations developed suddenly, though without pain (Fig 1). Unlike peripheral corneal ulceration, little or no inflammation was seen on the ocular surface at any time. Penetrating keratoplasties were performed and a treatment regimen was begun that included artificial tears (for dry eye) and 0.1% betamethasone topically four times a day. In addition, the 76 year old received 2 mg betamethasone intravenously once a day and 1 g flomoxef sodium intravenously twice a day for the first 3 days after surgery; this was followed by betamethasone (1 mg) and antibiotic (300 mg cefzinil; 100 mg, three times a day) taken internally for 1 week. At the time of writing, topical steroids (in both cases 0.1% betamethasone) continue to be used, and both patients are doing well 7 years (38 year old) and 5 years (76 year old) after surgery, with no sign of another perforation.

Figure 1

Paracentral corneal perforation (38 year old woman). There is little or no inflammation on the ocular surface. The anterior chamber is maintained by a therapeutic contact lens.

At the time of surgery, corneal tissue directly adjacent to the paracentral perforation, as well as tissue in the peripheral cornea, was obtained for histopathological study. In line with the lack of clinically detectable inflammation (Fig 1), light microscopic examination of both cases revealed only a few inflammatory cells surrounding the perforation and none in more peripheral areas (data not shown). In both patients’ corneas, electron microscopy documented many electron dense deposits in the extracellular stromal matrix (Fig 2). These amorphous deposits were intimately associated with collagen, and might represent degraded and/or aggregating collagen fibrils resulting from the melting process. Also detected were numerous atypical, thin fibrils (Fig 2; inset). These were fairly widely interspersed with normal diameter collagen though, on occasion, they appeared in groups. Interestingly, these abnormal electron microscopic features were observed not only in stroma right at the edge of the perforation, but in more peripheral regions of the cornea as well.

Figure 2

Corneal stroma surrounding the perforated area (76 year old woman). Electron dense material (arrow), possibly aggregated collagen, is located throughout the extracellular matrix. Also (inset), many atypical, thin fibres (arrowheads) are interspersed with collagen. Bar=200 nm (main figure) and 100 nm (inset).


The general clinical presentations of both these individuals, especially the lack of inflammation, are similar to previous reports of perforated paracentral corneas.2 Often, infiltrating macrophages and T cells that are associated with HLA class II antigens are located around corneal ulcers,3 and various cytokines and proteinases secreted from these cells are thought to trigger a corneal perforation. However, this pathogenesis is unlikely in our cases because little or no inflammatory cell infiltration was detected.

As in all connective tissues, collagen imparts to stroma by far the majority of its tensile strength. It follows that ultrastructural collagen alterations might cause our patients’ corneas to become somewhat fragile, predisposing them to non-inflammatory corneal perforations. With this in mind, it is noteworthy that the alterations of the corneal extracellular matrix we report here closely resemble those seen in ruptured hand tendons of rheumatoid arthritis patients.4 It is possible, therefore, that a similar structural weakening of both these connective tissues occurs and is associated with collagen abnormalities in long standing rheumatoid arthritis.