Article Text

Download PDFPDF

Peripheral ulcerative keratitis in rheumatoid arthritis: successful use of intravenous cyclophosphamide and comparison of clinical and serological characteristics
  1. A R Clewes1,
  2. J K Dawson2,
  3. S Kaye3,
  4. R C Bucknall1
  1. 1Rheumatic Diseases Unit, Royal Liverpool University Hospital, Liverpool, UK
  2. 2Department of Rheumatology, St Helens and Knowsley Hospitals NHS Trust, St Helens, Merseyside, UK
  3. 3Department of Ophthalmology, St Paul’s Eye Unit, Royal Liverpool University Hospital, Liverpool, UK
  1. Correspondence to:
    Dr A R Clewes
    Rheumatic Diseases Unit, Link 7c, Royal Liverpool University Hospital, Prescot Street, Liverpool, L7 8XP, UK;

Statistics from

Peripheral ulcerative keratitis (PUK) is a devastating complication of rheumatoid arthritis (RA) that can lead to rapid corneal destruction (corneal melt) and perforation with loss of vision (fig 1). It is well reported to herald systemic vasculitis and carries a high mortality rate if not aggressively treated.1 There is a paucity of information concerning the clinical and serological characteristics of these patients and no clear consensus about treatment.

Figure 1

 Note the peripheral circumferential damage from corneal ulceration (arrows).

We report the successful management of seven such patients with the use of intravenous cyclophosphamide and compare the serological and clinical characteristics with a comparative group of 150 rheumatoid patients without associated PUK, who attended an outpatient clinic.


All patients received systemic immunosuppressive treatment with intravenous cyclophosphamide and the dose was dependent on age, weight, renal function, and clinical response. All received some form of adjunctive ocular treatment including cyanoacrylate glue, bandage contact lenses, punctual occlusion, and corneal transplants. Two patients also received intravenous corticosteroid treatment with 1 g methylprednisolone (table 1).

Table 1

 Summary of characteristics

Student’s t test was used to compare the melt and comparative groups, the Mann-Whitney U test to compare quantitative data, and a χ2 squared test with Fisher’s exact test to compare frequencies. No difference in duration of disease, presence of rheumatoid nodules, Sjögren’s syndrome, or seropositivity for rheumatoid factor was found between the groups. Disease activity did not correlate with ocular disease and there was no active synovitis at presentation. The melt group were older (median age 71.9 v 58.9 years, p = 0.013) and were more inclined to be antinuclear antibody (ANA) positive (p = 0.03). Serum albumin levels were statistically lower (36.7 g/l v 40.83 g/l, p = 0.012) and immunoglobulin levels polyclonally increased (IgA 3.62 v 2.80 g/l, p = 0.035; IgG 3.93 v 1.42 g/l, p = 0.001; IgM 16.74 v 10.97 g/l, p = 0.005). A previous history of scleritis seemed to correlate with increased risk of PUK (p = 0.002), but the presence of sicca disease did not reach significance (p = 0.165).


No correlation between articular activity and PUK was found, but there was a statistically significant association with the presence of ANA, raised immunoglobulins, and reduced albumin concentrations when compared with the comparative group. This may represent increased subclinical disease activity. The pathogenic mechanisms behind PUK are still unclear but whatever the mechanisms, the important issue is that this condition needs aggressive treatment with immunosuppression.1–8 Foster et al compared the outcome of 34 patients with rheumatoid associated necrotising scleritis and PUK, half of whom received systemic immunosuppressive treatment and half non-immunosuppressive topical conventional treatment. In those not receiving immunosuppression, vasculitis developed in more than 50% of cases with a high mortality rate.1 Jayson and Jones,2 Mcavin et al,3 and Watson and Hayreh4 also suggested that these aggressive destructive ocular lesions herald systemic vasculitis and require aggressive immunosuppressive treatment. Azathioprine,5,6 methotrexate,7,8 and ciclosporin8 have been used with varying success, but cyclophosphamide1,6,9 has been widely used in destructive ocular inflammatory conditions and has a well established role in treating systemic vasculitis.10 Systemic and local steroids are used but we have concerns about the reduction in cornea healing, and steroids alone do not alter mortality.1

We found no incidence of infection or immunosuppressive complication. The vision of all patients was preserved and, to date, none of the patients who received intravenous cyclophosphamide have presented with evidence of systemic vasculitis and this is probably due to the use of immunosuppressive treatment.

We cannot emphasise enough that this is a condition that should be managed by both a rheumatologist and an ophthalmologist, both of whom need to be aware of the acute ocular concerns as well as the long term consequences with associated high mortality in those who do not receive immunosuppressive treatment.1–3


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.