Article Text

Randomised controlled trial of corticosteroid regimens in endothelial corneal allograft rejection
  1. Chief of Ophthalmology, El-Maghraby Eye and Ear Hospital, Khozam St Kilo 3 Makkah Road, PO Box 7344, Jeddah 21462, Saudi Arabia
  1. med_dir{at}
  1. Moorfields Eye Hospital, London EC1V 2PD
  1. f.larkin{at}

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,—In the abstract of their paper Huddeet al 1 conclude that “In treatment of graft rejection, additional systemic treatment with 500 mg of methylprednisolone yields no significant benefit over intensive local corticosteroids alone.”

The authors based their statement on the findings that a first episode of endothelial graft rejection could not be reversed in three of 19 patients (16%) treated with local steroids alone, while none of 17 failed to reverse when given additional systemic steroid.

Even though the benefit of the additional steroid therapy may not have been statistically significant because of the small numbers, there is no indication that with larger numbers it might not very well become significant. The authors did not, however, limit their dismissal to a lack ofstatistical significance of the outcome, but broadened it to the wider summary statement of “no significant benefit”. Their data do not support such a sweeping condemnation. Reversal of the first graft rejection episode without a single failure in 17 patients would certainly constitute a strong clinical argument in favour of additional treatment with systemical steroids.

The authors do themselves admit that the statistical power of the study was such that it would only have been able to detect a difference in outcome of the order of 40%. Do they then reason that a difference in outcome of less than 40% is to be regarded as clinically irrelevant? Why was such an arbitrary statistical straightjacket chosen for this study?

Considering that no adverse effects of the systemic steroid treatment were observed, I regard the outcome of the study of clinical relevance, particularly since an increased cure rate of the first (and possibly successive) rejection episode is also likely to affect the long term outcome of these grafts.

In my opinion, this study does not disprove the efficacy of additional systemic steroid treatment for initial episodes of graft rejection.



Editor,—Dr Teichmann is correct in stating that a larger sample size might possibly have demonstrated that graft recipients treated with systemic, in addition to topical, steroid have statistically significant improvement in outcome. However, as stated in the conclusion of the paper, we do not believe that our data are evidence of a major beneficial effect. If there is a small benefit, it is for readers to judge whether it justifies systemic steroid in addition to topical steroid. Weighing up possible benefits with risks, inconvenience, and cost is a decision so often encountered throughout therapeutic medicine.

We would make two further points in response. Firstly, in our study the rejection episode was reversed in a much higher proportion of patients than in previously reported studies; the power calculation used in planning the trial was based on these reports. Secondly, our analysis of combined graft survival and rejection-free survival in the two treatment groups (Fig 4 in the paper) took into account the reversal of rejection in all systemic steroid treated patients, yet indicated very similar outcomes (indeed, marginally superior survival in the topical treatment group, not statistically significant) at 24 months from recruitment, when follow up was terminated.

Linked Articles