Article Text

Autologous serum in epithelial defects: reply
  1. J Dart1,
  2. A Poon2
  1. 1Moorfields Eye Hospital, London, UK
  2. 2Royal Victorian Eye and Ear Hospital, Melbourne, Australia
  1. Correspondence to: John Dart, MA, DM, FRCS, Moorfields Eye Hospital, 162 City Road, London EC1V 2PD, UK; j.dart{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.

Mukerji and colleagues have criticised several aspects of our study and justified their views with unrepresentative quotations from the literature.1,2

Choice of time period before an epithelial defect can be described as “persistent”

There is no accepted definition of persistent epithelial defect (PED) that includes a time period. We favour the definition given in one text “when the epithelium fails to regrow over a defect within the expected time course.”3 For the purpose of a study the time course must be specified. Mukerji et al prefer 2 weeks4 whereas in another recent paper 10 days was chosen with the proviso that the study treatment could start earlier if there was progression to perforation.5 The penalty for patients at high risk of corneal melt, in the presence of a persistent defect, is substantial. In our study 2/13 patients had an epithelial defect for less than 2 weeks before inclusion in the study; Mukerji et al can ignore the data from these patients if they wish.

Washout period with preservative-free artificial tears

No “washout period” was used in this study, as we were trying to identify whether serum would have an additional effect on epithelial healing over preservative-free artificial tears and conventional therapy in patients with severe ocular surface disease. We were not trying to compare the efficacy of serum against preservative-free lubricants in healing PEDs. Conventional therapy was continued as serum alone could not fully address all the ocular surface problems, including aqueous tear deficiency and eyelid trauma, so that giving serum without continuing the other treatments may have caused harm to some of the patients. Other studies have adopted similar protocols with a PED defined as a defect that persisted “despite conventional treatment such as artificial tears or extended wear contact lenses” with no washout period4 or that allowed a reduction in the 10 day wash out period if corneal ulceration progressed.5 In a clinical study some ethics committees may not condone the treatment of persistent defects without non-preserved antibiotic prophylaxis against bacterial keratitis.

Keratoplasty patients being treated without waiting for the corneal epithelial defect to heal by itself

These patients had both failed previous grafts because of corneal perforation resulting from PEDs and one patient had recurrent epithelial breakdown, responding to the re-introduction of serum drops, after their initial withdrawal. We thought that, for some readers, this might be useful data to add to this descriptive study.

The rationale behind the use of 100% serum when previous studies have proved the efficacy of a 20% solution

The previous study Mukerji et al quote is, like ours and all the other studies on this subject, a descriptive study and no more proves that serum drops at 20% work than does our own with 50% and 100%. Other studies, all uncontrolled, have used from 20%–30% serum4,6 for both persistent defect and dry eye. All these studies use empirical regimens for serum concentration and dosing frequency. Unlike Mukerji et al, we have some patients (one reported in the study) who prefer 100% serum to 50%. Mukerji et al are probably also unaware of our study demonstrating that 100% serum is less toxic that 50% serum to cultured corneal epithelial cells7; this provides some laboratory data to support the use of 100% serum. Lastly, the manufacture of 100% serum requires less handling so reducing the risk of contamination.

The use of the slit lamp micrometer

Mukerji et al have missed the point that the treatment end point was closure of the defect, for which their proposed technique is inappropriate, rather than rate of closure, for which it would be ideal.

The deficiencies in our study are not the points that Mukerji et al have made but the fact that it was uncontrolled. The study was planned as a randomised controlled trial but could not be carried out as such for the reasons alluded to in the last paragraph. These regulatory problems are being overcome in the United Kingdom. We should all hope to see randomised treatment trials carried out in this area as well as the laboratory studies necessary to support the clinical application of serum.