Dear Editor,

An aim to provide an optimised keratoprosthesis, with excellent biointegration, and all other properties meeting ideal requirements, is one we share with the authors of the recent article entitled “Hydroxyapatite promotes superior keratocyte adhesion and proliferation in comparison with current keratoprosthesis skirt materials”.[1] However, the current paper includes some points that require clarification.

PHEMA (spelled in full: poly(2-hydroxyethyl methacrylate)is a non-toxic polymer of the toxic monomer HEMA, though cytotoxicity is still possible if non-reacted monomer has not been fully removed. The ' polyhydroxyethyl methacrylate' the authors obtained for their study was not fully described, and may have been contact lens blanks, unlikely to have been processed for 'implantable quality'. Whether the samples had been fully extracted was not stated, nor was the hydration of the samples when used for the study. Contact lens blanks are not designed for cell adhesion and the results of this study, with regard to PHEMA, are entirely predictable and have been previously reported.

The commercially available keratoprosthesis AlphaCor is made from a form of PHEMA, specifically modified for its intended purpose within the cornea. In particular, the AlphaCor OPTIC is made from a relatively low water content, but hydrated, microporous form, similar to the samples evaluated by the authors, specifically because it does not encourage cellular adhesion (epithelial coverage is not desired for this model, nor would adherent posterior cells and membranes be desirable).

In contradistinction, the biointegratable SKIRT region of AlphaCor is made from a macroporous form of PHEMA with a very high water content; this material, with its interconnecting channels, has been optimised to promote viable biocolonisation, which has been extensively described in the literature. The authors of the present article do concede that cells 'may behave differently in colonising a 3-D porous keratoprosthesis skirt': indeed they do. Further, very subtle modifications of the sponge structure significantly affect all aspects of biointegration.

Both early trial results, such as the preliminary cases cited by the authors, and current results for over 250 AlphaCor devices, have been extensively presented and made available to all device users. Histology now available from AlphaCor devices explanted from human recipients confirms that the biointegration process in humans is similar to that previously shown in the animal model, and maintained in the long term. As expected, specific inflammatory processes can cause localised reversal of biointegration in areas of stromal melting. Certainly, porosity itself does not prevent melting processes, as is also seen in relation to hydroxyapatite keratoprostheses and orbital implants.

There is no argument that keratoprosthesis materials and design require ongoing revision and improvement. The authors' findings in relation to hydroxyapatite are interesting, although as they note, this rigid material has its own limitations. Novel approaches are undergoing early evaluation and may offer benefits. However, at present, in our view, AlphaCor is a device worthy of consideration for those in whom a donor graft would fail.

Disclosure: Celia Hicks is Medical Director of CooperVision Surgical, manufacturer of AlphaCor. The Biomaterials and Polymer Research department of the Lions Eye Institute has a financial interest with CooperVision Surgical through support of departmental funding, travel and research.

References

1. J S Mehta, C E Futter, S R Sandeman, R G A F Faragher, K A Hing, K E Tanner, and B D S Allan. Hydroxyapatite promotes superior keratocyte adhesion and proliferation in comparison with current keratoprosthesis skirt materials. Br. J. Ophthalmol. 2005; 89: 1356 – 1362.

Dear Editor,

The article by Al-Hazmi et al.[1] states that combined trabeculotomy- trabeculectomy with mitomyocin C (CTTM) gave better results than trabeculotomy alone for primary congenital glaucoma (PCG) at the King Khaled Eye Specialist Hospital (KKESH) in Riyadh, Saudi Arabia, between 1982 and 2002. For moderate PCG the success rate is stated as 40% and 80% for trabeculotomy and CTTM, respectively. For severe PCG the stated success rate is 10% and 70% for trabeculotomy and CTTM, respectively. However, without more specifics regarding when the trabeculotomies were performed at KKESH, the authors cannot advocate CTTM over trabeculotomy for moderate and severe PCG.

As the article states, over the years at KKESH the success rate for trabeculotomy for PCG dramatically improved (29% from 1982-1990; 47% from 1991-94; 82% from 1995-2002) The authors attribute these improved results over the years to improved primary healthcare facilities within the kingdom, earlier referrals, better equipment availability, and surgeons becoming more adapt at surgical intervention. In contrast, CTTM for PCG was first performed at KKESH in 1994 with less of a “learning curve”; the success rate from 1994 -2002 was 72%. The complication rate, however, was higher for CTTM than for trabeculotomy.

Because initial trabeculotomy success for PCG at KKESH has dramatically increased with time, it is important to know how many of the reported trabeculotomy failures for moderate and severe PCG were from the earlier periods of the hospital. This information was not in the manuscript. It may be that trabeculotomy as currently performed at KKESH for moderate and severe PCG has a success rate similar to that of CTTM for the same patient population with less surgical complications.

References

1. Al-Hazmi A, Awad A, Zwaan Z, et al. Correlation between surgical success rate and severity of congenital glaucoma. Br J Ophthalmol 2005;89:449–453.

Dear Editor,

We read with interest the paper by Cazabon et al.[1] on the important emerging problem of sudden visual loss after removal of silicone oil. We have seen a similar pattern of visual loss in our own patients typically seen in macula on detachments associated with giant retinal tears. We have identified 12 cases between 2 units (St Thomas’, London and Sunderland Eye Infirmary), but 5 of these clearly describe onset of visual loss before oil removal (onset between 1 month and 5 months after oil insertion).[2] Results of investigations were very similar to those reported by Cazabon et al. In 4 of 5 pattern ERG was suggestive of macular dysfunction. The timing of onset of visual loss obviously alters the potential aetiology, which as stated is unknown.

In their paper, information on acuity for cases 2 and 3, between 1 week after oil insertion and oil removal is not provided. Did these cases have visual loss preceding oil removal? Developing cataract can obviously hinder interpretation of acuity measurements. In our cases the symptoms described did not fit with cataract (scotoma, red desaturation) and persisted if any cataract was removed.

We have seen a further case since this report, a 46 year old female with a giant retinal tear and macula-on retinal detachment affecting the right eye. Acuity reduced during period of tamponade from 6/6 2 weeks after oil insertion to 6/36+1, which did not recover after oil removal. She reported a central negative scotoma. Electrophysiology suggested macular dysfunction.

We have speculated that photo-toxicity may play a role, as oil transmits light more in the blue spectrum than aqueous.[3] The fat soluble macula pigments, lutein and zeaxanthin, are thought to protect the macula from photo-oxidative damage. Silicone oil has previously been reported to dissolve fat soluble elements from the retina.[4]

We measured the macular pigment optical density (MPOD) in this case using a modified confocal scanning laser ophthalmoscope and 2-wavelength autofluorescence technique 3 weeks after oil removal. The results showed a substantially reduced MPOD in the eye which had silicone oil compared to the fellow eye. Although the peak MPOD, at the foveal centre, of both eyes was similar (0.47 Right vs. 0.52 Left), the MPOD at half-degree, one-degree and 2-degrees eccentricity from the foveal centre was markedly lower in the eye which had silicone oil (0.12, 0.06, 0.02 respectively vs. 0.40, 0.22, 0.07).

Although MPOD varies greatly between individuals, there is usually high inter-ocular symmetry in normal eyes.[5] Further work is required to determine whether or not this relates to the visual loss and whether therapeutic supplementation, may reduce the risk of visual loss.

References

1. Cazabon S, Groenewald C, Pearce IA, Wong D. Visual loss following removal of intraocular silicone oil. Br J Ophthalmol 2005;89:799-802.

2. Herbert EN, Habib M, Steel D, Williamson TH. Central scotoma associated with intraocular silicone oil tamponade develops before oil removal. Graefe’s Arch Clin Exp Ophthalmol. DOI 10.1007/s00417-005-0076-6.

3. Azzolini C, Docchio F, Brancato R Trabucchi G. Interactions between light and vitreous fluid substitutes. Arch Ophthalmol. 1992;110:1468-1471.

4. Refojo MF, Leong FL, Chung H et al. Extraction of retinol and cholesterol by intraocular silicone oils. Ophthalmology 1998;95:614-8.

5. Bone RA, Sparrock JM.. Comparison of macular pigment densities in human eyes. Vision Res. 1971;11:1057-1064.

Dear Editor,

We thank Singh and Raj for their interest and submit the following, in response to their comments, concerning our report of early chorioretinal anastomosis in non-ischaemic CRVO[1]:

Of the six patients who underwent laser anastomosis two had a successful anastomosis following the first treatment at a single site (33%), two cases were successful after a second attempt at one site (66%), a further case had two treatments and the second treament was at two sites, one further patient had two single treatments, followed by a third treatment at two sites. This left all patients with a functioning anastomosis (100%) as confirmed with fluorescein angiography.

We considered that the intervention was justified as this group have a significant risk of visual deterioration. We did not feel that a sham laser procedure was justified and such a group would further have diluted results and limited recruitment to the different groups. As OCT constitutes an objective measure, this may not be necessary. However in a larger study this may be a practical possibility.

Finally, concerning the reference to Fuller et al.[2], this study focussed on anterior segment neovascularisation and data analysis centred on non-perfused (ischaemic) CRVO and excluded perfused (non-ischaemic) CRVO. It is therefore difficult to comment on this point. Natural history studies are best conducted on a large scale. In our series of untreated patients we did not observe any disc collaterals during the treatment period. This point is, however, of limited significance.

We hope that this information addresses the points raised.

Richard Antcliff, Eric Mayer, Tom Williamson and John Shilling.

References

1. Antcliff RJ, M.E., Williamson TH, Shilling JS., Early chorioretinal anastomosis in non-ischaemic CRVO: a randomised trial. Br J Ophthalmol., 2005. 89(6): p. 780-1.

2. Fuller, J.J., et al., Retinochoroidal collateral veins protect against anterior segment neovascularization after central retinal vein occlusion. Arch Ophthalmol, 2003. 121(3): p. 332-6.