TECHNICAL ASPECTS IN CONTROL AND EXPERIMENTAL GROUPS 1 Control Groups:

Once the technique of vitreolensectomy in the rabbit had been mastered, there were no particular technical difficulties thereafter up to this experimental stage. The rabbit eye tends to bleed readily, and this was kept to a minimum by raising the intraocular pressure, i.e. by raising the infusion bag. Peroperative fibrin formation was not a problem once low molecular weight heparin was used (all Control and

Experimental Group animals).

3.11.2 Experimental Group A:

Several technical difficulties were encountered whilst repairing retinal defects with octylcyanoacrylate and PVdF in an air filled eye. These relate both to introduction and application of the substrate, and to delivery of the adhesive.

Introducing the substrate through the sclerotomy invariably resulted in enlargement of the sclerotomy, with loss of an air-tight seal. This led to hypotony, loss of good visualisation of the posterior segment and an increased risk of vitreous haemorrhage. A preplaced 6/Dexon purse string suture sited round the sclerotomy helped to maintain an air-tight seal, but this was not always successful. Introducing the propylene substrate was much more problematic than the PVdF membrane, and so PVdF was used for all eyes.

Considerable difficulty was encountered in accurately placing the substrate over the retinal break, as it was hard to free the material from the jaws of the forceps, even using a blunt needle introduced through the third sclerotomy. Once on the retinal surface it was difficult to see the transparent material, but this problem was solved by colouring the substrate with ink from an indelible felt pen. Manoeuvring the substrate until it was accurately placed over the retinal defect was difficult, mainly because of surface tension effects. Additional retinal holes were sometimes created during this procedure.

The cyanoacrylate delivery system proved effective as the adhesive rarely polymerised within the tubing. Applying minute quantities was a problem as the drop of adhesive tended to flow back up the outside of the teflon tubing rather than remaining as a discrete drop at the end of the tube. If a thin film of preretinal fluid was present the octylcyanoacrylate tended to spread widely before polymerising;

great care was therefore taken in drying the preretinal space with the modified flute needle prior to delivering the adhesive.

The octylcyanoacrylate took 1-2 seconds to polymerise after application which was long enough, in most instances, to allow the tubing of the delivery system to be removed from the patch to prevent the patch and tubing from sticking together. Although evaluated in the in vitro experiments lophendylate was not used in any animals as this would have made the groups too small.

Air/fluid exchange was performed at the end of the operation. Care was taken to do this in as controlled a way as possible, but this was not always possible. While changing the air pump to the infusion line the eyes tended to collapse as soon as the intraocular pressure dropped because the rabbit eye has little scleral rigidity. When this happened the infusion fluid was run in quickly to restore intraocular pressure, to reduce the considerable risk of vitreous haemorrhage. The posterior segment was examined at the end of the procedure, and the findings noted. Air/fluid exchange was not performed in two animals. The reasons for this, and the consequences are reported in the next section.

3.11.3 Experimental Group B:

The same technical difficulties were encountered in introducing and applying the substrate as described above.

In this group of animals Tisseel was used as the adhesive and PVdF as the substrate. In only one animal was the 500 lU/ml strength thrombin used; the remainder were treated with 4 lU/ml. When using the stronger concentration of thrombin the needle of the Duploject blocked, as the fibrin clot formed very rapidly as the components mixed. Sequential delivery was tried, but the first component spread before the second could be delivered. Small, localised applications could not be made. All subsequent procedures were carried out using the Duploject system and 4 lU/ml strength thrombin. Even using the Duploject system it was difficult to apply minute amounts of Tisseel as the adhesive tended to spread widely before setting to form a fibrin clot. It was necessary to wait at least five minutes after application for the fibrin clot to form and adhere to the retina before undertaking air/fluid exchange. Meticulous drying of the preretinal space with the flute needle helped to make the clot stick to the retina. If the patch separated from the retinal surface during air/fluid exchange the fibrin was removed as extensively as possible with the vitrophage and the patch reapplied.

PVdF is hydrophobic, and it tended to float to the top of the drop of unclotted adhesive, floating away from the retinal surface and break.

3.11.4 Experimental Group C:

Patching using fibrin prepared from autologous plasma and bovine thrombin 4 lU/ml was only attempted in three eyes as many problems were encountered. Delivery was undertaken using the Duploject system and to aid visualisation of the adhesive fluorescein was added to the plasma component. In 1 eye the substrate would not adhere to the retina despite repeated attempts and waiting ten minutes for the clot to form. In two eyes the patch did remain adherent, but again this was after repeated attempts. In one of these eyes examination showed that the adhesive had spread very widely, forming a thin film covering virtually the whole of the inferior retina.