To compare, through a randomized, multicenter surgical trial, the postoperative tamponade effectiveness of intraocular silicone oil with that of an intraocular long-acting gas (initially sulfur hexafluoride [SF 6 ], later perfluoropropane [C 3 F 8 ]) for the management of retinal detachment complicated by proliferative vitreoretinopathy (PVR), using vitrectomy and associated techniques.

To evaluate the ocular complications that result from the use of silicone oil and gas.

The treatment of retinal detachment complicated by PVR remains controversial. Although some cases are managed successfully by pars plana vitrectomy and with temporary tamponade provided by intraocular gas, others eventually redetach with this technique. Preliminary reports indicate that prolonged tamponade with liquid silicone results in improved anatomical success, but the eventual visual outcome may be prejudiced by silicone-related complications, particularly glaucoma and keratopathy. The addition of hydraulic reattachment by simultaneous fluid/gas exchange to vitrectomy surgery has proved to be an important development. Although complications are few with these procedures, subsequent redetachment is frequent.

The Silicone Study was a randomized trial to investigate the relative merits of silicone oil or gas as tamponade modalities. All study patients underwent vitrectomy and were randomized intraoperatively either to silicone oil or to gas. Two groups of eyes were entered into the study: eyes that had not had a prior vitrectomy (Group 1) and those that had undergone previous vitrectomy outside the study (Group 2).

A critical element in the study was a standardized surgical procedure for PVR. This surgical procedure, intended to relieve retinal traction with vitrectomy techniques, was followed by assessment of the relief provided by an intraocular air tamponade. The eye was randomized to silicone oil or gas only after completion of the entire surgical procedure to eliminate investigator bias that might develop through knowledge of the treatment modality. Patients were examined 5 to 14 days following the randomization and again at 1, 3, 6, 12, 18, 24, and 36 months after that date. Repeated surgery was permitted for either treatment modality. The Fundus Photograph Reading Center staff processed and analyzed photographs taken at all the clinics, graded the preoperative severity of PVR on the basis of baseline visit photographs, and confirmed the macular status at followup visits.

End points of the study were visual acuity of 5/200 or greater and macular reattachment for 6 months following the final surgical procedure. The successful outcomes and complication rates of the two modalities were compared.

Eligibility criteria included but were not limited to PVR of Grade C-3 or greater according to the Retina Society Classification and visual acuity of light perception or better.