Transscleral cyclophotocoagulation versus cyclocryotherapy in treatment of neovascular glaucoma: a retrospective analysis

Sir, Neovascular glaucoma is often inadequately controlled on medical therapy and in many cases cyclodestructive procedures are used (Krupin et al. 1978; Robert et al. 1990; Oguri et al. 1998). We retrospectively compared the longterm effects on intraocular pressure (IOP) and possible complications of contact transscleral diode laser cyclophotocoagulation (TSCPC) versus cyclocryotherapy in 94 eyes of 94 patients with neovascular glaucoma. All the procedures had been carried out at our department between 1997 and 1999, and follow-up time was 12–36 months (average 31 months) after surgery. Diseases causing neovascular glaucoma included: diabetes mellitus (33); retinal vein occlusion (37); retinal artery occlusion (6), and others including trauma, uveitis and retinal detachment (18). Data were obtained from all subjects after 1 year, from 78 subjects after 2 years and from 64 subjects at the last follow-up visit after 3 years. Follow-up examinations included determination of best corrected visual acuity (VA), slit-lamp biomicroscopic and – if possible – ocular fundus examinations. Prior to the cyclodestructive procedures, panretinal photocoagulation for retinal ischaemia had been performed in 61 patients and vitreoretinal surgery had been carried out in 17 patients. During follow-up, retinal laser treatment was performed in 32 patients and vitreoretinal surgery was carried out in seven patients. Cyclocryotherapy was performed in 51 eyes of 51 patients (mean age 73 years) with neovascular glaucoma, and was used whenever IOP was considered to be too high for the eye despite maximum tolerated topical therapy, or when pain occurred in blind eyes. Standard surgical technique can be summarized as follows: 4–12 applications per eye; 180–360 degrees; at a distance of 2.5–4 mm from the surgical limbus; using a probe tip temperature of − 80 °, and a 60-second duration of application. In 43 eyes of 43 patients (mean age 69 years), diode laser contact cyclophotocoagulation was performed with a power of 1400–2000 mW for a duration of 2 seconds, with 20–35 applications over 270–360 degrees. A semiconductor diode laser system with a 600-µm diameter and quartz fibre optic was used 1.2 mm behind the surgical limbus. Laser applications were delivered to all four quadrants except in eight patients, where the temporal or nasal upper quadrant was saved. Power was reduced by 250 mW if there was any disruptive tissue reaction (the ‘pop’ effect) during two consecutive applications. Postoperative treatment was similar in both groups, and antiglaucomatous medication was continued according to the measured IOP. Retreatment was performed when ocular pain occurred or when IOP increased to a level considered potentially damaging to the optic disc. Hypotony was defined as IOP lower than 5 mmHg on two examinations. In the cyclocryotherapy group, the mean IOP in all patients decreased from 43.6 ± 13.4 mmHg (range 25–72 mmHg) preoperatively to 23.6 ± 11 mmHg (range 7–60 mmHg) 1 week postoperatively; 15.2 ± 9.2 mmHg (range 0–38 mmHg) after 1 month; 10.6 ± 9 mmHg (range 0–24 mmHg) after 6 months, and to 8.2 ± 11.5 mmHg (range 0–36 mmHg) after 1 year. At the 2-year follow-up visit, the mean IOP was 6.5 ± 6.6 mmHg (range 0–19 mmHg), and at the last control visit at 3 years, a mean IOP of 8.6 ± 7.4 mmHg (range 0–23 mmHg) was achieved. All mean IOP measurements obtained postoperatively were significantly lower (p < 0.05) than preoperative values (paired Student's t-test). Complications included a decrease in VA of two or more Snellen lines in 23 eyes, loss of light perception in 13 eyes, chronic hypotony in 17 eyes, and phthisis bulbi in seven eyes. Retreatment was performed in three eyes (once in two eyes and twice in one eye). In the TSCPC group, the mean IOP was reduced in all patients from 47 ± 10.3 mmHg (range 27–74 mmHg) preoperatively to 26.7 ± 10.2 mmHg (range 10–50 mmHg) 1 week postoperatively; 29.5 ± 12.5 mmHg (range 9–56 mmHg) after 1 month; 13.3 ± 12.7 mmHg (range 0–36 mmHg) after 6 months, and to 10.6 ± 11.9 mmHg (range 0–44 mmHg) after 12 months. At the 2-year follow-up visit, the mean IOP was 7.8 ± 8.8 mmHg (range 0–28 mmHg) and at the 3-year examination, a mean IOP of 7.5 ± 15 mmHg (range 0–30 mmHg) was found. At each time-point postoperative IOP was significantly lower (p < 0.05) than baseline readings. Retreatment was performed in three eyes. Visual acuity deteriorated by two or more Snellen lines in 17 eyes, loss of light perception occurred in eight eyes, chronic hypotony in nine eyes, and phthisis bulbi in six eyes. Thus, both cycloablative procedures lowered IOP considerably and significantly compared to baseline readings. A significant difference between both groups was noted only at the first postoperative month (with better IOP control provided by cryotherapy). Cyclocryotherapy-treated patients seemed to have a higher incidence of sight loss (not significant) and chronic hypotony (significant; p < 0.05) than TSCPC patients. Diode laser TSCPC and cyclocryotherapy are both known to be well tolerated, non-invasive procedures for lowering IOP in neovascular glaucoma. Filtering surgery is usually unsuccessful owing to conjunctival scarring due to previous surgery and neovascular tissue growth at the filtration site (Ophir & Porges 2000). Drainage devices are often used to lower IOP in these cases, but can only be inserted by experienced surgeons, and may be associated with severe complications (Assaad et al. 1999). Both cyclodestructive procedures and their potential benefits and complication rates have been described for different kinds of refractory glaucoma, but due to differences in research material, the results from different studies are not readily comparable. Our study showed that diode laser TSCPC was as effective as cyclocryotherapy in terms of IOP control and resulted in a lower rate of complications. We therefore recommend TSCPC as the better procedure in neovascular glaucoma.