Chorioretinal anastomosis using argon laser has been successful improving vision in one-third of eyes in the non-ischemic variant of the disorder in selected patients but has been associated with frequent complications (Browning and Antoszyk 1998; McAllister et al. 1998; McAllister and Constable 1995). This therapy is not generally used in ischemic CRVO because of a high complication rate in particular neovascularisation.

Grid macular laser therapy in BRVO can control macular oedema, with 60 % of treated cases retaining 20/40 vision or better at 3 years (Branch Vein Occlusion Study Group 1984) and reduce the incidence of vitreous haemorrhage by half (Branch Vein Occlusion Study Group 1984, 1986). In BRVO, the site of occlusion can easily be visualised and is thought to occur as a result of arteriosclerosis in the media of the artery giving rise to a compressive effect on the adjacent vein. At the arteriovenous crossing, the artery is generally located anterior to the occluded vein, within a common adventitial sheath. Charles attempted to relieve the blockage by dissecting the arteriole off the venule at the arteriovenous crossing in one patient (Osterloh and Charles 1988). Recently, the technique has been revisited with reported success in 10 out of 15 patients, with an average gain of four lines of vision (Opremcak and Bruce 1999). In a non-randomised comparative study, patients receiving intervention performed better with 75 % doubling their visual angles compared to 40 % with conventional treatment (Mason et al. 2004). Opremcak has designed an instrument for the blunt dissection of the arteriole from the venule by inserting a spatulated knife between the blood vessels. Potential complications include retinal tear or detachment, vitreous haemorrhage, retinal gliosis at the incision site, arcuate scotoma and cataract. Resolution of macular oedema following sheathotomy has been confirmed using optical coherence tomography (OCT) (Fujii et al. 2003). There are also theoretical reasons whereby the localised ischemia of the retina in BRVO might be improved by the PPV procedure because removal of the vitreous gel may allow oxygenation of the retina from other sites in the eye with good blood flow (Cringle et al. 1992). Indeed, some investigators have tried PPV and gas alone (Saika et al. 2001).

Usually, CRVO is associated with severe irreversible visual loss and with improvement of vision in only 20 % (The Central Vein Occlusion Study Group 1997). Incision of the optic nerve on the nasal side (radial optic neurotomy) has been described (Opremcak et al. 2001; Williamson et al. 2003; Garcia-Arumi et al. 2003) but is of unproven benefit as yet despite large case series (Opremcak et al. 2001, 2006). It is thought that the neurotomy may help blood flow in the central retinal vein by relieving pressure on the vein as it exits the lamina cribrosa, assuming that CRVO is a ‘neurovascular compression syndrome’, resulting from increased pressure within the confined space of the scleral outlet. Intravitreal pO₂ is severely reduced in CRVO and may be increased by vitrectomy (Williamson et al. 2009a).

Cystoid macular oedema (CMO) is a major cause of visual loss in RVO. Injections of varying doses of triamcinolone into the vitreous cavity (1–21 mg) have been investigated for reversing CMO in uveitis, postcataract surgery, retinal vein occlusion, telangiectasia and diabetes (Jonas et al. 2002, 2003, 2004, 2005a, b; Spandau et al. 2005; Alldredge and Garretson 2003; Antcliff et al. 2001; Benhamou et al. 2003; Bynoe and Weiss 2003; Conway et al. 2003; Martidis et al. 2001; Greenberg et al. 2002; Martidis et al. 2002; Degenring and Jonas 2003; Degenring et al. 2003; Williamson and O’Donnell 2005). Dexamethasone pellets can be inserted into the vitreous and have shown some efficacy for improved vision over 6 and 12 months (Haller et al. 2010), but as with triamcinolone, repeated injections are required.

The steroid stabilises the leaky vascular endothelium reducing the extracellular fluid accumulation perhaps by downregulating vascular endothelial growth factor (VEGF) (Jonas and Sofker 2001). In many cases, the feasibility of steroid is lessened because the duration of action is only a few months, requiring the administration of repeated injections in chronic conditions. Unfortunately, those repeat injections are associated with a reduced or absent response with time (Williamson and O’Donnell 2005). Dexamethasone has, however, been associated with less frequent IOP elevation than triamcinolone injection which in the later can be severe and require surgery for IOP control.

Anti-VEGF agents have been used widely for cystoid macular oedema from many causes including RVO. Bevacizumab reduces CMO and improves vision with low complication rates but requires repeat injection. Anti-VEGF agents are useful also for controlling neovascular glaucoma.

TPA may not cross the vasculature to enter the central retinal vein; therefore, investigators have attempted to insert TPA directly into the vein (Weiss 1998) using a 33-G cannula. The technique does not seem to have had a major effect on visual acuity, and vitreous haemorrhage is common (Weiss and Bynoe 2001). More recently, Weiss has added intravitreal steroid injections to try to improve success rates (Bynoe and Weiss 2003).