The idea that the macular neurosensory retina might be surgically detached and then reattached in such a manner that the fovea is relocated on the underlying retinal pigment epithelium (RPE) and choroid has been attributed to Lindsey et al.1 However, the 1983 ARVO abstract did not make clear that the main purpose of this exercise was to restore vision in eyes with subfoveal choroidal neovascular membranes (CNVMs) complicating age related macular degeneration (AMD), and the experimental and theoretical potential of the method was first realised clinically by Machemer and Steinhorst2 only in 1993. Their “macular rotation” procedure comprised induction of retinal detachment, vitrectomy and 360° peripheral retinotomy, removal of the CNVM (and any associated subretinal haemorrhage), and 30–80° rotation of the retina around the axis of the optic nerve before silicone oil injection and 360° endolaser retinopexy. Of the three eyes so treated, one showed a major visual improvement while proliferative vitreoretinopathy (PVR) compromised the outcome in the other two eyes. Modifications of the macular rotation method then followed3-6 but only Eckardt and colleagues from Frankfurt6 have succeeded in perfecting the technique, including minimising PVR and simultaneously counterrotating the globe to avoid a tilted image. A majority of the 30 eyes which they treated in 1997–8 showed an improvement in distance acuity, an improvement in near acuity, and maintenance or restoration of comfortable reading vision after silicone oil removal and cataract surgery with a mean follow up of 10.5 months. Notwithstanding the absence of a control series and independent visual assessment, 25–50° macular rotation for foveal relocation in AMD appears not only to have been validated, but also (given the difficulty and duration of the surgery) to have been vindicated by Eckardt's success. His results far exceed any remote expectation of spontaneous reversal of the generally progressive natural history of AMD (for example, improvement in distance acuity after 12 months in only 7% of the control group in another AMD study),7 and they provide a standard against which all future refinements of foveal relocation surgery must be judged. Eckardt's results also put into perspective the supposed merits of alternative therapies for subfoveal CNVMs in AMD for which randomised trials have demonstrated a statistically significant benefit. In the case of laser ablation of subfoveal CNVMs,8 for example, any prospect of future reading vision is instantly destroyed by the laser treatment, and the method has proved so intrinsically anti-Hippocratic (“first do no harm”) as to render immaterial the evidence base of overall long term benefit, at least in the UK and Ireland.9 In the case of photodynamic therapy (PDT),7 the comparative figure for improvement in distance acuity 12 months after 3 monthly therapy using verteporfin is 16.4%, the main benefit of PDT manifesting as a slowing of the rate of visual deterioration; no information is available on PDT concerning that most relevant of functional outcomes—reading capability. All the while we have also witnessed the false dawns created by other potential approaches to exudative AMD such as interferon, radiotherapy, and surgical excision of CNVMs.

The sustained benefits of macular rotation await confirmation through additional (and preferably properly controlled) series and extended follow up (especially to provide data on recurrence of CNVMs). However, work has been progressing in the meantime on exploring more practical approaches to foveal relocation. DeJuan and colleagues from the Wilmer Institute10 have described a method for obtaining limited foveal relocation (up to 15° or 1.5 mm) without recourse to peripheral retinotomy and thereby hopefully reducing the risk of PVR and obviating the need for extraocular muscle surgery. After inducing a retinal detachment in the vitrectomised eye posterior retinal redundancy is created by scleral shortening, and foveal relocation is induced by gas injection and postoperative posturing in a manner akin to that responsible for the macular folds which can complicate gas tamponade in externally buckled and vitrectomised eyes.11 Because forceps removal of a CNVM frequently involves extraction of an even wider area of RPE (D Wong, personal communication), the CNVM is left in situ and is destroyed by laser photocoagulation postoperatively provided the membrane is no longer subfoveal. Anecdotal reports of spectacular individual successes from such procedures have captured the imagination (not least of the tabloid press) and a relocation of 0.5 mm has also been achieved by the same method but without recourse to scleral shortening.12However, the overall outcomes13 from this technique have so far been disappointing, apparently reflecting (a) the difficulty in predicting the potential for visual recovery in individual cases (presumably depending on the degree of secondary photoreceptor apoptosis preoperatively and the health of the RPE), (b) the complications of postoperative photocoagulation (not least subfoveal recurrence) of the newly extrafoveal CNVM, and (c) the difficulty in predicting the precise location which the fovea will take up postoperatively. This last problem has now been addressed by Wong and Lois from Liverpool who report their results of direct intraoperative macular manipulation for foveal relocation in this issue of theBJO (p 352). A crucial element of their technique involves internal drainage of the subretinal infusate and patience in allowing the effect of the RPE pump to “kick in” to maintain the induced ectopia; two of the nine eyes in their series had useful reading vision 3 and 4 months after the foveal relocation. The significance to be attached to retention of vision and postsurgical reduction in vision in the other seven patients is difficult to judge in the absence of a control series but the authors' expectation of improved outcomes in the future with further refinement of the technique appears to be justified. Nevertheless, given the rapid pace of developments in foveal relocation surgery for exudative AMD, perhaps the time has now come for an international “tracker trial”14 of this surgery and to echo the recommendation of DeJuan and colleagues10 that continued investigation of such techniques only occurs “in the context of a clinical research effort approved by an independent institutional review board”.