In this issue, Garcia-Arumi et al report their experience with endoresection of 38 choroidal melanomas (see page 1040).1 Readers of this article might be wondering who in their right mind would have the temerity to perform such an operation, fragmenting this dangerous tumour using a vitreous cutter and causing countless malignant cells to swirl around the vitreous cavity. One can imagine no better way of disseminating melanoma to other parts of the eye and the rest of the body—surgeons must surely be killing their patients.

Eminent opinion leaders have expressed similar concerns about radiotherapy and even enucleation. Manschot histologically showed viable tumour cells after radiotherapy and vociferously advocated early enucleation.2 Observing a peak in mortality in the second postoperative year, Zimmerman hypothesised that enucleation disseminated melanoma cells into the circulation, accelerating metastatic death.3 Zimmerman was undoubtedly influenced by what I hereby term the “dandelion allegory.” This fear of iatrogenic tumour dissemination encouraged measures such as no-touch enucleation, pre-enucleation radiotherapy and even non-treatment of uveal melanoma.

The “Zimmerman–Manschot debate” and concerns about iatrogenic tumour seeding prompted the Collaborative Ocular Melanoma Study (COMS) in North America to perform multicentre, randomised, prospective studies of (1) enucleation versus iodine plaque radiotherapy and (2) enucleation alone versus enucleation with pre-enucleation radiotherapy. The COMS claimed to have found no statistical difference in survival between brachytherapy and enucleation and no benefit from pre-enucleation radiotherapy.4 5 It would …