Multiple islands of retinoblastoma; incidence rate and time span of appearance.

IN the days before successful irradiation of bilateral retinoblastoma, the opportunities for studying the dates of appearance and the rate of growth of multiple islands of this neoplasm occurring in the remaining eye were rare, and indeed could only have occurred either when parental permission for excision of the second eye was refused or in the rare event of spontaneous recovery. The rarity of retinoblastoma in the white race (16 8 in 100,000 eye patients), the even greater rarity of spontaneous recovery, and the fact that the survival of those children for whom treatment was declined was unlikely to exceed one year or 18 months, doubtless explain the absence of any account in the literature of the incidence rate and the period during which multiple retinoblastoma islands may appear. Such information is of importance in the follow-up and prognosis of children affected by this tragic disease. Hilgartner (1903) discovered that retinoblastoma was radiosensitive, and Lawson (1916), Knapp (1920), and Chase (1920) tried radiotherapy, using rather crude surface applicators of Columbia paste in which radium needles were set. The results of this technique were ineffective in destroying the neoplasm and often inflicted serious damage to the eye. So it was not till 1929, when Foster Moore (with whom I was privileged to be associated in this pioneer work) succeeded in destroying a retinoblastoma by the interstitial insertion of a radon seed, that a study of the subsequent fate of the second affected eye was possible. Since 1929 improved radiotherapeutic technique, particularly with the use of radio-active discs made to fit exactly the curvature of a child's sclera, and the selection of patients in whom only one-third or less of the retina was involved in the neoplasm, have raised the successful results to 96 per cent. These favourable results have made it possible to study the course of 43 patients treated and followed-up during a period of 25 years. Fig. 1 (overleaf) shows the ages (and therefore to some extent the span of time) at which multiple islands, two to five in number, appeared in 26 patients after the initial detection and successful treatment of this neoplasm at one or more sites. Absolute accuracy about the date of onset of the first island was only possible when a hitherto unaffected eye was being kept under observation, for in the majority when the child was first brought to hospital on account of one eye being almost full of retinoblastoma the second eye