Inferred natural history of uveal melanoma.

We have observed that the fatality rate of patients with posterior uveal melanomas rapidly increased from a very low rate prior to enucleation to a maximum of 8% per year during the second year after enucleation. These data suggest to us that events related to enucleation have a deleterious effect on survival, and we have postulated that these events are either the iatrogenic dissemination of tumor cells or an adverse effect on the immune-defense system, or both. Others have proposed that the relationship between enucleation and increased mortality is only happenstance. They would attribute the onset of symptoms to the rapid growth of the malignant tumor and thus claim that this brings the patient to enucleation soon after the tumor becomes malignant. We have studied 2105 cases on file in the REgistry of Ophthalmic Pathology and have found that enucleation was not always performed shortly after the onset of visual disturbance. In one third of the cases, enucleation was delayed until onset oment was not obtained until the tumor measured larger than 15 mm in diameter. Using Zelen's method to infer the natural history of uveal melanoma, we found that the mean time it took for a small tumor (less than 10 mm in diameter) to become a large tumor (greater than 15 mm) was approximately 7 years. The average age of patients treated for medium-sized tumors was 5 years less than that of patients treated for larger tumors. This is interpreted as evidence of an average delay of 5 years in the treatment of large melanomas. If this delay is taken into consideration, then the patients treated with medium-sized tumors had a worse survival during the first 7 years than did patients whose treatment was delayed until the tumor became large. After the seventh year, however, the survival was better. These findings support our hypothesis that the postoperative increase in mortality during the first several years, particularly among patients with tumors of medium size, was related to enucleation.