Threshold analysis of mortality outcomes in the collaborative ocular melanoma study (COMS)

Area of Science:

• Ophthalmology
• Oncology
• Radiotherapy

Background:

• The Collaborative Ocular Melanoma Study (COMS) previously reported comparable survival rates between plaque brachytherapy and enucleation for medium-sized choroidal melanomas.
• Detecting subtle differences in survival outcomes for uveal melanoma treatments presents a statistical challenge.

Purpose of the Study:

• To quantify the mortality differences needed to achieve statistical significance in survival outcomes between plaque brachytherapy and enucleation for choroidal melanoma.
• To assess the robustness of these mortality thresholds under various statistical conditions.

Main Methods:

• Reanalysis of 12-year mortality data from the COMS trial.
• Application of threshold analysis to determine the number of deaths required to alter the 95% confidence interval (CI) for the risk ratio.

Main Results:

• At 12 years, the risk ratio for melanoma-related deaths was 1.08 (95% CI 0.82-1.42) with 105 deaths in the brachytherapy arm and 98 in the enucleation arm.
• Achieving statistical significance would necessitate 31 additional deaths in the brachytherapy group or 23 fewer in the enucleation group.
• Shifting the results to favor brachytherapy would require 34 additional deaths in the enucleation group or 27 fewer in the brachytherapy group.

Conclusions:

• The COMS trial did not demonstrate a clinically significant survival advantage for either plaque brachytherapy or enucleation in medium choroidal melanomas.
• Substantial shifts in mortality data are required to reach statistical significance, indicating the difficulty in detecting modest treatment effects in uveal melanoma research.