To the Editor.
—I read with great interest in the January Archives an article by Donoso et al1 concerning correlation of survival times with various uveal melanoma risk factors. However, I am concerned about the continuing trend of these and other authors2-6 to use single two-dimensional measurements when estimating the size or growth of a three-dimensional tumor, eg, basal chord diameter, greatest linear dimension, greatest height, or nucleolar area in a tissue section. I believe that feasible estimates of tumor volume would represent size and growth more accurately.Most uveal melanomas when fresh cut resemble (though not exactly) the solid geometry of the outer segment of a sphere more closely than that of any other three-dimensional figure. Indeed, throughout nature most cells and tumors are known to seek roundness (sphericity) in an effort to minimize surface area or loss of nutrients.7 The volume of a sphere segment
O' Rourke J. The Mathematics of Melanoma Volume. Arch Ophthalmol. 1986;104(7):973-974. doi:10.1001/archopht.1986.01050190031016