Abstract
Purpose :
It can be difficult to determine if available superior tarsal tissue will be sufficient for repair of a
large lower eyelid defect. This study proposes a method to classify the shape of upper tarsal plates and to
estimate the area of tissue available for use in a tarsoconjunctival graft.
Methods :
A cross sectional study of 67 patients aged 18-39 was performed. The study was IRB
approved and HIPAA compliant. Photos of upper tarsal plates were taken and digitally analyzed by
ImageJ with the ObjectJ plugin to measure central tarsal height, width, graft area, and the lateral and
medial heights of the potential graft area. Graft area was estimated by the rectangular formula: central
height – 4 mm (the minimum tissue needed to preserve the eyelid margin) x width and compared to the
measured area. Plates whose areas were overestimated by >25% were defined as triangular, 15-25% as
ellipsoid and <15% as trapezoidal (Figure 1). A correlation was performed in Graphpad Prism 7 between
the amount overestimated (estimated area / measured area) and an equation comparing lateral and
medial heights to central tarsal height (curvature equation).
Results :
The most common eyelid shape was ellipsoid (44.8%), followed by trapezoidal (42.5%) and
triangular (12.7%). The range of the curvature equation was 1.46 to 2.34 (Figure 2). The curvature ratio
was strongly inversely correlated to the graft area ratio (r = -.91).
Conclusions :
We describe a new equation for characterizing the shape of tarsal plates. The equation
compares changes in the lateral and medial heights of the tarsus with its central height and is strongly
inversely correlated with the amount of area overestimated. We propose that when estimating graft area
by height x width, an assumption of a 20% overestimation is reasonable for all but the most rounded
tarsal plates.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.