ANATOMY &PHYSIOLOGY Computed Tomography Dimensions of the Lacrimal Gland in Normal Caucasian Orbits Diana A. Tamboli*, Matheson A. Harris, M.D.†, Jeffery P. Hogg, M.D.‡, Tony Realini, M.D., M.P.H.†, and Jennifer A. Sivak-Callcott, M.D.† * West Virginia University School of Medicine; †West Virginia University Eye Institute; and ‡Department of Radiology, West Virginia University Hospital, Morgantown, West Virginia, U.S.A. Introduction: CT is frequently used to assess the lacrimal gland, yet no published reports exist of normal dimensions using this modality. We retrospectively evaluated CT data from normal orbits and measured the lacrimal gland dimensions. Methods: Two-hundred ninety-three consecutive orbital CT scans (586 orbits) of 282 patients were identified. Caucasian patients aged 18 years and older without known orbital disease were included. Traumatized orbits were excluded. All repeated scans were excluded. Length and width of the lacrimal gland were measured in axial and coronal sections with GE Centricity Software. The primary outcome was a descriptive analysis of the distribution of dimensions in normal orbits. A random subset of patient scans was measured independently by observers at different levels of training, and intraclass correlation coefficients (ICCs) were established. Results: Three hundred orbits (150 right orbits and 150 left orbits) of 282 patients were included. Normal distributions occurred with axial and coronal length measurements. Mean (10th and 90th percentiles) lacrimal gland axial length in right orbits was 14.7 mm (10.9 mm and 18.3 mm) and 14.5 mm in left orbits (10.3 mm and 18.3 mm). Coronal length averaged 17.7 mm in right eyes (13.9 mm and 21.8 mm) and 16.9 mm in left eyes (12.8 mm and 20.8 mm). Axial and coronal width showed non-normal distributions. No statistically significant difference in dimensions was found between genders, and a significant inverse linear relation occurred between gland size and age. Agreement was significant among the observ- ers (ICC 0.7). Conclusion: This study presents a normal range of Cauca- sian lacrimal gland dimensions seen on CT. Lacrimal gland size decreases with age, and no gender difference exists. This range can aid in differentiating diseased and nondiseased lacrimal glands, especially when correlated with physical examination. This may be useful in clinical trials that involve lacrimal gland size. (Ophthal Plast Reconstr Surg 2011;27:453–456) L acrimal gland disease often manifests as enlargement, which is seen on physical examination, during surgery, or at autopsy. Both ultrasound and postmortem lacrimal gland size have been reported, 1,2 but no established dimensions exist of nondiseased lacrimal glands on computed tomography (CT), the most used orbital imaging modality. Contemporary CT, especially with con- trast enhancement, accurately depicts the boundaries of the lacri- mal gland. With widely used software, measurement of the gland dimensions can be done directly on the digital image in the clinic. To make this measurement meaningful, we set out to establish a normal range of lacrimal gland dimensions. METHODS A retrospective review was done of 293 consecutive orbital CT scans performed at WV University Hospital between October 2008 and December 2009. Scans from Caucasian patients, age 18 years and older, that had axial and coronal soft-tissue images, were included. Patients with known orbital disease or trauma were excluded. In unilateral trauma, the affected orbit was excluded, but the nontrauma- tized orbit was included if no CT evidence of orbital soft tissue or bony injury was seen. CT scans were all collected on a Toshiba Aquilion 64-slice scanner (Toshiba America Medical Systems, Inc., Tustin, CA, U.S.A.) by using a 2-mm slice-thickness orbit protocol. Centricity Software (GE Healthcare, Little Chalfont, Buckinghamshire, U. K.) was used to display and measure dimensions. As the palpebral and orbital lobes are difficult to distinguish on CT, the lacrimal gland was treated as one structure. In each of the axial and coronal soft-tissue series, the image in which the lacrimal gland appeared the largest was chosen. Two measurements of the lacrimal gland were made on each chosen image. In the axial images, the long dimension was measured from the most-posterior tip of the lacrimal gland to the most-anterior tip. The wide dimension was measured from the lateral edge to the medial edge of the gland at its widest location, perpendicular to the first measure- ment (Figure 1). In the coronal images, the long dimension was measured from the superior tip of the lacrimal gland to the inferior tip. The wide dimension was measured from the lateral edge to the medial edge of the lacrimal gland at its widest point, perpendicular to the long measurement (Figure 2). Image-Agreement Analysis. Four clinicians—a neuroradiologist, an oculoplastic faculty member, an oculoplastic fellow, and a fourth-year medical student—were given 30 randomly selected CT scans from the retrospectively reviewed images. Each observer chose the single image in which the lacrimal gland appeared the largest, in each of the axial and coronal series. The proportion of orbits for which the images selected were within 2 images of one another was determined. Measurement-Agreement Analysis. The same 4 clinicians performed the lacrimal gland measurements on 30 randomly selected axial and Accepted for publication February 4, 2011. Presented at the 2010 ASOPRS Fall Meeting poster session. Funding was provided by a Research to Prevent Blindness grant. The authors have no financial interest related to this manuscript. Address correspondence and reprint requests to Dr. Jennifer A. Sivak- Callcott, West Virginia University Eye Institute, Morgantown, WV 26505, U.S.A. E-mail: sivakj@wvuhealthcare.com DOI: 10.1097/IOP.0b013e31821e9f5d Ophthal Plast Reconstr Surg, Vol. 27, No. 6, 2011 453