Assessment of tear film osmolarity using the TearLab^™ osmometer in normal dogs and dogs with keratoconjunctivitis sicca

Objective: To evaluate repeatability and reproducibility of tear osmolarity measured using the TearLab^™ osmometer in normal dogs and to assess its diagnostic potential in dogs with keratoconjunctivitis sicca (KCS). Animals studied: Beagle dogs; six normal and five with KCS. Procedures: Tear osmolarity and Schirmer tear test-1 (STT-1) values were obtained at various times. Normal dogs were assessed for diurnal variation and repeatability and reproducibility of measurements. Dogs with KCS were evaluated before and after 5 months’ topical twice-daily therapy with 2% cyclosporine. Results: Mean ± SD tear osmolarity (mOsm/L) was significantly higher in normal dogs (337.4 ± 16.2) than in dogs with KCS before therapy (306.2 ± 18.0; P < 0.0001), but not following therapy with 2% cyclosporine (330.5 ± 13.7; P = 1.00). Osmolarity readings lower than 325.5 mOsm/L were suggestive of KCS (84.8% sensitivity and 87.1% specificity). In normal dogs, tear osmolarity readings were stable during the daytime (P = 0.99). Repeated measurements revealed high variability and typically poor-to-moderate repeatability and reproducibility, although this was improved by taking three successive measurements at each session. Considering combined data from all dogs, a positive correlation existed between STT-1 and tear osmolarity measurements (Pearson's correlation test, P = 0.04, r = 0.62). Conclusions: Canine tear osmolarity as determined by TearLab^™ osmometer was variable, required multiple readings to be informative, and differed from values reported for humans. Dogs with KCS had a lower tear osmolarity than did normal dogs, and this increased following cyclosporine therapy.