Predictability and limitations of non-invasive murine tonometry: comparison of two devices

Our purpose was to evaluate the accuracy, reproducibility and predictive ability of two non-invasive tonometers developed for intraocular pressure (IOP) measurements in the mouse. The prototype impact-rebound tonometer (I-R) and a prototype optical interferometry tonometer (OIT) utilizing a fiberoptic pressure sensor, were compared. Enucleated eyes from C57/BL6 mice were used for the calibration. The anterior chamber was cannulated and the IOP was adjusted in increments of 5 cm of H2O (open stopcock method). A calibration curve was generated for each individual eye along with a master calibration curve for all eyes. Two operators measured the IOP. The instruments were then used in alternating order to measure the IOP in C57/BL6 and in DBA2/J animals. The same eyes were subsequently cannulated and the error of the non-invasive tonometers was determined. Both tonometers yielded almost equivalent ex vivo calibration curves with individual R2 of 0.9878 and 0.9902 respectively. Both instruments were highly reproducible. In vivo the I-R tonometer underestimated while the OIT overestimated the IOP. This error was systematic and therefore predictable. The confidence intervals of this error were determined by comparing the IOP estimates provided by each tonometer with the measurements obtained invasively by cannulation in vivo. The 95% CI of the error were 2.36 mmHg for the I-R and 2.62 mmHg for the OIT respectively. Non-invasive tonometry in the mouse is feasible. Both non-invasive instruments provide accurate and reproducible measurements with the OIT requiring calibration curves for each individual investigator.