Abstract
Purpose: :
Accurate measurement of oxygen partial pressure (pO2) in the eye is important for understanding ocular physiology and disease. Two types of probes, fiberoptic and polarographic, have been used to measure pO2 in the human eyes. Recent studies using commercial versions of the two probe types reported different pO2 and oxygen gradients in the human vitreous chamber (Williamson, et al., Graefes Arch 2009; Holekamp, et al., AJO 2005, 2006). To obtain insight into the source of these differences, we compared measurements of pO2 made with these probes in vitro and in vivo.
Methods: :
Polarographic (Licox; Integra Life Science, Plainsboro, NJ) and fiber-optic probes (Oxylab pO2 Optode; Oxford Optronix, Oxford, UK) were placed into a liquid-filled chamber that was equilibrated with 100% N2, 1% O2 or 5% O2. Young albino rabbits were anesthetized, intubated, and maintained under normoxic condition and pO2 was measured in different regions of the vitreous chamber.
Results: :
The pO2 reported by the two probes in vitro were highly correlated (R2=0.997, P<0.001). A Bland-Altman plot detected a small difference between two systems; the Licox was more accurate at 5% O2, while the Oxylab was more accurate at low (0 and 1%) pO2. The response times of the two probes were differed greatly (Licox, >5 min, Oxylab, ~30 sec). It was difficult to obtain accurate O2 measurements with the Licox probe in vivo, since the probe had to be held in position in the eye for many minutes. The Licox probe also had a substantially larger O2-sensitive surface (Licox 18 mm2, Oxylab <1 mm2), providing poor spatial resolution in a small area like the vitreous chamber. The steep gradient of pO2 between the retina and the lens detected by the Oxylab probe was not detected using in the Licox probe. In the Licox probe, the oxygen and temperature sensors are widely separated; when the probe tip is inserted close to the retina, the temperature sensor is outside the eye, where it provides inaccurate temperature correction and is often contacted by the surgeon’s fingers.
Conclusions: :
Both methods accurately measured pO2 in vitro, however the long response time, larger size of the oxygen sensor and the separation of the oxygen and temperature sensors in the Licox probe make it unsuitable for measurement of oxygen distribution in small structures like the human eye. We suggest that previous measures of intraocular pO2 made using the Licox system are unlikely to be accurate.
Keywords: oxygen • electrophysiology: non-clinical • vitreous