Noninvasive measurement of tissue carbon dioxide tension using a fiberoptic conjunctival sensor: effects of respiratory and metabolic alkalosis and acidosis.

To evaluate potential clinical applications of a newly developed, noninvasive fiberoptic conjunctival carbon dioxide (PcjCO2) sensor designed to measure continuously tissue PCO2 in a vascular bed supplied by the internal carotid artery, we studied the effects of graded respiratory and metabolic alkalosis and acidosis on PcjCO2 in a hemodynamically stable canine model. Respiratory changes were induced by varying the frequency of ventilation and metabolic changes were induced by incremental infusions of sodium bicarbonate and hydrochloric acid. Continuous measurement of end-tidal carbon dioxide tension (PETCO2) was also performed. During respiratory alkalosis and acidosis, PcjCO2 values correlated well with PaCO2 (r = 0.96, n = 106); linear regression analysis of PcjCO2 vs. PaCO2 produced a slope of 1.01 and a y-intercept of 3.94 over a PaCO2 range of 12 to 76 torr. The mean PcjCO2-PaCO2 gradient was 4 +/- 3 (SD) torr. PETCO2 values also correlated well with PaCO2 (r = 0.91), as well as with PcjCO2 values (r = 0.91). Both PcjCO2 and PETCO2 showed a much weaker correlation with PaCO2 during metabolic alkalosis and acidosis, partly because the variation in PaCO2 was less. Moreover, the PcjCO2-PaCO2 gradient increased during the metabolic portion of the study up to a mean of 10 +/- 8 (SD) torr during metabolic acidosis, implying a build-up and/or lack of washout of CO2 from the conjunctival tissues, despite the normal physiologic range of PaCO2 values. We conclude that in a hemodynamically stable canine model, PcjCO2 and PETCO2 values correlate well with PaCO2 during pure respiratory alkalosis and acidosis; the correlation weakens significantly, however, with metabolic alterations in tissue CO2 levels.