Abstract

Anesthesia of rare exotic felids in captivity carries the added risk that exhaustive validation of monitoring techniques is not possible. One such technique, pulse oximetry, as an index of arterial oxygen saturation (SpO₂), is part of minimum monitoring requirements in human anesthesia. It can provide a continuous, noninvasive indication of blood oxygen saturation when access to arterial blood or blood gas analysis is unavailable. Although, validation in vivo is preferred, in vitro confirmation of basic principles upon which the technique is based is a viable alternative to repeated and often invasive studies on valuable animals. This study was initiated to determine if the absorbance spectra of hemoglobins from exotic felids were sufficiently different from those of human hemoglobin to affect the applicability of pulse oximetry to zoo cats presented for anesthesia.

Pulse oximetry is based on the Beer-Lambert law, which states that the concentration of a pure substance can be determined from the spectral absorbance of that substance (e.g., oxyhemoglobin) at a given wavelength if the pathlength across which the incident light travels is known. In vitro, the measurement is made across a sample of lysed red blood cells in a cuvette of known dimensions. In order to apply the technology in vivo, it is assumed that the pulsatile component of the signal represents the arterial blood, and that all other light absorbing moieties are not important. This condition results in a nonhomogeneity of the absorbing system which necessitates empirical calibration of instrument algorithms.¹ Most monitors currently in use employ calibration curves obtained from human volunteers, which, in theory, can affect the application of pulse oximetry to nonhuman species. Since arterial oxygen saturation is determined by algorithms that analyze plethysmographic waveforms of light absorption measured at both the visible red (660 nm) and near infrared wavelengths (940 nm), mM extinction coefficients for a given species at these wavelengths should be comparable to human extinction coefficients if the derived SpO₂ values are to be valid. Such comparisons have already been made between human adult, human fetal, and canine hemoglobin.^2,3 Using the dog as a model for arterial oxyhemoglobin desaturation accuracy, the absorbance spectra of canine hemoglobin was demonstrated to be almost identical to that of human hemoglobin.³ This is not unexpected, since the absorbance range used by pulse oximetry affects only the heme moiety of hemoglobin, which is virtually conserved among vertebrates.³ We have determined the light absorbance characteristics of hemoglobins that are relevant to pulse oximetry for those animal species commonly used for biomedical research purposes or presented for anesthesia in veterinary medicine (e.g., dog, cat, horse, cow, pig), and compared them to human hemoglobin.⁴ Comparative absorbance data for exotic feline hemoglobin has not been reported. The need exists to further characterize and validate pulse oximetry in exotic species.

Heparinized blood samples were obtained from adult, nonsmoking, human volunteers and domestic cats (F. catus) as controls. Blood samples of exotic felids Sumatran tiger (P. tigris sumatrae), lion (P. leo), lynx (L. lynx), leopard (P. pardus), cheetah (A. jubatus), serval (F. serval), bobcat (F. rufus) were collected at the Columbus Zoo under anesthesia initiated for other medical reasons. Blood samples were immediately shipped on ice after collection. Hemoglobin crude lysates are prepared by lysis of washed erythrocytes in ice-cold distilled water and cleared by centrifugation. Purified hemoglobin was obtained by ion exchange chromatography (DEAE cellulose) followed by gel filtration column chromatography (Sephadex). Total hemoglobin, methemoglobin, and carboxyhemoglobin of the eluted hemoglobin fraction were then determined by in vitro oximetry (OSM-3 hemoximeter) prior to storage under liquid nitrogen. For spectrophotometric analysis in the 600–1000 nm range (region of interest to pulse oximetry), the sample was diluted to about 1 mM based on total hemoglobin concentration determined by multiwavelength spectrophotometry. The precise concentration of the diluted sample was then determined by a standardized cyanmethemoglobin assay. The supernatants were then analyzed after oxygenation was achieved by aeration with 100% oxygen for 10 min (HbO₂), and after deoxygenation with sodium dithionite (Hb). Oxyhemoglobin was also analyzed at 350–750 nm after further dilution (about 0.017 mM) to minimize the effects of light scattering by impurities in solution. Final concentrations of the dilute samples were calculated from the maximal absorbance of the Soret band (415 nm). Spectral analysis of the samples was performed using a scanning spectrophotometer (Cary-14, Varian Instruments, Sunnyvale, CA) in a 1-cm pathlength cuvette. The apparent mM extinction coefficients were calculated using an OLIS 3820 on-line data system (Jefferson, GA). The mM extinction coefficients at 415, 541, 576, 660 and 940 nm for the felids were compared with those of human controls and of accepted human values.

Minor differences in the absorbance of both oxyhemoglobin and deoxyhemoglobin were found initially between human hemoglobin and that of the felids (data not shown). No differences were seen in HI or HbCO content of the samples that would affect extinction coefficients. Examination of dilute samples in the strongly absorbing Soret region, where the effects of light scattering are minimized, revealed no major differences in absorbencies. Adjustments for the effects of light scattering by purification of the hemoglobin solutions and careful sample handling resulted in extinction coefficients that were consistent with both human controls and with literature values (Table 1).

Table 1. Apparent millimolar (mM) extinction coefficients for oxyhemoglobin (HbO₂) and deoxyhemoglobin (Hb) of felids compared to those of humans

Literature values are indicated in parentheses.
^a ε_λ = extinction coefficient, which is defined as the optical density of the absorbing substance at wavelength λ in a concentration of 1 mmol/L and a pathlength of 1 cm.

The results of our studies demonstrate that the hemoglobin absorption spectra for the cats are consistent with those of human use of pulse oximetry in these species. Small quantitative differences in the apparent optical density between human hemoglobin and that of the cat became insignificant when the effects of light scattering were taken into account. These data are in agreement with other studies supporting the conservation of spectral properties among the mammalian hemoglobins.³ It should be noted, however, that if centrifugally cleared, hemoglobin solutions can produce light scattering artifact, then non-hemolyzed blood from species with different erythrocyte morphology may affect the accuracy of SpO₂ determinations. Non-hemolyzed blood does not obey the Beer-Lambert law.⁵ The effects due to scattering cannot be reliably separated from those due to absorption; thus, the necessity for the empirical calibration curve. It is not unreasonable to assume that given differences in erythrocyte morphology, concentration and aggregation properties, nonhuman species may present challenges to human calibration algorithms. However, the clinical applicability of these systems does not appear to be compromised.

Acknowledgments

This study was supported by a cooperative research grant from the Columbus Zoo and The Ohio State University.

Literature Cited

1.  Erhardt, W., C. Lendl, R. Hipp, G. von Hegel, and H. Wiesner. 1990. The use of pulse oximetry in clinical veterinary anesthesia. J. Ass. Vet. Anaesth. 17:30–31.

2.  Harris, A.P., M.J. Sendak, R.T. Donham, M. Thomas, and D. Duncan. 1988. Absorption characteristics of human fetal hemoglobin at wavelengths used in pulse oximetry. J. Clin. Monit. 4:175–177.

3.  Sendak, M.J., A.P. Harris, and M.D. Donham. 1988. Accuracy of pulse oximetry during arterial oxyhemoglobin desaturation in dogs. Anesthesiology. 68:111–114.

4.  Grosenbaugh, D.A., J.O. Alben, and W.W. Muir. 1997. Absorbance spectra of inter-species hemoglobins in the visible and near infrared regions. J. Vet. Emerg. Crit. Care. (accepted).

5.  Anderson, N.M. and P. Sekelj. 1967. Light-absorbing and scattering properties of non-haemolysed blood. Phys. Med. Biol. 12:173–184.