This article was published in an Elsevier journal. The attached copy
is furnished to the author for non-commercial research and
education use, including for instruction at the author’s institution,
sharing with colleagues and providing to institution administration.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
Author's personal copy
Available online at www.sciencedirect.com
Respiratory Physiology & Neurobiology 161 (2008) 92–94
Short communication
The oxygen affinity of sickle hemoglobin
Amgad Abdu a , José Gómez-Márquez b , Thomas K. Aldrich a,∗
a
Pulmonary Medicine Division, Montefiore Medical Center and Albert Einstein College of Medicine, 111 East 210th Street, Bronx, NY, United States
b Department of Medicine, Montefiore Medical Center and Albert Einstein College of Medicine, Bronx, NY, United States
Accepted 17 December 2007
Abstract
The right-shifted oxyhemoglobin dissociation curve of sickle cell disease (SCD) has been thought to result in abnormally low arterial oxygen
saturation (SO2 ), even when oxygen partial pressure (PO2 ) is normal. However, without polymer formation (minimal under normoxic conditions),
HbS oxygen affinity is normal. We hypothesized that in SCD, in vivo SO2 is normal when PO2 is normal. We retrospectively examined 50 blood
gas and COoximetry samples from SCD patients and from controls matched for pH, PO2 , and carboxyhemoglobin. Control data fell close to the
Severinghaus curve, as did non-hypoxemic (SO2 > 92.5%) SCD data. In contrast, hypoxemic (SO2 < 92.5%) SCD data fell well below the standard
curve. Thus, although SCD patients’ oxygen affinity is low under hypoxic conditions, it is normal at normal arterial SO2 . Therefore, a finding of
abnormally low saturation demonstrates that PO2 is abnormally low. Given our previous finding that pulse oximetry faithfully reflects saturation in
SCD, low pulse oximeter readings in SCD constitute reliable evidence of impaired gas exchange.
© 2007 Elsevier B.V. All rights reserved.
Keywords: Blood; Hemoglobin; Blood gases; Pulse oximetry; Anemia
1. Introduction
Many sickle cell disease (SCD) patients have abnormally
high P50 (partial pressure of oxygen at 50% saturation)
(Lonsdorfer et al., 1983). As a result, several investigators have
suggested that pulse oximetry readings may be abnormally low
in SCD, even when gas exchange is normal (Blaisdell et al.,
2000; Comber and Lopez, 1996; Homi et al., 1997). If so, then
a low pulse oximeter reading, e.g. 92%, may not necessarily indicate abnormal arterial oxygenation. However, in vitro,
the oxygen affinity of HbS depends upon the severity of its
polymerization; in the absence of polymer formation, sickle
hemoglobin (HbS) has normal oxygen affinity (Fabry et al.,
2001). Since HbS polymerization is a function of the severity
of hypoxemia, oxygen affinity would be expected to be normal
when arterial oxygenation is normal. We hypothesized that even
among patients with right-shifted P50s, the oxygen affinity of
hemoglobin remains normal at normal arterial PO2 .
To test this hypothesis, we retrospectively reviewed data for
PO2 , pH, base excesses, and oxygen saturations (SO2 ) of SCD
patients and controls matched for pH, PO2 , and carboxyhemoglobin (COHb) to determine whether the hypoxemic samples
∗
Corresponding author. Tel.: +1 718 920 6087; fax: +1 718 652 8384.
E-mail address: [email protected] (T.K. Aldrich).
1569-9048/$ – see front matter © 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.resp.2007.12.005
from SCD patients lie farther from the standard oxyhemoglobin
dissociation curve (OHDC) than do the matched controls, and
whether the well-oxygenated samples from SCD patients follow
the standard OHDC.
2. Methods
Using Montefiore’s Clinical Looking Glass (a database
derived from the Hospital’s clinical information system), we
identified 50 patients with SCD (40 SS genotype, 6 SC genotype, and 4 S␤0 Thalassemia genotype) hospitalized between 1
January 2000 and 30 June 2007, and not transfused for at least 6
weeks, who had arterial blood gases and COoximetry performed
on the same arterial or venous blood samples. We accepted
samples with pH between 7.3 and 7.55, PCO2 between 20 and
65 mmHg, PO2 between 30 and 165 mmHg, and methemoglobin
(metHb) less than 5%.
We confirmed the genotypes by reviewing the patients’
records. We tabulated the date, time, age, gender, pH, partial pressure of carbon dioxide (PCO2 ), base excess (BE),
PO2 , SO2 , oxyhemoglobin concentration (O2Hb), carboxyhemoglobin concentration (COHb), and metHb. We also noted
previous prescription of hydroxyurea. Once the SCD group was
constructed, we assembled a control group of blood samples with
arterial or venous blood gases and simultaneous COoximetry
Author's personal copy
A. Abdu et al. / Respiratory Physiology & Neurobiology 161 (2008) 92–94
93
Table 1
Characteristics of the two populations
SCD group
Control group
Number of subjects
Age (median, range)
Gender (W:M)
Genotype
50
30.5 (13–98)
23:27
40 SS; 6 SC; 4 S␤0 Thal
50
52 (6–68)
29:21
Number of samples
pH
PCO2 (mmHg)
PO2 (mmHg)
SO2 (%)
COHb (%)
MetHb (%)
50
7.401 ± .042 (7.30–7.48)
39.4 ± 5.9 (23–52)
74.0 ± 25.0 (31–162)
87.5 ± 13.5 (34–100)
4.03 ± 1.61 (1.1–8.4)
1.74 ± 1.01 (0.1–4.1)
50
7.40 ± .052 (7.30–7.53)
42.4 ± 7.81 (28–65)
74.0 ± 24.5 (31–163)
91.4 ± 10.3 (59–100)
3.82 ± 1.62 (0.7–8.1)
0.90 ± .0.58 (0–2.6)
from non-SCD patients, matching for COHb within 1% point,
PO2 within 10 mmHg, and pH within 0.1 unit. Table 1 shows
the characteristics of the two groups. The study was reviewed
and approved by Montefiore’s Institutional Review Board for
the Protection of Human Subjects.
For each sample with SO2 ≤ 96.5%, we used the method of
Severinghaus (1979) to calculate the expected PO2 , based on
pH, BE, and SO2 and calculated the right shift of the OHDC
as the difference between the predicted and measured PO2 (PO2
cannot be accurately predicted for SO2 > 96.5% (Severinghaus,
1979). For each sample, regardless of SO2 , we used the method of
Severinghaus (1979) to calculate the expected SO2 , based on pH,
PO2 , and base excess. We computed the difference between the
predicted and measured saturations, a measure of the “downshift” of the OHDC for that sample. Right shift cannot be
calculated at normal or high PO2 , where the OHDC is nearly
flat, so we reasoned that the downshift would be a more meaningful estimate of any abnormality of oxygen affinity than right
shift across the range of SO2 we studied.
For samples with SO2 < 92.5% (hypoxemic) and for samples
with SO2 > 92.5% (normoxemic), we compared the right- and
downshifts among SCD and non-SCD groups, using Student’s
t-test, accepting P < 0.05 as evidence of statistical significance.
We also performed simple and multiple regression analyses
of SO2 , pH, PCO2 , COHb, and metHb (and genotype for the
SCD patients) against the downshift for both SCD and control
samples.
3. Results
Fig. 1 plots the control and SCD patients’ data in comparison to the Severinghaus curve. It is evident that the controls and
the non-hypoxemic SCD patients’ data fall close to the standard curve, but that the hypoxemic SCD patients’ data fall well
below. Neither the control group’s nor the non-hypoxemic SCD
patients’ right- and downshifts differed significantly from zero.
In contrast, the hypoxemic (SO2 < 92.5%) SCD patients’ data
fell well to the right and below the standard curve (right shift
11.36 ± 8.96 mmHg and downshift 5.74 ± 6.16% points, both
significantly greater than control or non-hypoxemic SCD samples. Among the SCD patients, the severity of the downshift was
significantly negatively correlated with SO2 (Fig. 2).
Fig. 1. Positions of control and SCD samples on the oxyhemoglobin dissociation
curve (OHDC). The solid line is the normal OHDC, from Severinghaus (1979),
assuming pH 7.4 and normal base excess (BE). The control samples (corrected
for pH and BE) fall near the curve, as do the normally oxygenated SCD samples
(>∼85 mmHg). The poorly oxygenated SCD samples (on average) fall well to
the right and below the standard curve.
Fig. 2. Downshift (corrected for pH and BE) as a function of SO2 in SCD patients.
There is a strong negative correlation of downshift with SO2 .
In both control and SCD groups, regression analysis showed
significant negative correlation of the non-pH- and non-BEcorrected downshift with SO2 , but the magnitude of the effect
was more than 2.5 times as high for the SCD group than for controls. Among the SCD patients, addition of genotype as a second
regressor improved the model; the SS genotype was associated
with the largest and SC genotype with the least downshift.
4. Discussion
We show that, although SCD patients do indeed have elevated P50s, indicating reduced hemoglobin oxygen affinity, at
least under hypoxic conditions; they do not have abnormally low
oxygen affinity at normal arterial partial pressures. By matching
our control group for pH and COHb and by limiting the study to
relatively low metHb, we demonstrated that none of these factors
could be responsible for the different behavior of SCD and control OHDCs under hypoxic conditions. Examination of the data
Author's personal copy
94
A. Abdu et al. / Respiratory Physiology & Neurobiology 161 (2008) 92–94
from two previous smaller uncontrolled studies, conducted for
other purposes, show similar shifts in the SCD patients’ OHDCs
at low but not at normal PO2 (Ortiz et al., 1999; Fitzgerald and
Johnson, 2001), but that fact was not commented upon. Thus,
in SCD, when saturation is found to be abnormally low, it can
be assumed that PO2 is also abnormally low. Given our previous finding that pulse oximetry faithfully reflects saturation in
SCD (Ortiz et al., 1999), we conclude that a low pulse oximeter
reading in SCD should not be discounted.
Our findings are compatible with the known behavior of
sickle hemoglobin (HbS) in vitro. HbS polymerizes under
hypoxic conditions, coming out of solution and producing the
abnormal RBC deformability and bizarre shapes that are characteristic of the disease (Fabry et al., 2001). The polymer has
much lower affinity for oxygen than does hemoglobin A or than
does HbS when it is in solution. In fact, the P50 of nonpolymerized HbS is similar to that of HbA, and only when polymer
forms is there a right shift (Fabry et al., 2001). The elevated 2,3
DPG (Bookchin and Lew, 1996) and the elevated HbF levels
(Serjeant et al., 1996) of SCD are too small to effect more than
a negligible shift of the OHDC to the right or left, respectively.
References
Blaisdell, C.J., Goodman, S., Clark, J.F., Casella, J.F., Loughlin, G.M., 2000.
Pulse oximetry is a poor predictor of hypoxemia in stable children with sickle
cell disease. Arch. Pediatr. Adolesc. Med. 154, 900–903.
Bookchin, R.M., Lew, V.L., 1996. Pathophysiology of sickle cell anemia. Hematol. Oncol. Clin. North Am. 10, 1241–1253.
Comber, J.T., Lopez, B.L., 1996. Evaluation of pulse oximetry in sickle cell anemia patients presenting to the emergency department in acute vasoocclusive
crisis. Am. J. Emerg. Med. 14, 16–18.
Fabry, M.E., Desrosiers, L., Suzuka, S.M., 2001. Direct intracellular measurement of deoxygenated hemoglobin S solubility. Blood 98, 883–884.
Fitzgerald, R.K., Johnson, A., 2001. Pulse oximetry in sickle cell anemia. Crit.
Care Med. 29, 1803–1806.
Homi, J., Levee, L., Higgs, D., Thomas, P., Serjeant, G., 1997. Pulse oximetry in a cohort study of sickle cell disease. Clin. Lab. Haematol. 19, 17–
22.
Lonsdorfer, J., Bogui, P., Otayeck, A., Bursaux, A., Poyart, C., Cabannes, R.,
1983. Cardiorespiratory adjustments in chronic sickle cell anemia. Bull. Eur.
Physiopathol. Respir. 19, 339–344.
Ortiz, F.O., Aldrich, T.K., Nagel, R.L., Benjamin, L.J., 1999. Accuracy of pulse
oximetry in sickle cell disease. Am. J. Respir. Crit. Care Med. 159, 447–
451.
Serjeant, G., Serjeant, B., Stephens, A., Roper, D., Higgs, D., Beckford,
M., Cook, J., Thomas, P., 1996. Determinants of haemoglobin level in
steady-state homozygous sickle cell disease. Br. J. Haematol. 92, 143–
149.
Severinghaus, J.W., 1979. Simple, accurate equations for human blood O2 dissociation computations. J. Appl. Physiol. 46, 599–602.