- Email: [email protected]
Confirmation of spurious hypoxemia using continuous blood gas analysis in a patient with chronic myelogenous leukemia
Pergamon 01452126(95)00117-4
CONCISE
REPORT
CONFIRMATION OF SPURIOUS HYPOXEMIA USING CONTINUOUS BLOOD GAS ANALYSIS IN A PATIENT WITH CHRONIC MYELOGENOUS LEUKEMIA Barry A. Mizock*t,
Cory Franklin*t,
Phillip LindesmithS and Prabodh C. Shahs
*Division of Critical Care Medicine, Department of Medicine, Cook County Hospital, Chicago, Illinois, U.S.A.; TThe Chicago Medical School, Chicago, Illinois, U.S.A.; SOptex Biomedical Inc., Woodlands, Texas, U.S.A.; and 5Division of Hematology, Department of Medicine, Cook County Hospital, Chicago, Illinois, U.S.A. (Received 5 July 1995. Accepted 3 August 1995) Abstract-Patients with extreme leukocytosis or thrombocytosis who have hypoxemia on arterial blood gas analysis may demonstrate normal oxygen saturation using pulse oximetry. The most commonly invoked explanation for this phenomenon is oxygen consumption in the blood gas sample prior to analysis. However, others have challenged the premise that the hypoxemia is spurious. We describe a patient with extreme leukocytosis and hypoxemia in whom normoxia was confirmed by continuous blood gas analysis. Key words: Leukemia,
leukocytosis,
hypoxemia,
Introduction In 1979, Hess et al. reported on a series of 12 patients with leukemia or thrombocytosis who had low arterial oxygen tension (PaO*) in the absence of dyspnea or peripheral cyanosis [l]. They noted that the Pa02 in blood gas samples stored at room temperature fell progressively over time and that this decline was prevented by prompt icing. They also observed that the rate of decline in PaOz correlated with the leukocyte count as well as with the type and maturity of the proliferating cell. Others have confirmed the apparent artifactual nature of the hypoxemia and applied the terms ‘leukocyte larceny’ [2], or ‘oxygen steal’ [3] to this phenomenon. The postulated mechanism is thought to involve ongoing consumption of oxygen in the blood Abbreviations: Cl, cardiac index; CO, cardiac output; CVP, central venous pressure; PA, pulmonary artery pressure; PuC02, arterial carbon dioxide tension; PaOz, arterial oxygen tension; PAWP, pulmonary artery wedge pressure; pHa, arterial pH; PVR, pulmonary vascular resistance; .Sp02, oxygen saturation by pulse oximetry; SVR, systemic vascular resistance. Correspondence to: Barry A. Mizock M.D., Department of Medicine, Cook County Hospital, 1835 West Harrison Street, Chicago, IL 60612, U.S.A. (Tel: 312 572 3828; Fax: 312 572 3810).
arterial
blood gases, oximetry.
gas sample by leukocytes or platelets. Subsequent observations of normal oxygen saturation with pulse oximetry, as well as mitigation of hypoxemia when potassium cyanide was added to the specimen, lend credence to this mechanism [24]. Gartrell and Rosenstrauch presented data obtained from two patients with chronic myeloid leukemia which suggested that the observed hypoxemia was not spurious [5]. They noted that when serial determinations of Pa02 were performed on specimens stored at room temperature (but not on ice), a linear decline in PaOa was noted; regression analysis indicated low PaOz at time zero. The validity of oxygen saturation obtained with pulse oximetry (SpOz) in patients with leukemia based on high levels of methemoglobin commonly found in these patients has also been questioned (SpOZ exhibits an inverse relationship with the methemoglobin fraction)
[61. The difficulty
in accurately assessing oxygenation with extreme leukocytosis or thrombocytosis creates a number of diagnostic and therapeutic dilemmas for the clinician. However, recent innovations in technology may have resolved this problem. Continuous intra-arterial blood gas monitoring devices have been developed which combine fiberoptic technology with spectrophotometric chemistry. The device typically status in patients
1002
B. A. Mizock et al.
consists of a fiberoptic bundle which terminates in a sensor unit containing indicator chemicals; it is placed intra-arterially through the lumen of a conventional 20 gauge arterial catheter. A light signal is transmitted from the monitor along the fiberoptic bundle; the light absorbance of the reagents in the sensor unit is related to levels of pHa, PaC02 and PaO*. These optical changes are in turn transmitted back to a microprocessor unit that displays the information in real time [7]. We report the first case of extreme leukocytosis associated with low PaOz on intermittent arterial blood gas analysis in which normoxia was confirmed by a continuous blood gas monitor. Report The patient was a 26-year-old Korean female with a history of chronic myeloid leukemia for 8 years who was admitted with complaints of weakness, nausea and vomiting, and abdominal pain for 3 weeks. In 1991, she was in an accelerated phase of leukemia and was treated with one course of cytarabine and daunorubicin. She responded well and was converted back to a chronic phase. She had recently been treated with hydroxyurea as an outpatient but had been unable to take this medication for several weeks prior to admission because of nausea and vomiting. On examination, she was noted
to be alert and oriented x3. Vitals: BP 128178 mmHg, RR 24/min, HR 110 bpm, T 37.9”C. She exhibited marked pallor with puffy facies. No cyanosis was noted. She had good air entry bilaterally, S,, S2 were normal, no extra sounds, murmurs or jugular venous distension were present. The liver was enlarged with a span of 14 cm, firm in consistency. Palpation of the abdomen revealed massive splenomegaly down to the left iliac fossa; mild diffuse abdominal tenderness without rebound were also elicited. Bowel sounds were normoactive. The stool was guaiac negative. On admission, hemoglobin was 4.8 g/d1 (48 g/l), WBC 1.2 million/mm” (5% metas, 7% myelos, 10% promyelocytes, 37% blasts), leukocrit 40%, platelet count 60,00O/mm”, prothrombin time 17.5 s, partial thromboplastin time 43.2 s, Na 138 mmol/l, K 5.5 mmol/l (pseudohyperkalemia not ruled out), Cl 102 mmol/l, CO2 19.7 mmol/l, BUN 2.9 mmol/l, Cr 61.9 pmolil, glucose 3.1 mmol/l, uric acid 8.5 mg/dl (505.6 pmolil), Ca 2.4 mmol/l, total bilirubin 1.1 mg/dl (18.8 pmolil), direct bilirubin 0.1 mg/dl (1.7 pmol/l), LDH 2167 U/l, CPK 20 U/l, SGOT 59 U/l, SGPT 27 U/l, alk phos 256 U/l, GGT 73 U/l, urinalysis: lo-15 WBC, 2-4 RBC. Chest X-ray demonstrated borderline cardiomegaly and bilateral perihilar and basal interstitial densities; the EKG showed sinus tachycardia with T wave inversion Vi-V+ Arterial blood gases on room air were: pH 7.40, PaOz 28 mmHg
BIOSENTRY I UNIT 05 COOK COUNTY
12:oo 12/15/94
0o:oo
12:oo
TIME
Fig. 1. Tracing of continuous PaOz display. Crosses (+) represent values for intermittently obtained PaO*.
0o:oo
1003
Spurioushypoxemia Table 1. Comparison of Pa02 and Sa02 from arterial blood gases(ABG), continuous blood gas analysis (CBG) and pulse oximetry (SpOz) Date 12115 12116
Time
ABG
CBG
(PaO#a02) (Pa02/Sa02) 18:20 51.7 mmHg/87% 96 mmHgi96.9% 20:30 59.3 mmHg/90.5% 121 mmHg/98.2% 01:20 52.9 mmHg/87.9% 115 mmHg/97.6% 07:45 53.6 mmHgi88.3% 78 mmHg/93%
(3.7 kPa), PaCO* 35 mmHg (4.7 kPa), HC03 23 mmol/l, SaOz 48.8%. Pulse oximetry at that time revealed a saturation of 94%. The patient was subsequently transferred to the medical ICU where she received leukophoresis, hydroxyurea and packed red cell transfusions. Her abdominal pain worsened acutely. Broadspectrum antibiotics were administered and general surgery was consulted. A computerized tomographic scan of the abdomen revealed a large solid right adrenal mass with perinephric fluid collection. She subsequently developed respiratory distress which necessitated intubation and mechanical ventilation. Pulmonary artery catheterization showed: CO 5.8 l/min, CI 4.2 l/minlm2, SVR 397 dyne/s/em-5, PVR 205 dynelsicm-5, PA 42132 mmHg (mean 34 mmHg), PAWP 20 mmHg, CVP 16 mmHg. The patient developed wide complex tachycardia accompanied by hypotension. The tachycardia was felt to be supraventricular but did not respond to adenosine or cardioversion. The patient subsequently suffered cardiac arrest and could not be resuscitated. Autopsy was refused by the family. Prior to the acute deterioration which lead to her death, a BiosentryTM continuous intra-arterial blood gas sensor (Optex Biomedical, Inc., Woodlands, Texas, U.S.A.) was inserted into the right radial artery through the lumen of an indwelling arterial catheter. An in vitro calibration was performed prior to insertion using the method specified by the manufacturer. A number of arterial blood gases were obtained intermittently in the standard fashion through the arterial line. Each sample was immediately iced and analyzed within 3-4 min. The PaOz obtained was compared to the value for Pa02 displayed on the continuous monitor at the time the specimen was drawn (Fig. 1). The SpO2 shown on the pulse oximeter at the time of sampling was also compared to the SaOz derived from intermittent and continuous blood gas analysis. These data demonstrated that Pa02 obtained by the intermittent method was consistently lower than that from continuous blood gas analysis. In addition, Sp02 exhibited a reasonable degree of correlation with continuously derived Sa02, whereas Sp02 did not correlate with SaOz derived from intermittent blood gases. These data are shown in Table 1.
SPO2
93% 93% 95% 96%
WBC 871,200 688,800
Discussion Our data lend support to the premise that intermittent methods of arterial blood gas analysis may lead to spuriously low values for Pa02 in patients with extreme leukocytosis due to ongoing oxygen consumption in the period which precedes measurement. The question of whether cooling the specimen is effective in preventing this decline remains unresolved; prompt icing did not mitigate the fall in Pa02 in our patient. This corroborates similar observations in earlier studies [3, 8, 91. Several investigators have suggested using regression analysis of serial determinations of Pa02 in samples stored at room temperature as a means to assess oxygenation status [5, lo]. However, it is not clear whether the decay in Pa02 over time is linear as suggested, or exponential [lo]. Our data also support the use of pulse oximetry as a means to monitor oxygenation in patients with extreme leukocytosis, keeping in mind that SpO2 may be unreliable in the presence of elevated levels of methemoglobin [5]. Theoretically, continuous blood gas analysis presents an ideal method of assessing oxygenation status in hypoxemic patients with extreme leukocytosis or thrombocytosis, since there is no delay in measurement related to sampling. However, in vivo calibration of the sensor is unreliable in this setting so that questions remain regarding its accuracy. Although small studies performed in a variety of clinical settings have shown continuous blood gas analysis to be useful, its reliability and cost-effectiveness need to be confirmed in large controlled prospective trials [7]. References 1. Hess C. E., Nichols A. B., Hunt W. B. & Suratt P. M. (1979) Pseudohypoxemia secondary to leukemia and thrombocytosis. New Engl. J. Med. 301, 361. 2. Fox M. J., Browdy J. S. & Weintraub L. R. (1979) Leukocyte larceny: a causeof spurious hypoxemia. Am. J. Med. 67, 742.
3. Loke J. & Duffy T. P. (1984) Normal arterial oxygen saturation with the ear oximeter in patients with leukemia and leukocytosis. Cancer 53, 1767. 4. Rello J., Benito S., Triginer C. & Net A. (1989) False hypoxemia induced by leukocytosis [Letter]. Crit. Care Med. 17, 970.
1004
B. A. Mizock et al
5. Gartrell K. & Rosenstrauch W. (1993) Hypoxaemia in patients with hyperleukocytosis: true or spurious, and clinical implications. Leukemia Res. 17, 915. 6. Vaidya M. S., Das Gupta A., Pavri R. S., Baxi A. J. & Advani S. H. (1981) Acquired methemoglobinemia in leukemia: its etiopathogenesis and possible effects on red cell structure and function. Leukemia Res. 5, 265. 7. Lumsden T. & Marshall W. R. (1995) New developments in blood gas analysis: continuous intra-arterial determination of POz, pH, and PCOz. In Critical Care Monitoring: From Pre-hospital to the ICU (Levine R. L. & Fromm R. E. Jr, Eds), p. 223. Mosby, St Louis, IL, U.S.A.
8. Chillar R. K., Belman M. J. & Farbstein M. (1980) Explanation for apparent hypoxemia associated with extreme leukocytosis: leukocytic oxygen consumption. Blood 55, 922. 9. Shohat M., Schonfeld T., Zaizoz R., Cohen I. J. & Nitzan M. (1988) Determination of blood gases in children with extreme leukocytosis. Crit. Care Med. 16, 787. 10. Vincent F., Levy V., Bensousan T., Escudier B. & Leclercq B. (1994) Blood gas analysis in patient with hyperleucocytosis: validity of repeated measures on same sampling without ice? [Letter.] Leukemia Res. 18, 553.
CONCISE
REPORT
CONFIRMATION OF SPURIOUS HYPOXEMIA USING CONTINUOUS BLOOD GAS ANALYSIS IN A PATIENT WITH CHRONIC MYELOGENOUS LEUKEMIA Barry A. Mizock*t,
Cory Franklin*t,
Phillip LindesmithS and Prabodh C. Shahs
*Division of Critical Care Medicine, Department of Medicine, Cook County Hospital, Chicago, Illinois, U.S.A.; TThe Chicago Medical School, Chicago, Illinois, U.S.A.; SOptex Biomedical Inc., Woodlands, Texas, U.S.A.; and 5Division of Hematology, Department of Medicine, Cook County Hospital, Chicago, Illinois, U.S.A. (Received 5 July 1995. Accepted 3 August 1995) Abstract-Patients with extreme leukocytosis or thrombocytosis who have hypoxemia on arterial blood gas analysis may demonstrate normal oxygen saturation using pulse oximetry. The most commonly invoked explanation for this phenomenon is oxygen consumption in the blood gas sample prior to analysis. However, others have challenged the premise that the hypoxemia is spurious. We describe a patient with extreme leukocytosis and hypoxemia in whom normoxia was confirmed by continuous blood gas analysis. Key words: Leukemia,
leukocytosis,
hypoxemia,
Introduction In 1979, Hess et al. reported on a series of 12 patients with leukemia or thrombocytosis who had low arterial oxygen tension (PaO*) in the absence of dyspnea or peripheral cyanosis [l]. They noted that the Pa02 in blood gas samples stored at room temperature fell progressively over time and that this decline was prevented by prompt icing. They also observed that the rate of decline in PaOz correlated with the leukocyte count as well as with the type and maturity of the proliferating cell. Others have confirmed the apparent artifactual nature of the hypoxemia and applied the terms ‘leukocyte larceny’ [2], or ‘oxygen steal’ [3] to this phenomenon. The postulated mechanism is thought to involve ongoing consumption of oxygen in the blood Abbreviations: Cl, cardiac index; CO, cardiac output; CVP, central venous pressure; PA, pulmonary artery pressure; PuC02, arterial carbon dioxide tension; PaOz, arterial oxygen tension; PAWP, pulmonary artery wedge pressure; pHa, arterial pH; PVR, pulmonary vascular resistance; .Sp02, oxygen saturation by pulse oximetry; SVR, systemic vascular resistance. Correspondence to: Barry A. Mizock M.D., Department of Medicine, Cook County Hospital, 1835 West Harrison Street, Chicago, IL 60612, U.S.A. (Tel: 312 572 3828; Fax: 312 572 3810).
arterial
blood gases, oximetry.
gas sample by leukocytes or platelets. Subsequent observations of normal oxygen saturation with pulse oximetry, as well as mitigation of hypoxemia when potassium cyanide was added to the specimen, lend credence to this mechanism [24]. Gartrell and Rosenstrauch presented data obtained from two patients with chronic myeloid leukemia which suggested that the observed hypoxemia was not spurious [5]. They noted that when serial determinations of Pa02 were performed on specimens stored at room temperature (but not on ice), a linear decline in PaOa was noted; regression analysis indicated low PaOz at time zero. The validity of oxygen saturation obtained with pulse oximetry (SpOz) in patients with leukemia based on high levels of methemoglobin commonly found in these patients has also been questioned (SpOZ exhibits an inverse relationship with the methemoglobin fraction)
[61. The difficulty
in accurately assessing oxygenation with extreme leukocytosis or thrombocytosis creates a number of diagnostic and therapeutic dilemmas for the clinician. However, recent innovations in technology may have resolved this problem. Continuous intra-arterial blood gas monitoring devices have been developed which combine fiberoptic technology with spectrophotometric chemistry. The device typically status in patients
1002
B. A. Mizock et al.
consists of a fiberoptic bundle which terminates in a sensor unit containing indicator chemicals; it is placed intra-arterially through the lumen of a conventional 20 gauge arterial catheter. A light signal is transmitted from the monitor along the fiberoptic bundle; the light absorbance of the reagents in the sensor unit is related to levels of pHa, PaC02 and PaO*. These optical changes are in turn transmitted back to a microprocessor unit that displays the information in real time [7]. We report the first case of extreme leukocytosis associated with low PaOz on intermittent arterial blood gas analysis in which normoxia was confirmed by a continuous blood gas monitor. Report The patient was a 26-year-old Korean female with a history of chronic myeloid leukemia for 8 years who was admitted with complaints of weakness, nausea and vomiting, and abdominal pain for 3 weeks. In 1991, she was in an accelerated phase of leukemia and was treated with one course of cytarabine and daunorubicin. She responded well and was converted back to a chronic phase. She had recently been treated with hydroxyurea as an outpatient but had been unable to take this medication for several weeks prior to admission because of nausea and vomiting. On examination, she was noted
to be alert and oriented x3. Vitals: BP 128178 mmHg, RR 24/min, HR 110 bpm, T 37.9”C. She exhibited marked pallor with puffy facies. No cyanosis was noted. She had good air entry bilaterally, S,, S2 were normal, no extra sounds, murmurs or jugular venous distension were present. The liver was enlarged with a span of 14 cm, firm in consistency. Palpation of the abdomen revealed massive splenomegaly down to the left iliac fossa; mild diffuse abdominal tenderness without rebound were also elicited. Bowel sounds were normoactive. The stool was guaiac negative. On admission, hemoglobin was 4.8 g/d1 (48 g/l), WBC 1.2 million/mm” (5% metas, 7% myelos, 10% promyelocytes, 37% blasts), leukocrit 40%, platelet count 60,00O/mm”, prothrombin time 17.5 s, partial thromboplastin time 43.2 s, Na 138 mmol/l, K 5.5 mmol/l (pseudohyperkalemia not ruled out), Cl 102 mmol/l, CO2 19.7 mmol/l, BUN 2.9 mmol/l, Cr 61.9 pmolil, glucose 3.1 mmol/l, uric acid 8.5 mg/dl (505.6 pmolil), Ca 2.4 mmol/l, total bilirubin 1.1 mg/dl (18.8 pmolil), direct bilirubin 0.1 mg/dl (1.7 pmol/l), LDH 2167 U/l, CPK 20 U/l, SGOT 59 U/l, SGPT 27 U/l, alk phos 256 U/l, GGT 73 U/l, urinalysis: lo-15 WBC, 2-4 RBC. Chest X-ray demonstrated borderline cardiomegaly and bilateral perihilar and basal interstitial densities; the EKG showed sinus tachycardia with T wave inversion Vi-V+ Arterial blood gases on room air were: pH 7.40, PaOz 28 mmHg
BIOSENTRY I UNIT 05 COOK COUNTY
12:oo 12/15/94
0o:oo
12:oo
TIME
Fig. 1. Tracing of continuous PaOz display. Crosses (+) represent values for intermittently obtained PaO*.
0o:oo
1003
Spurioushypoxemia Table 1. Comparison of Pa02 and Sa02 from arterial blood gases(ABG), continuous blood gas analysis (CBG) and pulse oximetry (SpOz) Date 12115 12116
Time
ABG
CBG
(PaO#a02) (Pa02/Sa02) 18:20 51.7 mmHg/87% 96 mmHgi96.9% 20:30 59.3 mmHg/90.5% 121 mmHg/98.2% 01:20 52.9 mmHg/87.9% 115 mmHg/97.6% 07:45 53.6 mmHgi88.3% 78 mmHg/93%
(3.7 kPa), PaCO* 35 mmHg (4.7 kPa), HC03 23 mmol/l, SaOz 48.8%. Pulse oximetry at that time revealed a saturation of 94%. The patient was subsequently transferred to the medical ICU where she received leukophoresis, hydroxyurea and packed red cell transfusions. Her abdominal pain worsened acutely. Broadspectrum antibiotics were administered and general surgery was consulted. A computerized tomographic scan of the abdomen revealed a large solid right adrenal mass with perinephric fluid collection. She subsequently developed respiratory distress which necessitated intubation and mechanical ventilation. Pulmonary artery catheterization showed: CO 5.8 l/min, CI 4.2 l/minlm2, SVR 397 dyne/s/em-5, PVR 205 dynelsicm-5, PA 42132 mmHg (mean 34 mmHg), PAWP 20 mmHg, CVP 16 mmHg. The patient developed wide complex tachycardia accompanied by hypotension. The tachycardia was felt to be supraventricular but did not respond to adenosine or cardioversion. The patient subsequently suffered cardiac arrest and could not be resuscitated. Autopsy was refused by the family. Prior to the acute deterioration which lead to her death, a BiosentryTM continuous intra-arterial blood gas sensor (Optex Biomedical, Inc., Woodlands, Texas, U.S.A.) was inserted into the right radial artery through the lumen of an indwelling arterial catheter. An in vitro calibration was performed prior to insertion using the method specified by the manufacturer. A number of arterial blood gases were obtained intermittently in the standard fashion through the arterial line. Each sample was immediately iced and analyzed within 3-4 min. The PaOz obtained was compared to the value for Pa02 displayed on the continuous monitor at the time the specimen was drawn (Fig. 1). The SpO2 shown on the pulse oximeter at the time of sampling was also compared to the SaOz derived from intermittent and continuous blood gas analysis. These data demonstrated that Pa02 obtained by the intermittent method was consistently lower than that from continuous blood gas analysis. In addition, Sp02 exhibited a reasonable degree of correlation with continuously derived Sa02, whereas Sp02 did not correlate with SaOz derived from intermittent blood gases. These data are shown in Table 1.
SPO2
93% 93% 95% 96%
WBC 871,200 688,800
Discussion Our data lend support to the premise that intermittent methods of arterial blood gas analysis may lead to spuriously low values for Pa02 in patients with extreme leukocytosis due to ongoing oxygen consumption in the period which precedes measurement. The question of whether cooling the specimen is effective in preventing this decline remains unresolved; prompt icing did not mitigate the fall in Pa02 in our patient. This corroborates similar observations in earlier studies [3, 8, 91. Several investigators have suggested using regression analysis of serial determinations of Pa02 in samples stored at room temperature as a means to assess oxygenation status [5, lo]. However, it is not clear whether the decay in Pa02 over time is linear as suggested, or exponential [lo]. Our data also support the use of pulse oximetry as a means to monitor oxygenation in patients with extreme leukocytosis, keeping in mind that SpO2 may be unreliable in the presence of elevated levels of methemoglobin [5]. Theoretically, continuous blood gas analysis presents an ideal method of assessing oxygenation status in hypoxemic patients with extreme leukocytosis or thrombocytosis, since there is no delay in measurement related to sampling. However, in vivo calibration of the sensor is unreliable in this setting so that questions remain regarding its accuracy. Although small studies performed in a variety of clinical settings have shown continuous blood gas analysis to be useful, its reliability and cost-effectiveness need to be confirmed in large controlled prospective trials [7]. References 1. Hess C. E., Nichols A. B., Hunt W. B. & Suratt P. M. (1979) Pseudohypoxemia secondary to leukemia and thrombocytosis. New Engl. J. Med. 301, 361. 2. Fox M. J., Browdy J. S. & Weintraub L. R. (1979) Leukocyte larceny: a causeof spurious hypoxemia. Am. J. Med. 67, 742.
3. Loke J. & Duffy T. P. (1984) Normal arterial oxygen saturation with the ear oximeter in patients with leukemia and leukocytosis. Cancer 53, 1767. 4. Rello J., Benito S., Triginer C. & Net A. (1989) False hypoxemia induced by leukocytosis [Letter]. Crit. Care Med. 17, 970.
1004
B. A. Mizock et al
5. Gartrell K. & Rosenstrauch W. (1993) Hypoxaemia in patients with hyperleukocytosis: true or spurious, and clinical implications. Leukemia Res. 17, 915. 6. Vaidya M. S., Das Gupta A., Pavri R. S., Baxi A. J. & Advani S. H. (1981) Acquired methemoglobinemia in leukemia: its etiopathogenesis and possible effects on red cell structure and function. Leukemia Res. 5, 265. 7. Lumsden T. & Marshall W. R. (1995) New developments in blood gas analysis: continuous intra-arterial determination of POz, pH, and PCOz. In Critical Care Monitoring: From Pre-hospital to the ICU (Levine R. L. & Fromm R. E. Jr, Eds), p. 223. Mosby, St Louis, IL, U.S.A.
8. Chillar R. K., Belman M. J. & Farbstein M. (1980) Explanation for apparent hypoxemia associated with extreme leukocytosis: leukocytic oxygen consumption. Blood 55, 922. 9. Shohat M., Schonfeld T., Zaizoz R., Cohen I. J. & Nitzan M. (1988) Determination of blood gases in children with extreme leukocytosis. Crit. Care Med. 16, 787. 10. Vincent F., Levy V., Bensousan T., Escudier B. & Leclercq B. (1994) Blood gas analysis in patient with hyperleucocytosis: validity of repeated measures on same sampling without ice? [Letter.] Leukemia Res. 18, 553.