The authors have no relevant financial information or potential conflicts of interest to disclose.
Original Research Contribution
The Diagnostic Value of Serum Ischemia-modified Albumin Levels in Experimentally Induced Carbon Monoxide Poisoning and Their Correlation With Poisoning Severity
Article first published online: 19 JUN 2013
© 2013 by the Society for Academic Emergency Medicine
Academic Emergency Medicine
Volume 20, Issue 7, pages 652–658, July 2013
How to Cite
Academic Emergency Medicine 2013; 20:652–658 © 2013 by the Society for Academic Emergency Medicine
- Issue published online: 16 JUL 2013
- Article first published online: 19 JUN 2013
- Manuscript Accepted: 31 JAN 2013
- Manuscript Revised: 30 JAN 2013
- Manuscript Revised: 22 JAN 2013
- Manuscript Received: 4 DEC 2012
The objectives were to determine the diagnostic value of blood ischemia-modified albumin (IMA) levels in experimentally induced carbon monoxide (CO) poisoning and to analyze their correlation with poisoning severity.
Thirty-six female rats were randomly assigned to one of three groups: I (control group), II (low-dose CO poisoning group), and III (high-dose CO poisoning group). The control group was kept in room air, while groups II and III were exposed to 3 L/min of 3,000 ppm and 3 L/min of 5,000 ppm CO gas for 30 minutes, respectively. Serum carboxyhemoglobin (COHb), IMA, and malondialdehyde (MDA) levels; brain, heart, lung, liver, and kidney tissue MDA measurements; and histopathologic damage scores were then compared.
IMA levels were significantly higher in groups II and III than in group I. A moderate positive correlation was observed between COHb and IMA levels. There was a strong positive correlation between COHb levels and degree of damage in all organs, but IMA and MDA levels did not reflect a similar correlation.
Ischemia-modified albumin levels are higher in rats exposed to CO. This indicates that IMA levels can potentially be important in the diagnosis of exposure to CO or of CO poisoning. However, IMA levels are not a good biochemical marker in terms of determining the severity of poisoning.
El Valor Diagnóstico de los Niveles en Plasma de Albúmina Modificada por la Isquemia en la Intoxicación por Monóxido de Carbono Provocada de Forma Experimental y Sus Correlaciones con la Gravedad de la Intoxicación
Determinar el valor diagnóstico de los niveles en sangre de albúmina modificada por la isquemia (AMI) en la intoxicación por monóxido de carbono (CO) provocada de forma experimental y analizar sus correlaciones con la gravedad de la intoxicación.
Se asignaron de forma aleatoria 36 ratas hembras a uno de los 3 grupos: I (grupo control), II (grupo de intoxicación por CO a dosis baja) y III (grupo de intoxicación por CO a dosis alta). El grupo control se mantuvo en aire ambiental, mientras que los grupos II y III se expusieron a 3 L/min de 3.000 mg/L y 3 L/min de 5.000 mg/L de gas CO, durante 30 minutos, respectivamente. Se compararon los niveles de carboxihemoglobina (COHb), AMI y malondialdehido (MDA) en suero; las mediciones de MDA en tejido cerebral, cardiaco, pulmonar, hepático, y renal; y las puntuaciones de lesión histopatológica.
Los niveles de AMI fueron significativamente más elevados en los grupos II y III en compración con el grupo I. Se observó una correlación moderadamente positiva entre los niveles de COHb y AMI. Hubo una correlación positiva fuerte entre los niveles de COHb y el grado de lesión en todos los órganos, pero los niveles de AMI y MDA no reflejaron una correlación similar.
Los niveles de AMI son más altos en las ratas expuestas a CO. Esto indica que los niveles de AMI pueden potencialmente ser importantes en el diagnóstico de la exposición a CO o de la intoxicación por CO. Sin embargo, los niveles de AMI no son buenos marcadores bioquímicos en términos de determinar la gravedad de la intoxicación.
Carbon monoxide (CO) poisoning is a significant health problem seen frequently across the world. CO poisoning results in an estimated 50,000 emergency department (ED) visits in the United States annually and is one of the leading causes of death from poisoning.[1, 2] The affinity of hemoglobin for CO is 230 to 270 times greater than that of oxygen (O2). In CO poisoning, the requisite amount of oxygen cannot be transported to the tissues because of the carboxyhemoglobin (COHb) complex that arises from a combination of CO and Hb. In addition to causing hypoxia, the COHb complex can also lead to injury by impairing tissue perfusion through causing peripheral vasodilation and hypotension. Serious clinical findings, such as noncardiogenic pulmonary edema, rhabdomyolysis, disseminated intravascular coagulation, multiorgan failure, and acute tubular necrosis are known to appear due to hypoxia developing in association with CO poisoning.
The measurement of blood COHb concentrations is an important diagnostic technique, but there is only a weak correlation between blood COHb concentrations and organ damage in the clinical setting. COHb concentrations exceeding 60% generally result in death, while at lower concentrations clinical findings varying from mild symptoms to coma can be seen, and it is therefore not always possible to determine clinical symptoms and consequences using COHb concentrations. Various biochemical markers other than COHb have been and are still being investigated to identify beforehand the clinical results of this poisoning with its significant neurotoxic and cardiotoxic effects. The literature contains only one study evaluating whether blood ischemia-modified albumin (IMA) concentrations, which have been shown to rise in several ischemic conditions and have entered into clinical use in the early diagnosis of diseases such as myocardial ischemia and ischemic stroke in particular, can be used in the diagnosis of CO intoxication and associated organ injury.[7, 8] Our aim in this study was to determine whether there is a diagnostically significant change in serum IMA concentrations with exposure of differing intensities in acute CO poisoning experimentally induced in rats and whether there is a correlation between CO exposure-associated organ damage and IMA concentrations.
This was a randomized, controlled, nonblinded animal study, the protocol for which was approved by the Karadeniz Technical University Animal Care and Ethics Committee. In establishing our experimental model, the most frequently employed experimental intoxication protocol in previous studies, involving 3,000 and 5,000 ppm CO, was used.[9, 10]
Eighteen mature female Sprague Dawley rats (10 weeks old, weighing 240–280 g) were used. The animals were kept in steel cages until the day of the study at a room temperature of 22°C and were given water and standard rat chow. Only water was provided for the last 12 hours before the study. The CO gas to be used in the experimental protocol was ordered from HABAŞ Sınai veTıbbiGazlarEndüstrisi A.Ş. (İzmit, Turkey), in such a way as to include 3,000 and 5,000 ppm concentrations of CO. For the protocol, a special experimental bell jar, 100 × 40 × 50 cm in size, in the form of a two-part aquarium, capable of admitting gas from one side and with a discharge hole on the other, was prepared.
The 18 rats were randomized into three groups of six individuals each. Group I—control group rats were placed inside the bell jar and breathed room air for 30 minutes at room temperature; group II—low-dose CO exposure group rats were placed inside the experimental bell jar and exposure established with 30 minutes of respiration of a 3,000 ppm concentration of CO gas mixture at 3 L/min; group III—high-dose CO exposure group rats were placed inside the experimental bell jar and exposure was established with 30 minutes of respiration of a 5,000 ppm concentration of CO gas mixture at 3 L/min.
At the end of 30 minutes, general anesthesia was performed in all groups with the intramuscular administration of 50 mg/kg ketamine and 5 mg/kg xylazine. Laparotomy was performed and the abdominal aorta exposed; 4-mL blood specimens were collected through puncture of the abdominal aorta for IMA and COHb measurement. Euthanasia was subsequently performed. Blood specimens and tissue specimens (brain, heart, liver, lung, and kidney) were collected, COHb and IMA concentrations were measured, and histopathologic analysis was performed. Tissues were preserved in a 10% formaldehyde solution for histopathologic analysis.
Once blood samples were taken, serum and plasma specimens were centrifuged for 15 minutes at 1800 × g. The specimens were pipetted into Eppendorf tubes and stored at –80°C until analysis. Reduced cobalt to albumin binding capacity (IMA concentration) was analyzed using the rapid colorimetric method. The results are reported as absorbance units.
COHb measurements were performed automatically with a blood gas device (Rapidlab 1265, Bayer Health Care LLC, Pittsburgh, PA) in the experimental laboratory.
Blood Malondialdehyde (MDA) Measurement
MDA concentrations in plasma samples were measured using the thiobarbituric acid–reactive substance method described by Yagi. Tetramethoxypropane was used as a standard, and MDA concentrations were calculated as nmol/mL.
Tissue MDA Measurement
A piece of testis tissue was used to measure MDA concentrations. The sample was minced and homogenized in an ice-cold 1.15% KCl solution containing 0.05% Triton X-100 using an Ultra-TurraxT25 homogenizer (Janke and Kunkel IKA, Staufen, Germany). Tissue MDA concentrations were determined using the method described by Mihara and Uchiyama. Tetramethoxypropane was used as a standard. MDA concentrations were calculated as nmol/mL per gram of wet tissue.
Tissue specimens were fixed for 48 hours in a 10% formaldehyde solution for histopathologic analysis. Tissue parts from all groups were dehydrated by being passed through different degrees of alcohol and made transparent by being passed through a xylene solution. Paraffin tissue blocks were prepared and 5-μm-thick sections were taken for microscopic examination. Following deparaffinization, sections were stained with hematoxylin-eosin. Histopathologic analysis was performed by an experienced histologist blinded to the groups. Damage parameters for each organ were evaluated by reviewing at least five microscopic fields at 200 × and 400 × magnifications. An Olympus BX 51 light microscope (Olympus, Tokyo, Japan) was used for the evaluation of preparates. Preparates for all organs were scored semiquantitatively (0 = none, 1 = mild, 2 = average, 3 = severe).
Statistical analysis was performed using SPSS 15.0 (IBM SPSS, Armonk, NY). Kruskal Wallis analysis of variance, followed by post hoc testing using the Mann Whitney U-test with Bonferroni correction, was conducted to compare the groups. Spearman's correlation analysis was used to assess the relationship between biochemical parameters and histopathologic damage scores. Statistical significance was set at p < 0.05, with no adjustment for multiple testing.
|Parameters||Group I||Group II||Group III|
|COHb%||0.21a (0.09–0.24)||11.25* (9.32–13.22)||17.15† (14.02–21.57)|
|IMA (ABSU)||1.23a (1.14–1.32)||1.57* (1.51–1.65)||1.61† (1.58–1.75)|
|Serum MDA||0.66 (0.51–0.76)||0.61 (0.55–0.85)||0.81 (0.68–0.92)|
|Brain tissue MDA||867.70 (756.9–1044.9)||882.90 (774.2–958.8)||821.70 (703.6–950.1)|
|Heart tissue MDA||329.60 (240.4–380.6)||292.20 (236.2–323.1)||204.90 (163.1–240.9)|
|Kidney tissue MDA||744.60 (666.7–813.9)||802.90 (542.2–882.5)||641.80 (485.6–832.5)|
|Lung tissue MDA||406.80 (376.1–518.7)||364.40 (332.6–428.5)||382.40 (206.1–441.7)|
|Liver tissue MDA||661.90 (445.5–1088.9)||819.60 (451.1–1028.4)||685.20 (544.5–882.2)|
Correlation analysis between blood biochemical parameters in rats exposed to CO revealed a significant positive correlation only between COHb and IMA concentrations (r = 0.51, p = 0.03). The correlations between serum MDA and COHb (r = 0.45, p = 0.06) and between MDA and IMA (r = 0.14, p = 0.58) were not significant. There was no significant correlation between tissue MDA concentrations investigated for all organs and all biochemical parameters (IMA, COHb concentrations, and serum MDA concentrations; p > 0.05 for all evaluations).
The results of the histopathologic evaluations performed for groups I, II, and III and the median degrees of damage are shown in Table 2. Pronounced hemorrhage, leukocyte infiltration, and degeneration in the alveolar cells was present in the lung preparates from groups II and III (Figure 2A) compared to the control group. In the evaluation of brain tissue preparates there were more neuronal cells with hyperchromatic nuclei and vascular congestion in group III when compared to the control group. From the semiquantitative analysis, neurons with a hyperchromatic nucleus were more numerous in group II when compared to group III (Figure 2B). From the evaluation of heart tissue preparates, occasional spaces between myocytes and pronounced vascular congestion were present in group III when compared to the control group. Vascular congestion was greater in group II compared to the other groups (Figure 2C). In the evaluation of kidney tissue preparates, there was pronounced vascular congestion between the tubules in Groups II and III (Figure 2D), compared to the control group. In addition, there was leukocyte infiltration and degeneration in the tubular cells. No significant difference was observed between groups II and III in terms of vascular congestion, tubular cell degeneration, or leukocyte infiltration. The evaluation of liver preparates showed sinusoidal dilatation was more pronounced in group II compared to group III and the control group. Sinusoidal dilatation and pronounced vascular congestion between sinusoids were present in group II (Figure 2E).
|Location||Group I||Group II||Group III|
|Neuron cell degeneration||0.5 (0–1)||2 (1–2)||1 (1–2)|
|Vascular congestion||0 (0–1)||1.5 (1–2)||2 (2–3)|
|Total histologic injury||0.5 (0–2)*†||3 (2–4)*||3.5 (3–4)†|
|Degeneration in myocytes||0 (0–1)||2 (1–2)||2 (1–2)|
|Vascular congestion||0.5 (0–1)||3 (2–3)||3 (2–3)|
|Total histologic injury||1 (0–1)*†||4.5 (4–5)*||4.5 (4–5)†|
|Tubular cell degeneration||1 (0–1)||2 (1–2)||2 (1–2)|
|Medullar congestion||1 (0–1)||3 (2–3)||3 (2–3)|
|Leukocyte infiltration||0 (0–0)||2 (1–2)||1.5 (1–2)|
|Total histologic injury||1.5 (0–2)*†||6.5 (5–7)*||6 (5–7)†|
|Leukocyte infiltration||1 (0–1)||3 (2–3)||2 (1–3)|
|Hemorrhage||1.5 (1–2)||3 (2–3)||2.5 (2–3)|
|Alveolar cell degeneration||1 (0–1)||1.5 (1–2)||1 (1–2)|
|Alveolar edema||0 (0–0)||0 (0–1)||0.5 (0–1)|
|Total histologic damage||3.5 (2–4)*†||7 (6–8)*||6.5 (5–8)†|
|Sinusoidal dilatation||0 (0–1)||0 (0–1)||0.5 (0–1)|
|Vascular congestion||0 (0–0)||3 (2–3)||3 (2–3)|
|Total histologic damage||0 (0–1)*†||3 (3–3)*||3 (2–4)†|
The degrees of histologic damage in the organs were correlated with one another; heart, brain, and kidney damage scores were particularly significantly, positively correlated (Table 3). Results of examination of the correlations between histologic damage and COHb and IMA concentrations in the various organs are shown in Table 4. There was no significant correlation between serum and tissue IMA concentrations and organ histologic damage (p > 0.05 for all comparisons).
Our study investigated whether there are diagnostically valuable changes in serum IMA concentrations in different intensities of exposure to CO in acute CO poisoning induced experimentally and whether there is a correlation between CO exposure–dependent organ damage and serum IMA concentrations. The results show that serum IMA concentrations rise significantly with exposure to CO. Although this rise exhibited a correlation with blood COHb concentrations, there was no correlation between organ damage and IMA concentrations. This indicates that IMA concentrations have the potential to be used as a supplementary diagnostic tool in addition to COHb in the diagnosis of CO exposure or poisoning, but that IMA concentrations are not a good biochemical marker in determining the severity of poisoning on the basis of histopathologic damage. New studies analyzing different dimensions of this subject, which has only been the subject of limited research to date, are now needed. IMA can thus be confirmed as a useful assistant test in diagnosis, treatment planning, and maybe also prognosis under specific conditions in which COHb concentrations are known to be affected (such as delayed transportation or when blood COHb concentrations are incompatible with clinical condition, as in patients receiving oxygen therapy before reaching the hospital).
Carbon monoxide poisoning is a significant cause of morbidity and mortality. Prior studies have shown a correlation between severity of CO poisoning and CO poisoning–associated sequelae, with various clinical parameters such as COHb concentrations or the presence of neurologic symptoms.[14, 15] Although studies have reflected symptoms associated with COHb concentrations, there is controversy whether the correlation is not clear or whether there is a weak correlation between COHb concentrations and the clinical picture in patients with CO poisoning. This is because COHb concentrations exhibit a wide range of variations in patients presenting with CO poisoning. These broad variations may have various causes, such as concentration of CO exposure, duration of exposure, time until measurement for COHb concentration, the basal condition of the patient exposed to CO poisoning, individual sensitivity to CO poisoning, and support oxygen therapy administered during the period up to COHb measurement. For these reasons, there is still a need for an alternative biochemical parameter to COHb concentrations, a conventional marker with low long-term stability, in the diagnosis of CO poisoning and determination of prognosis. Although no studies have yet been performed directly for that purpose, IMA may be a more stable parameter compared with COHb, which is known to be easily affected by several agents cited above.
Similar or different problems are encountered in biochemical parameters employed for the early diagnosis of many diseases, and contemporary researchers are seeking an optimal parameter. IMA is one of the biochemical parameters on which these activities have focused. The measurement of albumin species that arise with changes in the metal-binding capacity of albumin in ischemic events is important and practicable in the diagnosis of many ischemic diseases. IMA measurement has been studied in the early diagnosis of myocardial ischemia in patients with acute chest pain. One broad study established that IMA had a negative predictive value as high as 90% in the exclusion of acute coronary syndrome in patients presenting to the ED with chest pain and that together with negative cardiac troponins and nondiagnostic electrocardiogram, the negative predictive value rises as high as 97.1%. In addition, IMA concentrations have been shown to be elevated in acute ischemic events such as pulmonary embolism, mesenteric ischemia, peripheral arterial occlusion, stroke, and acute cardiac arrest, and it has been suggested that they can be a diagnostic or prognostic marker in these diseases.[19, 20]
The only study in the literature, to our knowledge, investigating whether IMA concentrations are a valuable biochemical parameter in patients with CO poisoning is a recent one by Turedi et al. They compared 33 patients with CO poisoning with 49 healthy controls and determined higher blood IMA concentrations in the patients with CO poisoning both at time of presentation and at the third hour of treatment. COHb concentrations in patients with CO poisoning at time of presentation fell significantly at the third hour of treatment and their diagnostic value decreased. However, the study showed no significant difference between blood IMA concentrations measured at time of presentation and at the third hour of treatment. From that perspective, they reported that IMA concentrations are a sensitive parameter, and maybe a more stable one than COHb concentrations, in patients with CO poisoning. In our study, the first experimental rat study to evaluate IMA concentrations in CO poisoning, IMA concentrations in the group with low CO exposure and also in the group with high CO exposure were significantly higher than in the group with no exposure. The elevated IMA concentrations we determined in rats with CO poisoning are compatible with those of the clinical study by Turedi et al.
Carboxyhemoglobin and IMA concentrations were positively correlated in our study (r = 0.51, p = 0.03). This may be regarded as a reflection of oxidative stress and hypoxic status emerging with high blood COHb concentrations. However, no similar correlation was determined between COHb concentrations and MDA concentrations, regarded as a marker of oxidative stress. In contrast to our study, Turedi et al. determined no significant correlation between serum IMA concentrations and COHb concentrations in patients with CO poisoning (r = –0.19, p = 0.273). This should be investigated in future comprehensive studies with larger patient numbers.
Carbon monoxide exhibits its toxic effect by causing hypoxia and gives rise to various systemic and neurologic complications. Because of their high oxygen requirements, the brain, heart, and kidneys are the organs most sensitive to the hypoxic effects of CO exposure. Central nervous system activation is held responsible for the majority of symptoms in CO poisoning. In our study, in addition to the organs most affected by exposure to CO poisoning, we also evaluated the histopathologic effects on lung and liver tissue in rats with exposure to CO poisoning. In our histopathologic analysis, significantly higher histopathologic damage was detected in both the 3,000 ppm low CO and the 5,000 ppm high CO groups compared to the control group. Organs' degrees of histopathologic damage were correlated with one another, and heart, brain, and kidney damage scores in particular exhibited positive correlations with one another. This was interpreted as indicating a global hypoxic process in the entire body, not solely in the heart, brain, and kidneys. Examination of the correlation between histopathologic damage and biochemical parameters showed a positive correlation between the presence of COHb and organ damage, but there was no correlation between rising COHb concentrations and organ damage (no dose response), although IMA concentrations and MDA concentrations reflected no such relationship. In the clinical study by Turedi et al., when patients were classified on the basis of severity of clinical poisoning, COHb concentrations rose with increasing intensity of poisoning, but IMA concentrations did not change significantly in any poisoning severity-dependent manner. The results both from the clinical study by Turedi et al. and from our study based on histopathologic damage show that IMA concentrations are not a good biochemical marker in the determination of severity of CO poisoning.
This study was performed with a limited number of subjects due to ethics committee requirements. Therefore, some of our comparisons may be underpowered to detect significant differences, particularly between intervention groups, as opposed to comparisons between the intervention and control groups. For this reason, our results must be confirmed with larger sample sizes. The study also had some limitations in terms of the model employed. Our study was controlled, but may not mimic typical CO poisoning cases seen in clinical practice. Additionally, our study examined injury and biochemical changes arising as a result of exposure to high and low doses for only 30 minutes. Higher intoxication concentrations could not be achieved due to the exposure being kept limited and we had no means of measuring changes that might take place in the parameters we evaluated at such higher concentrations. CO exposure was only considered as a foundation in our study; no additional evaluations were performed considering the factors that affect COHb concentrations, such as the administration of support oxygen therapy before measurement, and the effects of these factors on COHb, IMA concentrations, and MDA concentrations.
Additionally, IMA is a new biomarker influenced significantly by a wide array of physiologic variables, including exercise and hydration. Although no procedure other than CO exposure was performed on the subjects during our study, we were not able to control all those variables that might possibly influence IMA concentrations. Finally, we used a noncommercial IMA test, which may be less reproducible than the standard commercial assay.
Serum ischemia-modified albumin concentrations are elevated in a significant manner in line with carbon monoxide exposure, and that increase exhibits a positive correlation with carboxyhemoglobin concentrations. This indicates that ischemia-modified albumin concentrations can potentially be used as an additional diagnostic tool to carboxyhemoglobin in the diagnosis of carbon monoxide exposure or poisoning. The absence of a significant correlation between histopathologic damage and ischemia-modified albumin shows that ischemia-modified albumin is not a good biochemical marker in terms of determining severity of poisoning.
- 1Carbon monoxide poisoning. Türkiye Klinikleri J Surg Med Sci. 2006; 2:74–80., .
- 3Carbon monoxide poisoning. In: Tintinalli JE, Kelen GD, Stapczynski JS. Emergency Medicine: A Comprehensive Study Guide, 6th ed. New York, NY: McGraw-Hill, 2004, pp 1238–42..