Carbon monoxide analysis method in human blood by Airtight Gas Syringe – Gas Chromatography – Mass Spectrometry (AGS-GC-MS): Relevance for postmortem poisoning diagnosis

Highlights

• Accuracy profile validation of a AGS-GC-MS method for CO determination in human blood.

• Validation range of 10 to 200 nmol/mL HS, relevant for postmortem poisoning cases.

• Method was applied to samples from fatal intoxications.

• Results show higher CO concentrations compared to previous approaches.

• Approach has potential to increase number and quality of CO poisoning diagnoses.

Abstract

Carbon monoxide is one of the most abundant toxic air pollutants. Symptoms of a CO intoxication are non-specific, leading to a high number of misdiagnosed CO poisoning cases that are missing in the disease statistics. The chemical nature of the molecule makes it difficult to detect for long periods and at low levels, thus requiring a very accurate and sensitive method. Current methods capable of accurate and sensitive analyses are available, however an inconsistency between results and symptoms are frequently reported.

Therefore, an improved method for the analysis of carbon monoxide in blood and in the headspace (HS) of the sampling tube with the use of Airtight Gas Syringe – Gas Chromatography – Mass Spectrometry (AGS-GC-MS) is hereby presented and validated, for CO concentrations in a range of 10–200 nmol/mL HS (2–40 μmol/mL blood). Analytical LOQ is found at 0.9 nmol/mL HS (0.18 μmol/mL blood) and LOD at 0.1 nmol/mL gas. Application to intoxicated samples from autopsies and comparison to previously published methods show that this method is more appropriate, since performed under fully controlled conditions. Results show higher CO concentrations compared to previous approaches, indicating that results might have been underestimating the true blood CO burden. Therefore, this approach has the potential to help reduce the misdiagnosed cases and the gap between measurement and diagnosis of CO poisonings.

Introduction

A very simple chemical structure (molecule built by two atoms), formation during incomplete combustion of hydrocarbons and high occurrence in fires, exhaust fumes of motor vehicles, industrial exhaust gases, cigarette smoke and wood-fired stoves – those are all characteristics of carbon monoxide (CO) [1]. This odourless, tasteless and colourless gas has been related to numerous hospitalizations and deaths, not only due to its high toxicity, but mostly because of its chemical characteristics: exposure to CO occurs without the awareness of an individual. It is inhaled through the lungs and from there directly transferred to the blood stream [2]. Current knowledge affirms that once diffused to blood, CO combines with the haemoglobin (Hb) present to form carboxyhaemoglobin (COHb) [3] and is also transported to the tissues [4]. Hb is the oxygen-carrying protein molecule present in red blood cells. One of the main characteristics of CO is its high affinity for Hb, being 200–250 times higher compared to the affinity of oxygen (O₂) [5]. This results in CO competing with and displacing O₂ from the binding sites on the haeme, leading to a reduced oxygen-carrying and -storage capacity of Hb [6]. The main organs suffering from the deriving hypoxia are the brain and heart, since they are the organs with the highest oxygen requirement [5].

Other known damages caused by CO include the inhibition of mitochondrial respiration, the excess-activation of platelets (resulting in inflammation-like effects), ischemic and anoxic brain injuries and the generation of free radicals, which are known to be mutagenic and tumour cells-promoters [7,8].

The severity of the damages caused to an individual exposed to CO is related to the quantity and time of exposure to CO. However, the symptoms of a CO poisoning, which include dizziness, nausea, headache and respiratory troubles, do not always present themselves immediately, but appear only after a certain time delay, and when they do, they are often attributed to other types of diseases or infections [4]. Therefore, it is of high importance to have accurate and reliable, but also rapid and simple methods to measure the levels of CO poisoning, especially in cases where the symptoms do not give a clear indication of the causes.

Due to the high affinity of CO to Hb, it is assumed that the majority of CO binds with Hb when introduced in the blood circulation, which resulted in COHb being the primary biomarker for CO poisonings [1]. CO is mainly eliminated unchanged through the lungs. Between 10 and 50% of CO in the organism is bound to tissue proteins, mainly myoglobine (Mb) and cytochrome c oxidase, and the rest is thought to be under bound form as COHb [9].

Until now, the most common technique used in clinical as well as post-mortem routine analyses is the measurement of the COHb-levels by CO-oximetry (blood analysis). CO-oximetry is a technique based on automated differential spectrophotometry, which measures the concentration of an analyte by relating it to the measured absorbance when exposed to light of different wavelengths, according to the Lambert-Beer-Law. With a CO-oximeter, the saturation levels of COHb (%), methaemoglobin (MetHb), oxyhaemoglobin (O₂Hb) and normal, non-carrying haemoglobin (HHb) are measured [10]. Pulse CO-oximeters (clinical finger monitoring) can additionally determine standard pulse oximeter parameters such as oxygen saturation, pulse rate and perfusion index [11]. The major advantage of pulse CO-oximetry is that the measurement is done continuously and is non-invasive, thus allowing the monitoring of the parameters in a clinical setting, without causing pain or damage to the patient. However, this technique cannot be used in postmortem samples, since an active blood circulation is needed to obtain results and clinical samples taken perimortem have an excessively significant sampling time delay.

A major drawback of spectrophotometric methods is the dependence on the optical state of the sample. Degradation of the sample due to storage as well as postmortem interferences, such as thermo-coagulation [12], contamination due to incomplete haemolysis, high lipid concentrations or thrombocytosis and putrefaction [13], can change the blood state and result in either an alteration of the measurement or the impossibility of the device to determine a value. Consequently, another biomarker of CO exposure should be investigated and its detection method should not be optical-based.

Therefore, techniques focused on direct CO rather than optical ones that focus on COHb, which are independent of the quality of the blood sample, have been investigated and developed. The most successful was found to be Gas Chromatography (GC) in combination with a variety of detection methods, such as thermo conductivity detector (TCD), flame-ionization detector (FID), Reduced Gas Analyzer (RGA) and Mass Spectrometry (MS).

In gas chromatographic CO detection, CO is released at gaseous state through a liberating agent, after lysis of the blood, and then analysed. Haemolysis is performed through the use of a haemolytic agent, the most common ones being saponine, Triton X-100 or other detergents. Liberation of the CO occurs through the reaction with a strong acid, which yields CO and water as the only products [12,[14], [15], [16], [17]]. As releasing agents, sulphuric acid (H₂SO₄), hydrochloric acid (HCl) and potassium ferricyanide (K₃Fe(CN)₆) are generally used. Other acids such as lactic acid [18], citric acid [18,19] or phosphoric acid [19] have also been tested.

For the gas chromatographic separation, a capillary column with a 5 Å molecular sieve has been found to be specific for the separation of CO from other interfering gases such as carbon dioxide (CO₂), nitrogen (N₂), oxygen (O₂) and methane (CH₄) [20].

Detection of the analyte is achieved with numerous detectors linked to the gas chromatograph. The first detector applied to CO-determination was TCD [15], later replaced by other detectors, such as FID [20]. For detection with FID, CO is chemically reduced to methane with a methanizer and then detected. This method is very sensitive and specific and was the most popular detection system used in conjunction with CO [[17], [18], [19],[21], [22], [23], [24]]. Nevertheless, one of its major drawbacks is the fact that the addition of a methanizer to the apparatus is needed, which limits the use of the instrument only for CO-analysis. Therefore, another type of detector employed was MS. The developed MS methods are more simple, rapid, accurate, reproducible, in addition to the versatility of the instrument, since it can be used for all types of analysis and hence is useful in laboratories for routine analyses [13,25,26]. Furthermore, MS allows for a higher power of identification: additionally to the retention time, the compounds are identified with the mass spectrum, which allows quantification with a stable labelled isotope as internal standard.

One issue regarding all measurement methods is the calibration. Calibration of the techniques was performed either with pure CO gas, which was diluted appropriately, or with the fortification of blood with CO to reach different COHb% saturation levels. In the latter, 100%-saturation was confirmed with either UV-spectrophotometry or CO-oximetry. Reliability can be debated though, considering that, first of all, the spectrophotometric methods used at that time were only detecting at several wavelengths, while modern CO-oximeters analyse the full spectrum, leading to a possible error in the obtained values. Secondly, these optical methods only measure the CO bound to Hb, not taking into account possible dissolved CO present in the sample that was not taken into account when building the calibration curve, which could shift the ‘real’ curve into higher levels of CO poisoning.

An alternative calibration method was developed firstly in 1993, where Cardeal et al. used the reaction of formic acid with sulphuric acid to form CO [19]. Varlet et al. went another step further by developing an approach which uses isotopically labelled formic acid (¹³HCOOH) to produce ¹³CO as internal standard for a Headspace (HS)-GC-MS method [26].

The HS-GC-MS approach with isotopically labelled formic acid used for building of the calibration curve shows the most accuracy, sensitivity, specificity and reproducibility. However, after development and validation [26,27], no further research was carried out in the field.

An additional issue involves the currently existing correlation between the COHb%-levels and the symptoms developed by patients, which do not always agree: patients were found to have an elevated COHb% saturation level, but showed no signs of CO-intoxication, while other patients with a low COHb%-level lost consciousness or suffered severe delayed consequences [28]. Thus, there seems to be a great fallacy in the understanding of the true role played by CO in poisoning cases. This might be due to an underestimation of the total CO measured with the current techniques and the neglect of the possible presence of CO in dissolved state and not bound to Hb, which can have major implications in the role of CO in the pathophysiology of a CO-poisoning.

Therefore, an improved approach by Airtight Gas Syringe (AGS) followed by GC-MS for CO determination is hereby presented, which not only shows improved sensitivity and lower costs, but also takes into account the total amount of CO present in blood by analysing the CO in blood and in the headspace of the blood tube used to store the sample, with high importance from both an analytical and clinical point of view. This constitutes the first step to acknowledge the significance of total CO in blood as alternative biomarker for CO exposure.