British Journal of Anaesthesia 91 (6): 797±9 (2003)
DOI: 10.1093/bja/aeg271
Real-time breath monitoring of propofol and its volatile metabolites during surgery using a novel mass spectrometric technique: a feasibility study G. R. Harrison1, A. D. J. Critchley2, C. A. Mayhew2* and J. M. Thompson2 1
Featherstone Department of Anaesthesia and Intensive Care, Queen Elizabeth Hospital, University Hospital Birmingham NHS Trust, Birmingham B15 2TH, UK. 2Molecular Physics Group, School of Physics and Astronomy, University of Birmingham, Birmingham B15 2TT, UK *Corresponding author. E-mail:
[email protected] Background. At present, there is no rapid method for determining the plasma concentration of i.v. anaesthetics. A solution might be the measurement of the anaesthetic concentration in expired breath and its relation to the plasma concentration. We used chemical ionization methods to determine whether an i.v. anaesthetic can be detected in the low concentrations (parts per billion by volume) in the expired breath of an anaesthetized patient. Method. Chemical ionization mass spectrometry can measure trace gases in air with high sensitivity without interference from major gases. We carried out a feasibility trial with a proton transfer reaction mass spectrometer (PTR-MS) to monitor the i.v. anaesthetic agent propofol and two of its metabolites in exhaled gas from an anaesthetic circuit. Exhaled gas was sampled via a 4 m long, unheated tube connected to the PTR-MS. Results. Propofol and its metabolites were monitored in real time in the expired breath of patients undergoing surgery. Conclusion. Routine measurement of i.v. agents, analogous to that for volatile anaesthetic agents, may be possible. Br J Anaesth 2003; 91: 797±9 Keywords: anaesthetics i.v., propofol; equipment, proton transfer reaction mass spectrometer; ions, chemical ionization; ventilation, breath analysis Accepted for publication: June 29, 2003
Propofol (2,6 di-isopropyl phenol) is a common i.v. anaesthetic. However, the plasma concentration of the agent cannot be continuously monitored during surgery, and therefore no direct assessment can be made of the plasma level that ensures adequate anaesthesia. We set out to show that soft chemical ionization techniques can measure propofol and its volatile metabolites (2,6 di-isopropyl quinone and 2,6 di-isopropyl quinol) in exhaled gas in real-time. The results of this study illustrate that soft ionization mass spectrometric techniques may be able to provide real-time point-of-care measurements.
Methods Compared with conventional electron impact ionization, mass spectrometry using soft chemical ionization allows
trace molecules in air to be ionized with minimal fragmentation so that the major gases found in air or used in anaesthesia can be left unionized. Sensitive trace-gas analysis techniques can be obtained, giving opportunities for expired breath analysis of multiple markers that can be used for non-invasive diagnosis of disease (metabonomics) and drug monitoring. Until recently, soft chemical ionization techniques for very low concentration gas analysis have been limited to large laboratory-based instruments1 that are not able to be used in ITUs, operating theatres or out-patient clinics. Recently, a portable, low-power proton transfer reaction mass spectrometer (PTR-MS) has been developed.2 The PTR-MS can be used to measure trace gases in expired breath by exploiting unique features of the reactions of protonated water, H3O+, with neutral molecules (M):
Ó The Board of Management and Trustees of the British Journal of Anaesthesia 2003
Harrison et al.
H3O+ + M ® MH+ + H2O Trace components (M) are selectively converted to ions (MH+) through reaction with H3O+. There is a linear relationship between the recorded MH+ signal and the concentration of M in the original gas mixture, so that the latter can be determined provided the mass transmission characteristics of the instrument are known. Detection levels for MH+ product ions are very low, so that sensitivities reaching the few tens of parts per trillion by volume level are possible. Importantly, the major gases present in breath (nitrogen, oxygen, and carbon dioxide) and gases such as nitrous oxide do not react with H3O+ because they have much less af®nity for H+ than does H2O, so no proton transfer reactions occur. Formation of MH+ is thus selective to trace gases and vapours, especially many volatile organic compounds (VOCs). For many VOC reactions MH+ is the sole product, so, from the mass spectrum, masses of the original neutral molecules M may be identi®ed. We conducted a small feasibility trial to measure propofol in expired breath during gynaecological surgery, using the PTR-MS, undertaken in Birmingham Women's Hospital NHS Trust. The chairman of the South Birmingham Local Research Ethics Committee gave approval for this study (Chair's action). The ®rst author anaesthetized the patients. A standard i.v. anaesthetic technique was used: induction was with fentanyl (1±2 mg kg±1), then propofol from a diprifusor (set at 8 mg ml±1) until loss of eyelash re¯ex. Tracheal intubation was facilitated with atracurium (0.5 mg kg±1) and the lungs were ventilated with a 50/50 mixture of nitrous oxide in oxygen. Muscle relaxation was maintained with atracurium. During surgery, propofol was infused to obtain a predicted plasma concentration of 8 mg ml±1 and this was gradually reduced to a predicted value of 4 mg ml±1 as surgery progressed. The PTR-MS took a continuous sample of gas from the expiratory limb of the anaesthetic circuit. The inlet ¯ow was set at approximately 15 ml min±1, which maintained the reaction region of the PTR-MS at a pressure of 2 mbar. VOCs in the gas sample whose proton af®nities were greater than that of water became charged by the transfer of a proton from H3O+. The protonated VOCs then entered the mass spectrometer region of the instrument where they were detected.
Results In pre-clinical testing, the PTR-MS was tested for measuring propofol at expected expired breath concentrations, by sampling headspace air above serum taken previously from an anaesthetized patient. Figure 1 illustrates the mass spectrum obtained, with background peaks subtracted. A large peak at 179 atomic mass units (amu) dominates; smaller peaks at 95, 137, and 193 amu are distinguishable. The 179 amu peak is protonated propofol. Those at 95 and 137 amu are fragment ions from collision-induced
Fig 1 Mass spectrum of the air sampled over serum taken from a patient anaesthetized with propofol. This spectrum represents the average of 10 scans with the background subtracted. The peaks, which can be uniquely de®ned to propofol, are identi®ed. The other mass peaks at 97, 101, 107, 111, 113, 115, 135, 139, and 141 amu remain unassigned.
dissociation (CID) of protonated propofol in the PTR-MS. The protonated quinone metabolite peak is at 193 amu. These results were veri®ed by sampling air above pure propofol, which provided a much greater signal, so that masses could be readily identi®ed. The presence of CID products are useful, because they were used in the experiments with the expired breath of patients (see below) to ensure that any mass at 179 amu is indeed a result of propofol and not that resulting from some other drug, for example neuromuscular blocking agents. In clinical feasibility tests, a 4 m long, unheated sampling tube of 1.59 mm internal diameter conveyed exhaled breath from each of the ®ve anaesthetized patients to the PTR-MS in the adjacent anaesthetic room. The tube was attached to the anaesthetic circuit expiratory limb, shortly after patients were taken into theatre, and all patients were anaesthetized before monitoring was started. Considerable losses within the sampling system, primarily on the tube's surface, were expected because laboratory measurements showed that the intensity of the propofol peak was dependent on the length of the sampling tube used. Despite these losses, propofol and its metabolites were measured in patients' expired breath. Figure 2 shows concentration variations early during surgery of propofol and its quinol and quinone metabolites in an anaesthetized patient's expired breath. That particular patient was very heavy (>120 kg). A rapid rise and a subsequent steady decline in propofol and metabolite concentrations in this patient's breath in the early phase of surgery are clearly seen.
Discussion We found that propofol and its metabolites could be measured in patients' exhaled gases during surgery. The intensity of the propofol peaks we observed in the mass spectra indicate that this technique could measure much
798
Detection of propofol in breath samples
Fig 2 Real-time propofol and metabolite concentration measurements for the ®rst 10 min of expired breath sampling from a patient undergoing surgery. This patient showed an initial surge of propofol shortly after the operation started, corresponding to approximately 50 p.p.b. by volume levels and then steadily decreasing to a value of about 5 p.p.b. by volume within about 10 min of the operation.
smaller concentrations of propofol, certainly down to those concentrations relating to the change between awareness and loss of consciousness. Of course, the mixed expired gas we measure will not give us a propofol concentration that is the same as alveolar gas, and therefore it is not easy to directly correlate this to the plasma concentration. We plan to address this issue in further trials and establish the relation between the propofol concentration in plasma and that in the expired breath. The investigation presented in this paper is a proof of concept only, that is that propofol could be measured in expired breath using a new mass spectrometric analytical technique in real-time.
Although these tests were on a very limited scale, we did ®nd biological variation between patients. Figure 2 represents the most dramatic changes in expired breath concentrations observed, for both propofol and its metabolites. In setting the infusion device, infusion rates are determined by a built-in algorithm requiring only the patient's weight and target plasma level as inputs. Our results suggest that studies of i.v. anaesthesia with propofol using PTR-MS allow an understanding of the variations in predicted values within and between patients. Routine propofol monitoring might be feasible in an analogous manner to that for volatile anaesthetic agents. With the latter, Minimum Alveolar Concentrations (MACs) are speci®ed for adequate anaesthesia.3 Perhaps, similar MACs could be determined for propofol?
Acknowledgements We are especially grateful to the medical and nursing staff in Birmingham Women's Hospital NHS Trust for facilitating these investigations. We wish to thank Mr S. G. Arkless, School of Chemical Sciences, University of Birmingham, for his technical support before and during the clinical studies. EPSRC provided ®nancial support for this work (GR/R06489).
References
799
1 Smith D, SÏpanÏel P, Thompson JM, et al. The selected ion ¯ow tube method for workplace analyses of trace gases in air and breath: its scope, validation and applications. Appl Occup Environ Hyg 1998; 13: 817±23 2 Lindinger W, Hansel A, Jordan A. On-line monitoring of volatile organic compounds at pptv levels by means of Proton-TransferReaction Mass Spectrometry (PTR-MS). Medical applications, food control and environmental research. Int J Mass Spectrom Ion Process 1988; 173: 191±241 3 Grant WJ. Medical Gases: Their Properties and Uses. Aylesbury, Bucks, UK: HM+M Publishers Ltd, 1978; 103
