- Email: [email protected]
Biomonitoring occupational sevoflurane exposure at low levels by urinary sevoflurane and hexafluoroisopropanol
Toxicology Letters 231 (2014) 154–160
Contents lists available at ScienceDirect
Toxicology Letters journal homepage: www.elsevier.com/locate/toxlet
Biomonitoring occupational sevoflurane exposure at low levels by urinary sevoflurane and hexafluoroisopropanol Maria Luisa Scapellato a, *, Mariella Carrieri a , Isabella Maccà a , Fabiola Salamon a , Andrea Trevisan a , Maurizio Manno b , Giovanni Battista Bartolucci a a b
Department of Cardiologic, Thoracic and Vascular Sciences, University of Padova, Via Giustiniani, 2, Padova 35128, Italy Department of Public Health, University of Napoli Federico II, Via Pansini, 5, Napoli 80131, Italy
H I G H L I G H T S
Sev-U and HFIP were correlated with individual Sevoflurane exposure. Sev-U and HFIP suffered from some critical variability factors. Sev-U seems influenced by peaks of exposure, HFIP by exposure on the previous day. Further research is needed to identify the best biomarker at low exposure levels. More work is needed to identify the biological limit values for Sev-U and HFIP.
A R T I C L E I N F O
A B S T R A C T
Article history: Received 30 July 2014 Received in revised form 30 September 2014 Accepted 6 October 2014 Available online 19 October 2014
This study aimed to correlate environmental sevoflurane levels with urinary concentrations of sevoflurane (Sev-U) or its metabolite hexafluoroisopropanol (HFIP) in order to assess and discuss the main issues relating to which biomarker of sevoflurane exposure is best, and possibly suggest the corresponding biological equivalent exposure limit values. Individual sevoflurane exposure was measured in 100 healthcare operators at five hospitals in north-east Italy using the passive air sampling device Radiello1, and assaying Sev-U and HFIP concentrations in their urine collected at the end of the operating room session. All analyses were performed by gas chromatography-mass spectrometry. Environmental sevoflurane levels in the operating rooms were also monitored continuously using an infrared photoacoustic analyzer. Our results showed very low individual sevoflurane exposure levels, generally below 0.5 ppm (mean 0.116 ppm; range 0.007–0.940 ppm). Sev-U and HFIP concentrations were in the range of 0.1–17.28 mg/L and 5–550 mg/L, respectively. Both biomarkers showed a statistically significant correlation with the environmental exposure levels (Sev-U, r = 0.49; HFIP, r = 0.52), albeit showing fairly scattered values. Sev-U values seem to be influenced by peaks of exposure, especially at the end of the operating-room session, whereas HFIP levels by exposure on the previous day, the data being consistent with the biomarkers’ very different half-lives (2.8 and 19 h, respectively). According to our results, both Sev-U and HFIP are appropriate biomarkers for assessing sevoflurane exposure at low levels, although with some differences in times/patterns of exposure. More work is needed to identify the best biomarker of sevoflurane exposure and the corresponding biological equivalent exposure limit values. ã 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Sevoflurane Hexafluoroisopropanol Biomarkers Biological monitoring Operating room personnel
1. Introduction * Corresponding author. Tel.: + 39 049 8211369; fax: +39 049 8212542. E-mail addresses: [email protected] (M.L. Scapellato), [email protected] (M. Carrieri), [email protected] (I. Maccà), [email protected] (F. Salamon), [email protected] (A. Trevisan), [email protected] (M. Manno), [email protected] (G.B. Bartolucci). http://dx.doi.org/10.1016/j.toxlet.2014.10.018 0378-4274/ ã 2014 Elsevier Ireland Ltd. All rights reserved.
The airborne dispersion of volatile anaesthetics during surgical procedures gives rise to some unavoidable exposure of the operating room personnel. The development of appropriate strategies for monitoring pollution has led to a better identification of the factors capable of influencing exposure to anaesthetics,
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
i.e. structural (shape and volume of buildings), system (ventilation, gas delivery and scavenging systems, type of anaesthesia equipment), behavioral (proper use and maintenance of equipment, compliance with work procedures), and anaesthesiological factors, such as the features of the patient’s reawakening phase (Byhahn et al., 2001; Virgili et al., 2002). This has led to the implementation of more effective preventive measures, such as enforcing action in the workplace to minimize exposure of health care workers and to provide them with adequate information, training and health surveillance (Smith, 2009). Developments in anaesthesiological procedures and in the usage and features of anaesthetic gases have also contributed, over the years’ to a drastic reduction in pollution levels. Several studies on the trends of anaesthetic pollution in operating rooms in Italian hospitals and elsewhere over the last 20 years have nonetheless indicated that critical situations were and are still quite common. The picture varies considerably, sometimes even between different operating rooms in the same hospital (Barker and Abdelatti, 1997; Byhahn et al., 2001; Cottica et al., 2001; Fustinoni et al., 2012; Henderson and Matthews, 2000; Hoerauf et al., 1997; Rovesti et al., 2003; Sartini et al., 2006; Scapellato et al., 2004, 2008; Trevisan et al., 2003; Virgili et al., 2002; Zaffina et al., 2012). The anaesthetic gases most commonly used in operating rooms are N2O and several latest-generation halogenated substances that are less toxic and more manageable from the anaesthesiological standpoint than their predecessors. Among the halogenated anaesthetics, sevoflurane – a fluorinated methylethylether – has the advantage of a low blood–gas partition coefficient that enables a more precise control of anaesthesia (Eger, 1993; Robinson et al., 1999). The major international organizations have yet to recommend environmental and biological limit values for sevoflurane. So, when examining environmental monitoring data, reference is normally made to the NIOSH (1977) recommendations for halogenated anaesthetics in general (which refer to products used in 1976, such as halothane and enflurane), i.e. the ceiling limits of 2 ppm when the halogenated substances are used alone, or 0.5 ppm when they are administered in association with N2O. As for biological monitoring, the best biomarker for assessing occupational exposure to sevoflurane is still a matter of debate. Sevoflurane is metabolized in the liver by the cytochrome P450 iso-enzyme CYP2E1 (Behne et al., 1999; Kharasch and Thummel, 1993; Kharasch et al., 1995a,b), forming inorganic fluorides and hexafluoroisopropyl alcohol (HFIP), an organic metabolite that is stable, non-reactive, rapidly conjugated with glucuronic acid, and excreted in the urine (Martis et al., 1981). Glucuronoconjugation is a rather specific metabolic pathway of sevoflurane, whereas other gaseous anaesthetics do not give rise to metabolites stable enough to undergo conjugation with glucuronic acid. That is why sevoflurane metabolites were first recommended for use in biomonitoring exposure. It is common knowledge that inorganic fluorides are non-specific indicators of exposure because their concentration in blood is also influenced by a person’s eating habits and hygiene. In addition, newly-formed fluorides are only partly eliminated through the kidney and may also be incorporated in the bone matrix, so variations in their blood concentration in exposed subjects are minimal (Haufroid et al., 2000; Hoerauf et al., 1997). Several authors have suggested using HFIP as a biological marker instead, because its level in urine correlate well with environmental sevoflurane concentration at both high and low exposure levels (Barbic et al., 2003; Haufroid et al., 2000; Imbriani et al., 2001). Accorsi et al. (2003) recommended also the unmodified sevoflurane in urine (Sev-U) as a valid biomarker. The aim of the present study was to investigate the correlation between sevoflurane environmental levels and the urinary concentrations of two biological indicators, sevoflurane itself
155
and its metabolite HFIP, in subjects occupationally exposed to the anaesthetic in operating rooms, and to discuss the main issues related to the identification of the best biomarker of sevoflurane exposure and with a view to suggesting the corresponding biological equivalent exposure limit values. 2. Materials and methods Environmental and biological measures of sevoflurane exposure were obtained for 100 operating room personnel (physicians, anaesthetists, instrumentalists and nurses) at five hospitals in the Veneto region of north-eastern Italy, 2 university hospitals (A and B) and 3 smaller general hospitals, (C–E), during operating room sessions for surgical procedures requiring inhalatory general anaesthesia with sevoflurane. The characteristics of the study population, grouped by hospital, are summarized in Table 1. All subjects provided an informed consent to their inclusion in the study. 2.1. Environmental monitoring The operators’ individual environmental exposure to sevoflurane was assessed using diffusive samplers (Radiello1) placed near their breathing zone. These devices are particularly suitable for use in the operating room because they are sterile and very small in size. Each sample was collected throughout an operating room shift of at least 4 h. The Radiello1 analysis was performed using the headspace gas chromatography–mass spectrometry (HS–GC–MS) method. The instrumentation used was an autosystem XL-Turbomass gas chromatograph equipped with an automatic headspace sampler HS 40XL (PerkinElmer), a 25 m Poraplot Q capillary column, and a 2.5 m column particle trap (Chromopack). The operating conditions were as follows. The vials were heated for 60 min at 45 C and analyzed by monitoring the three masses (EI+) 33.00, 131.00, 181.00. The method was validated in accordance with the International Conference on Harmonization (ICH, 1996). The performance characteristics were as follows: – the precision (inter assay variation, CV%) on 5 replicates in the
concentration range of 0.02–17 ppm for a 4-h exposure was within the range 2–10%, – the accuracy in the same range of concentrations was within 93 and 110%, – the analytical detection limit value for a 4-h exposure was 0.02 ppm, – the instrument response was linear, in the considered range. To examine the time course of sevoflurane concentrations, the environmental air levels of the anaesthetic were also monitored continuously at a fixed position using an infrared photoacoustic analyzer (monitor type 1302, Bruel & Kjær) coupled with a multipoint sampler (multi-point sampler and doser 1303, Innova Air Tech Instruments). As a preliminary check on the comparability of the two sampling systems, in some of the operating rooms sevoflurane measurements were taken simultaneously at the same Table 1 Characteristics of the study population (total and grouped by hospital). Hospital A B C D E Total population
Female
Male
Age in years (mean SD)
54 23 5 10 8
22 11 2 7 6
32 12 3 3 2
37.9 6.8 40.7 7.7 38.9 7.9 40.1 8.9 41.3 9.6
100
48
52
39.8 7.3
No. of subjects
156
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
fixed position (for 4 h) using both, the photoacoustic analyzer and Radiello1. For this purpose, the data from the photoacoustic analyzer were calculated as the mean of values recorded during a 4-h continuous monitoring. 2.2. Biological monitoring All operators were asked to empty their bladder before starting their work shift in the operating room. At the end of the work shift (lasting at least 4 h) a urine sample was collected from each subject in an area far from the operating room to avoid sample contamination. Within 2 min from micturition, 10 mL of urine were transferred to hermetically-sealed vials (volume 22 mL) for Sev-U determination, while the remaining volume was used for HFIP analysis. All samples were stored at 20 C until analysis, which was performed within 2 weeks from sampling. Sev-U and HFIP analysis were by the headspace GC-MS method using the instrumentation described above. For Sev-U, the vials were heated for 30 min at 30 C and analyzed by monitoring the same three masses (EI+) 33.00, 131.00, 181.00 as for the environmental sevoflurane analysis. The method was validated in the concentration range of 0.1–30 mg/L, these values usually cover even the worst pollution conditions encountered in the operating room. The performance characteristics are as follows: – the precision (inter assay variation, CV%) was within the range
0.4–10.4% on 5 replicates, – the accuracy was within 95 and 105%, – the analytical detection limit value was 0.1 mg/L, – the instrument response was linear in the considered range.
The urinary HFIP was analyzed after overnight enzymatic hydrolysis with b-glucoronidase at 37 C. The vials were heated at 95 C for 60 min and the samples were analyzed by monitoring the masses 99 and 129 m/z. The method was validated in the range of concentration 1–799 mg/L. The performance characteristics are as follows: – the precision (inter assay variation, CV%) was within the range
4.7–12.4% on 5 replicates, – the accuracy was within 97–102.7%, – the analytical detection limit value was 1 mg/L, – the instrument response was linear in the considered range.
In a smaller group of operators (n = 40) coming from operating room sessions lasting more than 6 h, urinary creatinine was assayed using a colorimetric test (Sigma Diagnostics Inc. St. Louis, MO, USA) to calculate the HFIP/creatinine ratio.
2.3. Statistical analysis Statistical analysis was carried out using the Microcal Origin and the StatsDirect software. The non parametric Kruskal–Wallis test was used to compare groups of workers (by hospital and by job titles). Linear regression coefficient were calculated to analyze the correlation between environmental sevoflurane values obtained concurrently using both the photoacustic analyzer and Radiello1. A non parametric correlation was also applied for confirmative purposes (Spearman’s test). The normality of data distribution was assessed by the Shapiro– Wilk test. Non-normally distributed environmental and biological monitoring values were analyzed by parametric tests after logarithmic transformation. Pearson’s coefficients were used for correlations between variables. In all tests, a p value lower than 0.05 was considered statistically significant. 3. Results The preliminary study on the comparability of the two sampling methods revealed a good correlation between 15 concurrent sevoflurane measurements carried out at the same fixed positions by photoacoustic analyzer and Radiello1 (r = 0.96, p < 0.0001). This result was also obtained by applying the non-parametric Spearman’s test for correlation analysis (Rho 0.77, p = 0.0013), thus confirming the reliability of the two sampling methods even when used for the environmental monitoring of sevoflurane at low levels. Table 2 shows the means, standard deviations, medians and ranges of the environmental and biological data recorded in the operating-room personnel (total and grouped by hospital). The individual concentrations recorded in the operators’ breathing zone were generally low. In all 100 samples, environmental sevoflurane levels were below the NIOSH limit of 2 ppm and only 9 samples exceeded 0.5 ppm. The urinary levels of both biomarkers were consistent with the environmental data and confirmed a low level of exposure, albeit with a wider range of concentrations. The same data analyzed by single hospital showed that the differences between the medians of operating room personnel exposure measured in the different hospitals were statistically significant (Kruskal–Wallis test). The individual airborne concentrations measured in the hospitals A, B and E showed a statistically significant higher exposure than those of the hospital D. The median values of the two biomarkers confirmed a statistically significant different exposure among the operators of different hospitals (see Table 2). Table 3 shows the data grouped by job title. The results of the Kruskal–Wallis test showed that the differences between the medians of different groups of subjects were statistically
Table 2 Individual sevoflurane levels measured with Radiello1, and urinary concentrations of sevoflurane and HFIP in operating room personnel (total and grouped by hospital). The different hospitals were named A, B, C, D, E. Hospital
A B C D E Total population
No. of subjects
54 23 5 10 8 100
Sevoflurane in urine (mg/L)
HFIP (mg/L)
Mean SD (median)
Range
Mean SD (median)
Range
Mean SD (median)
Range
0.097 0.174 (0.028)a 0.250 0.319 (0.107)a 0.038 0.034 (0.028) 0.0132 0.003 (0.012)a 0.030 0.012 (0.026)a
<0.010–0.912 0.010–0.940 0.015–0.096 <0.010–0.018 0.020–0.049
0.58 0.38 5.29 4.96 0.33 0.38 0.56 0.38 0.56 0.43
0.20–2.00 1.20–17.28 0.10–0.24 0.20–1.20 0.20–1.49
48.9 69.5 (27.3)a 133.2 150.6 (55.0)a 14.6 5.1 (13.0)a 16.7 16.2 (17.5)a 122.6 110.5 (110.0)a
1.0–360.0 1.0–550.0 10.0–20.0 1.0–50.0 1.0–290.0
0.116 0.213 (0.026)
<0.010–0.940
1.64 3.09 (0.62)
<0.10–17.28
69.2 101.7 (26.0)
<1.0–550.0
Sevoflurane in air (ppm)
(0.50)a (2.80)a (0.20)a (0.48)a (0.56)a
Sevoflurane in air: Overall, p value = 0.0085 (A vs D, p = 0.0032; B vs D, p = 0.0002; D vs E, p = 0.003). Sevoflurane in urine: Overall, p value < 0.0001 (A vs B, p < 0.0001; A vs C, p = 0.035; B vs C, p < 0.0001; B vs D, p < 0.0001; B vs E, p < 0.0001). HFIP: Overall, p value = 0.0023 (A vs B, p = 0.022; A vs C, p = 0.045; A vs D, p = 0.041; B vs C, p = 0.0026; B vs D, p = 0.0009; C vs E, p = 0.0131; D vs E, p = 0.0126). a Kruskal–Wallis test.
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
157
Table 3 Airborne sevoflurane levels and urinary concentrations of sevoflurane and HFIP in operating room personnel by job title. Job title
Anesthetists n = 19 Surgeons n = 25 Instrumentalists n = 14 Anesthetist nurses n = 9 Operating room nurses n = 33
Sev-U (mg/L)
Sevoflurane in air (ppm)
HFIP (mg/L)
Mean SD (median)
Range
Mean SD (median)
Range
Mean SD (median)
Range
0.290 0.353 (0.096)a 0.124 0.226 (0.016)a 0.116 0.218 (0.020) 0.093 0.070 (0.089) 0.095 0.159 (0.026)a
0.013–0.940 <0.010–0.777 <0.010–0.734 <0.010–0.189 0.010–0.558
2.18 3.49 (0.70) 2.26 3.35 (0.82) 2.59 4.80 (0.78) 0.87 0.52 (0.90) 2.39 4.41 (0.90)
0.20–13.21 0.20–12.72 <0.10–16.30 0.20–1.60 0.20–17.28
85.5 100.1 (49.4) 82.9 134.4 (35.8) 75.1 127.2 (23.0) 65.4 79.5 (26.00) 62.5 76.3 (31.7)
1.0–344.1 1.0–550.0 10.0–450.0 20.0–240.0 <1.0–217.1
Overall: p value = 0.0301 (Anesthetists vs Instrumentalists, p = 0.0019; Anesthetists vs Surgeons, p = 0.002; Anesthetists vs Operating room nurses, p = 0.0297). a Kruskal–Wallis test.
significant indicating that job title was an important variable influencing exposure to the anaesthetic. Anaesthetists revealed a statistically higher exposure to sevoflurane than surgeons, instrumentalists or operating room nurses, but not in comparison with anaesthetist nurses. There were no statistically significant differences between the groups of workers with regard to the levels of either biomarkers, however. Fig. 1A shows the correlation between individual environmental exposure data and urinary excretion of sevoflurane (log-transformed data). Although there were some very scattered points, where quite low environmental concentrations corresponded to very high biological values and vice versa, a statistically significant correlation emerged overall (r = 0.49, p < 0.0001). Moreover, all the high Sev-U concentrations concerned subjects working in the same operating room, where the photoacoustic
Fig. 1. Correlations between individual sevoflurane levels measured with Radiello1 and urinary concentrations of sevoflurane (A) or HFIP (B) at the end of the work shift in all subjects.
analyzer revealed very high peak levels of sevoflurane – even exceeding 20 ppm – towards the end of the operating room session, which was due to a leak in the supply circuits (data not shown) while during most of the session sevoflurane concentrations were very low. Taking the short half-life of sevoflurane (2.8 h) into account, even these short peaks could be responsible, therefore, for the high Sev-U values detected in these subjects. Fig. 1B shows the correlation between individual environmental sevoflurane concentrations and urinary HFIP levels (log-transformed data), with a regression coefficient of 0.52 (p < 0.0001). Here again, the scattering of the data may be partly explained by the substance’s relatively long half-life (19 h): subjects with high urinary levels of the metabolite and very low levels of environmental exposure may have been highly exposed on the previous day. Unfortunately, we do not have environmental data for the operating room sessions performed on the previous days. The correlation between airborne sevoflurane values and urinary HFIP levels did not improve significantly after adjusting for creatinine (data not shown). A statistically significant correlation was found between the two biomarkers (log-transformed data, r = 0.273, p = 0.006), albeit showing a marked variability in the values (Fig. 2). In order to verify the reliability of the two biomarkers also in the case of very low exposure, the data were divided into two subgroups according to the median value, equal to 0.026 ppm. Table 4 shows the linear relationships between log-log transformed values of airborne sevoflurane and urinary biomarkers (Sev-U and HFIP) in all subjects (see also Fig. 1) and in subjects with sevoflurane exposure below and above 0.026 ppm. The regression equations for subgroups with exposure above 0.026 ppm is similar to that obtained considering all the 100 subjects and the correlations were still statistically significant. In the groups with exposure below 0.026 ppm, the correlation between the
Fig. 2. Correlation between urinary concentrations of sevoflurane and HFIP measured at the end of the work shift in all subjects.
158
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
Table 4 Linear relationships between log-log transformed values of airborne sevoflurane (Sev-A, ppm) and urinary biomarkers (Sev-U and HFIP, mg/L): all subjects (1), subjects with sevoflurane exposure <0.026 ppm (2), subjects with sevoflurane exposure 0.026 ppp (3). Equation log log log log log log
Sev-U = 0.417 log Sev-A + 0.444 (1) HFIP = 0.640 log Sev-A + 2.310 (1) Sev-U = 0.576 log Sev-A + 0.837 (2) HFIP = 2.072 log Sev-A + 4.995 (2) Sev-U = 0.779 log Sev-A + 0.689 (3) HFIP = 0.606 log Sev-A + 2.255 (3)
No. of subjects
r
p
100 100 50 50 50 50
0.49 0.52 0.22 0.40 0.65 0.51
<0.0001 <0.0001 n.s. 0.0039 <0.0001 0.0001
n.s., not statistically significant.
environmental data and the HFIP values was statistically significant but quantitatively different, probably due to the values under the analytical detection limit. As for Sev-U, the correlation was not statistically significant. It should be underlined that at these very low exposure levels, the influence of other factors, such as intraand interindividual biological variability, can be more relevant. No statistically significant differences were found, instead, in the correlations between airborne sevoflurane exposure and biological monitoring data among the different groups of professionals (data not shown). 4. Discussion In the present study, environmental sevoflurane pollution in the hospitals monitored was very low, with average concentrations of the anaesthetic generally lower than those reported in the literature (Accorsi et al., 2003, 2005; Barbic et al., 2003; Haufroid et al., 2000; Imbriani et al., 2001). Pollution levels, though, were very variable and could differ among different hospitals and even between operating rooms at the same hospital. As already reported in other studies (Accorsi et al., 2005; Tankó et al., 2009; Trevisan and Gori, 1990; Virgili et al., 2002), we found that anaesthetists were more exposed to sevoflurane than other professionals, due to their prolonged presence near both the anaesthesiological equipment and the patient. The use of different sampling methods, both reliable, enabled us to obtain detailed information on both environmental pollution and health care workers’ individual exposure. Continuous monitoring with the photoacoustic analyzer identified environmental pollution peaks, while individual sampling by Radiello1 measured the total individual exposure of different professionals. Our study also identified a statistically significant correlation between individual environmental exposure and biomarkers’ levels. In both cases, however, there were some scattered points, where quite low environmental concentrations corresponded to very high biological values. This may be partly explained by the different half-lives of the two biomarkers. As mentioned earlier, HFIP levels, because of the long half-life (19 h), may also reflect exposure on the previous day. Some authors (Barbic et al., 2003; Hoerauf et al., 1997) found appreciable levels of this metabolite in the urine of many subjects even at the beginning of the work shift. Some operators who had been exposed to sevoflurane during the previous working day already had high HFIP values in their urine collected at the start of their operating room shift, showing that 24 h were not enough for a substantial, let alone complete, urinary HFIP elimination. Moreover, unlike Barbic et al. (2003), but in agreement with Haufroid et al. (2000) and Imbriani et al. (2001), we found that adjusting for creatinine did not improve the correlation between environmental sevoflurane data and urinary HFIP values. As for Sev-U, our findings suggest that this biomarker is particularly influenced by peaks of exposure (as detected by the photoacoustic analyzer), especially towards the end of the operating room session, probably because of the substance’s very
short half-life (2.8 h). This explanation is supported by the data of Fig. 2, showing a statistically significant correlation between the two biomarkers, but with a wide scattering of the values. The scattering can also be partly explained by interindividual biological variability, which is bound to be higher at very low sevoflurane exposures, as found indeed in our study (always below 1 ppm), and also by the presence of a cluster of subjects with low Sev-U levels but relatively high HFIP values, possibly due to residual exposure from the previous day. Based on our results, the calculated urinary concentrations of biomarker corresponding to an environmental sevoflurane exposure of 2 and 0.5 ppm were 318 and 131 mg/L, respectively, for HFIP (lower 95% CL: 133 and 72 mg/L, respectively) and 3.7 and 2.1 mg/L, respectively, for Sev-U (lower 95% CL: 2.0 and 1.4 mg/L, respectively). The values for HFIP are quite similar to those calculated by Imbriani et al. (2001), but very different from those proposed by Haufroid et al. (2000),Barbic et al. (2003) and Accorsi et al. (2005). As for Sev-U, our calculated biological equivalent exposure values are similar to those proposed by Accorsi et al. (2005), though, when we omitted subjects showing significant peak levels of sevoflurane exposure at the end of an operating session (as reported in the Section 3), our values became remarkably lower (1.9 and 1.3 mg/L for an environmental sevoflurane exposure of 2 and 0.5 ppm, respectively). So, considering the present biological monitoring data overall, both biomarkers, HFIP and Sev-U, have confirmed some critical features reported in the literature. Urinary HFIP has been suggested by several authors as a specific indicator of occupational exposure to sevoflurane for which the biological equivalent level corresponding to a given airborne concentration could be derived (Barbic et al., 2003; Haufroid et al., 2000; Imbriani et al., 2001). Values obtained from calculation by different authors, though, were rather different. Imbriani et al. (2001) for instance calculated a biological equivalent concentrations for HFIP of 488 and 160 mg/L, corresponding to the NIOSH limits of exposure to airborne sevoflurane of 2 and 0.5 ppm, respectively. Haufroid et al. (2000) extrapolated an intermediate biological index of 800 mg/L HFIP, corresponding to a 2 ppm concentration of sevoflurane, while Barbic et al. (2003) calculated a value of 2772.7 mg/L HFIP for a 2 ppm sevoflurane exposure. It should be emphasized, however, that the subjects of our study showed, on average, a much lower exposure levels than those of other studies (Accorsi et al., 2005; Barbic et al., 2003; Haufroid et al., 2000; Imbriani et al., 2001). For example, all our exposure data are similar to the subgroup of subjects identified by Imbriani et al. (2001) as “low exposed”. Accorsi et al. (2005) assessed both biomarkers considered in our study, and reported a statistically significant correlation with the environmental levels of sevoflurane for both, though HFIP data showed a less linear correlation with individual exposure, and a rather high intercept, than those of Sev-U. This suggested some degree of accumulation of the metabolite during the course of the workweek, although the authors were unable to find any differences in excretion on different weekdays, or between data
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
obtained in the early and the late parts of the two weeks of their survey. Taking the half-life of the two biomarkers into account (2.8 and 19 h for sevoflurane and HFIP, respectively), the authors suggested that end-of-work-shift urinary sevoflurane could be a biomarker reflecting (only) very recent exposure (i.e. from the same work shift), while HFIP could be a more stable exposure tracer, also reflecting various conditions capable of influencing in vivo sevoflurane metabolism, such as cigarette smoking, alcohol consumption and genetic metabolic polymorphisms. Since the enzyme mainly responsible for HFIP production is CYP2E1, and smoking and alcohol can be important modulators of its activity, the authors studied the influence of these variables. They indeed found that smoking habits were an important confounding factor for urinary HFIP but not for Sev-U. This is one of the reasons why they recommended using Sev-U rather than HFIP for the biological monitoring of sevoflurane exposure. When they calculated, for both biomarkers, the biological equivalent limits of exposure corresponding to the airborne sevoflurane concentrations of 2 and 0.5 ppm. they found for HFIP, but not for Sev-U, higher values than those found in the present study: 2660 and 820 mg/L for HFIP (lower 95% CL, 1610 and 630 mg/L, respectively) and 3.9 and 1.4 mg/L for Sev-U (lower 95% CL: 2.8 and 1.2 mg/L, respectively), respectively. In conclusion, both biomarkers used in this study to assess occupational sevoflurane exposure were found to be correlated with the levels of the anaesthetic measured in the air, although they both suffered from some critical variability factors that we were unable to fully elucidate. The current state of knowledge suggests that both biomarkers can be used for group biological monitoring purposes, providing that standardized procedures for urine sample collection are adopted and adequate information about the individual’s lifestyle, or other habits that are capable of influencing HFIP metabolism, is available. As for HFIP, given the relatively long half-life, its use might be extended to assess also the exposure of workers who were (only) involved in the morning shift on the previous day. As for Sev-U, instead, its measurement should only be used for monitoring exposure occurring during the work shift on same day. Besides, particular attention should be paid to urine sample collection and storage, as samples require to be sealed carefully within one or two minutes of collection to prevent any loss of the anaesthetic, and to detecting exposure peaks, especially if they occurred in the latter part of the operating room session. A continuous monitoring of airborne anaesthetics’ levels would be most helpful in this regard. Finally, further research is needed to reach consensus on which biomarker is best for the routine biomonitoring of sevoflurane exposure in the workplace and to establish definite biomarker’s levels to be recommended as biological equivalent exposure limit values for routine use in occupational health programmes. Conflict of interest The authors declare that there are no conflicts of interest. Transparency document The Transparency document associated with this article can be found in the online version. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.toxlet.2014.10.018.
159
References Accorsi, A., Valenti, S., Barbieri, A., Raffi, G.B., Violante, F.S., 2003. Proposal for single and mixture biological exposure limits for sevoflurane and nitrous oxide at low occupational exposure levels. Int. Arch. Occup. Environ. Health 76, 129–136. Accorsi, A., Morrone, B., Domenichini, I., Valenti, S., Raffi, G.B., Violante, F.S., 2005. Urinary sevoflurane and hexafluoro-isopropanol as biomarkers of low-level occupational exposure to sevoflurane. Int. Arch. Occup. Environ. Health 78, 369–378. Barbic, F., Bagnati, M., Basile, M., Zanoli, V., Carrettoni, L., Cassani, C., Porta, P., Fortina, A., 2003. L’esafluoroisopropanolo urinario nella valutazione dell’esposizione professionale a sevoflurano: aspetti metodologici e punti critici. G. Ital. Med. Lav. Ergon. 25, 95–97. Barker, J.P., Abdelatti, M.O., 1997. Anaesthetic pollution. Potential sources, their identification and control. Anaesthesia 52, 1077–1083. Behne, M., Wilke, H.J., Harder, S., 1999. Clinical pharmacokinetics of sevoflurane. Clin. Pharmacokinet. 36, 13–26. Byhahn, C., Wilke, H.J., Westpphal, K., 2001. Occupational exposure to volatile anaesthetics: epidemiology and approaches to reducing the problem. CNS Drugs 15, 197–215. Cottica, D., Bartolucci, G.B., Grignani, E., Locatelli, C., Sala, C., Scapellato, M.L., Sesana, G., 2001. Gas anestetici. In: Minoia, C., Perbellini, L. (Eds.), Monitoraggio Ambientale e Biologico Dell’Esposizione Professionale a Xenobiotici, Vol. 4. Morgan Edizioni Tecniche, Milano. Eger, E.I., 1993. New inhalation agents – desflurane and sevoflurane. Can. J. Anaesth. 40, 3–8. Fustinoni, S., Pasini, R., Strangi, F., Valla, C., Missineo, P., Margonari, M., Denaro, R., Bertazzi, P.A., 2012. Air and biomonitoring of occupational exposure to anesthetic gases in the health care workers of a large hospital in Milan. G. Ital. Med. Lav. Ergon. 34 (Suppl. 3), 278–279 (Article in Italian). Haufroid, V., Cardinal, S., Licot, C., Villalpando, M.L., Van Obbergh, L., Clippe, A., Lison, D., 2000. Biological monitoring of exposure to sevoflurane in operating room personnel by the measurement of hexafluoroisopropanol and fluoride in urine. Biomarkers 5, 141–151. Henderson, K.A., Matthews, I.P., 2000. Staff exposure to anaesthetic gases in theatre and non theatre areas. Eur. J. Anaesthesiol. 17, 149–151. Hoerauf, K.H., Koller, C., Taeger, K., Hobbhahn, J., 1997. Occupational exposure to sevoflurane and nitrous oxide in operating room personnel. Int. Arch. Occup. Environ. Health 69, 134–138. The International Conference on Harmonization (ICH), 1996. Validation of Analytical Procedure: Methodology (Q 2B), Food and Drug Administration, USA. ICH, Geneva, Switzerland. Imbriani, M., Zadra, P., Negri, S., Alessio, A., Maestri, L., Ghittori, S., 2001. Biological monitoring of occupational exposure to sevoflurane. Med. Lav. 92, 173–180 (Article in Italian). Kharasch, E.D., Thummel, K.E., 1993. Identification of cytochrome P450 2E1 as the predominant enzyme catalyzing human liver microsomal defluorination of sevoflurane, isoflurane and methoxyflurane. Anesthesiology 79, 795–807. Kharasch, E.D., Karol, M.D., Lanni, C., Sawchuk, R., 1995a. Clinical sevoflurane metabolism and disposition: I. Sevoflurane and metabolite pharmacokinetics. Anesthesiology 82, 1369–1378. Kharasch, E.D., Armstrong, A.S., Gunn, K., Artru, A., Cox, K., Karol, M.D., 1995b. Clinical sevoflurane metabolism and disposition: II. The role of cytochrome P450 2E1 in fluoride and hexafluoroisopropanol formation. Anesthesiology 82, 1379–1388. Martis, L., Lynch, S., Napoli, M., Woods, E.D., 1981. Biotransformation of sevoflurane in dogs and rats. Anaesth. Analg. 60, 186–191. NIOSH, 1977. Criteria for a Recommended Standard: Occupational Exposure to Waste Anesthetic Gases and Vapors. Publication no. 77–140. US DHEW, Cincinnati, Ohio. Robinson, B.J., Uhrich, T.D., Ebert, T.J., 1999. A review of recovery from sevoflurane anaesthesia: comparisons with isoflurane and propafol including metaanalysis. Acta Anaesthesiol. Scand. 43, 185–190. Rovesti, S., Ferrari, A., Faggiano, D., Vivoli, G., 2003. Monitoring occupational exposure to volatile anaesthetics in the operating theatre: environmental and biological measurements. Ann. Ig 17, 219–230 (Article in Italian). Sartini, M., Otaria, G., Dall’era, M., Spagnolo, A.M., Cristina, M.L., 2006. Nitrous oxide pollution in operating theatres in relation to the type of leakage and the number of efficacious air exchanges per hour. J. Prev. Med. Hyg. 47, 155–159. Scapellato, M.L., Trevisan, A., Carrieri, M., Maccà, I., Bonfiglio, E., Gori, G., Serraino, S., Bartolucci, G.B., 2004. Dose and effect biomarkers in sevoflurane exposure. G. Ital. Med. Lav. Ergon. 26, 99–101 (Article in Italian). Scapellato, M.L., Mastrangelo, G., Fedeli, U., Carrieri, M., Maccà, I., Scoizzato, L., Bartolucci, G.B., 2008. A longitudinal study for investigating the exposure level of anesthetics that impairs neurobehavioral performance. NeuroToxicology 29, 116–123. Smith, F.D., 2009. Management of exposure to waste anesthetic gases. AORN J. 91, 482–494. Tankó, B., Molnár, C., Budi, T., Peto, C., Novák, L., Fülesdi, B., 2009. The relative exposure of the operating room staff to sevoflurane during intracerebral surgery. Anesth. Analg. 109, 1187–1192. Trevisan, A., Borella Venturini, M., Carrieri, M., Giraldo, M., Maccà, I., Perini, M., Scapellato, M.L., Virgili, A., Bartolucci, G.B., 2003. Biological indices of kidney involvement in personnel exposed to sevoflurane in surgical areas. Am. J. Ind. Med. 44, 474–480.
160
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
Trevisan, A., Gori, G., 1990. Biological monitoring of nitrous oxide exposure in surgical areas. Am. J. Ind. Med. 17, 357–362. Virgili, A., Scapellato, M.L., Maccà, I., Perini, M., Carrieri, M., Gori, G., Saia, B., Bartolucci, G.B., 2002. Occupational exposure to anaesthetic gases in some hospitals in the Veneto Region. G. Ital. Med. Lav. Ergon. 24, 447–450 (Article in Italian).
Zaffina, S., Camisa, V., Poscia, A., Tucci, M.G., Montaldi, V., Cerabona, V., Wachocka, M., Moscato, U., 2012. Occupational exposure to sevoflurane in pediatric operating rooms: the multi-point sampling method for risk assessment. G. Ital. Med. Lav. Ergon. 34 (Suppl. 3), 266–268 (Article in Italian).
Contents lists available at ScienceDirect
Toxicology Letters journal homepage: www.elsevier.com/locate/toxlet
Biomonitoring occupational sevoflurane exposure at low levels by urinary sevoflurane and hexafluoroisopropanol Maria Luisa Scapellato a, *, Mariella Carrieri a , Isabella Maccà a , Fabiola Salamon a , Andrea Trevisan a , Maurizio Manno b , Giovanni Battista Bartolucci a a b
Department of Cardiologic, Thoracic and Vascular Sciences, University of Padova, Via Giustiniani, 2, Padova 35128, Italy Department of Public Health, University of Napoli Federico II, Via Pansini, 5, Napoli 80131, Italy
H I G H L I G H T S
Sev-U and HFIP were correlated with individual Sevoflurane exposure. Sev-U and HFIP suffered from some critical variability factors. Sev-U seems influenced by peaks of exposure, HFIP by exposure on the previous day. Further research is needed to identify the best biomarker at low exposure levels. More work is needed to identify the biological limit values for Sev-U and HFIP.
A R T I C L E I N F O
A B S T R A C T
Article history: Received 30 July 2014 Received in revised form 30 September 2014 Accepted 6 October 2014 Available online 19 October 2014
This study aimed to correlate environmental sevoflurane levels with urinary concentrations of sevoflurane (Sev-U) or its metabolite hexafluoroisopropanol (HFIP) in order to assess and discuss the main issues relating to which biomarker of sevoflurane exposure is best, and possibly suggest the corresponding biological equivalent exposure limit values. Individual sevoflurane exposure was measured in 100 healthcare operators at five hospitals in north-east Italy using the passive air sampling device Radiello1, and assaying Sev-U and HFIP concentrations in their urine collected at the end of the operating room session. All analyses were performed by gas chromatography-mass spectrometry. Environmental sevoflurane levels in the operating rooms were also monitored continuously using an infrared photoacoustic analyzer. Our results showed very low individual sevoflurane exposure levels, generally below 0.5 ppm (mean 0.116 ppm; range 0.007–0.940 ppm). Sev-U and HFIP concentrations were in the range of 0.1–17.28 mg/L and 5–550 mg/L, respectively. Both biomarkers showed a statistically significant correlation with the environmental exposure levels (Sev-U, r = 0.49; HFIP, r = 0.52), albeit showing fairly scattered values. Sev-U values seem to be influenced by peaks of exposure, especially at the end of the operating-room session, whereas HFIP levels by exposure on the previous day, the data being consistent with the biomarkers’ very different half-lives (2.8 and 19 h, respectively). According to our results, both Sev-U and HFIP are appropriate biomarkers for assessing sevoflurane exposure at low levels, although with some differences in times/patterns of exposure. More work is needed to identify the best biomarker of sevoflurane exposure and the corresponding biological equivalent exposure limit values. ã 2014 Elsevier Ireland Ltd. All rights reserved.
Keywords: Sevoflurane Hexafluoroisopropanol Biomarkers Biological monitoring Operating room personnel
1. Introduction * Corresponding author. Tel.: + 39 049 8211369; fax: +39 049 8212542. E-mail addresses: [email protected] (M.L. Scapellato), [email protected] (M. Carrieri), [email protected] (I. Maccà), [email protected] (F. Salamon), [email protected] (A. Trevisan), [email protected] (M. Manno), [email protected] (G.B. Bartolucci). http://dx.doi.org/10.1016/j.toxlet.2014.10.018 0378-4274/ ã 2014 Elsevier Ireland Ltd. All rights reserved.
The airborne dispersion of volatile anaesthetics during surgical procedures gives rise to some unavoidable exposure of the operating room personnel. The development of appropriate strategies for monitoring pollution has led to a better identification of the factors capable of influencing exposure to anaesthetics,
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
i.e. structural (shape and volume of buildings), system (ventilation, gas delivery and scavenging systems, type of anaesthesia equipment), behavioral (proper use and maintenance of equipment, compliance with work procedures), and anaesthesiological factors, such as the features of the patient’s reawakening phase (Byhahn et al., 2001; Virgili et al., 2002). This has led to the implementation of more effective preventive measures, such as enforcing action in the workplace to minimize exposure of health care workers and to provide them with adequate information, training and health surveillance (Smith, 2009). Developments in anaesthesiological procedures and in the usage and features of anaesthetic gases have also contributed, over the years’ to a drastic reduction in pollution levels. Several studies on the trends of anaesthetic pollution in operating rooms in Italian hospitals and elsewhere over the last 20 years have nonetheless indicated that critical situations were and are still quite common. The picture varies considerably, sometimes even between different operating rooms in the same hospital (Barker and Abdelatti, 1997; Byhahn et al., 2001; Cottica et al., 2001; Fustinoni et al., 2012; Henderson and Matthews, 2000; Hoerauf et al., 1997; Rovesti et al., 2003; Sartini et al., 2006; Scapellato et al., 2004, 2008; Trevisan et al., 2003; Virgili et al., 2002; Zaffina et al., 2012). The anaesthetic gases most commonly used in operating rooms are N2O and several latest-generation halogenated substances that are less toxic and more manageable from the anaesthesiological standpoint than their predecessors. Among the halogenated anaesthetics, sevoflurane – a fluorinated methylethylether – has the advantage of a low blood–gas partition coefficient that enables a more precise control of anaesthesia (Eger, 1993; Robinson et al., 1999). The major international organizations have yet to recommend environmental and biological limit values for sevoflurane. So, when examining environmental monitoring data, reference is normally made to the NIOSH (1977) recommendations for halogenated anaesthetics in general (which refer to products used in 1976, such as halothane and enflurane), i.e. the ceiling limits of 2 ppm when the halogenated substances are used alone, or 0.5 ppm when they are administered in association with N2O. As for biological monitoring, the best biomarker for assessing occupational exposure to sevoflurane is still a matter of debate. Sevoflurane is metabolized in the liver by the cytochrome P450 iso-enzyme CYP2E1 (Behne et al., 1999; Kharasch and Thummel, 1993; Kharasch et al., 1995a,b), forming inorganic fluorides and hexafluoroisopropyl alcohol (HFIP), an organic metabolite that is stable, non-reactive, rapidly conjugated with glucuronic acid, and excreted in the urine (Martis et al., 1981). Glucuronoconjugation is a rather specific metabolic pathway of sevoflurane, whereas other gaseous anaesthetics do not give rise to metabolites stable enough to undergo conjugation with glucuronic acid. That is why sevoflurane metabolites were first recommended for use in biomonitoring exposure. It is common knowledge that inorganic fluorides are non-specific indicators of exposure because their concentration in blood is also influenced by a person’s eating habits and hygiene. In addition, newly-formed fluorides are only partly eliminated through the kidney and may also be incorporated in the bone matrix, so variations in their blood concentration in exposed subjects are minimal (Haufroid et al., 2000; Hoerauf et al., 1997). Several authors have suggested using HFIP as a biological marker instead, because its level in urine correlate well with environmental sevoflurane concentration at both high and low exposure levels (Barbic et al., 2003; Haufroid et al., 2000; Imbriani et al., 2001). Accorsi et al. (2003) recommended also the unmodified sevoflurane in urine (Sev-U) as a valid biomarker. The aim of the present study was to investigate the correlation between sevoflurane environmental levels and the urinary concentrations of two biological indicators, sevoflurane itself
155
and its metabolite HFIP, in subjects occupationally exposed to the anaesthetic in operating rooms, and to discuss the main issues related to the identification of the best biomarker of sevoflurane exposure and with a view to suggesting the corresponding biological equivalent exposure limit values. 2. Materials and methods Environmental and biological measures of sevoflurane exposure were obtained for 100 operating room personnel (physicians, anaesthetists, instrumentalists and nurses) at five hospitals in the Veneto region of north-eastern Italy, 2 university hospitals (A and B) and 3 smaller general hospitals, (C–E), during operating room sessions for surgical procedures requiring inhalatory general anaesthesia with sevoflurane. The characteristics of the study population, grouped by hospital, are summarized in Table 1. All subjects provided an informed consent to their inclusion in the study. 2.1. Environmental monitoring The operators’ individual environmental exposure to sevoflurane was assessed using diffusive samplers (Radiello1) placed near their breathing zone. These devices are particularly suitable for use in the operating room because they are sterile and very small in size. Each sample was collected throughout an operating room shift of at least 4 h. The Radiello1 analysis was performed using the headspace gas chromatography–mass spectrometry (HS–GC–MS) method. The instrumentation used was an autosystem XL-Turbomass gas chromatograph equipped with an automatic headspace sampler HS 40XL (PerkinElmer), a 25 m Poraplot Q capillary column, and a 2.5 m column particle trap (Chromopack). The operating conditions were as follows. The vials were heated for 60 min at 45 C and analyzed by monitoring the three masses (EI+) 33.00, 131.00, 181.00. The method was validated in accordance with the International Conference on Harmonization (ICH, 1996). The performance characteristics were as follows: – the precision (inter assay variation, CV%) on 5 replicates in the
concentration range of 0.02–17 ppm for a 4-h exposure was within the range 2–10%, – the accuracy in the same range of concentrations was within 93 and 110%, – the analytical detection limit value for a 4-h exposure was 0.02 ppm, – the instrument response was linear, in the considered range. To examine the time course of sevoflurane concentrations, the environmental air levels of the anaesthetic were also monitored continuously at a fixed position using an infrared photoacoustic analyzer (monitor type 1302, Bruel & Kjær) coupled with a multipoint sampler (multi-point sampler and doser 1303, Innova Air Tech Instruments). As a preliminary check on the comparability of the two sampling systems, in some of the operating rooms sevoflurane measurements were taken simultaneously at the same Table 1 Characteristics of the study population (total and grouped by hospital). Hospital A B C D E Total population
Female
Male
Age in years (mean SD)
54 23 5 10 8
22 11 2 7 6
32 12 3 3 2
37.9 6.8 40.7 7.7 38.9 7.9 40.1 8.9 41.3 9.6
100
48
52
39.8 7.3
No. of subjects
156
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
fixed position (for 4 h) using both, the photoacoustic analyzer and Radiello1. For this purpose, the data from the photoacoustic analyzer were calculated as the mean of values recorded during a 4-h continuous monitoring. 2.2. Biological monitoring All operators were asked to empty their bladder before starting their work shift in the operating room. At the end of the work shift (lasting at least 4 h) a urine sample was collected from each subject in an area far from the operating room to avoid sample contamination. Within 2 min from micturition, 10 mL of urine were transferred to hermetically-sealed vials (volume 22 mL) for Sev-U determination, while the remaining volume was used for HFIP analysis. All samples were stored at 20 C until analysis, which was performed within 2 weeks from sampling. Sev-U and HFIP analysis were by the headspace GC-MS method using the instrumentation described above. For Sev-U, the vials were heated for 30 min at 30 C and analyzed by monitoring the same three masses (EI+) 33.00, 131.00, 181.00 as for the environmental sevoflurane analysis. The method was validated in the concentration range of 0.1–30 mg/L, these values usually cover even the worst pollution conditions encountered in the operating room. The performance characteristics are as follows: – the precision (inter assay variation, CV%) was within the range
0.4–10.4% on 5 replicates, – the accuracy was within 95 and 105%, – the analytical detection limit value was 0.1 mg/L, – the instrument response was linear in the considered range.
The urinary HFIP was analyzed after overnight enzymatic hydrolysis with b-glucoronidase at 37 C. The vials were heated at 95 C for 60 min and the samples were analyzed by monitoring the masses 99 and 129 m/z. The method was validated in the range of concentration 1–799 mg/L. The performance characteristics are as follows: – the precision (inter assay variation, CV%) was within the range
4.7–12.4% on 5 replicates, – the accuracy was within 97–102.7%, – the analytical detection limit value was 1 mg/L, – the instrument response was linear in the considered range.
In a smaller group of operators (n = 40) coming from operating room sessions lasting more than 6 h, urinary creatinine was assayed using a colorimetric test (Sigma Diagnostics Inc. St. Louis, MO, USA) to calculate the HFIP/creatinine ratio.
2.3. Statistical analysis Statistical analysis was carried out using the Microcal Origin and the StatsDirect software. The non parametric Kruskal–Wallis test was used to compare groups of workers (by hospital and by job titles). Linear regression coefficient were calculated to analyze the correlation between environmental sevoflurane values obtained concurrently using both the photoacustic analyzer and Radiello1. A non parametric correlation was also applied for confirmative purposes (Spearman’s test). The normality of data distribution was assessed by the Shapiro– Wilk test. Non-normally distributed environmental and biological monitoring values were analyzed by parametric tests after logarithmic transformation. Pearson’s coefficients were used for correlations between variables. In all tests, a p value lower than 0.05 was considered statistically significant. 3. Results The preliminary study on the comparability of the two sampling methods revealed a good correlation between 15 concurrent sevoflurane measurements carried out at the same fixed positions by photoacoustic analyzer and Radiello1 (r = 0.96, p < 0.0001). This result was also obtained by applying the non-parametric Spearman’s test for correlation analysis (Rho 0.77, p = 0.0013), thus confirming the reliability of the two sampling methods even when used for the environmental monitoring of sevoflurane at low levels. Table 2 shows the means, standard deviations, medians and ranges of the environmental and biological data recorded in the operating-room personnel (total and grouped by hospital). The individual concentrations recorded in the operators’ breathing zone were generally low. In all 100 samples, environmental sevoflurane levels were below the NIOSH limit of 2 ppm and only 9 samples exceeded 0.5 ppm. The urinary levels of both biomarkers were consistent with the environmental data and confirmed a low level of exposure, albeit with a wider range of concentrations. The same data analyzed by single hospital showed that the differences between the medians of operating room personnel exposure measured in the different hospitals were statistically significant (Kruskal–Wallis test). The individual airborne concentrations measured in the hospitals A, B and E showed a statistically significant higher exposure than those of the hospital D. The median values of the two biomarkers confirmed a statistically significant different exposure among the operators of different hospitals (see Table 2). Table 3 shows the data grouped by job title. The results of the Kruskal–Wallis test showed that the differences between the medians of different groups of subjects were statistically
Table 2 Individual sevoflurane levels measured with Radiello1, and urinary concentrations of sevoflurane and HFIP in operating room personnel (total and grouped by hospital). The different hospitals were named A, B, C, D, E. Hospital
A B C D E Total population
No. of subjects
54 23 5 10 8 100
Sevoflurane in urine (mg/L)
HFIP (mg/L)
Mean SD (median)
Range
Mean SD (median)
Range
Mean SD (median)
Range
0.097 0.174 (0.028)a 0.250 0.319 (0.107)a 0.038 0.034 (0.028) 0.0132 0.003 (0.012)a 0.030 0.012 (0.026)a
<0.010–0.912 0.010–0.940 0.015–0.096 <0.010–0.018 0.020–0.049
0.58 0.38 5.29 4.96 0.33 0.38 0.56 0.38 0.56 0.43
0.20–2.00 1.20–17.28 0.10–0.24 0.20–1.20 0.20–1.49
48.9 69.5 (27.3)a 133.2 150.6 (55.0)a 14.6 5.1 (13.0)a 16.7 16.2 (17.5)a 122.6 110.5 (110.0)a
1.0–360.0 1.0–550.0 10.0–20.0 1.0–50.0 1.0–290.0
0.116 0.213 (0.026)
<0.010–0.940
1.64 3.09 (0.62)
<0.10–17.28
69.2 101.7 (26.0)
<1.0–550.0
Sevoflurane in air (ppm)
(0.50)a (2.80)a (0.20)a (0.48)a (0.56)a
Sevoflurane in air: Overall, p value = 0.0085 (A vs D, p = 0.0032; B vs D, p = 0.0002; D vs E, p = 0.003). Sevoflurane in urine: Overall, p value < 0.0001 (A vs B, p < 0.0001; A vs C, p = 0.035; B vs C, p < 0.0001; B vs D, p < 0.0001; B vs E, p < 0.0001). HFIP: Overall, p value = 0.0023 (A vs B, p = 0.022; A vs C, p = 0.045; A vs D, p = 0.041; B vs C, p = 0.0026; B vs D, p = 0.0009; C vs E, p = 0.0131; D vs E, p = 0.0126). a Kruskal–Wallis test.
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
157
Table 3 Airborne sevoflurane levels and urinary concentrations of sevoflurane and HFIP in operating room personnel by job title. Job title
Anesthetists n = 19 Surgeons n = 25 Instrumentalists n = 14 Anesthetist nurses n = 9 Operating room nurses n = 33
Sev-U (mg/L)
Sevoflurane in air (ppm)
HFIP (mg/L)
Mean SD (median)
Range
Mean SD (median)
Range
Mean SD (median)
Range
0.290 0.353 (0.096)a 0.124 0.226 (0.016)a 0.116 0.218 (0.020) 0.093 0.070 (0.089) 0.095 0.159 (0.026)a
0.013–0.940 <0.010–0.777 <0.010–0.734 <0.010–0.189 0.010–0.558
2.18 3.49 (0.70) 2.26 3.35 (0.82) 2.59 4.80 (0.78) 0.87 0.52 (0.90) 2.39 4.41 (0.90)
0.20–13.21 0.20–12.72 <0.10–16.30 0.20–1.60 0.20–17.28
85.5 100.1 (49.4) 82.9 134.4 (35.8) 75.1 127.2 (23.0) 65.4 79.5 (26.00) 62.5 76.3 (31.7)
1.0–344.1 1.0–550.0 10.0–450.0 20.0–240.0 <1.0–217.1
Overall: p value = 0.0301 (Anesthetists vs Instrumentalists, p = 0.0019; Anesthetists vs Surgeons, p = 0.002; Anesthetists vs Operating room nurses, p = 0.0297). a Kruskal–Wallis test.
significant indicating that job title was an important variable influencing exposure to the anaesthetic. Anaesthetists revealed a statistically higher exposure to sevoflurane than surgeons, instrumentalists or operating room nurses, but not in comparison with anaesthetist nurses. There were no statistically significant differences between the groups of workers with regard to the levels of either biomarkers, however. Fig. 1A shows the correlation between individual environmental exposure data and urinary excretion of sevoflurane (log-transformed data). Although there were some very scattered points, where quite low environmental concentrations corresponded to very high biological values and vice versa, a statistically significant correlation emerged overall (r = 0.49, p < 0.0001). Moreover, all the high Sev-U concentrations concerned subjects working in the same operating room, where the photoacoustic
Fig. 1. Correlations between individual sevoflurane levels measured with Radiello1 and urinary concentrations of sevoflurane (A) or HFIP (B) at the end of the work shift in all subjects.
analyzer revealed very high peak levels of sevoflurane – even exceeding 20 ppm – towards the end of the operating room session, which was due to a leak in the supply circuits (data not shown) while during most of the session sevoflurane concentrations were very low. Taking the short half-life of sevoflurane (2.8 h) into account, even these short peaks could be responsible, therefore, for the high Sev-U values detected in these subjects. Fig. 1B shows the correlation between individual environmental sevoflurane concentrations and urinary HFIP levels (log-transformed data), with a regression coefficient of 0.52 (p < 0.0001). Here again, the scattering of the data may be partly explained by the substance’s relatively long half-life (19 h): subjects with high urinary levels of the metabolite and very low levels of environmental exposure may have been highly exposed on the previous day. Unfortunately, we do not have environmental data for the operating room sessions performed on the previous days. The correlation between airborne sevoflurane values and urinary HFIP levels did not improve significantly after adjusting for creatinine (data not shown). A statistically significant correlation was found between the two biomarkers (log-transformed data, r = 0.273, p = 0.006), albeit showing a marked variability in the values (Fig. 2). In order to verify the reliability of the two biomarkers also in the case of very low exposure, the data were divided into two subgroups according to the median value, equal to 0.026 ppm. Table 4 shows the linear relationships between log-log transformed values of airborne sevoflurane and urinary biomarkers (Sev-U and HFIP) in all subjects (see also Fig. 1) and in subjects with sevoflurane exposure below and above 0.026 ppm. The regression equations for subgroups with exposure above 0.026 ppm is similar to that obtained considering all the 100 subjects and the correlations were still statistically significant. In the groups with exposure below 0.026 ppm, the correlation between the
Fig. 2. Correlation between urinary concentrations of sevoflurane and HFIP measured at the end of the work shift in all subjects.
158
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
Table 4 Linear relationships between log-log transformed values of airborne sevoflurane (Sev-A, ppm) and urinary biomarkers (Sev-U and HFIP, mg/L): all subjects (1), subjects with sevoflurane exposure <0.026 ppm (2), subjects with sevoflurane exposure 0.026 ppp (3). Equation log log log log log log
Sev-U = 0.417 log Sev-A + 0.444 (1) HFIP = 0.640 log Sev-A + 2.310 (1) Sev-U = 0.576 log Sev-A + 0.837 (2) HFIP = 2.072 log Sev-A + 4.995 (2) Sev-U = 0.779 log Sev-A + 0.689 (3) HFIP = 0.606 log Sev-A + 2.255 (3)
No. of subjects
r
p
100 100 50 50 50 50
0.49 0.52 0.22 0.40 0.65 0.51
<0.0001 <0.0001 n.s. 0.0039 <0.0001 0.0001
n.s., not statistically significant.
environmental data and the HFIP values was statistically significant but quantitatively different, probably due to the values under the analytical detection limit. As for Sev-U, the correlation was not statistically significant. It should be underlined that at these very low exposure levels, the influence of other factors, such as intraand interindividual biological variability, can be more relevant. No statistically significant differences were found, instead, in the correlations between airborne sevoflurane exposure and biological monitoring data among the different groups of professionals (data not shown). 4. Discussion In the present study, environmental sevoflurane pollution in the hospitals monitored was very low, with average concentrations of the anaesthetic generally lower than those reported in the literature (Accorsi et al., 2003, 2005; Barbic et al., 2003; Haufroid et al., 2000; Imbriani et al., 2001). Pollution levels, though, were very variable and could differ among different hospitals and even between operating rooms at the same hospital. As already reported in other studies (Accorsi et al., 2005; Tankó et al., 2009; Trevisan and Gori, 1990; Virgili et al., 2002), we found that anaesthetists were more exposed to sevoflurane than other professionals, due to their prolonged presence near both the anaesthesiological equipment and the patient. The use of different sampling methods, both reliable, enabled us to obtain detailed information on both environmental pollution and health care workers’ individual exposure. Continuous monitoring with the photoacoustic analyzer identified environmental pollution peaks, while individual sampling by Radiello1 measured the total individual exposure of different professionals. Our study also identified a statistically significant correlation between individual environmental exposure and biomarkers’ levels. In both cases, however, there were some scattered points, where quite low environmental concentrations corresponded to very high biological values. This may be partly explained by the different half-lives of the two biomarkers. As mentioned earlier, HFIP levels, because of the long half-life (19 h), may also reflect exposure on the previous day. Some authors (Barbic et al., 2003; Hoerauf et al., 1997) found appreciable levels of this metabolite in the urine of many subjects even at the beginning of the work shift. Some operators who had been exposed to sevoflurane during the previous working day already had high HFIP values in their urine collected at the start of their operating room shift, showing that 24 h were not enough for a substantial, let alone complete, urinary HFIP elimination. Moreover, unlike Barbic et al. (2003), but in agreement with Haufroid et al. (2000) and Imbriani et al. (2001), we found that adjusting for creatinine did not improve the correlation between environmental sevoflurane data and urinary HFIP values. As for Sev-U, our findings suggest that this biomarker is particularly influenced by peaks of exposure (as detected by the photoacoustic analyzer), especially towards the end of the operating room session, probably because of the substance’s very
short half-life (2.8 h). This explanation is supported by the data of Fig. 2, showing a statistically significant correlation between the two biomarkers, but with a wide scattering of the values. The scattering can also be partly explained by interindividual biological variability, which is bound to be higher at very low sevoflurane exposures, as found indeed in our study (always below 1 ppm), and also by the presence of a cluster of subjects with low Sev-U levels but relatively high HFIP values, possibly due to residual exposure from the previous day. Based on our results, the calculated urinary concentrations of biomarker corresponding to an environmental sevoflurane exposure of 2 and 0.5 ppm were 318 and 131 mg/L, respectively, for HFIP (lower 95% CL: 133 and 72 mg/L, respectively) and 3.7 and 2.1 mg/L, respectively, for Sev-U (lower 95% CL: 2.0 and 1.4 mg/L, respectively). The values for HFIP are quite similar to those calculated by Imbriani et al. (2001), but very different from those proposed by Haufroid et al. (2000),Barbic et al. (2003) and Accorsi et al. (2005). As for Sev-U, our calculated biological equivalent exposure values are similar to those proposed by Accorsi et al. (2005), though, when we omitted subjects showing significant peak levels of sevoflurane exposure at the end of an operating session (as reported in the Section 3), our values became remarkably lower (1.9 and 1.3 mg/L for an environmental sevoflurane exposure of 2 and 0.5 ppm, respectively). So, considering the present biological monitoring data overall, both biomarkers, HFIP and Sev-U, have confirmed some critical features reported in the literature. Urinary HFIP has been suggested by several authors as a specific indicator of occupational exposure to sevoflurane for which the biological equivalent level corresponding to a given airborne concentration could be derived (Barbic et al., 2003; Haufroid et al., 2000; Imbriani et al., 2001). Values obtained from calculation by different authors, though, were rather different. Imbriani et al. (2001) for instance calculated a biological equivalent concentrations for HFIP of 488 and 160 mg/L, corresponding to the NIOSH limits of exposure to airborne sevoflurane of 2 and 0.5 ppm, respectively. Haufroid et al. (2000) extrapolated an intermediate biological index of 800 mg/L HFIP, corresponding to a 2 ppm concentration of sevoflurane, while Barbic et al. (2003) calculated a value of 2772.7 mg/L HFIP for a 2 ppm sevoflurane exposure. It should be emphasized, however, that the subjects of our study showed, on average, a much lower exposure levels than those of other studies (Accorsi et al., 2005; Barbic et al., 2003; Haufroid et al., 2000; Imbriani et al., 2001). For example, all our exposure data are similar to the subgroup of subjects identified by Imbriani et al. (2001) as “low exposed”. Accorsi et al. (2005) assessed both biomarkers considered in our study, and reported a statistically significant correlation with the environmental levels of sevoflurane for both, though HFIP data showed a less linear correlation with individual exposure, and a rather high intercept, than those of Sev-U. This suggested some degree of accumulation of the metabolite during the course of the workweek, although the authors were unable to find any differences in excretion on different weekdays, or between data
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
obtained in the early and the late parts of the two weeks of their survey. Taking the half-life of the two biomarkers into account (2.8 and 19 h for sevoflurane and HFIP, respectively), the authors suggested that end-of-work-shift urinary sevoflurane could be a biomarker reflecting (only) very recent exposure (i.e. from the same work shift), while HFIP could be a more stable exposure tracer, also reflecting various conditions capable of influencing in vivo sevoflurane metabolism, such as cigarette smoking, alcohol consumption and genetic metabolic polymorphisms. Since the enzyme mainly responsible for HFIP production is CYP2E1, and smoking and alcohol can be important modulators of its activity, the authors studied the influence of these variables. They indeed found that smoking habits were an important confounding factor for urinary HFIP but not for Sev-U. This is one of the reasons why they recommended using Sev-U rather than HFIP for the biological monitoring of sevoflurane exposure. When they calculated, for both biomarkers, the biological equivalent limits of exposure corresponding to the airborne sevoflurane concentrations of 2 and 0.5 ppm. they found for HFIP, but not for Sev-U, higher values than those found in the present study: 2660 and 820 mg/L for HFIP (lower 95% CL, 1610 and 630 mg/L, respectively) and 3.9 and 1.4 mg/L for Sev-U (lower 95% CL: 2.8 and 1.2 mg/L, respectively), respectively. In conclusion, both biomarkers used in this study to assess occupational sevoflurane exposure were found to be correlated with the levels of the anaesthetic measured in the air, although they both suffered from some critical variability factors that we were unable to fully elucidate. The current state of knowledge suggests that both biomarkers can be used for group biological monitoring purposes, providing that standardized procedures for urine sample collection are adopted and adequate information about the individual’s lifestyle, or other habits that are capable of influencing HFIP metabolism, is available. As for HFIP, given the relatively long half-life, its use might be extended to assess also the exposure of workers who were (only) involved in the morning shift on the previous day. As for Sev-U, instead, its measurement should only be used for monitoring exposure occurring during the work shift on same day. Besides, particular attention should be paid to urine sample collection and storage, as samples require to be sealed carefully within one or two minutes of collection to prevent any loss of the anaesthetic, and to detecting exposure peaks, especially if they occurred in the latter part of the operating room session. A continuous monitoring of airborne anaesthetics’ levels would be most helpful in this regard. Finally, further research is needed to reach consensus on which biomarker is best for the routine biomonitoring of sevoflurane exposure in the workplace and to establish definite biomarker’s levels to be recommended as biological equivalent exposure limit values for routine use in occupational health programmes. Conflict of interest The authors declare that there are no conflicts of interest. Transparency document The Transparency document associated with this article can be found in the online version. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.toxlet.2014.10.018.
159
References Accorsi, A., Valenti, S., Barbieri, A., Raffi, G.B., Violante, F.S., 2003. Proposal for single and mixture biological exposure limits for sevoflurane and nitrous oxide at low occupational exposure levels. Int. Arch. Occup. Environ. Health 76, 129–136. Accorsi, A., Morrone, B., Domenichini, I., Valenti, S., Raffi, G.B., Violante, F.S., 2005. Urinary sevoflurane and hexafluoro-isopropanol as biomarkers of low-level occupational exposure to sevoflurane. Int. Arch. Occup. Environ. Health 78, 369–378. Barbic, F., Bagnati, M., Basile, M., Zanoli, V., Carrettoni, L., Cassani, C., Porta, P., Fortina, A., 2003. L’esafluoroisopropanolo urinario nella valutazione dell’esposizione professionale a sevoflurano: aspetti metodologici e punti critici. G. Ital. Med. Lav. Ergon. 25, 95–97. Barker, J.P., Abdelatti, M.O., 1997. Anaesthetic pollution. Potential sources, their identification and control. Anaesthesia 52, 1077–1083. Behne, M., Wilke, H.J., Harder, S., 1999. Clinical pharmacokinetics of sevoflurane. Clin. Pharmacokinet. 36, 13–26. Byhahn, C., Wilke, H.J., Westpphal, K., 2001. Occupational exposure to volatile anaesthetics: epidemiology and approaches to reducing the problem. CNS Drugs 15, 197–215. Cottica, D., Bartolucci, G.B., Grignani, E., Locatelli, C., Sala, C., Scapellato, M.L., Sesana, G., 2001. Gas anestetici. In: Minoia, C., Perbellini, L. (Eds.), Monitoraggio Ambientale e Biologico Dell’Esposizione Professionale a Xenobiotici, Vol. 4. Morgan Edizioni Tecniche, Milano. Eger, E.I., 1993. New inhalation agents – desflurane and sevoflurane. Can. J. Anaesth. 40, 3–8. Fustinoni, S., Pasini, R., Strangi, F., Valla, C., Missineo, P., Margonari, M., Denaro, R., Bertazzi, P.A., 2012. Air and biomonitoring of occupational exposure to anesthetic gases in the health care workers of a large hospital in Milan. G. Ital. Med. Lav. Ergon. 34 (Suppl. 3), 278–279 (Article in Italian). Haufroid, V., Cardinal, S., Licot, C., Villalpando, M.L., Van Obbergh, L., Clippe, A., Lison, D., 2000. Biological monitoring of exposure to sevoflurane in operating room personnel by the measurement of hexafluoroisopropanol and fluoride in urine. Biomarkers 5, 141–151. Henderson, K.A., Matthews, I.P., 2000. Staff exposure to anaesthetic gases in theatre and non theatre areas. Eur. J. Anaesthesiol. 17, 149–151. Hoerauf, K.H., Koller, C., Taeger, K., Hobbhahn, J., 1997. Occupational exposure to sevoflurane and nitrous oxide in operating room personnel. Int. Arch. Occup. Environ. Health 69, 134–138. The International Conference on Harmonization (ICH), 1996. Validation of Analytical Procedure: Methodology (Q 2B), Food and Drug Administration, USA. ICH, Geneva, Switzerland. Imbriani, M., Zadra, P., Negri, S., Alessio, A., Maestri, L., Ghittori, S., 2001. Biological monitoring of occupational exposure to sevoflurane. Med. Lav. 92, 173–180 (Article in Italian). Kharasch, E.D., Thummel, K.E., 1993. Identification of cytochrome P450 2E1 as the predominant enzyme catalyzing human liver microsomal defluorination of sevoflurane, isoflurane and methoxyflurane. Anesthesiology 79, 795–807. Kharasch, E.D., Karol, M.D., Lanni, C., Sawchuk, R., 1995a. Clinical sevoflurane metabolism and disposition: I. Sevoflurane and metabolite pharmacokinetics. Anesthesiology 82, 1369–1378. Kharasch, E.D., Armstrong, A.S., Gunn, K., Artru, A., Cox, K., Karol, M.D., 1995b. Clinical sevoflurane metabolism and disposition: II. The role of cytochrome P450 2E1 in fluoride and hexafluoroisopropanol formation. Anesthesiology 82, 1379–1388. Martis, L., Lynch, S., Napoli, M., Woods, E.D., 1981. Biotransformation of sevoflurane in dogs and rats. Anaesth. Analg. 60, 186–191. NIOSH, 1977. Criteria for a Recommended Standard: Occupational Exposure to Waste Anesthetic Gases and Vapors. Publication no. 77–140. US DHEW, Cincinnati, Ohio. Robinson, B.J., Uhrich, T.D., Ebert, T.J., 1999. A review of recovery from sevoflurane anaesthesia: comparisons with isoflurane and propafol including metaanalysis. Acta Anaesthesiol. Scand. 43, 185–190. Rovesti, S., Ferrari, A., Faggiano, D., Vivoli, G., 2003. Monitoring occupational exposure to volatile anaesthetics in the operating theatre: environmental and biological measurements. Ann. Ig 17, 219–230 (Article in Italian). Sartini, M., Otaria, G., Dall’era, M., Spagnolo, A.M., Cristina, M.L., 2006. Nitrous oxide pollution in operating theatres in relation to the type of leakage and the number of efficacious air exchanges per hour. J. Prev. Med. Hyg. 47, 155–159. Scapellato, M.L., Trevisan, A., Carrieri, M., Maccà, I., Bonfiglio, E., Gori, G., Serraino, S., Bartolucci, G.B., 2004. Dose and effect biomarkers in sevoflurane exposure. G. Ital. Med. Lav. Ergon. 26, 99–101 (Article in Italian). Scapellato, M.L., Mastrangelo, G., Fedeli, U., Carrieri, M., Maccà, I., Scoizzato, L., Bartolucci, G.B., 2008. A longitudinal study for investigating the exposure level of anesthetics that impairs neurobehavioral performance. NeuroToxicology 29, 116–123. Smith, F.D., 2009. Management of exposure to waste anesthetic gases. AORN J. 91, 482–494. Tankó, B., Molnár, C., Budi, T., Peto, C., Novák, L., Fülesdi, B., 2009. The relative exposure of the operating room staff to sevoflurane during intracerebral surgery. Anesth. Analg. 109, 1187–1192. Trevisan, A., Borella Venturini, M., Carrieri, M., Giraldo, M., Maccà, I., Perini, M., Scapellato, M.L., Virgili, A., Bartolucci, G.B., 2003. Biological indices of kidney involvement in personnel exposed to sevoflurane in surgical areas. Am. J. Ind. Med. 44, 474–480.
160
M.L. Scapellato et al. / Toxicology Letters 231 (2014) 154–160
Trevisan, A., Gori, G., 1990. Biological monitoring of nitrous oxide exposure in surgical areas. Am. J. Ind. Med. 17, 357–362. Virgili, A., Scapellato, M.L., Maccà, I., Perini, M., Carrieri, M., Gori, G., Saia, B., Bartolucci, G.B., 2002. Occupational exposure to anaesthetic gases in some hospitals in the Veneto Region. G. Ital. Med. Lav. Ergon. 24, 447–450 (Article in Italian).
Zaffina, S., Camisa, V., Poscia, A., Tucci, M.G., Montaldi, V., Cerabona, V., Wachocka, M., Moscato, U., 2012. Occupational exposure to sevoflurane in pediatric operating rooms: the multi-point sampling method for risk assessment. G. Ital. Med. Lav. Ergon. 34 (Suppl. 3), 266–268 (Article in Italian).