- Email: [email protected]
The skin reservoir of sulphur mustard
Toxicology in Vitro 22 (2008) 1539–1546
Contents lists available at ScienceDirect
Toxicology in Vitro journal homepage: www.elsevier.com/locate/toxinvit
The skin reservoir of sulphur mustard I.J. Hattersley a,*, J. Jenner a, C. Dalton a, R.P. Chilcott b, J.S. Graham c a b c
Dstl Biomedical Sciences Department, Biomedical Systems, Dstl Porton Down, Salisbury, Wiltshire, SP4 0JQ, UK Chemical Hazards and Poisons Division (HQ), C/o NRPB, Chilton, Didcot, Oxfordshire OX11 0RQ, UK US Army Medical Research Institute of Chemical Defense, 3100 Ricketts Point Road, Aberdeen Proving Ground, MD 21010-5425, USA
a r t i c l e
i n f o
Article history: Received 27 February 2008 Accepted 6 June 2008 Available online 14 June 2008 Keywords: Dermal Skin Reservoir Sulphur mustard
a b s t r a c t Studies of the percutaneous reservoir of sulphur mustard (HD) formed during absorption carried out during WWI and WWII are inconclusive. More recent studies have indicated that a significant amount of unreacted HD remains in human epidermal membranes during percutaneous penetration studies in vitro. The present study investigated the nature and persistence of the HD reservoir formed during in vitro penetration studies using dermatomed slices of human and pig skin (0.5 mm thick). Amounts of 14C-HD that (a) penetrated, (b) remained on the surface, (c) were extractable from and (d) remained in the skin after extraction were estimated by liquid scintillation counting (confirmed using GC–MS analysis). The results demonstrated that there is a reservoir of HD in human and pig skin for up to 24 h after contamination of the skin surface in vitro with liquid agent. At least some of this reservoir could be extracted with acetonitrile, and the amounts of extracted and unextracted HD exceed the amount required to produce injury in vivo by at least 20 fold. The study demonstrated the presence of a reservoir whether the skin was covered (occluded) or left open to the air (unoccluded). The study concluded that the extractable reservoir was significant in terms of the amount of HD required to induce a vesicant response in human skin. The extractable reservoir was at least 20 times the amount required per cm2 estimated to cause a response in all of the human population, as defined by studies carried out in human volunteers during the 1940s. Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction Sulphur mustard (HD) was first used as a chemical warfare agent in World War 1 (WW1) and is a potent vesicant, damaging all the surfaces of the body including the skin, eyes and respiratory tract. An important aspect of HD toxicology is the latency period after contact with liquid or vapour before symptoms are manifest, which can be hours or days depending upon dosage. This latency has led researchers to propose the existence of a depot or reservoir of HD within the skin, which has long been debated. Several historical studies have reported conflicting data and conclusions with regard to the existence of such a reservoir. A WW1 study reported by Smith and co-workers (1919), established that there was evidence for a substantial extractable reservoir, while studies conducted during the Second World War (WW2), summarised by Renshaw (1947), indicated that, even following massive liquid contamination, there was no appreciable extractable reservoir of unreacted agent. Smith et al. (1919) showed that HD contamination may be transferred from an exposed individual to an unexposed individual via direct contact between the two. A volunteer was exposed to HD * Corresponding author. E-mail address: [email protected] (I.J. Hattersley). 0887-2333/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.tiv.2008.06.002
vapour and, 30 min after exposure, the skin of a second volunteer was placed in direct contact with the contaminated area on the first. The second volunteer subsequently developed a burn. Thus, it was clear that as agent may be transferred from a contaminated individual to an uncontaminated individual, for at least thirty minutes after exposure, unreacted HD must be present on or near the surface of the skin. Occlusion (covering) of the site of contamination also increased intensity of the burn. It was clear from this work that a reservoir of HD existed in human skin for at least 45 min post exposure, some of which could be extracted with a suitably penetrating solvent (kerosene). Studies conducted in the 1940s by Henriques and Moritz (1944), Cullumbine (1946) and Axelrod and Hamilton (1947) are summarised in a review by Renshaw (1947). Renshaw concluded that, based on the work of these authors, ‘‘even after massive contamination with liquid HD, human skin contains no appreciable reservoir of unreacted HD” and therefore ‘‘all therapeutic procedures based upon the neutralisation of free penetrated HD must necessarily be valueless”, in direct conflict with the conclusions of Smith et al. (1919) 20 years earlier. Our review of the original studies summarised by Renshaw revealed some inconsistencies between the investigators observations which may explain the differences in the conclusion of Renshaw’s review and the observations of Smith et al. (1919).
1540
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
Henriques and Moritz (1944) could not extract radiolabelled HD from the skin at 2 min post exposure, indicating that HD was rapidly fixed within the skin. However, the skin was cleaned with an ether soaked swab to remove any surface contamination prior to analysis, which may also have removed any HD reservoir. Cullumbine (1946) used a histological stain which formed an insoluble black complex with free HD but not with degradation products such as thiodiglycol. This showed ‘‘free” unreacted HD in the skin at 15 min post exposure but not at 30 min leading to the conclusion that ‘‘free HD, after penetration, did not exist in the skin except in the epidermis”. Using micro-autoradiography with 35S-HD Axelrod and Hamilton (1947) showed that 35S label was present within both the epidermis and the dermis, although it was never determined whether the label detected was free HD or a degradation product. A more recent study, Chilcott et al. (2001), investigated the existence of a HD reservoir in exposed skin (heat separated epidermal membranes). 35S labelled HD was used to determine absorption rate and dose distribution of liquid HD through human epidermal membranes. ‘‘Free” HD was determined by placing treated skin sections in ethyl ether for up to 1 week, to ensure all extractable radioactivity had been removed, and sampling the ether for 35S. ‘‘Fixed” or bound HD was determined by placing the skin samples (following ether extraction) into a tissue solubiser (Soluene 350: quaternary ammonium hydroxide in toluene) to determine the extent of label remaining within the sample (confirmed by GC–MS analysis). The results clearly demonstrated that an extractable HD reservoir (up to 36% of applied dose) was present within the epidermal membranes for at least 24 h post exposure. There are clear inconsistencies between studies conducted around the time of the WWI and those conducted during and after WW2. Therefore, it is important to further evaluate the possibility of the presence of an extractable sulphur mustard reservoir within the skin of an exposed individual so that future therapeutic techniques may be optimised accordingly. The aims of this study were to identify and quantify the extent of the sulphur mustard reservoir present within the skin, in vitro, both as unextractable and extractable HD, under both occluded and unoccluded conditions, for pig and human skin. A range of time points were investigated, from 5 min to 24 h post exposure. These data provided an estimate as to the feasibility of removing/extracting HD from the skin of an individual following contamination with HD and thus the possibility of effective prevention of or reduction in the severity of incapacitation caused as a result of exposure to sulphur mustard by suitable decontamination techniques.
2. Materials and methods The synthesis, use and destruction of HD in this study was conducted in accordance with the Chemical Weapons Convention (1997) to which the UK is a signatory state. Radiolabelled HD was synthesised at Dstl Porton Down and had a radiochemical purity >98%. Chemical purity of unlabelled HD was >98%. Radiolabelled and cold agents were mixed in appropriate proportions to give a nominal activity of approximately 10 kBq ll 1. Liquid scintillation counting (LSC) materials (Soluene-350TM, emulsifier-safe cocktail and opaque plastic vials) were purchased from Perkin–Elmer (Chandler’s Ford, Hampshire). All other chemicals were analytical grade and were purchased from the Sigma Chemical Company (Poole, Dorset). Animals were used in accordance with the Animals (Scientific Procedures) Act 1986. Twelve weanling pigs (large white strain, weight range 8–10 kg) were obtained from the Institute of Animal Health (Newbury). Animals were given access to food and water ad
libitum. To obtain fresh pig skin animals were sedated with HypnovelÒ (Midazolam, 6 ml i.m., 5 mg ml 1), anaesthetised with 5% Isofluorane and culled with an overdose of EuthatalTM (sodium pentobarbitol, 6 ml i.v., 200 mg ml 1). The whole dorsal skin flank (approximately 40 30 cm) was close-clipped and excised from each animal. Subcutaneous fat was removed and the skin was immediately dermatomed to a nominal thickness of 500 lm using a Zimmer air dermatome (Zimmer Inc., Dover, OH, USA), prior to mounting in diffusion cells. Whole human abdominal skin (from elective reduction surgery) was obtained with full patient consent (age range 29–43) and was free from overt pathology. Donors were female Caucasians. The skin samples were stored at 25 °C for up to 2 months. After thawing, each skin sample was cleaned of subcutaneous fat and was dermatomed to a nominal thickness of 500 lm using a Zimmer air dermatome, prior to mounting in diffusion cells. 2.1. Percutaneous absorption Experiments were performed with Franz-type glass diffusion cells with an area available for diffusion of 2.54 cm2. Each diffusion cell consisted of a dermatomed piece of either human or pig skin forming a barrier between an upper (donor) and lower (receptor) chamber. The receptor chambers were filled with 50% aqueous ethanol (5 ± 0.5 ml). The skin surface temperature within each diffusion cell was controlled by mounting the sets of six diffusion cells on metal plates heated by circulating water. Pilot studies using thermocouples to measure the skin surface temperature established the water temperature required to maintain skin surface at 32 °C (±2 °C). The heated plates were mounted on stirring plates enabling magnetic followers to ensure the receptor solution of each cell was continuously stirred. The skin was allowed to equilibrate with the receptor media for 12 h prior to HD dosing. Experiments were designed such that each control and treatment group contained samples of skin from six different individuals. Results are there fore expressed as mean ± SD of ‘n’ individual donors. A preliminary series of experiments monitored the deposition of 14C-HD in the skin of pigs and humans up to 3 h post contamination, under both occluded and unoccluded conditions. The experimental protocol for the main study assessed time points of 1/2–24 h using 6 samples of skin from different donors at each time point. The studies conducted over 24 h consisted of unoccluded human skin, occluded pig skin and unoccluded pig skin. The effects of occlusion on human skin up to 24 h post exposure was not investigated due to a limited supply of human tissue, although results are available for occluded human skin up to 3 h post exposure. Topical dosing was performed by the direct application of undiluted HD (20 ll = 25.4 mg) to the skin using a positive displacement pipette. This amount of HD has been previously shown to cover the surface of a 2.54 cm2 diffusion cell by autoradiography (Chilcott et al., 2000). For diffusion cells being evaluated under occluded conditions the donor chambers had a 2 2 cm2 of parafilm placed onto an inert (perfluorinated) cream (AG-7) that covered the rim of each donor chamber. Following application of the agent the cells were left for the predetermined length of time. The parafilm from each occluded cell was removed and placed into 20 ml acetonitrile. Each skin surface was gently swabbed with dry cotton wool which was then immersed in 20 ml acetonitrile. Each skin sample was removed and weighed before being placed into 20 ml acetonitrile for 24 h, after which it was air dried to remove excess solvent before being reweighed and placed into 10 ml Soluene-350TM (it was determined, by a preliminary investigation, that immersion in acetonitrile for 24 h was sufficient to remove any ‘free’ HD from the skin sample).
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
The receptor fluid was retained from each receptor chamber for sampling. ‘‘Standard” samples were prepared simultaneously from known quantities of 14C-HD. The amount of radioactivity in each sample was measured using a Wallac 1409 DSA LS counter (using the manufacturer’s 14Cquench curve library) set to exclude single-photon (non-radioactive) events. Counter efficiency was assessed as (CPM/DPM) 100 and was 89–90%. The amount of radioactivity was converted to amount of HD by comparison to standard samples prepared and measured simultaneously. Samples of the acetonitrile extracts of skin and extracts of the cotton wool swabs were retained at 25 °C until analysed by GC–MS.
1541
and unoccluded conditions and similarly for pig skin under both conditions. In these experiments the amount of 14C-HD extracted from the tissue was generally higher than that which remained in the tissue. These experiments showed that there was a reservoir in the skin at three hours after contamination and indicated that experiments were required to define the distribution of 14C-HD for 24 h after contamination. In the 3 h studies there was a larger reservoir of extractable HD from human skin (4–8%) than pig skin (0.4–1.5%) under occluded conditions (p < 0.05, 2 way ANOVA), however this difference was not significant under unoccluded conditions. 3.3. Twenty four hour experiments
2.2. Quantification of samples by GC–MS HD samples were analysed on a DB-WAX column (25 m 0.2 mm I.D., 0.2 lm film thickness) at 90 °C (1 min), 90– 260 °C (linear gradient at 20 °C/min) and 260 °C (5.5 min) in an Agilent Technologies 6890 gas chromatograph. Samples were eluted with a helium (99.999%) carrier gas (13 psi head pressure at 250 °C; splitless) and quantified using a flame photometric detector set-up for phosphorus and sulphur detection. Confirmation of randomly selected positive samples was made using an Agilent Technologies 6890 gas chromatograph with a 5973 inert mass selective detector in electron impact mode (GC– MS(EI)). This system chromatograph used a DB–5MS column (25 m 0.2 mm I.D., 0.33 lm film thickness), helium carrier, 17 psi splitless injector at 250 °C, and a elution temperature gradient of 40 °C (1 min), 40–160 °C (linear gradient at 20 °C/min) and 260– 320 °C (linear gradient at 20 °C/min) and 320 °C (2.0 min). The mass spectrometer used an ion source temperature of 230 °C, ionisation: EI (70 eV), ionisation current: 60 lA and a mass scanning range of 40–550 m/z. Data acquisition was started after a 4 min solvent delay post injection. Identification of sulphur mustard was determined against authentic sulphur mustard analytical standard spectra and the NIST02 mass spectral library database. Quantitation of sulphur mustard in the acetonitrile extracts was determined against authentic sulphur mustard analytical standards in the selective ion monitoring mode utilising: 109, 111, 158 and 160 m/z over a calibration range of 1–10 lg/ml. 2.3. Statistical analysis Where required statistical significance was determined using 1 way and 2 way ANOVA and 1 sample t-tests and were carried out using Graphpad Prism (Graphpad Prism version 4.00 for windows, GraphPad Software, San Diego, CA) analysis of amounts of HD. Differences were judged as significant at p < 0.05.
3. Results 3.1. Diffusion cell experiments The first series of experiments measured the distribution of 14CHD in the skin over 3 h post application of 14C-HD to the surface, under both occluded and unoccluded conditions. The second was conducted in an identical fashion to the first, except that distribution was monitored out to 24 h post exposure. 3.2. Three hour experiments (data not shown) In the first 3 h after application of 14C-HD the majority of the agent remained on the surface of the skin and could be removed by swabbing. This was true for human skin under both occluded
Unoccluded experiments demonstrated that the majority of the applied agent (>70%) remained on the surface of skin for up to three hours post exposure. However, the amount recoverable from the surface dropped steadily to less than 5% of the applied dose by 8 h in human skin (Fig. 1a) and by 12 h in pig skin (Fig. 2a). Under occluded conditions (pig only) the amount recoverable from the surface of the skin remained almost constant throughout 24 h (70–90%) (Fig. 3a). In these experiments the amount of agent which was extractable from the skin was lower than the amount which remained unextractable for both pig and human skin, under both occluded and unoccluded conditions. Even so, the amount of extractable HD was significantly greater than that required to cause vesication at certain time points, under unoccluded conditions in both human and pig skin (p < 0.05, 1 sample t-test, see Figs. 1 and 2). Total amount of HD in the tissue (extracted + unextracted) was 1–4% of the applied dose over the 24 h for human skin, under unoccluded conditions (Fig. 1b). The amount of agent in the pig tissue was higher under unoccluded conditions, ranging from 5% to 20% (Fig. 2b). The amount present in the pig tissue under occluded conditions was between 3% and 5% of the applied dose throughout the 24 h (Fig. 3b). Total recoveries of HD from diffusion cell systems were higher under occluded conditions than unoccluded conditions between 4 and 24 h post exposure for pig (Figs. 2 and 3a). This is due to evaporation from the unoccluded cells. After 4 h less than 50% of the initial applied dose could be recovered under unoccluded conditions for both pig and human skin, however, under occluded conditions in excess of 80% of the applied dose could be recovered at 24 h post exposure. The fraction of the applied dose which could be extracted from the skin using acetonitrile was less than 10% in all cases (human and pig, occluded and unoccluded conditions). Extractable fractions were similar in pig and human skin during the 24 h studies, generally being in the region of about 1% of the initial applied dose with no significant change over time (p > 0.05, 1 way ANOVA). The fraction of the applied dose not extracted from the skin was 0–3%, although in the case of the 24 h unoccluded pig skin study this was as high as 10–15% for the first 4 h, before dropping to approximately 4% by 24 h. The proportion of applied HD which had penetrated the skin at the end of the experiment was less than 2.5% for all 3 h studies (pig and human, occluded and unoccluded). Over 24 h the amount penetrated under unoccluded conditions was comparable for both pig and human skin (1.5% and 2%, respectively), however, it was as high as 7% for the pig skin under occluded conditions. Occluded conditions exhibited a higher level of penetration for both human and pig skin, being approximately double that of the unoccluded conditions over the 3 h time period and in excess of 4 fold higher over 24 h. The weight lost by skin samples upon extraction by acetonitrile was constant across the whole of this study, being generally in the region of 20–25% of the initial weight of the skin in most cases.
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
surface
30
penetrated
Applied Dose
25
Amount HD recovered (mg)
120
extractable unextractable
A
100
20
80
15
60
10
40
5
20
0
3
0 0.5
6
12
18
24
12
B
10
Recovered HD (% applied dose)
1542
8
2
6
* 1
* 4
* *
* *
*
2
* 0
0 0.5
6
12
18
24
Time (hours) Fig. 1. Recovery of HD at various time points up to 24 h after application of 20 lL of liquid to the surface of human skin under unoccluded conditions. Graph A shows the amount of HD, and percentage of applied dose, recovered from the receptor fluid i.e. penetrated the skin (cross hatched), remaining on the surface of the skin recovered by swabbing (solid), extracted from the skin with acetonitrile (open) and not extracted (hatched). Graph B shows only extracted and unextracted material for clarity. Points are mean ± standard deviation: n = 6 diffusion cells. The broken horizontal line on graph B indicates the dose (10 lg cm 2) at which all the human population shows a reaction to HD (Harvey and Anderson, 1934a,b). Asterisks above bars indicates that the amount of HD extracted significantly exceeds this dose (p < 0.05, 1 sample t-test).
There were no apparent differences between pig and human skin and between occluded and unoccluded conditions in this respect. 3.4. Chemical analysis of HD in skin samples Samples from the acetonitrile extracts of the surface swabs and skin from the unoccluded experiments were subject to GC–MS analysis. The results correlated well with the estimates of amounts of HD recovered made using radiolabel (correlation coefficients over the four experiments r2 = 0.858 ± 0.087; mean ± SD). The majority of the 14C label recovered from the surface of the skin was identified as HD, with only very small percentages (generally less than 1%) being identified as thiodiglycol, the hydrolysis product of HD. Of the material extracted from the skin, a large percentage remained as 14C HD, however, levels were higher than those from the surface swabs (means were in the range 18–53% from human and 8–67% from pig).
4. Discussion This study has shown that there is a reservoir of HD in human and pig skin for up to 24 h after contamination of the skin surface in vitro with liquid agent. At least some of this reservoir can be extracted with an organic solvent (acetonitrile) and at certain time points the amounts of extracted and unextracted HD significantly exceed the amount required to produce injury in vivo. This is supported by the evidence of the GC/MS analysis which identified free HD in the samples of acetonitrile used for extraction. 4.1. Recovery of HD after contamination of the skin surface The results of this study indicate that under occluded conditions the majority of an applied dose of liquid HD may be recovered from the surface of the skin (pig or human) up to 24 h post exposure by swabbing with cotton wool. Similarly, under unoccluded condi-
1543
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
surface
30
unextractable penetrated
Applied Dose
100
20
80
15
60
10
40
5
20
0
0 6
12
18
24
6
24
B 5
20
4
16
3
12
*
2
*
*
Recovered HD (% applied dose)
Amount HD recovered (mg)
25
120
extractable
A
8
* *
1
4
0
0 6
12
18
24
Time (hours) Fig. 2. Recovery of HD at various time points up to 24 h after application of 20 lL of liquid to the surface of pig skin under unoccluded conditions. Graph A shows the amount of HD, and percent applied dose, recovered from the receptor fluid i.e. penetrated the skin (cross hatched), remaining on the surface of the skin recovered by swabbing (solid), extracted from the skin with acetonitrile (open) and not extracted (hatched). Graph B shows only extracted and unextracted material for clarity. Points are mean ± standard deviation: n = 6 diffusion cells. The broken horizontal line on graph B indicates the dose (10 lg cm 2) at which all the human population shows a reaction to HD (Harvey and Anderson, 1934a,b). Asterisks above bars indicates that the amount of HD extracted significantly exceeds this dose (p < 0.05, 1 sample t-test).
tions, a combination of evaporation and swabbing may account for the bulk of an applied dose. Thus, in excess of 80% of an applied dose of liquid HD either evaporated or remained on the skin surface and was unavailable for diffusion, in agreement with previous studies. Therefore, 20% or less of the initial percutaneous dose of liquid HD will be available to penetrate the skin and cause vesication. Such an absorbed dose is more than sufficient to cause vesication. Studies reported by Anderson (1936) and Harvey and Anderson (1934a,b) established that in a population of Indian and British troops all the subjects produced a skin reaction to a dose of 10 lg cm2. Previous studies in this laboratory have demonstrated that an applied dose of liquid HD (20 ll) will spread to fill the area available for diffusion (2.54 cm2) within 1 h of application (Chilcott et al., 2000). Thus, in the diffusion cells used in this study, a minimum vesicating dose would equate to 25 lg, which is much lower than the mg quantities of HD absorbed in the present study. Although the amounts extracted and unextracted at certain time points were significantly larger (p < 0.05) than the dose caus-
ing vesication in all exposed men (10 lg cm 2, Andersen et al., 1936), the variability in the data was too high to be able to draw meaningful conclusions about the kinetics of these reservoirs. This study investigated the distribution of HD between the surface, the skin tissue and the receptor chamber of the diffusion cells over the first 3 h, and in a subsequent series of experiments over 24 h after contamination of the skin surface. In the 3 h experiments the proportion of the applied dose which was extractable from the skin with acetonitrile was 4–10% (1–2.5 mg) in human skin, under occluded and unoccluded conditions. For pig skin, however, the extractable dose was lower, being approximately 1.5% (0.375 mg). This was significant for occluded conditions (p < 0.05) but not unoccluded conditions. The extractable reservoir of HD did not change significantly over the 24 h time course. Over all the time points up to 24 h a maximum of 5% of the applied dose (1.25 mg) was extractable from the unoccluded pig skin which contrasts with previous results from this laboratory reported by Chilcott et al. (2000), who used human epidermal membranes
1544
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
30
A
120
penetrated
Applied Dose
unextractable extractable
25
100
20
80
15
60
10
40
5
20
0
0 6
12
18
24
2
8
B 6
1
Recovered HD (% applied dose)
Amount HD recovered (mg)
surface
4
2
0
0 6
12
18
24
Time (hours) Fig. 3. Recovery of HD at various time points up to 24 h after application of 20 lL of liquid to the surface of pig skin under occluded conditions. Graph A shows the amount of HD recovered from the receptor fluid i.e. penetrated the skin (cross hatched), remaining on the surface of the skin recovered by swabbing (solid), extracted from the skin with acetonitrile (open) and not extracted (hatched). Graph B shows only extracted and unextracted material for clarity. Points are mean ± standard deviation: n = 6 diffusion cells. The broken horizontal line on graph B indicates the dose (10 lg cm 2) at which all the human population shows a reaction to HD (Harvey and Anderson, 1934a,b). The extracted amounts of HD were not significantly greater than this dose.
and demonstrated an ether extractable reservoir of up to 36% of the applied dose. There could be several reasons for this disparity. The method of removing HD from the surface of the skin was slightly different. Chilcott et al. (2000) used a series of ethanol washes (5) to remove any remaining HD at the end of the experiment however, in this study cotton wool swabs were used, which may have removed a portion of the stratum corneum due to the mechanical action of the swabbing process (thus removing a portion extractable reservoir present in the upper portion of the skin, resulting in an under estimation of the extractable material). Similarly, the use of a solvent to remove HD from the skin surface could also remove a portion of the reservoir. Another difference between the two studies was the solvent used to extract the HD from the skin after the surface had been wiped. Acetonitrile was used in this study whereas diethyl ether was used by Chilcott et al. (2000). This may result in different extraction efficiencies due to the different partition coefficients of the HD between the solvents and the skin. Since it is known that
HD rapidly hydrolyses in aqueous environments, any unreacted HD in the skin after longer than a few minutes must be dissolved in the lipid domains of the tissue. In the skin this would be the specialised lipids of the stratum corneum and/or the lipids of the cell membranes in the viable epidermis and superficial dermis. This is indirectly shown by the fact that organic solvents are effective at removing the absorbed HD from the skin. Thus, Smith et al. (1919) have demonstrated that HD can be removed from the cutaneous reservoir using kerosene (paraffin). Chilcott et al. have shown that diethyl ether can remove some of this HD and this study has shown that acetonitrile can also remove a significant amount of HD from the skin following a neat liquid dose. Despite the limitation of all these studies it is clear that there is a solvent extractable reservoir present in the skin up to at least 24 h post exposure, which contrasts with the conclusions of Renshaw (1947), based on the work reported by Henriques and Moritz (1944), Cullumbine (1946) and Axelrod and Hamilton (1947). It is also important to note that the solvent extraction removed more
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546 Table 1 Percentage loss in weight of skin produced by extracting with acetonitrile in the 3 h study Time/mins
5 10 30 60 120 180
Human
Pig
Occluded
Unoccluded
Occluded
Unoccluded
22.84 ± 7.54 16.46 ± 6.19 19.93 ± 8.62 23.74 ± 7.87 20.60 ± 4.71 17.72 ± 5.49
23.15 ± 4.93 22.59 ± 6.15 24.23 ± 6.07 21.65 ± 4.97 23.77 ± 8.99 21.98 ± 8.72
23.55 ± 10.61 15.56 ± 9.11 27.83 ± 8.08 7.39 ± 3.48 40.91 ± 9.72 37.13 ± 8.27
23.15 ± 4.93 22.59 ± 6.15 24.23 ± 6.07 21.65 ± 4.97 23.77 ± 8.98 21.98 ± 8.72
Data are expressed as mean ± SD of 4–6 pieces of skin.
Table 2 Percentage loss in weight of skin produced by extracting with acetonitrile in the 24 h study Time/hours
0.5 1.0 2.0 4.0 6.0 8.0 10.0 12.0 16.0 20.0 24.0
Human
Pig
Unoccluded
Occluded
Unoccluded
33.80 ± 10.66 27.83 ± 6.54 29.06 ± 5.83 25.48 ± 3.16 31.31 ± 5.18 28.54 ± 3.47 n/a 26.00 ± 3.33 22.23 ± 3.31 20.22 ± 11.93 16.79 ± 11.97
n/a n/a n/a 22.39 ± 7.10 10.32 ± 11.59 18.90 ± 5.45 17.10 ± 7.66 25.88 ± 21.42 n/a n/a 11.41 ± 5.60
19.42 ± 3.32 16.72 ± 3.33 21.27 ± 5.03 18.62 ± 3.88 22.56 ± 4.54 17.05 ± 4.01 n/a 13.34 ± 6.19 16.27 ± 7.08 13.60 ± 5.49 11.72 ± 4.05
than the HD from the skin, since the average weight of the skin was reduced by 7–40% by the extraction process (Tables 1 and 2). This reduction in weight probably represents lipid material, water and ethanol co-extracted with the HD, but at this time the precise identity of the co-extractant is not known. Clearly, further work is required to determine the best solvent system to extract HD from the skin, and to determine the fraction, if any, of the HD remaining in the tissue in this study that was covalently linked to tissue constituents. 4.2. Fate of the absorbed fraction The in vitro preparation used in the present study is only able to reproduce some of the conditions prevalent within the skin during percutaneous absorption of HD. It has been shown previously that the physical processes of partitioning into the skin compares well to that measured by surface disappearance measurement in man (Chilcott et al., 2000). This preparation does not seek to maintain the skin in a metabolically viable state and the results should be viewed in this light. Since the major route of metabolism for HD is spontaneous hydrolysis, the absence of other metabolic pathways is not likely to materially affect the conclusions drawn from this study. The effects of perfusing the skin with blood is modelled here by ensuring the receptor fluid does not become saturated with HD (HD hydrolysis rapidly in the receptor fluid maintaining an effective ‘‘sink” for the parent molecule). This in vitro system therefore effectively models a maximally perfused area of skin. Similarly, the use of 50:50 ethanol/water as the receptor fluid ensures that the partitioning of the HD into the receptor fluid from the skin is not rate limiting. The fraction of the applied dose absorbed into the skin either penetrated into the receptor chamber (equivalent to the blood in vivo), or remained in the skin. Of the fraction remaining in the skin, when it is removed from the diffusion cell, part was extractable
1545
with acetonitrile. The fraction of the absorbed dose which remained in the skin after extraction contained the agent, some of which has been previously demonstrated to be covalently bound to the tissue, part of which would cause vesication in vivo. The 1–3% of applied dose that remained non-extractable in the present study between 6 and 24 h is consistent with previous studies (Chilcott et al., 2000) which have shown that in the region of 2% of an applied dose of HD is unextractable, and normally regarded as fixed in the skin. This small proportion which remains bound in the skin even following extraction with a solvent is most likely that which is directly responsible for vesication and thus development of a burn. Two percent of an applied dose of 20 ll (0.4 ll or 0.51 mg) equates to at least 20 times the minimum vesicating dose of 10 lg cm 2 for HD determined by Anderson (1936) and Harvey and Anderson (1934a) by applying HD in benzene to the skin of humans. Previous studies in this laboratory have shown a larger proportion of the absorbed dose is retained in the skin over a 24 h period (Chilcott et al., 2000, 2001). The earlier studies were conducted with epidermal membranes which had been prepared by heat separation from human skin, whereas the current study has used split thickness slices of skin prepared using a dermatome. Each of these preparations has different merit as a model of how the skin behaves in vivo. The epidermal membranes isolate the part of the skin generally believed to contain the barrier layer to the penetration of chemicals whereas the dermatomed slice includes all the tissue between the outer layers of the skin and the blood. Though the reason for the differences in distribution of the HD within the two preparations is not known, the contribution of the additional dermal tissue in the split thickness skin and the possible removal of surface lipid during heat separation of the epidermis cannot be discounted. Further investigation of these differences may prove valuable in understanding the initial events involved in HD absorption by the skin. The mechanism by which HD produces vesication remains unknown. However, it is known that HD is an alkylating agent (Papirmeister et al., 1991; Somani and Babu, 1989) which binds to nucleophilic sites within the cell, such as those found in DNA, RNA and protein, causing inter and intra strand cross links (Mol et al., 1989). HD preferentially binds to the N-7 of guanine, however, it will also bind to N-3 and O-6 of guanine as well as N-1 of adenine. A comparison between the rates at which these reactions occur to the rate at which HD is absorbed into the various layers of the skin would provide a valuable insight into the time after contamination the removal of any unreacted agent from the skin would be beneficial. These studies were conducted using a liquid agent challenge. The behaviour of HD in the skin after exposure to vapour remains to be defined. The penetration of vapour into the skin is similar to liquid, but since the agent partitions from the air into the stratum corneum the thermodynamic equilibria may be different and could generate a reservoir with different characteristics. In conclusion, the proportion of the applied dose which is present in the skin post exposure (extractable + unextractable) is generally less than 10%. The majority of any dose or contamination may be removed by simple swabbing or solvent washing (Chilcott et al., 2000, 2001). However, the amount remaining in the skin is in excess of the amount required to produce vesicant burns. Some of this agent can be extracted using organic solvents. Any physical or chemical means that is capable of removing this depot has the potential to decrease HD-induced pathology, even beyond the 2-min efficacy window (Sidell et al., 1997) of conventional decontamination procedures. The presence of an extractable reservoir of HD is therefore significant, if a suitable solvent could be incorporated into decontamination procedures for this purpose then burn intensity and period of incapacitation could conceivably be reduced,
1546
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
although to what extent is unclear. It is recommended that further studies are conducted both in vitro and in vivo to identify candidate solvents as well as to elucidate the relative merits of solvent extraction of HD following exposure. Conflict of interest statement The authors declare that there are no conflicts of interest. Acknowledgement This work was supported by the US Army Medical Research and Materiel Command under Contract # W91ZLK-04-P-1273. References Anderson, J.S., 1936. Report on the sensitivity to mustard gas of British troops in India and of Indian troops under cold weather conditions in northern India. CDRE (India) Report, 138. Axelrod, D.J., Hamilton, J.G., 1947. Radio-graphic studies of the distribution of lewisite and mustard gas in skin and eye tissue. Am. J. Pathol. 23, 389–441. Chilcott, R.P., Jenner, J., Carrick, W., Hotchkiss, S.A.M., Rice, P., 2000. Human skin absorption of bis-2-(chlorethyl)sulphide (sulphur mustard) in vitro. J. Appl. Toxicol. 20, 349–355.
Chilcott, R.P., Jenner, J., Hotchkiss, S.A., Rice, P., 2001. In vitro skin absorption and decontamination of sulphur mustard: comparison of human and pig-ear skin. J. Appl. Toxicol. 21, 279–283. Cullumbine, H., 1946. The mode of penetration of the skin by mustard gas. Br. J. Dermatol. Syphillis 58, 291–294. Harvey, W.G., Anderson, J.S., 1934a. Report on the sensitivity to mustard gas of British troops in India and of Indian troops. CDRE (India) Report, 110. Harvey, W.G., Anderson, J.S., 1934b. Length of service in India and the sensitivity to mustard gas of Europeans. CDRE (India) Report, 118. Henriques, F.C., Moritz, A.R., 1944. The mechanism of cutaneous injury by mustard gas. An experimental study using mustard prepared with radioactive sulfur Office of Scientific Research Division 9 OSRD-3620, Washington. Mol, M.A., van de Ruit, A.M., Kluivers, A.W., 1989. NAD+ levels and glucose uptake of cultured human epidermal cells exposed to sulfur mustard. Toxicol. Appl. Pharmacol. 98, 159–165. Papirmeister, B., Feister, A.J., Robinson, S.I., Ford, R.D., 1991. Medical Defense against Mustard Gas. CRC Press, Boca Raton, FL. Renshaw, B., 1947. Observations on the role of water in the susceptibility of human skin to injury by vesicant vapors. J. Inv. Dermatol., 75–85. Sidell, F.R., Urbanetti, J.S., Smith, W.J., Hurst, C.G., 1997. Vesicants. In: Sidell, F.R., Takafuji, E.T., Franz, D.R. (Eds.), Textbook of military medicine, Part I: Warfare, weaponry, and the casualty – medical aspects of chemical and biological warfare. Office of the Surgeon General at TMM Publications, Borden Institute, Walter Reed Army Medical Center, Washington, DC, pp. 197–228. Smith, H.W., Clowes, G.H.A., Marshall, E.K., 1919. On dichloroethylsulfide (mustard gas) IV The mechanism of absorption by the skin. J. Pharmacol. Exp. Therapeutics 13, 1–30. Somani, S.M., Babu, S.R., 1989. Toxicodynamics of sulphur mustard. Int. J. Clin. Pharmacol. Therapy Toxicol. 27, 419–435.
Contents lists available at ScienceDirect
Toxicology in Vitro journal homepage: www.elsevier.com/locate/toxinvit
The skin reservoir of sulphur mustard I.J. Hattersley a,*, J. Jenner a, C. Dalton a, R.P. Chilcott b, J.S. Graham c a b c
Dstl Biomedical Sciences Department, Biomedical Systems, Dstl Porton Down, Salisbury, Wiltshire, SP4 0JQ, UK Chemical Hazards and Poisons Division (HQ), C/o NRPB, Chilton, Didcot, Oxfordshire OX11 0RQ, UK US Army Medical Research Institute of Chemical Defense, 3100 Ricketts Point Road, Aberdeen Proving Ground, MD 21010-5425, USA
a r t i c l e
i n f o
Article history: Received 27 February 2008 Accepted 6 June 2008 Available online 14 June 2008 Keywords: Dermal Skin Reservoir Sulphur mustard
a b s t r a c t Studies of the percutaneous reservoir of sulphur mustard (HD) formed during absorption carried out during WWI and WWII are inconclusive. More recent studies have indicated that a significant amount of unreacted HD remains in human epidermal membranes during percutaneous penetration studies in vitro. The present study investigated the nature and persistence of the HD reservoir formed during in vitro penetration studies using dermatomed slices of human and pig skin (0.5 mm thick). Amounts of 14C-HD that (a) penetrated, (b) remained on the surface, (c) were extractable from and (d) remained in the skin after extraction were estimated by liquid scintillation counting (confirmed using GC–MS analysis). The results demonstrated that there is a reservoir of HD in human and pig skin for up to 24 h after contamination of the skin surface in vitro with liquid agent. At least some of this reservoir could be extracted with acetonitrile, and the amounts of extracted and unextracted HD exceed the amount required to produce injury in vivo by at least 20 fold. The study demonstrated the presence of a reservoir whether the skin was covered (occluded) or left open to the air (unoccluded). The study concluded that the extractable reservoir was significant in terms of the amount of HD required to induce a vesicant response in human skin. The extractable reservoir was at least 20 times the amount required per cm2 estimated to cause a response in all of the human population, as defined by studies carried out in human volunteers during the 1940s. Ó 2008 Elsevier Ltd. All rights reserved.
1. Introduction Sulphur mustard (HD) was first used as a chemical warfare agent in World War 1 (WW1) and is a potent vesicant, damaging all the surfaces of the body including the skin, eyes and respiratory tract. An important aspect of HD toxicology is the latency period after contact with liquid or vapour before symptoms are manifest, which can be hours or days depending upon dosage. This latency has led researchers to propose the existence of a depot or reservoir of HD within the skin, which has long been debated. Several historical studies have reported conflicting data and conclusions with regard to the existence of such a reservoir. A WW1 study reported by Smith and co-workers (1919), established that there was evidence for a substantial extractable reservoir, while studies conducted during the Second World War (WW2), summarised by Renshaw (1947), indicated that, even following massive liquid contamination, there was no appreciable extractable reservoir of unreacted agent. Smith et al. (1919) showed that HD contamination may be transferred from an exposed individual to an unexposed individual via direct contact between the two. A volunteer was exposed to HD * Corresponding author. E-mail address: [email protected] (I.J. Hattersley). 0887-2333/$ - see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.tiv.2008.06.002
vapour and, 30 min after exposure, the skin of a second volunteer was placed in direct contact with the contaminated area on the first. The second volunteer subsequently developed a burn. Thus, it was clear that as agent may be transferred from a contaminated individual to an uncontaminated individual, for at least thirty minutes after exposure, unreacted HD must be present on or near the surface of the skin. Occlusion (covering) of the site of contamination also increased intensity of the burn. It was clear from this work that a reservoir of HD existed in human skin for at least 45 min post exposure, some of which could be extracted with a suitably penetrating solvent (kerosene). Studies conducted in the 1940s by Henriques and Moritz (1944), Cullumbine (1946) and Axelrod and Hamilton (1947) are summarised in a review by Renshaw (1947). Renshaw concluded that, based on the work of these authors, ‘‘even after massive contamination with liquid HD, human skin contains no appreciable reservoir of unreacted HD” and therefore ‘‘all therapeutic procedures based upon the neutralisation of free penetrated HD must necessarily be valueless”, in direct conflict with the conclusions of Smith et al. (1919) 20 years earlier. Our review of the original studies summarised by Renshaw revealed some inconsistencies between the investigators observations which may explain the differences in the conclusion of Renshaw’s review and the observations of Smith et al. (1919).
1540
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
Henriques and Moritz (1944) could not extract radiolabelled HD from the skin at 2 min post exposure, indicating that HD was rapidly fixed within the skin. However, the skin was cleaned with an ether soaked swab to remove any surface contamination prior to analysis, which may also have removed any HD reservoir. Cullumbine (1946) used a histological stain which formed an insoluble black complex with free HD but not with degradation products such as thiodiglycol. This showed ‘‘free” unreacted HD in the skin at 15 min post exposure but not at 30 min leading to the conclusion that ‘‘free HD, after penetration, did not exist in the skin except in the epidermis”. Using micro-autoradiography with 35S-HD Axelrod and Hamilton (1947) showed that 35S label was present within both the epidermis and the dermis, although it was never determined whether the label detected was free HD or a degradation product. A more recent study, Chilcott et al. (2001), investigated the existence of a HD reservoir in exposed skin (heat separated epidermal membranes). 35S labelled HD was used to determine absorption rate and dose distribution of liquid HD through human epidermal membranes. ‘‘Free” HD was determined by placing treated skin sections in ethyl ether for up to 1 week, to ensure all extractable radioactivity had been removed, and sampling the ether for 35S. ‘‘Fixed” or bound HD was determined by placing the skin samples (following ether extraction) into a tissue solubiser (Soluene 350: quaternary ammonium hydroxide in toluene) to determine the extent of label remaining within the sample (confirmed by GC–MS analysis). The results clearly demonstrated that an extractable HD reservoir (up to 36% of applied dose) was present within the epidermal membranes for at least 24 h post exposure. There are clear inconsistencies between studies conducted around the time of the WWI and those conducted during and after WW2. Therefore, it is important to further evaluate the possibility of the presence of an extractable sulphur mustard reservoir within the skin of an exposed individual so that future therapeutic techniques may be optimised accordingly. The aims of this study were to identify and quantify the extent of the sulphur mustard reservoir present within the skin, in vitro, both as unextractable and extractable HD, under both occluded and unoccluded conditions, for pig and human skin. A range of time points were investigated, from 5 min to 24 h post exposure. These data provided an estimate as to the feasibility of removing/extracting HD from the skin of an individual following contamination with HD and thus the possibility of effective prevention of or reduction in the severity of incapacitation caused as a result of exposure to sulphur mustard by suitable decontamination techniques.
2. Materials and methods The synthesis, use and destruction of HD in this study was conducted in accordance with the Chemical Weapons Convention (1997) to which the UK is a signatory state. Radiolabelled HD was synthesised at Dstl Porton Down and had a radiochemical purity >98%. Chemical purity of unlabelled HD was >98%. Radiolabelled and cold agents were mixed in appropriate proportions to give a nominal activity of approximately 10 kBq ll 1. Liquid scintillation counting (LSC) materials (Soluene-350TM, emulsifier-safe cocktail and opaque plastic vials) were purchased from Perkin–Elmer (Chandler’s Ford, Hampshire). All other chemicals were analytical grade and were purchased from the Sigma Chemical Company (Poole, Dorset). Animals were used in accordance with the Animals (Scientific Procedures) Act 1986. Twelve weanling pigs (large white strain, weight range 8–10 kg) were obtained from the Institute of Animal Health (Newbury). Animals were given access to food and water ad
libitum. To obtain fresh pig skin animals were sedated with HypnovelÒ (Midazolam, 6 ml i.m., 5 mg ml 1), anaesthetised with 5% Isofluorane and culled with an overdose of EuthatalTM (sodium pentobarbitol, 6 ml i.v., 200 mg ml 1). The whole dorsal skin flank (approximately 40 30 cm) was close-clipped and excised from each animal. Subcutaneous fat was removed and the skin was immediately dermatomed to a nominal thickness of 500 lm using a Zimmer air dermatome (Zimmer Inc., Dover, OH, USA), prior to mounting in diffusion cells. Whole human abdominal skin (from elective reduction surgery) was obtained with full patient consent (age range 29–43) and was free from overt pathology. Donors were female Caucasians. The skin samples were stored at 25 °C for up to 2 months. After thawing, each skin sample was cleaned of subcutaneous fat and was dermatomed to a nominal thickness of 500 lm using a Zimmer air dermatome, prior to mounting in diffusion cells. 2.1. Percutaneous absorption Experiments were performed with Franz-type glass diffusion cells with an area available for diffusion of 2.54 cm2. Each diffusion cell consisted of a dermatomed piece of either human or pig skin forming a barrier between an upper (donor) and lower (receptor) chamber. The receptor chambers were filled with 50% aqueous ethanol (5 ± 0.5 ml). The skin surface temperature within each diffusion cell was controlled by mounting the sets of six diffusion cells on metal plates heated by circulating water. Pilot studies using thermocouples to measure the skin surface temperature established the water temperature required to maintain skin surface at 32 °C (±2 °C). The heated plates were mounted on stirring plates enabling magnetic followers to ensure the receptor solution of each cell was continuously stirred. The skin was allowed to equilibrate with the receptor media for 12 h prior to HD dosing. Experiments were designed such that each control and treatment group contained samples of skin from six different individuals. Results are there fore expressed as mean ± SD of ‘n’ individual donors. A preliminary series of experiments monitored the deposition of 14C-HD in the skin of pigs and humans up to 3 h post contamination, under both occluded and unoccluded conditions. The experimental protocol for the main study assessed time points of 1/2–24 h using 6 samples of skin from different donors at each time point. The studies conducted over 24 h consisted of unoccluded human skin, occluded pig skin and unoccluded pig skin. The effects of occlusion on human skin up to 24 h post exposure was not investigated due to a limited supply of human tissue, although results are available for occluded human skin up to 3 h post exposure. Topical dosing was performed by the direct application of undiluted HD (20 ll = 25.4 mg) to the skin using a positive displacement pipette. This amount of HD has been previously shown to cover the surface of a 2.54 cm2 diffusion cell by autoradiography (Chilcott et al., 2000). For diffusion cells being evaluated under occluded conditions the donor chambers had a 2 2 cm2 of parafilm placed onto an inert (perfluorinated) cream (AG-7) that covered the rim of each donor chamber. Following application of the agent the cells were left for the predetermined length of time. The parafilm from each occluded cell was removed and placed into 20 ml acetonitrile. Each skin surface was gently swabbed with dry cotton wool which was then immersed in 20 ml acetonitrile. Each skin sample was removed and weighed before being placed into 20 ml acetonitrile for 24 h, after which it was air dried to remove excess solvent before being reweighed and placed into 10 ml Soluene-350TM (it was determined, by a preliminary investigation, that immersion in acetonitrile for 24 h was sufficient to remove any ‘free’ HD from the skin sample).
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
The receptor fluid was retained from each receptor chamber for sampling. ‘‘Standard” samples were prepared simultaneously from known quantities of 14C-HD. The amount of radioactivity in each sample was measured using a Wallac 1409 DSA LS counter (using the manufacturer’s 14Cquench curve library) set to exclude single-photon (non-radioactive) events. Counter efficiency was assessed as (CPM/DPM) 100 and was 89–90%. The amount of radioactivity was converted to amount of HD by comparison to standard samples prepared and measured simultaneously. Samples of the acetonitrile extracts of skin and extracts of the cotton wool swabs were retained at 25 °C until analysed by GC–MS.
1541
and unoccluded conditions and similarly for pig skin under both conditions. In these experiments the amount of 14C-HD extracted from the tissue was generally higher than that which remained in the tissue. These experiments showed that there was a reservoir in the skin at three hours after contamination and indicated that experiments were required to define the distribution of 14C-HD for 24 h after contamination. In the 3 h studies there was a larger reservoir of extractable HD from human skin (4–8%) than pig skin (0.4–1.5%) under occluded conditions (p < 0.05, 2 way ANOVA), however this difference was not significant under unoccluded conditions. 3.3. Twenty four hour experiments
2.2. Quantification of samples by GC–MS HD samples were analysed on a DB-WAX column (25 m 0.2 mm I.D., 0.2 lm film thickness) at 90 °C (1 min), 90– 260 °C (linear gradient at 20 °C/min) and 260 °C (5.5 min) in an Agilent Technologies 6890 gas chromatograph. Samples were eluted with a helium (99.999%) carrier gas (13 psi head pressure at 250 °C; splitless) and quantified using a flame photometric detector set-up for phosphorus and sulphur detection. Confirmation of randomly selected positive samples was made using an Agilent Technologies 6890 gas chromatograph with a 5973 inert mass selective detector in electron impact mode (GC– MS(EI)). This system chromatograph used a DB–5MS column (25 m 0.2 mm I.D., 0.33 lm film thickness), helium carrier, 17 psi splitless injector at 250 °C, and a elution temperature gradient of 40 °C (1 min), 40–160 °C (linear gradient at 20 °C/min) and 260– 320 °C (linear gradient at 20 °C/min) and 320 °C (2.0 min). The mass spectrometer used an ion source temperature of 230 °C, ionisation: EI (70 eV), ionisation current: 60 lA and a mass scanning range of 40–550 m/z. Data acquisition was started after a 4 min solvent delay post injection. Identification of sulphur mustard was determined against authentic sulphur mustard analytical standard spectra and the NIST02 mass spectral library database. Quantitation of sulphur mustard in the acetonitrile extracts was determined against authentic sulphur mustard analytical standards in the selective ion monitoring mode utilising: 109, 111, 158 and 160 m/z over a calibration range of 1–10 lg/ml. 2.3. Statistical analysis Where required statistical significance was determined using 1 way and 2 way ANOVA and 1 sample t-tests and were carried out using Graphpad Prism (Graphpad Prism version 4.00 for windows, GraphPad Software, San Diego, CA) analysis of amounts of HD. Differences were judged as significant at p < 0.05.
3. Results 3.1. Diffusion cell experiments The first series of experiments measured the distribution of 14CHD in the skin over 3 h post application of 14C-HD to the surface, under both occluded and unoccluded conditions. The second was conducted in an identical fashion to the first, except that distribution was monitored out to 24 h post exposure. 3.2. Three hour experiments (data not shown) In the first 3 h after application of 14C-HD the majority of the agent remained on the surface of the skin and could be removed by swabbing. This was true for human skin under both occluded
Unoccluded experiments demonstrated that the majority of the applied agent (>70%) remained on the surface of skin for up to three hours post exposure. However, the amount recoverable from the surface dropped steadily to less than 5% of the applied dose by 8 h in human skin (Fig. 1a) and by 12 h in pig skin (Fig. 2a). Under occluded conditions (pig only) the amount recoverable from the surface of the skin remained almost constant throughout 24 h (70–90%) (Fig. 3a). In these experiments the amount of agent which was extractable from the skin was lower than the amount which remained unextractable for both pig and human skin, under both occluded and unoccluded conditions. Even so, the amount of extractable HD was significantly greater than that required to cause vesication at certain time points, under unoccluded conditions in both human and pig skin (p < 0.05, 1 sample t-test, see Figs. 1 and 2). Total amount of HD in the tissue (extracted + unextracted) was 1–4% of the applied dose over the 24 h for human skin, under unoccluded conditions (Fig. 1b). The amount of agent in the pig tissue was higher under unoccluded conditions, ranging from 5% to 20% (Fig. 2b). The amount present in the pig tissue under occluded conditions was between 3% and 5% of the applied dose throughout the 24 h (Fig. 3b). Total recoveries of HD from diffusion cell systems were higher under occluded conditions than unoccluded conditions between 4 and 24 h post exposure for pig (Figs. 2 and 3a). This is due to evaporation from the unoccluded cells. After 4 h less than 50% of the initial applied dose could be recovered under unoccluded conditions for both pig and human skin, however, under occluded conditions in excess of 80% of the applied dose could be recovered at 24 h post exposure. The fraction of the applied dose which could be extracted from the skin using acetonitrile was less than 10% in all cases (human and pig, occluded and unoccluded conditions). Extractable fractions were similar in pig and human skin during the 24 h studies, generally being in the region of about 1% of the initial applied dose with no significant change over time (p > 0.05, 1 way ANOVA). The fraction of the applied dose not extracted from the skin was 0–3%, although in the case of the 24 h unoccluded pig skin study this was as high as 10–15% for the first 4 h, before dropping to approximately 4% by 24 h. The proportion of applied HD which had penetrated the skin at the end of the experiment was less than 2.5% for all 3 h studies (pig and human, occluded and unoccluded). Over 24 h the amount penetrated under unoccluded conditions was comparable for both pig and human skin (1.5% and 2%, respectively), however, it was as high as 7% for the pig skin under occluded conditions. Occluded conditions exhibited a higher level of penetration for both human and pig skin, being approximately double that of the unoccluded conditions over the 3 h time period and in excess of 4 fold higher over 24 h. The weight lost by skin samples upon extraction by acetonitrile was constant across the whole of this study, being generally in the region of 20–25% of the initial weight of the skin in most cases.
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
surface
30
penetrated
Applied Dose
25
Amount HD recovered (mg)
120
extractable unextractable
A
100
20
80
15
60
10
40
5
20
0
3
0 0.5
6
12
18
24
12
B
10
Recovered HD (% applied dose)
1542
8
2
6
* 1
* 4
* *
* *
*
2
* 0
0 0.5
6
12
18
24
Time (hours) Fig. 1. Recovery of HD at various time points up to 24 h after application of 20 lL of liquid to the surface of human skin under unoccluded conditions. Graph A shows the amount of HD, and percentage of applied dose, recovered from the receptor fluid i.e. penetrated the skin (cross hatched), remaining on the surface of the skin recovered by swabbing (solid), extracted from the skin with acetonitrile (open) and not extracted (hatched). Graph B shows only extracted and unextracted material for clarity. Points are mean ± standard deviation: n = 6 diffusion cells. The broken horizontal line on graph B indicates the dose (10 lg cm 2) at which all the human population shows a reaction to HD (Harvey and Anderson, 1934a,b). Asterisks above bars indicates that the amount of HD extracted significantly exceeds this dose (p < 0.05, 1 sample t-test).
There were no apparent differences between pig and human skin and between occluded and unoccluded conditions in this respect. 3.4. Chemical analysis of HD in skin samples Samples from the acetonitrile extracts of the surface swabs and skin from the unoccluded experiments were subject to GC–MS analysis. The results correlated well with the estimates of amounts of HD recovered made using radiolabel (correlation coefficients over the four experiments r2 = 0.858 ± 0.087; mean ± SD). The majority of the 14C label recovered from the surface of the skin was identified as HD, with only very small percentages (generally less than 1%) being identified as thiodiglycol, the hydrolysis product of HD. Of the material extracted from the skin, a large percentage remained as 14C HD, however, levels were higher than those from the surface swabs (means were in the range 18–53% from human and 8–67% from pig).
4. Discussion This study has shown that there is a reservoir of HD in human and pig skin for up to 24 h after contamination of the skin surface in vitro with liquid agent. At least some of this reservoir can be extracted with an organic solvent (acetonitrile) and at certain time points the amounts of extracted and unextracted HD significantly exceed the amount required to produce injury in vivo. This is supported by the evidence of the GC/MS analysis which identified free HD in the samples of acetonitrile used for extraction. 4.1. Recovery of HD after contamination of the skin surface The results of this study indicate that under occluded conditions the majority of an applied dose of liquid HD may be recovered from the surface of the skin (pig or human) up to 24 h post exposure by swabbing with cotton wool. Similarly, under unoccluded condi-
1543
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
surface
30
unextractable penetrated
Applied Dose
100
20
80
15
60
10
40
5
20
0
0 6
12
18
24
6
24
B 5
20
4
16
3
12
*
2
*
*
Recovered HD (% applied dose)
Amount HD recovered (mg)
25
120
extractable
A
8
* *
1
4
0
0 6
12
18
24
Time (hours) Fig. 2. Recovery of HD at various time points up to 24 h after application of 20 lL of liquid to the surface of pig skin under unoccluded conditions. Graph A shows the amount of HD, and percent applied dose, recovered from the receptor fluid i.e. penetrated the skin (cross hatched), remaining on the surface of the skin recovered by swabbing (solid), extracted from the skin with acetonitrile (open) and not extracted (hatched). Graph B shows only extracted and unextracted material for clarity. Points are mean ± standard deviation: n = 6 diffusion cells. The broken horizontal line on graph B indicates the dose (10 lg cm 2) at which all the human population shows a reaction to HD (Harvey and Anderson, 1934a,b). Asterisks above bars indicates that the amount of HD extracted significantly exceeds this dose (p < 0.05, 1 sample t-test).
tions, a combination of evaporation and swabbing may account for the bulk of an applied dose. Thus, in excess of 80% of an applied dose of liquid HD either evaporated or remained on the skin surface and was unavailable for diffusion, in agreement with previous studies. Therefore, 20% or less of the initial percutaneous dose of liquid HD will be available to penetrate the skin and cause vesication. Such an absorbed dose is more than sufficient to cause vesication. Studies reported by Anderson (1936) and Harvey and Anderson (1934a,b) established that in a population of Indian and British troops all the subjects produced a skin reaction to a dose of 10 lg cm2. Previous studies in this laboratory have demonstrated that an applied dose of liquid HD (20 ll) will spread to fill the area available for diffusion (2.54 cm2) within 1 h of application (Chilcott et al., 2000). Thus, in the diffusion cells used in this study, a minimum vesicating dose would equate to 25 lg, which is much lower than the mg quantities of HD absorbed in the present study. Although the amounts extracted and unextracted at certain time points were significantly larger (p < 0.05) than the dose caus-
ing vesication in all exposed men (10 lg cm 2, Andersen et al., 1936), the variability in the data was too high to be able to draw meaningful conclusions about the kinetics of these reservoirs. This study investigated the distribution of HD between the surface, the skin tissue and the receptor chamber of the diffusion cells over the first 3 h, and in a subsequent series of experiments over 24 h after contamination of the skin surface. In the 3 h experiments the proportion of the applied dose which was extractable from the skin with acetonitrile was 4–10% (1–2.5 mg) in human skin, under occluded and unoccluded conditions. For pig skin, however, the extractable dose was lower, being approximately 1.5% (0.375 mg). This was significant for occluded conditions (p < 0.05) but not unoccluded conditions. The extractable reservoir of HD did not change significantly over the 24 h time course. Over all the time points up to 24 h a maximum of 5% of the applied dose (1.25 mg) was extractable from the unoccluded pig skin which contrasts with previous results from this laboratory reported by Chilcott et al. (2000), who used human epidermal membranes
1544
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
30
A
120
penetrated
Applied Dose
unextractable extractable
25
100
20
80
15
60
10
40
5
20
0
0 6
12
18
24
2
8
B 6
1
Recovered HD (% applied dose)
Amount HD recovered (mg)
surface
4
2
0
0 6
12
18
24
Time (hours) Fig. 3. Recovery of HD at various time points up to 24 h after application of 20 lL of liquid to the surface of pig skin under occluded conditions. Graph A shows the amount of HD recovered from the receptor fluid i.e. penetrated the skin (cross hatched), remaining on the surface of the skin recovered by swabbing (solid), extracted from the skin with acetonitrile (open) and not extracted (hatched). Graph B shows only extracted and unextracted material for clarity. Points are mean ± standard deviation: n = 6 diffusion cells. The broken horizontal line on graph B indicates the dose (10 lg cm 2) at which all the human population shows a reaction to HD (Harvey and Anderson, 1934a,b). The extracted amounts of HD were not significantly greater than this dose.
and demonstrated an ether extractable reservoir of up to 36% of the applied dose. There could be several reasons for this disparity. The method of removing HD from the surface of the skin was slightly different. Chilcott et al. (2000) used a series of ethanol washes (5) to remove any remaining HD at the end of the experiment however, in this study cotton wool swabs were used, which may have removed a portion of the stratum corneum due to the mechanical action of the swabbing process (thus removing a portion extractable reservoir present in the upper portion of the skin, resulting in an under estimation of the extractable material). Similarly, the use of a solvent to remove HD from the skin surface could also remove a portion of the reservoir. Another difference between the two studies was the solvent used to extract the HD from the skin after the surface had been wiped. Acetonitrile was used in this study whereas diethyl ether was used by Chilcott et al. (2000). This may result in different extraction efficiencies due to the different partition coefficients of the HD between the solvents and the skin. Since it is known that
HD rapidly hydrolyses in aqueous environments, any unreacted HD in the skin after longer than a few minutes must be dissolved in the lipid domains of the tissue. In the skin this would be the specialised lipids of the stratum corneum and/or the lipids of the cell membranes in the viable epidermis and superficial dermis. This is indirectly shown by the fact that organic solvents are effective at removing the absorbed HD from the skin. Thus, Smith et al. (1919) have demonstrated that HD can be removed from the cutaneous reservoir using kerosene (paraffin). Chilcott et al. have shown that diethyl ether can remove some of this HD and this study has shown that acetonitrile can also remove a significant amount of HD from the skin following a neat liquid dose. Despite the limitation of all these studies it is clear that there is a solvent extractable reservoir present in the skin up to at least 24 h post exposure, which contrasts with the conclusions of Renshaw (1947), based on the work reported by Henriques and Moritz (1944), Cullumbine (1946) and Axelrod and Hamilton (1947). It is also important to note that the solvent extraction removed more
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546 Table 1 Percentage loss in weight of skin produced by extracting with acetonitrile in the 3 h study Time/mins
5 10 30 60 120 180
Human
Pig
Occluded
Unoccluded
Occluded
Unoccluded
22.84 ± 7.54 16.46 ± 6.19 19.93 ± 8.62 23.74 ± 7.87 20.60 ± 4.71 17.72 ± 5.49
23.15 ± 4.93 22.59 ± 6.15 24.23 ± 6.07 21.65 ± 4.97 23.77 ± 8.99 21.98 ± 8.72
23.55 ± 10.61 15.56 ± 9.11 27.83 ± 8.08 7.39 ± 3.48 40.91 ± 9.72 37.13 ± 8.27
23.15 ± 4.93 22.59 ± 6.15 24.23 ± 6.07 21.65 ± 4.97 23.77 ± 8.98 21.98 ± 8.72
Data are expressed as mean ± SD of 4–6 pieces of skin.
Table 2 Percentage loss in weight of skin produced by extracting with acetonitrile in the 24 h study Time/hours
0.5 1.0 2.0 4.0 6.0 8.0 10.0 12.0 16.0 20.0 24.0
Human
Pig
Unoccluded
Occluded
Unoccluded
33.80 ± 10.66 27.83 ± 6.54 29.06 ± 5.83 25.48 ± 3.16 31.31 ± 5.18 28.54 ± 3.47 n/a 26.00 ± 3.33 22.23 ± 3.31 20.22 ± 11.93 16.79 ± 11.97
n/a n/a n/a 22.39 ± 7.10 10.32 ± 11.59 18.90 ± 5.45 17.10 ± 7.66 25.88 ± 21.42 n/a n/a 11.41 ± 5.60
19.42 ± 3.32 16.72 ± 3.33 21.27 ± 5.03 18.62 ± 3.88 22.56 ± 4.54 17.05 ± 4.01 n/a 13.34 ± 6.19 16.27 ± 7.08 13.60 ± 5.49 11.72 ± 4.05
than the HD from the skin, since the average weight of the skin was reduced by 7–40% by the extraction process (Tables 1 and 2). This reduction in weight probably represents lipid material, water and ethanol co-extracted with the HD, but at this time the precise identity of the co-extractant is not known. Clearly, further work is required to determine the best solvent system to extract HD from the skin, and to determine the fraction, if any, of the HD remaining in the tissue in this study that was covalently linked to tissue constituents. 4.2. Fate of the absorbed fraction The in vitro preparation used in the present study is only able to reproduce some of the conditions prevalent within the skin during percutaneous absorption of HD. It has been shown previously that the physical processes of partitioning into the skin compares well to that measured by surface disappearance measurement in man (Chilcott et al., 2000). This preparation does not seek to maintain the skin in a metabolically viable state and the results should be viewed in this light. Since the major route of metabolism for HD is spontaneous hydrolysis, the absence of other metabolic pathways is not likely to materially affect the conclusions drawn from this study. The effects of perfusing the skin with blood is modelled here by ensuring the receptor fluid does not become saturated with HD (HD hydrolysis rapidly in the receptor fluid maintaining an effective ‘‘sink” for the parent molecule). This in vitro system therefore effectively models a maximally perfused area of skin. Similarly, the use of 50:50 ethanol/water as the receptor fluid ensures that the partitioning of the HD into the receptor fluid from the skin is not rate limiting. The fraction of the applied dose absorbed into the skin either penetrated into the receptor chamber (equivalent to the blood in vivo), or remained in the skin. Of the fraction remaining in the skin, when it is removed from the diffusion cell, part was extractable
1545
with acetonitrile. The fraction of the absorbed dose which remained in the skin after extraction contained the agent, some of which has been previously demonstrated to be covalently bound to the tissue, part of which would cause vesication in vivo. The 1–3% of applied dose that remained non-extractable in the present study between 6 and 24 h is consistent with previous studies (Chilcott et al., 2000) which have shown that in the region of 2% of an applied dose of HD is unextractable, and normally regarded as fixed in the skin. This small proportion which remains bound in the skin even following extraction with a solvent is most likely that which is directly responsible for vesication and thus development of a burn. Two percent of an applied dose of 20 ll (0.4 ll or 0.51 mg) equates to at least 20 times the minimum vesicating dose of 10 lg cm 2 for HD determined by Anderson (1936) and Harvey and Anderson (1934a) by applying HD in benzene to the skin of humans. Previous studies in this laboratory have shown a larger proportion of the absorbed dose is retained in the skin over a 24 h period (Chilcott et al., 2000, 2001). The earlier studies were conducted with epidermal membranes which had been prepared by heat separation from human skin, whereas the current study has used split thickness slices of skin prepared using a dermatome. Each of these preparations has different merit as a model of how the skin behaves in vivo. The epidermal membranes isolate the part of the skin generally believed to contain the barrier layer to the penetration of chemicals whereas the dermatomed slice includes all the tissue between the outer layers of the skin and the blood. Though the reason for the differences in distribution of the HD within the two preparations is not known, the contribution of the additional dermal tissue in the split thickness skin and the possible removal of surface lipid during heat separation of the epidermis cannot be discounted. Further investigation of these differences may prove valuable in understanding the initial events involved in HD absorption by the skin. The mechanism by which HD produces vesication remains unknown. However, it is known that HD is an alkylating agent (Papirmeister et al., 1991; Somani and Babu, 1989) which binds to nucleophilic sites within the cell, such as those found in DNA, RNA and protein, causing inter and intra strand cross links (Mol et al., 1989). HD preferentially binds to the N-7 of guanine, however, it will also bind to N-3 and O-6 of guanine as well as N-1 of adenine. A comparison between the rates at which these reactions occur to the rate at which HD is absorbed into the various layers of the skin would provide a valuable insight into the time after contamination the removal of any unreacted agent from the skin would be beneficial. These studies were conducted using a liquid agent challenge. The behaviour of HD in the skin after exposure to vapour remains to be defined. The penetration of vapour into the skin is similar to liquid, but since the agent partitions from the air into the stratum corneum the thermodynamic equilibria may be different and could generate a reservoir with different characteristics. In conclusion, the proportion of the applied dose which is present in the skin post exposure (extractable + unextractable) is generally less than 10%. The majority of any dose or contamination may be removed by simple swabbing or solvent washing (Chilcott et al., 2000, 2001). However, the amount remaining in the skin is in excess of the amount required to produce vesicant burns. Some of this agent can be extracted using organic solvents. Any physical or chemical means that is capable of removing this depot has the potential to decrease HD-induced pathology, even beyond the 2-min efficacy window (Sidell et al., 1997) of conventional decontamination procedures. The presence of an extractable reservoir of HD is therefore significant, if a suitable solvent could be incorporated into decontamination procedures for this purpose then burn intensity and period of incapacitation could conceivably be reduced,
1546
I.J. Hattersley et al. / Toxicology in Vitro 22 (2008) 1539–1546
although to what extent is unclear. It is recommended that further studies are conducted both in vitro and in vivo to identify candidate solvents as well as to elucidate the relative merits of solvent extraction of HD following exposure. Conflict of interest statement The authors declare that there are no conflicts of interest. Acknowledgement This work was supported by the US Army Medical Research and Materiel Command under Contract # W91ZLK-04-P-1273. References Anderson, J.S., 1936. Report on the sensitivity to mustard gas of British troops in India and of Indian troops under cold weather conditions in northern India. CDRE (India) Report, 138. Axelrod, D.J., Hamilton, J.G., 1947. Radio-graphic studies of the distribution of lewisite and mustard gas in skin and eye tissue. Am. J. Pathol. 23, 389–441. Chilcott, R.P., Jenner, J., Carrick, W., Hotchkiss, S.A.M., Rice, P., 2000. Human skin absorption of bis-2-(chlorethyl)sulphide (sulphur mustard) in vitro. J. Appl. Toxicol. 20, 349–355.
Chilcott, R.P., Jenner, J., Hotchkiss, S.A., Rice, P., 2001. In vitro skin absorption and decontamination of sulphur mustard: comparison of human and pig-ear skin. J. Appl. Toxicol. 21, 279–283. Cullumbine, H., 1946. The mode of penetration of the skin by mustard gas. Br. J. Dermatol. Syphillis 58, 291–294. Harvey, W.G., Anderson, J.S., 1934a. Report on the sensitivity to mustard gas of British troops in India and of Indian troops. CDRE (India) Report, 110. Harvey, W.G., Anderson, J.S., 1934b. Length of service in India and the sensitivity to mustard gas of Europeans. CDRE (India) Report, 118. Henriques, F.C., Moritz, A.R., 1944. The mechanism of cutaneous injury by mustard gas. An experimental study using mustard prepared with radioactive sulfur Office of Scientific Research Division 9 OSRD-3620, Washington. Mol, M.A., van de Ruit, A.M., Kluivers, A.W., 1989. NAD+ levels and glucose uptake of cultured human epidermal cells exposed to sulfur mustard. Toxicol. Appl. Pharmacol. 98, 159–165. Papirmeister, B., Feister, A.J., Robinson, S.I., Ford, R.D., 1991. Medical Defense against Mustard Gas. CRC Press, Boca Raton, FL. Renshaw, B., 1947. Observations on the role of water in the susceptibility of human skin to injury by vesicant vapors. J. Inv. Dermatol., 75–85. Sidell, F.R., Urbanetti, J.S., Smith, W.J., Hurst, C.G., 1997. Vesicants. In: Sidell, F.R., Takafuji, E.T., Franz, D.R. (Eds.), Textbook of military medicine, Part I: Warfare, weaponry, and the casualty – medical aspects of chemical and biological warfare. Office of the Surgeon General at TMM Publications, Borden Institute, Walter Reed Army Medical Center, Washington, DC, pp. 197–228. Smith, H.W., Clowes, G.H.A., Marshall, E.K., 1919. On dichloroethylsulfide (mustard gas) IV The mechanism of absorption by the skin. J. Pharmacol. Exp. Therapeutics 13, 1–30. Somani, S.M., Babu, S.R., 1989. Toxicodynamics of sulphur mustard. Int. J. Clin. Pharmacol. Therapy Toxicol. 27, 419–435.