Laser ablation inductively coupled plasma mass spectrometry as a novel clinical imaging tool to detect asbestos fibres in malignant mesothelioma.

Rationale: Malignant pleural mesothelioma is an extremely aggressive and incurable malignancy associated with prior exposure to asbestos fibres. Difficulties remain in relation to early diagnosis, notably due to impeded identification of asbestos in lung tissue. This study describes a novel laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) imaging approach to identify asbestos within mesothelioma models with clinical significance. Methods: Human mesothelioma cells were exposed to different types of asbestos fibres and prepared on plastic slides for LA-ICP-MS analysis. No further sample preparation was required prior to analysis, which was performed using an NWR Image 266 nm laser ablation system coupled to an Element XR sector-field ICP mass spectrometer, with a lateral resolution of 2 μ m. Data was processed using LA-ICP-MS ImageTool v1.7 with the final graphic production made using DPlot software. Results: Four different mineral fibres were successfully identified within the mesothelioma samples based on some of the most abundant elements that make up these fibres (Si, Mg and Fe).


Rationale:
Malignant pleural mesothelioma is an extremely aggressive and incurable malignancy associated with prior exposure to asbestos fibres. Difficulties remain in relation to early diagnosis, notably due to impeded identification of asbestos in lung tissue. This study describes a novel laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) imaging approach to identify asbestos within mesothelioma models with clinical significance.
Methods: Human mesothelioma cells were exposed to different types of asbestos fibres and prepared on plastic slides for LA-ICP-MS analysis. No further sample preparation was required prior to analysis, which was performed using an NWR Image 266 nm laser ablation system coupled to an Element XR sector-field ICP mass spectrometer, with a lateral resolution of 2 μm. Data was processed using LA-ICP-MS ImageTool v1.7 with the final graphic production made using DPlot software.
Results: Four different mineral fibres were successfully identified within the mesothelioma samples based on some of the most abundant elements that make up these fibres (Si, Mg and Fe). Using LA-ICP-MS as an imaging tool provided information on the spatial distribution of the fibres at cellular level, which is essential in asbestos detection within tissue samples. Based on the metal counts generated by the different types of asbestos, different fibres can be identified based on shape, size, and elemental composition. Detection of Ca was attempted but requires further optimisation.
Conclusions: Detection of asbestos fibres in lung tissues is very useful, if not necessary, to complete the pathological dt9iagnosis of asbestos-related malignancies in the medicolegal field. For the first time, this study demonstrates the successful application of LA-ICP-MS imaging to identify asbestos fibres and other mineral fibres within mesothelioma samples. Ultimately, high-resolution, fast-speed LA-ICP-MS analysis has the potential to be integrated into clinical workflow to aid earlier detection and stratification of mesothelioma patient samples.

| INTRODUCTION
Malignant pleural mesothelioma (MPM) is a rare, extremely aggressive and incurable malignancy associated with occupational and environmental exposure to asbestos and other mineral fibres (MF). 1 MPM has a long latency period with an average of 30-60 years, with a median survival time of 6-12 months and a 5-year survival of <5%. 2 Despite asbestos being banned in most Western countries, the global incidence of MPM is expected to increase over the next two decades, mainly due to the long latency period, continuous mining and usage in certain countries, as well as environmental exposure. 1 Difficulties remain in relation to presymptomatic diagnosis, particularly due to lack of validated biomarkers and impeded identification of asbestos fibres in lung tissue. 2 Being able to accurately identify these MF within samples is not only essential in aiding early diagnosis of MPM, but it also plays a key role in linking this diagnosis to asbestos exposure, which has high implications in legal, social and political matters. 3 Asbestos is an umbrella term that comprises various polyfilamentous silicate minerals, which, once broken into fibrils, can be easily inhaled, in most cases lodging into the lining of the lungs and triggering a chronic inflammatory cascade. 2 Asbestos consists of two main classes; the serpentines (of which chrysotile is the most common type) and the amphiboles, including crocidolite, amosite, tremolite, anthophyllite and actinolite. 4 The potency of these MF to induce asbestos-related diseases (ARD), such as asbestosis, lung cancer and MPM, varies based on their differences in chemical and physical structures, presented briefly in Table 1.
By geological classification, asbestos fibres are defined as having aspect ratio ≥3:1, length ≥5 μm, and width <0.25 μm. 6 This does not account for cleavage fragments, or high potency shorter fibres. 9 This asbestos classification was created to facilitate the counting of the MF using standard microscopy methods and was not based on the potency of these segments or shorter fibres to induce ARD. 10 According to the latest guidelines published by the British Thoracic Society, the current investigation workflow involves an urgent chest X-ray of symptomatic patients, followed by a pleural evaluation computerised tomography (CT) scan, immunohistochemistry of biopsies or cytology-type specimens, and establishing the prospect of occupational exposure or para exposure to asbestos fibres. 11 However, it is essential to detect MF in lung tissues to complete the pathological diagnosis of MPM in correlation to asbestos exposure. The routine quantification method uses phase contrast microscopy to count only the fibres that adhere to the definition mentioned above. Therefore, current test methods exclude certain types of MF such as cleavage fragments, short or thin fibres, or naturally occurring MF (e.g. erionite), despite all of these being previously reported to cause ARD. 9,12,13 More established methods of asbestos identification and quantification, such as scanning electron microscopy (SEM) and transmission electron microscopy (TEM), generally employ chemical digestion of the tissues in order to eliminate the organic matrix, thus losing all the spatial information offered by imaging methods such as optical and polarised light microscopy. 14 However, these imaging methods cannot differentiate between asbestiform and nonasbestiform fibres and fail to detect low concentrations of MF or fibres that are less than 0.20 μm in diameter. 3 Moreover, methods of differentiating MF based on surface charge are not relatable since non-asbestiform amphiboles are virtually identical to asbestiform amphibole fibres in morphology, crystal structure, refractive index, and chemical composition. 9 Given the similarities in chemical structures between the MF (i.e. silicates), and also the proportional discrepancies of main elements such as Fe, Ca and Mg between each fibre type, elemental analysis has gained popularity in recent years compared with optical T A B L E 1 Classification, common names and physicochemical properties of the mineral fibres used in this work Long, straight, coarse fibres. Can be flexible.

Non-asbestiform
Wollastonite 8 Unspecified CaSiO 3 Soluble in concentrated HCl. Composed of chains of silica tetrahedra connected side by side through calcium.
Bladed crystal masses, single crystals can show an acicular particle shape and usually it exhibits a white colour methods. One of the first reports used an in-air micro-particle induced X-ray emission (in-air micro-PIXE) system to identify the location of asbestos bodies in lung tissue sections. 15 However, this has only been achieved with limited spatial resolution. Sub-micrometer lateral resolution has been reported in a study by Pascolo et al who employed soft X-ray imaging and X-ray fluorescence (XRF) microscopy to investigate elemental lateral distribution of asbestos bodies in lung tissue. 16 Given the complexity of the techniques used in the study, the limitations revolve around the applicability in a clinical setting and it is presented more as a research tool on chemical interactions between fibres and cells.
Laser ablation inductively coupled plasma mass spectrometry (LA-ICP-MS) is one of the most versatile techniques for the analysis of trace elemental distribution due to its exceptional limits of detection for a vast range of metallic analytes (μg g −1 ), ease of sample preparation, excellent lateral resolution and relatively good acquisitions times, making it the ideal candidate for clinical settings.
Furthermore, due to technical advancements in instrumentation, cellular analysis with a relatively short time of acquisition is possible.
It has gained popularity in recent years and it is continuing to develop through the introduction of novel chamber designs and geometries.
For example, earlier designs allowed for resolutions between 5 and 100 μm 17 compared with recent advancements capable of cellular analysis with 1-2 μm spot size 18,19 or even sub-micrometre spot diameter recently reported by Vanhaecke et al. 20 Given these features, LA-ICP-MS imaging is an established technique in monitoring elemental distribution in certain neurodegenerative conditions such as Alzheimer's disease 21 and Parkinson's disease. 22 LA-ICP-MS has also been employed as an imaging tool in cancer research 23 and as a drug monitoring tool. 24 To our knowledge, this is the first application of LA-ICP-MS imaging to spatially resolve asbestos and other MF at cellular resolution. We aimed to identify some of the most common MF, amosite, chrysotile and crocidolite, as well as a non-asbestiform control, wollastonite, within MPM models using a laser ablation system that consists of a dual-volume ablation cell with an integrated ICP torch, which allows for an increase in absolute sensitivity and faster analysis times. 18

| Mineral fibres
The

| LA-ICP-MS imaging
Analysis was performed using an NWR Image 266 nm laser ablation system (Elemental Scientific Lasers, Bozeman, MT, USA) coupled to an Element XR sector-field ICP mass spectrometer (Thermo Scientific, Bremen, Germany). 25 The laser ablation system used consists of a two-volume ablation cell and dual concentric injector (DCI). 18

| RESULTS AND DISCUSSION
To assess the potential of LA-ICP-MS to be ultimately integrated within the clinical setting as a MF imaging tool for patients suspected of MPM, preliminary studies had first to be carried out on cell lines exposed to MF.
In this work, chrysotile, amosite, crocidolite and wollastonite were selected on the panel of MF to be detected based on the metal content. The panel was carefully selected to include both the most common type of asbestos (i.e. chrysotile), 2 and the MF with the highest carcinogenicity and bioresistance (i.e. crocidolite). 6 The MF were identified by LA-ICP-MS based on the Fe, Mg and Si content.
Method optimisation included the selection of the right isotopes for F I G U R E 1 NCI-H28 cells cultured in a fibre-free environment. Left column shows the microscope perspective of the areas prior to ablation. Middle column presents the elemental distribution as analysed by LA-ICP-MS imaging. Intensity bar was adjusted in the right-hand column to normalise against the cellular background. A, Low 57 Fe distribution within the mesothelioma cells; B, high 24 Mg intensity counts ; C, the significantly low overall 29   also used as a matrix (data not presented). There was no significant difference in either 24 Mg, 57 Fe or 29 Si intensity signals generated by the two cellular models. The elemental intensity generated by the NCI-H28 cells is presented in Figures 1A-1C. Note the signal F I G U R E 3 NCI-H28 cells cultured with amosite fibres. Left column shows the microscope perspective of the areas prior to ablation. Areas with fibres visible using optical microscopy were selected for proof of concept. Middle column presents the elemental distribution as analysed by LA-ICP-MS imaging. The intensity bar was adjusted in the right-hand column to focus on the fibres. As expected, higher 57 Fe distribution is presented in (A) than for serpentine fibres. Note the shorter fibres and fibre amosite fibre fragments spread all over the sample as shown using LA-ICP-MS imaging. B, The 24 Mg counts within the sample. Note that the engulfed amosite fibres are not visible in the microscopic image. C, 29 Si distribution within the matrix and amosite fibres [Color figure can be viewed at wileyonlinelibrary.com] intensity generated by 24 Mg (Figure 1B), present at detectable and relatively homogeneous levels within the biological matrix, making it a suitable choice to aid correlation of the LA-ICP-MS and microscopic images.

| Chrysotile
Chrysotile is known for its long, curled appearance and short biopersistence in the lungs. 5 The slightly decreased carcinogenicity of chrysotile compared with that of other asbestos fibres is mainly reported by many researchers. 34 Given the sample preparation in this proof of concept study, the localisation of the fibres varies between the samples, supporting the hypothesis that LA-ICP-MS can ablate and detect MF regardless of the positioning. One such example is presented in Figure 3B where a short amosite fibre has been engulfed by the cell, hindering visualisation simply by light microscopy.

| Crocidolite
Importantly, the difference in nominal composition between the elements that make up each of the MF selected in the panel is mirrored in the intensity generated by analysis. For instance, the Fe content in chrysotile presented earlier in Figure 2A compared with crocidolite ( Figure 4A) brings the work a step closer to differentiating between the fibres in blind analysis based solely on elemental composition. As magnesium is a very common impurity in crocidolite, its signal can easily provide spatial information of the position of the crocidolite fibres ( Figure 4B). Likewise, a high 29 Si signal is shown in Figure 4C, both in longer crocidolite fibres and in the shorter fragments.

| Wollastonite
Wollastonite was selected as a negative control MF due to its similar chemical and physical properties to those of established asbestos fibres. It is a calcium inosilicate mineral with no established association to MPM. Mined in a similar fashion to other MF, wollastonite might contain other metal impurities including Fe, 8 which could explain the slight 57 Fe signal present in Figure 5A. On the other hand, no magnesium was present in the wollastonite samples, as shown by the LA-ICP-MS images ( Figure 5B). As expected, a high 29 Si signal was generated in the case of this MF ( Figure 5C). Notably, some impurities present in the optical image were proven not to be wollastonite, confirming the high sensitivity of this technique.

| CONCLUSIONS
The poor treatment response and short-term survival of MPM cases are largely attributed to late diagnosis and lack of a standardised screening method. Moreover, current detection of asbestos fibres by microscopy is inadequate, with numerous fibres being undetectable, even in patients with heavy asbestos exposure. 3 Our study is the first to use LA-ICP-MS to spatially resolve multiple MF within a cellular background based on the presence of three main elemental components. The use of a low-dispersion ablation cell provided high-resolution, high-speed analysis, demonstrating that LA-ICP-MS has the potential to screen for the presence of asbestos in more biologically complex models such as human tissue samples. Further investigation to look at detection and differentiation of asbestos bodies with different elemental compositions would improve the technique and facilitate identification of fibre subsets. As a proof of concept study, the present work aimed to prove the ability of LA-ICP-MS to detect these MF against a cellular background using a low lateral resolution (2 μm).
The work sets the basis for future multi-element analysis carried out using a time-of-flight mass spectrometer, which provides pseudosimultaneous detection of the full mass range.
We succeeded in mapping amosite, chrysotile, crocidolite and wollastonite distribution within 2D cytospins based on the metal content. Further work will focus on more biologically complex matrices such as 3D models and patient samples to confirm the presence of asbestos. These preliminary data suggest that ultimately this instrument can be used within a clinical setting to aid early diagnosis of MPM and patient outcome.