Inhalation of crystalline α-quartz (silica) dust in many types of workplaces is well known to cause silicosis, kidney disease, autoimmune disease, lung cancer and increased susceptibility to tuberculosis. The development of pulmonary fibrosis and cancer following silica inhalation have been much studied and well-described, and the mechanisms leading to these conditions have been investigated using animal models and related in vivo and in vitro experiments. Ordinarily, such studies are performed using crystalline silica of high purity. Previous measurements of respirable dust at unconventional oil and gas well drilling sites during hydraulic fracturing operations indicated that some workers were exposed to levels of silica dust that exceeded established workplace exposure limits (see Fedan (2020) for review). Importantly, fracking sand is a type of industrial sand and some industrial sand-exposed worker populations are at an increased risk for silicosis and lung cancer (Hughes et al., 2001; McDonald et al., 2005; Steenland and Sanderson, 2001; Rando et al., 2018; Vacek et al., 2019). With the emergence of hydraulic fracturing as a mainstay method for oil and gas retrieval, and a rise in the number of workers potentially exposed to dust generated during the handling of sand used in the formulation of fracking fluid, the question arose as to whether the toxicity of inhaled fracking sand dust (FSD) resembles that observed after inhalation of crystalline silica.

The previous eight papers in this series (Fedan et al., 2020; Olgun et al., 2020; Russ et al., 2020; Sager et al., 2020; Krajnak et al., 2020; Sriram et al., 2020; Anderson et al., 2020) described chemical and biophysical properties of nine FSDs, their pulmonary effects after intratracheal instillation, and the effects of inhaled FSD 8 in the pulmonary, cardiovascular and immune systems, brain and kidney. The main findings made in the preceding studies are highlighted in this report. The reader is directed to the individual investigations for full descriptions of the results.

A second purpose of this report is an attempt to reconcile the finding that, at least in the lungs where comparisons may be drawn, the profile of many responses to inhaled FSD in the rat model differ appreciably from those that have been established for inhalation of reference crystalline silica, i.e., MIN-U-SIL® 5 (MIN-U-SIL). That is, the toxicity of inhaled crystalline silica reported in the literature and inhaled FSD 8 are not equivalent at the post-exposure time points investigated, with FSD 8 being milder in its pulmonary effects.

The reader is referred to the individual reports for detailed descriptions of the methods used in the seven individual investigations. Male rats were exposed to FSDs by intratracheal instillation or by inhalation; following exposure, a panel of in vivo and in vitro experiments were performed, as will be described below.

Several of the in vivo and in vitro experiments performed in this series of experiments to examine the effects of intratracheally-administered FSDs and FSD 8 inhalation have not been performed previously using MIN-U-SIL. Although it was not the purpose of these studies to compare side-by-side the effects of the two dusts, some of the effects of FSD observed in the present studies can be compared with the effects of MIN-U-SIL in the literature.

First, the chemical compositions of MIN-U-SIL and FSD dusts differ in subtle ways that affect their bioactivity. X-ray crystallography indicated that FSDs are predominantly crystalline silica as α-quartz; appreciable amounts of several other “impurities” were identified as comprising the remaining constituents. In contrast, MIN-U-SIL, often regarded as pure, crystalline silica, contains Fe, Al, Mg, Na, K, Ca and Ti in trace amounts (1.7% total). These elements, except for Ti, were identified in FSD using EDS. The small amounts of non-silica impurities in MIN-U-SIL have been generally regarded as without biological consequence. However, their presence has been shown to be sufficient to affect the bioactivity of MIN-U-SIL, as their removal with hydrofluoric acid decreased its hemolytic activity (Pavan et al., 2017). In contrast, washing MIN-U-SIL with hydrochloric acid removes surface iron contaminants and greatly increases silica toxicity as reflected in lung weight, phospholipidosis, and changes in lung protein concentration (Miles et al., 1994).

Second, an appreciable difference exists in the rates of clearance of FSD 8 and MIN-U-SIL from the lungs. Whereas FSD was cleared with a t_{1/ 2} of ~11.3 d after a single inhalation exposure, the t_1/2 for MIN-U-SIL particles was markedly longer, i.e., ~82 d, after a 20-d inhalation exposure (Porter et al., 2004). It cannot be ruled out with present data that the slower clearance of MIN-U-SIL particles is due to the longer exposure period, but this is unlikely, as lung burden overload did not occur under these exposure conditions (Porter et al., 2001). Rather, it would appear that FSD is cleared faster from the lung than MIN-U-SIL because of the chemical properties conferred on the particles attributable to the presence of greater quantities of impurities. The impurities in FSD 8 may render the dust with advantageous toxicokinetic characteristics that favor more rapid clearance to the lymph nodes compared to MIN-U-SIL particles. A major route of particle clearance from the lung is to the lymph nodes (Harmsen et al., 1985; Choi et al., 2010). Redistribution to the lymph nodes is important in workplace silica exposures. In multiple worker autopsy studies, silicosis of the lymph nodes precedes silicosis in the lung parenchyma (Cox-Ganser et al., 2009; Honma et al., 2007; Taeger et al., 2011). This lymph node damage is hypothesized to increase the risk of silicosis in the lung parenchyma by impairing silica clearance (Cox-Ganser et al., 2009). In addition, lymph node deposition of FSD may contribute to immune stimulation and potential further particle redistribution via the lymphatics. The potential role of lymphatic clearance in the extrapulmonary effects of FSD is an important area for future research.

Third, in a rat model of silicosis employing MIN-U-SIL, inflammatory changes in the lung increased at a rapid rate during the 36 to 116 d post-exposure interval following 40 or 60 but not 20 d of silica dust exposure (Porter et al., 2002; Langley et al., 2004). Therefore, in the present study, whether inflammatory changes were also evident in the lung 90 d following FSD inhalation was evaluated. No comparable changes were detected. In comparison to MIN-U-SIL, the greater abundance of impurities in FSD may have mitigated the inflammatory responses. However, inhalation exposure to FSD 8 was limited to 4 d.

Fourth, surface properties (Warheit et al., 2007) and the presence of surface metals in association with silica particles affect its bioactivity. After inhalation exposure of rats to FSD, no evidence of an inflammatory response was found. It is very likely that the greater abundance of metals in FSD compared to MIN-U-SIL plays a critical role in the pulmonary responses. Inasmuch as several minerals bind to FSD 8 and to MIN-U-SIL, and both dusts contain a nearly identical metal content profile (although metal percentages differ), it is not possible to ascribe one or more metals or cellular pathways to the pathways responsible for these differences. The presence of aluminum blunts the lung inflammatory response to silica (Nolan et al., 1981; Begin et al., 1986, 1987; Donaldson et al., 2001; Duffin et al., 2001). In contrast, the presence of iron has been associated with increased and decreased effects of silica particles (Schins et al., 2002; Ghiazza et al., 2011; Pavan and Fubini, 2017). In a study in which minerals in MIN-U-SIL were extracted with hydrofluoric acid to reduce metal impurities and the hemolytic activity of neat and acid-treated silica on erythrocytes were compared, the hemolytic activity of the dust was decreased (Pavan et al., 2017). Binding of metal salts or the polymer, polyvinyl-pyridine-N-oxide, to the surface of MIN-U-SIL decreased its hemolytic activity (Nolan et al., 1981). Similarly, MIN-U-SIL that was washed in HCl to remove surface iron had greatly enhanced toxicity relative to unwashed MIN-U-SIL (Miles et al., 1994). As both aluminum and iron are present in FSD, the preponderance of the overall pulmonary effects of FSD must reflect the influence and interplay of the effects of the metals. Inasmuch as the other metals in FSD have not been reported to influence toxicity appreciably in studies using MIN-U-SIL, it seems possible, at least tentatively, that effects of these two metals determine the pulmonary response to FSD. Alternatively, pathways stimulated by the other metals in FSD may participate in blunting FSDs’ relative toxic effects in ways that have not yet been described. In addition, quartz with an occluded surface is less toxic than DQ12, a reference material that is 87% quartz and 13% amorphous silica but has an unoccluded surface (Creutzenberg et al., 2008; International Programme on Chemical Safety, 2000). It is also unknown how the metals in FSD influence the toxicity profile of the dust on the cardiovascular and immune systems, brain and kidney, especially in comparison with MIN-U-SIL, the effects of which on these organ systems have not been investigated.

Given the differences in the characteristics and bioactivities of FSD and MIN-U-SIL discussed above, the question arises as to whether FSD 8 is unique in its toxicological profile among other silica-containing dusts inhaled in the workplace. Chen et al. (2005) compared the risk of silicosis among tin miners, tungsten miners, and pottery workers in relation to inhaled cumulative total dust and cumulative respirable silica dust. The risk of silicosis in pottery workers was less than that of the miners, even though they experienced a higher cumulative total dust exposure; moreover, the latency for silicosis was longer in pottery workers than in the other two cohorts. The authors concluded both the characteristics of the silica dust and the cumulative exposure to respirable silica influence silicosis risk (Chen et al., 2005). A likely hypothesis to explain these findings was set forth from the investigation of Harrison et al. (2005), who proposed that the degree of surface occlusion by alumino-silicate affects the bioactivity of silica dusts. The investigators proposed that the extent of surface coating of silica particles influences toxicity. Using energy-dispersive X-ray spectroscopy (EDS) to map the extent of surface occlusion by clay coating, the dusts inhaled by pottery workers and tin and tungsten miners were found to be 45%, 18% and 13%, respectively. These findings buttress the view of these investigators that the differences in susceptibility to silicosis in the three worker cohorts may be attributable to the degree of surface occlusion by clay. EDS was performed on FSD particles, and the results indicated that surface “coating” of particles was heterogeneous, both between individual FDS particle types and within individual particles in each sample (the degree of surface occlusion was not quantified in this study). This was in contrast to MIN-U-SIL particles, which gave a single EDS spectrum irrespective of the location of the x-ray beam (although trace minerals are present), indicating that the dust particles were not surface-occluded.

There are several limitations of this study. Sands utilized for fracking purposes are derived from many locations and they are not geologically equivalent (Benson and Wilson, 2015). Processing of the sand for use, e. g., washing with water, and environmental conditions will add a variety of exogenous factors to the product that could affect its bioactivity, depending on locale. The amount and chemical nature of the “impurities,” both exogenous and endogenous, would vary among sands obtained from different vendors. The findings and conclusions of this study are, therefore, not applicable to fracking sands from all sources, as the bioactivity of FSD could be linked to its geographical source. Unconventional gas extraction workers are not breathing “pure” crystalline silica dust (MIN-U-SIL) in the industrial setting, and, in light of the differences observed in this study, potential lung disease in these workers would be best modeled using sand dusts obtained at the drilling sites.

The FSDs used in this study may have some similarities to some other workplace respirable quartz samples. A previous study demonstrated that two workplace quartz samples from sandstone milling and sand feeder stock were less inflammatory, less cytotoxic and had less surface reactivity than DQ12 silica (Clouter et al., 2001). Comparing the physicochemical properties of other types of industrial sand dusts with the physiochemical properties of fracking sand dust is important because workers exposed to industrial sand have increased risk for silicosis and lung cancer (Hughes et al., 2001; McDonald et al., 2005; Steenland and Sanderson, 2001; Rando et al., 2018; Vacek et al., 2019). Recently, industrial sand workers exposed to low concentrations of industrial sand for long time periods were found to have a greater risk for silicosis than workers with comparable cumulative exposures received at a high dose over a shorter time period (Vacek et al., 2019). It is certainly possible that the physico-chemical properties of FSD and similar sand dusts change the timing of disease onset relative to MIN-U-SIL. Workers are not exposed to MIN-U-SIL and most cases of silicosis in workers develop over decades, a process that cannot be replicated in short-lived rodents. However, a more rapidly progressive form of pneumoconiosis has recently been reported under some mining conditions and crystalline silica is implicated in this process (Cohen et al., 2016). In short, humans get silicosis from many different workplace silica exposures and it is important to study factors that enhance the risk or facilitate rapid disease progression. Our study may provide clues regarding some of those potential factors.

Another limitation of these studies is the fact that the FSD used in these experiments was “aged.” By this it is meant that silica that has undergone fracturing (i.e., "freshly fractured"), such as during preparation of hydraulic fracturing fluid, gives rise to free radicals that exacerbate its toxicity in the lungs compared to silica that is “aged” with the passing of time (Vallyathan et al., 1988, 1995; Dalal et al., 1990; Shoemaker et al., 1995; Castranova et al., 1996). The FSDs that were utilized in the present study were collected at gas well drilling sites and were stored until animal exposures were conducted several months later. During the interval between collection and experimentation the FSD became aged. On the other hand, workers exposed to recently generated FSD on site would have inspired freshly-fractured FSD particles, not aged particles. The significance of this difference is that this study did not completely emulate workplace exposures. Most investigations of MIN-U-SIL toxicity also utilize aged dust, and it is against this background that FSD vs. MIN-U-SIL comparisons are being made. Additional study of the bioactivity of fresh vs. aged FSD is warranted.

Silicosis and lung cancer are risks in industrial sand workers (Hughes et al., 2001; Vacek et al., 2019; Rando et al., 2018), and our findings can inform investigations into etiology of these diseases in this cohort, especially from the standpoint of the differences our studies revealed between FSDs and MIN-U-SIL in terms of their physico-chemical properties. The dusts generated during sand quarrying and processing have not, to our knowledge, been characterized to the extent done in the present investigations. However, there are clear differences in the chemical compositions of the sand dusts arising at fracking sites, which reflect to a certain degree the origins of the fracking sands. Assuming that the dusts generated during quarrying also contain minerals other than silica, the presence of the minerals could influence the onset of silicosis in workers.

A number of investigations are suggested by the present studies to address questions and gaps that arose during these experiments. The abrogated pulmonary fibrotic response to FSD treatments in comparison to MIN-U-SIL treatment does not necessarily suggest that a fibrotic response to FSD would not have developed in time. It is possible that a fibrotic response is delayed in onset but that it would eventually occur following a longer exposure period to FSD. Side-by-side exposures to MIN-U-SIL and FSD in animal models could reveal time-course differences in the progression of inflammation.

Exposures of workers at drilling sites are not continuous and, perhaps, exposure “bursts” in the animals over extended periods of time might better mimic work site conditions experienced by the itinerant workers at drilling rigs.

Whereas the particles of FSD 8 were respirable when generated for animal inhalation exposures, the particle sizes of the nine neat FSDs given intratracheally were heterogeneous and contained non-respirable sized particles. Doses were based on dust weight. Using inhalation delivery of respirable particles generated from the nine FSDs could provide a more accurate assessment and possible detection of differences in the toxic potencies of the dusts, allowing additional evaluation of the roles of silica and metal contents in determining bioactivity.

It is intuitive that inhalation delivery of FSD 8, even on a sub-chronic basis, might be expected to induce changes in lung performance, but marked alterations were not observed. Effects in the cardiovascular system, immune system, kidneys, and brain were, however, observed. Whether other organs, such as liver and gut, are affected is worthy of consideration. As well, the mechanisms by which FSD 8 provoked responses in organs distant from the lung are not apparent at present.

The presence of minerals associated with silica dust (Pavan and Fubini, 2017; Pavan et al., 2017) and FSDs (Fedan et al., 2020) affects their toxicity. It might be predicted that removal of minerals from FSDs using chemical means (HCl and HF treatment) might change the toxicity profile from what was observed in the present studies to that described for crystalline silica.

In summary, a comprehensive analysis of the biological effects of FSD on several organ systems has been made using a rat intratracheal instillation and inhalation exposure model and in vivo and in vitro models. Two major findings made were that the bioactivities of the FSD have differences compared with MIN-U-SIL following short-term testing in rats. In addition, the toxicities of FSD extended to non-pulmonary organ systems, which, for the most part, were greater than those seen in the lungs. These findings are relevant to the understanding of the potential risk of cases of silicosis in the unconventional gas extraction and industrial sand industries.