We presented novel findings on the cytotoxic effects of flavored aerosols emitted from CBD-containing e-cigarettes. Our study aimed to examine the acute in vitro effects of several commercially available products using an established ALI model. Consistent with our pilot study (Leigh and Goniewicz, 2020), we confirmed that CBD when vaporized with e-cigarette refill solutions (PG + CBD) shows the cytotoxic effects on bronchial epithelial cells (Figure 1). Additionally, we found that PG + CBD refill solutions resulted in increased release of several pro-inflammatory cytokines, including IL-1β, CXCL1, and CXCL10 (Figure 1). These results are consistent with other CBD studies showing increase cytotoxicity and suppression of viability of cells exposure to CBD (Cerretani et al., 2020; Urasaki et al., 2020). These results are important at a time when products containing CBD are being widely marketed as goods with potential health benefits. While the majority of the CBD containing products on the market are sold as a tincture or edibles to be taken orally or used topically with limited scientific research justifying their use, we have shown that these products may have potential adverse respiratory effects when inhaled as e-cigarette aerosols.

Consistent with findings from our study examining in vitro effects of flavored nicotine-containing refill solutions (Leigh et al., 2016), different flavors in CBD-containing products showed different cytotoxic effects. We presented novel findings showing that in each of the five examined flavors without CBD, there are significant differences from air control in the measured cytotoxicity assays (Figure 1). Importantly, the PG-only solvent control also showed increased cytotoxicity compared to the air control (Figure 1), suggesting that solvents used in e-cigarettes may also independently contribute to the cytotoxic effects of the aerosols emitted from those devices.

An important finding from our study is that some commercially purchased products were more cytotoxic than others. This is likely a consequence of the differing flavoring chemical present in the various refill solutions. While the exact flavoring compound(s) responsible for this increase cytotoxicity was not determine in this study, we have noticed that some flavoring chemicals were only present in those flavored products that showed increased cytotoxicity, such as ligustrazin in F2 (Supplementary Table 2). Additionally, new compounds may be created when flavored refill solutions are heated inside e-cigarette devices. There is a need for further research identifying which flavoring chemicals are responsible for increasing cytotoxicity in flavored refill solutions. Identification of such highly cytotoxic flavoring chemicals may inform development of product standards and future products regulation to assure consumer safety.

Another important result of our study is the observed cumulative effect of flavorings and CBD present in all tested refill solutions. We observed a significant increase in cytotoxicity for all flavors as well as increased release of pro-inflammatory cytokines for most flavors, when comparing flavored refill solutions with and without CBD (Figure 1). Flavored refill solutions with CBD also resulted in a significant increase in cytotoxicity and production of pro-inflammatory mediators as compared to the unflavored controls containing only CBD (PG + CBD, Figure 1). Importantly, we did not observe similar effects for nicotine in our previous study (Leigh et al., 2016) since addition of nicotine to PG only solutions did not significantly affect the toxicity of the aerosol from previously tested refill solution. In contrast to nicotine, addition of CBD to flavored refill solutions amplify the observed biological responses. However, it should be noted that aerosol characteristics are highly linked to the device used, if other devices were used these results may not necessarily be the same.

In this study, we measured several pro- and anti-inflammatory mediators and found a significant increase in pro-inflammatory mediators IL-1β, CXCL1, CXCL2, and CXCL10 as well as a significant decrease in anti-inflammatory mediator IL-10 when H292 cells were exposed to flavored CBD refill solutions. While determining the mechanism behind these pro- and anti-inflammatory cytokines/chemokines was not our primary hypotheses, we believe the increase inflammation observed may be a result of co-administration of the solvent, propylene glycol, commonly used in refill solutions along with CBD. These results differ from that of (Urasaki et al., 2020) that found CBD was only cytotoxic in its pure form and not when diluted in oil. However, in our study, we used a different solvent which may explain these results. One of the important limitations of our study is that even though we purchased products of the same brand with and without CBD from the same supplier, we found some differences in flavoring composition of those products (Supplementary Table 2). This indicates that either there were additional additives present in CBD-containing refill solutions, that there may have been some inconsistencies between batches or that the flavor name present on the packaging does not necessarily use the same flavoring chemicals consistently to make that flavor profile. It should be noted that all commercial refill solutions were only qualitatively analyzed for flavoring chemicals. Future quantitive studies should be conducted to verify the presents and concentration of each of these flavoring compounds. Additionally, dose-responses experiments may be needed to precisely measure potential toxic effects of those products.

Another limitation of this study is that only one commercial product, in one concentration was used as the unflavored CBD containing refill solutions. Future studies should examine multiple unflavored CBD products, in multiple concentrations to determine if cytotoxic effects observed in our study are dose dependent. Although we tested five different flavored refill solutions with and without CBD, all products selected for our study came from a single manufacturer. Additional research should examine a large number of flavored CBD containing refill solutions from multiple manufacturers to verify our findings. Finally, our study only used a single physiologically relevant cell line to examining cytotoxicity and inflammatory endpoints. This culture model does not mimic the pseudostratified columnar epithelial cells of the human bronchial epithelial tissue comprising of goblet, cilia, and basal cells. While this method is useful for surveying acute effects of inhaled mixtures, it may not necessarily correlate to the respiratory effects seen with long-term use of these products in humans. Future studies using 3D culture and observational human trials would be useful to further verify these studies and understand the mechanisms involved.

A final limitation of this study is that no quantitative measurements of dosimetry were performed other than observed weight of refill solutions before and after use (Supplementary Table 3). With these measurements, we observed that the eGO e-cigarette with the CE4 tank had inconsistent delivery between runs with on average more aerosol delivered to the trials with refill solutions not containing CBD. We would expect the observed effects for CBD containing refill solutions would be further increased if similar amounts of aerosol were delivered to these trials. Future studies should include measurements of physical parameters such as contact angle, surface tension, and viscosity as well as other physical measurement such as quantitative chemical analysis and pH of refill solutions. Additionally, identification of other physical changes in the exposure system including pH and osmotic concentration would be useful to eliminate these as possible causes of observed results. The osmotic concentration of pure PG is upwards of 13,000 mOsm. Normal osmotic concentration for most vertebrate cells range from 260 to 320 mOSM/kg (Ian, 2000). As a result, it is possible some of the effects observed in this study are a result of a difference in osmotic concentration as demonstrated previous studies (Zhang et al., 2019).

In summary, the results of our in vitro study suggest potential harmful respiratory effects of flavorings and CBD when inhaled simultaneously with aerosols emitted from e-cigarettes. As use of cannabis-derived vaping products are increasing, studies are urgently needed to evaluate potential health consequences in users of these substances, particularly respiratory effects on chronic inhalation of flavored CBD-containing vaping products.