Epidemiological and toxicological risk assessments of ortho-toluidine for bladder cancer

Abstract

1. Introduction

Ortho-toluidine (OT) is a human carcinogen that was first listed as a Group 1 carcinogen by the International Agency for Research on Cancer (IARC) in 2012.¹ OT is absorbed through inhalation, ingestion, and/or skin contact, and OT exposure targets the urinary organs in humans. In 2016, the Japan Society for Occupational Health (JSOH)² proposed a change in the classification of OT from Group 2A to Group 1. In 1991, the JSOH had recommended 1 part per million (ppm) (4.38 mg/m³) as the occupational exposure limit-mean (OEL-M) for OT based on OT's noncarcinogenic health effects and a skin absorption notation.²

A case-series analysis from Japan reported that 5 male workers (1 retired) from a small factory of a company that produces organic dyes and pigment intermediates were diagnosed with bladder cancer (BCa) in 2014-2015 after complaining of gross hematuria.^3,4 Six aromatic amines had been used in the factory: OT, aniline (AN), p-toluidine, o-anisidine, 2,4-dimethylaniline, and o-chloroaniline. According to the inspection report by the Ministry of Health, Labour and Welfare preceding this case-series analysis, high concentrations of OT and its metabolites were detected in the urine of workers at the factory even though the average concentration of OT in the workplace air was significantly lower, at 0.003 ppm, than the 1-ppm OEL-M, indicating that dermal absorption played a significant role.³

In this mini-review, we report integrative evidence for the health risk assessment of OT, focusing on bladder cancer analyzed in studies of humans, experimental animals, and OT's skin permeability.

2. Clinical studies and epidemiology

Various epidemiological studies have assessed the relationships between BCa (incidence or mortality) and OT exposure among primarily OT-exposed workers but also other subjects. We review the reports of a Japanese case series⁴ and retrospective cohort studies performed in 2 different fields by the US National Institute for Occupational Safety and Health (NIOSH)^5,-7 and another publication by a Japanese ortho-toluidine study group.⁸ All data were appropriately adjusted for confounding factors such as co-exposed chemicals and tobacco smoking.

The NIOSH study series began in 1988, and a study reported in 1991 by Ward et al⁶ investigated a chemical plant in the United States where OT, AN, hydroquinone, and toluene had been used to synthesize a rubber antioxidant product since 1957. Ward et al⁶ reported a standardized incidence ratio (SIR) of 6.48 for BCa in 708 workers exposed to OT and AN, and an increased risk of BCa that was strongly associated with a greater than 10-year exposure and a greater than 20-year latency. In another investigation conducted in 1990, the mean (SD) of 28 personal air sampling concentrations was reported to be 412 (366) μg/m³, whereas the mean (SD) post-shift urinary OT levels in exposed nonsmokers was 80.1 (94.0) μg/L.⁷ In addition, the mean post-shift urinary OT value was 35 times higher in the OT-exposed workers (n = 42; 98.7 [119] μg/L) compared with the non-exposed workers (n = 25; 2.8 [1.4] μg/L).⁷ Using the exposure categories revised by NIOSH, Carreón et al⁵ reassessed an SIR of 5.84 for BCa in 962 workers exposed to OT and AN. Although the study workers were exposed to both OT and AN, Carreón et al concluded that OT was likely associated with the risk of BCa because it is a more potent carcinogen than AN in animals¹ and was found in the urine of workers at higher levels compared with AN.⁷

After the first report⁴ of 5 cases of BCa, 4 more male workers at the same factory in Japan and 1 male worker at another factory of the same company were also diagnosed with BCa in 2016 and 2017 in health checkups focusing on the detection of urinary tract cancer in Japan. All 5 workers had been exposed primarily to OT and, to a lesser extent, to AN, 2,4-xylidine (dimethylaniline; MX), p-toluidine, o-anisidine), and/or o-chloroaniline. All of those workers had been engaged in drying and packing a product made from OT.⁴

In 2021, Nakano et al⁸ observed that a series of workers with BCa had a high prevalence of a clinical history of bladder cystitis, gross hematuria, microscopic hematuria (red blood cells ≥5/high-power field), and nuclear matrix protein-22 (a marker of BCa) in their urine compared with the nonexposed workers. In addition, the prevalence rates of cystitis and bladder lesion–related symptoms in the study's aromatic amine–exposed BCa-free workers were significantly higher than those of the nonexposed workers.⁸ Since no exposure concentrations of each aromatic amine were available, Nakano et al calculated the surrogate exposure levels of each aromatic amine by using the working process, exposure potency, and duration of exposure. Overall, the SIR for BCa in OT-exposed workers was 56.8 (95% CI, 27.7-104.3), and 0-10-year lagged analyses showed clear dose–response relationships between the SIR and the surrogate exposure levels (Figure 1). On the other hand, the SIRs for AN- and MX-exposed workers were 57.5 (95% CI, 28.0-105.5) and 57.7 (95% CI, 28.2-105.9), respectively. However, these dose–response relationships for AN and MX were unclear compared with that for OT. There was a bias of smoking in the SIR calculated at 1.32-1.69, and the estimated SIR adjusted for smoking was 33.5-42.8. Nakano et al concluded that OT may be associated with an increased risk of BCa. Continuous monitoring for bladder cystitis and hematuria in OT-exposed workers is important for the early detection of BCa.

Figure 1

The age-adjusted standardized incidence ratio (SIR) for bladder cancer in 98 men exposed to aromatic amines, including 9 cases evaluated by Nakano et al.⁸

Open in new tabDownload slide

In a NIOSH series reported by Park et al⁹ the excess lifetime risk of BCa from OT exposure in the rubber industry at 1 part per billion (ppb) was estimated to range from 1 to 7 per thousand. Park et al conducted a quantitative risk assessment based on cumulative estimated OT exposures using previously assigned ranks of exposure intensity and reported actual exposure in jobs, together with a direct historical environmental sampling for OT.⁹ Their study included several assumptions, however. First, that a linear model could be fitted to the relationship between OT exposure and BCa incidence even at low concentrations. Second, that urinary OT would be due to exposure from the air rather than percutaneous exposure. And third, the effects of exposure to tobacco and AN were negligible. For these reasons, there are limitations to the study's present interpretation. A urinary biological monitoring index for the excess lifetime risk of BCa due to exposure to OT from inhalation and/or dermal contact is needed.

3. Animal studies

Dietary exposure to OT induces bladder cancer in rats,¹ and recent 4-week rat carcinogenicity studies further demonstrated that OT induces both simple hyperplasia with increased cell proliferative activity and the formation of γ-H2AX, a novel DNA damage marker for predicting potential carcinogenicity in rats.^10,11 These findings indicate that OT, like most bladder carcinogens, initiates the urothelial carcinogenic process beginning with simple hyperplasia and progressing to nodular hyperplasia and papillary hyperplasia, which can ultimately develop into papilloma and urothelial carcinoma. OT is currently classified by the IARC as a Group 1 carcinogen.¹

Our research group also conducted an experimental study of acetoaceto-o-toluidide (AAOT) and obtained robust evidence that AAOT is a potent bladder carcinogen. We first observed that AAOT induced simple hyperplasia with increased cell proliferative activity and γ-H2AX formation in the bladder urothelium of rats in a 4-week carcinogenicity study.¹² Our subsequent study revealed that AAOT promoted urinary bladder carcinogenesis induced by N-butyl-N-(4-hydroxybutyl)nitrosamine in a dose–response manner in a 36-week rat 2-stage urinary bladder carcinogenesis model.¹³ Importantly, in that study, urinary analyses of AAOT and its metabolites showed that AAOT, OT, and downstream metabolites of OT increased in a dose-dependent manner in the urine of male and female rats administered AAOT in their diet. Notably, OT was the most abundant urinary metabolite, present at least 1 order of magnitude higher than AAOT and the other OT metabolites found in the urine. These findings indicate that AAOT is primarily converted into OT and excreted in the urine, suggesting that OT metabolized from AAOT plays a pivotal role in the bladder carcinogenicity of AAOT.

Although results obtained from in vitro experiments cannot be simply extrapolated to complex in vivo responses, our earlier results demonstrated that the toxicities of AAOT in rat MYP3 cells and human 1T1 cells were similar, and the toxicities of OT in rat MYP3 cells and human 1T1 cells were also similar.¹² This suggests that the toxicity profiles of AAOT and OT in rat and human cells might not differ significantly.

Figure 2

Metabolic pathway in liver and potential modes of action for AAOT/OT with regard to rat urinary bladder carcinogenesis. AAOT, acetoaceto-o-toluidide; OT, ortho-toluidine.

Open in new tabDownload slide

Multiple mechanisms by which OT induces urinary bladder cancer have been proposed, including metabolic activation to reactive metabolites that bind DNA and proteins, mutagenicity, oxidative DNA damage, chromosomal damage, and cytotoxicity.¹ In an investigation of the mechanisms of AAOT and OT in rats, we carried out a comprehensive analysis of DNA adducts (DNA adductome), and we identified a number of common oxidative DNA adducts, including 8-hydroxy-2'-deoxyguanosine (8-OHdG), in the urothelium of rats administered AAOT and OT in their diet.¹¹ Our group's 4-week rat bladder carcinogenicity study also revealed that administration of the antioxidant apocynin significantly inhibited the cell proliferation, DNA damage, and oxidative stress induced by OT. These findings strongly imply that oxidative stress plays a crucial role in the development of urinary cancer caused by OT.¹¹

A 4-week carcinogenicity study documented that reactive oxidative stress (ROS) can activate the oncogene JUN. Gene expression analyses revealed that the expression of JUN and its downstream target genes was significantly increased in the urothelium of male rats administered AAOT in their diet.¹³ This suggests that the induction of ROS and the consequent overexpression of JUN and its downstream target genes are critical factors in the bladder carcinogenicity of AAOT and OT.¹³

It is also well established that a metabolic activation of OT and the subsequent genotoxic effects in the bladder urothelium occur in both experimental animals and humans. The role of the body's liver metabolism is thus of paramount importance when considering the carcinogenesis of OT. We used humanized-liver mice (NOG-TKm30 mice) established via human hepatocyte transplantation to compare differences in the OT-induced expression of metabolic enzymes in human and mouse liver cells.¹⁴ The comparison demonstrated significant differences in the expression of hepatic metabolic enzymes induced by OT in human and mouse liver cells, resulting in differing metabolisms of OT. For example, the expression of CYP3A4 was markedly increased in the livers of humanized-liver mice, whereas the expression of Cyp2c29 (human CYP2C9/19) was increased in the livers of the NOG-TKm30 mice. These differences could have profound implications for the carcinogenicity of compounds metabolized by the liver and are crucial for extrapolating data from animal models to humans.

Figure 2 depicts the metabolic pathways in liver and potential modes of action for OT-mediated rat urinary bladder carcinogenesis, based on observations from our group's recent studies and results presented in the literature. OT is the key carcinogenic metabolite of AAOT. N-hydroxyl-o-toluidine (N-hydroxyl-OT), a metabolite of OT, covalently binds to DNA bases, leading to DNA adduct formation in the bladder urothelium, causing DNA damage and mutagenicity. This is a well-established carcinogenic mode of action for OT. However, although N-acetyl anthranilic acid (NAATA) is a downstream metabolite of 2-aminobenzyl alcohol (2ABA) and N-acetyl-OT, the current data do not support a role of NAATA in 2ABA- and N-acetyl-OT-mediated DNA damage and urothelial hyperplasia.

In another pathway, the OT metabolites N-acetyl-OT and 4-amino-m-cresol are further metabolized to N-acetyl-4-amino-m-cresol, which generates ROS. ROS can cause DNA and protein adduct formation, contributing to OT-induced carcinogenesis. ROS are also associated with the overexpression of JUN and its downstream target genes, which may further contribute to the carcinogenicity of OT and AAOT. The metabolism of OT thus generates the metabolites N-hydroxyl-OT, which directly causes DNA damage, and N-acetyl-4-amino-m-cresol, which generates ROS. OT metabolites thereby promote DNA damage and cell proliferation, resulting in bladder carcinogenesis.

In summary, our experimental studies provide convincing evidence that AAOT is a potent bladder carcinogen that may contribute to bladder cancers. Although epidemiological studies in human populations are invaluable for identifying carcinogenic factors, they face significant challenges. It often takes many years for cancer to develop after exposure to a carcinogen, necessitating long-term observation to identify the substance as a carcinogen. In addition, it is far better to identify potential human carcinogens using animal models before they are identified as human carcinogens in epidemiological studies. As the list of chemicals associated with workplace exposure continues to grow, the identification of chemicals that may trigger occupational cancer becomes increasingly critical, and understanding their carcinogenic mechanisms is essential for effective cancer prevention. The use of experimental models remains an indispensable approach for identifying the causes of occupational cancer and elucidating their mechanisms of action.

4. Studies of skin permeability

As mentioned above, dermal absorption has played a significant role in the development of bladder cancer caused by exposure to aromatic amines, including OT, that were used in chemical plants in Japan. In fact, from the perspective of the physical properties of the chemicals, all 6 aromatic amines used in chemical plants have extremely low vapor pressures, making them unlikely to evaporate into the air. Dermal exposure is therefore considered an important exposure route. The current knowledge regarding the skin permeability of the 6 aromatic amines used in chemical plants (OT, AN, p-toluidine, o-anisidine, 2,4-dimethylaniline, and o-chloroaniline) remains extremely limited.

Figure 3

The permeation test using 3-dimensional (3D) skin was performed with slight modifications to the method described by Abdallah et al,¹⁵ fundamentally following the principle of the Franz cell diffusion assay. Briefly, a culture cup containing 3D skin (culture surface area approximately 0.9 cm²) was inserted into a receptor bottle filled with phosphate-buffered saline (PBS). A total of 200 μL of PBS containing [¹⁴C]-labeled o-toluidine (OT), p-toluidine (PT), aniline (AN), anisidine (ANS), 2,4-dimethylaniline (DMA), o-chloroaniline (OCA), and dimethylformamide (DMF), at a final concentration of 5 μM each, was applied to 3D skin (the radioactivity of each added solution, calculated based on the specific activity, ranged from approximately 1500 Bq to 3700 Bq). Subsequently, the PBS in the receptor bottle was sampled at specified intervals, and the permeated amount was quantified by counting the β-rays using a liquid scintillation counter.

Open in new tabDownload slide

Our research group used a 3-dimensional (3D) reconstructed human skin model to investigate the skin permeability of these aromatic amines. The 3D skin model we used consists of stratified layers of human epidermal cells, replicating the basic structure of the epidermis from the basal layer to the granular layer, spinous layer, and stratum corneum. As shown in Figure 3, approximately 70% to nearly 80% of the 6 aromatic amines, based on the initial amount applied to the skin being 100%, had permeated through the 3D skin within 8 hours. In contrast, the permeation rate of dimethylformamide, which is well known as an industrial chemical that often poses problems due to its percutaneous absorption, was ~20% at the same time point. Based on the skin permeability within 8 hours, we observed that these 6 aromatic amines permeated through the 3D skin approximately 3 to 4 times faster than dimethylformamide. Despite some variations between the substances, the skin permeability rate of the 6 aromatic amines is generally considered to be high¹⁶.

The main barrier to the skin absorption of chemicals is the outermost layer of the epidermis, that is, the stratum corneum, which is a generally hydrophobic (lipophilic) layer. Chemicals with higher lipophilicity are thus thought to penetrate the stratum corneum more easily. The octanol-water partition coefficient, which indicates the degree of lipophilicity of a substance, ranges from 0.9 to 1.9 for the aforementioned 6 aromatic amines, whereas it is −1.0 for dimethylformamide. The lower skin permeability of dimethylformamide compared with aromatic amines can therefore be attributed to dimethylformamide's difficulty in penetrating the stratum corneum of 3D skin.

Korinth et al^17,18 investigated the permeability of OT and AN in excised human skin, and reported that the permeation rates after 24 hours were approximately 50% and 40%, respectively. This suggests that the skin permeation rates obtained using 3D skin models are about 5 to 6 times faster than those in actual human skin, and it also suggests that calculations based on results from 3D skin models can be used to determine the permeation rates of chemicals in human skin. These data highlight the potential use of 3D skin models, which are easier to handle, for estimating dermal exposure levels to chemicals.

The skin permeability of all 6 aromatic amines is considered to be similarly high, but when their genotoxic potentials were tested in a human urothelial cell line using γ-H2AX as a marker of DNA damage, clear differences were observed.¹⁰ As expected, OT, which is recognized as a carcinogen in both humans and animals, exhibited strong γ-H2AX generation. Surprisingly, o-chloroaniline and 2,4-dimethylaniline showed even stronger γ-H2AX generation than OT. The overall order of the inductive potency of γ-H2AX was o-chloroaniline = 2,4-dimethylaniline > o-toluidine > o-anisidine > p-toluidine = AN. Prolonged dermal exposure to OT along with other genotoxic aromatic amines over many years may contribute to the development of bladder cancer.

Funding

None declared.

Conflicts of interest

None declared.

References

Author notes