Registration Dossier

Toxicological information

Carcinogenicity

Currently viewing:

Administrative data

Description of key information

Three Klimisch 2 studies were identified investigating the incidence of tumours after oral administration of biphenyl to rats (Umeda et al., 2002) and mice (Umeda et al., 2005; Bionetics Research Labs, 1968). Further, one Klimisch 3 study (Ambrose et al., 1960) and three Klimisch 4 studies were identified (Innes et al., 1969; Pecchiai and Saffiotti, 1957, Imai et al., 1983). These studies also investigated possible carcinogenic effects after oral administration. An increased incidence of tumour formation was observed in the urinary bladder of male rats (Umeda et al., 2002) and in the liver of female mice (Umeda et al., 2005). The former observation appears to be related to urinary calculi formation whereas the latter is thought to be related to peroxisome proliferation. Based on the study of Umeda et al. (2002), a NOAEL of 60 mg/kg bw/d could be derived for tumour formation. However, because of the facts that biphenyl does not exert genotoxicity and the non-genotoxic effects involved in cancer formation in rodents have been demonstrated to be of little, if any, relevance to humans, biphenyl should not be considered as a carcinogen in humans, and no classification for carcinogenicity is proposed.

Key value for chemical safety assessment

Carcinogenicity: via oral route

Endpoint conclusion
Endpoint conclusion:
adverse effect observed
Dose descriptor:
NOAEL
60 mg/kg bw/day
Study duration:
chronic
Species:
rat
Quality of whole database:
acceptable

Carcinogenicity: via inhalation route

Endpoint conclusion
Endpoint conclusion:
no study available

Carcinogenicity: via dermal route

Endpoint conclusion
Endpoint conclusion:
no study available
60

Additional information

Based on the available animal data, the lowest NOAEL for lesion formation of 60 mg/kg bw/day, (i.e., 1500 ppm in the diet) and NOAEL of 38 mg/kg bw/day (i.e., 500 ppm in the diet) for general toxicity could be obtained from the oral 2- year rat study of Umeda et al. (2002). In this study, neoplastic lesions (tumours) were observed in the urinary bladder of male rats at the highest dose of 338 mg/kg bw/day (i.e., 4500 ppm in the diet). No tumours were observed in female rats at any of the doses tested. It was suggested that bladder tumours in male rats are induced secondary to calculi formation triggered by a different metabolic pathway compared to female rats. The increased cell replication caused by chronic physical injury and irritation from bladder stones was identified as the underlying factor in induction of bladder tumours. Non- neoplastic kidney lesions were observed in males and females that accompanied renal effects at high doses of biphenyl.

Experimental evidence of a mechanism underlying male predominance in calculus formation in the bladder of biphenyl- exposed rats has been published (Ohnishi et al., 2000; see IUCLID section 7.9.3. and CSR section 5.10.). These authors demonstrated that the male bladder calculi were composed of potassium 4-hydroxy-biphenyl-o-sulfate (4-HBPOSK), whereas the female calculi were composed of 4-hydroxy-biphenyl (4-HBP) and KHSO4to which 4-HBPOSK was further hydrolyzed only in female urine, and that a series of irreversible and stable metabolic pathways from biphenyl to 4-HBPOSK resulted in formation of calculus in the male rat, whereas long- term calculus formation was prevented by reversible hydroxylation of 4-HBPOSK in female urine. The significantly increased urinary pH found in the 4500 ppm- dosed males may also facilitate the formation of bladder calculus because co- administration of biphenyl and KHCO3to male rats was reported to result in the formation of urine crystals of 4-HBPOSK through the increased urinary pH (Ohnishi et al., 2001; see IUCLID section 7.9.3. and CSR section 5.10.). Clearly, the results discussed above demonstrate that there exists a close causal association between threshold of bladder calculi formation and tumour induction predominantly in male rats.

Bladder cancer formation resulting from chronic irritation of the bladder epithelium caused by the presence of urinary calculi has been identified as a threshold effect which is also time- related. Cohen (1995 a, b; 1998) reported that this mechanism of bladder cancer formation may be less important to humans, upright organisms in which any urinary crystals formed will soon obstruct the urinary tract and cause severe pain leading to consultation with an urologist and removal of the crystals. In horizontal organisms, calculi stay present for most of their life span, and therefore cancer formation is expected to occur due to substantially longer periods of irritation that stimulates cell division and epithelial hyperplasia.

In a 2- year oral study in BDF1 mice (Umeda et al., 2005), an increased incidence of liver adenomas, but not carcinomas, was observed in female mice. The increased tumour incidence in the female mice was moreover not dose dependent. In male BDF1 mice, there was no increased incidence of liver tumours, but rather a decrease in tumour incidence was noted with increased biphenyl dose. Non- neoplastic lesions were observed in both males and females. The reported NOAEL was 87 mg/kg bw/day (i.e., 667 ppm in the diet) in females and 720 mg/kg bw/day in males (i.e., 6000 ppm). 

Overall, evidence suggests that the increased tumour incidence observed in the Umeda et al. (2005) study is unique to the female of the mouse strain used (i.e., BDF1). No carcinogenic responses or non- cancer effects were found in female ddY mice exposed to 5000 ppm biphenyl in the diet for 2 years (Imai et al., 1983) and B6C3F1 and B6AKF1 mice exposed to 517 ppm biphenyl (Innes et al., 1969; Bionetics Research Labs 1968). In the 18- month limit test (Bionetics Research Labs, 1968), mice were exposed by oral gavage to 215 mg/kg biphenyl from day 7 to 28 and thereafter to 517 ppm nominal in the diet. No significant increase in the number of tumours was observed in the rather small number of animals (18/sex).

It was suggested that induction of peroxisome proliferation in the liver of female mice could play a role in the liver tumorigenesis in BDF1 females, as investigated in a 90- day study by Umeda et al. (2004); however, the data are inadequate to confirm the activation of PPAR as the cause of liver tumours as the peroxisomal vacuoles were present at very high dose of 16,000 ppm biphenyl, much higher than the tumorigenic dose, but not at 8,000 ppm biphenyl, also a tumorigenic dose in the cancer bioassay. Thus, there is no consistent correlation between doses of biphenyl that produced peroxisome proliferation and liver tumours.

An alternative MOA involving sustained regenerative and proliferative changes in the livers of high- dose treated BDF1 females in response to cytotoxicity from the sex- specific, prolonged exposure to reactive metabolites of biphenyl was recently realized. The hepatocellular toxicity was demonstrated in the cancer study of Umeda et al. (2005) by increases in liver enzymes (i.e., ALT, and AST) in plasma of female mice at tumorigenic biphenyl doses, but not in female mice at non- tumorigenic biphenyl doses. In male mice, there was no increased ALT or AST measured at any dose of administered biphenyl. As a result of cytotoxicity in the liver of biphenyl- treated animals, significant regenerative cellular proliferation had to occur to compensate for the damaged tissue. There is a general correlation (particularly at higher doses) between occurrence of hepatotoxicity and/or regenerative/proliferative lesions and development of tumours for numerous chemicals for example carbon tetrachloride. Tumorigenesis through this threshold MOA is believed to require persistent hepatocellular cytotoxicity and regenerative cellular proliferation for tumour formation. 

Recent work by Zhang, Clark et al. (2013) looking at the absorption, metabolism and elimination (AME) of biphenyl helped clarify gender differences in BDF1 mice and showed differences in the rates of elimination of the biphenyl- derived radioactivity from blood between sexes and not in the proportion of biphenyl metabolites formed (refer to section 7.1 of the IUCLID and section 5.1 of the CSR). The blood area under the curve (AUC) for biphenyl metabolites was 1793 µg h g-1for males and 3102 µg h g-1for females, indicating that elimination of biphenyl metabolites in female BDF1 mice is substantially slower compared to BDF1 males and that females would overall be exposed to higher cumulative levels of biphenyl reactive metabolites over time compared to males. This finding is consistent with the observed increased incidence of liver tumours in BDF1 females at high biphenyl doses in the Umeda et al., 2005 cancer bioassay. The prolonged exposure to reactive biphenyl metabolites in BDF1 females would result in sex- specific liver injury, increased activities of plasma enzymes and regeneration of liver tissue. Overall, Zhang et al., (2013) findings are consistent with the threshold MOA for liver tumours in the BDF1 females where the accumulation of small number of minor monohydroxylated and dihydroxylated metabolites of biphenyl over time from slower elimination compared to males has the potential to induce liver injury. Although BDF1 males produced the same proportion of these mono- and dihydroxylated metabolites, they were eliminated from the body faster compared to the females, thus not leading to liver enzyme increases in serum and increase in liver tumour incidence seen in the cancer bioassay. Liver injury would thus be attributed to gender- specific elimination differences in the BDF1 mouse strain, but not in other strains or species.

Overall, the facts that biphenyl does not exert genotoxic activity and that the non-genotoxic effects associated with cancer formation in rodents at high doses can be demonstrated to be of no or limited relevance to humans support waiving of the possible classification of biphenyl as carcinogen.


Justification for selection of carcinogenicity via oral route endpoint:
a weight of evidence approach was used - no single key study is assigned, with Umeda et al., 2002 arriving at lowest NOAEL

Carcinogenicity: via oral route (target organ): digestive: liver; urogenital: urinary bladder

Justification for classification or non-classification

In the available studies on Biphenyl, there is evidence of tumour formation at high (oral) doses in male rats (bladder) and female mice (liver). Biphenyl is not genotoxic and through the additional mechanistic work it has been demonstrated that the tumour formation has a clear threshold and appears to be species, strain and sex specific. Given the absence of genotoxicity, the threshold mechanism of action and the specificity of the tumour formation to species, strain and sex, the relevance of these tumours to humans is considered minimal and, as a consequence, no classification for carcinogenicity is proposed for this substance.