Plasma estradiol/progesterone ratio, increase leads to Persistent vaginal cornification

The biological plausibility of the key event relationship is supported by aging rodent females. Reproductive senescence in rodents results from initial centrally mediated changes with alterations in hypothalamic function (Gore, 2000; Kermath, 2012). Female rats and mice proceed through sequential reproductive stages. Female rats undergo a transition from regular estrous cycles to irregular cycles (up to the age of 3- to 7-month-old), followed by persistent estrus (characterized by PVC), then repetitive pseudopregnancy (also called persistent diestrus) and finally anestrus (Finch, 2014; Shirai, 2015). However, there are species and strain differences in the pattern of changes and timing in the onset of reproductive senescence (Vidal, 2017). The tree different pathways of reproductive senescence are represented in Fig. 17 (Finch, 2014) and the pattern and the age at cycle cessation in different strains of rats and mice are available in Table 5 (Nelson, 1982). PE begins to be observed in Sprague-Dawley rats by 6 to 7 months of age (Vidal, 2017) or even earlier (Elridge, 1999). In contrast, Han Wistar rats are reported to have cycle irregularities after 6 months of age and move mainly into persistent diestrus (Mitchard and Klein 2016). 

Figure 17. Alternate trajectories of rodent reproductive senescence from (Finch, 2014)

Table 5. Incidence and vaginal cytological status of prolonged cycles in different strains of mouse and rat ranked by age (Nelson, 1982)

In aged rodents showing persistent vaginal cornification, the hormonal profile is characterized by sustained E2 and low P (Finch 2014). This hormonal change is explained by the histopathology of ovary showing numerous follicular cysts (producing estradiol) and lack of corpora lutea (producing progesterone).

The link between increased E2/P4 ratio and PVC has been demonstrated in different strains of both species. In a comparative study between Donryu and Fischer-344 rats, Nagaoka et al., sequentially followed estrous cycles by vaginal smear as well as E2 and P plasma levels, up to the age of 15 months. The E2/P4 ratio was higher in Donryu rats than in F-344 rats from 8 month of age and was about 7 times higher (p < 0.01) at 12-month. In Donryu rats persistent estrous (measured as PVC) appeared in 17% of the 5-month animals, the prevalence increased with age (90% of the 10-month animals while almost all F-344 rats showed a normal estrous cycle up to 8 months, and only a few cases showed persistent estrous thereafter). The authors concluded that the sustained increased E2/P4 ratio observed in Donryu rats may explained the high spontaneous occurrence of uterine endometrial adenocarcinoma in Donryu rats compared to F-344 rats (Nagaoka, 1994). The link between increased E2/P4 and PVC was also showed in aging Long Evans rats (Lu, 1979 and C578L/6J mice (Nelson, 1981) in which increased E2/P4 ratio was observed in PVC rats and mice compared to younger cycling animals.

Atrazine

Several studies in Sprague-Dawley (SD) female rats have shown that prolonged administration of atrazine by diet induced earlier increase of the number of animals displaying PVC. However, only one publication reports the investigation of both the KE upstream and KE downstream events SD rats and Fischer-344 rats.

In a 6-month repeated dose toxicity studies (0, 25, 50 or 400 ppm), the number of animals with persistent estrous at 400 ppm was increased compared to controls from three months onwards (20/90 vs 10/90 at 3 months; 50/90 vs 26/90 at 6 months). A significant increase in percent of total days with estrous was noted in the 400ppm group after three months of treatment (Eldridge, 1999, Simpkins 2011, see also Fig. 18). In a 24-month chronic toxicity studies with interims kills in SD rats (0, 70, 400 ppm with interim kills every 3 months) and Fischer-344 rats (0, 10, 70, 200 or 400 ppm) a dose-dependent increased in the percent days in estrous stage was noted from the 70 ppm SD group after nine months of treatment reaching statistically significance only at time point 9-month). There was no effect of 400ppm atrazine on cyclicity in female Fischer-344 rats. Plasma E2 concentrations were significantly increased from 70 ppm onwards at 3 months in SD rats but not in Fischer rats. No other significant effects on E2 and progesterone were seen in both species (Wetzel, 1994; US EPA 2000; USEPA, 2018). However, when calculating the E2/P4 ratio from the data reported in Wetzel, 1994, a dose-dependent increased of E2/P4 ratio is noted at 3- and 9-month time points in SD female rats while no effect was observed in female Fischer-344 rats (Table 6).

Figure 18.  (b) Dose- and time-dependent effect of 0, 25, 50, or 400 ppm atrazine administered in the diet on the percent days in estrous in female SD rat (b) Comparison of the dose- and the time-dependent effects of atrazine administered at dietary concentrations of 0, 70, or 400 ppm on the percent days in estrous in female SD rats (c) Atrazine administered at dietary concentrations of 70 or 400 ppm had no effect on the percent days in estrous in female Fischer 344 rats (Simpkins, 2011)

Table 6. E2/P4 ratio calculated from Wetzel, 1994 data

These different studies are thoroughly assessed in different available documents of US EPA evaluation (Issues Pertaining to Atrazine Cancer Risk Assessment, US EPA 2000 https://archive.epa.gov/scipoly/sap/meetings/web/html/062700_mtg.html,)

GnRH antagonist Cetrorelix

After 30-day intramuscularly treatment of Cetrorelix (depot formulation corresponding to 20–24 µg/kg per day) in SD females (proestrous at day 0) Estradiol levels in the group given Cetrorelix were significantly reduced (P < 0.01) by 30% on day 4 of the experiment but returned to control values after day 10. While progesterone levels decreased in the treated rats by more than 50% on day 4 and remained at this significantly lower level (P < 0.01) through the entire experiment, resulting in an increased E2/P4 ratio (Fig. 21).

 The treatment induced prolonged (4–6 days) diestrous phase followed from day 6 by persistent estrous smears in most rats (75.5% of the smears examined between days 6 and 30 showed estrous) (Horvath, 2004, see also Fig 19).

Figure 19. Effect of Cetrorelix pamoate on serum E2 and P4 in female rats during the treatment (b1 and c1: serum E2 and P in control female rats at different stages of estrous cycle (Horvath, 2004)

Dose and temporal concordance

• Atrazine: The dose and temporal concordance is well demonstrated in the only study available investigating both the up and downstream KEs. However only two dose levels were tested (Table 7).
• Cetrorelix: only one dose was tested and the experiment lasted only 30 days, which significantly limit the evaluation of the dose and temporal concordance (Table 7).

Table 7. Dose and temporal concordance studies

Dose and temporal concordance

See Annex B.3.

• Only two stressors have been investigated. The dose and temporal concordance should be further substantiated with other stressors.
• E2 bioavailability in uterus was not measure directly in any of the studies reported: the empirical evidence of the KER relies on the indirect measurement of the downstream KE (i.e., the surrogate event PVC). It should be highlighted that Atrazine was negative in a uterotrophic bioassay performed in ovariectomized SD females which is not considered as an uncertainty but rather supports that atrazine does not exhibit a direct estrogenic activity. An intact HPG axis is necessary which is not the case in ovariectomized females.
• Sex hormone measurements are rarely carried out in female rodents' studies, which limits the investigation of the empirical evidence of this KER.
• Characteristics of reproductive aging in the female rats varies among strains (Chapin, 1996; Finch, 2014). This could explain the discrepancies of the results observed between SD and Fischer female rats exposed to atrazine.
• The internal quality of the primary research study used to substantiate the empirical evidence has not been evaluated (recommendation).