Chemotherapy decreases the mortality of cancer patients, but it is also associated with irreversible ovarian toxicity [18, 19]. Administration of a GnRH agonist has been proposed to be a non-invasive method to protect the ovarian reserve from chemotherapy, although the efficacy of this approach is still controversial [8, 13, 19–25]. By suppressing the hypothalamic-pituitary-ovarian axis, the GnRH agonist preserves more primordial follicles, which are not vulnerable to chemotherapeutic agents . The recommended pretreatment before chemotherapy requires approximately two weeks due to the flare-up effect of gonadotropin concentration in the first week. This flare-up effect is not acceptable in two types of patients, including those suffering from rapidly progressive cancer and who cannot wait for the initiation of chemotherapy, and patients for whom an elevated gonadotropin level may promote the progress of some hormone-susceptible cancers such as breast and ovarian cancer. GnRH antagonist, which is also reported to act as a protective agent, rapidly down regulates the gonadotropin level without flare-up effect, but there are no long-acting GnRH antagonists available [9, 11, 26–29]. The combination of GnRH agonist and antagonist treatment is assumed to have a rapid long-acting suppression effect, avoiding the initial gonadotropin activation. Until now, four studies investigating the combination of these drugs have reported that the hormone changes are beyond those observed using controlled ovarian hyperstimulation protocols. Mardesic et al. reported a down regulation of LH 96 h after the combination treatment, and von Woff et al. found an incomplete reduction in the hormone flare-upin the combination group [12, 15]. Roth et al. and Mueller et al. detected flare-up in both groups [13, 14]. In our study, an increasing tendency was observed for E2, LH and FSH on day 3, but only the level of LH increased significantly. This outcome is possibly due to either the concentration peak not being observed at the right time or the animals that are not at the same stage of estrous at the beginning of treatments. Co-treatment with GnRHant in the long term or the short term significantly prevented the hormone flare-up, and the gonadotropin level rapidly fell to a low level in 5 days. This conflicting result could be attributed to differences between human and rodents, the dosage and the administration strategy. Rats have a higher concentration of GnRH receptors in the ovary, and pituitary-ovarian  desensitisation can be completed quickly; thus, rats may show a different response to the treatment compared to humans.
Another finding in our study was that the combination (lng) treatment significantly preserved more primordial follicles than observed in the GnRHa or GnRHant only groups. Danforth et al. reported that in a murine model, the antagonist did not protect ovaries from chemotherapy-induced toxicity, and it even depleted primordial follicles . No such effect was observed in the present study, but the antagonist preserved the follicles similarly to the agonist. This result is consistent with other studies performed using the antagonist [9, 11, 26–29]. The combination (lng) treatment enhanced the protective effect, which was confirmed by all of our results. The combination treatment decreased the E2 serum concentration to 96.6±5.8 pg/ml (mean±SD) on day 8, and in the GnRHa and GnRHant groups, the E2 levels were 159.6±8.5 pg/ml and 151.9±10.4 pg/ml, respectively. In addition, 80% of the rats in the combination group resumed a normal oestrous cycle within 10 days after chemotherapy, while in the GnRHa and GnRHant only groups, 20% and 40% resumed a normal cycle, respectively. These results indicated that the ovarian function was maximally suppressed and quickly resumed with the combination treatment. Although Johnson et al. report that the primordial follicle pool is renewable, this characteristic remains to be confirmed . Therefore, quantitative measurement of follicles is the best way to evaluate the potential fertility. Consistently, the combination treatment maximally preserved the primordial follicles. The pregnancy rate was not examined because the animals were all sacrificed for histological analysis. Nevertheless, the effect of the combination (sht) treatment was not significantly different compared to that in the GnRHa or GnRHant groups. The enhanced protective effect could be attributed to the total dosage of the two drugs without any calibration, or it could be related to the synergistic reaction between them. Our work suggests two possible methods to rapidly and effectively protect the ovary from chemotherapy-induced damage: one is to add GnRHant in the short term; the other is to increase the dosage of GnRHa. The underlying mechanism requires further investigation.