For several years, detection of autoantibodies has been recommended in clinical practice for women with infertility including the detection of ANA, APA and ATA. A link of APA with recurrent pregnancy loss has been established, and treatment based on anticoagulation such as subcutaneous heparin is effective . As an organ-specific autoantibody, ATA affect pregnancy outcome negatively by damaging the thyroid function, and the thyroxin replacement therapy is efficacious in preventing foetal loss for patients with recurrent miscarriage [1, 2]. However, the potential correlation between ANA and IVF/ICSI outcome and specific medication are less reported.
Previous reports showed that ANA were relevant to adverse reproductive events including recurrent spontaneous abortion, endometriosis, infertility, IVF failure and ovarian dysfunction [6–10]. One recent study proposed that ANA might impair oocyte quality and embryo development, leading to reduced pregnancy and implantation rates . A further exploration  similarly suggested that IVF outcomes were markedly poorer in ANA + women, and this effect became worse with an increased level of serum ANA.
In this study, we found that the presence of ANA in serum predicted an adverse IVF/ICSI outcome, primarily reflecting the lower rates of MII oocytes and normal fertilisation and the reduced rates of implantation and clinical pregnancy, as well as the increased rates of abnormal fertilisation and early miscarriage. The cleavage rate and the perfect and available embryo rates were not significantly different between the ANA + women and the controls, which was a finding that did not coincide with that of Ying Y et al. . We found that ANA-titre ≤1:320 women had as poor IVF/ICSI outcomes as the ANA-titre > 1:320 women, and only the abnormal fertilisation rate in the IVF cycles in the ANA-titre > 1:320 women was significantly higher. These results did not suggest the titre-dependent effect of ANA on IVF/ICSI outcomes, in agreement with Hasegawa et al.  and Taniguchi et al. .
In general terms, abnormal autoimmune conditions may impair all stages of fertility, leading to ovarian and testicular failure, implantation failure or pregnancy loss through different putative mechanisms . It has been proposed that APA and A-β2-GPI could result in thrombosis of placental blood vessels, dysfunctions of trophoblasts in the peri-implantation period or an imbalance of maternal hormones [23–25]. Unlike APA and A-β2-GPI, the mechanism by which ANA determine reproductive failure remains speculative. Hasegawa et al.  held that ANA did not target specific organs and only presented as an abnormal degree of autoimmunity, based on their findings that the adverse IVF outcome was not ANA-titre dependent. One recent well designed study  demonstrated that the presence of APA, ANA or ATA in recipients using donor oocytes had no negative effect on pregnancy, implantation or miscarriage, which suggested that implantation failure might be because of poor-quality oocytes that may lead to a subsequent embryo development disorder. One in vivo test , in which anti-centromere antibodies were microinjected into mouse oocytes, showed the anti-centromere antibody could interfere with chromosome congression in the pro-metaphase. Shirota K et al.  considered that anti-centromere antibodies might infiltrate oocytes and lead to centromere dysfunction during meiosis and mitosis and impair the transition from MI to MII during oocyte maturation. Ying Y et al.  has confirmed that ANA exist in follicular fluid and embryos in ANA + patients, and serum and follicular fluid ANA negatively correlated with the number of high-quality embryos. The embryos co-cultured with IgG extracted from ANA + women were found to be severely impaired or even died .
Ando et al.  administered low-dose prednisolone (5 mg/d) or dexamethasone (0.5 mg/d) daily during the entire IVF cycle until the pregnancy test was performed in 51 lVF-ET cycles of patients positive for ANA, anti-DNA antibody, and/or lupus anti-coagulant (LAC), as well as 29 IVF-ET cycles of patients negative for any antibodies and discovered significant increases of pregnancy and implantation rates in the antibodies-positive patients with corticosteroid treatment but not in the antibody-negative patients. Hasegawa et al.  administered prednisolone (10 mg/d) plus low-dose aspirin (81 mg/d) to ANA + and/or APA + women from the first day of COH until pregnancy was confirmed by ultrasonography and discovered that the ANA + women with treatment had significantly better outcomes of IVF-ET (40.6% pregnancy rate and 20.3% implantation rate). Taniguchi et al.  administered prednisolone (15–60 mg/d) starting from the first day after OPU for 5 days in 56 IVF-ET cycles of 24 ANA + women and 167 IVF-ET cycles of 96 ANA-women and found the implantation rate and clinical pregnancy rate improved significantly in the ANA + woman but not in the ANA-women, which coincides with the findings of Ando et al. One previous prospective study administered prednisone plus aspirin for 4 weeks before IVF treatment to 52 women positive for ACA, ANA, anti-dsDNA antibody, rheumatoid factor, and/or LAC and ultimately obtained a satisfactory clinical pregnancy rate (32.7%) . The present study pretreated ANA + women with prednisone (10 mg/d) plus low-dose aspirin (100 mg/d) (i.e., P + A) for three months before IVF treatment and observed that the ANA + cycles with P + A had markedly more 2PN, high-quality and available embryos, and an increased implantation rate. Our study also showed that the ANA titre was not relevant to the IVF/ICSI outcome and the adjuvant treatment was not closely related to the reduction of the serum ANA titre, which is in agreement with Hasegawa et al. . In this study and that of Hasegawa et al. , women positive for autoantibodies were administered a 10-mg daily dose of prednisone or prednisolone, and this dosage is too low to reduce the autoantibody titres. Thus the corticosteroid effect may be derived from another mechanism such as an anti-inflammatory action or the regulation of immune cells as in the reduction of NK cells . Low-dose aspirin for its anti-thrombotic effect may reduce uterine and intraovarian vascular resistance, improve blood perfusion and increase oocyte maturation, the high-quality embryos rate, and the implantation rate . Although these five studies including the present work demonstrated a beneficial effect of corticosteroid or corticosteroid plus aspirin therapy, an ideal protocol for this adjuvant therapy (i.e., the indication of the patients, the drug selection and dosage, and the timing of commencement and end) requires further investigation.
In this study, we observed the detrimental effect of ANA on IVF/ICSI outcome and the beneficial effect of prednisone plus low-dose aspirin adjuvant treatment for ANA + patients. The comparison between the first IVF/ICSI cycle to a subsequent one would unavoidably yield bias, which is the limitation of our study. To minimise the bias, we rigorously selected the subjects, and there was a long interval between the first and the second IVF/ICSI cycles of at least 3 months. In our future work, we will attempt to perform a placebo-controlled, double blind, and prospective study.