Discussion
Although various strategies have been attempted to improve endometrial growth in women with thin endometrium, the optimal means of EP in these women is yet to be determined. The aim of our study was to evaluate the effect of TAM in patients with a thin endometrium in FET cycles. By univariate and multivariate analysis, our results found a superior pregnancy outcome following FET in a TAM ovarian stimulation cycle compared to HRT cycle in women with thin endometrium.
In our study we observed that the durations of endometrium preparation were 6 day shorter in the TAM FET group (10 days) than in the HRT FET group (16 days). This phenomenon is interesting but hard to explain. We hypothesize that ovarian stimulation with TAM may stimulate a faster development of the dominant follicle(s). More prospective studies are needed to explore its clinical relevance.
The result of this study showed that endometrial lining of the current HRT cycle differ from prior cycle (8.15±1.11 mm vs <7mm). Different from previous cycle, prolonged duration, increased dosage and transvaginal estrogen treatment were used in the present study. The study of Liu showed that effectiveness of estrogen treatment was significantly associated with treatment duration (12). Our study was in line with this study.
Our results showed that endometrium thickness was much more improved in TAM group than HRT group. TAM has been reported to improve endometrium thickness while ovulation induction. In the study of Kasey Reynolds, endometrial thickness was improved when the OI (ovulation induction) drug been switched from CC to TAM (34).Wang et al compared TAM and CC in ovarian stimulation cycle in combination with HMG (human menopausal gonadotrophin) and found that TAM treated patients had a significantly increased endometrial thickness (28).
The literature on the treatment of thin endometrium with TAM in FET cycle is rather sparse. Chen X et al used TAM in 3 women who showed a repeated unresponsive thin endometrium and resulted in endometrium expansion to at least 7.7mm and conception(29). Soon after, another study by the same group showed TAM increased the endometrial thickness from 6.5 to 8.8mm in women with thin endometrium undergoing FET(30). Ke H et al explore the effect of TAM in patients with a thin endometrium in FET cycle. Their study included 226 women who had an EMT of less than 7.5mm in their previous cycles. After the use of TAM, EMTs were significantly improved. When stratified by different previous endometrium preparation protocols, EMT increased from 6.11±0.98mm to 7.87±1.48 in NC group, from 6.24±1.01 to 8.22±1.67 in HRT group and from 6.34±1.03 to 8.05±1.58mm in OI group(31). In line with the previous data, our result showed that the EMT in TAM group was 8.98±2.02mm which was significantly thicker than the EMT in HRT groups (8.15±1.11mm).
Some authors found that TAM exposure promotes endometrial cell proliferation through estrogen and non-estrogen pathway. In estrogen signaling pathway, TAM stimulate the GPER (GPR30 Estrogen Receptor), which in turn, activates the SF1 (Steroidogenic Factor 1), which is a transcription factor that induce aromatase expression in endometrial cells (35). In non-estrogen signaling pathway, a study found that TAM up-regulates the expression of Ki67 and IGF-1,markers of proliferation(36).
Although the first studies by Chen X et al only include 3 recurrent thin endometrium patients who successfully conceived with one miscarriage after tamoxifen treatment(29), their following studies consisting of 61 thin endometrium women showed that the clinical pregnancy rate, early miscarriage rate and implantation rate were 44.3%(27/61)、7.4% (2/27)、24.2%(32/132) respectively. The authors concluded that in patients of recurrent thin endometrium, tamoxifen treatment may be a successful alternative approach(30). The study of Ke H et al showed that when stratified by different etiologies of thin endometrium, patients with PCOS obtained the highest rate of clinical pregnancy (60%) and live birth (55.56%) per transfer. The clinical pregnancy rate and live birth rate in patients with history of intrauterine adhesion were 6/18(33.33%) and (5/18, 27.78%) respectively, while for patients with history of uterine curettage, the clinical pregnancy rate and live birth rate were (39/101, 38.61%) and (32/101,31.68%) respectively(31). However, there were no control groups in the studies of Chen X et al and Ke H et al. Furthermore, they are limited by their small sample size. Different from the previous study, our study compared the effect of TAM with the mostly commonly used protocol- HRT protocol rather than described the reproductive outcome only so that this effect can be analyzed precisely. Overcoming methodological shortcomings in above-mentioned studies and adjusting for a number of important confounders, our study demonstrated that not only the clinical pregnancy rate and the implantation rate but also the ongoing pregnancy rate and live birth rate of TAM group were significantly higher than HRT group.
Despite the improvement of endometrium thickness, another possible mechanism for improved reproductive outcomes with the use of TAM during FET is that TAM might potentially improve corpus luteum function. Wu reported lower spontaneous abortion rates in patients with luteal-phase dysfunction treated with TAM as compared with CC(37).
Cost-effectiveness is a major concern for patients during their infertility treatment. TAM is much cheaper than estrogen agents. In China, TAM costs $42 for the whole cycle whereas estradiol val­erate costs more than $11 per day supply and cannot be withdrawn until 8-10 weeks gestation. Future studies analyzing cost efficiency of these two regimens are needed.
Long-term safety of offspring after IVF treatment is of paramount importance. The short half-life of TAM (86h) and administration of this agent in the early follicular phase from cycle day 5 to day 10 leave a sufficient period of time for complete washout before embryo implantation. Furthermore, previous research suggested that TAM administration during the first trimester did not increase the risk of major congenital anomalies (38).
We recognize there are some important limitations of our retrospective study. First, the retrospective nature of our analysis may subject to selection bias regarding the type of FET protocol (TAM vs.HRT). In this regard, we exclude multiple types of FETs following the same fresh IVF/ICSI cycle. Secondly, endometrial receptivity assay was not performed on both groups of patients. Further studies on this subject are needed to investigate the details of the effect of TAM on the endometrium.
Our study also has some strength. First, our study is the largest in this area by far with 100% follow-up and provides new insight into current practice patterns. Secondly, the laboratory conditions and IVF procedures did not been changed throughout the study period.