Endometrial preparation was started in all recipient patients from day 2 of menses. Estradiol valerate 6 mg/d (Progynova^TM, 2 mg Tablet, Zydus Cadila, Pharmaceuticals, India) was started on the second day of the menstrual cycle to achieve an endometrial thickness of > 7.0 mm. In case inadequate endometrial thickness was observed, the dose of estradiol valerate was increased up to 12 mg to achieve > 7 mm endometrial thickness. On the oocyte retrieval day of the donor, a progesterone injection of 100 mg/ml/day (Gestone^TM Injection, Ferring Pharmaceuticals, India) was administered to the recipient along with estradiol valerate. Day 5 embryo transfer was performed with fresh embryos. 2 good-quality blastocysts were transferred. Estradiol valerate and progesterone were continued at the same dose for 2 wk following the transfer. Serum β human chorionic gonadotropin was performed 2 wk later; in the case of positive values, the same treatment was continued until 12 wk of gestation.

2.4. Outcome measures
The primary outcome was the live birth rate in recipients. Secondary outcomes related to donors included M2 oocyte proportion percent at oocyte retrieval, duration of stimulation, total consumption of gonadotropins, blastulation rate, and good-quality blastocyst proportion. Secondary outcomes related to recipients were biochemical pregnancy, implantation rate, and clinical pregnancy rate.

2.5. Ethical considerations
The study protocols were approved by Sharda Narayan Ethics Committee, Mau, India (SNH/Ethics/001), and informed and written consent was taken from each participant.

2.6. Statistical analysis
MS Excel spreadsheet was used to code and record data. SPSS v23 (IBM Corp. Released 2015. IBM SPSS Statistics for Windows Version 23.0. Armonk, NY: IBM Corp.) was used for data analysis. Means/standard deviations and median/IQRs for continuous variables, forms were used for expressing descriptive analysis and categorical variables expressed as frequencies and percentages, independent sample t test was used for group comparison of continuously distributed data when comparing 2 groups. In case of uneven distribution, appropriate non-parametric tests, Wilcoxon test/Kruskal Wallis test were used for comparisons. Categorical data were analyzed using the Chi-squared test for group comparisons Binary logistic regression method was used to ascertain the significant predictors of dichotomous outcomes and linear regression for continuous ones. The individual coefficients/odds ratios and significance for each of the variables were first performed using univariable regression then, the backward stepwise variable selection was used to choose the best set of predictor variables. Finally, multivariable linear/logistic regression was performed to get the final model. Statistical significance was kept at p < 0.05.

3. Results
A total of 250 oocyte donors fulfilling all the criteria of oocyte donation as per Indian Council of Medical Research guidelines, who received an OS with 2 protocols, were studied. The pregnancy outcomes in 384 recipients who received oocyte donation or embryo donation (using donor sperm and donor oocyte) from donors using 2 different OS protocols were studied. Cycle characteristics of recipient groups are depicted in table I. The mean age of recipients was 38.80 ± 6.37 and 38.39 ± 6.32 yr in both groups, respectively; BMI and Anti Mullerian hormone were comparable in both groups (p = 0.47, p = 0.17). Sperm source (self or donor) was statistically significant (p = 0.019) with more self-sperm ICSI in group 2. Indication for IVF had a significant difference (p = 0.023) in both groups. All oocytes retrieved from both donor groups were inseminated by ICSI. Out of the 181 recipients in group 1, donor sperm was used in 77 recipients, while out of the 203 recipients in group 2, donor sperm was used in 66 recipients. 2 good-quality blastocysts were transferred on day 6 of progesterone supplementation in both groups. On analysis, as depicted in table II, biochemical pregnancy was 74% vs. 73.4% in groups 1 and 2, respectively implantation and clinical pregnancy were comparable in both groups. The implantation rate with twin sacs was significantly lower in group 2. Out of the 110 ongoing pregnancies, 107 were delivered in group 1. There were 123 deliveries from 130 ongoing pregnancies in group 2. Table III depicts the regression models for all pregnancy outcomes in both univariable and multivariable analyses. Univariable and multivariable regression was done using group 1, group 2, sperm source, age, BMI, recurrent implantation failure, and endometrial thickness as predictors and pregnancy outcomes as a dependent. Group 1 categorical variable is being used as the reference category.
On comparing group 1 (GnRH antagonist) with group 2, univariable analysis provided unadjusted OR 0.97 (0.61-1.53, p = 0.893) for biochemical pregnancy, 1.02 (0.42-2.49, p = 0.962) for implantation, 1.78 (0.57-5.61, p = 0.322) for clinical pregnancy, 2.67 (0.80-8.90, p = 0.111) for ongoing pregnancy, and 0.49 (0.12-1.96, p = 0.316) for live birth. Multivariable regression analysis of the above dependent variables was not statistically different. Unadjusted OR 0.57 (0.33-1.00 p = 0.049) for implantation rate of twin sacs was significantly less in group 2 (MPA). Multivariable adjusted OR for implantation rate provided a statistically significant difference in twin sac implantation. However, live birth and the number of live-born fetuses were not statically different in both groups. Multivariable adjusted OR 0.57 (0.31-1.05, p = 0.073) for live-born fetuses showed no significant difference as depicted in table IV. Demographic details and cycle characteristics of donor groups are in table V. The mean age, and BMI of donors showed no significant difference in both groups. The duration of stimulation was 11.03 ± 0.80 days in group 1 and 11.0 ± 0.74 days in group 2. Gonadotropin consumption in group 1 was 2428 ± 278 and in group 2 was 2251 ± 141 (p = 0.400). The M2 oocytes retrieved in both groups were 14.60 ± 8.26 and 13.40 ± 6.53, respectively. The Blastulation rate was better in group 2 but not statistically significant (p = 0.165). All-day 5 embryos reaching the morphological stage of 3AA and above (according to the Gardner staging system) were labeled as good-quality blastocysts. Good-quality blastocysts were comparable in both groups. Regression analysis was performed for M2 oocyte proportion percent and good-quality embryo proportion percent as the dependent variable. Group 1, group 2, age, BMI, stimulation dose, and retrieved oocyte were taken as predictors (Table VI). M2 oocyte proportion percent, univariable analysis 0.79 (-1.74 to 3.32, p = 0.538) did not show a significant difference in both the groups. On further multivariable analysis, though there was a slightly higher M2 oocyte proportion percent in group 2, the difference was not statistically significant 1.08 (-1.47 to 3.62, p = 0.406). Retrieved oocyte number independently affected the M2 oocyte proportion percent, after adjusting confounding factors in both univariable 0.01 (0.00-0.01, p = 0.002) and multivariable 0.01 (0.00-0.01, p = 0.005), however, regression analysis for good-quality blastocyst proportion was comparable using both univariable and multivariable analysis in between the groups. It can be concluded from the result of regression analysis that embryological outcomes especially M2 oocyte proportion and good-quality blastocysts proportion were quite comparable in both the groups.
The mean stimulation cost for rFSH and GnRH antagonist protocol in group 1 was 1,107.74$ while in group 2 with rFSH and MPA protocol, it was 868.82$. A difference of 239.74$ was observed in stimulation costs, MPA protocol being more cost-effective. With a monthly per capita income of 153$, OS with MPA appears to be a more effective and convenient option as shown in figure 1.








4. Discussion
The use of GnRH antagonist vs. MPA had a comparable live birth in both groups. In this study, the authors found embryological characteristics to be comparable in both groups. Although live birth was comparable, with 2 embryos transferred in both the groups, twin sac implantations were significantly less in group 2. Comparative studies of a progestin for prevention of LH surge, and its subsequent embryological and pregnancy outcomes in OS of patients followed by FET in the next cycle have shown similar results and comparable outcomes (2, 8). There is a study comparing GnRH antagonist versus MPA in oocyte donation cycles. Our results are by the previous report of Beguería and colleagues (where the duration of stimulation, as well as the total gonadotropin dose up to trigger, was similar in GnRH antagonist and MPA-treated patients) (9).
As per the definition of premature LH surge, no premature LH surge was observed with the use of progestins to inhibit spontaneous ovulation during OS (10) or with the use of GnRH antagonists (11). In our study, MPA was started from day 2 of the menses along with rFSH injection. As a result, there was consistent LH suppression with no incidence of premature LH surge. In previous studies, there was no case of premature LH surge reported in PCOS patients receiving daily MPA 10 mg as well as in normal responders co-administered with either 4 mg or 10 mg MPA per day during OS (12, 13). We observed that embryological characteristics were similar in both groups. M2 oocyte proportion showed slightly higher retrieval in group 2 but it was not statistically significant when multivariable regression analysis was performed, unlike the finding of significantly higher yield of M2 oocyte as stated in a retrospective study of donor oocyte cycle (14). It was a small study on 13 oocyte donors. Our finding is in accordance, with the earlier comparative study where MPA was comparable to GnRH antagonist in terms of the number of mature oocytes retrieved at ovum pick up in oocyte donation cycles and was non-inferior to GnRH antagonist. This study comparing both drugs in the oocyte-donation cycle had comparable M2 retrieval but showed negative pregnancy outcomes in recipients (9); however, the intention-to-treat multilevel analyses had a p-value = 0.05 for live birth rate. In our study, live births were not significantly different in groups. The effectiveness of MPA on the clinical pregnancy rates, implantation rates, and live-birth rates was previously studied and was found to be comparable (2, 13).
For proper interpretations of the results of pregnancy outcomes, one should keep in mind that, in most studies, a comparison of MPA was done with GnRH agonist or other progestins in short GnRH agonist protocols, rather than a GnRH antagonist protocol (2, 13). The current study, however, analyzed the results of the use of MPA in a GnRH antagonist protocol. The use of this protocol is recommended in oocyte donation cycles to reduce the incidence of ovarian hyperstimulation syndrome (15, 16). In addition, we administered rFSH to all patients and donors as observed in previous studies (15, 9), few authors used hMG, having a small amount of urinary human chorionic gonadotropin (HCG), (13, 2). Moreover, we used the GnRH agonist trigger for donors to minimize the risk of ovarian hyperstimulation syndrome, as used in some studies (17, 18), while in some studies the trigger was performed with HCG (2) or by HCG and triptorelin (14).
Although vitrification of embryos is mandatory when used in patients, thereby increasing the cost of treatment, its cost-effectiveness, flexibility, and convenience cannot be denied in the oocyte donation cycle where FET is the protocol of choice. In India, most of the estimated 1.2 billion people have to pay healthcare expenses out of their own pockets. A < 15% of the population has any kind of healthcare coverage, especially in rural areas. With a per capita income of 153$ (2019-2020), any possible cost-effective alternative which is as efficient and comparable to conventional treatment protocols should be further explored. In this study, we have analyzed both self-sperm and donor-sperm as sperm sources that could have an impact on pregnancy outcomes. The difference in indication for IVF also could have contributed to pregnancy outcomes.
Previous studies have heterogeneity in the protocols used, so interpretations of results cannot be properly validated. Limited studies are comparing GnRH antagonist vs. MPA in oocyte donation cycles. Further comparative studies that take into account all confounding factors with the use of uniform and most feasible protocol, preferably prospective randomized controlled trials on larger populations, are required to validate the pregnancy outcomes with the use of MPA in OS protocols as opposed to those of conventional protocols during ART procedures, especially with oocyte donation cycle.

Conclusion
Based on the findings of this study, the authors conclude that MPA may be as effective as GnRH antagonists to achieve positive reproductive and pregnancy outcomes in the oocyte donation cycle.

Acknowledgments
The current study is not funded by any government or private organization. All authors have unanimously given consent and approval for the study with no conflict of interest.

Conflict of Interest
None to declare.