Effectiveness of hormone add-on strategies in ovarian stimulation for women with poor ovarian response: a systematic review and network meta-analysis of randomized controlled trials.

PURPOSE: Despite various stimulation protocols and adjuvant treatments, optimizing the reproductive success of women with poor ovarian response (POR) undergoing assisted reproduction techniques (ART) remains a challenge. The aim was to evaluate and compare the ovarian stimulation and reproductive outcomes of hormonal add-ons in women with POR undergoing ART.

METHODS

MEDLINE, LILACS, SCOPUS, EMBASE, Scielo.br, PROSPERO, CINAHL, PsycINFO, AMED, Clinicaltrials.gov, ICTRP, Cochrane Library, and conference proceedings were searched for randomized controlled trials without temporal, geographic, and language restrictions. We included only RCTs that allocated women with POR, according to Bologna criteria, undergoing ART to at least two different ovarian stimulation protocols, with the treatment arm including at least one hormonal add-on. A random-effect network meta-analysis was performed for mixed multiple treatment comparisons to rank available add-ons by the surface under the cumulative ranking curve area (SUCRA). Primary outcome was the live birth rate.

RESULTS

Twenty-two studies (4131 women) were directly and indirectly compared. Concerning the live birth rate, no significant differences among add-ons were noted (very low evidence), with testosterone (SUCRA = 34.0%) showing the highest probability of the best treatment. For the clinical pregnancy rate, according to SUCRA ranking, human growth hormone (SUCRA = 46.3%) and testosterone (SUCRA = 44.6%) had increased chances of being ranked first, with growth hormone being significantly more efficacious than estrogens (OR 3.46 [95% CI 1.59 to 7.53]; low evidence) while recombinant luteinizing hormone was significantly less efficacious (OR 0.50 [95% CI 0.26 to 0.96]; very low evidence). Regarding the overall number of retrieved oocytes and the mean number of metaphase II oocytes, human growth hormone was confirmed best ranked (SUCRA = 53.2% and SUCRA = 67.9%). Letrozole had significantly less gonadotropins used than controls (SMD - 7.02 [95% CI - 12.82 to - 1.22]; low evidence) (SUCRA = 67.0%) and the smallest stimulation duration (SUCRA = 52.0%).

CONCLUSION

Low to very-low evidence shows that women with POR undergoing controlled ovarian stimulation may benefit from adding human growth hormone or testosterone for improved reproductive outcomes. However, additional high-quality randomized controlled trials are needed to overcome the limitations of the current literature.

TRIAL REGISTRATION: PROSPERO CRD42024618797, date: 25 November 2024.