Introduction

The global use of advanced techniques such as assisted reproductive technology (ART) has expanded, leading to noteworthy advancements in the treatment of infertility. However, despite these achievements, complex challenges persist in comprehending the intricate process of implantation and enhancing the outcomes of ART. Further research and advancements are necessary to address these complexities and improve the overall success of ART procedures⁽¹⁾.

It is crucial to conduct additional research on current therapeutic approaches to augment the success rates of ART due to the consistently low worldwide frequencies of embryo implantation and gestation⁽²⁾. Among the intricate aspects of ART, implantation is one of the most vulnerable and complex processes⁽³⁾. It depends on numerous local and systemic elements, including immune agents and hormonal cues⁽⁴⁾. Before being implanted, the embryo needs to produce substances that encourage the attachment site, whereas the decidua, conversely, should emit substances that support the differentiation and initial growth of the embryo. Consequently, exploring and understanding these factors are crucial for improving the success of ART^(5,6). Synchronization of endometrial receptivity and embryonic competence in a timely manner is of utmost importance for successful embryo implantation. It is crucial that the mentioned mediators operate within a normal range and at the appropriate time to ensure optimal conditions for this process⁽⁷⁾. Achieving adequate endometrial growth is a vital requirement for successful implantation. However, the precise understanding of the factors influencing endometrial growth remains limited and requires further investigation⁽⁸⁾.

In recent times, significant numbers of studies have focused on investigating angiogenesis and vascularization within the endometrium. These studies have revealed that women with thin endometrium often exhibit poor uterine receptivity, which is potentially attributable to compromised blood flow impedance through the endometrium⁽⁹⁾. Over the past 10 years, several approaches have been examined to improve endometrial thickness among poor endometrial responders^{(10,11,12,13,14)}. Along these lines, sildenafil citrate (SC) has been extensively examined in previous clinical trials to gage its impact on enhancing the success rate of ART. This is attributed to its well-documented vasodilatory and antithrombotic activities, making it one of the most thoroughly researched medications in this regard^(15,16). Numerous clinical trials have been conducted, and a mounting body of evidence suggests a beneficial effect of SC on ART outcomes^(17,18,19).

To date, 28 randomized clinical trials (RCTs) have been performed to investigate the effect of SC on ART outcomes. However, despite these efforts, a definitive conclusion regarding its efficacy remains elusive, and prior studies have not yielded strong evidence⁽²⁰⁾. Considering the substantial debate surrounding the effectiveness of SC, we conducted a contemporary systematic review and meta-analysis of RCTs specifically targeted at appraising the influence of SC on various clinical endpoints of ART.

Methods

This investigation was conducted following the guidelines outlined in the preferred reporting items for systematic reviews and meta-analyses statement⁽²¹⁾ and the Cochrane handbook for systematic reviews of interventions⁽²²⁾. This investigation was registered in the international prospective register of systematic reviews under the identifier CRD42023433884.

Databases and Search Strategy

To identify potential studies, a systematic search was conducted using several databases, including PubMed, Scopus, Embase, PubMed, and the Cochrane Central Register of Controlled Trials. The search encompassed the period from the inception of each database to April 15, 2023. In addition, we manually inspected the reference lists of pertinent published studies with the intention of discovering any extra suitable RCTs. The search query comprisedthe following: (“in vitro fertilization” OR “intracytoplasmic sperm injection” OR “IVF” OR “Embryo Transfer” OR “ICSI” AND “Sildenafil” OR “SC” OR “Hydroxyhomosildenafil” OR “Revatio” OR “Homosildenafil” OR “Acetildenafil” OR “Viagra”). Supplement Table 1 provides comprehensive information regarding the search strategy employed, including the specific terms used and database-specific indexing terminology. No limitations were set based on the year of publication or language during the search process.

Inclusion and Exclusion Criteria

In our review, studies were analyzed if they met the following conditions: (i) the study comprised subfertile patients undergoing ART, (ii) the intervention group received SC alone or in combination with other agents, (iii) the comparator (control) group received a placebo, no treatment or an active agent other than SC, (iv) the study reported at least one of the desired endpoints such as endometrial thickness, chemical pregnancy, clinical pregnancy, live birth or miscarriage and (v) the study design was an RCT. In contrast, our review excluded studies that were in any of the following categories: (i) the study was not an RCT, (ii) the study involved animals, and (iii) the study lacked adequate information on the study methodology or findings.

Study Selection, Data Collection, and Quality Assessment

Following the retrieval of all citations, duplicates were eliminated. Subsequently, the titles/abstracts of the residual citations were examined, and any unrelated citations were excluded. Then, the full-text citations of the residual ones were screened to establish their final eligibility. Two coauthors independently selected the studies, and any disagreements were resolved via consensus.

Data collection involved two distinct categories. The initial category encompassed the fundamental attributes of the eligible RCTs, such as the name of the primary author, year of publication, country, study arms, participant count, specifics of treatment arms, and reported results. The subsequent category comprised the clinical endpoints, namely, endometrial thickness, rate of chemical pregnancy, and rate of clinical pregnancy. The task of collecting data was carried out by two sets of coauthors who worked independently, and disputes were resolved via consensus within each pair.

The Cochrane risk of bias tool was used to assess the methodological quality of RCTs⁽²³⁾. Each of the seven domains was assessed as unclear, low, or high risk of bias. Two coauthors completed the study appraisal independently, and disputes were resolved by discussion with a third co-author.

Statistical Analysis

The Mantel-Haenszel method was used to compute the risk ratio (RR) and its 95% confidence interval (CI) for the contrasting findings of chemical and clinical pregnancy rates.

In contrast, the standardized mean difference (SMD) was used to compute the continuous result of the average thickness of the endometrium, along with its 95% CI, employing the inverse-variance method. The random-effects model was employed for all calculations. The heterogeneity of the RCTs was assessed visually through forest plots and statistically using the chi-square-based Q statistic and I² value. Significant heterogeneity was established when the p-value of the chi-square-based Q statistic measured <0.10 or I² measured >50%. Stata software (version 17) was used to perform all statistical analyses. Subgroup analyses were conducted to explore the impact of SC on ART outcomes, considering related study features. These features included intervention type (e.g., SC alone, SC with clomiphene citrate (CC), SC with estradiol valerate, and SC with letrozole), population type (e.g., recurrent implantation failure (RIF) or other infertility categories), and control type (e.g., CC, estradiol, letrozole, or placebo), which were considered potential sources of variation.

Results

Summary of the Database Screening

In total, 935 records were found across all databases after eliminating duplicates. Of these, 89 records underwent a complete text examination, and among them, 61 records were excluded for specific reasons. Eventually, 28 RCTs were included in the meta-analysis (Figure 1).

Summary of Baseline Characteristics of Eligible RCTs

All RCTs had a parallel design and were performed between 2010 and 2023. The RCTs were conducted in various countries, including Egypt (n=14)^{(24,25,26,27,28,29,30,31,32,33,34,35,36,37)}, India (n=5)^{(38,39,40,41,42)}, Iran (n=4)^{(17,19,43,44)}, Iraq (n=3)^(45,46,47), Russia (n=1)⁽⁴⁸⁾, and Korea (n=1)⁽⁴⁹⁾. Seven RCTs compared SC with placebo, ten RCTs compared the combination of SC and CC with CC, 7 RCTs compared the blend of SC and estradiol valerate with estradiol valerate, three RCTs compared the combination of SC and letrozole with letrozole, and one RCT compared SC with placebo. Table 1 displays the range of sample sizes, with participant numbers varying from 44 to 850. Among the seven RCTs, the focus was on women with recurrent implantation failure. On the other hand, the remaining RCTs examined different forms of infertility, including patients with narrow endometria or undetermined causes of infertility.

Summary of the Study Quality of the Included RCTs

Supplement Table 2 presents an overview of the bias risk of the RCTs that were included. With the exception of 15 RCTs, all other RCTs were determined to have a low risk regarding random generation. Uncertainty regarding random allocation was observed in 13 RCTs. For allocation concealment, 14 RCTs were deemed to have a high risk of bias, whereas eight RCTs had an unclear risk of bias. Performance bias was rated as a high and unclear risk of bias in 16 and four RCTs, respectively. Detection bias was rated as high and unclear risk of bias in 10 and 7 RCTs, respectively. Attrition bias was rated as a high and unclear risk of bias in four and six RCTs, respectively. Fifteen RCTs were concluded to exhibit an unclear risk of bias for the domain of reporting bias.

Meta-analysis of the Rate of Chemical Pregnancy

Chemical pregnancy data were pooled from 11 RCTs with a total of 982 individuals (485 cases and 497 controls). The probability of chemical pregnancy (RR=1.54; 95% CI: 1.27, 1.87) was significantly higher in the intervention group (i.e., SC alone or combination with other treatments) than in the placebo group (i.e., any other active treatment, no treatments, or placebo), without significant heterogeneity (I²=0%; p=0.679) (Figure 2). A counter funnel plot and Egger’s test showed no sign of publication bias (β=0.83, 95% CI: -1.05, 2.73; p=0.343, Supplement Figure 1). The sensitivity analysis indicated that the calculated combined risk RR ranged between 1.50 (95% CI: 1.23, 1.83) and 1.62 (95% CI: 1.32, 1.97). This suggests that none of the individual RCTs had a substantial influence on the overall effect size (Supplement Figure 2).

The impact size was more pronounced in a subgroup of RCTs that provided subcutaneous injections (SC) to RIF women (n=3 RCTs, RR=2.08, 95% CI: 1.22, 3.54, I²=0%), compared with women with other infertility types (n=8 RCTs, RR=1.48, 95% CI: 1.20, 1.81, I²=0%) (Supplement Figure 3). There was no distinction between the subgroup of RCTs that administered SC doses of ≤50 mg (n=5 RCTs, RR: 1.66, 95% CI: 1.25, 2.20, I²=0%) and those that administered SC doses of more than 50 mg (n=6 RCTs, RR=1.45, 95% CI: 1.12, 1.89, I²=0%) (Supplement Figure 4).

Three RCTs with participants experiencing RIF were conducted to compare subcutaneous SC intervention with placebo. Additionally, four RCTs compared the combination of SC and CC with CC alone, three RCTs compared the blend of SC and estradiol valerate with estradiol valerate alone, and one RCT compared the combination of SC and letrozole with letrozole alone. The forest plots, using random-effects analysis, for all the RCTs that compared SC with placebo showed a noteworthy improvement in the chemical pregnancy rate in the SC intervention group compared with the control arm (RR=2.08, 95% CI: 1.22, 3.54, I²=0%). Furthermore, a significantly substantial change was noticed in the subgroup of RCTs that contrasted the blend of SC and CC with CC alone (RR=1.47, 95% CI: 1.15, 1.88, I²=0%). However, no significant change was observed in the subgroup of RCTs that compared the blend of SC and estradiol valerate with estradiol valerate alone (RR=1.31, 95% CI: 0.85, 2.01, I²=6.42%) (Figure 2).

Meta-analysis of the Rate of Clinical Pregnancy

Data on clinical pregnancy were gathered from 21 RCTs involving 2.816 patients (1.401 cases and 1.415 controls). The intervention group, which received either SC treatment alone or in combination with other therapies, exhibited a significantly higher likelihood of clinical pregnancy (RR=1.43; 95% CI: 1.29, 1.59) compared with the control group, which consisted of other active interventions, no intervention, or placebo. No significant heterogeneity was identified (I²=2.36%; p=0.716) (Figure 3). A counter funnel plot and Egger’s test indicated no evidence of publication bias (β=0.74, 95% CI: -0.01, 1.50; p=0.054, Supplement Figure 5). Sensitivity analysis demonstrated that the combined RR estimates varied from 1.40 (95% CI: 1.27, 1.55) to 1.51 (95% CI: 1.33, 1.71), indicating that no individual RCT exerted a significant influence on the overall effect size (Supplement Figure 6).

No significant changes were noticed between the subgroup of RCTs that provided SC treatment to RIF women (n=6 RCTs, RR=1.63, 95% CI: 1.13, 2.33, I²=0%) and those with different infertility causes (n=15 RCTs, RR=1.41, 95% CI: 1.26, 1.56, I²=72%) (Supplement Figure 7). The subset of RCTs that used SC doses of ≤50 mg (n=9 RCTs, RR=1.62, 95% CI: 1.36, 1.93, I²=0%) showed a stronger effect size compared with those that used SC doses of more than 50 mg (n=12 RCTs, RR=1.32, 95% CI: 1.17, 1.49, I²=0%) (Supplement Figure 8).

Six RCTs assessed the use of SC (n=183 patients) in contrast to a placebo (n=183 patients). Furthermore, seven RCTs investigated the combination of SC and CC (n=890 patients) compared with CC alone (n=893 patients). Six RCTs examined the combined administration of estradiol valerate and SC versus estradiol valerate alone, encompassing 203 participants in the experimental group and 209 participants in the control group. Lastly, two RCTs compared the combined use of letrozole and SC versus letrozole alone, involving 125 participants in the experimental group and 125 participants in the control group.

Pooling the outcomes of six RCTs that compared the rates of successful pregnancies between the SC (study compound) and placebo groups revealed a considerably higher likelihood of achieving clinical pregnancy in the SC group (RR=1.59, 95% CI: 1.10, 2.30, I²=0%). Women who received both SC and CC had significantly higher chances of clinical pregnancy compared with those who received monotherapy CC (RR=1.35, 95% CI: 1.19, 1.53, I²=0%). Remarkably, women who received both SC and estradiol valerate experienced a notable increase in clinical pregnancy rates compared with those who only received estradiol valerate (RR=1.55, 95% CI: 1.08, 2.22, I²=0%). Conversely, no statistically significant change was observed between the intervention and control groups in a subset of RCTs that administered the combination of letrozole and SC versus letrozole alone (n=2, RR=1.78, 95% CI: 0.98, 3.24, I²=45.42%) (Figure 3).

Meta-analysis of the Diameter of the Endometrial Thickness

Data regarding the thickness of the endometrium were gathered from ten RCTs involving 748 participants, with 374 being cases and 374 being controls. The findings revealed that women who received an intervention, either subcutaneous administration alone or in combination with other treatments, had a considerably higher average endometrial thickness (SMD=0.77, 95% CI: 0.20, 1.34; I²=92.72%) than the control group. The control group included various interventions such as other active treatments, no intervention, or a placebo (Figure 4). A counter funnel plot and Egger’s test were conducted, which indicated no indication of publication bias (β=5.40, 95% CI: -9.02, 19.84; p=0.413). Sensitivity analysis demonstrated that the combined RR estimates varied from 0.58 (95% CI: 0.12, 1.03) to 0.89 (95% CI: 0.39, 1.47), implying that no individual RCT had a substantial impact on the overall effect size (Supplement Figure 9).

There was a notable distinction observed between a specific group of RCTs that used SC treatment in patients experiencing RIF (consisting of 4 RCTs, with a SMD of 1.33, 95% CI: 0.15, 2.51, and an I² value of 94.35%) and another group that administered SC treatment to patients with different forms of infertility (comprising 6 RCTs, with an SMD of 0.43, 95% CI: -0.02, 0.88, and an I² value of 83.35%, Supplement Figure 10). The magnitude of the effect was more pronounced in the subset of RCTs that used SC doses exceeding 50 mg (including 7 RCTs, with an SMD of 0.65, 95% CI: 0.08, 1.22, and an I² value of 89.89%) than in those that employed SC doses of 50 mg (consisting of 3 RCTs, with an SMD of 1.09, 95% CI: -0.42, 2.67, and an I² value of 96.37%, Supplement Figure 11).

Four RCTs conducted a comparison between SC administration and placebo, with a total of 127 patients in the case group and 127 participants in the control group. Among them, one RCT compared the combined use of SC and CC (n=45 patients) with the use of CC alone (n=45 patients). Furthermore, another RCT contrasted the combined use of SC and letrozole with the use of letrozole alone, with 75 cases and 75 controls. Moreover, an additional RCT compared SC (n=35 patients) with G-CSF (n=35 patients). Lastly, one RCT compared the combined use of estradiol valerate and SC (n=40 patients) with the use of estradiol valerate alone (n=40 patients).

After the intervention, patients who were administered SC experienced an increase in endometrial thickness in comparison with those who received a placebo (n=4 RCTs, SMD=1.33, 95% CI: 0.15, 2.51, I²=94.35%). The average endometrial thickness was significantly higher in women who received both SC and CC than in those who only received CC (n=1 RCT, SMD=0.92, 95% CI: 0.49, 1.36). No notable change between patients who received a blend of SC and estradiol valerate versus those who received estradiol valerate alone was noted (n=1 RCT, SMD=0, 95% CI: -0.43, 0.43). Additionally, women who received SC exhibited a significantly higher endometrial thickness compared with those who received G-CSF (n=1 RCT, SMD=0.67, 95% CI: 0.19, 1.14) (Figure 4).

Discussion

Understanding the specific function and operational method of SC in the implantation process is an intricate issue. However, various potential theories have been proposed. Initially, SC hinders the activity of phosphodiesterase 5 (PDE5), an enzyme responsible for breaking down cyclic guanosine monophosphate (cGMP). Through the application of SC, the levels of cGMP remain heightened, resulting in the relaxation of blood vessels and an augmented blood circulation toward the endometrium⁽⁵⁰⁾. Second, SC could influence vasoactive cytokines responsible for governing the growth of the uterine lining or the attachment of the embryo⁽¹⁹⁾. Third, the use of SC can improve the readiness of the uterus by aiding in the growth of spiral arteries and augmenting the flow of arterial blood within the uterine region. Fourth, SC may increase natural killer cell activity in addition to the role of SC in promoting endometrial growth facilitation⁽⁵¹⁾. The fifth point is that SC can trigger the angiogenic reactions of the vascular endothelial growth factor. This involvement in the formation of new blood vessels and heightened vascular penetrability in the middle secretory stage is fundamental to the achievement of satisfactory implantation⁽⁵²⁾.

In our current systematic review and meta-analysis, we examined 28 RCTs involving 3.426 women with subfertility. Among these participants, 1.702 were assigned to receive the intervention, whereas 1.724 were placed in the control group. All individuals underwent ART. Based on the collective findings, it is evident that the use of SC, either alone or in conjunction with other active treatments such as CC, estradiol valerate, or letrozole, can lead to higher chemical pregnancy rate, clinical pregnancy rate, and endometrial thickness compared with the control group. However, it is essential to approach these findings with concern because the quality of the RCTs included was not up to par. When subgroup analysis was conducted, it was found that women with RIF experienced the most significant increase in chemical and clinical pregnancy rates compared with patients with other infertility causes. Various RCTs included infertile patients with unexplained infertility or narrow endometrium. Considering the various sources of infertility investigated in the eligible RCTs and the inadequate number of available RCTs, it was impossible to conduct a proper subgroup analysis. Therefore, it is important to be cautious when interpreting the results of this analysis.

Furthermore, due to inadequate information on the methodologies used in the included studies, we could not perform subgroup analysis considering various potential factors contributing to heterogeneity, such as the specific dose of SC, the stage of embryos during transfer, or the type of protocol employed^(24,25,26). Moreover, although we did identify a noteworthy distinction between women who were administered SC and those who received a placebo, the available evidence of endometrial thickening was of extremely poor quality, exhibiting imprecise results and significant heterogeneity. Lastly, none of the RCTs included in our analysis offered data on live birth rates or side effects, resulting in a lack of conclusive evidence regarding differences between the groups in these aspects.

Prior to our investigation, many systematic reviews and meta-analyses had examined the effectiveness of various treatments (such as gonadotropin-releasing hormone agonist⁽⁵³⁾, aspirin⁽¹²⁾, human chorionic gonadotrophin⁽⁵⁴⁾, and a mixture of alpha-tocopherol plus pentoxifylline⁽⁵⁵⁾ in preparing the endometrium for patients experiencing ART. Nonetheless, the existing evidence was insufficient to establish a specific procedure for preparation of the endometrium. We came across only one meta-analysis that assessed the clinical utility of vasodilation-promoting agents in patients undergoing ART. This analysis spanned six data sources and encompassed 15 studies and 1.326 patients. All meta-analyzed RCTs contrasted a vasodilator with a control group comprising no treatment or placebo. Three of the RCTs examined the clinical utility of SC. Two of the RCTs judged the efficacy of monotherapy SC compared with placebo, while the other RCT assessed the blend of SC and estradiol valerate compared with no treatment. The meta-analysis of the two articles comparing SC alone with placebo indicated no discernible differences between the two treatments. Likewise, an investigation comparing the blend of SC and estradiol valerate indicated no significant changes between the two interventions⁽²⁰⁾. In addition to these three RCTs, we identified 25 additional RCTs that evaluated the efficacy of SC either alone or in conjunction with other active therapies. The most extensive RCT to date, conducted in Egypt, examined the effectiveness of combining SC and CC compared with CC alone and was well-executed.

At Ain-Shams University Women’s Hospital, an RCT was conducted involving 850 women diagnosed with PCOS who had previously experienced failure with CC. The study compared two groups, one receiving SC treatment and the other receiving CC treatment. The SC group demonstrated a greater thickness of the endometrial lining and had a higher likelihood of achieving a clinical pregnancy than the CC group⁽²⁹⁾.

The methodological quality and similarity of the studies included have a sizable influence on the reliability of the meta-analysis results. In our study, the quality of the trials included was comparatively poor, but there was little variation in the analysis of clinical and chemical pregnancy rates. However, we found high heterogeneity in the analysis of endometrial thickness. Although we did not find any statistical heterogeneity in the main outcomes across the trials, we cannot disregard the potential impact of clinical heterogeneity on the study findings. While most RCTs examined in the review had clearly defined criteria for including and excluding participants, a small number of RCTs failed to specify their precise criteria. In addition, the RCTs lacked consistent or unreported information about the primary source of infertility and the history of RIF or repeated abortion among the women included in the research. There was also variation across the RCTs with respect to the methods used for luteal stage stimulation, endometrial setup, and ovarian hyperstimulation regimens, with some studies failing to report these specific details. Moreover, there were notable variations in the dosage of SC, a crucial element that significantly impacts the efficacy of the given medication, across the various studies. Unfortunately, most studies did not report miscarriage rates, drug-related adverse events, or pregnancy outcomes, thereby limiting our understanding of these important factors.

Regarding the effects of SC on frozen-thawed embryo transfer (FET) cycles in the setting of ART, our results from included studies advocate that SC may positively affect endometrial receptivity, potentially enhancing blood flow to the uterus and improving the implantation rates of embryos during FET⁽¹⁹⁾. The medication is thought to exert influence by relaxing uterine smooth muscles and increasing blood flow, which could contribute to a more favorable environment for embryo implantation⁽⁵⁶⁾. However, research on this topic is ongoing, and while some studies have shown promising results, further investigation is needed to establish the optimal dosage, timing, and overall efficacy of SC in the context of FET cycles⁽⁵⁷⁾.

There were various shortcomings in our review. First, the majority of the RCTs we analyzed depicted inadequate methodological quality and harbored a small number of patients. These factors can affect the reliability of the studies. Second, there was significant variability in some results, especially with regard to endometrial thickness. However, we did not perform subgroup analyses to investigate the potential causes of this variability, such as the specific dose of SC, stage of embryo transfer, or type of protocol used. This was because the analyzed studies did not offer sufficient detailed data about their methods.

Conclusion

The present systematic review and meta-analysis revealed that luteal administration of SC, whether orally or vaginally, either alone or as an adjuvant therapy, may enhance endometrial thickness and clinical pregnancy rate in women experiencing ART. Considering the methodological limitations of the analyzed RCTs, this evidence does not yet endorse the routine use of SC in clinical practice. To establish its efficacy and safety more conclusively, future high-quality RCTs with larger sample sizes are necessary. Future RCTs should prioritize different elements, including processing methods, embryo stage, embryo quality, dosage, administration timing, and the composition of the control group. These studies aim to pinpoint particular patient groups that would derive the greatest advantages from this drug administration and establish the ideal dose, timing, and type of SC that would produce the most favorable outcomes while minimizing potential adverse events.

Ethics

Peer-review: Internally and externally peer-reviewed.

Authorship Contributions

Concept: S.B., A.A-Z., Design: S.B., A.A-Z., Data Collection or Processing: S.B., M.A., M.S.A., H.T.S., W.H.A., E.B., A.Y.A., I.A.B., A.A., O.A., A.A-Z., Analysis or Interpretation: S.B., M.A., O.A., A.A-Z., Literature Search: M.A., M.S.A., H.T.S., W.H.A., E.B., A.Y.A., I.A.B., A.A., O.A., Writing: S.B., A.A-Z.

Conflict of Interest: No conflict of interest was declared by the authors.

Financial Disclosure: The authors declared that this study received no financial support.