Role of flibanserin in managing hypoactive sexual desire disorder in women: A systematic review and meta-analysis

Abstract

Background:

Flibanserin, approved for the treatment of hypoactive sexual desire disorder (HSDD) in females, has demonstrated diverse therapeutic and adverse effect (AE) prospects in the extant randomized controlled trials (RCTs). This meta-analysis aimed to characterize the outcomes of flibanserin use in these patients comprehensively.

Methods:

RCTs involving women with HSDD receiving flibanserin in the intervention arm and placebo in the control arm were sought after throughout the electronic databases. The primary outcomes were the changes from baseline in satisfying sexual events (SSE) per month and sexual desire score per month measured using an electronic diary (eDiary).

Results:

From 478 initially screened articles, data from 8 RCTs involving 7906 women with HSDD were analyzed. In premenopausal women, flibanserin 100 mg was superior to placebo in improving the number of SSE per month (mean difference, MD 0.69, 95% CI [0.39, 0.99]), eDiary sexual desire score (MD 1.71, 95% CI [0.43, 2.98]), Female Sexual Function Index (FSFI) desire domain (FSFI-d) score (MD 0.30, 95% CI [0.29, 0.31]), FSFI total score (MD 2.51, 95% CI [1.47, 3.55]), Female Sexual Distress Scale-Revised (FSDS-R) Item 13 score (MD −0.30, 95% CI [−0.31, −0.29]), and FSDS-R total score (MD −3.30, 95% CI [−3.37, −3.23]). Compared to placebo, a higher number of premenopausal women using flibanserin 100 mg achieved improvements in the Patient’s Global Impression of Improvement score (OR 1.93, 95% CI [1.58, 2.36], P < .00001) and responded positively at Patient Benefit Evaluation (PBE) (odds ratio, OR 1.76, 95% CI [1.34, 2.31], P < .0001). Postmenopausal women receiving flibanserin 100 mg also benefited in terms of the number of SSE per month, FSFI-d and total scores, FSDS-R Item 13 and total scores, and PBE response. Although flibanserin use was associated with higher risks of dizziness, fatigue, nausea, somnolence, and insomnia, these adverse events were mild in nature; the serious AEs and severe AEs were comparable between the flibanserin and placebo groups.

Conclusion:

While flibanserin has demonstrated efficacy in the treatment of HSDD in both pre- and postmenopausal women, its therapeutic advantages may be overshadowed by the higher likelihood of AEs.

1. Introduction

Female sexual dysfunction (FSD) is common, even more common than male sexual dysfunction. There is limited data on the incidence and prevalence of FSD.^[1] The fourth International Consultation on Sexual Medicine consensus statement asserts that, across a variety of assessment methods, the prevalence of women’s sexual complaints, regardless of age, is on the order of 40% to 50%.^[2] Hypoactive sexual desire disorder (HSDD) is the most prevalent sexual complaint in women. HSDD is defined in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Revised, as persistent deficient sexual fantasies and desire for sexual activity that causes marked distress or interpersonal difficulty.^[3] In the Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition, HSDD has been replaced by Female Sexual Interest/Arousal Disorder, merging arousal and desire disorders.^[4] It is estimated that about one in 10 women have HSDD. The prevalence varies with age; 8.9% of women ages 18 to 44, 12.3% ages 45 to 64, and 7.4% >65 years have HSDD.^[5] Despite its high prevalence, effective pharmacotherapy for HSDD is still limited. Only two medications, flibanserin and bremelanotide, have been approved by the United States Food and Drug Administration (US FDA) for treating HSDD in premenopausal women. Bremelanotide is a melanocortin 3 and 4 receptor agonist, taken as a subcutaneous injection before sexual activity. Flibanserin is a 5-hydroxytryptamine (5-HT) 1A receptor agonist, 5-HT2A antagonist, taken as a bedtime tablet.^[1,5,6] The US FDA approved flibanserin, supported by findings from 3 randomized placebo-controlled trials involving premenopausal women with HSDD. These trials demonstrated statistically significant enhancements in sexual desire, desire-related distress, and the frequency of satisfying sexual events (SSE) when administered at a dosage of 100 mg (daily at bedtime) compared to a placebo.^[6] Common adverse events of flibanserin include dizziness, somnolence, nausea, fatigue, insomnia, and dry mouth.^[1,5,6] Coadministration of alcohol may increase the incidence of sedation-, hypotension-, and syncope-related adverse events of flibanserin.^[7]

Many randomized controlled trials (RCTs) and observational studies have reported the efficacy and safety of flibanserin in pre- and postmenopausal women with HSDD, and there are marked heterogeneities in the efficacy and safety outcomes data. Although some meta-analyses and pooled analyses have been conducted to summarize the efficacy and safety of flibanserin, they have not included all available RCTs to date. Those meta-analyses did not perform separate analyses of flibanserin’s efficacy and safety in pre- and postmenopausal women and for flibanserin 50 and 100 mg.^[8–12] Consequently, it became imperative to conduct a new meta-analysis incorporating the findings of the latest RCTs to gain a more comprehensive understanding of the efficacy and safety of flibanserin in women with HSDD.

2. Methods

This meta-analysis complied with the guidelines outlined in the Cochrane Handbook for Systematic Reviews of Interventions and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) checklists.^[13] The predetermined protocol has been officially recorded in PROSPERO, with a registration number CRD42023471856, and is accessible online. Since ethical approval was already granted for the individual studies, there wasn’t any need for additional approval for this meta-analysis.

A thorough investigation was conducted by searching multiple databases and registers, such as MEDLINE (via PubMed), Scopus, Google Scholar, Cochrane Central Register, and ClinicalTrials.gov. The search covered the period from these sources’ inception until January 20, 2024. The search strategy utilized a Boolean approach with the terms (flibanserin) AND (hypoactive sexual desire disorder) OR (female sexual interest/arousal disorder); the search terms were applied to titles only. A thorough and careful search was conducted to find any recently published or unpublished clinical trials in English. This search included examining references within the RCTs included in this study and relevant journals.

The selection of RCTs for this meta-analysis was based on the PICO criteria. The patient population (P) consisted of women with HSDD; the intervention (I) was the administration of flibanserin for managing HSDD; the control (C) included women receiving a placebo; and the outcomes (O) included the change from baseline (CFB) to study end in the number of SSE, among other outcomes. This analysis included RCTs with pre- or postmenopausal women aged ≥ 18 years and a diagnosis of HSDD. The trials had at least 2 treatment arms/groups, one receiving flibanserin and the other receiving a placebo. Exclusion criteria excluded animal or healthy human trials, nonrandomized trials, retrospective studies, conference proceedings, letters to editors, case reports, and articles lacking data with outcomes of interest.

The primary efficacy outcomes of this meta-analysis were the CFB to study end in the SSE per month and sexual desire score per month measured using an electronic diary (eDiary). Additional efficacy outcomes encompassed the CFB in the Female Sexual Function Index (FSFI) desire domain (FSFI-d) score, FSFI total score, Female Sexual Distress Scale-Revised (FSDS-R) Item 13 score, FSDS-R total score, Patient’s Global Impression of Improvement (PGI-I), and Patient Benefit Evaluation (PBE). Safety outcomes included any adverse event (AEs), drug-related AEs, serious AEs, severe AEs, AEs leading to study discontinuation, and common AEs, including dizziness, headache, fatigue, nausea, somnolence, and insomnia. The analyses of the efficacy outcomes were stratified according to the menstrual status of the subjects (premenopausal or postmenopausal) and the dose of flibanserin. Safety outcomes were stratified according to the dose of flibanserin.

Data extraction was independently conducted by 5 review authors using standardized data extraction forms, with details provided elsewhere.^[14] The handling of missing data has also been elaborated upon in the same source.^[14] Five authors independently performed the risk of bias assessment using the risk of bias tool in Review Manager (RevMan) Web 2023 Version. Specific biases have been outlined in the same source.^[14]

Outcomes were expressed as mean differences (MD) for continuous variables. The results were reported as odds ratios (ORs) or risk ratios (RRs) with 95% confidence intervals (CIs) for binary outcomes. The data was aggregated, employing random effect models to analyze the primary and secondary outcomes. The RevMan web platform (2023 version, developed by Cochrane, based in Oxford, UK) enabled the comparison of MD for primary and secondary outcomes between the flibanserin (experimental drug) and placebo groups in the included studies. The results were reported as 95% confidence intervals (CIs). Forest plots, created using RevMan web software, portrayed outcomes, with the left side favoring flibanserin and the right side favoring the placebo. A significance level of P < .05 was used. The results included forest plots incorporating data from at least 2 RCTs.

The evaluation of heterogeneity was initially performed by analyzing forest plots. Afterward, a Chi² test was conducted with N-1 degrees of freedom and a significance level of 0.05 to ascertain the statistical significance. Additionally, the I² test was utilized in the further analysis.^[15] The details of interpreting I² values have already been elaborated elsewhere.^[14]

The GRADE methodology assessed the quality of evidence about each meta-analysis outcome.^[16] The process of creating the summary of findings table and evaluating the quality of evidence as “high,” “moderate,” “low,” or “very low” has been previously described.^[14] Publication bias was evaluated using funnel plots, in which studies falling outside the inverted funnel plot indicated the presence of substantial publication bias.^[17]

3. Results

3.1. Search results

The study selection process is illustrated in Figure 1. Initially, 478 articles were identified; following the screening of titles and abstracts and subsequent full-text reviews, the number of studies considered for this meta-analysis was narrowed to 10. An exhaustive assessment resulted in the incorporation of 8 RCTs encompassing 7906 individuals diagnosed with HSDD, all of whom satisfied the specified inclusion criteria.^[18–25] One study was excluded as it was an open-label extension study that included women who had completed one of the previous 5 clinical trials of flibanserin and had no placebo control group.^[26]

Figure 1.

Flowchart on study retrieval and inclusion in the meta-analysis.

3.2. Study characteristics

All of the 8 RCTs included in this meta-analysis were phase 3 studies. Six studies were conducted among premenopausal women;^[18–22,25] and 2 among postmenopausal women.^[23,24] In one study, the women with HSDD who showed a predefined response to flexible-dose flibanserin (50 mg or 100 mg/d) during a 24-week open-label period were randomized to either flibanserin or placebo for the double-blind period.^[19] In other studies, study subjects never took flibanserin before the randomization. Three studies used flibanserin 100 mg every night at bedtime, and one used flibanserin 50 to 100 mg every night at bedtime in the single intervention arm; these intervention arms were grouped as flibanserin 100 mg.^{[19,20,23,24]} Two studies had 2 intervention arms and used flibanserin 50 mg every night at bedtime (grouped as flibanserin 50 mg) and flibanserin 100 mg every night at bedtime (grouped as flibanserin 100 mg).^[18,22] One study had 3 intervention arms and used flibanserin 25 mg twice a day (omitted from the analysis), flibanserin 50 mg every night at bedtime (grouped as flibanserin 50 mg), and flibanserin 50 mg twice a day (grouped as flibanserin 100 mg).^[21] The remaining study also had 3 intervention arms: flibanserin 25 mg twice a day (grouped as flibanserin 50 mg), flibanserin 50 mg twice a day (omitted from the analysis), and flibanserin 100 mg every night at bedtime (grouped as flibanserin 100 mg).^[25]

All the studies were of 24-week duration. Along with other outcomes, SSE was one of the primary outcomes of all the included studies. The details of the included and excluded studies are shown in Tables 1 and 2, respectively.

Table 1.

The basic characteristics of the included randomized controlled trials and participants.

Trial registration no. Trial Name, Study place	Study ID	Menopausal status of the participants	Study arms	N	Age (yr, mean ± SD)	HSDD duration (m, mean ± SD)	Primary efficacy outcomes	Follow-up duration
NCT00360529, VIOLET Study, United States and Canada	DeRogatis 2012[18]	Premenopausal	Placebo	295	35.5 ± 7.0	59.7 ± 46.0	SSE and eDiary sexual desire score	24 weeks
			Flibanserin 50 mg q.h.s.	295	36.3 ± 7.5	54.5 ± 40.1
			Flibanserin 100 mg q.h.s.	290	35.6 ± 7.2	55.8 ± 43.4
NCT00277914, ROSE Study, United States and Canada	Goldfischer 2011[19]	Premenopausal	Placebo	170	38.4 ± 6.7	63.0 ± 47.0	SSE and eDiary sexual desire score	24 weeks (after a 24 week open-label period)
			Flibanserin 50–100 mg q.h.s.	163	37.3 ± 7.0	58.0 ± 49.0
NCT00996164, BEGONIA Trial, United States	Katz 2013[20]	Premenopausal	Placebo	547	36.6 ± 7.8	49.5 ± 44.7	FSFI-d score and SSE	24 weeks
			Flibanserin 100 mg q.h.s.	543	36.5 ± 8.0	49.2 ± 40.3
NCT00360243, United States	NCT00360243[21]	Premenopausal	Placebo	349	NA	NA	SSE and eDiary sexual desire score	24 weeks
			Flibanserin 25 mg b.i.d.	337	NA	NA
			Flibanserin 50 mg q.h.s.	363	NA	NA
			Flibanserin 50 mg b.i.d.	336	NA	NA
NCT00491829, Multiple countries in Europe	NCT00491829[22]	Premenopausal	Placebo	318	NA	NA	SSE	24 weeks
			Flibanserin 50 mg q.h.s.	311	NA	NA
			Flibanserin 100 mg q.h.s.	316	NA	NA
NCT01057901, PLUMERIA Study, United States and Canada	Portman 2017[23]	Postmenopausal	Placebo	369	58.8 ± 5.3	56.8 ± 50.0	SSE and FSFI-d score	24 weeks (discontinued by the sponsor, 45.3% completed week 16)
			Flibanserin 100 mg q.h.s.	376	56.4 ± 5.4	62.2 ± 58.5
NCT00996372, SNOWDROP Trial, United States	Simon 2014[24]	Postmenopausal	Placebo	480	55.5 ± 5.3	61.6 ± 51.3	SSE and FSFI-d score	24 weeks
			Flibanserin 100 mg q.h.s.	467	55.4 ± 5.4	59.5 ± 46.0
NCT00360555, DAISY Study, Study, United States and Canada	Thorp 2012[25]	Premenopausal	Placebo	398	36.2 ± 6.6	64.9 ± 49.6	SSE and eDiary sexual desire score	24 weeks
			Flibanserin 25 mg b.i.d.	396	34.9 ± 7.0	59.2 ± 44.9
			Flibanserin 50 mg b.i.d.	392	35.7 ± 6.9	63.9 ± 44.2
			Flibanserin 100 mg q.h.s.	395	35.4 ± 6.9	60.0 ± 44.7

b.i.d. = Twice a day; q.h.s. = Every night at bedtime.

eDiary = electronic diary, FSFI-d = Female Sexual Function Index Desire Domain, HSDD = Hypoactive Sexual Desire Disorder, SSE = satisfying sexual events.

Table 2.

The basic characteristics of the excluded published articles and participants.

Study ID	Study duration	Reason of exclusion	Study arms	N	Main results
Jayne 2012[26]	52 weeks	- Open-label extension study of previous 5 clinical trials. - Had no placebo arm.	Flibanserin 50–100 mg/d	1723	- Somnolence, sedation, fatigue, dizziness, nausea, and vomiting were reported by 15.8%, 1.6%, 7.6%, 6.9%, 6.3%, and 1.4% of participants, respectively. - At study end, 42% of baseline non-remitters had improved their FSFI score to remission level.

FSFI = Female Sexual Function Index.

3.3. Risk of bias in the included studies

The bias risk across the 8 studies included in the meta-analysis is depicted in Figure S1, Supplemental Digital Content, http://links.lww.com/MD/M936. All 8 studies (100%) exhibited low risks of performance bias, detection bias, and reporting bias; all (100%) had a high risk of other biases. Six studies (75%) exhibited low risks of selection bias, and 2 (25%) exhibited unclear risks. Regarding attrition bias, 5 studies (62.5%) showed low risks, and 3 (37.5%) exhibited unclear risks. Publication bias was assessed through funnel plots given in Figure S2, Supplemental Digital Content, http://links.lww.com/MD/M937.

3.4. Grading of the results

The grades for the certainty of the evidence of the key outcomes of this meta-analysis are given in the SoF table (Table S1, Supplemental Digital Content, http://links.lww.com/MD/M938).

3.5. Effect of flibanserin on the primary efficacy outcomes

3.5.1. Satisfying sexual events

In premenopausal women, flibanserin was superior to placebo in increasing the number of SSE at both 100 mg (MD 0.69, 95% CI [0.39, 0.99], P < .00001, I² = 27% [mild heterogeneity], moderate certainty of evidence) (Fig. 2A) and 50 mg (MD 0.32, 95% CI [0.04, 0.59], P = .02, I² = 0% [not important heterogeneity]) (Fig. 2B) doses. In postmenopausal women, the superiority of flibanserin 100 mg over placebo in increasing the number of SSE was also observed (MD 0.37, 95% CI [0.14, 0.61], P = .002, I² = 0% [not important heterogeneity], high certainty of evidence) (Fig. 2C). Sensitivity analysis performed by omitting each trial to search for the sources of heterogeneity of flibanserin 100 mg on SSE in premenopausal women showed that none of the trials could substantially influence the pooled effect.

Figure 2.

Forest plot highlighting the change from baseline in (A) SSE in premenopausal women, Flibanserin 100 mg vs Placebo; (B) SSE in premenopausal women, Flibanserin 50 mg vs Placebo; (C) SSE in postmenopausal women, Flibanserin 100 mg vs Placebo; (D) e-Diary sexual desire score in premenopausal women, Flibanserin 100 mg vs Placebo; (E) e-Diary sexual desire score in premenopausal women, Flibanserin 50 mg vs Placebo.

3.5.2. eDiary sexual desire score

Flibanserin 100 mg outperformed placebo in improving sexual desire score measured using eDiary (MD 1.71, 95% CI [0.43, 2.98], P = .009, I² = 0% [not important heterogeneity], moderate certainty of evidence) (Fig. 2D) in premenopausal women. Such benefit was not observed with flibanserin 50 mg over placebo (MD 0.67, 95% CI [−0.71, 2.05], P = .34, I² = 0% [not important heterogeneity]) (Fig. 2E) in the same population.

3.6. Effect of flibanserin on secondary efficacy outcomes

3.6.1. FSFI desire domain and total scores

Compared to placebo, greater increments in FSFI desire domain scores were observed in premenopausal women with flibanserin 100 mg (MD 0.30, 95% CI [0.29, 0.31], P < .00001, I² = 0% [not important heterogeneity]) (Fig. 3A), and flibanserin 50 mg (MD 0.25, 95% CI [0.06, 0.45], P = .01, I² = 0% [not important heterogeneity]) (Fig. 3B). A similar benefit was observed with flibanserin 100 mg in postmenopausal women (MD 0.25, 95% CI [0.06, 0.44], P = .01, I² = 0% [not important heterogeneity]) (Fig. 3C).

Figure 3.

Forest plot highlighting the change from baseline in (A) FSFI-d score in premenopausal women, Flibanserin 100 mg vs Placebo; (B) FSFI-d score in premenopausal women, Flibanserin 50 mg vs Placebo; (C) FSFI-d score in postmenopausal women, Flibanserin 100 mg vs Placebo; (D) FSFI total score in premenopausal women, Flibanserin 100 mg vs Placebo; (E) FSFI total score in premenopausal women, Flibanserin 50 mg vs Placebo; (F) FSFI total score in postmenopausal women, Flibanserin 100 mg vs Placebo.

Flibanserin 100 mg was better than placebo in improving FSFI total scores in both premenopausal (MD 2.51, 95% CI [1.47, 3.55], P < .00001, I² = 90% [high heterogeneity]) (Fig. 3D), and postmenopausal women (MD 1.42, 95% CI [0.56, 2.29], P = .001, I² = 0% [not important heterogeneity]) (Fig. 3F). Such a beneficial effect was not observed with flibanserin 50 mg dose in premenopausal women (MD 0.74, 95% CI [−0.52, 1.99], P = .25, I² = 80% [high heterogeneity]) (Fig. 3E).

3.6.2. FSDS-R item 13 and total scores

Flibanserin 100 mg also outperformed placebo in reducing FSDS-R Item 13 scores in both premenopausal (MD −0.30, 95% CI [−0.31, −0.29], P < .00001, I² = 0% [not important heterogeneity]) (Fig. 4A), and postmenopausal women (MD −0.20, 95% CI [−0.39, −0.01], P = .04, I² = 0% [not important heterogeneity]) (Fig. 4C). The efficacy of flibanserin 50 mg in reducing FSDS-R Item 13 scores in premenopausal women was similar to placebo (MD −0.10, 95% CI [−0.29, 0.09], P = .31, I² = 0% [not important heterogeneity]) (Fig. 4B).

Figure 4.

Forest plot highlighting the change from baseline in (A) FSDS-R Item 13 score in premenopausal women, Flibanserin 100 mg vs Placebo; (B) FSDS-R Item 13 score in premenopausal women, Flibanserin 50 mg vs Placebo; (C) FSDS-R Item 13 score in postmenopausal women, Flibanserin 100 mg vs Placebo; (D) FSDS-R total score in premenopausal women, Flibanserin 100 mg vs Placebo; (E) FSDS-R total score in premenopausal women, Flibanserin 50 mg vs Placebo; (F) FSDS-R total score in postmenopausal women, Flibanserin 100 mg vs Placebo.

In premenopausal women, flibanserin was superior to placebo in reducing FSDS-R total scores at both 100 mg (MD −3.30, 95% CI [−3.37, −3.23], P < .00001, I² = 0% [not important heterogeneity]) (Fig. 4D) and 50 mg (MD −1.40, 95% CI [−2.53, −0.27], P = .02, I² = 0% [not important heterogeneity]) (Fig. 4E) doses. A similar benefit of flibanserin 100 mg over placebo was also observed in postmenopausal women (MD −2.42, 95% CI [−3.68, −1.16], P = .0002, I² = 0% [not important heterogeneity]) (Fig. 4F).

3.6.3. Patient’s global impression of improvement

Postmenopausal women treated with flibanserin 100 mg showed a similar change in PGI-I score compared to those treated with placebo (MD −0.15, 95% CI [−0.44, 0.14], P = .32, I² = 56% [moderate heterogeneity]) (Fig. 5A). In premenopausal women, a higher number of subjects achieved improvements in the PGI-I score in the flibanserin 100 mg arm (OR 1.93, 95% CI [1.58, 2.36], P < .00001, I² = 32% [mild heterogeneity]) (Fig. 5B), and flibanserin 50 mg arm (OR 1.39, 95% CI [1.11, 1.74], P = .004, I² = 0% [not important heterogeneity]) (Fig. 5C) than in the placebo arm.

Figure 5.

Forest plot highlighting the (A) Change from baseline in PGI-I score in postmenopausal women, Flibanserin 100 mg vs Placebo; (B) Proportions of premenopausal women with improvement in PGI-I score, Flibanserin 100 mg vs Placebo; (C) Proportions of premenopausal women with improvement in PGI-I score, Flibanserin 50 mg vs Placebo; (D) Proportions of premenopausal women responded positively at PBE, Flibanserin 100 mg vs Placebo; (E) Proportions of premenopausal women responded positively at PBE, Flibanserin 50 mg vs Placebo; (F) Proportions of postmenopausal women responded positively at PBE, Flibanserin 100 mg vs Placebo. PBE = Patient Benefit Evaluation.

3.6.4. Patient benefit evaluation

The percentage of premenopausal women who responded positively (“yes”) when asked whether they had experienced a meaningful benefit from the study medication was higher in the flibanserin 100 mg arm (OR 1.76, 95% CI [1.34, 2.31], P < .0001, I² = 59% [moderate heterogeneity]) (Fig. 5D), and flibanserin 50 mg arm (OR 1.41, 95% CI [1.12, 1.78], P = .004, I² = 0% [not important heterogeneity]) (Fig. 5E) than in the placebo arm. For the same outcome, flibanserin 100 mg was better than placebo in postmenopausal women (OR 1.54, 95% CI [1.24, 1.90], P < .0001, I² = 0% [not important heterogeneity]) (Fig. 5F).

3.7. Safety outcomes

The safety outcomes of flibanserin versus placebo are summarized in Table 3. The number of study subjects experiencing any AEs was higher in the flibanserin 100 mg arm (RR 1.37, 95% CI [1.14, 1.65], P = .009, I² = 93% [high heterogeneity], very low certainty of evidence), and flibanserin 50 mg arm (RR 1.18, 95% CI [1.03, 1.36], P = .02, I² = 72% [moderate heterogeneity]) (Fig. 5E) than in the placebo arm. More study subjects on flibanserin 100 mg experienced drug-related AEs (RR 2.33, 95% CI [1.96, 2.76], P < .00001, I² = 0% [not important heterogeneity]) and AEs leading to study discontinuation (RR 2.08, 95% CI [1.39, 3.10], P = .0003, I² = 50% [moderate heterogeneity]) than those on placebo. Serious AEs and severe AEs were comparable between flibanserin and placebo groups.

Table 3.

The results of safety outcomes of flibanserin vs placebo in the meta-analysis.

Safety variables	Flibanserin 100 mg vs Placebo				Flibanserin 50 mg vs Placebo
	No. of RCTs (participants)	Pooled effect size, RR [95% CI]	I2 (%)	P	No. of RCTs (Participants)	Pooled effect size, RR [95% CI]	I 2 (%)	P
AEs – All	8 (5809)	1.37 [1.14, 1.65]	93	.009	4 (2725)	1.18 [1.03, 1.36]	72	.02
Drug-related AEs	2 (2034)	2.33 [1.96, 2.76]	0	<.00001	–	–	–	–
Serious AEs	7 (5476)	1.23 [0.77, 1.95]	0	.38	4 (2725)	1.46 [0.89, 2.41]	0	.14
Severe AEs	4 (3112)	1.36 [0.87, 2.12]	34	.18	–	–	–	–
AEs leading to study discontinuation	5 (3697)	2.08 [1.39, 3.10]	50	.0003	–	–	–	–
Dizziness	8 (5809)	3.95 [2.40, 6.51]	70	<.00001	4 (2725)	2.39 [1.61, 3.55]	0	<.0001
Headache	7 (4722)	1.12 [0.92, 1.36]	0	.25	4 (2725)	0.97 [0.76, 1.23]	0	.79
Fatigue	6 (4117)	1.91 [1.30, 2.81]	53	.01	4 (2725)	1.08 [0.80, 1.46]	0	.61
Nausea	8 (5809)	2.37 [1.69, 3.33]	54	<.00001	4 (2725)	1.50 [1.05, 2.16]	18	.03
Somnolence	7 (5476)	4.04 [3.15, 5.18]	0	<.00001	4 (2725)	2.08 [1.30, 3.35]	34	.002
Insomnia	3 (1663)	1.69 [1.09, 2.62]	0	.02	–	–	–	–

AE = adverse event, CI = confidence interval, RCT = randomized controlled trials, RR = risk ratio.

More subjects taking flibanserin 100 mg (RR 3.95, 95% CI [2.40, 6.51], P < .00001, I² = 70% [moderate heterogeneity], low certainty of evidence) and flibanserin 50 mg (RR 2.39, 95% CI [1.61, 3.55], P < .0001, I² = 0% [not important heterogeneity]) experienced dizziness than those who took a placebo. Compared to the placebo group, a similar number of subjects in the flibanserin 100 mg and flibanserin 50 mg groups experienced headache. Fatigue (RR 1.91, 95% CI [1.30, 2.81], P = .01, I² = 53% [moderate heterogeneity], low certainty of evidence), nausea (RR 2.37, 95% CI [1.69, 3.33], P < .000001, I² = 54% [moderate heterogeneity], moderate certainty of evidence), somnolence (RR 4.04, 95% CI [3.15, 5.18], P < .000001, I² = 0% [not important heterogeneity], moderate certainty of evidence), and insomnia (RR 1.69, 95% CI [1.09, 2.62], P = .02, I² = 0% [not important heterogeneity], moderate certainty of evidence) were more frequent in the flibanserin 100 mg group than in the placebo group. Subjects in the flibanserin 50 mg group also exhibited a higher risk of somnolence than the placebo group (RR 2.08, 95% CI [1.30, 3.35], P = .002, I² = 34% [mild heterogeneity]).

4. Discussion

The present meta-analysis incorporated the results of RCTs of flibanserin in HSDD that are available to date. It highlights the efficacy and adverse events of flibanserin compared to a placebo used to manage HSDD in pre-and postmenopausal women. Flibanserin is more effective than placebo in treating HSDD, although its use is associated with higher risks of AEs.

US FDA approved flibanserin in August 2015 as the first medication to treat acquired, generalized HSDD in premenopausal women.^[27] Flibanserin is a non-hormonal oral medicine that acts on the brain to increase sexual desire. The mechanism by which flibanserin improves sexual desire and related distress is not known; this may be attributed to its high affinity for 5-HT1A and 5-HT2A receptors, displaying agonist activity on 5-HT1A and antagonist on 5-HT2A, resulting in lowering of serotonin in the brain as well as an effect on increasing norepinephrine and dopamine neurotransmitters. Serotonin exerts an inhibitory influence on sexual activity by reducing dopamine release in the prefrontal cerebral cortex. Both dopamine and norepinephrine are essential components in the process of sexual excitation. Norepinephrine serves as a stimulant in sexual arousal, while dopamine contributes to the enhancement of desire. By modulating these monoamines in patients with HSDD, flibanserin improves sexual desire and arousal.^[28]

In premenopausal women, flibanserin use was associated with 0.69 and 0.32 additional SSE per month at 100 and 50 mg daily doses; the number was 0.37 in postmenopausal women with flibanserin 100 mg. The improvement is not as robust as expected. US FDA report on the approval of flibanserin 100 mg mentioned that, on average, treatment with flibanserin 100 mg increased the number of satisfying sexual events by 0.5 to one additional event per month over placebo.^[27] The result of the present meta-analysis is within the range mentioned by the FDA for premenopausal women. The efficacy of 50 mg of flibanserin is far less in the same population. The drug has minimal efficacy in postmenopausal women and is not yet approved for treating HSDD in them. In addition, a mean of 1.71 eDiary sexual desire score was increased among the premenopausal women treated with flibanserin 100 mg; flibanserin at 50 mg daily doses was proven ineffective in this regard. Moreover, the FSFI desire domain score, which is extensively used as a measure of sexual desire in women, showed the superiority of flibanserin over placebo in both premenopausal (mean increase 0.30 for 100 mg and 0.25 for 50 mg) and postmenopausal women (0.25 for 100 mg). Similar trends were observed for the FSFI total score using flibanserin 100 mg in premenopausal and postmenopausal women. Flibanserin 100 mg also decreased the FSDS Item 13 score, the distress score related to sexual desire, by 0.3 premenopausal women and 0.2 postmenopausal women over placebo. Flibanserin (100 and 50 mg) also positively impacted FSDS-R total score reduction. The successful treatment of HSDD is directly associated with patient-reported outcomes (PGI-I and PBE) that capture women’s perspectives on improving their disease. Significantly more premenopausal women receiving flibanserin (both 100 mg and 50 mg) reported that their condition had improved (PGI-I). Similarly, more premenopausal and postmenopausal women reported a meaningful benefit from study medication (PBE) than those who received a placebo, suggesting that flibanserin provides a meaningful benefit to women with HSDD. The clinical efficacy of flibanserin 100 mg in improving FSFI-d and total scores, FSDS-R item 13 and total scores, and positive responders during PBE was better in premenopausal women than in postmenopausal women.

Flibanserin use was associated with higher risks of AEs and drug-related AEs, although these were of mild to moderate intensity. Serious AEs and severe AEs were equally low in flibanserin and placebo groups, although more study subjects taking flibanserin discontinued the study due to AEs. The risks of somnolence (RR 4.04) and dizziness (RR 3.95) were highest among the AEs with 100 and 50 mg doses. Risks for fatigue, nausea, and insomnia were increased by flibanserin 100 mg but not flibanserin 50 mg. In an open-label extension study after RCTs of flibanserin, 1.2% of participants reported serious AEs over 52 weeks of follow-up, and 10.7% of participants discontinued treatment due to AEs. Additionally, 15.8%, 1.6%, 7.6%, 6.9%, 6.3%, and 1.4% of participants reported somnolence, sedation, fatigue, dizziness, nausea, and vomiting, respectively.^[26] Moreover, severe AEs included hypotension and syncope, which are amplified with concurrent alcohol intake.^[28] Appropriate patient selection, avoidance of concomitant alcohol and other drugs that might increase the risks of flibanserin-associated AEs, and taking the drug at bedtime may reduce the AEs.

The advisory committee of the US FDA acknowledged the small treatment effects and substantial safety concerns but considered the unmet medical need while recommending FDA approval of flibanserin. Alongside, issues related to drug-drug interactions, consideration of nonpharmacologic approaches to HSDD with less importance, concerns about “medicalizing” low sexual desire, and the change from HSDD to FSIAD are drawing attention to the regulatory authorities for continuing the approval status of the drug.^[29]

4.1. Strengths and limitations

The main strength of this meta-analysis is the inclusion of a large population from a fairly good number of studies. The general quality of the included trials was good; all were phase 3 RCTs and double-blind trials. The certainty of the evidence generated is moderate to high for primary outcomes. We have separately analyzed the efficacy outcomes comparing flibanserin 100 and 50 mg with the placebo in pre- and postmenopausal women. Moreover, sub-analyses of safety outcomes compared flibanserin 100 and 50 mg with the placebo; such reports are missing in the previous meta-analyses. Furthermore, the meta-analysis included results of all RCTs available to date. There are also several limitations. Only 2 studies were conducted among postmenopausal women. In one study, there was a 24-week open-label period before the double-blind period, which might have reduced drug response.^[19] Flibanserin 25 mg b.i.d. in one study^[25] and 50 mg b.i.d. in another study^[21] were included in the flibanserin 50 and 100 mg arms, respectively. The b.i.d. dosing might have different efficacy and adverse effect potentials than the once-daily nighttime dosing. The certainty of the evidence generated was very low for any AEs and low for dizziness and fatigue.

5. Conclusion

In this meta-analysis, flibanserin showed modest benefits in treating HSDD, with the cost of higher AE risks, although the AEs were mild in nature. Premenopausal women with HSDD were more benefited than premenopausal women with the condition. The higher dose (flibanserin 100 mg) was more effective in treating HSDD but imparted higher risks of AEs. The continued efficacy and safety of flibanserin for long-term use still remain to be established.

Author contributions

Conceptualization: A.B.M. Kamrul-Hasan.

Data curation: A.B.M. Kamrul-Hasan, Mohammad Abdul Hannan, Fatema Tuz Zahura Aalpona, Lakshmi Nagendra, Shahjada Selim, Deep Dutta.

Formal analysis: A.B.M. Kamrul-Hasan, Lakshmi Nagendra, Deep Dutta.

Investigation: A.B.M. Kamrul-Hasan, Mohammad Abdul Hannan, Muhammad Shah Alam, Fatema Tuz Zahura Aalpona, Lakshmi Nagendra.

Methodology: A.B.M. Kamrul-Hasan, Muhammad Shah Alam, Lakshmi Nagendra, Shahjada Selim, Deep Dutta.

Project administration: A.B.M. Kamrul-Hasan, Shahjada Selim.

Resources: A.B.M. Kamrul-Hasan, Mohammad Abdul Hannan, Muhammad Shah Alam, Fatema Tuz Zahura Aalpona, Lakshmi Nagendra, Shahjada Selim, Deep Dutta.

Software: A.B.M. Kamrul-Hasan, Fatema Tuz Zahura Aalpona, Lakshmi Nagendra, Deep Dutta.

Supervision: A.B.M. Kamrul-Hasan, Deep Dutta.

Validation: A.B.M. Kamrul-Hasan, Deep Dutta.

Visualization: Mohammad Abdul Hannan, Muhammad Shah Alam, Fatema Tuz Zahura Aalpona, Shahjada Selim.

Writing – original draft: A.B.M. Kamrul-Hasan.

Writing – review & editing: A.B.M. Kamrul-Hasan, Mohammad Abdul Hannan, Muhammad Shah Alam, Fatema Tuz Zahura Aalpona, Lakshmi Nagendra, Shahjada Selim, Deep Dutta.

Supplementary Material