CADTH Reimbursement Review

Ospemifene (Osphena)

Sponsor: Duchesnay Inc.

Therapeutic area: Dyspareunia, vaginal dryness

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Stakeholder Input

Clinical Review

Executive Summary

An overview of the submission details for the drug under review is provided in Table 1.

Introduction

Hormonal changes, particularly a decrease in estrogen, are often associated with signs and symptoms in post-menopausal women that have an impact on health-related quality of life (HRQoL) as well as physical, mental, and sexual health.¹ Many post-menopausal women experience vulvar and vaginal atrophy (also called vulvovaginal atrophy) (VVA) due to physiologic changes in the female genital anatomy that result from aging and a lack of ovarian estrogen production in menopause. Women with VVA will typically present with vaginal dryness, pruritis, burning, pain, or dyspareunia as self-reported symptoms. Genitourinary syndrome of menopause is a newer, broader term that encompasses VVA as well as other genitourinary symptoms and may not be limited to patients who are sexually active.² Symptoms of genitourinary syndrome of menopause can be grouped as genital symptoms, sexual symptoms, and urinary symptoms.¹ While no report of incidence of VVA among Canadians is available, a study that included 1,016 Canadians reported a prevalence of 34%³; however, the prevalence of VVA is likely significantly underreported because many patients assume that the symptoms experienced during menopause are normal changes associated with aging. Previous literature suggests that 60% to 90% of post-menopausal patients may suffer from VVA and experience significant deficits in their quality of life because of it.⁴

The Society of Obstetricians and Gynaecologists of Canada (SOGC) recommends vaginal lubricants or moisturizers as first-line management options for genitourinary syndrome of menopause, particularly if the patient’s concerns are limited to vaginal dryness or dyspareunia.² Treatment for VVA typically includes estrogen hormonal therapies administered vaginally as creams, tablets, capsules, or a ring. The clinical expert consulted for this review indicated that the majority of women do not get adequate relief from vulvovaginal symptoms from systemic estrogen alone; local vaginal estrogen is still required.

The drug under review by CADTH is ospemifene 60 mg tablets for oral administration.⁵ Ospemifene is a selective estrogen receptor modulator (SERM) that binds to estrogen receptors, eliciting both antagonistic and agonistic effects and increasing the cellular maturation and mucification of the vaginal epithelium.⁵ In Canada, ospemifene is indicated in post-menopausal women for the treatment of moderate to severe dyspareunia and/or vaginal dryness, which are symptoms of VVA, a component of genitourinary syndrome of menopause. The sponsor has requested that ospemifene be reimbursed according to the approved Health Canada indication. The objective of this CADTH Reimbursement Review is to perform a systematic review of the beneficial and harmful effects of ospemifene (60 mg) for the treatment of moderate to severe dyspareunia and/or vaginal dryness, which are symptoms of VVA in post-menopausal women.

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient groups that responded to CADTH’s call for patient input and from a clinical expert consulted by CADTH for the purpose of this review.

Patient Input

Input was received from 1 patient group, Women’s Health Coalition Alberta (WHC). WHC advocates, raises awareness, and provides education about the urogynecological and reproductive health of patients of all ages. WHC noted the overall lack of awareness and understanding of urogynecological health, the limited therapeutic options for peri- and post-menopausal conditions (e.g., post-menopausal VVA), and the potential inequity in accessing preferred treatments when they are not reimbursed by public drug plans. WHC emphasized that the clinical and psychological impacts caused by untreated menopausal conditions are often overlooked and dismissed. Further, it expressed the expectation that a suitable treatment option for patients would improve their health outcomes and potentially raise clinician awareness of the importance of treating menopausal conditions.

To provide additional background on lived experience, values, and preferences of patients with VVA, patient group websites were searched for information about original experiences of patients with VVA. Healthtalk.org is a non-profit organization that has collected hundreds of stories from patients with any health condition.⁶ Information from video interviews with 13 British patients about VVA was available through Healthtalk.org, and obtained, assessed, and synthesized by the CADTH review team. The interviewed patients reported vaginal dryness, decline in libido (contributing to a decline in sexual activity), and urinary problems as some of the common complications they experienced after entering menopause. Interviewed patients also described the importance of sex in a relationship and how decreased sexual activity attributed to VVA symptoms may add significant complications to a relationship over time. In the interviews, some patients indicated they were made aware of the lack of knowledge regarding the effects of hormone replacement therapies, and of the fact that treatment with such therapies may not prevent thinning of the vaginal wall. The thinning of vaginal tissue was reported as causing severe discomfort for many patients, resulting in vaginal tears and bleeding. Patients also described how the decline in estrogen they had experienced affected the pelvic floor, bladder, uterus, vagina, or bowel, sometimes leading to urinary and bowel problems. Patients also reported difficulties with incontinence. Negative impacts on quality of life were experienced by many patients.

Clinician Input

Input From Clinical Experts Consulted by CADTH

The following input was provided by a clinical specialist with expertise in the diagnosis and management of post-menopausal patients with VVA.

The clinical expert indicated that women with VVA will typically present with vaginal dryness, pruritis, burning, pain, or dyspareunia; the goal of treatment is to provide relief of these symptoms. In the experience of the clinical expert, in most cases, currently available treatments are effective in providing relief of VVA symptoms. Because most available treatments for the symptoms of VVA are administered intravaginally, ospemifene offers an alternative route of administration, as an orally administered tablet; however, the clinical expert also noted that some patients may prefer a local therapy over systemic therapy due to hesitancy around the use of hormonal treatments.

The clinical expert consulted for this review noted that despite the utility of vaginal moisturizers and lubricants, women with VVA will generally experience more effective symptomatic relief from local vaginal estrogen. The clinical expert noted that additional therapeutic options have recently become available for the treatment of genitourinary syndrome of menopause, such as intravaginal prasterone and orally administered ospemifene (a SERM). These treatments represent a departure from traditional estrogen-based management strategies and substantially widen the scope of options available to women with VVA. The clinical expert relayed that the main adverse effect of ospemifene is hot flashes, which may be a significant barrier to widespread use in women with post-menopausal symptoms. The clinical expert suggested that taking this and other factors into account, it is unlikely that ospemifene will become first-line therapy.

The clinical expert consulted for this review felt that the majority of patients with genitourinary syndrome of menopause are anticipated to benefit from a therapeutic drug with estrogen receptor agonist properties, such as ospemifene. Feedback from the expert indicated that the patients most in need of intervention are those with more severe symptoms, and ospemifene provides an additional option versus traditional vaginal estrogen therapy. Additionally, the clinical expert felt that the oral route of administration for ospemifene may be especially suited for women who are unable to self-administer vaginal medication, such as due to severe pain or mobility limitations.

The clinical expert stated that women who do not report symptoms would generally not be diagnosed with VVA in clinical practice, and that patients will generally self-identify based on their description of symptoms (clinical history). Alternatively, patients seen for urogynecological issues, such as vaginal prolapse or urinary incontinence, may have VVA identified by history and visual inspection. The expert noted that clinical history and visual inspection of the vulva during a physical exam would be the usual methods for identifying patients with VVA; however, given the rise in telemedicine, the expert anticipated that more diagnoses will be made based on clinical history alone, and that this may also be considered a reasonable approach.

The clinical expert did not identify a specific subgroup of patients who would be less suited for treatment with ospemifene beyond those with any contraindication to ospemifene.

The clinical expert consulted by CADTH indicated that in clinical practice, subjective patient-reported improvement in symptoms is the primary outcome used to determine whether a patient is responding to treatment. The expert noted that improvement of symptoms is most often correlated with visual inspection at examination, although subjective symptoms are clinically more meaningful than appearance. Moreover, any improvement in vulvovaginal symptoms would be considered a clinically meaningful response, according to the clinical expert, who noted that this may include decrease in the sensation of vaginal dryness, decreased vaginal burning/pain, decreased frequency of urinary tract infections or bladder urgency or irritation, and decreased dryness and pain during intercourse. Histologic examination is generally not performed or required, based on the experience of the clinical expert.

Based on feedback from the clinical expert, there is no strict schedule for when treatment response needs to be assessed. The expert suggested that it would be reasonable to assess response approximately 3 months to 6 months after initiating treatment, then at 6 months to 12 months, and yearly thereafter.

Regarding discontinuation of treatment, the clinical expert stated that a patient may discontinue treatment if they wish, though symptoms may return after some time. They noted that the assessment of the risks and benefits is subjective, given that the condition is ultimately a quality of life issue.

The clinical expert indicated that ospemifene would most likely be prescribed in an outpatient ambulatory clinic setting by a family physician or gynecologist, with patients self-administering the medication at home. Because genitourinary syndrome of menopause is very common, the clinical expert believed that a vast majority of clinicians with experience in the treatment of women’s health issues would be suitable prescribers for pharmacologic treatment.

Clinician Group Input

Input from clinician groups was not received for the review of ospemifene.

Drug Program Input

Input was obtained from the drug programs that participate in the CADTH reimbursement review process. The following were identified as key factors that could potentially affect the implementation of a CADTH recommendation for ospemifene:

• considerations for initiation of therapy

• considerations for continuation or renewal of therapy

• considerations for discontinuation of therapy

• considerations for prescribing of therapy

• generalizability of trial populations to the broader populations in the jurisdictions

• system and economic issues.

Clinical Evidence

Pivotal Studies and Protocol Selected Studies

Description of Studies

A total of 5 phase III, double-blind, placebo-controlled, randomized controlled trials (RCTs) that assessed ospemifene 60 mg were included in the systematic review: Study 310 (N = 544, excluding the ospemifene 30 mg treatment group), Study 821 (N = 919), Study 231 (N = 631), Study 718 (N = 426), and Study 310X (N = 118 patients who continued from Study 310). Studies 310, 821, and 231 were designed to assess the efficacy and safety of ospemifene 60 mg over 12 weeks; Study 718 was designed to assess the efficacy and long-term safety of ospemifene 60 mg over 52 weeks; and Study 310X was a 52-week, long-term safety extension (LTSE) of Study 310 that assessed only safety outcomes. Most of the trials were conducted from 2006 to 2009 (studies 310, 821, and 718); Study 231 was conducted between 2016 and 2017. The trials primarily recruited patients in the US. No patients were studied in Canada. All of the studies enrolled post-menopausal women between 40 years and 80 years of age who had 5% or fewer superficial cells in the Maturation Index (MI) of the vaginal smear and a vaginal pH level greater than 5.0. In addition, studies 310, 821, and 231 included patients who identified at least 1 moderate to severe symptom of VVA that was considered the most bothersome.

Studies 310, 821, 231, and 718 included the following as co-primary end points assessed at week 12: percentage of vaginal superficial and vaginal parabasal cells on a vaginal smear and vaginal pH. Studies 310, 821, and 231 also included severity of the most bothersome symptom (MBS) of VVA as a co-primary end point. Secondary end points assessed in the 12-week studies included urinary symptoms using the Urinary Distress Inventory – Short Form (UDI-6) and sexual function (studies 821 and 231) using the Female Sexual Function Index (FSFI). HRQoL, mental health-related outcomes, bone mineral density, and adherence were identified as outcomes of interest to this review, but were not assessed in any of the included studies. The majority of patients included in studies 310, 821, 231, and 718 were 55 years of age and older and White. The proportion of patients who had previous experience with hormonal treatment varied significantly between the studies, ranging from 3% to 61% of patients. Of the 544 patients in Study 310, 222 (41%) reported vaginal dryness as the MBS and 242 (44%) reported vaginal pain with sexual activity (dyspareunia) as the MBS. In Study 821, 314 (34%) of patients reported vaginal dryness as their MBS and 605 (66%) reported dyspareunia as their MBS. Study 231 required patients to have vaginal dryness as their MBS. Study 718 did not report assessments of MBS at baseline. Baseline characteristics for Study 310X were limited to demographic information.

Efficacy Results

The efficacy of ospemifene was presented by 4 of the 5 included studies (all except Study 310X). A summary of key efficacy results is provided in Table 2.

Change in the severity of symptoms of VVA following 12 weeks of treatment was measured using the VVA questionnaire and evaluated in studies 310, 821, and 231 as a co-primary end point. Further, a formal minimal important difference (MID) was not identified in the published literature; however, the clinical expert consulted by CADTH indicated that any reduction in symptom severity was considered clinically meaningful because that is a primary goal of treatment. Each of these studies evaluated the change in vaginal dryness in patients who identified it as the MBS of VVA. In Study 310, the mean changes in severity of vaginal dryness at week 12 were –1.26 (standard deviation [SD] = 1.03) and –0.84 (SD = 1.00) for the ospemifene and placebo treatment groups, respectively, indicating that patients randomized to ospemifene reported a greater reduction in symptom severity compared to patients randomized to placebo (P = 0.021). |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| All patients included in Study 231 reported moderate or severe vaginal dryness as the MBS of VVA at baseline. The mean changes in symptom severity from baseline to week 12 were –1.29 (SD = 1.01) for ospemifene and –0.91 (SD = 0.96) for placebo (P < 0.0001). In Study 821, change from baseline in vaginal dryness was assessed in the dryness stratum (patients indicating vaginal dryness as the MBS of VVA at baseline). In contrast to the results of Study 310 and 231, Study 821 did not demonstrate a statistically significant difference in the reduction of severity of vaginal dryness compared to placebo, based on mean differences of –1.3 (SD = 1.08) for ospemifene and –1.1 (SD = 1.02) for placebo (P = 0.080) at week 12. Given the failure to demonstrate an improvement based on vaginal dryness as the MBS of VVA, the efficacy of ospemifene could not be concluded in the dryness stratum.

Patients who identified dyspareunia as the MBS of VVA were also enrolled in Study 310 and Study 821. In Study 310, this included 142 patients of 544 patients (26%) from the overall intention-to-treat (ITT) population. This population informed the analysis of the change from baseline to week 12 in severity of dyspareunia as the MBS of VVA as part of the co-primary end point for severity of the MBS of VVA. In Study 310, the mean changes from baseline to week 12 in severity of dyspareunia were –1.2 (SD = 1.3) in the ospemifene treatment group and –0.9 (SD = 1.1) in the placebo treatment group, which corresponded to a greater reduction in the severity of dyspareunia with ospemifene compared to placebo (P = 0.023). In Study 821, the co-primary end points in the dyspareunia stratum were analyzed independently from those in the dryness stratum. In Study 821, the mean changes from baseline in severity of dyspareunia were –1.5 (SD = 1.08) in the ospemifene treatment group and –1.2 (SD = 1.12) in the placebo treatment group. Therefore, a greater reduction in the severity of dyspareunia with ospemifene compared to placebo was demonstrated (P = 0.0001).

Cytology measurements included the percentage of parabasal cells and the percentage of superficial cells from a vaginal smear. The changes from baseline to week 12 in the percentages of parabasal cells and superficial cells were co-primary end points in studies 310, 821, 231, and 718. These outcomes provide an objective assessment of the signs of VVA and are considered standard in clinical trials; however, the clinical expert consulted by CADTH indicated that they are not particularly relevant to clinicians, given that they are rarely assessed in clinical practice. A reduction in the percentage of parabasal cells and increase in the percentage of superficial cells correspond with an improvement in VVA. Studies 310, 821 (dryness stratum and dyspareunia stratum), 231, and 718 demonstrated differences in the change in the percentages of parabasal cells and superficial cells in favour of ospemifene compared to placebo. Because these outcomes are not typically used in clinical practice, the clinical expert was unable to quantify a clinically meaningful improvement. Further, a formal MID was not identified in published literature.

The changes from baseline to week 12 in the percentages of parabasal cells and superficial cells were reported as follows:

• In Study 310, the mean changes from baseline in the percentages of parabasal cells were –30.1 (SD = 37.93) and 3.98 (SD = 35.21) for the ospemifene and placebo treatment groups, respectively, in favour of ospemifene (P < 0.001). The mean changes in the percentages of superficial cells were 10.8 (SD = 15.66) for the ospemifene group and 2.18 (SD = 8.39) for the placebo group, in favour of ospemifene (P < 0.001).

• In Study 231, the least squares (LS) mean changes in the percentages of parabasal cells were –23.7 (standard error [SE] = 1.4) and –1.9 (SE = 1.4) for the ospemifene and placebo treatment groups, respectively, in favour of ospemifene; the treatment-group difference was –21.8 (95% confidence interval [CI], –25.7 to –18.0; P < 0.0001). The LS mean changes in the percentages of superficial cells were 7.8 (SE = 0.7) and 0.6 (SE = 0.7) for the ospemifene and placebo treatment groups, respectively, in favour of ospemifene; the treatment-group difference was 7.2 (95% CI, 5.2 to 9.1; P < 0.0001).

• In the dryness stratum of Study 821, the LS mean changes in the percentages of parabasal cells were –31.7 (SE = 2.11) for ospemifene and –3.9 (SE = 2.18) for placebo, respectively; the treatment-group difference was –27.8 (95% CI, –33.75 to –21.90; P < 0.0001). The change in the percentage of superficial cells was reported as a median (range) because the analysis of covariance (ANCOVA) assumptions were not met. The median changes at week 12 were 7.0 (range = –4 to 65) for ospemifene and 0.0 (range = –11 to 57) for placebo in favour of ospemifene (P < 0.0001).

• In the dyspareunia stratum of Study 821, the LS mean changes in the percentages of parabasal cells for ospemifene and placebo were –40.3 (SE = 1.56) and –0.4 (SE = 1.57), respectively; the treatment-group difference was –39.9 (95% CI, –44.15 to –35.63; P < 0.0001). The change in the percentage of superficial cells was reported as a median (range) because the ANCOVA assumptions were not met. The median changes at week 12 were 7.0 (range = –6 to 79) for ospemifene and 0.0 (range = –5 to 85) for placebo, in favour of ospemifene (P < 0.0001).

• A non-parametric method of analysis was used in Study 718 because the assumptions of ANCOVA were not met. The medians of the changes in the percentages of parabasal cells were –40 (95% distribution-free CI, –55.0 to –30.0) for ospemifene and 0 (95% distribution-free CI, 0.0 to 10.0) for placebo, in favour of ospemifene (P < 0.0001). The medians of the changes in the percentages of superficial cells were 5 (95% CI, 5.0 to 7.0) for ospemifene and 0 (0.0 to 0.0) for placebo, in favour of ospemifene (P < 0.0001).

Study 718 also assessed the percentages of parabasal and superficial cells at weeks 26 and 52 as secondary outcomes. The results of both outcome assessments were similar to the results at week 12; however, the study was not powered to detect a difference in secondary outcomes, and the assessments were not controlled for multiplicity.

Vaginal pH was assessed in studies 310, 821, 231, and 718 as the change from baseline to week 12. This was a co-primary end point in each of the 4 studies. As with cytology assessments, vaginal pH is often measured in clinical trials, but is not particularly relevant to clinicians because it is rarely assessed in clinical practice. The clinical expert consulted by CADTH was unable to quantify a clinically meaningful improvement vaginal pH, and a formal MID was not identified in published literature. However, vaginal pH greater than 5.0 is an indicator of vaginal atrophy; therefore, a reduction in pH is suggestive of an improvement in VVA.

• In Study 310, the mean changes from baseline to week 12 in vaginal pH were –1.0 (SD = 1.1) for ospemifene and –0.1 (SD = 08) for placebo, in favour of ospemifene (P < 0.001).

• In Study 231, the LS mean changes from baseline to week 12 were –1.01 (SE = 0.04) for ospemifene and –0.29 (SE = 0.04) for placebo, corresponding to a treatment-group difference of –0.72 (95% CI, –0.84 to –0.59; P < 0.0001) in favour of ospemifene.

• In Study 821, the LS mean changes from baseline to week 12 in vaginal pH were –0.95 (SE = 0.07) and –0.94 (SE = 0.05) for ospemifene in the dryness and dyspareunia strata, respectively. The LS mean changes from baseline in the placebo treatment groups were –0.25 (SE = 0.07) and –0.07 (SE = 0.05) in the dryness and dyspareunia strata, respectively. The difference in the change in vaginal pH was in favour of ospemifene for both strata (P < 0.0001).

• In Study 718, the mean changes from baseline to week 12 in vaginal pH were –1.21 (SD = 0.912) for ospemifene and –0.16 (SD = 0.945) for placebo, corresponding to a treatment-group difference of –0.97 (95% CI, –1.17 to –0.77; P < 0.0001) in favour of ospemifene. The analyses at week 26 and week 52 were based on observed cases, which yielded similar results to those reported at week 12.

Urinary symptoms were assessed as a secondary outcome using the UDI-6 in studies 310 and 821 by domain score and total score, and by the total score in Study 231. No change in urinary symptoms, as measured by the UDI-6, were observed in any of the analyses. Sexual function was assessed as a secondary outcome in studies 821 and 231 using the FSFI. The FSFI is commonly used in clinical trials and is a validated tool for the measurement of women’s overall sexual function. The clinical expert consulted by CADTH indicated that the domains of the FSFI are clinically relevant, but sexual function is typically evaluated informally in clinical practice. Overall, the results of the FSFI were inconsistent between studies or did not demonstrate an improvement in sexual function compared to placebo, with the exception of the pain domain. The treatment-group differences in the change from baseline to week 12 for the pain domain were 0.58 (95% CI, 0.327 to 0.838) in Study 821 (all patients) and 0.45 (95% CI, 0.11 to 0.80) in Study 231, suggesting an improvement in favour of ospemifene. This result is aligned with a reduction in severity of dyspareunia demonstrated in the trials.

Harms Results

A summary of key safety results is provided in Table 3. No deaths were reported in any of the included studies, and specific serious adverse events (SAEs) were infrequently reported. No SAEs were reported by patients who received ospemifene in Study 310; 1.5% of patients who received placebo reported at least 1 SAE. The proportions of patients reporting at least 1 SAE in studies 821 and 231 were similar between treatment groups (1.3% versus 1.5% in Study 821 and 1.6% versus 1.0% in Study 231 for ospemifene versus placebo). In Study 718, 4.9% of patients in the ospemifene group and 6.5% of patients in the placebo treatment group reported at least 1 SAE. During the 12-week treatment period of studies 310, 821, and 231, patients who received ospemifene reported adverse events (AEs) at a similar or slightly higher frequency than patients who received placebo (60%, 63%, and 35% of patients who received ospemifene versus 52%, 51%, and 33% of patients who received placebo in studies 310, 821, and 231, respectively). Similar results were observed during the 52-week treatment period of Study 718, although the frequency of AEs was higher overall than in the 12-week studies. In addition, AEs were reported more frequently by those who received ospemifene compared to placebo (64% versus 45%) during the 52-week treatment period of Study 310X (including 12 weeks in Study 310), although this is likely biased in favour of placebo due to the high rate of discontinuation from study in the placebo treatment group. Specific AEs were not reported in more than 9% of patients in the 12-week studies or 13% of patients in Study 718. The most commonly reported AE in each of the 4 studies was hot flashes. Hot flashes were consistently reported more frequently by patients who received ospemifene (i.e., by 6% to 8% of patients who received ospemifene and by 3% to 3% of patients who received placebo). Vaginal infections, vaginal discharge, and muscle spasms were also reported as AEs more frequently by those in the ospemifene group versus the placebo group. Overall, patients who withdrew from treatment due to an AE were similar between treatment groups in the 12-week trials (2% to 5% for ospemifene and 3% to 5% for placebo). In studies 718 and 310X, withdrawals due to AEs were more frequent in the ospemifene treatment groups (14% and 6%, respectively) than in the placebo groups (10% and 2%, respectively). The rates of specific AEs leading to discontinuation were infrequent; however, hot flashes were the only AEs that led to treatment discontinuation for at least 1 patient who received ospemifene in every study.

The following notable ha rms were included in the CADTH systematic review protocol: vaginal hemorrhage, abnormal genital bleeding, cervical dysplasia, breast mass, endometrial hyperplasia, uterine polyps, cardiovascular disorders (e.g., thromboembolic and hemorrhagic stroke, coronary heart disease), breast cancer, uterine cancer, deep vein thrombosis (DVT), and pulmonary embolism. In studies 310, 821, 231, and 718, a total of |||||||| patients in the ospemifene treatment groups and |||||||| patients in the placebo treatment groups reported vaginal hemorrhage. Uterine polyps were reported by 6 patients and 1 patient for ospemifene and placebo, respectively. Cervical dysplasia was reported in |||||||| patients for ospemifene and |||||||| patients for placebo, and breast mass was reported in 7 patients in both the ospemifene and placebo groups. Endometrial hyperplasia was reported in 1 patient who received ospemifene, and breast cancer was reported in 1 patient who received placebo. A total of 2 patients reported DVT, both of whom were in ospemifene treatment groups. No patients reported experiencing abnormal genital bleeding, uterine cancer, pulmonary embolism, or other cardiovascular disorders (e.g., thromboembolic and hemorrhagic stroke, coronary heart disease). In Study 310X, vaginal hemorrhage, |||||||||||||||||||||||| and breast mass were reported by 1 patient (each) in the ospemifene treatment group. No other notable harms were reported.

Table 3: Summary of Key Safety Results From Pivotal and Protocol Selected Studies

AE = adverse event; OSP = ospemifene; PBO = placebo; SAE = serious adverse event; WDAE = withdrawal due to adverse event.

^aNo patients reported abnormal genital bleeding, cardiovascular disorders (e.g., thromboembolic and hemorrhagic stroke, coronary heart disease), uterine cancer, or pulmonary embolism.

Source: Clinical study reports.^7-11

Critical Appraisal

Each study used a mix of objective clinical outcomes and subjective patient-reported outcomes. Objective outcomes included cytology assessments (percentage of parabasal and superficial cells) and vaginal pH based on clinical results obtained from a vaginal smear. While commonly used in clinical trials, the objective outcomes are not typically used in clinical practice, according to feedback from the clinical expert. Subjective outcomes were patient reported and included the VVA questionnaire to assess the symptoms of VVA, the UDI-6 to assess urinary symptoms, and the FSFI to assess sexual function. Although the clinical expert consulted by CADTH indicated that the self-reported outcomes are considered clinically relevant in practice to measure treatment response, published MIDs were not identified for these outcome measures in post-menopausal women. Therefore, it is unclear whether the reported between-group differences are clinically meaningful. Further, evidence of validity, reliability, and responsiveness of the VVA questionnaire was not identified for this review; nor was the validity of treating the ordinal data as continuous. These factors make it difficult to interpret the results. Additionally, secondary outcomes (UDI-6, FSFI, and any outcomes reported after week 12) were not controlled for multiplicity; therefore, they are subject to type I error.

In all studies, the primary efficacy analyses were performed using the ITT population, and supportive analyses were performed in the per-protocol (PP) and modified intention-to-treat (mITT) populations (Study 231 only). All of the supportive analyses performed were consistent with the primary analyses, with the exception of vaginal dryness as the MBS of VVA in Study 310, for which statistical significance was not demonstrated in the PP population. The sponsor attributed the lack of statistical significance for the supportive analysis to the small sample size, which is likely a contributing factor; however, the results of the analysis of vaginal dryness as the MBS of VVA in Study 310 remain uncertain.

In Study 718, patients with VVA were identified based on MI and vaginal pH without a requirement for self-reported symptoms of VVA. This introduces uncertainty about the generalizability of the patient population to post-menopausal patients with moderate to severe vaginal dryness or dyspareunia. Otherwise, the eligibility criteria used in the included studies were generally considered appropriate and reflective of post-menopausal patients with VVA, although restrictive (70% of patients in Study 231 failed screening; data were not reported in the other included studies). Most notably, patients with comorbidities, such as a history of cancer or cardiovascular disorders, were excluded from the trials, leading to uncertainty regarding the generalizability of the safety results. Lastly, evidence informing the efficacy of ospemifene is primarily based on patients receiving treatment for up to 12 weeks. Supportive efficacy data based on clinical outcomes were available for up to 52 weeks; however, the evidence is weak and not based on clinically relevant outcomes (symptom severity), causing uncertainty in the long term efficacy. Additionally, safety evidence in patients who received treatment for up to 52 weeks was available, but subject to high and imbalanced discontinuation rates. Moreover, patients are expected to continue treatment for more than 1 year, but there is no evidence of safety beyond this time point.

Indirect Comparisons

Description of Studies

The sponsor-submitted indirect treatment comparison (ITC)¹² was included in this review along with an additional ITC (Li et al.) identified in the literature search.¹³ Both ITCs conducted a systematic review and network meta-analysis (NMA) to evaluate the comparative efficacy and safety of ospemifene versus other alternative therapies in the treatment of VVA. Both ITCs used a Bayesian framework for NMA analysis.

In the sponsor-submitted ITC, 27 RCTs were eligible, 5 of which involved ospemifene. Other treatments investigated included a conjugated estrogens vaginal cream (Premarin), an estradiol vaginal insert (Vagifem), an estradiol soft gel vaginal insert (Imex), an estradiol vaginal ring (Estring), and a prasterone vaginal ovule (Intrarosa). The sample size of the included trials ranged from 21 patients to 826 patients, and the mean age ranged from 56 years to 63 years. The eligible RCTs recruited primarily post-menopausal women with moderate to severe genitourinary symptoms, and the majority of the trials were 12 weeks in duration (range = 12 weeks to 14 weeks). For the NMA, the sponsor only included RCTs with the following treatments: ospemifene 60 mg oral daily (Osphena), estradiol vaginal cream 0.02 mg (Estrace), estradiol transdermal patch 14 mcg (Estradiol patch), estradiol vaginal cream 2 mg and 7.5 mg (Estring), estriol vaginal pessary 0.5 mg (Estriol pessary), estradiol vaginal capsule 4 mcg and 10 mcg (Imvexxy), dehydroepiandrosterone (DHEA) vaginal suppository 6.5 mg (Intrarosa), lubricants, conjugated estrogens vaginal cream 0.3 mg or 0.63 mg (Premarin), promestriene vaginal cream 10 mg, or estradiol vaginal insert 10 mcg (Vagifem). The sponsor noted that the majority of trials were at low risk of bias; however, 4 RCTs were at high risk of bias from blinding.

In the Li et al. ITC, 29 RCTs were eligible, with 8,311 participants (sample sizes ranged from 180 patients to 909 patients). Five treatments were investigated: laser therapy, vaginal estrogen (vaginal estrogen therapies pooled together), ospemifene, vaginal DHEA, and moisturization and/or lubrication. The mean age of participants ranged from 51 years to 65 years, and the duration of the trials ranged from 6 weeks to 52 weeks. Neither the severity nor the duration of symptoms was described by the authors.

Efficacy Results

Sponsor-Submitted Indirect Treatment Comparison

For the outcome of mean difference (MD) in change from baseline to follow-up in MBS score for vaginal dryness, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

For the outcome of MD in change from baseline to follow-up in MBS score for dyspareunia, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

For the outcome of MD in change from baseline to follow-up for combined MBS score for vaginal dryness and dyspareunia, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

For the outcome of MD in change in percentage of parabasal cells, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

For the outcome of MD in change in percentage of superficial cells, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

For the outcome of MD in reduction of vaginal pH, ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||

Li et al. Indirect Treatment Comparison

In the Li et al. ITC, there was no difference between ospemifene and vaginal estrogens for the outcomes of mean difference in change in vaginal dryness (MD = –2.9; 95% credible interval [CrI], –13 to 8.1), dyspareunia (MD = 8.0; 95% CrI, 0.2 to 17), or sexual function (MD = 1.5; 95% CrI, –2.7 to 5.6). The reduction in vaginal pH was smaller for ospemifene versus vaginal estrogens (MD = 0.31; 95% CrI, 0.05 to 0.58). There was no difference in the reduction in percentage of parabasal cells for ospemifene compared with vaginal estrogens (MD = 2.2; 95% CrI, –9.5 to 15).

Harms Results

In the sponsor-submitted ITC, there was no difference in the risk of treatment-emergent adverse events (TEAEs) for ospemifene versus conjugated estrogens vaginal cream (relative risk [RR] = 1.07; 95% CrI, 0.93 to 1.24) or versus estradiol vaginal tablet (RR = 1.11; 95% CrI, 0.95 to 1.28). There was no difference in the risk of serious TEAEs for ospemifene versus conjugated estrogens vaginal cream (RR = 0.75; 95% CrI, 0.02 to 31) or versus estradiol vaginal tablet (RR = 0.87; 95% CrI, 0.15 to 4.17). There was no difference in the risk of urinary tract infection (UTI) between ospemifene and estradiol vaginal tablet (RR = 2.55; 95% CrI, 0.23 to 35). The risk of headaches was lower for ospemifene compared with estradiol vaginal ring (RR = 0.00; 95% CrI, 0.00 to 0.04), while there was no difference compared with conjugated estrogens vaginal cream (RR = 0.74; 95% CrI, 0.38 to 1.42) or estradiol vaginal tablet (RR = 1.43; 95% CrI, 0.24 to 8.50). There was no difference in the risk of discontinuation due to AEs for ospemifene versus estradiol vaginal ring (RR = 1.26; 95% CrI, 0.28 to 1.52), conjugated estrogens vaginal cream (RR = 0.97; 95% CrI, 0.31 to 2.69), or estradiol vaginal tablet (RR = 0.94; 95% CrI, 0.31 to 2.45).

Critical Appraisal

The sponsor-submitted ITC provided a clear rationale and objective, and was generally well conducted aside from the following limitations of note. Heterogeneity in effect sizes (based on I² > 50%) was observed for some comparisons; however, it was not explored further through meta-regression with suspected effect modifiers. The extent to which eligible studies satisfied the similarity assumption was unclear. While patient and study characteristics were broadly similar, the appropriateness of combined different doses in nodes, differences in placebo use across trials, and the unclear extent of prior VVA treatment make it challenging to assess the similarity of the eligible studies. Subgroup or sensitivity analyses did not result in different results than the base case and were generally not able to explain heterogeneity, although decisions regarding the methodology were not adequately described. In the analysis of safety outcomes, there were wide CrIs and low event rates (resulting in extremely low RRs) for some comparisons; these outcomes make it challenging to assess comparative safety (e.g., for headache or UTI). Further, for some efficacy outcomes, there were a limited number of trials for some nodes, resulting in wide and overlapping CrIs. This makes it difficult to draw conclusions around comparative efficacy for ospemifene and relevant comparators.

The Li et al. ITC described the study objective and study selection process. Concerns were identified with respect to study selection. Specific eligibility criteria were not provided (e.g., based on severity of symptoms), and the authors did not provide explicit criteria around specific relevant interventions or comparators. Information about disease severity and duration of symptoms was not extracted, making it challenging to assess whether the similarity assumption was satisfied. Further, since severity of symptoms was not provided, it is unclear how relevant the population was for the present review. The authors converted the continuous outcomes into a 0 to 100 scale because different outcome scales were used across studies; however, they did not provide details about how this was carried out or whether it was appropriate. Not all comparators in the Li et al. ITC were relevant to this review. The comparison of ospemifene to vaginal estrogens was relevant. However, the Li et al. ITC combined all vaginal estrogens into 1 node (including different drugs and dosage forms; for example, conjugated estrogens, estradiol 4 mcg or 10 mcg vaginal capsule). Some of the vaginal estrogens included in the vaginal estrogen node were not comparators of interest for this review (e.g., estriol cream). Given that there may be differences between different vaginal estrogen products in terms of efficacy and safety, the appropriateness of combining these treatments into 1 node is uncertain. It further makes it challenging to draw conclusions around the comparative efficacy and safety of ospemifene versus individual relevant treatments. A description of model fit was not provided; therefore, it is unclear if model fit was adequate. Network diagrams were not provided in the Li et al. ITC, and it was unclear how many studies contributed to specific comparisons. Heterogeneity (I² > 50%) was observed for some outcomes involving ospemifene (dyspareunia, vaginal pH, parabasal cells), which could be explained by age (vaginal pH) or dose (change in the percentage of parabasal cells), but could not be explained for other outcomes.

Other Relevant Evidence

Description of Studies

Study 312, a multi-centre, open-label, phase III LTSE of Study 310, has been summarized to provide additional evidence regarding the long-term safety of oral daily doses of ospemifene 60 mg for the treatment of VVA in post-menopausal women without a uterus. During this extension study, all patients received ospemifene 60 mg per day irrespective of their treatment assignment in the initial 12-week Study 310. The duration of treatment was 52 weeks followed by a 4-week post-treatment follow-up period, for a total of 68 weeks (including the initial 12 weeks of Study 310).^14,15 The baseline characteristics of those who continued into the LTSE were similar to those in the core study in terms of age, race, ethnicity, and body mass index (BMI).

Of the 826 post-menopausal women randomized to Study 310, 301 women (36.4%) enrolled in the open-label extension study, Study 312. Overall, 117 patients (38.7%) discontinued from the study. The most common reasons for discontinuation were patient decision or withdrawal of consent (13.2%), AEs (12.3%), and loss to follow-up (5.6%).

Efficacy Results

Efficacy was not assessed in Study 312.

Harms Results

During the 52-week treatment period, 73.1% of patients reported at least 1 TEAE, and 4% of patients reported at least 1 SAE. The most common AEs were sinusitis (8%), UTI (8.6%), and hot flashes (10.3%). None of the specific SAEs were reported in more than 2 patients. AEs leading to treatment discontinuation were reported in 34 patients (11.3%); hot flashes, reported by 2% of patients, were the most frequent AE leading to discontinuation. ||||||||||||||||||||||||||||||||||||||||||||||||||||||| breast mass (n = 1), and hemorrhagic stroke (n = 1) were the only notable harms reported in Study 312, all of which were infrequent.

Critical Appraisal

Study 312 had several limitations resulting from the overall design, including the lack of a comparison group to provide context and control for potential confounders. Additionally, the open-label design may influence the perception of improvement by patients and clinicians, which could affect the reporting of harms. Among the enrolled patients, 117 patients (38.7%) discontinued prematurely from the study, which may have resulted in safety outcomes being reported. Because the patients who took part in Study 312 were originally from the parent studies, and the eligibility criteria remained the same, it is reasonable to expect that the same limitations to generalizability are relevant to the open-label extension study. For instance, given that the participants were predominantly White (92.4%), the results from these trials may not be generalizable to other racial groups that may commonly be seen at some centres in Canada. Since this open-label extension safety study focuses on a very specific patient population (post-menopausal women with no uterus), it would be best to compare the safety results with similar studies to get a more accurate idea of the safety profile among the general population. The treatment duration was 52 weeks, which might not be a sufficient time frame over which to observe and note all potential safety issues.

Conclusions

Five studies were summarized as part of the CADTH systematic review: 4 phase III, double-blind RCTs that assessed the efficacy and safety of ospemifene 60 mg compared to placebo over 12 weeks (studies 310, 821, 231) and over 52 weeks (Study 718) in post-menopausal women with VVA, as well as a double-blind, placebo-controlled LTSE of Study 310 (Study 310X) that provided evidence of safety over up to 52 weeks of treatment. The studies included in this review demonstrated a beneficial effect of ospemifene compared to placebo over a 12-week treatment period in post-menopausal women who had self-reported moderate to severe vaginal dryness or dyspareunia VVA symptoms. Although the efficacy of ospemifene versus placebo in relieving vaginal dryness was demonstrated in Study 231, there was inconsistency in this finding across the studies due to lack of statistical significance for this outcome in the primary analysis of Study 821, as well as in a supportive analysis performed in the PP population of Study 310. Across the included studies, the clinical benefit was estimated using a patient-reported outcome, the VVA questionnaire. The exact clinical interpretation of the difference in score of the VVA questionnaire is unclear, particularly due to the lack of a recognized MID, a lack of sufficient validation of the questionnaire, and the nature of an ordinal score; however, the magnitude of observed change in symptom scores was similar between the different trial populations of similar eligibility criteria. The self-assessment of individual patients may suffer from recall bias in both the ospemifene and placebo groups. However, the observed clinical benefit was supported by objective measures of VVA, namely a reduction in the percentage of parabasal cells, an increase in the percentage of superficial cells, and a reduction in pH.

Evidence from 2 ITCs summarized for this review suggests there is no difference between ospemifene and other treatments for symptoms of VVA in terms of comparative effectiveness. This finding should be interpreted with caution due to uncertainty associated with the ITCs; however, it highlights the limitation of a lack of direct comparative evidence for active treatment options. Other gaps in the evidence include the absence of assessments of both HRQoL, and symptom relief beyond 12 weeks of treatment. Evidence assessing the safety of ospemifene was available for up to 52 weeks on treatment. No deaths were reported, and SAEs were reported infrequently. Overall, the safety profile of ospemifene was acceptable based on the included trials, with the exception of the frequency of hot flashes and uncertainty around the risk of thromboembolic events. Further study is warranted to obtain long-term safety data, including evidence of safety beyond 1 year.

Introduction

Disease Background

Hormonal changes, particularly a decrease in estrogen, are often associated with signs and symptoms in post-menopausal women that have an impact on HRQoL as well as physical, mental, and sexual health.¹ Many post-menopausal women experience VVA due to physiologic changes in the female genital anatomy that result from aging and a lack of ovarian estrogen production in menopause. The vulva and vagina are particularly susceptible to changes related to menopause because there are estrogen receptors in the vulva, vaginal, bladder, urethra, and muscles of the pelvic floor.² The clinical expert consulted by CADTH indicated that with menopause, the vulva loses much of its collagen and adipose tissue, and glandular secretions are diminished; the vaginal surface thins, loses its elasticity, and is more easily injured, with decreased fluid secretion; and changes in urethral and vaginal flora and pH can render menopausal women more susceptible to UTIs and vaginal infections. Women with VVA will typically present with vaginal dryness, pruritis, burning, pain, and dyspareunia as self-reported symptoms.

Genitourinary syndrome of menopause is a newer, broader term that encompasses VVA as well as other genitourinary symptoms and may not be limited to patients who are sexually active.² Symptoms of genitourinary syndrome of menopause can be grouped as: genital symptoms, including dryness, burning, itching, irritation, and bleeding; sexual symptoms, including dyspareunia and other sexual dysfunctions; and urinary symptoms, including urgency, dysuria, and recurrent UTIs.¹ Women may present with some or all of the signs and symptoms of genitourinary syndrome of menopause.¹⁶ Signs of genitourinary syndrome of menopause can be observed through physical examination conducted by an experienced health care provider, given that there may be changes in the colour, size, and integrity of the anatomy of the vagina. There may also be signs of decreased lubrication and an increase in vaginal pH; typically, a pH of greater than 5.0 would be considered abnormal.² The clinical expert consulted by CADTH indicated that assessments of vaginal pH, parabasal cells, and superficial cells are not typical in Canadian clinical practice, although these are often evaluated in clinical trials. In women with vaginal atrophy, an increase in parabasal cells and decrease in superficial cells can be observed. Moreover, as women age, the proportion of parabasal cells will continue to increase, and the MI may eventually consist entirely of parabasal cells.¹⁶ According to guidelines published by SOGC, in the absence of treatment, genitourinary syndrome of menopause will evolve chronically in most women and progress to functional and structural urogenital tissue changes that can be difficult to reverse even with treatment.²

Of note, genitourinary syndrome of menopause has been described as being defined by the presence of symptoms; however, not all women with signs of atrophy identified through pelvic examination are symptomatic.¹⁶ While no report of the incidence of VVA among Canadians is available, a study that included 1,016 Canadians reported a prevalence of 34%.³ However, estimates regarding the prevalence of patients who suffer from VVA or genitourinary syndrome of menopause may be underreported. Many patients will not report changes they experience during menopause because they will associate changes to normal aging. Previous literature suggests that 60% to 90% of post-menopausal patients may suffer from VVA and experience significant deficits in their quality of life because of it.⁴ Due to underreporting, it may be important for health care providers to take the initiative and ask post-menopausal patients about symptoms related to genitourinary syndrome of menopause to identify the condition as early as possible and provide optimal care.²

Standards of Therapy

An ideal treatment for genitourinary symptoms of menopause would provide complete symptomatic relief from the urogenital and/or vulvovaginal changes experienced by women in menopause, thereby improving patients’ quality of life, with minimal adverse effects and long-term health risks. SOGC recommends that patients try vaginal lubricants and/or vaginal moisturizers as first-line management options for genitourinary syndrome of menopause, particularly if patient concerns are limited to vaginal dryness or dyspareunia.² The clinical expert indicated that these treatments may include hyaluronic acid gel or polycarbophil gel. Second-line treatment for VVA typically includes the administration of local estrogen to reverse the effects of estrogen withdrawal, or non-hormonal treatments to counter the effects of VVA.² Several formulations exist for local vaginal estrogen, including creams, a hormone-releasing ring, and tablets. The clinical expert indicated that the majority of women who receive systemic estrogen for other menopausal symptoms do not get adequate relief from vulvovaginal symptoms from systemic estrogen alone; local vaginal estrogen is still required. The clinical expert noted that some alternative supplements, such as phytoestrogens, black cohosh, and dong quai, have been studied, but have not been found to be effective compared with placebo. They also noted that the benefit of vitamin D supplementation is unclear.

Estrogen treatment favourably alters patients’ physiology to treat the underlying disease and targets disease symptoms. The clinical expert noted that estrogen improves blood supply to vulvovaginal tissues, restoring normal vaginal flora and pH, improving symptoms, and reducing the risk of urogenital infections. However, Health Canada has issued a black box warning for vaginal estrogen therapies for several disease risks (myocardial infarction, stroke, invasive breast cancer, pulmonary emboli, and DVT, for most products) based on evidence for oral estrogen-plus-progestin therapy and oral estrogen-alone therapy. These therapies are also contraindicated in patients with known or suspected estrogen-dependent malignant neoplasia and patients with a known, suspected, or past history of breast cancer, also based on evidence for systemic therapies. According to the clinical expert, it is possible for some patients with these contraindications to be treated with vaginal estrogen; however, the product monograph warnings can lead to hesitancy. Estrogen hormonal therapies may be administered to patients vaginally as creams, tablets, capsules, or a ring. Some patients may prefer products other than vaginal creams, because they can be messy.

Drug

The drug under review by CADTH is ospemifene (60 mg) tablets for oral administration. It is recommended that ospemifene be administered consistently, once daily with food.⁵ In Canada, ospemifene is indicated in post-menopausal women for the treatment of moderate to severe dyspareunia and/or vaginal dryness, which are symptoms of VVA, a component of genitourinary syndrome of menopause. The sponsor has requested that ospemifene be reimbursed according to the approved Health Canada indication.

Ospemifene is a SERM that acts by binding to estrogen receptors. In some tissues, ospemifene acts as an agonist that activates estrogenic pathways; in other tissues, it acts as an antagonist by causing a blockade of estrogenic pathways.⁵ Ospemifene has an effect on estrogen receptors in the vagina, increasing the cellular maturation and mucification of the vaginal epithelium.

Stakeholder Perspectives

Patient Group Input

This section was prepared by CADTH staff based on the input provided by patient groups.

Input was received from 1 patient group, WHC. WHC advocates, raises awareness, and provides education about the urogynecological and reproductive health of patients of all ages. WHC noted the overall lack of awareness and understanding of urogynecological health, the limited therapeutic options for peri- and post-menopausal conditions (e.g., post-menopausal VVA), and the potential inequity in accessing preferred treatments when they are not reimbursed by public drug plans. WHC emphasized that the clinical and psychological impacts caused by untreated menopausal conditions are often overlooked and dismissed. Further, it expressed the expectation that a suitable treatment option for patients would improve their health outcomes and potentially raise clinician awareness of the importance of treating menopausal conditions.

To provide additional background on the lived experiences, values, and preferences of patients with VVA, patient group websites were sought. Healthtalk.org is a non-profit organization that has collected hundreds of stories from patients with any health condition.⁶ Information from video interviews with 13 British patients about VVA was available through Healthtalk.org and was obtained, assessed, and synthesized by the CADTH review team. The interviewed patients reported vaginal dryness, decline in libido (contributing to a decline in sexual activity), and urinary problems as some of the common complications they experienced after entering menopause. Interviewed patients also described the importance of sex in a relationship and how decreased sexual activity (attributed to VVA symptoms) can add significant complications to a relationship over time. In the interviews, some patients indicated they were made aware of the lack of knowledge regarding the effects of hormone replacement therapies, and that treatment with hormone replacement therapies may not prevent thinning of the vaginal wall. Thinning of vaginal tissue was reported as causing severe discomfort for many patients, resulting in vaginal tears and bleeding. Patients also described how the decline of estrogen they experienced had affected their pelvic floor, bladder, uterus, vagina, and bowel, sometimes leading to urinary and bowel problems. Patients also reported difficulties with incontinence. Negative impacts on quality of life were experienced by many patients.

Clinician Input

Input From the Clinical Expert Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise regarding the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol; assisting in the critical appraisal of clinical evidence; interpreting the clinical relevance of the results; and providing guidance on the potential place in therapy). The following input was provided by 1 clinical specialist with expertise in the diagnosis and management of post-menopausal patients with VVA.

Unmet Needs

According to the clinical expert consulted by CADTH, treatment for VVA typically includes the administration of local estrogen to reverse the effects of estrogen withdrawal or of non-hormonal treatments to counter the effects of VVA. The clinical expert indicated that women with VVA will typically present with vaginal dryness, pruritis, burning, pain, and/or dyspareunia; the goal of treatment is to provide relief of these symptoms. In the experience of the clinical expert, in most cases, currently available treatments are effective in providing relief of VVA symptoms. Because most available treatments for symptoms of VVA are administered intravaginally, ospemifene offers an alternative route of administration (as an orally administered tablet); however, the clinical expert also noted that some patients may prefer a local therapy over a systemic therapy due to hesitancy around the use of hormonal treatments.

Place in Therapy

The clinical expert consulted for this review noted that despite the utility of vaginal moisturizers and lubricants, women with VVA will generally experience more effective symptomatic relief from vaginal local estrogen. While estrogen can be delivered through oral, transdermal, or vaginal routes of administration, the vaginal route has traditionally been the most effective for vulvovaginal symptoms. The clinical expert noted that some studies have indicated that the majority of women receiving systemic estrogen (alone) for other menopausal symptoms do not get adequate relief of VVA symptoms; local vaginal estrogen is still required. They also indicated that several formulations exist for local vaginal estrogen, including creams, a hormone-releasing ring, and tablets. The clinical expert noted that more recently, additional therapeutic options have become available for the treatment of genitourinary syndrome of menopause, such as intravaginal prasterone and orally administered ospemifene (a SERM). These represent a departure from traditional estrogen-based management strategies and substantially widen the scope of options available to women with VVA.

The clinical expert consulted by CADTH relayed that ospemifene is a SERM that has specific estrogen receptor agonist activity on vaginal tissues as well as on the bones, with partial agonist activity on the uterus. The clinical expert indicated that by targeting the estrogen receptors, ospemifene provides symptomatic relief from VVA caused by the decline in estrogen levels in menopause. It was also noted that hot flashes represent the main adverse effect of ospemifene; the clinical expert believed that this could be a significant barrier to widespread use in women with menopausal symptoms. The clinical expert suggested that, considering this factor and others, it is unlikely that ospemifene will become first-line therapy.

Patient Population

The clinical expert consulted for this review believed that the majority of patients with genitourinary syndrome of menopause are anticipated to benefit from a therapeutic drug with estrogen receptor agonist properties, such as ospemifene. Feedback from the expert indicated that the patients most in need of intervention are those with more severe symptoms, and that ospemifene provides an additional option to traditional vaginal estrogen therapy. Additionally, the clinical expert believed that the oral route of administration for ospemifene may be especially suited for women who are unable to self-administer vaginal medication (e.g., due to severe pain or mobility limitations).

The clinical expert stated that women who do not report symptoms would generally not be diagnosed with VVA in clinical practice. The expert indicated that patients will generally self-identify based on their description of symptoms during clinical history. Alternatively, patients seen for urogynecological issues, such as vaginal prolapse or urinary incontinence, may have VVA identified at the time of assessment by history and visual inspection. The expert noted that clinical history and visual inspection of the vulva on physical exam would be the usual methods for identifying patients with VVA; however, given the rise in telemedicine, the expert anticipated that diagnoses will increasingly be made based on clinical history alone, which may also be considered a reasonable approach.

The clinical expert stated that patients with any contraindication to ospemifene would not be suitable for treatment with this medication. The clinical expert noted that these patients include those with a hypersensitivity or allergy to the medication or its components, undiagnosed vaginal bleeding, active venous or arterial thromboembolic disease or history of such disease, and those with a known or suspected estrogen-dependent tumour.

Assessing Response to Treatment

The clinical expert consulted by CADTH indicated that in clinical practice, subjective patient-reported improvement in symptoms is the primary outcome used to determine whether a patient is responding to treatment. The expert noted that improvement of symptoms is most often correlated with visual inspection at examination; however, subjective symptoms are more clinically meaningful than appearance. Moreover, any improvement in vulvovaginal symptoms would be considered a clinically meaningful response, according to the clinical expert, who noted that this may include decrease in sensation of vaginal dryness, decreased vaginal burning and/or pain, decreased frequency of UTIs or bladder urgency and/or irritation, and decreased dryness and pain during intercourse. Histologic examination is generally not performed or required, based on the experience of the clinical expert.

Based on feedback from the clinical expert, there is no strict schedule for when treatment response needs to be assessed. The expert suggested that it would be reasonable to assess response to treatment approximately 3 months to 6 months after initiation, then again at 6 months to 12 months, and yearly thereafter.

Discontinuing Treatment

The clinical expert consulted by CADTH stated that a patient may discontinue treatment if she wishes; however, symptoms may return after some time. The expert noted that the assessment of the risks and benefits is subjective, given that the treatment is meant to improve quality of life. In less common cases, VVA may be very severe, with erosion and ulceration of the vaginal mucosa. In such cases, pharmacologic treatment will be strongly recommended, according to the clinical expert, and may be particularly important in women with pessaries.

Prescribing Conditions

According to the clinical expert consulted by CADTH, ospemifene would most likely be prescribed in an outpatient ambulatory clinic setting by a family physician or gynecologist. The clinical expert indicated patients would self-administer the medication at home. Given that genitourinary syndrome of menopause is very common, the clinical expert believed that the vast majority of clinicians with experience in the treatment of women’s health issues would be suitable prescribers.

Additional Considerations

The clinical expert noted that vaginal prasterone is also a recent addition to traditional vaginal estrogen treatments, and along with ospemifene, may cause a shift in the management approach, as described previously.

Clinician Group Input

Input for the review of ospemifene was not received from any clinician groups.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may affect their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical experts consulted by CADTH are summarized in Table 5.

Table 5: Summary of Drug Plan Input and Clinical Expert Response

AE = adverse event.

Clinical Evidence

The clinical evidence included in the review of ospemifene is presented in 3 sections. The first section, the Systematic Review, includes pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those studies that were selected according to an a priori protocol. The second section includes indirect evidence from the sponsor and indirect evidence selected from the literature that met the selection criteria specified in the review. The third section includes sponsor-submitted LTSEs and additional relevant studies that were considered to address important gaps in the evidence included in the systematic review.

Systematic Review (Pivotal and Protocol Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of ospemifene (60 mg) for the treatment of moderate to severe dyspareunia and/or vaginal dryness — which are symptoms of VVA, a component of genitourinary syndrome of menopause — in post-menopausal women.

Methods

Studies selected for inclusion in the systematic review include pivotal studies provided in the sponsor’s submission to CADTH and Health Canada as well as those meeting the selection criteria presented in Table 6. Outcomes included in the CADTH review protocol reflect outcomes considered to be important to patients, clinicians, and drug plans.

Table 6: Inclusion Criteria for the Systematic Review

AE = adverse event; DVT = deep vein thrombosis; HRQoL = health-related quality of life; RCT = randomized controlled trial; SAE = serious adverse event; vs. = versus; WDAE = withdrawal due to adverse event.

^aThese outcomes were identified as being of particular importance to patients in the input received by CADTH from patient groups.

The literature search for clinical studies was performed by an information specialist using a peer-reviewed search strategy according to the PRESS Peer Review of Electronic Search Strategies checklist.²⁰

Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946—) through Ovid and Embase (1974—) through Ovid. All Ovid searches were run simultaneously as a multi-file search. Duplicates were removed using Ovid deduplication for multi-file searches, followed by manual deduplication in Endnote. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concept was ospemifene (Osphena/Senshio). The following clinical trial registries were searched: the US National Institutes of Health’s clinicaltrials.gov, WHO’s International Clinical Trials Registry Platform search portal, Health Canada’s Clinical Trials Database, and the European Union Clinical Trials Register.

No filters were applied to limit the retrieval by study type. Retrieval was not limited by publication date or by language. Conference abstracts were excluded from the search results. See Appendix 1 for the detailed search strategies.

The initial search was completed on November 22, 2021. Regular alerts updated the search until the meeting of the CADTH Canadian Drug Expert Committee on March 23, 2022.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the CADTH tool Grey Matters: A Practical Tool For Searching Health-Related Grey Literature.²¹ Included in this search were the websites of regulatory agencies (US FDA and European Medicines Agency). Google was used to search for additional internet-based materials. See Appendix 1 for more information on the grey literature search strategy.

The drug sponsor was contacted for information regarding unpublished studies.

Two CADTH clinical reviewers independently selected studies for inclusion in the review based on titles and abstracts, according to the predetermined protocol. Full-text articles of all citations considered potentially relevant by at least 1 reviewer were acquired. Reviewers independently made the final selection of studies to be included in the review, and differences were resolved through discussion.

Findings From the Literature

A total of 5 studies were identified from the literature for inclusion in the systematic review (Figure 1). The included studies are summarized in Table 7. A list of excluded studies is presented in Appendix 3.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Table 7: Details of Included Studies

BMI = body mass index; DB = double blind; DBP = diastolic blood pressure; ECG = electrocardiogram; FSFI = Female Sexual Function Index; FSH = follicle stimulating hormone; MBS = most bothersome symptom; MI = Maturation Index; RCT = randomized controlled trial; SBP = systolic blood pressure; SERM = selective estrogen receptor modulator; SIL = squamous intraepithelial lesion; UDI-6 = Urinary Distress Inventory – Short Form; VVA = vulvovaginal atrophy.

Note: Eight additional reports were included: Bachmann (2010),²² Nappi (2015),²³ Portman (2013),²⁴ Portman (2014),²⁵ Constantine (2015),²⁶ Archer (2019),²⁷ Goldstein (2019),²⁸ and Goldstein (2014).²⁹

^aStudies 310, 821, and 231 specified that the mammogram was obtained at screening or within 9 months before randomization.

^bStudies 821 and 231 specified that endometrial thickness was determined by a centrally read ultrasound. (Note for Study 231: mammograms obtained within 9 months of screening were to be available.)

^cPatients in Study 718 were required to have an intact uterus.

^dIn studies 231 and 718, if there was any uncertainty about the time of the last spontaneous bleeding, the post-menopausal status was confirmed with FSH levels >  40 IU/L (both studies) and estradiol levels < 0.20 nmol/L (Study 718 only).

^eExcluded Bethesda system (2001) classifications included: atypical squamous cells of undetermined significance; HPV high-risk positive; atypical squamous cells — could not exclude high-grade SIL; atypical glandular cells (endocervical, endometrial, not otherwise specified); low-grade SIL; high-grade SIL; carcinoma; unsatisfactory specimen.

^fStudy 231: Patients were excluded if they had a vaginal infection and refused treatment or the infection did not respond to treatment.

^gEstimated glomerular filtration rate of less than 60 mL/min/1.73 m³ based on the Modification of Diet in Renal Disease Study.

^hProhibited medications included: dietary supplements and herbal therapies with assumed clinically significant estrogenic vaginal effects; any form of local vaginal hormonal products; any form of oral or transdermal estrogen or progestin; any form of progestin implants (subdermal or intrauterine) or estrogen implants (pellets); any form of estrogen-alone or progestin-alone injectable drug therapy; sex hormones or medications that were expected to affect clinically significant sex hormone levels; any SERMs; any vaginal lubricant or moisturizer other than that provided by the sponsor for use in the study; systemic fluconazole, rifampicin, rifabutin, carbamazepine, phenytoin, or St John’s wort.

ⁱMaturation value (MV) is defined as MV = (S × 1) + (I × 0.5) + (P × 0), where S = % of superficial cells; I = % of intermediate cells; and P = % of parabasal cells.

Source: Clinical Study Reports.^7-11

Table 8: Details of Included Studies — Extension Study 310X

BMI = body mass index; DB = double blind; LTSE = long-term safety extension; NA = not available.

Note: One additional report was included: Simon (2013).³⁰

Source: Clinical Study Report.¹¹

Description of Studies

The primary objective of studies 310, 821, and 231 was to assess the efficacy, safety, and tolerability of ospemifene 60 mg once daily compared to placebo in the treatment of symptoms of VVA in post-menopausal women. More specifically, Study 310 assessed the treatment of symptoms of VVA broadly; Study 821 assessed the treatment of moderate to severe vaginal dryness and moderate to severe vaginal pain associated with sexual activity; and Study 231 assessed the treatment of vaginal dryness as a symptom of VVA due to menopause. The primary objective of Study 718 was to assess the long-term safety of ospemifene 60 mg once daily in the treatment of VVA in post-menopausal women with an intact uterus.

All of the included studies were phase III, double-blind, placebo-controlled RCTs that enrolled post-menopausal women with VVA. Studies 310, 821, and 718 were conducted from 2006 to 2009, whereas Study 231 was conducted from 2016 to 2017. The 4 pivotal trials were conducted primarily in the US, with the exception of Study 718, which was conducted in Europe (Belgium, Denmark, Finland, and Sweden). No study sites were located in Canada. A total of 2,798 patients were enrolled in the 4 pivotal trials, Study 310 (N = 826), Study 821 (N = 919), Study 231 (N = 627), and Study 718 (N = 426). Study 310 included 3 treatment arms (ospemifene 30 mg, ospemifene 60 mg, and placebo); however, the ospemifene 30 mg treatment arm was not reported in this review because the dosage does not align with the Health Canada–approved dose. Patients in studies 310, 821, and 231 were randomized to ospemifene 60 mg once daily or placebo at a 1:1 ratio, respectively, using a randomization code (Study 310 and 821) or web- or voice-based interactive response technology (Study 231). Randomization was stratified by uterine status in Study 310, by MBS (vaginal dryness or vaginal pain associated with sexual activity) reported on the vaginal atrophy symptom questionnaire taken at randomization (Study 821), and by severity of vaginal dryness (moderate or severe) and uterine status (Study 231). In Study 718, patients were randomized to ospemifene 60 mg once daily or placebo at a 6:1 ratio, respectively, using permuted block randomization stratified by centre.

The duration of the double-blind treatment period was 12 weeks in studies 310, 821, and 231, and it was 52 weeks in Study 718. Each study was preceded by a 4-week or 6-week screening period during which pre-defined study entry criteria were reviewed, and each included a 2-week or 4-week follow-up period. Study 231 originally included a 92-week safety assessment period following week 12; however, this phase of the study was discontinued following a protocol amendment on November 2, 2016. Study 821 also originally planned to include a long-term open-label extension study, but this was not initiated.

Study 310X was an LTSE study following Study 310. Patients with an intact uterus who completed the protocol for Study 310 were eligible. Patients without an intact uterus had the option of continuing into Study 312. Study 312 is summarized in the Other Relevant Evidence section. Excluding the ospemifene 30 mg treatment arm, a total of 118 patients were included. Patients remained on the treatment they were assigned to in Study 310, and blinding was maintained. The treatment period was 52 weeks in total (which included 12 weeks of the core study), followed by a 4-week follow-up period.

Populations

Inclusion and Exclusion Criteria

Patients included in the 4 pivotal trials were required to be between the ages of 40 years and 80 years and post-menopausal, to have 5% or fewer superficial cells in the MI of the vaginal smear, and to have a vaginal pH greater than 5.0. In Study 310, patients were included if they had at least 1 moderate or severe symptom of VVA (vaginal dryness, dyspareunia, vaginal and/or vulvar irritation or itching, difficult and/or painful urination, or vaginal bleeding with sexual activity). In Study 821, patients were included if they self-reported moderate to severe vaginal dryness or dyspareunia as the MBS of VVA. In Study 231, patients were included if they self-reported moderate to severe vaginal dryness as the MBS of VVA. Patients were not required to report VVA as the MBS to be included in Study 718. Patients in studies 310, 821, and 231 were required to have been hysterectomized or have an intact uterus with a double-layer endometrial thickness of less than 4 mm at screening. Patients in Study 718 were required to have an intact uterus.

Patients were excluded from the 4 pivotal studies if they had evidence of hyperplasia, cancer, or other pathology from an endometrial biopsy at screening, clinically significant gynecological findings other than vaginal atrophy (e.g., vaginal prolapse of grade 2 or higher), or history or evidence of malignancy, or if they consumed more than 14 drinks containing alcohol per week. Additionally, patients were excluded if they used certain treatments within a specified number of days before the initial screening visit (Table 7).

In Study 310, 812 patients were excluded if they had a BMI of greater than or equal to 37 kg/m²; in Study 231, they were excluded if they had a BMI greater than or equal to 38.5 kg/m²; and in Study 718, they were excluded if they had a BMI greater than or equal to 30 kg/m². In studies 310, 812, and 718, patients were excluded if they were currently using heparin, itraconazole, ketoconazole, or digitalis alkaloids. Additionally, patients in Study 812 and Study 718 were excluded if they were currently using HIV antivirals, clarithromycin, telithromycin, or nefazodone.

To be eligible for Study 310X, patients were required to have an intact uterus, to have met the inclusion and exclusion criteria for Study 310, and to have completed the protocol for Study 310. Patients were excluded if they had clinically significant abnormal findings at the week 12 end-of-study visit.

Baseline Characteristics

A summary of baseline characteristics is provided in Table 9. In studies 310, 821, 231, and 718, the mean age of patients was between 58.5 years (SD = 6.4) and 62.9 years (SD = 6.5), and the majority of patients were over the age of 55 years (54.0% to 58.3% were 55 years to 64 years; 16.9% to 34.9% were at least 65 years). Most patients in the included studies were White (84.7% to 100.0%), with Black patients representing 0% to 12.1% of patients; Asian patients represented fewer than 3% of patients in the 4 studies. Further, in Study 718, 49% of patients were from Finland, 28% were from Belgium, 15% were from Denmark, and 8.0% were from Sweden. The mean BMI ranged from 26.0 kg/m² (SD = 4.4) to 27.3 kg/m² (SD = 4.5) in studies 310, 821, and 231, and was not reported in Study 718. From 40.9% to 56.8% of patients in studies 310, 821, and 231 had an intact uterus; all patients were required to have an intact uterus in Study 718. Information about hot flashes was not reported in Study 718, but in studies 310 and 821, 59.0% to 63.0% of patients reported experiencing hot flashes at baseline. In contrast, 8.3% to 8.9% of patients in Study 231 reported experiencing hot flashes at baseline. The majority of patients (86.6% to 90.7%) in studies 310 and 821 did not report a UTI in the past 6 months. Information about UTIs at baseline was not reported in Study 231 or Study 718.

The proportion of patients who had previous experience with systemic or vaginal hormonal treatment varied between the studies. In the ospemifene and placebo treatment groups, respectively, 19.6% and 18.7% of patients in Study 310, 61.3% and 55.0% of patients in Study 821, 2.9% and 2.5% of patients in Study 231, and 57.0% and 52.4% of patients in Study 718 had previously received hormonal treatment. The proportions of patients in Study 821 who had previously used vaginal hormone products in the ospemifene and placebo treatment groups were 31.3% and 24.1%, respectively, as well as 34.2% and 39.7% in the ospemifene and placebo treatment groups in Study 718, respectively.

Baseline measurements for the MBS of VVA, cytology, and vaginal pH are provided in Table 10. Of the 544 patients in Study 310, 222 patients (41%) reported vaginal dryness as the MBS; 242 patients (44%) reported vaginal pain with sexual activity (dyspareunia) as the MBS; 67 patients (12%) reported vaginal and/or vulvar irritation or itching as the MBS; 4 patients (1%) reported difficult and/or painful urination as the MBS; and 5 patients (1%) reported vaginal bleeding with sexual activity as the MBS. Vaginal dryness was reported as moderate by 52% and 60% of patients in the ospemifene and placebo treatment groups, respectively, and as severe by 45% and 40% of patients in the ospemifene and placebo treatment groups, respectively. Most of the patients (70% and 71% in the ospemifene and placebo treatment groups, respectively) reported severe dyspareunia. Additionally, irritation or itching was reported as moderate by 57% and 65% of patients in the ospemifene and placebo treatment groups, respectively, and as severe by 30% and 24% of patients in the ospemifene and placebo treatment groups, respectively.

In Study 821, 314 patients (34%) reported vaginal dryness as their MBS, and 605 patients (66%) reported dyspareunia as their MBS. In Study 231, patients were required to have vaginal dryness as their MBS. Study 718 did not report assessments of MBS at baseline. More than 50% of patients in both studies reported severe vaginal dryness, and treatment groups were balanced by severity of MBS. In Study 821, treatment groups were balanced by severity of dyspareunia, with 66% (ospemifene) and 67% (placebo) of patients reporting severe dyspareunia.

The mean percentage of parabasal cells, mean percentage of superficial cells, and vaginal pH were also reported in the 4 studies and were balanced between treatment groups.

The baseline characteristics available for Study 310X were limited (Table 11); however, the characteristics of those who continued into the LTSE were similar to those in the core study in terms of age, race, and BMI.

Table 9: Summary of Baseline Characteristics (ITT Population)

BMI = body mass index; Gr. = grade; HRT = hormone replacement therapy; ITT = intention to treat; NA = not available; NR = not reported; OSP = ospemifene; PBO = placebo; SD = standard deviation; UTI = urinary tract infection; VVA = vulvovaginal atrophy.

Source: Clinical Study Reports.^7-10

Table 10: Summary of Baseline Measurements (ITT Population)

Dry = dryness stratum; Dys = dyspareunia stratum; ITT = intention to treat; NA = not applicable; NR = not reported; OSP = ospemifene; PBO = placebo; SD = standard deviation.

^aIn Study 821, vaginal dryness as the most bothersome symptom at baseline was assessed in the dryness stratum; vaginal pain with sexual activity as the most bothersome symptom at baseline was assessed in the dyspareunia stratum.

Source: Clinical Study Reports.^7-10

Table 11: Summary of Baseline Characteristics — Extension Study 310X (ITT Population)

BMI = body mass index; ITT = intention to treat; OSP = ospemifene; PBO = placebo; SD = standard deviation.

Source: Clinical Study Report.¹¹

Interventions

In each of the included studies, the intervention employed was ospemifene 60 mg tablets, administered orally once daily in the morning with food. In studies 310, 821, and 231, the duration of treatment was 12 weeks and patients were instructed to use non-hormonal vaginal lubricant as needed. Patients were provided with a vaginal lubricant (K-Y jelly) and asked to record their use of it in a diary. Of note, patients in Study 821 were advised not to use vaginal lubricant and to refrain from sexual intercourse within 24 hours before any clinic visit. In Study 718, the duration of treatment was 52 weeks, and patients were not permitted to use non-hormonal vaginal lubricant during the first 12 weeks of treatment, but they were able to use it freely after week 12.

The studies were double blind and placebo controlled, using matching placebo tablets for oral administration. The ospemifene and placebo tablets were identical in size, weight, and colour, and could not be visually differentiated.

A summary of concomitant medication use reported by patients in the included studies is presented in Table 12. The proportions of patients reporting any concomitant medication use were not available for Study 310 or Study 231. In the dryness stratum of Study 821, 94% and 88% patients randomized to ospemifene and placebo, respectively, reported any concomitant medication use. In the dyspareunia stratum of Study 821, 93% and 94% of patients randomized to ospemifene and placebo, respectively, reported concomitant medication use. The most frequently reported concomitant medications in studies 310 and 821 were multivitamins, calcium, fish oil, acetylsalicylic acid, ibuprofen, and levothyroxine sodium. In Study 231, the most frequently reported concomitant medications were levothyroxine, omeprazole, acetylsalicylic acid, and ibuprofen. In Study 718, which reported prior and concomitant medication use, 91% and 87% of patients randomized to ospemifene and placebo, respectively, reported any concomitant medication use. The most frequently reported medications used by patients in Study 718 were ibuprofen, acetaminophen, simvastatin, and estradiol.

Outcomes

A list of efficacy end points identified in the CADTH review protocol that were assessed in the clinical trials included in this review is provided in Table 13. These end points are further summarized in the discussion that follows. A detailed discussion and critical appraisal of the outcome measures are provided in Appendix 4.

Table 12: Concomitant Medications (ITT Population)

OSP = ospemifene; ITT = intention to treat; NOS = not otherwise specified; NR = not reported; PBO = placebo.

^aReported by greater than or equal to 10.0% of patients in either treatment group.

^bReported as prior and concomitant medications.

Source: Clinical Study Reports.^7-10

Table 13: Summary of Outcomes of Interest Identified in the CADTH Review Protocol

FSFI = Female Sexual Function Index; MBS = most bothersome symptom; NA = not applicable; UDI-6 = Urinary Distress Inventory – Short Form; VVA = vulvovaginal atrophy.

^aFor the co-primary outcome, the MBS was vaginal dryness or dyspareunia in Study 310 and Study 821. In Study 231, the MBS was vaginal dryness. Of note, Study 310 included patients reporting any of the symptoms of VVA (vaginal dryness, vaginal and/or vulvar irritation or itching, dysuria, dyspareunia, and vaginal bleeding associated with sexual activity) as moderate or severe. The Study 821 inclusion criteria indicated that patients must report either vaginal dryness or dyspareunia as the MBS, and in Study 231, the MBS had to be vaginal dryness.

^bWeek 8 assessments were conducted in Study 231 only.

^cAs noted, the MBS reported by patients was either vaginal dryness (Study 821 and 231) or dyspareunia (Study 821).

^dWeek 4 assessments were conducted in studies 310, 821, and 231 only.

^eWeek 24 and week 52 assessments were conducted in Study 718 only.

^fReported as a total score.

Source: Clinical Study Reports.^7-10

Symptoms

VVA and urogenital symptoms were identified as outcomes of importance to patients and the clinical expert consulted by CADTH. Specifically, the following symptoms were of interest to this review: vulvar and vaginal pain, vaginal dryness, dyspareunia, vaginal and/or vulvar irritation and/or itching, incontinence, and genitourinary prolapse.

VVA Symptoms

Vulvar and vaginal symptoms (i.e., vaginal dryness, vaginal and/or vulvar irritation or itching, dysuria, dyspareunia, and vaginal bleeding associated with sexual activity) were assessed using the VVA questionnaire. The VVA questionnaire is completed by the patient and based on a 1-month recall or recall since the last study visit. The first part of the questionnaire asks patients if they have had any of the symptoms of VVA in the past month (or since the last visit); if they have, it asks them to indicate the severity of the most severe episode using 1 of 4 options: none, mild, moderate, or severe, which correspond to scores of 0, 1, 2, or 3, respectively. In the second part of the questionnaire, patients are asked to identify which symptom of the symptoms they rated as moderate or severe in the first part of the questionnaire is the most bothersome. In Study 310 and 821, the investigator or qualified study personnel reviewed the questionnaire with patients at study visits. Study 231 did not specify whether the questionnaire was reviewed with patients.

In the studies that used the VVA questionnaire (studies 310, 821, and 231), symptoms were assessed based on:

• the severity of the MBS of VVA identified by the patient (note: to be eligible for Study 821, the patient must have reported moderate or severe vaginal dryness or dyspareunia as their MBS at screening and randomization; to be eligible for Study 231, the patient must have reported moderate to severe vaginal dryness at screening and have indicated vaginal dryness as their MBS at screening and randomization)

• the severity of each symptom of VVA.

The change from baseline in the most bothersome VVA symptom was assessed in Study 310 (any VVA symptom at week 4 and week 12), Study 821 (vaginal dryness or dyspareunia at week 4 and week 12), and Study 231 (vaginal dryness at weeks 4, 8, and 12). The changes from baseline in all symptoms reported were also assessed in these studies (at weeks 4 and 12 in Study 310 and 821, and at weeks 4, 8, and 12 in Study 231). The VVA questionnaire was not employed in Study 718.

All assessments of the change in VVA symptoms were made using the scores associated with the symptom severity rating. A change from baseline of –3, –2, –1, 0, or 1 was interpreted to indicate the following changes in symptom severity:

• –3 indicated severe (3) to none (0)

• –2 indicated severe (3) to mild (1), or moderate (2) to none (0)

• –1 indicated severe (3) to moderate (2), moderate (2) to mild (1), or mild (1) to none (0)

• 0 indicated no change

• + 1 indicated none (0) to mild (1), mild (1) to moderate (2), or moderate (2) to severe (3).

The use of the MBS approach involving patients’ self-reported and -rated VVA symptoms is recommended by the FDA for standardizing patient-reported outcome measures for clinical studies of VVA treatments for post-menopausal women.³¹ No evidence of the validity, reliability, and responsiveness of the MBS approach using the VVA questionnaire in post-menopausal women was identified. Additionally, a formal MID for the VVA questionnaire was not identified.

Urinary Symptoms

The presence or absence of urinary symptoms (urgency, frequency, incontinence, retention, and pain or discomfort) were assessed using the UDI-6. Patients completed the questionnaire at study visits, and study personnel entered the date of completion and responses in the electronic case report form. The questionnaire asked patients if they experienced the following symptoms (yes/no): frequent urination, urine leakage related to the feeling of urgency, urine leakage related to physical activity, coughing, sneezing, small amounts of urine leakage (drops), difficulty emptying their bladder, and pain or discomfort in the lower abdominal or genital area. If the patient selected yes, they were asked to rate how much the symptom bothered them based on a 4-point scale: not at all (1), slightly (2), moderately (3), or greatly (4). If the symptom was absent, it was assigned a score of 0. The scores of the 6 items were summed to obtain a total score. No evidence of validity, reliability, responsiveness, or MID was estimated for the UDI-6 in post-menopausal patients with VVA-associated symptoms.

The changes from baseline in urinary symptoms were assessed at week 4 and week 12 in studies 310 and 821. In Study 231, the change from baseline in urinary symptoms was reported as a total score from the questionnaire, which ranged from 0 to 26, where lower values indicated less urinary distress.

Health-Related Quality of Life

HRQoL was not assessed in any of the included studies.

Sexual Function

Sexual function was evaluated in Study 821 and Study 231 using the FSFI. The FSFI (Rosen et al., 2000) was developed as a brief, multi-dimensional self-report instrument for assessing key dimensions of sexual function in women. The scale consists of 19 items that assess sexual function over the past 4 weeks and yield domain scores in 6 areas: sexual desire, arousal, lubrication, orgasm, satisfaction, and pain. Of the 19 items, 2 questions correspond to the desire domain, 4 correspond to arousal, 4 correspond to lubrication, and 3 correspond to each of orgasm, satisfaction, and pain.

A domain score is obtained by summing the score of each of the individual items of a domain, which is then multiplied by the domain factor. The domain factor for desire is 0.6; arousal and lubrication have a domain factor of 0.3 each; and the remaining domains (orgasm, satisfaction, and pain) have a domain factor of 0.4 each. The total score of the FSFI is derived from the sum of the 6 domain scores. The total score for the FSFI can range from 2.0 to 36.0 points. For each FSFI domain and total score, a higher score indicates a better rating of sexual function.

In these studies, patients completed the questionnaire at study visits. The questionnaires were then reviewed by study personnel, who entered the responses into the electronic case report form.

The FSFI has been translated into more than 20 languages and adapted in more than 30 countries.^32,33 It has also been studied for use with multiple populations, including women in different age groups with diverse medical conditions and various sexual dysfunctions.^32,33 Evidence of its validity and reliability has been demonstrated; however, none of the studies was specifically designed to test the psychometric properties of the FSFI in post-menopausal women with symptoms of VVA. The generalizability of the FSFI to this population is uncertain. Additionally, no evidence of responsiveness and no formal MID were identified.

Mental Health-Related Outcomes

Outcomes related to mental health, such as depression, anxiety, mood, and cognition, were not evaluated in any of the included studies.

Bone Mineral Density

Studies 310, 821, 231, and 718 did not assess outcomes related to bone mineral density.

Cytology

All of the included studies evaluated the change from baseline to week 12 in the percentage of parabasal cells in the MI of the vaginal smear. The changes from baseline to week 12 in the percentage of superficial cells were evaluated as well. Vaginal smear samples were taken from middle third of the lateral vaginal wall and obtained by the investigator throughout the study. Vaginal smear samples were evaluated at a central pathology laboratory by a qualified pathologist, who performed the cell count that was used to determine the percentage of parabasal cells and superficial cells per sample.

Vaginal pH

The change in vaginal pH from baseline to week 12 was evaluated in all of the included studies. It was also evaluated as a change from baseline to week 4 (Study 821 and Study 231), week 8 (Study 231), and weeks 26 and 52 (Study 718), as indicated in Table 13. Vaginal pH was measured by the investigator using a pH indicator strip by pressing the indicator strip against the middle third of the vaginal wall. In Study 821 and Study 231, 2 types of pH strips with different pH ranges were used. The first had a pH range from 4 to 7; if the measurement was outside of that range, the second pH strip (with a pH range from 2 to 9) was to be used. Patients were advised not to have sexual intercourse in the 24 hours before the measurement of vaginal pH. Additionally, patients in Study 821 were advised not to use vaginal lubricant 24 hours before measurement.

Adherence

None of the included studies assessed outcomes related to adherence as an efficacy outcome.

Harms

Safety outcomes were evaluated in all included studies. Study 310X was an LTSE of Study 310 that evaluated only safety outcomes.

Statistical Analysis

General Considerations

In Study 310 and Study 821, all efficacy and safety analyses were conducted in the ITT population. In addition, the primary efficacy analyses were performed in the PP population and were considered supportive. In Study 231, all efficacy analyses were conducted in the ITT population. The primary efficacy analyses were also conducted in mITT and PP populations and considered supportive. In Study 231, safety analyses were conducted in the safety population. In Study 718, efficacy analyses were conducted in the ITT population and safety analyses were conducted in the safety population.

Primary Outcomes

Studies 310, 821, and 231 employed 4 co-primary end points evaluated as the change from baseline to week 12 in the:

• percentage of parabasal cells in the MI of the vaginal smear

• percentage of superficial cells in the MI of the vaginal smear

• vaginal pH

• severity of the MBS of VVA.

In Study 310, the severity of the MBS of VVA end point was evaluated by symptom. The change in the severity of vaginal dryness was evaluated in patients who reported vaginal dryness as their MBS. Similarly, the change in the severity of dyspareunia was evaluated in patients who reported dyspareunia as their MBS. The evaluation of the co-primary variable of MBS of VVA was performed using the evaluation of these 2 symptoms (vaginal dryness and dyspareunia).

In Study 821, patients were stratified by their self-reported MBS (vaginal dryness or dyspareunia) at randomization and each stratum was analyzed as a separate experiment. To demonstrate a statistically significant improvement of ospemifene compared to placebo for the treatment of vaginal dryness, each co-primary end point was required to be in favour of ospemifene when analyzed in patients included in the dryness stratum. Similarly, to demonstrate a statistically significant improvement of ospemifene compared to placebo for the treatment of dyspareunia, each co-primary end point was required to be in favour of ospemifene when analyzed in patients who were included in the dyspareunia stratum.

Study 231 included only patients who reported vaginal dryness as their MBS; therefore, the co-primary end point of severity of the MBS of VVA was specific to vaginal dryness as a VVA symptom.

Study 718 included the same co-primary end points, with the exception of severity of MBS of VVA, for a total of 3 co-primary end points.

Power Calculation

A summary of the sample size and power calculations for studies 310, 821, 231, and 718 is provided in Table 14. Studies 310, 821, and 231 based the power calculations on the assumption of a 2-sided alpha level of 5%. The alpha level was not reported for Study 718. Additionally, Study 310 assumed a dropout rate of 15%; dropout rate assumptions were not reported in the other studies.

In Study 310, a sample size of 795 was required for greater than 99% power to detect a difference of 20% (SD = 40.3) in the change in parabasal cells, for 98% power to detect a difference of 5% (SD = 12.4) in the change in superficial cells, for 99% power to detect a difference of 0.5 (SD = 1.1) in vaginal pH, and for 81% power to detect a difference of 0.4 (SD = 0.94) in the change in severity of the MBS of VVA.

Sample size and power calculations were conducted separately for the dryness and dyspareunia strata in Study 821. A sample size of 250 patients in the vaginal dryness stratum and 500 patients in the dyspareunia stratum was required for 99% power to detect a difference of 34.0% (SD = 35.2) in the change of parabasal cells, a difference of 8.6% (SD = 12.1) in the change of superficial cells, and a difference of 0.90 (SD = 0.98) in vaginal pH. The sample size of 250 patients provided 90% power to detect a difference of 0.42 (SD = 0.99) in the MBS of vaginal dryness in vaginal dryness stratum. In the dyspareunia stratum, 500 patients provided 80% power to detect a difference of 0.30 (SD = 1.18) in the MBS of dyspareunia.

In Study 231, a sample size of 600 patients was required for greater than or equal to 90% power to detect a difference in each of the co-primary end points. The magnitude of the difference that could be detected was not reported; however, the effect size was available (Table 14). In Study 718, 350 patients (50 to receive placebo and 300 to receive ospemifene) were considered sufficient to provide supportive evidence of efficacy for the primary efficacy end points. Additional information was not provided.

Table 14: Power Calculations

MBS = most bothersome symptom; NR = not reported; OSP = ospemifene; PBO = placebo; vs. = versus; VVA = vulvovaginal atrophy.

^aIt was estimated that 40% of patients would declare vaginal dryness as their MBS.

^bPower calculations were performed for each stratum. Both the vaginal dryness stratum and the dyspareunia stratum had 99% power to detect the indicated differences based on the same size for each stratum.

^cThe sample size calculation was based on an estimated effect size rather than difference in means.

^dThe power was not reported, but the Clinical Study Report stated that “the sample size is considered to be sufficient to provide supportive evidence of efficacy for the primary efficacy end points.”

Source: Clinical Study Reports.^7-10

Statistical Test or Model

A summary of the statistical analysis of efficacy end points is described in Table 15. In Study 310 and Study 821, the co-primary outcomes regarding the change in parabasal cells, change in superficial cells, and change in vaginal pH were analyzed using ANCOVA unless the assumptions were violated, in which case a non-parametric approach was used. Treatment and study centre were included as fixed effects for the ANCOVA model in Study 310 and Study 821, as well as uterus status in Study 310 only. The change in severity of the MBS of VVA was analyzed using the Cochran-Mantel-Haenszel approach, controlling for study centre and uterus status in Study 310, with only study centre in Study 821. In Study 231, the change in parabasal cells, change in superficial cells, and change in vaginal pH were analyzed using a mixed-effects model for repeated measures (MMRM), with treatment, week, treatment by week interaction, and study centre as fixed effects and baseline value as a covariate. The change in severity of the MBS of VVA in Study 231 was analyzed using a generalized estimating equations approach, with the same fixed effects and covariate as the MMRM analysis. In Study 718, the changes in parabasal and superficial cells were analyzed using the Cochran-Mantel-Haenszel approach with centre as a stratification factor. Further, the change in vaginal pH was analyzed using the ANCOVA model with treatment and study centre as fixed effects and baseline value as a covariate.

In Study 310, multiplicity was addressed using a step-down approach. Initially, ospemifene 60 mg versus placebo with vaginal dryness as the MBS was evaluated. If statistical significance was demonstrated for each of the 4 co-primary outcomes, dyspareunia as the MBS was evaluated at 60 mg. To demonstrate effectiveness of ospemifene in the treatment of VVA, the 4 co-primary outcomes must demonstrate statistically significant improvements compared to placebo. The evaluation was stopped if statistical significance was not achieved for any of the co-primary outcomes. Study 231 and Study 718 also did not report any methods to control for multiplicity. In Study 821, to demonstrate effectiveness, all 4 co-primary outcomes were required to demonstrate statistical significance of ospemifene 60 mg versus placebo. Each stratum was analyzed independently and in a similar manner.

Secondary outcomes were not controlled for multiplicity in any of the 4 pivotal studies.

Data Imputation Methods

In Study 310 and Study 821, a last observation carried forward (LOCF) approach was used for missing data in the ITT population when conducting efficacy analyses, with the exception of urinary symptoms (only observed cases were reported). Baseline assessments were carried forward for patients lost to follow-up after the baseline visit. For patients without a baseline measurement, a change score of 0 was used. Study 821 also noted that patients without a baseline or post-baseline measurement were excluded from the corresponding analysis.

In Study 310, missing week 4 data for patients who discontinued the study early with missing week 4 records were handled as follows:

• If the early termination visit occurred at or before 35 days from randomization visit, the missing week 4 records were replaced by early termination records.

• If the early termination visit occurred after 35 days from the randomization visit, the missing week 4 records were not replaced.

An LOCF approach was also used in Study 718 for the analysis of the co-primary end points when data were missing due to patient discontinuation or other causes. Patients with no post-baseline observations were excluded from the analysis (LOCF was not used).

All analyses performed in Study 231 were based on observed data, given that missing data were not imputed.

Subgroup Analyses

None of the included studies analyzed the data by the subgroups of interest to this review (i.e., by severity of atrophy or by prior treatment experience).

Sensitivity Analyses

In Study 231, a sensitivity analysis of the co-primary end point of MBS of vaginal dryness was conducted using an MMRM method in which categorical data were treated as continuous data. The sensitivity analysis was performed to evaluate the robustness of the generalized estimating equations approach used for the primary analysis.

Sensitivity analyses were not performed in studies 310, 821, 718, or 310X.

Secondary Outcomes of the Studies

A summary of the statistical analysis of secondary efficacy end points is included in Table 15. Secondary end points were analyzed in a similar manner to the co-primary end points in studies 310, 821, 231, and 718, with the following exceptions.

In Study 821, secondary end points were analyzed by stratum and as a combined dataset. Continuous secondary end points were analyzed using ANCOVA. The analyses were performed in a similar manner to the primary end point when analyzed by stratum. When analyzed as a combined dataset, the model included stratum rather than study centre. Additionally, an analysis including treatment by stratum interaction was also conducted for generalizability across the strata. For categorical secondary end points, or the change in severity of MBS by symptom, analyses were conducted using a Cochran-Mantel-Haenszel row mean score controlling for study centre (for analyses by stratum) or for stratum (for the combined dataset).

In Study 310 and Study 821, urinary symptoms were reported descriptively as a change from baseline and based on observed values for each time point. In Study 231, ANCOVA with baseline score as a covariate was used to analyze the UDI-6 as a change from baseline in total score and the FSFI as a change from baseline in total score and domain scores.

Table 15: Statistical Analysis of Efficacy End Points

ANCOVA = analysis of covariance; FSFI = Female Sexual Function Index; LOCF = last observation carried forward; MBS = most bothersome symptom; MMRM = mixed-effects model for repeated measures; UDI-6 = Urinary Distress Inventory – Short Form; VVA = vulvovaginal atrophy.

^aAssumptions: normality of errors, homogeneity of variances, and equality of slopes among treatment groups.

Source: Clinical Study Reports.^7-10

Analysis Populations

In studies 310, 821, 231, and 718, all efficacy analyses were conducted on the ITT population. The ITT population included all randomized patients who received at least 1 dose of study medication (ospemifene or placebo).

An mITT population was used in Study 231 to conduct a supportive analysis of the primary end points. The mITT population included ITT patients who also met the following inclusion criteria: 5% or fewer superficial cells in the MI at screening, vaginal pH greater than 5.0 at screening and at randomization, and moderate to severe vaginal dryness as the self-reported MBS of VVA at screening and at randomization.

In studies 310, 821, and 231, the primary efficacy analysis was also performed on the PP population as supportive analyses. The PP population was determined before breaking the study blind. The PP population included patients from the ITT population who had completed at least 10 weeks of treatment and the end-of-study assessments, had taken at least 85% of the study medication (Study 231 specified within 12 weeks), did not have a vaginal infection or any other medical condition that would confound the primary efficacy assessment, and did not have any other major protocol violations (Study 231 specified within 12 weeks). Major protocol violations were reviewed on a case-by-case basis to determine eligibility for the PP dataset. An example of a violation that would exclude a patient is the use of concomitant hormonal medication.

In Study 310 and Study 821, all safety analyses were conducted on the ITT population using an as-treated approach; i.e., patients were assessed based on the treatment they received rather than on the treatment group to which they were randomized. For this review, these populations will be referred to as the safety population. Similarly, in Study 231 and Study 718, safety analyses were performed in the safety population, which was defined as all randomized patients who received at least 1 dose of a study drug; the analyses were performed according to the treatment received. In addition, it was noted that patients who received both ospemifene and placebo were included in the ospemifene treatment group.

Study 310X

All analyses were done on the ITT population. The ITT population was defined as all patients who entered the study and received at least 1 dose of study medication in Study 310X. All 180 enrolled patients were included in the ITT population.

Results

Patient Disposition

The disposition of patients in studies 310, 821, 231, and 718 is summarized in Table 16. Approximately 15% of patients in Study 310, 11% of patients in Study 821, and 11% of patients in Study 231 discontinued from the studies. Discontinuation rates were similar between the ospemifene and placebo treatment groups. In each of these studies, AEs were the most common reasons for discontinuation, followed by patient decision or withdrawal of consent. The proportion of patients who withdrew from the study was balanced by reason between the treatment groups, except for those withdrawing due to AEs (5.4% versus 3.1% for ospemifene versus placebo, respectively) and, in Study 821, patient decision (1.7% versus 4.2% for ospemifene versus placebo, respectively). In Study 718, 18% of patients discontinued, and the proportion of patients who withdrew was lower in the placebo group (12.7%) compared with the ospemifene group (19.0%). The most common reasons for discontinuing were AEs (9.5% versus 13.2%, placebo versus ospemifene, respectively) and patient request (1.6% versus 3.9%). Of the 14 patients who withdrew by patient request, 2 patients felt better with hormone replacement therapy, 4 patients had unresolved hot flashes, 2 patients complained of lack of effect, and 1 patient had surgery to remove an ovarian cyst and did not want further gynecological examinations. The other 5 patients withdrew for personal reasons or did not wish to provide further comment.

The number of patients in each study population analyzed is summarized in Table 16. In Study 310, reasons for exclusion from the PP population were limited to protocol deviations. This included no moderate or severe MBS at baseline (15 patients and 9 patients randomized to ospemifene and placebo, respectively), more than 5% superficial cells at baseline (7 patients and 10 patients randomized to ospemifene and placebo, respectively), and inadequate washout of medications before study (6 patients and 1 patient randomized to ospemifene and placebo, respectively). Other reasons included pH of 5 or less at baseline, and taking prohibited, hormonally active medications during study, baseline endometrial thickness of 4 mm or greater, homogenous or heterozygous for Factor V Leiden, a BMI of 37 or greater at baseline, and drug dispensing errors.

In Study 821, 11.4% of patients overall were excluded from the PP population. The most common reasons were less than 85% compliance with the treatment protocol (11.7% and 11.2% of patients in the ospemifene and placebo treatment groups, respectively) and treatment durations of less than 70 days (8.9% and 7.7% of patients in the ospemifene and placebo treatment groups, respectively). Other reasons for exclusion from the PP were infrequent and balanced between the treatment groups, with the exception of foreign organisms detected in the vaginal smear at baseline or visit 4, which occurred in 3.2% of patients in the ospemifene treatment group and 0.2% of patients in the placebo treatment group.

In Study 231, 14% and 16% of patients randomized to ospemifene and placebo, respectively, were excluded from the mITT set because they had more than 5% superficial cells in the MI at baseline or no measurement at baseline. Other reasons for exclusion were infrequent, occurring in fewer than 1% of patients. This warranted exclusion from the PP as well. Other reasons for exclusion from the PP included: treatment duration less than 10 weeks (9% of patients in both treatment groups), vaginal infection (4% and 2% of patients randomized to ospemifene and placebo, respectively), and incorrect dispensing of the study drug at any visit (1% and 2% of patients randomized to ospemifene and placebo, respectively).

Table 16: Patient Disposition

ITT = intention to treat; mITT = modified intention to treat; OSP = ospemifene; NR = not reported; PBO = placebo; PP = per protocol.

^aThis count does not include 2 patients in the placebo treatment group who discontinued due to treatment-emergent adverse events and who are accounted for in the summary of harms.

^bThree hundred and 16 patients were randomized to ospemifene. Three patients did not receive the study drug and 4 patients who were randomized to placebo received ospemifene in error.

^cThree hundred and 15 patients were randomized to placebo, and 4 received ospemifene in error. Additionally, 1 patient was excluded from the safety population because they were enrolled at 2 different sites at the same time.

Source: Clinical Study Reports.^7-10

The safety population was identical to the ITT population in Study 718.

A summary of the patient disposition in Study 310X is available in Table 17. Of the 464 patients who completed Study 310, 118 patients (25%) enrolled in extension Study 310X. Overall, 23% of patients discontinued from study. The proportion of patients who withdrew was lower in the ospemifene group (17%) compared with the placebo group (31%). The most common reasons for discontinuation were patient decision or withdrawal of consent (11% overall) and AEs (4% overall). The proportion of patients who decided to discontinue was lower in the ospemifene group (7%) compared than in the placebo group (16%). The proportion of patients who discontinued due to AEs was greater in the ospemifene group (6%) compared with the placebo group (2%). Other reasons for discontinuing were reported in fewer than 3% of patients overall, including loss to follow-up, major protocol violation, significant non-compliance with treatment or study procedures, and other.

Exposure to Study Treatments

A summary of duration of exposure and compliance with study treatments is provided for studies 310, 821, 231, and 718 in Table 18, and for Study 310X in Table 19. In each of the 12-week studies (studies 310, 821, and 231), patients were exposed to treatment for approximately 80 days to 82 days, and the exposures were similar between treatment groups. In the 52-week study (Study 718), the median number of days on which patients were exposed to ospemifene was 362 (range = 8 to 392); for patients in the placebo group, the median number was 363 (range = 61 to 378).

In studies 310, 821, 231, and 718, the compliance rate, defined as the number of doses taken divided by the number of doses that should have been taken for the duration of the treatment period, was greater than 92% in all treatment groups. The compliance rates in studies 310 and 718 were lower in the ospemifene treatment groups relative to the placebo treatment groups (93% versus 96% in Study 310, and 95% versus 99% in Study 718).

In Study 310X, the mean durations of treatment were 254 days (SD = 70) in the ospemifene treatment group and 232 days (SD = 93) in the placebo treatment group. Further, the mean compliance rates were 85% (SD = 22) in the ospemifene treatment group and 93% (SD = 17) in the placebo treatment group. When analyzed as medians (ranges), the reported treatment durations and compliance rates were similar between treatment groups.

Table 17: Patient Disposition — Extension Study 310X

ITT = intention to treat; NR = not reported; OSP = ospemifene; PBO = placebo.

Source: Clinical Study Report.¹¹

Table 18: Exposure to Study Treatments (ITT Population)

ITT = intention to treat; OSP = ospemifene; PBO = placebo; SD = standard deviation.

^aIn Study 231, the duration of treatment was reported in the safety population. All other studies reported duration of treatment in the ITT population.

Source: Clinical Study Reports.^7-10

Table 19: Exposure to Study Treatments (ITT Population) — Study 310x

ITT = intention to treat; OSP = ospemifene; PBO = placebo; SD = standard deviation.

Source: Clinical Study Report.¹¹

Efficacy

Only the efficacy outcomes and analyses of subgroups identified in the review protocol are reported in this section. See Appendix 3 for detailed efficacy data.

Symptoms

In studies 310, 821, 231, and 718, the VVA questionnaire was used to assess the symptoms of VVA; however, the symptoms were analyzed using different approaches. The change from baseline in symptoms of VVA reported as moderate or severe at baseline, the change from baseline in symptoms of VVA reported as the MBS at baseline, and in general, the change from baseline in symptoms of VVA (broadly) were considered clinically relevant and have been summarized. Further, although the VVA questionnaire evaluated the 5 symptoms of VVA and they were reported in the included studies, the results summarized in this section focus on 3 of the symptoms (vaginal dryness, dyspareunia, and vaginal irritation or itching), as per the CADTH systematic review protocol.

Most Bothersome Symptoms of VVA (Vaginal Dryness, Vaginal Pain Associated With Sexual Activity, Vaginal and/or Vulvar Irritation and/or Itching)

One of the co-primary end points in studies 310, 821, and 231 was the change from baseline in severity of vaginal dryness, reported as the self-reported MBS of VVA, assessed at week 12. The MBS of VVA was not assessed in Study 718. A summary of the results is provided in Table 20.

In Study 310, the analysis of the co-primary end point for MBS of vaginal dryness included patients in the ITT population who identified vaginal dryness as the MBS of VVA at baseline. Of this subset of patients, 51.7% and 59.6% of patients in the ospemifene and placebo treatment groups, respectively, reported moderate vaginal dryness as the MBS of VVA at baseline. Similarly, 44.9% of patients on ospemifene and 39.4% on placebo reported severe vaginal dryness as the MBS at baseline. At week 12, the mean changes in the severity of the symptom were –1.26 (SD = 1.03) and –0.84 (SD = 1.00) for the ospemifene and placebo treatment groups, respectively, indicating a greater reduction in symptom severity with ospemifene compared to placebo (P = 0.021). |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||| Of note, 29.7% of patients in the ospemifene group and 18.3% in the placebo group reported no symptoms of vaginal dryness at week 12.

Patients in Study 821 were stratified at baseline by MBS. The analysis of the co-primary end point for MBS of vaginal dryness included only patients from the dryness stratum. At baseline in both the ospemifene and placebo treatment groups of the dryness stratum, 49% of patients reported moderate vaginal dryness as the MBS and 51% reported severe vaginal dryness as the MBS. At week 12, there was no difference between treatment groups in the change in severity of vaginal dryness as the MBS of VVA, based on a mean change of –1.3 (SD = 1.08) for ospemifene and –1.1 (SD = 1.02) for placebo (P = 0.080).

All patients included in Study 231 reported moderate or severe vaginal dryness as the MBS of VVA at baseline. The proportions of patients reporting moderate dryness were 47.3% and 45.5% for ospemifene and placebo, respectively. Severe dryness at baseline was reported by 52.7% of patients in the ospemifene group and 54.5% in the placebo group. The mean change in symptom severity from baseline to week 12 was –1.29 (SD = 1.01) for ospemifene and –0.91 (SD = 0.96) for placebo. This corresponded to an odds ratio of 2.23 (95% CI, 1.62 to 3.06; P < 0.0001).

The supportive analyses performed in Study 821 (PP population) and Study 231 (PP population and mITT population) were consistent with the primary analysis (P = 0.0143 and P < 0.0001, respectively).

Table 20: Change From Baseline in Vaginal Dryness, Reported as the MBS of VVA at Baseline (ITT Population)

CI = confidence interval; CMH = Cochran-Mantel-Haenszel; ITT = intention to treat; MBS = most bothersome symptom; OR = odds ratio; OSP = ospemifene; PBO = placebo; SD = standard deviation; vs. = versus; VVS = vulvovaginal atrophy.

Notes: In Study 310, P values for treatment comparisons (each active vs. placebo) are from the CMH row mean score test controlling for uterine status (intact or hysterectomized) and pooled centre.

In Study 821, the P value was computed using CMH row mean score test controlling for centre.

In Study 231, odds ratio was exponential of the mean of cumulative log odds ratio. To calculate the odds ratio, 95% CI, and P value, the generalized estimating equations method was used.

Source: Clinical Study Reports.^7-9

In studies 310 and 821, the change from baseline to week 12 in the MBS of dyspareunia was also a co-primary end point. This was evaluated in patients who reported dyspareunia as their MBS. In Study 310, the statistical significance for the assessment of vaginal dryness (described previously) needed to be demonstrated before performing the analysis for dyspareunia as the MBS of VVA. A summary of the results is provided in Table 21. Of note, Study 231 included only patients whose MBS was vaginal dryness; therefore, this outcome does not apply. As previously noted, Study 718 did not assess the MBS of VVA.

In Study 310 and Study 821, the majority of patients included in the analyses rated dyspareunia as severe at baseline (Study 310: 70% and 71% for ospemifene and placebo, respectively; Study 821: 66.3% and 67.2% for ospemifene and placebo, respectively). In Study 310, the mean change from baseline to week 12 in the severity of dyspareunia was –1.2 (SD = 1.3) in the ospemifene treatment group and –0.9 (SD = 1.1) in the placebo treatment group. In Study 821, the mean change from baseline in the severity of dyspareunia was –1.5 (SD = 1.08) in the ospemifene treatment group and –1.2 (SD = 1.12) in the placebo treatment group. The difference in the change in severity of dyspareunia was greater for patients in the ospemifene group than for those in the placebo group in both studies (Study 310, P = 0.023; Study 821, P = 0.0001). Additionally, 28.3% and 18.9% of patients in the ospemifene and placebo treatment groups of Study 310, respectively, reported no symptoms of dyspareunia at week 12. In Study 821, 38.0% of patients in the ospemifene group and 28.1% of patients in the placebo group reported no symptoms of dyspareunia at week 12.

The supportive analyses performed in the PP populations of Study 310 and Study 821 were consistent with the primary analyses (P = 0.004 and P = 0.0004, respectively).

Table 21: Change From Baseline in Dyspareunia, Reported as the MBS of VVA at Baseline (ITT Population)

CMH = Cochran-Mantel-Haenszel; ITT = intention to treat; OR = odds ratio; OSP = ospemifene; PBO = placebo; SD = standard deviation; vs. = versus.

Study 310: The P values for treatment comparisons (each active vs. placebo) are from the CMH row mean score test controlling for uterine status (intact uterus or hysterectomized) and pooled centre.

Study 821: The P value was computed using CMH row mean score test controlling for centre.

Source: Clinical Study Reports.^7,8

Symptoms of VVA (Vaginal Dryness, Dyspareunia) Reported as Moderate or Severe at Baseline

The change from baseline in VVA symptoms reported as moderate or severe at baseline was assessed in Study 310 and Study 821. These were secondary outcomes in both studies and not adjusted for multiplicity. Summaries of the results for vaginal dryness and dyspareunia are provided in Table 22 and Table 23, respectively. The change from baseline in VVA symptoms reported as moderate or severe at baseline was also assessed in Study 231, but was reported as the MBS of VVA, which was a requirement for inclusion in the study. The results for Study 231 are summarized in Table 20.

In Study 310 and Study 821, |||||||||||||||||||||||| of the patients included in the analysis of change from baseline in vaginal dryness (Table 22) reported vaginal dryness as moderate at baseline, and |||||||||||||||| reported it as severe at baseline. The proportion of patients reporting moderate dryness was |||||||| in Study 310 |||||||||||||||| for ospemifene and placebo, respectively) than in Study 821 ||||||||||||||| for ospemifene and placebo, respectively) and Study 231 (|||||||||||||||| for ospemifene and placebo, respectively).

In Study 310, the change from baseline |||||||||||||||| in the severity of vaginal dryness was |||||||| in patients in the ospemifene treatment group than in the placebo treatment group ||||||||, based on a mean change of |||||||||||||||| for ospemifene and |||||||||||||||| for placebo. |||||||||||||||||||||||||||||||| of patients in the ospemifene treatment group and |||||||| of patients in the placebo treatment group reported ||||||||||||||||||||||||||||||||||||||||.

In Study 821, the change from baseline |||||||||||||||| in the severity of vaginal dryness was |||||||||||||||| for ospemifene and |||||||||||||||| for placebo (ospemifene versus placebo, |||||||||||||||||||||||||||||||| of patients in the ospemifene treatment group and |||||||| of patients in the placebo treatment group reported ||||||||||||||||||||||||||||||||||||||||.

Table 22: Change From Baseline in Vaginal Dryness, Reported as Moderate or Severe at Baseline (ITT Population)

CMH = Cochran-Mantel-Haenszel; ITT = intention to treat; LOCF = last observation carried forward; OSP = ospemifene; PBO = placebo; SD = standard deviation; vs. = versus.

Note: Missing values were replaced through the LOCF for ITT patients.

^aP values for treatment comparisons (OSP vs. PBO) are from the CMH row mean score test controlling for stratum (Study 310 = uterus status [intact or hysterectomized uterus] and pooled centre; Study 821 = dryness stratum and dyspareunia stratum). P values were not adjusted for multiplicity (i.e., type I error rate was not controlled for).

Source: Clinical Study Reports.^7,8

The results for the change from baseline in dyspareunia reported as moderate or severe at baseline, as reported in studies 310, 821, and 231, are provided in Table 23. Of the patients included in the analysis in Study 310 and Study 821, |||||||| of patients reported |||||||| dyspareunia at baseline. In Study 231, |||||||||||||||| of patients reported |||||||| dyspareunia at baseline. Of note, Study 310 included patients if they had at least 1 moderate or severe symptom of VVA (vaginal dryness, dyspareunia, vaginal and/or vulvar irritation or itching, difficult and/or painful urination, or vaginal bleeding with sexual activity), and Study 821 included patients with moderate to severe vaginal dryness or dyspareunia as the MBS of VVA.

In Study 310, the change in severity of dyspareunia from baseline ||||||||||||||||||||||||||| in the ospemifene treatment group and |||||||||||||||||||||||| in the placebo treatment group ||||||||||||||||. The proportions of patients included in the analysis reporting no symptoms of dyspareunia at week 12 were |||||||||||||||| in the ospemifene and placebo treatment groups, respectively. In Study 821, the change in severity of dyspareunia from baseline |||||||||||||||||||||||||||||||||||| in the ospemifene treatment group and |||||||||||||||||||||||| in the placebo treatment group. The proportions of patients included in the analysis reporting no symptoms of dyspareunia at week 12 were |||||||||||||||||||||||| in the ospemifene and placebo treatment groups, respectively.

In Study 231, the change from baseline in severity of dyspareunia was |||||||||||||||||||||||||||||||||||||||| in the ospemifene treatment group and |||||||||||||||||||||||||||||||| in the placebo treatment group. This corresponded to ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||.

Table 23: Change From Baseline in Dyspareunia, Reported as Moderate or Severe at Baseline (ITT Population)