Results of the search

We retrieved 139 records from the search of the electronic database and no other records from other sources (Fig.1). A total of 94 records were screened after duplicates were removed. We reviewed full copies of 13 and assessed them for eligibility. We identified four articles as possibly meeting the review inclusion criteria, and nine of them were ineligible for inclusion. One article was a non-randomized controlled trial that evaluated the clinical response of elagolix-treated women who did not achieve the primary outcome [22]. Two reviews, one on predictors of response to elagolix with add-back therapy and the other on medical treatment of uterine leiomyoma, were relevant to our research query [4, 23]. There was no outcome of interest in the four papers as two papers [3, 11] on elagolix pharmacotherapy and pharmacodynamics and another two more papers [14, 24] on drug-drug interactions were written. Adenomyosis was the topic of two more publications [25, 26]. We attempted to contact the trial authors for the full article but received no response. Therefore, we included four trials.

Four randomized controlled trials with 1949 participants were included in the study [27,28,29,30]. All four trials reported the primary outcome. All trials were sponsored by AbbVie [27,28,29,30].

All four studies were carried out in 323 locations across the United States, Puerto Rico, and Canada. One trial recruited participants from clinic settings [27]. The other three trials did not mention the location from which participants were recruited [28,29,30]. Three studies included premenopausal women aged 18 to 51 at the screening time [28,29,30], while one study recruited participants aged 2049 [27]. They underwent ultrasonography-confirmed diagnosis of uterine fibroids and heavy menstrual bleeding, as characterized by more than 80ml of menstrual blood loss per menstrual cycle for at least two cycles. The trials excluded participants due to a complex ovarian cyst, cancer, pelvic inflammatory disorder, osteoporosis history, or metabolic bone disease. Participants who had a myomectomy or hysterectomy for symptomatic uterine fibroid were exempted from the study [27,28,29,30].

Participants in the trials were randomized to the intervention and comparison groups. Two identical, double-blind, randomized, placebo-controlled, six-month phase 3 trials (Elaris Uterine Fibroids 1 and Elaris Uterine Fibroid 2) have been reported in one trial [29]. Elaris Uterine Fibroid-1 and Elaris Uterine Fibroid-2 participants were later randomized or pooled into a new study [30] to look at the long-term twelve-month safety and efficacy of elagolix with or without estradiol/norethindrone acetate. The meta-analysis included four trials that evaluated the primary outcomes. Three trials compared elagolix with placebo [27,28,29], and four trials compared to elagolix with estradiol/norethindrone acetate [27,28,29,30]. Only one trial compared elagolix to placebo at different doses of 100mg bd, 200mg bd, 300mg bd, 400mg qd, and 600mg qd [27]. One study was compared to placebo at doses of 300mg bd and 600mg qd [28]. Another trial was compared elagolix to placebo at a dose of 300mg bd [29].

In a comparison of elagolix to elagolix with estradiol/norethindrone acetate, one trial compared it at a dose of 0.5mg estradiol/0.1mg norethindrone acetate [27], while two trials compared it at a dose of 1.0mg estradiol/0.5mg norethindrone acetate [29, 30]. In one trial, elagolix was compared to elagolix with estradiol/norethindrone acetate at two doses: 0.5mg estradiol/0.1mg norethindrone acetate and 1.0mg estradiol/0.5mg norethindrone acetate [28]. All medications are taken orally as tablets or capsules. The duration of treatment differed between trials compared to elagolix versus placebo, as only one trial was three months [27], and the other two trials were six months [28, 29]. In contrast, the length of treatment differed between trials when comparing elagolix to elagolix with estradiol/norethindrone acetate, with a three-month [27], a six-month [28, 29], and a twelve-month [30] period.

The validated alkaline hematin method was used to quantify and evaluate the primary outcome in all four trials [27,28,29,30]. Any spotting or bleeding episodes on a sanitary pad were reported at the time of screening or during treatment. Participants kept an electronic daily bleeding diary (eDiary) and assessed bleeding patterns using the validated Mansfield-Voda-Jorgenson Menstrual Bleeding Scale [31]. All studies were followed up to at least three-months duration. The primary outcome was measured during the last month of the treatment period.

All four trials reported all secondary outcomes except for one study [27], which did not record bone mineral density due to a limited study time and a small sample size per group. Reduction of uterine and fibroid volume was calculated using trans abdominal or transvaginal ultrasound. The mean percentage change from baseline to the end of the treatment month was recorded.

The Uterine Fibroid Symptom and Quality of Life questionnaires cumulative score were used to measure symptom severity reduction and change in health-related quality of life in women with symptomatic uterine fibroids. It was a disease-specific, self-administered, validated questionnaire. There were 37 questions in all, split into two parts. The first part consisted of an 8-item symptom severity scale. The second part consisted of a 29-item health-related quality of life subscale with six domains (concern, behaviors, energy/mood, power, self-consciousness, and sexual function). All items are rated on a 5-point scale, with symptom intensity items ranging from not at all to a great deal, and health-related quality of life items ranging from none of the time to all of the time. The cumulative score for each of the two components was determined by adding the symptom intensity and health-related quality of life subscale scores and translating them to a 0-to-100-point scale. Higher overall health-related quality of life scores indicated better quality of life, while lower symptom severity scores indicate better quality of life.

The percentage of increase in hemoglobin concentration from baseline to the last month of treatment was reported in all trials. Loss of bone mineral density was assessed by dual-energy x-ray absorptiometry scans of the lumbar spine, total hip, and femoral neck. The mean percentage change in bone mineral density from baseline to the last month of treatment was recorded in three studies [28,29,30]. Any adverse events were recorded beginning with the first dose of the study drug and continuing for up to 30days after completing the last dose of the study drug. All four trials identified common adverse events such as hot flushes, headaches, nausea, and fatigue. In this review, only two trials documented adverse events such as abdominal pain, dizziness, and hypertension [27, 28]. Other non-significant adverse events identified in clinical trials will not be addressed in this review.

The assessment risk of bias is shown in Figs.2 and 3. Figure2 shows the proportion of studies assessed as low, high or unclear risk of bias for each risk of bias indicator. Figure3 shows the risk of bias indicators for individual studies. The details of these trials are found in the table of characteristics of included studies (Table 1).

Risk of bias graph: review authors judgements about each risk of bias item presented as percentages across all included studies

Risk of bias summary: review authors judgements about each risk of bias item for each included study

Only one trial, with 271 participants, was reported to have been recruited in a clinic setting, while the other three were not [27]. The method of randomization was not reported in all four trials [27,28,29,30]. Thus, we judged random sequence generation as having an unclear risk of bias. Allocation concealment was not mentioned and regarded as unclear in four trials [27,28,29,30].

Participants, care provider, investigator and outcome assessor were masked in all four trials. The details on blinding were not recorded in all four trials, but the outcomes were unlikely to be influenced as it was objectively collected and measured using standardized methods [27,28,29,30]. Therefore, they are judged as having a low risk of bias.

More than 80% of participants completed the studies in two trials [27, 30]. Meanwhile, 74.4% of participants in one trial completed the study [28]. Approximately 129 of the 571 participants failing to complete the analysis due to hypoestrogenism side effects (n=39), withdrawal (n=38), loss of follow up (n=25), noncompliance (n=11), lack of efficacy (n=3), surgery (n=4) and other (n=9) [28]. About 78% of 791 participants completed studies in Elaris Uterine Fibroid-1 and Elaris Uterine Fibroid-2 [29]. The study drug was discontinued by similar proportions of women in both treatment groups (16.5% for elagolix with estradiol/norethindrone acetateand 19.4% for elagolix alone), with the most common primary reason being lost to follow-up (5.0%and 5.1%,respectively) in one trial [29]. Missing data were evenly balanced across groups, and the reasons were similar. The most common reasons for missing outcome data included withdrawal, noncompliance, loss to follow up, hypoestrogenism side effects, pregnancy, and surgery, which led to discontinuation.

All four trials reported the outcomes as specified in their methods section[27,28,29,30]. The outcomes listed in the registered protocol were those reported. Although changes in bone mineral density were assessed as an exploratory parameter, one trial did not report due to the short duration of the study and the relatively small sample size per group [27]. We graded it as having a low risk of bias.

We discovered that women with asymptomatic anemia and a hemoglobin level of less than 12g/dl at screening or during the study period were advised to take iron supplements in two trials [27, 30]. This could have an influence on the hemoglobin level at the end of the treatment period. Thus, we judged it as having a high risk of bias. We detected no other potential source of bias in the other two trials [28, 29].

There would be two comparisons evaluated in this review, i.e., comparing elagolix versus placebo and comparing elagolix versus estradiol/norethindrone acetate.

Elagolix has increased the number of patients with a reduction of menstrual blood loss of less than 80ml (RR 4.81, 95% CI 2.45 to 9.45; I2 statistic=89%; P<0.001; four trials, 869 participants; moderate quality evidence) (Fig.4, Table 2) [27,28,29] or more than 50% from baseline (RR 4.87, 95% CI 2.55 to 9.31; I2 statistic=87%; P<0.001; four trials, 869 participants; moderate quality evidence) (Fig.5, Table 2) [27,28,29] compared to placebo. The sensitivity analysis did not change the cumulative effect estimate. Table 3 showed the subgroup analysis for reduction of menstrual blood loss of less than 80ml or more than 50% reduction from baseline stratified by frequency of drug administration, uterine and fibroid volume (Additional file 1).

Comparison between elagolix and placebo for the outcome reduction of menstrual blood loss of less than 80ml

Comparison between elagolix and placebo for the outcome reduction of menstrual blood loss of more than 50%

For the secondary outcomes, elagolix has increased the number of patients with improved hemoglobin level (RR 2.46, 95% CI 1.93 to 3.13; I2 statistic=0%; P<0.001; four trials, 554 participants; moderate quality evidence) [27,28,29], reduced the mean percentage change in uterine volume (MD 34.50, 95% CI 43.48 to 25.53; I2 statistic=63%; P<0.001; four trials, 783 participants; moderate quality evidence) [27,28,29], fibroid volume (MD 31.39, 95% CI 44.69 to 18.09; I2 statistic=65%; P<0.001; four trials, 750 participants; moderate quality evidence) [27,28,29], severity of symptoms (MD 31.54, 95% CI 41.85 to 21.22; I2 statistic=96%; P<0.001; four trials, 814 participants; low quality evidence) [27,28,29], and improved health-related quality of life (MD 30.64, 95% CI 20.14 to 41.15; I2 statistic=95%; P<0.001; four trials, 812 participants; low quality evidence) [27,28,29] (Additional file 1, Table 2) compared to placebo.

Elagolix has reduced bone mineral density in lumbar spine (MD 2.82, 95% CI 3.30 to 2.35; I2 statistic=0%; P<0.001; three trials, 574 participants; moderate quality evidence) [28, 29], total hip (MD 1.97, 95% CI 2.37 to 1.57; I2 statistic=46%; P<0.001; three trials, 574 participants; moderate quality evidence) [28, 29] and femoral neck (MD 1.92, 95% CI 2.61 to 1.23; I2 statistic=34%; P<0.001; three trials, 574 participants; moderate quality evidence) [28, 29] (Fig.6, Table 2) compared to placebo.

Comparison between elagolix and placebo for the outcome of bone mineral density (A: lumbar spine, B: total hip, C: femoral neck)

There was no significant of severe, serious or adverse event led to discontinuation of elagolix treatment. Elagolix has increased the number of patients with side effect of hot flush (RR 7.47, 95% CI 4.99 to 11.18; I2 statistic=8%; P<0.001; four trials, 890 participants; moderate quality evidence) [27,28,29] and headache (RR 1.88, 95% CI 1.25 to 2.83; I2 statistic=0%; P<0.001; four trials, 890 participants; low quality evidence) [27,28,29] (Fig.7, Table 4) compared to placebo.

Comparison between elagolix and placebo for the outcome of adverse events (A: hot flush, B: headache)

B) Comparison between elagolix and elagolix with estradiol/norethindrone acetate.

There was no difference in menstrual blood loss of less than 80ml (RR 1.08, 95% CI 1.00 to 1.16; I2 statistic=56%; P=0.070; five trials, 1365 participants; moderate quality evidence) (Fig.8, Table 5) [27,28,29,30] or more than 50% reduction from baseline between the elagolix (RR 1.08, 95% CI 1.01 to 1.15; I2 statistic=43%; P=0.020; five trials, 1365 participants; high quality evidence) (Fig.9, Table 5) [27,28,29,30] and elagolix with estradiol/norethindrone acetate. The sensitivity analysis did not change the cumulative effect estimate. Table 6 showed the subgroup analysis for reduction of menstrual blood loss of less than 80ml or more than 50% reduction from baseline stratified by dosage and uterine volume (Additional file 1).

Comparison between elagolix and elagolix with estradiol/norethindrone acetate for the outcome reduction of menstrual blood loss of less than 80ml

Comparison between elagolix and elagolix with estradiol/norethindrone acetate for the outcome reduction of more than 50% menstrual blood loss

Foe secondary outcomes, there was no difference improvement in hemoglobin level between elagolix (RR 0.99, 95% CI 0.80 to 1.22; I2 statistic=68%; P=0.930; five trials, 899 participants; low quality evidence) [27,28,29,30] and elagolix with estradiol/norethindrone acetate. However, elagolix has reduced mean percentage change in uterine volume (MD 17.47, 95% CI 27.54 to 7.40; I2 statistic=58%; P<0.001; five trials, 1250 participants; moderate quality evidence) [27,28,29,30], fibroid volume (MD 23.18, 95% CI 28.98 to 17.38; I2 statistic=0%; P<0.001; five trials, 1208 participants; high quality evidence) [27,28,29,30], symptoms severity (MD 9.05, 95% CI 9.68 to 8.43; I2 statistic=0%; P<0.001; five trials, 1288 participants; high quality evidence) [27,28,29,30], and increased health-related quality of life (MD 9.94, 95% CI 5.82 to 14.06; I2 statistic=76%; P<0.001; five trials, 1287 participants; low quality evidence) [27,28,29,30] (Additional file 1, Table 5) compared to elagolix with estradiol/norethindrone acetate.

Elagolix has reduced bone mineral density in the lumbar spine (MD 2.63, 95% CI 3.12 to 2.14; I2 statistic=49%; P<0.001; four trials, 1126 participants; moderate quality evidence [28,29,30], and total hip (MD 1.93, 95% CI 2.56 to 1.31; I2 statistic=75%; P<0.001; four trials, 1126 participants; very low quality evidence) [28,29,30] except femoral neck (MD 0.77, 95% CI 1.84 to 0.30; I2 statistic=78%; P=0.160; four trials, 1126 participants; very low quality evidence) [28,29,30] (Fig.10, Table 5) compared to elagolix with estradiol/norethindrone acetate.

Comparison between elagolix and elagolix with estradiol/norethindrone acetate for the outcome of bone mineral density (A: lumbar spine, B: total hip, C: femoral neck)

There was no difference of severe, serious or adverse event led to discontinuation between elagolix treatment and elagolix with estradiol/norethindrone acetate. Elagolix has increased the number of patients with side effect of hot flush (RR 2.67, 95% CI 2.30 to 3.10; I2 statistic=0%; P<0.001; five trials, 1403 participants; moderate quality evidence) [27,28,29,30], reduced the number of patients with risk of nausea (RR 0.63, 95% CI 0.43 to 0.91; I2 statistic=0%; P=0.010; five trials, 1403 participants; low quality evidence) [27,28,29,30] and fatigue (RR 0.43, 95% CI 0.23 to 0.80; I2 statistic=0%; P=0.008; five trials, 1403 participants; low quality evidence) [27,28,29,30] (Fig.11, Table 7) compared to elagolix with estradiol/norethindrone acetate.

Comparison between elagolix and elagolix with estradiol/norethidrone acetate for the outcome adverse event (A: hot flush, B: nausea, C: fatigue)

Read the original:
Elagolix treatment in women with heavy menstrual bleeding associated with uterine fibroid: a systematic review and meta-analysis - BMC Women's Health...

Comments are closed.