LIBERTY 2: Efficacy & Safety Study of Relugolix in Women With Heavy Menstrual Bleeding Associated With Uterine Fibroids: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for LIBERTY 2: Efficacy & Safety Study of Relugolix in Women With Heavy Menstrual Bleeding Associated With Uterine Fibroids (NCT03103087), a randomized phase 3 study. The primary outcome is Percentage Of Participants Who Achieved A Menstrual Blood Loss (MBL) Volume Of < 80 mL And A ≥ 50% Reduction From Baseline MBL Volume With Relugolix Plus E2/NETA.

2. Clinical Question

LIBERTY 2 was designed to evaluate whether participants met a combined response definition: menstrual blood loss volume below 80 mL and a reduction of at least 50% from baseline. The endpoint therefore reflects both reaching a specified final bleeding-volume threshold and making a substantial reduction relative to an individual's starting value. A participant had to meet both conditions to count as a responder.

The study also included a group that received relugolix with placebo for 12 weeks, then relugolix with estradiol and norethindrone acetate for 12 weeks. That delayed co-administration group was part of the trial, but the registry record states that the prespecified primary efficacy analyses compared the full-duration combination group with placebo. The primary result therefore addresses that comparison, not a comparison involving the delayed-treatment group.

The registry record describes women with heavy menstrual bleeding associated with uterine fibroids as the study population. Further detail about eligibility characteristics or the extent to which participants represent other populations is not reported in the registry record.

3. Trial Design

LIBERTY 2 is recorded as a completed Phase 3, randomized, parallel-group trial with triple masking and placebo control. The official title describes an international study and a double-blind design; the registry design fields report triple masking. Enrollment was 382. The study evaluated relugolix co-administered with and without low-dose estradiol and norethindrone acetate in women with heavy menstrual bleeding associated with uterine fibroids.

Three groups are described. Group A received relugolix with estradiol and norethindrone acetate for 24 weeks. Group B received relugolix with matching placebo for 12 weeks, followed by relugolix with estradiol and norethindrone acetate for 12 weeks. Group C received matching placebos for 24 weeks. This structure allowed the study to include a delayed co-administration regimen as well as a continuous placebo comparison.

The primary efficacy comparison was Group A versus Group C. The registry specifies that only those groups are presented for the primary analysis because that was the study's prespecified primary efficacy comparison. The primary analysis population is described as all randomized participants who received any amount of study drug. The arm-specific numbers contributing to the primary endpoint were 125 for Group A and 129 for Group C.

The primary timeframe extended from baseline to the last 35 days of treatment, up to Week 24. Blood loss was assessed from returned feminine products collected at clinical visits using the alkaline hematin method. The original sample-size justification, detailed allocation procedures, and fuller descriptions of the masking procedures are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: A total of 382 premenopausal women aged 18 to 50 years old (inclusive) with heavy menstrual bleeding (≥ 160 milliliters \[mL\] during 1 cycle or ≥ 80 mL per cycle for 2 menstrual cycles as measured by the alkaline hematin method) associated with uterine fibroids were randomized. One participant was randomized in error before eligibility confirmed.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All randomized participants who received any amount of study drug.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: All participants who were enrolled and randomized in the study.

6. Primary Endpoint

The registry lists the primary outcome as Percentage Of Participants Who Achieved A Menstrual Blood Loss (MBL) Volume Of < 80 mL And A ≥ 50% Reduction From Baseline MBL Volume With Relugolix Plus E2/NETA, measured over From Baseline up to the last 35 days of treatment (up to Week 24).

The primary endpoint is the percentage of participants whose menstrual blood loss volume was below 80 mL and whose volume had fallen by at least 50% from baseline. Both requirements had to be satisfied over the final 35-day treatment interval, up to Week 24, for a participant to be classified as a responder. The measurement used returned feminine products collected at clinical visits and analyzed by the alkaline hematin method.

This is a binary responder endpoint: each participant is counted as either meeting or not meeting the response definition. The resulting arm-level percentage gives a direct account of how frequently the combined threshold was attained in each treatment group. It is not a measure of the average change in blood loss, and it does not describe how far above or below the threshold participants who did not respond were.

Combining an absolute volume threshold with a relative reduction makes the endpoint depend both on the final measured volume and on change from baseline. The registry record uses this outcome as the primary measure of efficacy and presents a 95% confidence interval for each arm. The endpoint is specific to the defined measurement interval and response criteria; it should not be interpreted as a summary of every aspect of bleeding, symptoms, or longer-term outcomes.

7. Results Posted in the Registry

• Relugolix Plus E2/NETA (Group A): 71.2% (89/125) versus 14.7% (19/129) on control; risk difference 56.5 percentage points (95% CI 46.5 to 66.5), risk ratio 4.83 (3.14–7.43); unadjusted two-proportion z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Percentage Of Participants With Amenorrhea Over The Last 35 Days Of Treatment

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Relugolix Plus E2/NETA (Group A): 50.4% (63/125) versus 3.1% (4/129) on control; risk difference 47.3 percentage points (95% CI 38.0 to 56.6), risk ratio 16.25 (6.10–43.32); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Relugolix Plus E2/NETA (Group A) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage Of Participants With A Hemoglobin Level ≤ 10.5 g/dL At Baseline Who Achieved An Increase Of > 2 g/dL From Baseline At Week 24

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Relugolix Plus E2/NETA (Group A): 61.3% (19/31) versus 5.4% (2/37) on control; risk difference 55.9 percentage points (95% CI 37.3 to 74.5), risk ratio 11.34 (2.86–44.92); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Relugolix Plus E2/NETA (Group A) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage Of Participants With A Maximum Numerical Rating Scale (NRS) Score ≤ 1 For Uterine Fibroid-Associated Pain Over The Last 35 Days Of Treatment

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Relugolix Plus E2/NETA (Group A): 47.1% (32/68) versus 17.1% (14/82) on control; risk difference 30.0 percentage points (95% CI 15.6 to 44.4), risk ratio 2.76 (1.61–4.73); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Relugolix Plus E2/NETA (Group A) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted numeric secondary comparisons cover several related but distinct outcomes: amenorrhea during the last 35 days of treatment, hemoglobin improvement among participants with low baseline hemoglobin, and a low maximum numerical rating scale score for fibroid-associated pain. For each of these comparisons, the reported treatment difference is positive and the sponsor analysis reports p < 0.001. These outcomes offer additional perspectives on bleeding, blood-related status, and pain, but they do not redefine the primary responder endpoint.

The remaining listed secondary outcomes include change in menstrual blood loss volume, primary fibroid volume, uterine volume, a quality-of-life bleeding and pelvic discomfort score, and bone mineral density. Their numerical results are not reported in the registry record information used here, so no conclusion about their magnitude or direction can be drawn from the available result summaries.

Secondary outcomes can help contextualize a primary result by assessing related outcomes and other aspects of the study. They are not necessarily interchangeable with the primary endpoint: amenorrhea is a different bleeding outcome, and pain or hemoglobin outcomes reflect different measures and analysis populations. Their relationship to the primary result is therefore supportive context rather than a direct replication of the primary test.

The registry lists 43 posted secondary outcomes and 27 analyses overall. When many outcomes are examined, the chance of finding apparently compelling results by chance can increase. The registry record information does not report a multiplicity adjustment for the secondary comparisons. Their unadjusted p-values therefore carry less confirmatory weight than a prespecified primary analysis with a defined testing strategy. They should be read as evidence about the specific outcomes and analyses reported, with attention to the broader set of outcomes examined.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Percentage Of Participants With Amenorrhea Over The Last 35 Days Of Treatment
• Percent Change From Baseline At Week 24 In MBL Volume
• Percentage Of Participants With A Hemoglobin Level ≤ 10.5 g/dL At Baseline Who Achieved An Increase Of > 2 g/dL From Baseline At Week 24
• Percentage Of Participants With A Maximum Numerical Rating Scale (NRS) Score ≤ 1 For Uterine Fibroid-Associated Pain Over The Last 35 Days Of Treatment
• Percent Change From Baseline At Week 24 In Primary Uterine Fibroid Volume
• Percent Change From Baseline At Week 24 In Uterine Volume
• Change From Baseline At Week 24 In UFS-QoL Bleeding And Pelvic Discomfort Scale Score As Measured By The UFS-QoL (Q1, Q2, Q5)
• Percent Change From Baseline At Week 12 In Bone Mineral Density At The Lumbar Spine (L1 to L4) As Assessed By DXA

9. Statistical Interpretation

10. Statistical Methods Explained

How is a responder percentage formed?

The endpoint is binary because each participant either satisfies both response criteria or does not. The responder percentage is the count meeting the definition relative to the number assessed in the treatment group. It provides an interpretable summary of the proportion reaching the stated threshold, but it hides variation within the nonresponder group. A participant who narrowly misses one condition and a participant whose result is far from both conditions are both classified as nonresponders.

The observed percentages are 71.2% for relugolix plus estradiol/norethindrone acetate and 14.73% for placebo. The arm-specific confidence intervals describe uncertainty in those estimated proportions. Their widths reflect, among other things, the number of participants and the spread of binary outcomes. They are not ranges containing a fixed share of individual participant responses.

What does the risk difference express?

The risk difference compares the two responder proportions by subtracting the comparator proportion from the intervention proportion. The posted unadjusted recalculation is 0.565, with a 95% confidence interval from 0.465 to 0.665. This form of contrast is on a proportion scale. The registry's reported treatment difference is 56.47 with a 95% confidence interval from 46.45 to 66.49, which expresses the contrast in percentage points.

These estimates are not interchangeable with a relative measure such as a risk ratio or odds ratio. A difference focuses on the absolute separation between observed proportions, while relative measures compare them multiplicatively or through odds. The primary sponsor-reported result is the Cochran-Mantel-Haenszel treatment difference, so that analysis is the central registry-reported comparison.

Why is the reported analysis different from the simple recalculation?

The recalculation is an unadjusted calculation based on posted arm summaries. The registry reports a Cochran-Mantel-Haenszel analysis. In general, this method can combine comparisons across strata while accounting for the stratification used in an analysis. The registry entries available for these results do not identify the strata or give further model specifications, so the precise implementation cannot be reconstructed from the result alone.

The reported analysis and the simple recalculation are close in their stated treatment-difference values when expressed in their respective scales, and the registry record does not flag a discrepancy. The sponsor's result includes its stated confidence interval and p-value and should be distinguished from the unadjusted recalculation rather than treated as if it were produced by the same method.

What does a confidence interval convey?

A confidence interval gives a range of values reasonably compatible with the data and the statistical method used, under its assumptions. It reflects sampling uncertainty in an estimate; it does not include every possible source of uncertainty, such as limitations in endpoint measurement or incomplete outcome data. A narrower interval indicates greater statistical precision for that estimate than a wider interval, all else being equal.

For this binary endpoint, the arm-level intervals describe the responder percentages, while the interval around the treatment difference describes uncertainty in the contrast. Whether an interval includes a null value is one way to assess compatibility with no difference, but the interval's range and the clinical meaning of the endpoint also matter. The registry reports p-values for the treatment comparisons, but p-values do not replace the effect estimate or its interval.

Why can over-dispersion matter?

General explanation: over-dispersion occurs when observed variability is greater than a statistical model expects. If a method assumes less variability than is present, its standard errors may be too small and its intervals too narrow. That can make results appear more precise than warranted. For a binary responder outcome, the possibility of variation across centers or participant subgroups can matter to model choice, but the registry record does not report an over-dispersion assessment or a specific adjustment for it. This general issue should not be mistaken for evidence that over-dispersion occurred in this trial.

What can the p-values tell the reader?

The reported primary p-value is < 0.001 for the Cochran-Mantel-Haenszel analysis; the unadjusted recalculation reports p < 0.001. A p-value describes how unusual data at least this discrepant would be under a stated null model and its assumptions. It is not the probability that the null hypothesis is true, does not describe the size of the treatment difference, and does not establish whether the result generalizes beyond the analyzed participants.

The statistical plan in full, including the treatment of missing values, covariate or stratum details, and any multiplicity procedure, is not reported in the registry record. Those omissions limit how completely the reported analysis can be independently assessed. The calculations should therefore be interpreted as summaries tied to the methods and populations actually identified in the registry, not as a substitute for a complete protocol or statistical analysis plan.

11. Confidence Intervals and Precision

The primary arm estimates have 95% confidence intervals of 62.42% to 78.95% for relugolix plus estradiol/norethindrone acetate and 9.11% to 22.04% for placebo. The reported difference has a 95% confidence interval from 46.45 to 66.49 on the percentage-point scale. The recalculated unadjusted risk difference has a 95% confidence interval from 0.465 to 0.665 on the proportion scale.

These intervals show that the observed percentages and contrast are estimates rather than exact population values. The primary contrast interval spans values above zero, and its lower and upper limits convey a range of plausible treatment differences under the analysis assumptions. The width gives a useful sense of precision, but it does not capture every limitation: it does not quantify possible selection effects, measurement issues, missing-data bias, or uncertainty about how well the trial population represents another setting.

Confidence intervals are also not prediction intervals for an individual participant. The arm-level intervals refer to estimated group percentages. The contrast interval refers to the estimated difference between the groups. The registry gives 95% intervals for these results, but the method for each arm-level interval is not specified in the record information used. Accordingly, the reported limits are best interpreted as the record's stated uncertainty summaries, not as a complete description of all sources of uncertainty.

The subgroup-defined hemoglobin outcome has smaller analyzed groups than the primary endpoint, with 31 participants in the intervention group and 37 in placebo. Its arm-level intervals are correspondingly broad: 42.19% to 78.15% and 0.66% to 18.19%. This illustrates why subgroup results may be less precise than results based on the full primary analysis groups. The interval width should be considered alongside the outcome definition and the number of participants contributing data.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning illustrations express sample requirements as participants per arm for binary outcomes. They are not total event counts. The calculations use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. They use the observed effect as a reference and calculate planning figures for detecting the stated fraction of that effect. The trial's original sample-size justification and design assumptions are not reported in the registry record.

The 80% power column means the estimated number of participants per arm under which the calculation gives an 80% probability of detecting the specified effect, assuming the planning model is correct. The 90% power column represents a higher target probability of detecting that effect and generally requires more participants. These are planning quantities, not observed trial results. They do not describe the probability that the primary finding is true.

Rows based on 100% of the observed effect use the observed contrast as the planning target. Rows based on 75% or 50% use smaller target effects. Smaller effects are harder to distinguish from random variation, so the planning requirement rises as the target effect gets smaller. The figures are conditional on the calculation assumptions and are not an assertion that the trial used those assumptions to select its enrollment.

These sample-size estimates are not observed power. Observed power is not a useful substitute for examining the estimate and its confidence interval: after a trial, the confidence interval directly shows the precision of the estimated effect. The planning illustrations are most useful for understanding how the required per-arm sample size changes when the target effect or desired power changes, not for rating the trial after its results are known.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.147; treatment proportion p₁ = 0.712; average p̄ = 0.430.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7001 = 1.3721.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5750 = 0.4840.
4. n = (1.3721 + 0.4840)² / (0.5647)² = 10.8, rounded up to 11 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.147286821705426                          # control proportion
p1 <- 0.712                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(3103087)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 11 (80% power), 14 (90% power)
 75% of observed effect: n per arm = 19 (80% power), 25 (90% power)
 50% of observed effect: n per arm = 40 (80% power), 52 (90% power)
power.prop.test: 10.8 
Simulated power at n = 11 per arm: 0.821 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry's own primary analysis uses the Cochran-Mantel-Haenszel method and reports a treatment difference of 56.47, with a 95% confidence interval from 46.45 to 66.49 and p < 0.001. The unadjusted recalculation from the posted summary statistics reports a risk difference of 0.565, with a 95% confidence interval from 0.465 to 0.665 and p < 0.001.

The difference in displayed scales is important: the registry result is stated in percentage points, while the recalculation is stated as a proportion. The estimates describe the same direction and closely aligned magnitude when read on their respective scales. The registry record does not identify a discrepancy between the sponsor analysis and the recalculation.

The recalculation is useful as a transparent check based on the posted arm-level summaries, but it is not a replacement for the sponsor's specified analysis. The registry's Cochran-Mantel-Haenszel result takes precedence as the study's reported analysis because it is the sponsor-reported method and estimate. Its full implementation cannot be independently reproduced from the registry result alone because further analysis details are not reported in the registry record.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when statistical tests are conducted for multiple outcomes, comparisons, timepoints, or subgroups. Each test can produce a false-positive finding even when there is no true effect. As the number of tests increases, the overall chance of at least one false-positive result can rise unless the analysis plan controls that risk. This is a general statistical concern, not a claim that any particular finding in LIBERTY 2 is false.

The registry lists 44 outcome measures, 43 posted secondary outcomes, and 27 analyses. It identifies a prespecified primary efficacy comparison between relugolix plus estradiol/norethindrone acetate and placebo. The multiplicity strategy, including whether a hierarchy, adjustment, or other testing procedure was used for secondary outcomes, is not reported in the registry record. Thus, the secondary p-values should not be treated as if they were necessarily adjusted or independently confirmatory.

The primary endpoint population is described as all randomized participants who received any amount of study drug. The record does not provide the detailed missing-data method for participants without an evaluable Week 24 or early-termination measurement. It also does not state how missing values were classified for the responder endpoint, whether imputation was used, or how sensitivity analyses addressed missingness. These details matter because the binary classification depends on the endpoint measurement over a specific collection interval.

The registry reports arm denominators for the primary endpoint and for the posted secondary comparisons, but those counts alone do not explain why participants did or did not contribute data. The record information also does not give full details on the analysis population for every outcome, deviations from assigned treatment, or protocol exclusions. These omissions constrain interpretation of possible missing-data or analysis-population effects. They do not establish that such problems occurred; they indicate that their handling cannot be evaluated from the registry record.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: cholecystitis (0.8%), cholecystitis acute (0.8%), intervertebral disc degeneration (0.8%), intervertebral disc protrusion (0.8%), radius fracture (0.8%).

The safety record gives serious adverse event counts for each treatment group. One participant in Group A was reported as affected among 126 at risk; two in the delayed co-administration group were affected among 126; and four in the placebo group were affected among 129. No deaths were reported in any group. These are participant counts in the registry's safety summary, not estimates of a causal treatment effect.

The record also gives counts for other adverse events: 31 among 126 in Group A, 67 among 126 in Group B, and 42 among 129 in Group C. The registry's summary lists cholecystitis, acute cholecystitis, intervertebral disc degeneration, intervertebral disc protrusion, radius fracture, and road traffic accident among the top serious adverse event terms, each reported as 0.8%. It does not give group-specific counts for those listed terms in the information used.

Adverse event counts alone do not show whether an event was caused by study treatment, how severe or prolonged it was, or whether treatment was stopped. The serious-event figures should be read as descriptive safety reporting over the trial, not as a complete comparison of all possible harms or as a basis for individual medical decisions.

17. What the Result Does, and Does Not, Mean

The primary result describes the proportion of participants who met a specific combined menstrual blood loss response definition during the final treatment interval. In the registry's primary comparison, the observed responder percentage was higher in the relugolix plus estradiol/norethindrone acetate group than in placebo. The confidence intervals and reported treatment difference quantify uncertainty in that group-level comparison under the analyses used.

This result does not mean that every participant in the intervention group met the response definition, or that every participant in placebo failed to meet it. It does not describe the size of change among each participant, the time course of response, or outcomes after the stated treatment period. It also does not answer every question about symptoms, fibroid size, quality of life, or safety.

The endpoint combines a final blood loss threshold with a relative reduction from baseline. Its interpretation is therefore tied to that definition and the measured 35-day interval. A participant who fails either component is not a responder under the endpoint, even if another aspect of bleeding changed. The reported comparison should not be generalized automatically to different treatment regimens, populations, durations, or endpoints.

Statistical significance is not the same as clinical importance. The p-value describes compatibility with a null hypothesis under the analysis assumptions, while the effect estimate and confidence interval describe the magnitude and precision of the observed contrast. The registry record does not provide a threshold for determining what magnitude of difference is clinically important, nor does it report full details on missing data and multiplicity.

The results are trial evidence about the analyzed groups and should be interpreted with the design, endpoint, analysis methods, and reporting limits in view. They do not by themselves establish long-term outcomes or individual treatment choices. No medical advice follows from the statistical summaries alone.

18. Why This Trial Matters Statistically

This trial illustrates how a clinical study can define response using both an absolute threshold and a change from baseline. That design creates an endpoint that is straightforward to classify, but it also makes the result dependent on both the measurement interval and the starting value. Readers need the operational definition, not only a percentage, to understand what a responder represents.

The registry also makes the distinction between an arm-specific percentage and a treatment contrast important. The arm percentages show the observed response in each group; the treatment difference directly compares those proportions. Confidence intervals for both kinds of estimates help communicate precision and keep attention on the range of values compatible with the data rather than a single point estimate.

LIBERTY 2 additionally demonstrates why the sponsor's reported analysis should be distinguished from an unadjusted recalculation. The registry reports a Cochran-Mantel-Haenszel treatment difference, while the recalculation uses the posted arm summaries. Their close alignment is informative, but the methods are not identical and the full sponsor analysis cannot be reconstructed from the registry summary alone.

Finally, the trial has multiple secondary outcomes and analyses. Their breadth can provide useful context, yet it raises questions about multiplicity, reporting, and the weight to give individual secondary p-values. The registry record identifies the primary comparison but does not report the multiplicity procedure, detailed missing-data approach, or original sample-size justification. Statistical interpretation therefore depends both on what the results say and on what methodological details remain unavailable.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record provides a concise summary rather than a full protocol or statistical analysis plan. It does not report the trial's original sample-size justification, detailed allocation procedures, full handling of missing endpoint measurements, or the specific strata and implementation details for the Cochran-Mantel-Haenszel analysis. These omissions limit independent reconstruction of the sponsor's analysis and assessment of how all design decisions were made.

The primary analysis population is defined as randomized participants who received any amount of study drug, but the record information does not explain the status of participants without usable measurements or the rules used to classify them. The reported arm denominators describe the analyzed numbers but do not reveal reasons for missing data or establish whether missingness could have affected the result.

The registry lists many outcome measures and analyses, yet the multiplicity procedure is not reported. This is particularly relevant when interpreting secondary p-values, which may not have the same confirmatory status as the prespecified primary comparison. Numerical results for several listed secondary outcomes are not available in the result information, preventing evaluation of their direction and precision.

The primary comparison presented is Group A versus Group C. Although Group B is described in the trial design and safety results, no primary efficacy comparison with that delayed regimen is reported in the record information used. The findings therefore cannot be used to infer an efficacy contrast for Group B.

Finally, the primary endpoint is a specific blood loss responder measure assessed through returned feminine products over a defined interval. It does not capture every dimension of heavy menstrual bleeding or establish outcomes beyond the study timeframe.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: LIBERTY 2: Efficacy & Safety Study of Relugolix in Women With Heavy Menstrual Bleeding Associated With Uterine Fibroids — NCT03103087.
• Registry record last updated 2022-04-20; status last verified 2021-09.
• Results first posted 2022-04-20.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

LIBERTY 2 (NCT03103087) was a completed Phase 3 randomized, parallel-group, placebo-controlled study with triple masking and enrollment of 382. It evaluated relugolix with estradiol and norethindrone acetate, including a delayed co-administration group, in women with heavy menstrual bleeding associated with uterine fibroids.

The primary efficacy comparison was continuous relugolix plus estradiol/norethindrone acetate versus placebo. The endpoint was the percentage meeting both the menstrual blood loss volume threshold below 80 mL and a reduction of at least 50% from baseline during the last 35 days of treatment, up to Week 24. The registry-reported Cochran-Mantel-Haenszel treatment difference was 56.47, with a 95% confidence interval from 46.45 to 66.49 and p < 0.001.

The registry also reports secondary outcomes and safety counts. Details on the original sample-size justification, multiplicity procedure, and full missing-data approach are not reported in the registry record.