Multicenter Study in Postmenopausal Women With Osteoporosis, ALVOBOND: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Multicenter Study in Postmenopausal Women With Osteoporosis, ALVOBOND (NCT05395091), a randomized phase 3 study. The primary outcome is Percent Change From Baseline in LS BMD at Month 12 to Demonstrate Comparable Efficacy of AVT03 and Prolia®..

2. Clinical Question

The trial’s stated aim was to compare efficacy, safety, immunogenicity, pharmacodynamic profiles, and pharmacokinetic profiles of AVT03 and US-Prolia in postmenopausal women with osteoporosis. The primary endpoint was percent change from baseline in lumbar spine bone mineral density at Month 12. That endpoint addresses a measured change in bone density; it is not itself a fracture outcome.

The registry also lists secondary outcomes concerning bone mineral density at the hip and femoral neck, new morphometric vertebral fractures, a bone-turnover marker, adverse events, injection-site reactions, routine safety parameters, and anti-drug and neutralizing antibodies. The record lists 11 outcome measures and 9 posted secondary outcomes. These additional measures broaden the types of efficacy and safety information recorded, but their detailed results are not included in the primary comparison.

The central statistical question is whether the observed group estimates and the uncertainty around their difference are compatible with the trial’s stated goal of demonstrating comparable efficacy. The registry record does not report a comparability margin or the detailed decision rule used for that purpose. The reported estimate and confidence interval can therefore be described, but they do not alone establish the trial’s formal conclusion about comparability.

3. Trial Design

ALVOBOND (NCT05395091) is recorded as a phase 3, randomized, quadruple-masked, parallel-design study. Its official title describes a repeat-dose, two-arm, multicenter comparison of AVT03 and US-Prolia in postmenopausal women with osteoporosis. The listed sponsor is Alvotech Swiss AG. The registry reports enrollment of 532 and a status of completed.

The study began on 2022-08-23, and the recorded primary completion date is 2024-05-07. These dates describe the registry record; they do not supply details about the conduct of visits, the timing of database lock, or the timing of analysis. Those details are not reported in the registry record.

The two primary groups are labeled AVT03 and Prolia in the posted results. The primary endpoint population is described as evaluable subjects with data at Month 12, and the reported group counts for that endpoint are 239.0 for AVT03 and 237.0 for Prolia. The safety record uses groups with 266 at risk in each treatment group for the Baseline to Month 12 period. These are different reported analysis populations and should not be treated as interchangeable.

The safety record also lists Month 12 to Month18 groups labeled AVT03, Prolia/AVT03, and Prolia/Prolia. These labels indicate that the registry reports additional follow-up groups, including groups with different treatment sequences. The record does not provide enough detail to infer the full sequence allocation, the protocol’s rationale for these groups, or how the follow-up data were incorporated into the primary comparison. The original design assumptions, including any sample-size justification, 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.

Baseline to Month 12

Reasons for not completing this period:

Month 12 to Month 18

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The number of evaluable subjects with data at Month 12 per group.

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.

6. Primary Endpoint

The registry lists the primary outcome as Percent Change From Baseline in LS BMD at Month 12 to Demonstrate Comparable Efficacy of AVT03 and Prolia®., measured over Baseline to Month 12.

The primary endpoint is percent change from baseline in lumbar spine bone mineral density (LS BMD) at Month 12. It describes the relative change in a measured bone-density quantity between the starting measurement and the Month 12 measurement. The registry specifies the timeframe as baseline to Month 12 and reports the outcome as a least-squares mean with a standard error.

Bone mineral density is a continuous measurement, so its change can be summarized with a group mean and a comparison of group means. The lumbar spine endpoint gives a common, numerical basis for comparing the groups at the specified follow-up time. The trial description states that the comparison was intended to demonstrate comparable efficacy of AVT03 and Prolia.

The endpoint is not the same as a clinical fracture outcome. The registry lists incidence of new morphometric vertebral fractures among secondary outcomes, but the primary result presented is a change in LS BMD. A result for a continuous surrogate measurement should be interpreted as evidence about that measurement, rather than as a direct estimate of fracture risk or of every aspect of clinical benefit.

The primary endpoint estimate is based on evaluable subjects with Month 12 data in each group. How evaluability was defined, how missing Month 12 measurements were handled, and which covariates or model terms were used to obtain the least-squares means are not reported in the registry record. These details matter when interpreting how the analyzed observations represent the randomized groups.

7. Results Posted in the Registry

• Prolia: mean 5.18 versus 5.30 on control; difference -0.12 (95% CI -2.53 to 2.29), standardized effect d = -0.01; z-test p = 0.922.

8. Statistical Interpretation

9. Statistical Methods Explained

The posted primary outcome is a continuous measure, reported as a least-squares mean percent change from baseline in LS BMD. For a continuous endpoint, the mean describes the average observed change in a group, while a between-group mean difference expresses the estimated contrast between group averages. The posted values are 5.3 for AVT03 and 5.18 for Prolia; the reported contrast is -0.12 for Prolia relative to AVT03.

What does a least-squares mean represent?

A least-squares mean is generally a model-based group mean. Depending on the model, it can reflect adjustment for specified covariates or other design features. It is not necessarily identical to a simple arithmetic average of all recorded individual measurements. The registry identifies the outcome summary as a least-squares mean, but the model specification, covariates, and estimation method are not reported in the registry record. It is therefore not possible to determine from the record which adjustments, if any, produced these estimates.

How should the standard errors be read?

The group results include standard errors of 0.87 for AVT03 and 0.872 for Prolia. A standard error describes the sampling uncertainty of an estimated mean: in general, a smaller standard error indicates a more precisely estimated mean, while a larger one indicates more uncertainty. It is not the same as the spread of individual participant measurements. The record does not report the individual-level distribution of LS BMD changes.

The posted comparison also gives a pooled standard deviation of 13.44 and a standardized effect estimate of -0.009. The pooled standard deviation is an estimate of the outcome’s variability used in the recalculation. Standardizing the difference expresses it relative to that variability. These are recalculated, unadjusted quantities based on the posted summary statistics, rather than a substitute for a sponsor’s full analysis. The registry record indicates no discrepancy between the reported comparison and the recalculation.

What does the confidence interval add?

A confidence interval gives a range of values compatible with the data under the statistical procedure used to form it. Repeated use of a procedure with the stated confidence level would, under its assumptions, produce intervals that cover the target quantity at the stated rate. It is not a range containing a known proportion of individual outcomes, nor does it assign a probability to the fixed true difference after the interval has been calculated.

For the reported Prolia-minus-AVT03 difference, the interval extends from -2.534 to 2.294. Its inclusion of zero indicates that the data do not distinguish a zero difference from differences in either direction at the conventional two-sided significance level used for this interval. Whether that interval is sufficiently narrow to support comparability depends on a prespecified clinical margin. Such a margin is not reported in the registry record.

What does the p-value test?

The reported unadjusted p-value is 0.922. In general, a p-value quantifies how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the test assumptions held. It does not measure the size or clinical importance of an effect. It also does not show that the null hypothesis is true. The record does not report the detailed test specification or whether any multiplicity adjustment was planned, so the p-value should be read as the posted unadjusted result rather than as a complete account of the analysis strategy.

Why do analysis populations matter?

The endpoint description refers to evaluable subjects with Month 12 data. The safety record instead gives at-risk populations, with 266 in each group for Baseline to Month 12. These denominators answer different questions and arise from different reported populations. A randomized comparison can be affected by which participants are included in an analysis, particularly if outcome availability is related to treatment or outcome. The registry record does not report the criteria for evaluability, a missing-data method, or a primary intention-to-treat analysis population.

In general, missing-data procedures make assumptions about the unobserved measurements, and different assumptions can lead to different estimates. A complete interpretation of the primary analysis would require the protocol or statistical analysis plan to describe those assumptions, the model, and any sensitivity analyses. Those details are not reported in the registry record. The posted result is therefore best understood as the estimate for the reported evaluable population, with the available uncertainty, rather than as a full reconstruction of every planned analysis.

10. Confidence Intervals and Precision

The primary comparison reports a mean difference of -0.12, with a 95% confidence interval from -2.534 to 2.294. This interval is the main direct description of precision for the between-group contrast: it shows that the estimate is uncertain across a range extending on both sides of zero. The group standard errors, 0.87 for AVT03 and 0.872 for Prolia, describe uncertainty around the reported group means, not the spread of individual responses.

A confidence interval is more informative than the point estimate alone because it indicates which differences remain compatible with the data under the method used. The interval does not establish that the true difference is zero. It also does not, by itself, establish that possible differences are acceptably small. For that conclusion, the interval must be assessed against a clinically justified margin specified before analysis. The registry record does not report such a margin.

Precision is tied to the analysis population and model. The endpoint population is described as evaluable subjects with Month 12 data, and the record gives group counts of 239.0 and 237.0. The safety denominators are different. Because the registry does not report the full analysis model, the handling of missing observations, or sensitivity analyses, the posted interval cannot be assessed against those unreported choices.

The recalculated comparison is an unadjusted calculation from posted summary statistics. It complements the reported values but does not replace a prespecified analysis. The registry record indicates no discrepancy. Overall, the confidence interval gives a clear account of the uncertainty in the posted mean difference, while the unreported margin and analysis details limit what can be concluded about comparability.

11. 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 required participants per arm, not total events. They use the observed effect as a starting point and consider planning for 100% of that effect, 75% of the observed effect, and 50% of the observed effect. For each effect assumption, the displayed columns describe planning for 80% power and 90% power: in general terms, power is the chance that a planned study will detect an effect of the specified size under the planning assumptions, if that effect is truly present.

The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. The required participants per arm for 80% power are 196,829 when planning for 100% of the observed effect, 349,918 when planning for 75%, and 787,315 when planning for 50%. For 90% power, the corresponding figures are 263,498, 468,441, and 1,053,991.

Smaller effects are harder to distinguish from random variation, so planning to detect a smaller effect requires more participants. The large planning figures reflect the very small observed standardized effect used in this illustration. These calculations are planning illustrations, not observed power for ALVOBOND. They do not describe the trial’s original design: the registry record does not report the original sample-size assumptions or justification.

12. 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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.120; pooled standard deviation σ = 13.437.
2. n = 2 × (1.96 + 0.842)² × 13.437² / -0.120² = 196,828.7, rounded up to 196,829 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
delta <- -0.12                      # observed difference in means
sigma <- 13.4371316013086                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(5395091)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 196,829 (80% power), 263,498 (90% power)
 75% of observed effect: n per arm = 349,918 (80% power), 468,441 (90% power)
 50% of observed effect: n per arm = 787,315 (80% power), 1,053,991 (90% power)
power.t.test: 196829.6 
Simulated power at n = 196,829 per arm: 0.800 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists 11 outcome measures and 9 posted secondary outcomes. The secondary outcomes span additional bone-density measures, vertebral fracture incidence, a bone-turnover marker, adverse events, injection-site reactions, routine safety measures, and antibody outcomes. Considering several outcomes can increase the chance of at least one apparently notable result if each is tested separately. In general, multiplicity procedures can control that risk, but the registry record does not report a testing hierarchy, adjustment method, or confirmatory status for the secondary analyses.

The primary comparison has a posted unadjusted p-value of 0.922. Whether that value was intended to be interpreted without adjustment, or within a broader testing strategy, is not reported in the registry record. The registry reports 0 analyses for this endpoint record, so detailed analysis outputs beyond the posted summary comparison are not available in the record.

The endpoint population is defined as evaluable subjects with data at Month 12. The record does not report the exact evaluability criteria, the number or reasons for missing endpoint observations, or how missing measurements were handled. It also does not state whether the primary analysis used an intention-to-treat population, a per-protocol population, or another formally defined set beyond the description of evaluable subjects.

These omissions matter because population definitions and missing-data assumptions can influence a treatment comparison. General approaches include analyzing all randomized participants under a specified estimand, using a model that accommodates incomplete observations under stated assumptions, and conducting sensitivity analyses for alternative assumptions. The registry record does not report which approaches were planned or performed. It also does not report the detailed model, covariate adjustment, or the protocol’s analysis plan. Consequently, the posted estimate can be interpreted as the result for the stated evaluable population, but the robustness of that result to missingness, analysis choices, and multiplicity cannot be determined from the registry record.

14. 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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: iron deficiency anaemia (0.8%), cardiac failure congestive (0.8%), deafness bilateral (0.8%), cholecystitis (0.8%), appendicitis (0.8%).

For Baseline to Month 12, the registry reports serious adverse events in 9 of 266 at-risk participants in the AVT03 group and 10 of 266 in the Prolia group. The same groups each had 266 at risk for other adverse events, with 178 affected in each group. The record reports 3 deaths for AVT03 and 1 for Prolia during this period.

These are counts with their reported risk denominators; they are not estimates of a causal treatment effect. The registry also lists Month 12 to Month18 follow-up groups: serious events were reported in 6 of 242 at risk for AVT03, 3 of 122 for Prolia/AVT03, and 2 of 122 for Prolia/Prolia. Other adverse events affected 99 of 242, 39 of 122, and 39 of 122, respectively. Deaths in those groups were 1, 0, and 0.

Serious-event terms listed include iron deficiency anaemia, cardiac failure congestive, deafness bilateral, cholecystitis, appendicitis, and femur fracture, each reported as 0.8%. These summaries do not establish cause, comparative risk, or event timing beyond the reported periods. Further details about event adjudication and statistical comparisons are not reported in the registry record.

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

The posted primary result describes Month 12 change in LS BMD among evaluable participants with data at that time. The group least-squares means were 5.3 for AVT03 and 5.18 for Prolia, and the reported Prolia-minus-AVT03 difference was -0.12. The confidence interval from -2.534 to 2.294 conveys substantial uncertainty about the precise difference, including values on either side of zero.

The result does not prove that the treatments are identical. A failure to distinguish a difference from zero is not the same as demonstrating equivalence or comparability. A formal conclusion about comparability depends on the prespecified margin and analysis rule; neither is reported in the registry record. The estimate and interval are therefore evidence about the observed endpoint, not a complete reconstruction of the trial’s formal decision.

The primary outcome is a bone-density measure, not a direct measure of fractures or every dimension of clinical effect. Although the registry lists fracture, safety, immunogenicity, pharmacodynamic, and pharmacokinetic outcomes, the primary result does not resolve those questions. Nor does the endpoint alone describe longer-term outcomes beyond its baseline-to-Month 12 timeframe.

The result also applies to the reported evaluable population and the analysis as described in the registry. Details about the analysis model, missing data, and sensitivity analyses are not reported in the registry record. Those omissions limit assessment of how robust the estimate is to analytic choices and incomplete observations. The posted safety counts provide descriptive event information, but they do not by themselves establish that an event was caused by treatment or that the groups differ in safety.

In short, the registry reports similar point estimates for the primary measure, alongside a confidence interval that leaves uncertainty about the size of the difference. The record does not provide the additional prespecified criteria needed to translate this into a formal conclusion of comparable efficacy. Interpretation should stay within the endpoint, timeframe, population, and detail actually reported.

16. Why This Trial Matters Statistically

Comparative studies of a proposed biosimilar and a reference product require attention not only to whether a conventional test detects a difference, but also to how precisely the comparison excludes differences that would matter. ALVOBOND’s stated objective was to demonstrate comparable efficacy, making the confidence interval and any prespecified comparability margin central to interpretation. The registry reports the interval but does not report the margin or the full decision rule.

The continuous LS BMD endpoint allows the comparison to be expressed as a difference in group means, with standard errors and a confidence interval. This makes the estimated contrast and its uncertainty visible. At the same time, an endpoint result cannot stand in for every clinical and biological question in the study. The registry lists outcomes on fractures, safety, antibodies, and pharmacokinetic and pharmacodynamic profiles, which address different dimensions of comparison.

The record also illustrates why analysis populations matter. The primary endpoint is based on evaluable Month 12 data, while safety summaries use at-risk populations. Clear reporting of eligibility criteria, missing-data handling, analysis models, and sensitivity analyses helps readers judge whether the comparison is robust and how it relates to the randomized groups. These details are not reported in the registry record.

Statistically, the key lesson is that a near-zero point estimate and a large p-value do not by themselves establish similarity. The uncertainty interval, the clinical margin, multiplicity strategy, and analysis population all shape the meaning of a comparative result. The posted numbers support a focused description of the Month 12 LS BMD comparison, while the registry’s omissions limit broader conclusions.

17. 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.

The registry record does not report the comparability margin or the formal decision rule for the stated objective of demonstrating comparable efficacy. Without those criteria, the confidence interval cannot be assessed against the threshold that would define an acceptable difference. The record also does not describe the full statistical analysis plan, including model specification, covariate adjustment, or whether the reported test was the prespecified primary test.

The primary endpoint population is described as evaluable subjects with Month 12 data, but the criteria for evaluability and the handling of missing measurements are not reported in the registry record. The posted endpoint counts are not the same as the safety at-risk counts, and the record does not explain the difference. No sensitivity analyses or alternative analysis populations are described.

The registry lists multiple outcome measures and secondary outcomes, but it does not report a multiplicity strategy or the inferential status of those analyses. Detailed results for the listed secondary outcomes are not included in the primary comparison. This limits assessment of the wider efficacy and safety profile from this record.

The safety summaries provide counts by period and group, but the record does not give detailed event-level context, adjudication procedures, or statistical comparisons. The follow-up group labels include different treatment sequences, while the full protocol details and analysis of those groups are not reported. Finally, the original sample-size justification and design assumptions are not reported in the registry record; the planning illustrations are separate from the trial’s own design rationale.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Multicenter Study in Postmenopausal Women With Osteoporosis, ALVOBOND — NCT05395091.
• Registry record last updated 2025-05-29; status last verified 2025-05.
• Results first posted 2025-05-29.
• 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.

21. Record Summary

ALVOBOND (NCT05395091) is a completed, phase 3, randomized, quadruple-masked, parallel, multicenter study in postmenopausal women with osteoporosis. It compared AVT03, described as the proposed biosimilar for Prolia, with Prolia.

The primary endpoint was percent change from baseline in LS BMD at Month 12. The reported least-squares means were 5.3 for AVT03 and 5.18 for Prolia; the Prolia-minus-AVT03 difference was -0.12, with a 95% confidence interval from -2.534 to 2.294 and an unadjusted p-value of 0.922.

The posted estimates describe the primary endpoint but do not alone establish comparability. The registry record does not report a comparability margin, detailed analysis plan, or missing-data methods. Safety summaries include serious adverse events, other adverse events, and deaths for the reported follow-up groups.