Evaluation of Efficacy, Safety and Tolerability of Aldafermin in a Phase 2b, Randomized, Double-blind, Placebo-controlled, Multi-center Study In Subjects With Nonalcoholic Steatohepatitis and Stage 2/3 Fibrosis (ALPINE 2/3): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Evaluation of Efficacy, Safety and Tolerability of Aldafermin in a Phase 2b, Randomized, Double-blind, Placebo-controlled, Multi-center Study In Subjects With Nonalcoholic Steatohepatitis and Stage 2/3 Fibrosis (ALPINE 2/3) (NCT03912532), a randomized phase 2 study. The primary outcome is Liver Fibrosis Response After Administration With Aldafermin in Participants With Nonalcoholic Steatohepatitis and Stage 2/3 Fibrosis.

2. Clinical Question

3. Trial Design

This was a randomized, parallel-group, Phase 2b study of aldafermin, identified as NGM282 in the official title, versus placebo. The registry describes the trial as double-blind and placebo-controlled, and records masking as quadruple. It was conducted at multiple centers. The record lists enrollment of 171 participants and a completed study. The listed start date is 2019-05-16, and the primary completion date is 2021-03-04.

Participants were assigned to aldafermin 0.3 mg, 1.0 mg, or 3.0 mg, or to placebo. All four arm descriptions specify subcutaneous injection. The treatment period described in the study summary and official title was 24 weeks. Results are posted for 43 participants in the 0.3 mg group, 42 in the 1.0 mg group, 43 in the 3.0 mg group, and 43 in the placebo group. The primary outcome population is described as the Intent-to-Treat Analysis Set, although the registry record does not provide further details about its membership or handling of missing measurements.

The primary outcome is described as liver fibrosis response assessed using biopsies at baseline and Week 24, read by a central pathologist using NASH CRN criteria. The record also lists secondary outcomes concerning liver fat content and liver fat normalization or response. Three secondary outcomes are posted, while the registry lists five outcome measures overall. The complete outcome definitions and analysis details beyond the listed descriptions are not reported in the registry record.

The record does not report the trial’s original sample-size justification, planned allocation ratio, interim analyses, or full statistical analysis plan. These omissions limit assessment of whether the enrollment and analysis choices followed a prespecified design or how the posted comparisons fit into the full trial plan.

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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Liver fibrosis response was assessed in the Intent-to-Treat Analysis Set.

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: Baseline characteristics were assessed in the Intent-to-Treat Analysis Set.

6. Primary Endpoint

The registry lists the primary outcome as Liver Fibrosis Response After Administration With Aldafermin in Participants With Nonalcoholic Steatohepatitis and Stage 2/3 Fibrosis, measured over Baseline up to Week 24.

The primary outcome is titled as liver fibrosis response after aldafermin administration in participants with NASH and stage 2/3 fibrosis. Liver biopsies were collected at baseline and Week 24 and read centrally using NASH CRN criteria. Those criteria define fibrosis stages from F0 through F4, with F0 indicating no fibrosis and F4 indicating cirrhosis. The registry defines response as improvement in fibrosis by at least one stage, with no worsening of steatohepatitis, at Week 24 relative to baseline.

Biopsy-based assessment is relevant to this endpoint because the outcome definition depends on histologic fibrosis stage and steatohepatitis status. Central reading is part of the registry’s stated measurement process. In general, using a consistent reading framework can help make assessments across participants more comparable, though the record does not describe reader agreement or procedures for resolving disagreements.

A key reporting issue is that the posted primary results are means on a score scale, accompanied by standard errors. They are not presented as counts or proportions of participants who met the registry’s stated response definition. The scale’s precise construction and the relationship between its mean and the binary response criterion are not reported in the registry record. As a result, the posted means should not be interpreted as response rates, and the response definition alone does not clarify what a difference in the reported mean represents.

7. Results Posted in the Registry

• Aldafermin 0.3 mg: mean 0.24 versus 0.23 on control; difference 0.01 (95% CI -0.13 to 0.15), standardized effect d = 0.03; z-test p = 0.890.
• Aldafermin 1.0 mg: mean 0.25 versus 0.23 on control; difference 0.02 (95% CI -0.12 to 0.16), standardized effect d = 0.06; z-test p = 0.782.
• Aldafermin 3.0 mg: mean 0.27 versus 0.23 on control; difference 0.04 (95% CI -0.14 to 0.22), standardized effect d = 0.09; z-test p = 0.664.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary results are reported as means for a continuous score, with standard errors. A mean describes the average value in a group; a standard error describes uncertainty in the estimated group mean, rather than the spread of individual participant values. The reported standard errors are 0.04 for the 0.3 mg and 1.0 mg groups, 0.07 for the 3.0 mg group, and 0.06 for placebo. The registry record does not report standard deviations for the endpoint alongside these means.

What do the pairwise comparisons estimate?

Each recalculated contrast compares one aldafermin group with placebo. The posted mean differences are 0.01 for 0.3 mg, 0.02 for 1.0 mg, and 0.04 for 3.0 mg. These are unadjusted differences formed from the group summaries. The corresponding confidence intervals are recalculated from posted summary statistics; they are not identified as estimates from the sponsor’s multiple-comparison procedure. Each interval includes zero, so the summaries are compatible with a range of possible differences, including no difference, under this interval calculation.

The recalculated unadjusted p-values are 0.89, 0.782, and 0.664 for the respective dose comparisons. A p-value, in general, describes how incompatible the observed data are with a specified null model, assuming that model and the calculation are appropriate. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. These p-values are unadjusted and should not be mistaken for a multiplicity-adjusted result.

Why does multiplicity matter?

The study compares three dose groups with a shared placebo group, creating multiple opportunities to identify a difference. In general, when several hypotheses are tested, the chance of at least one apparently notable finding can increase unless the analysis accounts for the family of comparisons. The sponsor’s posted analysis is described as a “Multiple comparison procedure step” and gives p = 0.553 across the four groups. The record does not name the procedure in greater detail, specify its testing sequence or adjustment method, or provide adjusted pairwise estimates. The unadjusted pairwise calculations and sponsor-reported procedure therefore answer different inferential questions and should remain distinct.

What does the dispersion information mean?

Standard errors quantify uncertainty in estimated means, not variability among individual observations. In general, the standard deviation describes participant-level spread, while the standard error is influenced by both spread and sample size. Reconstructing variability from a standard error requires assumptions and adequate information about the underlying analysis. The recalculated comparisons use the posted summaries and do not report an approximate standard deviation, but the record does not give all details of the analysis model or the endpoint’s score construction. A reader should not treat these reconstructed contrasts as a substitute for a complete prespecified model.

Over-dispersion is a general concern when observed variability is larger than a model assumes. In general, failing to allow for over-dispersion can make standard errors too small and confidence intervals too narrow. The registry record does not report a specific over-dispersion assessment for this continuous endpoint, and there is not enough information to claim that over-dispersion was present or absent.

How should the score be understood?

The outcome description defines fibrosis response using a stage change and absence of worsening steatohepatitis, while the reported result is a mean on a score scale. General statistical interpretation depends on the outcome’s measurement scale and the meaning of a unit change. The registry record does not specify how the mean score was constructed from the biopsy assessment or how it maps to the response definition. Consequently, the reported mean differences can be compared numerically across groups, but a clinical interpretation of the score-unit difference is not supported by the information in the record.

The primary population is identified as the Intent-to-Treat Analysis Set. In general, an intention-to-treat approach aims to analyze participants according to randomized assignment, preserving the comparison created by randomization. The record does not describe the exact analysis-set rules, whether all randomized participants contributed outcome data, or how missing biopsies or other missing measurements were handled. The meaning of the label should therefore not be expanded beyond what the registry states.

Finally, the sponsor’s reported p-value is the posted inferential result from a multiple-comparison procedure, whereas the recalculated means, mean differences, confidence intervals, and p-values are simple unadjusted calculations from the posted summaries. The record marks no discrepancy between these results. The recalculated values are useful for understanding the size and precision of the pairwise contrasts, but they do not supply missing protocol details or replace the sponsor’s reported procedure.

10. Confidence Intervals and Precision

The recalculated confidence intervals for all three dose-versus-placebo mean differences include zero. This means the posted summaries do not pin down a single direction for each contrast: values compatible with the intervals include a difference favoring either group as well as no difference. The intervals are broad relative to the point estimates, so the estimated differences are not precise enough to settle their direction or magnitude. No confidence interval for the sponsor’s multiple-comparison result is reported in the registry record.

Precision is not the same as statistical significance or clinical importance. In general, a confidence interval gives a range of values compatible with the data and the method used to calculate it, under the method’s assumptions. A wide interval indicates greater uncertainty; an interval including zero does not show that the true difference is exactly zero. The observed mean differences—0.01, 0.02, and 0.04—are point estimates, and should be read together with their uncertainty rather than in isolation.

The group standard errors are 0.04, 0.04, 0.07, and 0.06 for the 0.3 mg, 1.0 mg, 3.0 mg, and placebo groups, respectively. These summarize uncertainty around group means, not the spread of individual scores. The registry does not report a fuller explanation of the score’s scale or a sponsor-reported confidence interval for the primary comparison procedure. Thus, precision can be discussed for the recalculated pairwise mean differences, but it cannot resolve the uncertainty about how those mean differences map to the registry’s response definition.

The confidence intervals are based on the posted summary statistics and are unadjusted pairwise intervals. They should not be interpreted as simultaneous intervals covering all dose comparisons with a stated family-wide confidence level. The record does not provide adjusted confidence intervals for the multiple-comparison procedure.

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 sample-size illustrations are planning calculations based on the observed effect, not the trial’s own design justification. For this continuous endpoint, the entries are participants per arm, rather than total events. 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 original design assumptions and the registry’s sample-size justification are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm for a design with an 80% chance of detecting the specified effect under those assumptions, if that effect and the model assumptions were correct. The 90% power column gives the corresponding number for a 90% chance. These are prospective planning quantities, not descriptions of the probability that this completed trial found a result.

The entries consider planning for 100%, 75%, and 50% of the observed effect. The estimates increase as the effect being targeted becomes smaller: a weaker contrast is harder to distinguish from sampling variation and generally requires more participants. The same planning logic applies to both power columns, with the higher power target requiring more participants for a given assumed effect. The calculated values can become very large when the observed contrast is small, so they are sensitive to the effect estimate used as their starting point.

These figures are not observed power. They are illustrations tied to the observed effect and stated planning assumptions; they do not show what the study’s actual power was, nor do they establish whether the original design was adequate. The registry record does not report the trial’s own 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 = 0.010; pooled standard deviation σ = 0.334.
2. n = 2 × (1.96 + 0.842)² × 0.334² / 0.010² = 17,550.1, rounded up to 17,551 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 <- 0.00999999999999998                      # observed difference in means
sigma <- 0.33436506994601                     # 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(3912532)
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 = 17,551 (80% power), 23,495 (90% power)
 75% of observed effect: n per arm = 31,201 (80% power), 41,769 (90% power)
 50% of observed effect: n per arm = 70,201 (80% power), 93,979 (90% power)
power.t.test: 17551.1 
Simulated power at n = 17,551 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. Analyses Reported by the Sponsor

The sponsor-reported analysis is described as a “Multiple comparison procedure step” and gives p = 0.553 for the four groups. The registry does not report an estimate, confidence interval, or further explanation of the procedure’s exact form. This is the sponsor’s posted inferential result and is distinct from the three recalculated unadjusted dose-versus-placebo comparisons.

The recalculated mean differences are 0.01 for 0.3 mg, 0.02 for 1.0 mg, and 0.04 for 3.0 mg, with unadjusted p-values of 0.89, 0.782, and 0.664. Those calculations use posted group summaries and do not account for the multiplicity procedure described in the sponsor analysis. They provide a direct numerical comparison with placebo, but should not be presented as the sponsor’s adjusted analysis.

The registry record indicates no discrepancy between the posted results and the recalculated summaries. That does not make the procedures interchangeable: one is a sponsor-reported multiple-comparison result, while the others are unadjusted pairwise calculations. The record does not provide enough detail to reconstruct the sponsor procedure or determine its adjusted dose-specific estimates.

14. Multiplicity, Missing Data and Analysis Populations

Several dose groups are compared with placebo, so multiplicity is relevant to interpretation. In general, evaluating several hypotheses can raise the chance of a false-positive result across the family of tests if no adjustment is made. The registry reports one sponsor analysis labeled “Multiple comparison procedure step,” with p = 0.553 for the four groups. It does not report the precise procedure, adjustment strategy, testing order, or adjusted dose-specific estimates. The recalculated pairwise p-values are explicitly unadjusted and cannot be treated as the multiplicity-controlled sponsor result.

The outcome was assessed in the Intent-to-Treat Analysis Set, but the registry record does not give detailed analysis-set rules. It does not state how missing biopsies, incomplete Week 24 assessments, withdrawals, or other missing outcome values were handled. The posted group sizes are available, but those figures alone do not explain whether all randomized participants had endpoint measurements or whether the analyzed population differed from enrollment. Missing-data assumptions and sensitivity analyses are not reported in the registry record.

The record also does not report the original statistical analysis plan, a prespecified hierarchy of outcomes, or whether interim analyses were planned or conducted. The primary outcome is described as a response based on biopsy-stage improvement without worsening steatohepatitis, while posted results are means with standard errors. The record does not clarify how this difference in presentation relates to the planned endpoint analysis. It lists three secondary outcomes and five outcome measures overall, but does not give a complete multiplicity strategy across primary and secondary outcomes.

These missing details do not invalidate the posted numerical results, but they constrain how fully the inferential framework can be evaluated. In particular, the unadjusted recalculations should be interpreted as descriptive pairwise contrasts rather than as a complete account of the trial’s planned testing strategy.

15. 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: suicide attempt (2.4%), cardiac hypertrophy (2.4%), obesity (2.4%), small intestinal obstruction (2.4%), osteoarthritis (2.4%).

The serious adverse event counts were 1 among 43 participants in the aldafermin 0.3 mg group, 4 among 41 in the 1.0 mg group, 1 among 43 in the 3.0 mg group, and 3 among 43 in the placebo group. The safety denominators differ from the primary outcome group sizes for the 1.0 mg group, so the posted safety figures should be read with their own denominators. The registry also reports deaths of 0, 1, 0, and 0 in those groups, respectively.

These are descriptive counts. They do not, by themselves, establish that an event was caused by treatment or allow a causal comparison between groups. The registry lists several serious adverse event terms with percentages, including suicide attempt, cardiac hypertrophy, obesity, small intestinal obstruction, osteoarthritis, and uterine cancer. Those terms are not mapped to treatment groups in the listed summary, so their group-specific distribution cannot be inferred from the record.

The registry’s “other” adverse event counts are 30 among 43, 34 among 41, 38 among 43, and 36 among 43 for the respective groups. The record does not provide enough detail in these summary fields to characterize event severity, timing, recurrence, or relatedness. Safety counts should therefore be interpreted as reported event summaries rather than as a complete assessment of comparative risk.

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

The posted primary results show mean scores of 0.24, 0.25, and 0.27 for the three aldafermin doses, compared with 0.23 for placebo. Recalculated unadjusted mean differences versus placebo are 0.01, 0.02, and 0.04. The confidence intervals for these contrasts all include zero, and the corresponding unadjusted p-values are 0.89, 0.782, and 0.664. These results describe the observed group summaries and the uncertainty around the pairwise mean differences; they do not establish that the groups are identical or that a potentially meaningful difference is impossible.

The endpoint requires particular caution. The registry defines a fibrosis response in terms of improvement by at least one histologic stage without worsening of steatohepatitis, but reports primary results as means on a score scale. The record does not explain how that score is constructed or how its mean relates to the response definition. Therefore, the means cannot be read as the number or proportion of participants who met the response criterion, and the numerical contrasts cannot be translated into response rates from the posted information.

The sponsor’s posted multiple-comparison result is p = 0.553. It is a different analysis from the recalculated unadjusted pairwise contrasts, and the record does not provide enough detail to recover adjusted estimates or confidence intervals. Neither that p-value nor the unadjusted p-values quantify clinical importance. Statistical results need to be read alongside endpoint definition, precision, and the analysis plan, some of which are not described in the registry record.

The findings are limited to this randomized Phase 2b study, its recorded treatment groups, and the 24-week assessment period. They do not determine what results would occur in other populations or over other durations. They also do not provide medical guidance. The registry record does not report a complete account of missing-data handling, the original sample-size justification, or the full statistical analysis plan, which further limits conclusions about the design and robustness of the analysis.

17. Why This Trial Matters Statistically

This trial offers an instructive example of why the endpoint definition and the reported statistical measure must be read together. The primary outcome description defines a biopsy-based response using change in fibrosis stage and steatohepatitis status. Yet the posted results are means and standard errors on a score scale. Without an explanation of how that scale represents the response definition, readers can compare the group means numerically but cannot interpret them as response proportions or translate a score difference into a clinical response.

The trial also illustrates the distinction between a sponsor-reported multiple-comparison result and unadjusted pairwise calculations. The sponsor reports a multiple-comparison procedure result for the four groups, while the recalculated estimates compare each dose with placebo without adjustment. These results have different inferential roles. Treating them as interchangeable would obscure how multiple testing was handled and what each p-value addresses.

Precision is another central statistical issue. The recalculated differences are small in the units reported, and their confidence intervals include zero. This combination does not establish absence of an effect; it indicates that the estimated contrasts remain uncertain. Planning illustrations based on the observed effects further demonstrate, in general, how a small target difference can imply large sample-size requirements, while also depending strongly on the chosen effect and assumptions.

Finally, the record shows the value of reporting analysis details. The analysis population is named, but missing-data methods, the original sample-size rationale, the full multiplicity strategy, and the precise relationship between endpoint definition and score are not reported. These are not merely technical details: they shape how confidently readers can interpret the estimate, its precision, and the scope of inference.

18. 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.

Several limits follow from the way the primary outcome is recorded. The definition describes a biopsy response based on improvement in fibrosis stage without worsening steatohepatitis, while the posted primary results are mean scores with standard errors. The scale construction and the mapping between a mean score and the defined response are not reported in the registry record. This prevents interpreting the posted means as response rates or assigning a clear clinical meaning to a score-unit contrast.

The record does not provide the original sample-size justification, full statistical analysis plan, or detailed procedure underlying the sponsor’s multiple-comparison result. It also does not report missing-data rules, sensitivity analyses, or exact analysis-set membership. Although the primary population is named as the Intent-to-Treat Analysis Set, the record does not explain whether all randomized participants contributed outcome measurements or how unavailable biopsies were addressed.

The sponsor result is reported as a p-value without an accompanying estimate or confidence interval, and the procedure is identified only as a multiple-comparison step. The recalculated pairwise intervals and p-values are unadjusted. They help describe the posted group summaries but cannot substitute for the full sponsor analysis or establish how its multiplicity procedure affected dose-specific inference.

The posted safety summary supplies event counts and some serious-event terms, but does not provide group mappings for those terms in the listed summary or full details on timing, severity, or relatedness. The outcome description concerns assessment through Week 24, and the record does not support broader conclusions beyond that study period. These limitations affect interpretation of the registry record; they do not, on their own, show that the trial conduct or results were invalid.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Evaluation of Efficacy, Safety and Tolerability of Aldafermin in a Phase 2b, Randomized, Double-blind, Placebo-controlled, Multi-center Study In Subjects With Nonalcoholic Steatohepatitis and Stage 2/3 Fibrosis (ALPINE 2/3) — NCT03912532.
• Registry record last updated 2025-07-03; status last verified 2025-03.
• Results first posted 2025-07-03.
• 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.

22. Record Summary

The registry record describes a completed, randomized, double-blind, placebo-controlled, multicenter Phase 2b study of aldafermin in participants with histologically confirmed NASH and stage 2/3 fibrosis. Enrollment was 171, and treatment was administered by subcutaneous injection for 24 weeks across three aldafermin dose groups and a placebo group.

The primary outcome concerns liver fibrosis at baseline and Week 24. Posted group results are means with standard errors; recalculated unadjusted dose-versus-placebo differences are 0.01, 0.02, and 0.04, and all corresponding confidence intervals include zero. The sponsor-reported multiple-comparison result is p = 0.553. The record does not explain how the mean score relates to its stated biopsy-response definition or report the full analysis plan, sample-size rationale, and missing-data methods.