Safety, Tolerability, and Efficacy of Monotherapy and Combination Regimens in Participants With Nonalcoholic Steatohepatitis (NASH): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety, Tolerability, and Efficacy of Monotherapy and Combination Regimens in Participants With Nonalcoholic Steatohepatitis (NASH) (NCT03987074), a randomized phase 2 study. The primary outcome is Percentage of Participants Experiencing Treatment-Emergent Adverse Events (TEAEs).

2. Clinical Question

3. Trial Design

The registry record describes NCT03987074 as a randomized, open-label, parallel-group phase 2 study of monotherapy and combination regimens in participants with NASH. The recorded enrollment is 109. The study status is completed; the recorded start date is 2019-07-29 and the primary completion date is 2020-07-13. Gilead Sciences is listed as the sponsor.

The regimens included semaglutide alone and combinations of semaglutide with firsocostat 20 mg, cilofexor 30 mg, cilofexor 100 mg, or firsocostat 20 mg plus cilofexor 30 mg. Semaglutide was dose-escalated from 0.24 mg to 2.4 mg over 16 weeks. Each arm description specifies treatment for 24 weeks. Because masking was none, participants and study personnel were not blinded according to the registry record.

The primary objective was to evaluate safety and tolerability. The primary outcome was the percentage experiencing TEAEs during the specified follow-up period. The safety groups reported for this outcome contain 21.0 or 22.0 participants per arm. Those arm-level counts are not the same item as the overall enrollment recorded by the registry. The record does not report the original sample-size justification, the allocation procedure in greater detail, or whether any interim analyses were planned. These design details should not be inferred from the posted results.

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: 209 participants were screened. 109 participants were enrolled. 1 participant was enrolled but was not randomized and was not included in the analysis.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Safety Analysis Set included participants who received at least 1 dose 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: The Safety Analysis Set included participants who received at least 1 dose of study drug.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants Experiencing Treatment-Emergent Adverse Events (TEAEs), measured over First dose date up to Week 24 plus 30 days.

The primary endpoint is the percentage of participants experiencing treatment-emergent adverse events (TEAEs), measured from the first dose through Week 24 plus 30 days. A TEAE was defined as an adverse event with onset on or after the study drug start date and no later than 30 days after permanent discontinuation, or an adverse event leading to premature discontinuation of study drug. Adverse events were assessed according to Common Terminology Criteria for Adverse Events (CTCAE) version 5.0.

This endpoint summarizes whether participants experienced at least one event meeting the registry definition during the observation period. It is used to describe the safety and tolerability experience across treatment regimens. It does not by itself describe how many events each participant experienced, how long events lasted, or their severity distribution. Those details are not reported in the primary endpoint summary. The denominator is the Safety Analysis Set: participants who received at least 1 dose of study drug.

The endpoint is a percentage rather than a continuous measurement. Its arm-level value therefore reflects the proportion of the safety group with a TEAE, not an average event count. The registry record also lists serious adverse events separately. Serious events are not interchangeable with the broader TEAE outcome: an event may be treatment-emergent without being classified as serious.

7. Results Posted in the Registry

• Semaglutide + Firsocostat 20 mg: 86.4% (19/22) versus 81.0% (17/21) on control; risk difference 5.4 percentage points (95% CI -16.7 to 27.5), risk ratio 1.07 (0.82–1.39); unadjusted two-proportion z-test p = 0.631.
• Semaglutide + Cilofexor 30 mg: 81.8% (18/22) versus 81.0% (17/21) on control; risk difference 0.9 percentage points (95% CI -22.4 to 24.1), risk ratio 1.01 (0.76–1.35); unadjusted two-proportion z-test p = 0.942.
• Semaglutide + Cilofexor 100 mg: 72.7% (16/22) versus 81.0% (17/21) on control; risk difference -8.2 percentage points (95% CI -33.3 to 16.8), risk ratio 0.90 (0.65–1.25); unadjusted two-proportion z-test p = 0.523.
• Semaglutide + Firsocostat 20 mg + Cilofexor 30 mg: 90.5% (19/21) versus 81.0% (17/21) on control; risk difference 9.5 percentage points (95% CI -11.4 to 30.5), risk ratio 1.12 (0.87–1.43); unadjusted two-proportion z-test p = 0.378.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: a participant either experienced a TEAE meeting the stated definition during follow-up or did not. The reported arm percentages express the number of participants with an event as a fraction of the safety group. This is different from an event rate based on person-time or a count of repeated events. The registry record identifies the Safety Analysis Set as participants who received at least 1 dose of study drug.

What does a risk difference represent?

A risk difference compares the event proportions in two arms by subtracting the comparator-arm proportion from the other arm’s proportion. A positive estimate indicates a higher observed event proportion in the combination arm; a negative estimate indicates a lower observed proportion. The posted comparisons use semaglutide alone as the comparator. The estimates are 0.054 for semaglutide plus firsocostat 20 mg, 0.009 for semaglutide plus cilofexor 30 mg, -0.082 for semaglutide plus cilofexor 100 mg, and 0.095 for semaglutide plus firsocostat 20 mg plus cilofexor 30 mg. These estimates are on the proportion scale and should be considered alongside their confidence intervals.

Why do the confidence intervals matter?

A confidence interval conveys uncertainty around an estimated difference. Narrower intervals indicate greater precision, while wider intervals indicate less precision. These comparisons have intervals that extend below and above zero. Thus, the posted data do not pin down the direction of the underlying difference. The interval is not a statement that every value within it is equally likely, and it does not account for any analysis choices that were not reported in the registry record.

How should the p values be read?

The registry record gives unadjusted p values of 0.631, 0.942, 0.523, and 0.378 for the four comparisons, respectively. A p value describes how compatible the observed data are with a specified null model under the assumptions of the test. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of a difference. The record does not report the statistical test details or a multiplicity strategy. The posted p values should therefore be read as unadjusted comparisons, not as a complete account of confirmatory inference.

Why can sparse outcomes be uncertain?

When arm sizes are limited, a small change in the number of participants with an event can have a noticeable effect on the observed percentage. This can leave estimates imprecise even when the percentages appear separated. General statistical principle: binary outcomes with uncommon events may yield especially unstable estimates in small groups. Here, the overall TEAE proportions are not uncommon, but the arm groups remain small and the confidence intervals are broad. The record does not report a model-based adjustment, stratification factors, or other methods that might affect precision.

What is not specified?

The registry record does not report the original statistical analysis plan, the exact test used for the posted comparisons, missing-data procedures, or whether analyses were adjusted for baseline characteristics. It also does not report an efficacy analysis among the posted results. No inference about those choices should be made from the numerical summaries alone. The registry record identifies randomized allocation and no masking, but does not give further detail on sequence generation or concealment.

10. Confidence Intervals and Precision

The risk-difference confidence intervals are broad in relation to their point estimates. For semaglutide plus firsocostat 20 mg versus semaglutide alone, the estimate was 0.054 and the interval ranged from -0.167 to 0.275. For semaglutide plus cilofexor 30 mg, the estimate was 0.009 and the interval ranged from -0.224 to 0.241. For semaglutide plus cilofexor 100 mg, the estimate was -0.082 and the interval ranged from -0.333 to 0.168. For semaglutide plus firsocostat 20 mg plus cilofexor 30 mg, the estimate was 0.095 and the interval ranged from -0.114 to 0.305.

All four intervals include zero. Each therefore includes values consistent with no difference, while also extending to differences in either direction. The data are not precise enough to establish a clear direction for any comparison. This does not mean the regimens are equivalent: demonstrating equivalence would require a prespecified margin and an analysis designed for that question, neither of which is reported in the registry record.

Confidence intervals are useful because they pair an estimate with its sampling uncertainty. They do not describe all sources of uncertainty, such as unreported analytic choices or missing-data handling. With arm sizes of 21.0 or 22.0 in the endpoint results, chance variation can materially affect percentages and comparative estimates. The registry record does not report confidence intervals for the arm percentages themselves. The displayed uncertainty applies to risk differences, not to every possible interpretation of the safety data.

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 figures are participants per arm, not total events. The calculation is a planning illustration based on the observed effect, not the trial’s own original design assumptions. The registry record’s original sample-size justification is not reported in the registry record. The listed planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect equal to the stated fraction of the observed effect.

The column for 80% power describes the participants per arm estimated to give an 80% chance of detecting the specified effect under those planning assumptions, if the effect and model assumptions hold. The 90% power column uses a higher target chance of detection and therefore requires more participants per arm for the same assumed effect. These are design-planning quantities; this is not observed power.

Rows describe planning for 100%, 75%, or 50% of the observed effect. A smaller assumed effect is harder to distinguish from random variation, so it requires more participants. Because the observed differences are uncertain and the confidence intervals are broad, these figures should be read as illustrations of how sample-size requirements depend on the assumed effect, not as a recommended design or a guarantee of future precision. The sample-size curve represents that relationship across effect assumptions. It does not add evidence about the treatment comparisons or establish what the completed study was designed to detect.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.810; treatment proportion p₁ = 0.864; average p̄ = 0.837.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5229 = 1.0249.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5215 = 0.4389.
4. n = (1.0249 + 0.4389)² / (0.0541)² = 731.7, rounded up to 732 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 <- 0.80952380952381                          # control proportion
p1 <- 0.863636363636364                          # 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(3987074)
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 = 732 (80% power), 980 (90% power)
 75% of observed effect: n per arm = 1,345 (80% power), 1,800 (90% power)
 50% of observed effect: n per arm = 3,122 (80% power), 4,179 (90% power)
power.prop.test: 731.7 
Simulated power at n = 732 per arm: 0.804 (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.

13. Multiplicity, Missing Data and Analysis Populations

The registry record posts a primary safety outcome and four unadjusted comparisons of combination regimens with semaglutide alone. The record does not report a multiplicity adjustment or a prespecified hierarchy for interpreting these comparisons. When multiple comparisons are made, the chance of at least one apparently unusual result can increase; general statistical practice may address this with an adjustment or a testing hierarchy. Whether such a plan existed for this study is not reported in the registry record. The reported p values should be understood as unadjusted rather than as multiplicity-controlled results.

The registry record lists 2 outcome measures, but no secondary outcomes are posted in the information analyzed here, and the reported count of analyses is 0. The record therefore does not give a broader set of secondary results with which to assess consistency across outcomes. The trial title refers to safety, tolerability, and efficacy, but efficacy results are not among the posted outcomes analyzed. No efficacy conclusion can be drawn from the safety percentages.

The primary outcome population is specified as the Safety Analysis Set, including participants who received at least 1 dose. The registry record does not report how missing outcome data were handled, whether participants with incomplete follow-up were included in a particular way, or whether sensitivity analyses were conducted. It also does not report an intention-to-treat analysis population or a per-protocol population. Randomized allocation is recorded, but the published safety endpoint uses the stated treated-participant population. This distinction matters because a safety population defined by treatment received is not automatically the same as every participant randomized.

Additional analysis details—including covariate adjustment, subgroup analyses, interim monitoring, and the original statistical analysis plan—are not reported in the registry record. Their absence from the posted record does not establish that they were or were not used. Interpretation should therefore remain anchored to the observed arm-level proportions and the explicitly unadjusted risk-difference comparisons.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: diarrhoea (4.8%), vomiting (4.8%), pancreatitis (4.5%).

The registry record separately reports participants with serious adverse events and deaths by arm. Serious adverse events were reported for 1 participant receiving semaglutide alone, 0 receiving semaglutide plus firsocostat 20 mg, 0 receiving semaglutide plus cilofexor 30 mg, 1 receiving semaglutide plus cilofexor 100 mg, and 0 receiving semaglutide plus firsocostat 20 mg plus cilofexor 30 mg. The corresponding at-risk group sizes were 21, 22, 22, 22, and 21. Deaths were reported as 0 in every arm.

These counts are a separate safety summary and should not be confused with the broader TEAE percentages. The registry record also lists Diarrhoea and Vomiting at 4.8% and Pancreatitis at 4.5% among the top serious adverse event terms. Those terms are not assigned to individual arms in the posted summary. The counts are descriptive; with small groups and few serious events, they do not establish that the regimens have the same or different serious-event risk. Further detail on event timing, severity grades, relatedness, and management is not reported in the registry record.

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

The results describe the proportion of participants in each treatment arm who experienced a TEAE during the registry-defined follow-up. The observed percentages ranged from 72.7% with semaglutide plus cilofexor 100 mg to 90.5% with semaglutide plus firsocostat 20 mg plus cilofexor 30 mg; the semaglutide-alone group was 81.0%. The comparisons with semaglutide alone have broad confidence intervals that all include zero.

These data do not establish that a combination regimen increases or decreases TEAEs relative to semaglutide alone. They also do not establish equivalence or rule out potentially important differences. The study’s small arm groups leave substantial uncertainty, and the registry record does not report a multiplicity adjustment or details of missing-data handling. The unadjusted p values are not a substitute for considering the size and precision of the estimated differences.

The endpoint covers TEAEs, not all aspects of tolerability or safety. The separate serious adverse event counts and the absence of reported deaths provide additional descriptive information, but do not resolve uncertainty about uncommon outcomes. The registry record does not provide enough detail to characterize all event severity, duration, relatedness, or clinical consequences.

The study title includes efficacy, but the posted results described for this analysis concern safety. They do not establish efficacy, comparative benefit, or an overall benefit-harm balance. The findings should be interpreted as a snapshot of the recorded safety outcomes in the study population and follow-up period, not as a basis for treatment decisions.

16. Why This Trial Matters Statistically

This study illustrates a common challenge in early clinical research: several regimens can be compared, but limited group sizes leave the estimates uncertain. Arm-level TEAE percentages may look meaningfully separated, yet the risk-difference confidence intervals extend across zero and into both positive and negative values. Statistical interpretation therefore depends not only on the observed percentages but also on the uncertainty around their differences.

The randomized, parallel-group design offers a structured comparison of combination regimens with semaglutide alone. However, open-label conduct means treatment assignment was not masked, and the registry record does not report further details about outcome assessment or analysis procedures. The primary outcome is a broad participant-level measure of any qualifying TEAE, while serious events are listed separately. These measures answer related but distinct safety questions.

The record also demonstrates why planning illustrations should not be confused with observed evidence. Sample-size calculations based on the observed effect can show how much larger a future study might need to be under specified assumptions, but they do not alter the precision of the completed study’s estimates. In this case, the broad intervals and unadjusted comparisons are central to interpreting the findings. The multiplicity plan, missing-data methods, and original sample-size rationale are not reported in the registry record, limiting assessment of the full analysis strategy.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The main limitation is imprecision: each treatment arm has 21.0 or 22.0 participants in the primary endpoint results, and the risk-difference confidence intervals are broad. The registry record does not report confidence intervals for the arm-specific TEAE percentages, limiting direct assessment of their precision. The comparisons are unadjusted, and no multiplicity strategy is reported despite several combination arms being compared with the same comparator.

The study was open-label, with no masking recorded. The registry record does not report missing-data methods, details of the statistical analysis plan, or whether outcome assessment involved procedures to reduce the influence of knowledge of treatment assignment. It does not report the original sample-size justification or interim analysis plan. Those omissions limit evaluation of the prespecified approach and how closely the posted results reflect it.

The primary endpoint captures whether a participant had a qualifying TEAE, but the posted summary does not provide a full account of event duration, severity, relatedness, or recurrent events. The registry record’s serious-event summary is also sparse, with few reported serious events and no deaths in the arms. Finally, efficacy outcomes are not posted among the results. These limitations mean that safety percentages alone cannot support a comprehensive comparison of the regimens.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Safety, Tolerability, and Efficacy of Monotherapy and Combination Regimens in Participants With Nonalcoholic Steatohepatitis (NASH) — NCT03987074.
• Registry record last updated 2021-07-15; status last verified 2021-06.
• Results first posted 2021-07-15.
• 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

NCT03987074 is a completed randomized, open-label, parallel-group phase 2 study sponsored by Gilead Sciences. It enrolled participants with NASH and compared semaglutide alone with four semaglutide-based combination regimens. The primary objective was safety and tolerability, measured in the Safety Analysis Set as the percentage with a TEAE from first dose through Week 24 plus 30 days.

TEAE percentages ranged from 72.7% to 90.5% across the arms, with 81.0% in the semaglutide-alone group. The unadjusted risk-difference confidence intervals for the combination comparisons all included zero and were broad. The registry record reports few serious adverse events and 0 deaths in each arm. Original design assumptions, multiplicity handling, and missing-data methods are not reported in the registry record.