DUR-928 in Subjects With SARS-CoV-2 With Acute Lung, Liver or Kidney Injury: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for DUR-928 in Subjects With SARS-CoV-2 With Acute Lung, Liver or Kidney Injury (NCT04447404), a randomized phase 2 study. The primary outcome is Composite Endpoint of Alive and Free of Organ Failure at Day 28.

2. Clinical Question

3. Trial Design

The registry describes a Phase 2, randomized, double-blind, placebo-controlled, parallel-group study. It lists quadruple masking. The available record does not identify which groups of people were masked, so the masking label should not be used to infer that detail. Randomization describes the intended assignment process; it does not guarantee that both arms will contain participants, as this record illustrates.

The study was sponsored by Durect. The registry lists a start date of 2020-09-21 and a primary completion date of 2021-03-04. Its status is terminated. The record says the study ended early with only 2 subjects enrolled, both assigned to active treatment. Thus, although the study included a placebo arm in its design, the realized enrollment left that arm empty.

The primary outcome was a participant count for a composite endpoint assessed at Day 28. The registry also lists secondary outcomes concerning survival, intensive-care status, and hospital status at specified follow-up times. It lists 4 outcome measures, while the posted information indicates that no analyses are available. The study’s protocol details, including the intended allocation ratio and the reason for termination, are not reported in the registry record.

These design features matter for interpretation. Randomization and masking are elements of the intended design, but they cannot overcome the absence of a comparator group in the enrolled sample. With only two enrolled participants, both in one arm, the record is descriptive of those participants and does not support a randomized estimate contrasting DUR-928 with placebo.

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: Study was terminated early with only 2 subjects enrolled, both assigned to active treatment. Both subject deaths were assessed as Not Related to study treatment occurring on Day 19 and Day 20 (15 and 16 days after last dose, respectively).

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: Study terminated early due to lack of enrollment.

6. Primary Endpoint

The registry lists the primary outcome as Composite Endpoint of Alive and Free of Organ Failure at Day 28, measured over Day 28.

The primary endpoint was the count of participants alive and free of organ failure at Day 28. The composite required participants to be free of mechanical ventilation, free of renal replacement therapy, and free of acute liver failure. A participant would therefore need to satisfy all components to meet the endpoint as described.

Composite endpoints are used in clinical research to represent a broader clinical state when several related outcomes are relevant. Combining components can capture more than one way a participant may experience serious illness or recovery. It can also make interpretation less straightforward: components may differ in importance, frequency, or clinical meaning. The registry description does not report a component-specific analysis or any weighting among the components.

The posted primary result is a count of 0.0 in the DUR-928 arm, where 2.0 participants were enrolled. The registry’s narrative says the study ended early and that both participants died on Day 19 and Day 20, before the Day 28 endpoint assessment. There were no placebo participants. Consequently, this endpoint result describes the enrolled DUR-928 arm but cannot establish how the outcome would compare with placebo. Details of endpoint adjudication and handling of incomplete or unavailable assessments are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: each participant either meets the composite endpoint or does not. The registry reports a count of participants, rather than a continuous measurement such as a laboratory value. For the DUR-928 arm, the posted count is 0.0 among 2.0 enrolled participants. The placebo arm has 0.0 participants, so there is no observed placebo outcome distribution to compare with the active-treatment outcome.

What a two-arm comparison would require

In general, a randomized comparison estimates how outcomes differ between participants assigned to the intervention and participants assigned to the comparator. For a binary endpoint, analysts might describe the event frequency in each arm and estimate a contrast between those frequencies. The contrast could be expressed using different statistical measures, depending on the analysis plan. Whatever measure is selected, it requires outcome information from both groups. Here, no placebo participants enrolled, and the registry indicates that the binary comparison is not estimable. A numerical treatment contrast should not be reconstructed from an empty comparator arm.

Why uncertainty matters

In general, an estimate summarizes the observed data, while a confidence interval communicates uncertainty around that estimate under the assumptions of the statistical method. A narrow interval would indicate greater precision than a wide interval, but neither an interval nor a treatment-effect estimate for the primary comparison is posted in this record. With no participants in one arm, the comparative uncertainty cannot be described as though both arms had contributed data. The absence of an interval is not a finding of no difference; it reflects that the comparison could not be estimated.

The registry lists no analyses. It therefore does not report a fitted model, an adjusted estimate, a test statistic, or a p-value for the primary comparison. None should be inferred from the arm counts. In particular, it would be inappropriate to treat the observed outcomes in the DUR-928 arm as if they were a direct estimate of the difference from placebo. The record also does not report whether any planned statistical method was changed after early termination.

Counts and the composite definition

Counts are straightforward to describe but may conceal detail. The primary count refers to participants who met the full composite definition, not to a separate count for each component. The registry description names mechanical ventilation, renal replacement therapy, and acute liver failure as components, but does not provide component-specific results. It also does not report further rules for how these components were assessed, how events were adjudicated, or how a participant with incomplete component information would be classified.

In general, a composite can increase the breadth of an endpoint by grouping related outcomes, but interpretation depends on the definition and the contribution of its components. A total count alone does not show which component determined a participant’s status. For this record, the reported deaths before Day 28 are relevant to the “alive” component, while separate component-level status for the other elements is not detailed in the posted primary result.

Planning illustrations are not observed evidence

Any sample-size calculation based on an observed effect is a planning illustration, not an estimate of trial power or a replacement for a valid treatment comparison. The registry record does not report a trial-specific sample-size justification. With no comparator participants and no estimable two-arm effect, a planning exercise based on an observed contrast would not establish what the study could have detected. In general, sample-size planning depends on assumptions about the outcome, the effect to detect, the statistical method, and the allocation between groups. Those assumptions should not be mistaken for observed trial findings.

The registry’s stated planning assumptions describe a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. Those are assumptions for an illustration, not evidence that the study enrolled according to that plan. The record does not report the intended sample size, the rationale for the design assumptions, or a trial-specific power calculation.

Overall, the statistical information supports a limited descriptive reading: the primary endpoint count in the enrolled DUR-928 arm is posted, and the study narrative reports early termination after two enrollments. It does not support a comparative efficacy estimate, a p-value, or a conclusion that the intervention and placebo have similar or different outcomes.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("DUR-928")
n      <- c(2)
events <- c(0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  DUR-928  2  0  0.0  0.0  84.2  0.0  65.8  0.8  70.8

11. Multiplicity, Missing Data and Analysis Populations

The registry lists secondary outcomes of being alive at Days 28 and 60, alive and out of the intensive care unit at Day 28, and alive and out of the hospital at Days 28 and 60. These are distinct outcomes and follow-up times. The record lists 4 outcome measures overall, but no analyses are posted. It does not report a prespecified strategy for controlling multiplicity, such as an order for testing outcomes or an adjustment for testing several outcomes. Without such details and without posted analyses, the secondary outcomes cannot be used to make inferential claims.

Multiplicity is a general statistical concern when several hypotheses are assessed: the chance of obtaining at least one apparently notable result can increase as more tests are considered. Methods for addressing this depend on the planned analysis. The registry record does not report which methods, if any, were planned. This absence should be described as missing information, not interpreted as proof that no plan existed.

The record also does not report a detailed missing-data strategy, rules for handling participants who do not complete follow-up, or a description of analysis populations such as intention-to-treat or per-protocol sets. The primary outcome information says the study was terminated early and only two subjects enrolled; both were assigned to active treatment. It does not provide a separate accounting of missing component assessments or follow-up data. Those details are not reported in the registry record.

Because the placebo arm had no participants and no analyses are posted, there is no basis for evaluating the consequences of missing outcomes within a two-arm analysis or for assessing whether an analysis population altered the result. Any account of imputation, exclusions, or sensitivity analyses would be speculative.

12. 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: respiratory failure (50.0%), brain injury (50.0%).

The adverse-event information lists 2 participants at risk and 2 affected by serious adverse events in the DUR-928 group. It also lists 2 participants at risk and 2 affected by other adverse events in that group, and 2 deaths. The placebo group has 0 participants at risk and no affected participants or deaths listed. These are counts from the enrolled groups, not a comparison of event rates between treatments.

The registry lists Respiratory failure at 50.0% and Brain Injury at 50.0% among the top serious-event terms. These percentages should be read as the values recorded for those terms; the available information does not give further detail about the denominator or event-level grouping. They should not be converted into additional counts or treated as estimates of risk in a broader population.

The registry narrative says both deaths were assessed as Not Related to study treatment. This is the attribution recorded for those cases. It does not by itself prove or disprove a causal relationship, and it does not provide a controlled safety comparison. With no placebo participants and only two participants enrolled overall, the adverse-event figures are descriptive and highly limited. They cannot establish how event frequencies would differ between DUR-928 and placebo.

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

The registry record documents a Phase 2 randomized study that was intended to evaluate DUR-928 against placebo in people with SARS-CoV-2 and acute lung, liver, or kidney injury. It reports an early-terminated study with two enrolled participants, both assigned to DUR-928. For the primary composite endpoint at Day 28, the recorded participant count in the DUR-928 arm is 0.0. The narrative reports that both participants died before Day 28.

These observations are not a treatment-effect estimate. The placebo arm enrolled no participants, so there is no observed comparator outcome. The posted results do not tell whether the primary endpoint would have occurred more or less often with DUR-928 than with placebo. They also do not establish that the arms would have had the same outcomes. The comparison is simply not estimable from this enrollment.

The record’s statements about deaths and adverse events describe what was recorded for the enrolled participants. The reported assessment that the deaths were Not Related to study treatment is an attribution in the record; it is not a randomized comparison and should not be treated as a general conclusion about safety. The adverse-event counts likewise cannot be compared meaningfully with a group containing no participants.

In general, small studies can be useful for generating information, but their results are uncertain and may not represent outcomes in a broader population. Here, the limited enrollment and absent comparator sharply constrain what can be learned about relative efficacy or safety. A lack of an estimable treatment contrast is not evidence of no effect, nor does it demonstrate an effect. It means the registry record does not contain the data needed to answer the comparative question.

The result is best understood as a record of a study that began but ended before a placebo-controlled analysis could be made. It should not be presented as proof that DUR-928 works, does not work, is safe, or is unsafe.

14. Why This Trial Matters Statistically

This study is statistically informative as an example of the difference between a trial’s intended design and the data it ultimately produces. The registry identifies randomization, double blinding, placebo control, and a parallel-group structure. Yet both enrolled participants were assigned to DUR-928, leaving no placebo participants. The intended comparison therefore did not become an estimable comparison in the posted results.

That distinction matters because design labels alone do not guarantee that a treatment effect can be estimated. Randomization is a method for balancing groups through assignment, but an empty arm supplies no outcome data for a contrast. Masking can reduce some sources of bias in a study, but it cannot solve the absence of a comparator group. Statistical methods can describe uncertainty only when the necessary information is available.

The record also illustrates why counts should be interpreted with their denominators and study context. A primary endpoint count in one arm may be factually accurate while still being insufficient to answer the clinical question. The early termination and the small enrollment are central to interpreting the count, as is the lack of posted comparative analyses.

For clinical researchers, the case underscores the value of distinguishing descriptive results from inferential claims. It also shows why registry records should be read for both outcome values and the information needed to interpret them: enrollment by arm, analysis availability, endpoint definition, and missing methodological details. The main statistical conclusion is not a direction or size of treatment effect, but that this record does not permit a placebo-controlled estimate.

15. 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 limitations specific to this study constrain interpretation. The registry says the study was terminated early after only 2 subjects enrolled, and both were assigned to active treatment. This left the placebo arm without participants. The result is therefore not a comparison of randomized groups, despite the randomized design stated in the record.

The primary outcome is a composite, but only an overall participant count is posted. The record does not give separate results for mechanical ventilation, renal replacement therapy, or acute liver failure. It also does not report detailed endpoint-assessment or adjudication procedures. This prevents examination of which components contributed to the recorded endpoint status.

No analyses are posted, and the registry indicates that a two-arm comparison for the binary endpoint is not estimable. There is no reported effect estimate or confidence interval for the treatment contrast. The record does not report the protocol’s sample-size justification, the reason for termination, a detailed statistical analysis plan, or a specific plan for missing data and analysis populations. These details are not reported in the registry record and cannot be inferred from the observed counts.

The safety information is also limited by the small enrollment and the absence of a placebo group. Recorded serious and other adverse-event figures describe only the enrolled DUR-928 participants. The listed serious-event terms do not, in the available information, provide a basis for a controlled safety comparison. The recorded attribution of the deaths as Not Related to study treatment should be reported as an assessment in the registry, not extended into a broader causal conclusion.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: DUR-928 in Subjects With SARS-CoV-2 With Acute Lung, Liver or Kidney Injury — NCT04447404.
• Registry record last updated 2022-08-05; status last verified 2022-07.
• Results first posted 2022-08-05.
• 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.

19. Record Summary

The registry record describes NCT04447404, a Phase 2 randomized, double-blind, placebo-controlled, parallel-group study sponsored by Durect. It evaluated DUR-928 in subjects with SARS-CoV-2 and acute lung, liver, or kidney injury. The study is listed as terminated, with a start date of 2020-09-21 and a primary completion date of 2021-03-04.

Only 2 subjects enrolled, both assigned to DUR-928; no participants were assigned to placebo. The primary endpoint was the count alive and free of organ failure at Day 28, with a posted count of 0.0 in the DUR-928 arm. The record reports deaths on Day 19 and Day 20 and says both were assessed as Not Related to study treatment. No analyses are posted, and a two-arm comparison is not estimable from the enrolled groups.