A Phase I/Ib Study of AZD9150 (ISIS-STAT3Rx) in Patients With Advanced/Metastatic Hepatocellular Carcinoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Phase I/Ib Study of AZD9150 (ISIS-STAT3Rx) in Patients With Advanced/Metastatic Hepatocellular Carcinoma (NCT01839604), a na phase 1 study. The primary outcome is Number of Participants With Dose Limiting Toxicities During Cycle 1.

2. Clinical Question

This is a safety and early-development question rather than a confirmatory comparison of AZD9150 with another treatment. The study also listed preliminary assessment of tumour response and pharmacokinetic measures, including Cmax and Tmax, among its secondary outcomes. Those objectives indicate the kinds of activity and drug-exposure information the study sought to assess, but the registry record does not report numerical results for them in the information available for this analysis.

3. Trial Design

The registry describes a Phase I/Ib, open-label, multicentre study of AZD9150 in patients with advanced or metastatic hepatocellular carcinoma. The phase field lists PHASE1. AstraZeneca is named as sponsor. The recorded enrollment is 58, and the study status is completed. The registry dates the start to 2013-05 and primary completion to 2015-02.

The study had a dose-escalation part, Part A, and a dose-expansion part, Part B. The record lists five dose groups: AZD9150 1mg/kg, 1.5 mg/kg, 2 mg/kg, 2.5 mg/kg, and 3 mg/kg. It does not describe a separate comparator group. Allocation is listed as not applicable, masking as none, and the model as single-group. In practical terms, participants and investigators were not masked to treatment, and the dose groups do not constitute randomized controls.

The primary outcome was the number of participants with dose-limiting toxicities during Cycle 1. The primary assessment window was Cycle 1 (22 days), with the cycle defined as 3 loading doses on Days 1, 3, and 5 followed by 3 weekly doses on Days 8, 15, and 22. The stated population was all patients who received at least 1 dose of AZD9150.

The registry also lists pharmacokinetic evaluation of Cmax and Tmax after single administrations and preliminary evaluation of tumour response as secondary outcomes. Four outcome measures are listed, but the numerical primary endpoint result and the detailed results for those secondary outcomes are not reported in the registry record. The record lists no analyses. It does not report the protocol’s dose-escalation decision rules, dose-selection rationale, or other design details beyond the study description and outcome definitions.

Because the study was open-label and lacked a randomized control group, the design is suited to describing observations within the treated study population, not to isolating a causal treatment effect against an alternative. The dose groups may also differ in ways not described by the registry record. Their counts should therefore be treated as descriptive dose-group observations rather than as evidence from a controlled comparison.

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: Note that 39 patients is the number of patients who were actually assigned to treatment out of the total enrolled. The ecpansion phase consisted of the 3mg/kg group only.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Safety: All patients who received at least 1 dose of AZD9150

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 Number of Participants With Dose Limiting Toxicities During Cycle 1, measured over DLT assessment window - Cycle 1 (22 days).

The primary endpoint counts participants with dose-limiting toxicities during Cycle 1. Its unit is participants with DLT, and its parameter is a number rather than a rate or an average. The assessment window is Cycle 1 (22 days), defined as 3 loading doses on Days 1, 3, and 5 followed by 3 weekly doses on Days 8, 15, and 22. The stated safety population includes all patients who received at least 1 dose of AZD9150.

A dose-limiting-toxicity count is used in early-phase drug development to describe whether serious tolerability concerns meeting the study’s DLT definition occurred during the specified assessment period. It can help characterize safety at the doses studied, but the count alone does not describe the type, timing, severity, or duration of each event. Those details depend on the endpoint definition and the accompanying event information. The DLT definition itself is not reported in the registry record.

The registry record does not report the observed number of participants with DLT for this endpoint. Thus, the endpoint’s purpose and assessment window can be described, but its numerical result cannot be interpreted from this record. The serious adverse event counts reported by dose group are related safety information, but serious adverse events and dose-limiting toxicities are not interchangeable endpoint labels. The registry record does not establish that every serious adverse event met the DLT definition, or that every DLT was a serious adverse event.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a count: the number of participants with dose-limiting toxicities during a defined period. This is a binary participant-level outcome in the sense that each participant either is or is not counted as having the event under the study definition. The reported endpoint unit is participants with DLT, and the parameter is a number. It is not an average measurement such as a laboratory value, nor is it a time-to-event endpoint.

What does a participant count tell us?

A count describes how many people met the endpoint definition in the population being assessed. When the denominator is also known, a count can be placed in context as a fraction of the assessed population. That fraction is often more interpretable across groups of different sizes than the count alone. In this case, the registry record identifies the primary safety population as all patients who received at least 1 dose of AZD9150, but it does not report the primary endpoint count or its outcome-specific denominator. Consequently, the primary DLT result cannot be converted into or interpreted as a proportion from the registry record.

Counts in small groups can be sensitive to the experience of individual participants. An observed count is the data for the group, not a precise statement about what would necessarily be observed in another set of participants. General statistical practice is to consider uncertainty as well as the point estimate. No confidence interval for the primary DLT endpoint is reported in the registry record. A confidence interval, in general, expresses the range of values compatible with the data under a specified statistical method; it does not resolve design limitations or turn an uncontrolled observation into a controlled comparison.

Why does the assessment window matter?

The endpoint is restricted to Cycle 1 (22 days), defined as 3 loading doses on Days 1, 3, and 5, followed by 3 weekly doses on Days 8, 15, and 22. A count within this window answers a narrower question than a count over the entire period of follow-up. It does not, by itself, describe toxicities that occur later. The registry’s time window also gives the endpoint a consistent period of assessment, but the record does not provide the detailed DLT criteria or rules for handling participants whose assessment may have been incomplete.

Early-phase safety studies commonly use dose-limiting toxicity counts to help characterize tolerability during dose evaluation. That is a general explanation of the endpoint’s role, not a description of an unreported protocol rule. The registry record does not report how dose escalation was governed, what threshold would have prompted a change, or how a dose was selected for expansion. Those details should not be inferred from the dose groups alone.

How should dose groups be compared?

The study is described as single-group, with allocation listed as not applicable and masking as none. The listed dose groups are therefore not randomized, like-for-like comparator arms. In general, a randomized comparison helps balance measured and unmeasured characteristics between groups; this design does not provide that protection for comparisons between dose levels. Participants may differ for reasons that the registry record does not describe, and the groups may reflect different stages or parts of the study. For that reason, the adverse-event counts by dose are descriptive. They do not isolate the effect of dose or support a causal dose-response conclusion.

The serious adverse event counts have denominators reported for each listed group. Interpreting a count alongside its denominator helps avoid treating a count from a smaller group as directly comparable to a count from a larger group. Even with denominators, these observations remain imprecise and non-randomized. The registry does not report confidence intervals or a formal statistical comparison for the serious adverse event counts.

What do the planning assumptions mean?

Any sample-size calculations associated with the statistical material are planning illustrations, not the study’s reported design justification and not observed power. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is a stated fraction of the observed effect. The registry record does not report that these assumptions governed the trial, and this study’s registry design is not a randomized two-arm comparison. A planning calculation describes a hypothetical design under specified assumptions. It does not establish what the study was designed to detect unless the protocol confirms that those assumptions were used.

Observed power is not a substitute for estimating uncertainty in the results. In general, a study’s observed data and their precision are more informative than a retrospective power calculation based on the observed effect. No numerical planning estimates or inferential results are stated in the registry information, and no numerical primary DLT result is available for recalculation.

What can recalculation establish?

A recalculated result, when available, is an unadjusted calculation from posted summary statistics. It may help readers check a simple arithmetic relationship, but it does not replace the sponsor’s analysis when a discrepancy is identified. No primary DLT summary count is reported in the registry record, and no statistical analyses are listed. There is therefore no basis in the record for recalculating the primary endpoint, its uncertainty, or a treatment comparison. The listed adverse event counts can be read as posted descriptive counts, but they do not permit an adjusted causal analysis.

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("AZD9150 1mg/kg", "1.5 mg/kg", "2 mg/kg", "2.5 mg/kg", "3 mg/kg")
n      <- c(6, 6, 6, 5, 14)
events <- c(0, 1, 1, 2, 3)
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  AZD9150 1mg/kg  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0
ARM  1.5 mg/kg  6  1  16.7  0.4  64.1  3.0  56.4  3.7  57.9
ARM  2 mg/kg  6  1  16.7  0.4  64.1  3.0  56.4  3.7  57.9
ARM  2.5 mg/kg  5  2  40.0  5.3  85.3  11.8  76.9  11.8  77.7
ARM  3 mg/kg  14  3  21.4  4.7  50.8  7.6  47.6  7.8  48.1

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists four outcome measures, including the primary DLT outcome and secondary outcomes for pharmacokinetics and preliminary tumour response. It does not report a multiplicity strategy describing how inferential conclusions across outcomes or analyses would be controlled. The record also lists no analyses. In general, examining multiple outcomes can increase the chance of a misleading positive finding if statistical testing is conducted repeatedly without an appropriate plan. That general concern cannot be resolved for this study from the registry information.

The primary endpoint population is defined as all patients who received at least 1 dose of AZD9150. The registry does not report how many participants contributed to the primary outcome, whether all eligible participants had complete Cycle 1 assessment, or how missing or incomplete endpoint data were handled. The DLT definition and any rules for classifying participants with incomplete follow-up are also not reported in the registry record. Without those details, the reader cannot assess whether missing endpoint information could affect the count.

The adverse event listings report group-specific numbers at risk and numbers affected for serious and other adverse events, but those data do not document missing-data handling for the primary outcome. The overall enrollment of 58 should not be assumed to be the denominator for every outcome. Nor should the available adverse-event denominators be treated as the primary endpoint denominator. The registry record does not report an analysis population beyond the stated primary safety population definition, nor does it describe sensitivity analyses or methods for handling missing observations.

12. Safety Results

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

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: abdominal pain (20.0%), hepatorenal failure (20.0%), aspartate aminotransferase increased (20.0%), liver abscess (14.3%), peritoneal haemorrhage (14.3%).

The registry reports serious adverse events by dose group. The AZD9150 1mg/kg group had 0 of 6 participants affected; the 1.5 mg/kg group had 0 of 6; the 2 mg/kg group had 2 of 7; the 2.5 mg/kg group had 2 of 5; and the 3 mg/kg group had 4 of 15. These counts describe the participants recorded as having serious adverse events in each group. They are not the primary DLT outcome, whose numerical result is not reported.

The registry also lists selected serious adverse event terms with percentages: abdominal pain, hepatorenal failure, and aspartate aminotransferase increased at 20.0%; liver abscess, peritoneal haemorrhage, and alanine amonitransferase increased at 14.3%. The record does not give enough context in these entries to reconstruct the relevant denominator or determine whether participants experienced more than one listed event. The percentages should therefore be read as posted, not reinterpreted as group-specific rates.

Serious adverse event counts do not establish that an event was caused by AZD9150. The study had no randomized control group, and the registry does not report a treatment-relatedness analysis for these events. Death counts are not reported in the registry record.

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

The registry record establishes that AZD9150 was studied in an open-label, multicentre Phase I/Ib study with dose escalation and dose expansion in patients with advanced or metastatic hepatocellular carcinoma. It defines the primary safety question as the number of participants with DLT during Cycle 1 (22 days), and it gives the dosing schedule used to define that assessment window. It also reports serious adverse event counts for the listed dose groups and selected serious event terms with percentages.

Those details do not establish the numerical result of the primary endpoint. The count of participants with DLT is not reported in the registry record. The serious adverse event counts cannot fill that gap because they refer to a different safety measure. Nor do the counts show whether the study met any dose-escalation criterion: the registry does not report the DLT definition, escalation rules, or a dose-selection decision.

The study’s design also limits what can be concluded across dose groups. Because the record describes no randomized comparator and lists allocation as not applicable, observed differences in group counts cannot be interpreted as effects caused by dose. Participant characteristics, timing, and other factors that could influence event counts are not described sufficiently to support such an inference. The open-label design further means that masking cannot be assumed for participants or investigators.

The registry identifies tumour response and pharmacokinetic evaluation as secondary objectives, but numerical results for those outcomes are not reported in the information available for this analysis. Thus, the record does not support a conclusion about the extent of tumour response, drug exposure, or a treatment effect compared with another option. A registry description of objectives is not evidence that an objective was achieved.

In general, early-phase studies provide descriptive information that can inform later research, while controlled studies are needed to estimate comparative effects with stronger causal interpretation. That distinction is general and does not imply a particular clinical decision. The most defensible reading is limited to the recorded study design, endpoint definitions, enrollment, and posted adverse event data, with the missing primary numerical result and analysis details kept explicit.

14. Why This Trial Matters Statistically

This study is statistically informative as an example of how early-phase safety questions differ from confirmatory comparative questions. Its primary endpoint is a participant count for dose-limiting toxicities in a tightly defined Cycle 1 (22 days) window. The focus on a specified early treatment period is relevant to dose evaluation, but the count is useful only when its outcome definition, denominator, and result can be interpreted together. The registry record supplies the window and population definition but not the numerical primary result.

The dose groups also illustrate why descriptive safety counts are not equivalent to randomized comparisons. The registry reports serious adverse event counts at several dose levels, but allocation is listed as not applicable and the study is single-group. The dose-level counts therefore summarize what was recorded in each group; they do not establish that dose caused any difference in events. That distinction matters especially when groups are small and their composition or study-stage context is not reported.

Another statistical lesson is the importance of keeping related but distinct outcomes separate. DLT is the primary endpoint, while serious adverse events are a separate safety listing. Without a reported DLT count or a stated relationship between those definitions, the serious-event summaries cannot be used as a proxy for the primary endpoint. Similarly, the stated secondary aims for pharmacokinetics and tumour response do not convey results in the absence of reported outcome values.

Finally, transparent registry reporting is necessary for readers to assess precision, missingness, and analysis choices. The record identifies the study’s broad purpose and several outcome definitions, but it does not report the primary result, analyses, multiplicity approach, or missing-data handling. The resulting interpretation must remain descriptive and bounded by what is recorded, rather than filling those gaps with assumptions.

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.

The principal trial-specific limitation is that the registry record does not report the numerical result for the primary DLT endpoint. It gives the assessment period and safety population definition, but not the number of participants with DLT or the number assessed for that endpoint. The DLT criteria themselves are also not reported, limiting interpretation of what events qualified.

The study was open-label and single-group, with no randomized comparator. The dose-group observations are not like-for-like control comparisons, and the registry does not describe participant characteristics by dose, the sequence or rules for dose escalation, or how participants entered dose expansion. Therefore, differences in adverse event counts should not be interpreted as evidence of a dose effect.

The registry reports serious adverse event counts, but does not provide event-level context sufficient to understand timing, relatedness, recurrence, or overlap among participants. It also does not report death counts. The selected serious event terms and percentages lack enough accompanying denominator detail in the record to reconstruct group-specific rates. These limitations restrict the safety interpretation to the posted summaries.

Although tumour response and pharmacokinetic outcomes are listed, their numerical findings are not reported in the information available for this analysis. The record also lists no analyses and does not report a multiplicity plan, missing-data methods, sensitivity analyses, or detailed analysis populations beyond the stated primary safety population. These omissions prevent assessment of statistical precision and robustness for the primary and secondary outcomes.

Finally, the enrollment figure of 58 is not a substitute for outcome-specific denominators. The adverse-event groups and the primary endpoint population may not have the same number of participants, and the registry does not provide the primary outcome denominator. Conclusions must therefore be limited to the design and descriptive safety information actually recorded.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Phase I/Ib Study of AZD9150 (ISIS-STAT3Rx) in Patients With Advanced/Metastatic Hepatocellular Carcinoma — NCT01839604.
• Registry record last updated 2017-03-06; status last verified 2016-05.
• Results first posted 2017-03-06.
• 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

This completed, open-label, multicentre Phase I/Ib study evaluated AZD9150 in patients with advanced or metastatic hepatocellular carcinoma. It included dose escalation and dose expansion, with listed dose groups from 1mg/kg to 3 mg/kg. Recorded enrollment was 58; the study start was 2013-05 and primary completion was 2015-02.

The primary endpoint was the number of participants with dose-limiting toxicities during Cycle 1 (22 days), among patients who received at least 1 dose. Its numerical result is not reported in the registry record. Serious adverse event counts are reported by dose group, but the study was not randomized and those counts do not establish a dose effect or replace the primary endpoint.

Pharmacokinetic evaluation and preliminary tumour response were listed as secondary outcomes. The registry record lists no analyses and does not report numerical results for those outcomes in the information available for this analysis. Interpretation is consequently limited to the study design, endpoint definitions, and posted safety summaries.