Study of TAC-101 as Second Line Treatment in Patients With Advanced Hepatocellular Carcinoma Who Received Sorafenib as First Line Therapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of TAC-101 as Second Line Treatment in Patients With Advanced Hepatocellular Carcinoma Who Received Sorafenib as First Line Therapy (NCT00687596), a randomized phase 2 study. The primary outcome is Overall Survival (OS).

2. Clinical Question

The study asked whether oral TAC-101, used as second-line treatment after sorafenib, could slow tumor activity in people with advanced hepatocellular carcinoma. Overall survival was the primary outcome: the time from randomization to death, with participants still alive censored at their last known date alive. The study also assessed safety and secondary outcomes including radiologic progression-free survival, time to progression, tumor markers, and exposure-response measures.

This is a description of the posted results, not a treatment recommendation. The registry record does not report a formal statistical analysis plan or the rationale used to choose the original enrollment target. It also does not provide a reported primary-outcome analysis beyond the summary comparison shown in the posted results.

3. Trial Design

The registry describes a phase 2, randomized, double-blind, placebo-controlled, international, multicenter, parallel-group study. Its stated purpose was to assess TAC-101 as second-line treatment after sorafenib in advanced hepatocellular carcinoma, with attention to tumor activity and safety. Enrollment was 52, and the study status is recorded as terminated.

The study began on 2008-08-01, and primary completion was recorded as 2010-05-10. The primary outcome timeframe was from randomization to death, or up to 1 year after the last participant was randomized, whichever came earlier; the registry gives a maximum duration of up to 21.3 months. Participants still alive were censored at the date they were last known to be alive, based on last contact or last follow-up visit.

The primary outcome was summarized in the safety population, defined in the registry record as randomized participants who received at least one dose of double-blind study medication. The results were summarized by treatment received. The posted primary-outcome arms each contain 26 participants. This analysis population definition matters: it is not simply a statement that every randomized participant was included, because receipt of at least one dose was also required.

The intervention description states that TAC-101 was given orally at 20 mg/day within 1 hour after morning meals on Days 1 to 14, followed by a recovery period. The description ends mid-sentence, and the full treatment schedule is not reported in the registry record. The placebo description also gives a matching oral tablet schedule, but its text is truncated. Details of randomization procedures, allocation concealment, site-level conduct, adherence, and reasons for termination are not reported in the registry record. The listed design features therefore describe the design as registered, not undocumented implementation details.

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: The study was stopped after only 52 of the required 220 participants (26 of 110 participants in each treatment group) were randomized. Sponsor terminated the study due to safety concerns and thus, stopped enrollment and study treatment on 06-May-2009, but continued follow-up of all the treated participants for survival till 10 May 2010 and thus, the same was considered as study completion date.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Analysis was performed on safety population that included all randomized participants who received at least one dose of double-blind study medication and was summarized by the treatment received.

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 analysis included all randomized participants who received any dose of study medication and analyzed by the treatment received.

6. Primary Endpoint

The registry lists the primary outcome as Overall Survival (OS), measured over From the date of randomization to date of death or up to 1 year post last participant was randomized, whichever was earlier (maximum duration: up to 21.3 months).

Overall survival measures the time from randomization until death. In this study, participants who remained alive were censored at the last date they were known to be alive, using the last contact date or last follow-up visit. The registry specifies a follow-up window ending at the earlier of death or 1 year after the last participant was randomized, with a maximum duration of up to 21.3 months.

Overall survival is used in clinical research because it is a direct time-to-event outcome: the event is death, and the timing of that event is part of the outcome. Censoring allows information from participants who have not had the event by their last known follow-up to be retained, while recognizing that their later survival time is unknown. This general explanation does not establish which statistical method was used in this study.

The posted outcome is expressed as a mean in months, with a standard deviation for each treatment group. The registry record does not report a survival-curve analysis, hazard ratio, or other time-to-event model for the primary endpoint. A mean and standard deviation provide a summary of the posted values, but do not by themselves describe the entire pattern of event timing or censoring. That distinction is important when interpreting an outcome whose definition is time to death.

The primary question in the posted numerical comparison is therefore a difference between the reported group means, not a reported hazard ratio or a complete survival-distribution comparison. The endpoint is clinically interpretable, but the summary statistics alone give a limited view of how survival experience varied over follow-up.

7. Results Posted in the Registry

• TAC-101: mean 8.23 versus 8.35 on control; difference -0.12 (95% CI -3.04 to 2.80), standardized effect d = -0.02; z-test p = 0.936.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Radiologic Progression-free Survival (PFS)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • TAC-101: mean 4.95 versus 4.82 on control; difference 0.13 (95% CI -2.46 to 2.72), standardized effect d = 0.03; z-test p = 0.922.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between TAC-101 and the control. The interval gives the range of effects that remain compatible with the data.

Time To Progression (TTP)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • TAC-101: mean 4.14 versus 4.09 on control; difference 0.05 (95% CI -2.26 to 2.36), standardized effect d = 0.01; z-test p = 0.966.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between TAC-101 and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline in Plasma Levels of the Tumor Marker Alpha-fetoprotein (AFP)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • TAC-101: mean 2,779 versus 571.70 on control; difference 2,208 (95% CI -6,895 to 11,310), standardized effect d = 0.14; z-test p = 0.635.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between TAC-101 and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary summaries cover radiologic progression-free survival, time to progression, and change from baseline in plasma alpha-fetoprotein. The first two mean estimates are close between groups, and their unadjusted confidence intervals include zero. The alpha-fetoprotein mean difference is larger in its measurement units, but its confidence interval also includes zero and the observed values are highly variable. These patterns do not provide a clear directional signal that differs substantially from the primary mean comparison.

Secondary endpoints address related but distinct questions. Progression-free survival includes progression or death, whereas time to progression concerns progression; tumor-marker change is a laboratory measure rather than a direct survival outcome. Agreement or disagreement among these outcomes can be informative, but none should be treated as a substitute for the primary endpoint. The registry record lists additional secondary outcomes concerning adverse events, AFP-L3, TAC-101 exposure, and IGF-2, but detailed arm summaries for those measures are not reported in the registry record in the results.

The displayed secondary comparisons are unadjusted. When several outcomes are examined, the chance of finding an apparently notable result by chance can increase. This is the general multiplicity problem. The registry record does not report a multiplicity-control plan or identify a confirmatory hierarchy for the secondary outcomes. As a result, individual secondary p values, where posted, should be treated as exploratory rather than as independent confirmation of a treatment effect.

The registry record lists 15 secondary outcome measures, but it does not list reported analyses for the three secondary summaries with numerical comparisons. The absence of a listed analysis plan makes it difficult to determine which analyses were prespecified, how missing observations were handled, or whether all measured outcomes are represented in the posted results. The broad secondary pattern is therefore useful as context, but it carries less evidential weight than a prespecified, fully reported primary analysis.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Radiologic Progression-free Survival (PFS)
• Time To Progression (TTP)
• Number of Participants With Treatment-Emergent Adverse Events (TEAEs) and Treatment-Emergent Serious Adverse Events (TESAEs)
• Change From Baseline in Plasma Levels of the Tumor Marker Alpha-fetoprotein (AFP)
• Percent Change From Baseline in Plasma Levels of the Tumor Marker Alpha Fetoprotein-L3 (AFP-L3)
• Maximum Exposure-response Analysis: Maximum Plasma Concentration (Cmax) of TAC-101
• Overall Exposure-response Analysis: Area Under Curve (AUC0-inf) of TAC-101
• Changes From Baseline in Plasma Insulin-like Growth Factor-2 (IGF-2)

9. Statistical Interpretation

10. Statistical Methods Explained

The posted continuous summaries compare the mean outcome in the TAC-101 group with the mean in the placebo group. The difference is expressed as TAC-101 minus placebo. A negative difference means the TAC-101 mean is lower; a positive difference means it is higher. The estimates are descriptive comparisons of the posted group summaries and should not be confused with an adjusted model unless an adjustment is explicitly reported. No adjusted analyses are listed in the registry record.

What do the mean and standard deviation describe?

A mean is the arithmetic average of the recorded outcome values in a group. The standard deviation describes how spread out those values are around the mean. A larger standard deviation indicates more variation among observations, although it does not show the shape of that variation or whether unusual observations influence the mean. For overall survival, the registry reports mean months and standard deviations, even though the endpoint is defined by time to death and includes censoring.

Standard deviations are not standard errors. A standard deviation describes variation among participants, whereas a standard error describes uncertainty in an estimated quantity such as a group mean. The posted comparison includes standard errors for the two means, but the registry record does not explain the exact inferential model underlying every summary. No additional modeling assumptions should be inferred from the summary alone.

How should the mean difference be read?

The primary estimate is -0.12 months for TAC-101 relative to placebo. The sign reflects the order of subtraction, not a judgment about clinical importance. The estimate is the observed difference in the reported means; it is not proof that the population-level difference is exactly -0.12 months. Sampling variation means that another study could produce a different estimate, and the confidence interval describes uncertainty around this estimate under the method used to construct it.

The standardized difference, -0.022, expresses the contrast in relation to the pooled standard deviation listed in the comparison. Standardization can help describe the magnitude on a scale that is less dependent on the outcome units. It does not resolve the limitations of the underlying endpoint summary, and it should not be treated as a separate clinical outcome.

What does the unadjusted p value say?

The primary unadjusted p value is 0.936. In general, a p value measures how incompatible the observed data are with a specified null hypothesis under the assumptions of the test. A large p value does not establish the null hypothesis, prove that treatments are equivalent, or quantify the probability that the observed result arose by chance. It also does not measure the size or importance of an effect. The term unadjusted indicates that the posted value is not presented as adjusted for multiple comparisons; an adjustment method is not reported in the registry record.

The confidence interval and p value answer related but different questions. The interval gives a range of estimates compatible with the data and method, while the p value summarizes evidence against a specified null. Neither can establish clinical importance without context such as a prespecified meaningful-difference threshold, which is not reported in the registry record.

Why does censoring matter for survival outcomes?

In a time-to-event outcome, some participants may not have experienced the event by the end of their observed follow-up. Their event time is not known; the available information is that they were alive through their last known contact. Censoring is a way of retaining that partial follow-up information rather than treating the last contact as the event. General time-to-event methods use event timing and censoring information, and their validity depends on assumptions about how censoring relates to the event process.

The registry definition states how living participants were censored, but the posted primary summary gives group means and standard deviations. A mean comparison does not reveal the full distribution of event times, the number and timing of censored observations, or how censoring was handled in an inferential model. A survival curve, event count, or hazard ratio is not reported in the registry record, so those methods or quantities should not be presumed.

Why can over-dispersion matter?

General explanation: over-dispersion occurs when outcome variation is greater than a statistical model assumes. If a model understates variability, standard errors can be too small and confidence intervals too narrow, making evidence appear more precise than it is. The posted standard deviations indicate considerable variation in the measured outcomes, but the registry record does not report an over-dispersion assessment or a model designed to address it. Over-dispersion should therefore be understood as a general methodological concern, not as a documented feature of this analysis.

The same general principle applies to skewness and influential observations. Means and standard deviations can be sensitive to extreme values, especially for outcomes with a long right tail. The registry record does not provide distributions or participant-level values, so those features cannot be evaluated from the posted summaries. No claim about the presence of skewness or outliers is warranted.

What is and is not known about the analysis?

The registry reports a primary outcome, an analysis population definition, group summaries, a mean difference, a confidence interval, and an unadjusted p value. It does not report the original statistical analysis plan, missing-data procedures beyond the stated survival censoring definition, interim analyses, covariate adjustment, or a rationale for the chosen sample size. The registry also lists no reported secondary analyses for the detailed secondary outcomes discussed in the posted summaries. These omissions limit what can be inferred about prespecification and analytic choices.

11. Confidence Intervals and Precision

The 95% confidence interval for the primary mean difference extends from -3.044 to 2.804 months. Because it includes zero, the posted data do not distinguish the estimate from no mean difference at the conventional two-sided significance level associated with this interval. More importantly, the interval also includes differences in both directions. It therefore expresses uncertainty about direction as well as magnitude.

A confidence interval is a range produced by a statistical method that, under its assumptions, is intended to cover the true effect at the stated confidence level over repeated sampling. It is not a range containing a specified probability that the true effect lies inside it after the data have been observed. Its width reflects both the amount of information and the variability in the outcome, as well as the method used to calculate the interval.

The point estimate of -0.12 months is much closer to zero than the endpoints of the interval. The posted interval is therefore consistent with the observed estimate being only one of many effects that remain compatible with these data. Without a prespecified threshold for a clinically important difference, precision cannot be classified as adequate or inadequate for a decision; that threshold is not reported in the registry record.

The standard deviations, 5.406 months for TAC-101 and 5.351 months for placebo, show substantial within-group spread relative to the estimated contrast. The confidence interval captures uncertainty in the difference, not the range of individual survival times. It also does not show the distribution of censoring or the timing of deaths. Participant-level follow-up and a reported time-to-event analysis would be needed to examine those aspects.

The interval should not be read as proof of no effect. Nor does its inclusion of zero alone establish that the treatments are alike. It indicates that the posted mean comparison is not precise enough to support a confident directional conclusion about the underlying mean difference.

12. 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 illustrations derived from the observed effect, not the original trial’s sample-size justification. The original design assumptions are not reported in the registry record. The stated illustration 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. For time-to-event calculations, the reported planning quantity is the total number of events; for other calculations, it is participants per arm.

The 80% power column describes a hypothetical design that, under its assumptions, would have an 80% chance of detecting the specified effect if that effect were the true one. The 90% power column describes the corresponding hypothetical chance of detection at that power target. These are planning probabilities, not guarantees about what a future study will find. The figures are not observed power.

The rows that plan for 100% of the observed effect use the full observed estimate as the target effect. The rows for 75% and 50% of the observed effect use progressively smaller targets. Smaller effects are harder to distinguish from random variation, so they require more participants or, for time-to-event planning, more events under the stated assumptions. The very small observed primary standardized difference helps explain why planning from the full observed primary effect produces very large illustrative requirements.

These calculations should not be mistaken for a recommendation about a future trial. They do not incorporate an independently chosen clinically meaningful effect, loss to follow-up, adherence, or other design features unless those are included in the stated assumptions. The registry record does not report such additions or the original sample-size calculation. The figures are best read as conditional illustrations of how sample requirements change when the target effect changes, rather than as evidence about the adequacy of the completed study.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = -0.120; pooled standard deviation σ = 5.379.
2. n = 2 × (1.96 + 0.842)² × 5.379² / -0.120² = 31,536.2, rounded up to 31,537 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 <- -0.119999999999999                      # observed difference in means
sigma <- 5.37857030259901                     # 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(687596)
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 = 31,537 (80% power), 42,218 (90% power)
 75% of observed effect: n per arm = 56,065 (80% power), 75,055 (90% power)
 50% of observed effect: n per arm = 126,145 (80% power), 168,872 (90% power)
power.t.test: 31537.1 
Simulated power at n = 31,537 per arm: 0.798 (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.

14. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates multiple outcomes, time points, subgroups, or analytic approaches. In general, the more opportunities there are to test hypotheses, the greater the chance that at least one result may appear unusual even when no corresponding effect exists. Common approaches include defining one primary hypothesis in advance, controlling error across a family of comparisons, or explicitly treating secondary analyses as exploratory. The registry record does not report a multiplicity adjustment or a testing hierarchy.

The study identifies overall survival as the primary outcome and lists 15 secondary outcome measures. The detailed secondary results include several distinct types of measures, including progression outcomes and biomarkers. The registry record gives unadjusted p values for the posted comparisons and does not state that they were corrected for multiplicity. Consequently, a secondary result should not be interpreted as confirmatory solely because its unadjusted p value appears small or because it points in a particular direction.

Missing-data handling is also only partly described. For overall survival, participants still alive were censored at their last known date alive. That is part of the endpoint definition, not a complete account of missing-data methods for every outcome. The registry record does not report how missing observations were handled for progression, tumor-marker, or other secondary outcomes, or whether sensitivity analyses were performed. It does not provide the amount or pattern of missing data for those endpoints.

The primary outcome population is the safety population: randomized participants who received at least one dose of double-blind study medication, summarized by treatment received. The registry record does not report a separate estimate for all participants analyzed according to randomized assignment. These population definitions can differ if treatment received or receipt of a dose differs from randomized allocation. The impact of that distinction cannot be assessed from the posted summaries alone.

The registry record also does not report interim analyses, stopping boundaries, the original statistical analysis plan, or a sample-size justification. The study status is recorded as terminated, but the reason for termination is not reported in the registry record. These gaps mean that prespecification, selective reporting, and the consequences of missing data cannot be evaluated from the reported summaries.

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: deep vein thrombosis (15.4%), atrial thrombosis (7.7%), gastrointestinal haemorrhage (7.7%), fatigue (7.7%), cardiac arrest (3.8%).

The posted safety counts distinguish serious adverse events, other adverse events, and deaths. In the TAC-101 group, 12 of 26 participants were reported with a serious adverse event, 25 of 26 with another adverse event, and 2 deaths were recorded. In the placebo group, the corresponding counts were 11 of 26, 24 of 26, and 6 deaths. These counts describe what was recorded in the groups; they do not establish that study treatment caused an event or a death.

The registry lists deep vein thrombosis at 15.4%, atrial thrombosis at 7.7%, gastrointestinal haemorrhage at 7.7%, fatigue at 7.7%, cardiac arrest at 3.8%, and left ventricular failure at 3.8% among the top serious adverse event terms. The posted term percentages should be read as registry-reported frequencies, not as comparative treatment effects. The registry record does not provide event-level narratives or establish causality for these terms.

Safety counts are not a substitute for a detailed assessment of event severity, timing, duration, recurrence, or relationship to treatment. The record identifies treatment-emergent adverse events and serious events as outcomes, but detailed information on those features is not reported in the registry record. The death counts also should not be assumed to represent deaths caused by treatment; the cause and attribution are not reported.

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

The posted primary result is a mean overall-survival estimate of 8.23 months for TAC-101 and 8.35 months for placebo, with a mean difference of -0.12 months. The confidence interval from -3.044 to 2.804 months is broad enough to include effects in either direction. The unadjusted p value of 0.936 does not provide evidence against a no-difference null under the comparison used, but it does not prove that there is no difference.

The result does not establish equivalence, non-inferiority, or a clinically unimportant difference. Those conclusions require a prespecified margin and an analysis designed around that margin. Neither is reported in the registry record. The interval indicates uncertainty; it cannot identify which compatible value is the true effect or whether any value would matter clinically.

Nor does the result establish that TAC-101 slows tumor activity or that it fails to do so. The registry’s primary endpoint is overall survival, while tumor progression and biomarkers are among the secondary outcomes. The posted summaries are not a complete account of the evidence on every outcome, and the registry record does not report a full time-to-event model for overall survival.

The study status is recorded as terminated. That status alone does not explain why the study ended, whether it ended as planned, or whether a particular result influenced the decision. The reason is not reported in the registry record. It would be inappropriate to infer a clinical or statistical rationale from status alone.

In practical statistical terms, this is a small randomized study with a near-zero primary mean difference and substantial uncertainty around that estimate. Its posted results are descriptive and do not support a firm conclusion about the underlying treatment effect. They also do not replace a complete assessment of the trial’s prespecified methods, participant-level follow-up, or safety details, which are not reported in the registry record.

17. Why This Trial Matters Statistically

This trial illustrates how a randomized design and a nearly identical pair of group means can coexist with considerable uncertainty. A point estimate close to zero may appear straightforward, but interpretation depends on the confidence interval, variability, analysis population, endpoint definition, and completeness of the statistical reporting. Here, the primary confidence interval spans negative and positive mean differences, so the point estimate alone would overstate the certainty of the result.

It also demonstrates why a non-significant comparison is not evidence of equivalence. The unadjusted p value is 0.936, but an equivalence claim would require a prespecified margin and an appropriate design and analysis. Those details are not reported in the registry record. The distinction is central to sound interpretation of clinical trials: absence of evidence for a difference is not automatically evidence that differences are small enough to be unimportant.

The trial also highlights a challenge with reporting time-to-event outcomes as means and standard deviations. Overall survival is defined using event timing and censoring, yet the registry does not report a survival curve, hazard ratio, or full time-to-event model. Such summaries can convey group averages but do not expose the timing and censoring patterns needed to understand the entire survival experience.

Finally, planning calculations based on the observed effect can be unstable when the estimate is very small. The resulting illustrations can become very large, but they are not the trial’s original design rationale and do not establish what sample size would be appropriate for a future study. A future target should come from a clinically justified effect and explicit design assumptions, none of which are reported in the registry record. The case underscores the value of prespecified analysis plans and complete reporting.

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.

The primary outcome summary is limited to group means, standard deviations, a mean difference, a confidence interval, and an unadjusted p value. The registry record does not report a survival curve, hazard ratio, event count, or participant-level event and censoring information. Because overall survival is a time-to-event endpoint, the posted mean comparison does not reveal the full pattern of survival times or follow-up.

The primary analysis used the safety population and summarized participants according to treatment received. A separate intention-to-treat result is not reported in the registry record. Without information on deviations from randomized assignment, treatment exposure, and outcomes among participants excluded from the safety population, the implications of that population definition cannot be assessed.

The study status is terminated, but the reason for termination is not reported in the registry record. The treatment descriptions are truncated, so the complete schedules and treatment discontinuation rules are not available in the record excerpt. Randomization procedures, allocation concealment, adherence, and site-level conduct are also not reported.

The registry record does not report the original sample-size rationale, a prespecified meaningful-difference threshold, the statistical analysis plan, or interim monitoring procedures. It does not describe missing-data methods for the secondary outcomes or a multiplicity strategy. These gaps limit assessment of whether analyses were prespecified and how reliably secondary findings should be interpreted.

Finally, the posted safety counts do not provide detailed event timing, severity beyond the serious-event classification, treatment attribution, or causes of death. Counts and listed event terms describe reported occurrences, but they do not establish causal relationships. The small arm sizes also mean that uncommon events and differences in event frequencies may be estimated imprecisely, although the registry record does not provide uncertainty intervals for those safety comparisons.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Study of TAC-101 as Second Line Treatment in Patients With Advanced Hepatocellular Carcinoma Who Received Sorafenib as First Line Therapy — NCT00687596.
• Registry record last updated 2024-09-19; status last verified 2024-08.
• Results first posted 2022-02-01.
• 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 randomized, double-blind, placebo-controlled phase 2 study of TAC-101 as second-line treatment after sorafenib for advanced hepatocellular carcinoma. Enrollment was 52, and the study status is terminated. The primary outcome was overall survival, defined from randomization to death, with living participants censored at their last known date alive.

Mean overall survival was 8.23 months with TAC-101 and 8.35 months with placebo. The mean difference was -0.12 months, with a 95% confidence interval from -3.044 to 2.804 months and an unadjusted p value of 0.936. The interval includes differences in either direction, so the posted estimate is uncertain and does not establish equivalence or no effect.

Secondary summaries and safety counts provide additional context, but their interpretation is limited by unadjusted comparisons and incomplete reporting of analysis methods. The reason for termination, original sample-size rationale, full analysis plan, and several details needed to assess missingness and multiplicity are not reported in the registry record.