Cabozantinib and Pembrolizumab for the First-Line Treatment of Advanced Liver Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cabozantinib and Pembrolizumab for the First-Line Treatment of Advanced Liver Cancer (NCT04442581), a na phase 2 study. The primary outcome is Number of Participants With Objective Response (Complete or Partial Response).

2. Clinical Question

Statistically, the design is descriptive rather than a controlled treatment comparison. A response count from a single group can characterize outcomes among enrolled participants, but it cannot by itself separate a treatment-associated effect from other explanations. The registry record also lists disease control, progression-free survival, overall survival, and adverse events as secondary outcomes, but the posted information does not give numerical results for those outcomes.

3. Trial Design

This was a phase II, single-group trial of cabozantinib and pembrolizumab for first-line treatment of advanced hepatocellular carcinoma. The sponsor was the University of Washington. The registry lists enrollment of 2, a status of terminated, a start date of 2021-04-20, and a primary completion date of 2021-12-13. The reason for termination is not reported in the registry record.

All participants were assigned to the described treatment group: cabozantinib S-malate and pembrolizumab. Cabozantinib was taken by mouth once daily on days 1–21, and pembrolizumab was given intravenously over 30 minutes on day 1. Cycles repeated every 21 days for up to 2 years, unless disease progression or unacceptable toxicity occurred. The registry lists masking as none. Separately, the primary endpoint description says response was assessed by blinded central review; this does not make the overall trial a masked comparison.

Because the design has no separate control arm, it does not provide a randomized or concurrent comparison against cabozantinib alone, pembrolizumab alone, or another treatment. The registry’s description of the combination’s rationale is background, not a result from a control group in this study. Enrollment of 2 also sharply limits how much can be inferred about the range of outcomes in the intended population.

The registry lists five outcome measures: objective response as the primary outcome, and disease control, progression-free survival, overall survival, and adverse events as secondary outcomes. It reports no analyses in the record. The statistical plan, including any prespecified inferential approach, is not reported in the registry record. The record does not state why the trial ended or whether its early termination was connected to enrollment, safety, efficacy, or operational considerations.

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.

Stage 1

Stage 2

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 Objective Response (Complete or Partial Response), measured over Assessed at baseline, every 2 cycles until cycle 4, then every 3 cycles until disease progression, up to 10 months 16 days..

The primary endpoint is the number of participants with an objective response, defined as complete response or partial response. Assessments were planned at baseline and at intervals described in the registry: every 2 cycles until cycle 4, then every 3 cycles until disease progression, up to 10 months 16 days. Target lesions were evaluated by cross-sectional imaging using CT or MRI and Response Evaluation Criteria in Solid Tumors version 1.1, with blinded central review described.

In the registry’s definitions, complete response means disappearance of all target lesions. Partial response requires at least a 30% decrease in the sum of the longest diameters of target lesions, using the baseline sum as the reference. The endpoint excludes stable disease from the objective-response count. The registry also describes overall response rate as the percentage of complete and partial responses, but the primary outcome is titled as a participant count.

As a general statistical concept, an objective-response endpoint gives a categorical summary of whether participants meet specified imaging criteria during assessment. It can be useful for describing tumor response without requiring a time-to-event outcome. A count does not convey the duration of response, and it is not interchangeable with disease control, which also includes stable disease. In this record, the primary response count is not reported, so the endpoint’s observed value and its precision cannot be assessed from the posted information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation depends on the question the design can answer and on the information actually reported. This trial is described as phase II and single group, with allocation not applicable and masking none. Its primary endpoint is a participant count for objective response. The registry record does not report the outcome count, confidence interval, or analysis method, so the intended endpoint can be described but its numerical result cannot be evaluated.

What does a single-group design estimate?

As a general statistical concept, a single-group study can describe outcomes among people who received the study treatment. It does not directly estimate a difference between treatment strategies because there is no separate comparison group. A response count, if reported, would be a description of the enrolled participants under the stated assessment process. It would not by itself establish what would have happened to comparable participants under another treatment.

That distinction matters because a treatment comparison requires a contrast. In a randomized study, random assignment is used to make groups comparable in expectation, subject to the study’s conduct and size. This record does not describe randomization or a control arm. The registry’s background statement about possible comparison with the medicines alone is not a result from such arms in this study.

How should the response endpoint be read?

General explanation: a categorical response endpoint classifies each assessed participant according to prespecified criteria. Here, the registry identifies RECIST version 1.1 and target-lesion imaging, and defines complete and partial response. The objective-response count includes those two categories, not stable disease. Disease control is listed separately and includes complete response, partial response, and stable disease. These endpoints answer different descriptive questions and should not be treated as interchangeable.

The record describes imaging by CT or MRI and blinded central review. Central review can be used to apply assessment criteria consistently, while blinding reviewers can reduce the influence of expectations on interpretation. The registry separately lists overall masking as none. Those statements concern different aspects of the study: the endpoint description identifies blinded review, while the design field does not describe masking of the trial overall.

What do uncertainty and precision mean?

General explanation: an estimate is more informative when its uncertainty is also reported. A confidence interval expresses a range of values compatible with the data under a specified statistical procedure and its assumptions. With small samples, estimates are generally less precise, and an interval may be wide. A confidence interval cannot be reconstructed from the registry information because the primary response count and interval are not reported. No interval should be inferred from enrollment alone.

The same caution applies to safety. Counts and percentages in a small group can change substantially with the experience of an individual participant. A listed percentage is a description of the registry entry, not a precise estimate of the rate in a broader population. There is no control group from which to estimate a comparative safety difference.

What do the planning assumptions mean?

The record describes sample-size illustrations as planning calculations, not as the trial’s original design. 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 expressed as a fraction of the observed effect. For time-to-event calculations, the stated target is a total number of events, while other calculations are expressed as participants per arm.

These are hypothetical planning assumptions, not evidence that the trial had a particular power. The trial itself is recorded as single group, so an equal-allocation two-arm illustration does not describe its actual allocation. The calculations are not observed power, and they do not supply missing primary outcome results. General explanation: planning calculations describe what a hypothetical design might require under specified assumptions; they do not show that those assumptions were used in the actual trial. The registry does not report the study’s own sample-size justification or statistical plan.

What is not available for assessment?

The registry record reports no analyses. It does not state how missing assessments were handled, which analysis population was intended, whether any interim analyses were planned, or how multiplicity was addressed. These omissions prevent a full evaluation of endpoint handling and inferential procedures. They should be treated as unreported details, not as evidence that a particular method was or was not used.

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("Treatment (Cabozantinib S-malate, Pembrolizumab)")
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  Treatment (Cabozantinib S-malate, Pembrolizumab)  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 one primary outcome, objective response, and four secondary outcomes: disease control, progression-free survival, overall survival, and adverse events. Several outcomes can raise a multiplicity issue when formal hypothesis tests are conducted, because multiple tests create more opportunities for chance findings. General statistical practice may address this through a prespecified testing strategy or adjustment, but no such strategy is reported in the registry record. The record also reports no analyses, so there are no posted test results to interpret for multiplicity.

The registry describes when response assessments were to occur, but does not report how missing or unevaluable imaging assessments were handled. It does not say whether the primary analysis included all enrolled participants, only participants with evaluable imaging, or another analysis population. The handling of withdrawals, treatment discontinuation, or incomplete follow-up is not reported in the registry record.

For time-to-event outcomes, the record lists progression-free survival and overall survival, but does not provide numerical results or details of censoring and follow-up. General explanation: time-to-event analyses depend on how events, censoring, and the observation period are defined. Those details cannot be evaluated from the posted information. The registry record does not report an interim-analysis plan, missing-data method, analysis population, or full statistical plan. Their absence from the record does not establish what procedures were used in practice.

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: general weakness/poor performance (50.0%), necrotizing fasciitis (50.0%).

The registry reports safety figures for the treatment group. For serious adverse events, 2 participants were at risk and 2 were affected. For other adverse events, the record also lists 2 at risk and 2 affected. One death is recorded. These are separate entries in the safety summary and should not be added together to create a total: the record does not establish whether categories overlap.

The listed serious adverse-event terms are General Weakness/Poor Performance and Necrotizing fasciitis, each reported as 50.0%. The registry does not provide the individual participant counts for these terms in the information. The percentages should therefore be read as registry-listed summaries, not as evidence of a comparative risk or a causal relationship to treatment.

General statistical explanation: adverse-event counts describe what was recorded among participants under observation. They do not, by themselves, show that treatment caused an event. A comparison group and fuller information about event definitions, follow-up, and participant-level overlap would be needed to interpret differences in risk. With enrollment of 2 and no separate comparator, these safety entries are descriptive and have limited precision. The reason for the recorded death is not reported in the registry record.

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

The registry record establishes that this phase II study evaluated cabozantinib and pembrolizumab together in a single group intended to include patients with advanced hepatocellular carcinoma not eligible for local therapy. It records enrollment of 2 and gives a treatment schedule, assessment approach, outcome list, and selected safety entries. Those details describe the trial’s scope and conduct as recorded; they do not supply an efficacy conclusion.

The primary endpoint is the count of participants with complete or partial response according to RECIST version 1.1. That count is not reported in the registry record. Numerical results for the listed secondary outcomes are also not reported in the information available for this analysis. Thus, the posted record does not show the observed objective-response result, disease-control result, progression-free survival, or overall survival.

Even if a response count were available, a single-group result would describe the participants in that group rather than estimate a randomized difference against another treatment. General statistical explanation: without a comparator, observed outcomes cannot be separated from other influences through a treatment contrast. The small enrollment also means that group summaries would provide limited information about the distribution of outcomes in a broader population.

The safety entries show that adverse events were recorded, including the serious-event terms listed in the registry, and that one death was recorded. They do not establish that the study treatment caused any event. The record does not report the cause of death, the reason the trial was terminated, or a comparative safety analysis.

Accordingly, the record can be read as a description of a small, terminated, single-group study with missing posted primary and secondary outcome results. It does not support a claim that the combination does or does not work, nor does it establish comparative benefit or harm. Those conclusions would require outcome data and a design and analysis capable of addressing the relevant comparison.

14. Why This Trial Matters Statistically

This trial matters statistically because its design and the availability of results place clear boundaries on interpretation. A phase II label does not replace the need to examine the actual design, enrollment, endpoint, and reported analysis. Here, the study is single group, enrollment is 2, and the primary objective-response count is not reported in the registry record. These features mean that neither a numerical response estimate nor a controlled treatment contrast can be assessed from the posted information.

The endpoint itself illustrates why definitions matter. RECIST version 1.1 distinguishes complete response, partial response, stable disease, and progression using imaging criteria. The primary objective-response endpoint counts only complete and partial responses, while the secondary disease-control outcome also includes stable disease. A clear distinction among these outcomes is necessary when interpreting any reported count.

The trial also illustrates the difference between planning calculations and observed evidence. The registry describes planning illustrations under assumptions including a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Those assumptions are not a description of the single-group trial’s actual allocation, and the calculations are not observed power. The study’s own sample-size justification is not reported in the registry record.

Finally, safety summaries require careful reading when the group is small and lacks a comparator. Counts and listed percentages can describe recorded events, but they do not show causation or quantify comparative risk. The record’s termination status adds context, while its reason is not reported. Together, these points make the study a useful example of how registry design details, missing outcomes, and limited enrollment constrain statistical conclusions.

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 record restrict interpretation. Enrollment was 2, and the trial is described as single group. There is no separate comparator arm, so the design cannot estimate a randomized contrast between the combination and another treatment. The registry lists the study as terminated, but the reason for termination is not reported in the registry record.

The central limitation is that the primary objective-response count is not reported. The record also does not provide numerical results for the listed secondary outcomes. Without those values, the magnitude and precision of the observed outcomes cannot be assessed. The registry record reports no analyses, and the statistical plan, analysis populations, handling of missing assessments, and any interim-analysis plan are not reported.

The primary endpoint description specifies RECIST version 1.1, imaging of target lesions, and blinded central review, but the posted information does not show how many participants had evaluable assessments or how incomplete assessments were handled. The registry’s overall masking field is none; the endpoint description separately says review was blinded. The record does not give further details that would reconcile or expand on those design features.

The safety information is limited to the listed group summaries and serious-event terms. The record does not provide enough detail to assess participant-level overlap among events or to determine the cause of the recorded death. With no comparator, these figures cannot establish comparative safety. The record also does not report a reason for termination, so no explanation for the small enrollment or absent outcome results should be assumed.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Cabozantinib and Pembrolizumab for the First-Line Treatment of Advanced Liver Cancer — NCT04442581.
• Registry record last updated 2023-04-05; status last verified 2023-03.
• Results first posted 2023-04-05.
• 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 describes a phase II, single-group study of cabozantinib S-malate and pembrolizumab as first-line treatment for advanced hepatocellular carcinoma. The sponsor was the University of Washington. Enrollment was 2; the study is listed as terminated, with a start date of 2021-04-20 and a primary completion date of 2021-12-13. The reason for termination is not reported in the registry record.

The primary outcome was the number of participants with complete or partial response under RECIST version 1.1. The registry describes blinded central review and a CT or MRI assessment schedule. The primary outcome count and numerical secondary-outcome results are not reported. The record lists disease control, progression-free survival, overall survival, and adverse events as secondary outcomes, and reports no analyses.

Safety summaries list 2 participants at risk and 2 affected for both serious and other adverse events, one death, and two serious-event terms each reported as 50.0%. These figures are descriptive and do not provide a treatment comparison or establish causation.