IMMULAB - Immunotherapy With Pembrolizumab in Combination With Local Ablation in Hepatocellular Carcinoma (HCC): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for IMMULAB - Immunotherapy With Pembrolizumab in Combination With Local Ablation in Hepatocellular Carcinoma (HCC) (NCT03753659), a na phase 2 study. The primary outcome is Objective Response Rate (ORR) According to RECIST 1.1.

2. Clinical Question

The study addresses a response question at a defined point in the treatment sequence: after pembrolizumab and before ablation. The registry rationale is that response at this stage could test whether pre-interventional treatment is associated with conversion or downstaging of borderline candidates for local ablation. This endpoint therefore concerns radiological tumour response before the local procedure, not recurrence after ablation or survival.

The registered study also lists time to recurrence, recurrence-free survival, overall survival, adverse events, and molecular biomarker analyses among its secondary outcomes. These address different questions and time horizons. They do not substitute for a reported primary endpoint result. The objective response result, including the number of participants with complete or partial response and the number assessed, is not reported in the registry record.

3. Trial Design

IMMULAB is registered as a phase II, multicenter, prospective, open-label, single-group trial in patients with early-stage HCC who were candidates for local ablation. The registry lists an enrollment of 30 and a completed status. The recorded start date is 2019-05-09 and the primary completion date is 2024-04-30.

The intervention combined pembrolizumab with a local ablation approach. The arm description specifies pembrolizumab 200mg intravenously every 3 weeks on day 1 of cycle 1 and cycle 2. The local procedure was scheduled for day 1 of cycle 3 and could be radiofrequency ablation, microwave ablation, brachytherapy, or a combination of transarterial chemoembolisation with one of those procedures. The description also specifies pembrolizumab administration 2 days after local ablation. Further dosing details are not reported in the registry record.

There was no random allocation and no masking. The design is therefore not a randomized comparison: outcomes in the single group cannot establish how the same participants would have fared under a different treatment strategy. The trial can describe observed outcomes in its enrolled group, but comparisons with an external population would require information and assumptions beyond the registry record.

The primary endpoint was assessed after 6 weeks (2 cycles) of treatment, before local ablation. The registry describes response assessment by radiological imaging using RECIST 1.1. It gives no primary response result, analysis population, or detailed account of how missing response assessments were handled. The registry record also does not report a sample-size justification, interim analysis plan, or a more detailed statistical analysis plan.

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

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 Objective Response Rate (ORR) According to RECIST 1.1, measured over After 6 weeks (2 cycles) of treatment.

The primary endpoint is objective response rate (ORR) according to RECIST 1.1, expressed as a count of participants. A participant is classified as having an overall response when the assessment meets the criteria for complete response or partial response. Complete response is defined as disappearance of all lesions. Partial response requires at least 30% decrease in the sum of the longest diameter of target lesions, with no new non-target lesions and no progression in non-target lesions.

The endpoint is assessed after 6 weeks (2 cycles) of pembrolizumab and before local ablation. This timing makes the measure relevant to the registered question about pre-interventional response, rather than response after the local procedure. The registry rationale states that assessing ORR at this point could test the hypothesis that pre-interventional pembrolizumab may be associated with conversion or downstaging of borderline candidates for local ablation. It also notes that ORR is generally accepted as an efficacy endpoint in one-armed trials without a control group.

A response count alone does not convey the precision of the estimate. Interpretation would ordinarily depend on the number assessed, the number responding, and an uncertainty interval, as well as how participants without an evaluable assessment were treated. Those primary response results and details are not reported in the registry record. Thus, the endpoint definition and its assessment time are available, but the magnitude and precision of the observed ORR cannot be read from the record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a single-group design estimate?

In a single-group study, participants receive the registered intervention strategy and outcomes are described within that group. For a binary endpoint such as response, the basic observed quantity is the number of participants meeting the response definition among those evaluated. A proportion can be useful for describing the group, but it is not a randomized treatment contrast. Without a concurrent comparator, differences from another group cannot be attributed to treatment using the trial design alone.

IMMULAB is registered as single-group, with no random allocation and no masking. The registry therefore does not support a like-for-like comparison between alternative strategies. A response estimate, if reported, would describe the enrolled and assessed participants under the trial conditions. It would not by itself show what would have happened to those same participants under another approach.

How should a binary response endpoint be read?

ORR combines complete and partial responses under RECIST 1.1. The criteria classify changes in lesions using radiological assessment. Complete response requires disappearance of all lesions; partial response requires at least 30% decrease in the sum of the longest diameter of target lesions, alongside the stated conditions for non-target lesions and new lesions. These rules provide an operational definition, but the registry record does not report the response observations or the number included in the endpoint analysis.

For a binary outcome, the count and denominator are both essential. The count indicates how many participants met the endpoint; the denominator indicates how many were evaluated or otherwise included. A confidence interval describes the uncertainty around an estimated response proportion. General statistical principle: greater uncertainty is expected when information is limited, while the interval also depends on the observed response pattern. No primary response estimate or interval is reported in the registry record, so neither the result nor its precision can be evaluated.

Why does endpoint timing matter?

The registered assessment occurs after 6 weeks (2 cycles) of pembrolizumab and before local ablation. This timing isolates the endpoint conceptually from the subsequent local procedure: the primary response question concerns the period before ablation. It is not the same outcome as recurrence after treatment or overall survival, which are listed separately as secondary outcomes. An endpoint measured at a particular time answers a question about that time and should not be treated as a direct measure of later outcomes.

What can a confidence interval add?

General statistical principle: an estimate is a best summary of the observed data under a specified method, while a confidence interval communicates how much uncertainty is compatible with those data under that method. It does not remove bias from study design, missing observations, or measurement choices. In a single-group trial, an interval around a response estimate would express uncertainty about the group outcome; it would not create the missing control comparison.

How should planning illustrations be understood?

General statistical principle: sample-size calculations depend on design choices and assumptions about the effect to detect, significance level, allocation, and statistical method. A planning illustration based on an observed effect is retrospective and should not be mistaken for the original trial's design justification. It also is not observed power. The registry record does not report the trial's own sample-size justification. The assumptions described for planning illustrations are not evidence that those assumptions were used to design this trial.

Over-dispersion is a general concern for count outcomes when variability exceeds what a chosen probability model expects. It can make standard errors too small if the model does not account for the extra variation. That issue is not the main structure of this primary endpoint, which is defined as a binary response count, and no model-based primary analysis is reported in the registry record. The record lists no analysis entries for the outcome results.

Other methodological details that would affect interpretation include the handling of unevaluable imaging, whether response assessments were reviewed centrally, and the exact analysis population. Those details are not reported in the registry record. Their absence means that the endpoint definition is clearer than the analysis implementation and resulting estimate.

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("Pembrolizumab With Local Ablation")
n      <- c(30)
events <- c(4)
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  Pembrolizumab With Local Ablation  30  4  13.3  3.8  30.7  5.3  29.7  5.5  29.8

11. Multiplicity, Missing Data and Analysis Populations

The registry lists several secondary outcomes in addition to the primary ORR endpoint: time to recurrence according to RECIST 1.1, recurrence-free survival according to RECIST 1.1, overall survival, adverse events, and molecular biomarker analyses. The record does not report how statistical testing across these outcomes was organized or whether any adjustment for multiple testing was planned. A multiplicity strategy is not reported in the registry record.

Multiplicity matters when several outcomes or comparisons are tested because the chance of at least one apparently notable result can increase as testing expands. General statistical principle: an adjustment or a prespecified hierarchy can help control how results are interpreted across multiple tests. Without a reported plan, the evidential status of any secondary findings cannot be judged from the registry record alone. No secondary analysis results are included in the record.

Missing data handling is also not reported. The primary outcome population field is empty, and the record does not state how participants without an evaluable scan were treated, whether they were excluded from the response analysis, or whether another rule was used. These choices can affect a binary response estimate because the denominator and classification of unassessed participants matter.

The registry record does not describe intention-to-treat, per-protocol, or other analysis populations for the primary endpoint. It also does not report an interim analysis plan or detailed statistical analysis plan. These omissions limit assessment of prespecification and of how deviations, incomplete assessments, or multiple outcomes were handled. They do not change the registered endpoint definition, but they leave important features of its analysis unspecified.

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: acute kidney injury (10.0%), ascites (6.7%), hepatic failure (6.7%), encephalopathy (6.7%), anemia (3.3%).

The registry reports that, among 30 participants at risk in the pembrolizumab with local ablation group, 14 had at least one serious adverse event and 28 had at least one other adverse event. The recorded death count is 4. These are participant counts, not event totals; a participant may have experienced more than one event. The record does not state whether the deaths or adverse events were judged related to treatment.

The listed serious adverse-event terms and percentages are acute kidney injury, 10.0%; ascites, 6.7%; hepatic failure, 6.7%; encephalopathy, 6.7%; anemia, 3.3%; and myocardial infarction, 3.3%. They are selected serious-event listings, not a complete account of every clinical event or its timing. The registry record does not provide event-level detail, duration, or the basis for attribution.

Seriousness and treatment relatedness are different concepts. A serious adverse event is classified by seriousness criteria, while causality requires a separate assessment. These figures describe reported safety observations in the study group; they do not establish that the intervention caused an event. The absence of a comparator also means that these counts cannot be interpreted as comparative risks.

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

The registry record gives a clear outline of a phase II, single-group study combining pembrolizumab with local ablation in early-stage HCC. It defines a primary radiological response endpoint at a specific point before ablation and records participant counts for adverse events and deaths. These features allow readers to understand the intended clinical question and the broad treatment sequence.

The record does not include the primary ORR result. It therefore does not show the observed number or proportion of participants with complete or partial response, nor does it show an uncertainty interval. The enrollment figure cannot be substituted for the endpoint denominator, because the number with an evaluable assessment is not reported. No conclusion about the magnitude or precision of the primary response outcome follows from the endpoint definition alone.

Even if an ORR estimate were available, the single-group design would describe outcomes under the trial strategy rather than provide a randomized comparison. The study has no concurrent control group, and the record does not report a causal estimate relative to another treatment strategy. The endpoint was assessed before local ablation, so it should also be distinguished from recurrence and survival outcomes listed as secondary measures.

The safety counts should likewise be read as descriptive reports from this group. The record does not state whether events were treatment-related and does not provide comparative safety data. Overall, the posted information supports a description of the design, endpoint definition, and selected safety findings. It does not establish the efficacy of the intervention or its causal relationship to adverse outcomes.

14. Why This Trial Matters Statistically

This trial illustrates a common statistical challenge in early-phase clinical research: a single-group study can characterize outcomes and provide a basis for further investigation, but it cannot by design make the same controlled comparison as a randomized trial. That distinction is especially important when an endpoint such as ORR is described as a valid measure for a one-armed trial. Endpoint validity and causal interpretation are related but separate issues.

The timing of the primary endpoint is also central. IMMULAB defines response after pembrolizumab and before local ablation, aligning the measure with the stated question about pre-interventional response and possible conversion or downstaging. This makes clear what the endpoint is intended to capture, while also distinguishing it from later recurrence and survival outcomes.

The registry record demonstrates why endpoint definitions alone are not enough for interpretation. Knowing the RECIST 1.1 criteria does not reveal the number of participants who responded, the number assessed, or the uncertainty around the result. Reporting those elements, alongside analysis populations and missing-data methods, is necessary for readers to judge the estimate and its limitations.

Finally, the safety record emphasizes the importance of separating event counts from causal claims. Participant-level counts can describe what was recorded in a study group, but without attribution information and a comparison group they do not quantify comparative risk. These distinctions are broadly relevant when interpreting open-label, single-group phase II studies.

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 most consequential limitation of the registry information is that the primary response result is not reported in the registry record. The endpoint definition and assessment time are stated, but no response count, denominator, estimate, or confidence interval is available. The number of participants with evaluable imaging is also not reported.

The single-group, open-label design has no randomized comparator, so it cannot isolate the effect of pembrolizumab from other explanations for observed outcomes. The registry identifies candidates for local ablation but does not provide detailed eligibility criteria or further characteristics of the primary outcome population. This limits assessment of how broadly any group-level result would apply.

Important analysis details are not reported, including the handling of missing or unevaluable assessments, the analysis population, and a detailed statistical analysis plan. The registry record also does not describe multiplicity handling for the several secondary outcomes. The absence of these details limits assessment of whether outcome analyses followed prespecified rules.

For safety, the record provides participant counts, selected serious-event terms, and a death count, but does not report treatment attribution, event timing, or event-level clinical detail. There is no control group for comparative safety interpretation. These limitations are specific to the available registry record; they do not determine what may be available in other trial reporting.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: IMMULAB - Immunotherapy With Pembrolizumab in Combination With Local Ablation in Hepatocellular Carcinoma (HCC) — NCT03753659.
• Registry record last updated 2026-03-23; status last verified 2026-03.
• Results first posted 2026-03-23.
• 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

IMMULAB is a completed, multicenter, prospective, open-label phase II single-group trial of pembrolizumab combined with local ablation for early-stage HCC. The registry lists enrollment of 30, a start date of 2019-05-09, and a primary completion date of 2024-04-30.

The primary endpoint is ORR according to RECIST 1.1 after 6 weeks (2 cycles) of pembrolizumab and before local ablation. The response result is not reported in the registry record. Listed secondary outcomes include recurrence, survival, adverse events, and biomarker analyses.

The registry reports 14 participants with at least one serious adverse event, 28 with at least one other adverse event, and 4 deaths among 30 at risk. It does not report whether events or deaths were attributed to treatment. No randomized comparison is available.