Safety Study of Liver Natural Killer Cell Therapy for Hepatoma Liver Transplantation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety Study of Liver Natural Killer Cell Therapy for Hepatoma Liver Transplantation (NCT01147380), a non-randomized phase 1 study. The primary outcome is Side Effect of Cadaveric Donor Liver NK Cell Infusion.

2. Clinical Question

The study also listed anti-tumor effect against hepatocellular carcinoma and anti-HCV effect, if applicable, among its secondary outcomes. Those aims are distinct from the primary safety question. The registry record lists no analysis results for them, so the record does not establish how either outcome was assessed or what was observed.

3. Trial Design

This was a phase 1, nonrandomized, parallel-group study with no masking. The registry record lists enrollment of 18 and describes a small-dose group and a large-dose group. In the adverse-event record, six participants are listed as at risk in the small-dose group and 12 in the large-dose group. These are different group sizes, and assignment was not randomized.

The intervention used natural killer cells obtained from donor liver perfusate. The cells were cultured and stimulated with IL-2 before being given to a liver transplant recipient associated with the same donor. The small-dose description gives an inoculation range between 10 and 100 million cells; the large-dose description gives a range between 100 and 1000 million cells. The group labels therefore reflect different intended cell-dose ranges, not random assignment to competing treatments.

The study objective was to evaluate feasibility and safety of adoptive transfer of activated natural killer cells for liver transplant recipients with hepatocellular carcinoma. The stated primary endpoint concerned side effects during one year. Listed secondary outcomes concerned infusion-related toxicity, anti-HCC effect, and anti-HCV effect if applicable.

The registry record lists the study as completed, with a start date of 2010-06 and primary completion in 2014-12. It does not report a protocol-level account of how participants entered each dose group, whether dose escalation rules were used, or how follow-up was handled beyond the stated endpoint timeframe. The nonrandomized design means that differences between the groups cannot be interpreted as a randomized dose effect.

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 Side Effect of Cadaveric Donor Liver NK Cell Infusion, measured over 1 year.

The primary endpoint is titled “Side Effect of Cadaveric Donor Liver NK Cell Infusion.” It is specified as a number of participants, with a timeframe of one year. The description says the study would monitor patient condition clinically, record any noticed side effects, and report them to the data safety monitoring committee. This is a safety-focused endpoint: it is intended to describe whether participants experienced side effects during follow-up, rather than to estimate an anti-tumor effect.

Counting participants with an event is a straightforward way to summarize safety observations. It indicates how many people in a group were recorded as affected, rather than how many separate episodes occurred. The registry record does not give a more detailed endpoint definition, such as explicit criteria for classifying a side effect, severity grading, or rules for attributing an event to the infusion. Those details are not reported in the registry record.

The registry’s adverse-event entries list zero serious and zero other affected participants in each dose group. These entries are relevant safety observations, but the registry record does not include a formal analysis of the primary endpoint or clarify whether the adverse-event categories capture every side effect encompassed by the primary endpoint. The recorded counts should therefore be read within the limits of the stated categories and available follow-up information.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record describes a count endpoint: the number of participants with a side effect. For a binary participant-level outcome, each person is classified as having or not having the event under the endpoint definition. The resulting group summary is a count of affected participants together with the number at risk. It is different from an event count, which could count repeated episodes in the same participant. The registry record does not say whether repeated events were counted separately for the primary endpoint.

For the posted adverse-event categories, the record gives both the number at risk and the number affected in each group. The small-dose group has six participants at risk, and the large-dose group has 12. The affected counts are zero for serious events and zero for other events in each group. These observations can be described directly without fitting a statistical model. The registry record lists no formal analyses, so there is no reported model-based estimate or inferential test to interpret.

In general, a confidence interval gives a range of values compatible with the observed data under a specified statistical method and assumptions. It communicates uncertainty around an estimate; it is not a guarantee that the true value lies in that range. A confidence interval for an event frequency would be useful for understanding how much uncertainty remains when the observed count is zero. No such interval is reported in the registry record, and no interval should be inferred from the counts alone in this description.

In general, a hypothesis test evaluates how compatible data are with a stated null hypothesis under a statistical model. A p-value is not the probability that the null hypothesis is true, nor does it measure clinical importance. The registry record lists no hypothesis tests or p-values for the safety outcome. There is therefore no basis to characterize the primary result as statistically significant or not statistically significant.

The group labels refer to different cell-dose ranges, but dose assignment was nonrandomized. Randomization is a general design method that can balance known and unknown prognostic factors across groups in expectation. Without random assignment, differences between groups may reflect who received each dose as well as the dose itself. A direct comparison of their observed counts cannot isolate a causal dose effect. The registry record does not report adjustment methods or a plan for controlling confounding.

Unequal group sizes also matter for descriptive precision. A count based on a larger group generally carries different sampling uncertainty from the same count based on a smaller group. Neither group’s zero count proves the absence of risk. General statistical methods can quantify uncertainty, but their interpretation depends on the endpoint definition, follow-up completeness, and assumptions about how participants were observed. The record does not report enough detail to assess those assumptions or to reconstruct a formal estimate.

The study is described as parallel-group, meaning that participants were assigned to one of the dose groups rather than being described as crossing between both treatments. A parallel structure does not itself make a comparison randomized. The lack of masking is also relevant in general: knowledge of treatment can influence observation, reporting, or classification of subjective outcomes. The registry record says masking was none, but it does not describe a blinded event-adjudication process.

The planning assumptions described in the record refer to a two-sided significance level of 0.05, equal allocation, and a normal approximation for an illustration based on an observed effect. Such planning assumptions are not the same as the study’s actual design or an observed treatment effect. The record does not report a sample-size justification for the trial, and no planning result should be interpreted as observed power. Nor does the presence of planning assumptions establish that the trial used a particular hypothesis test for its outcomes.

More broadly, statistical precision and clinical interpretation are separate. A precise estimate can still be affected by bias or limited applicability, while a descriptive count can be informative without supporting a causal conclusion. For this study, the available results support reporting the recorded counts and group sizes. They do not support a claim that a dose caused a particular safety outcome, or a claim about the frequency of events beyond the participants and categories recorded.

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("Small Dose", "Large Dose")
n      <- c(6, 12)
events <- c(0, 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  Small Dose  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0
ARM  Large Dose  12  0  0.0  0.0  26.5  0.0  24.2  0.2  24.7

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists four outcome measures and names three secondary outcomes: infusion-related toxicity, anti-HCC effect, and anti-HCV effect if applicable. The primary outcome concerns side effects of cadaveric donor liver natural killer cell infusion. No analysis results are listed, so the record does not show which outcomes were analyzed, how many comparisons were made, or whether any outcome was selected for emphasis after data collection.

Multiplicity arises when several outcomes, comparisons, or repeated tests are considered. In general, conducting many tests can increase the chance of at least one apparently notable result under a null model. Statistical plans may address this through prespecified outcome hierarchies or adjustment procedures. The registry record does not report a multiplicity strategy, outcome hierarchy beyond identifying a primary endpoint, or adjustment method. There are no reported p-values to assess in this regard.

Missing-data handling is also not reported in the registry record. The adverse-event entries give numbers at risk, but do not state whether every participant completed the one-year monitoring period, how withdrawals or incomplete follow-up were treated, or whether any observations were missing. No imputation method or sensitivity analysis is reported. Without those details, the counts cannot establish that all participants contributed complete follow-up.

The record does not report analysis populations, such as whether results were intended to include every enrolled participant, only those who received an infusion, or another defined group. It also does not describe protocol deviations or exclusions from analysis. The denominator in an adverse-event summary should therefore be understood as the number listed at risk in that category, not as proof of a particular analysis-population rule.

Interim analyses, stopping rules, and a detailed statistical analysis plan are not reported in the registry record. These omissions limit assessment of how outcomes and analyses were prespecified. They do not change the descriptive counts, but they constrain conclusions about completeness and inferential interpretation.

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 0% reporting threshold.

Most frequent non-serious adverse event terms

The registry’s adverse-event entries report serious-event counts and other-event counts separately. In the small-dose group, six participants are listed at risk, with zero affected by serious events and zero affected by other events. In the large-dose group, 12 participants are listed at risk, also with zero affected in each category. The registry record lists no serious adverse-event terms.

These counts describe recorded participants, not the number of events per person and not a complete measure of every possible safety outcome. The record does not give event descriptions for the other-event category, severity details, attribution assessments, or participant-level follow-up. It also does not say whether every participant completed the full one-year monitoring period. Those details are not reported in the registry record.

Zero observed events should not be interpreted as proof that no risk exists. A small study can record no events while leaving uncertainty about uncommon outcomes. The counts also do not establish equivalence between dose groups: the study was nonrandomized, the groups have different sizes, and they are not a like-for-like control comparison. The appropriate reading is descriptive—no affected participants were recorded in these listed categories—rather than a definitive estimate of comparative safety.

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

The posted results provide a limited descriptive safety signal: the adverse-event record lists zero participants affected by serious events and zero affected by other events in both dose groups. The record identifies six participants at risk in the small-dose group and 12 in the large-dose group. The endpoint was intended to monitor side effects over one year, but the registry does not provide a formal analysis of that endpoint or details establishing complete follow-up for every participant.

This result does not establish that the infusion has no risks. In general, a study with no recorded events can still be compatible with a range of underlying event frequencies, particularly when the number of participants is limited. Without a reported confidence interval, the precision of the event-frequency estimates cannot be assessed from the registry record.

It also does not show that the small dose and large dose have equivalent safety. The study was nonrandomized and had no masking, and the dose groups are not a like-for-like control comparison. Participant characteristics, clinical circumstances, monitoring, or event recording could differ between groups. The available record does not provide the details needed to examine those possibilities.

The study’s stated aims included feasibility and possible anti-tumor and anti-HCV effects, but the registry record lists no analyses. The adverse-event counts do not answer whether the intervention affected hepatocellular carcinoma or hepatitis C outcomes, and no such effect should be inferred from them. The registry record also does not report a detailed statistical plan or full endpoint definitions.

Accordingly, the result is best treated as an early-phase descriptive observation. It documents the adverse-event counts recorded in the registry, while leaving substantial questions about precision, ascertainment, follow-up, and generalizability unresolved. It is not a definitive estimate of safety or a causal comparison between dose levels.

14. Why This Trial Matters Statistically

This study is statistically informative because it illustrates how a zero-event result can be meaningful to report without being conclusive. Counts of zero are observations, not proof that an event is impossible. Their interpretation depends on how many participants were observed, how completely they were followed, and how events were defined and recorded. The registry record provides group sizes and adverse-event counts, but not the detail needed to resolve all of those questions.

The design also highlights the distinction between a dose-group description and a controlled comparison. The study is nonrandomized, so the small-dose and large-dose groups may differ for reasons beyond their intended cell-dose ranges. A numerical contrast between groups, even if calculated, would not by itself separate dose effects from differences in participant selection or clinical circumstances. This is a general issue in interpreting nonrandomized treatment groups.

Early-phase studies often focus on feasibility and safety rather than a definitive estimate of benefit. In this study, the stated objective included both feasibility and safety, while anti-tumor and anti-HCV effects were listed as secondary outcomes. The record lists no analyses, so it cannot support conclusions about those secondary questions. Clear separation of aims matters: a safety count is not evidence of efficacy, and an unreported efficacy analysis cannot be reconstructed from the stated objective.

Finally, this record shows why transparent reporting of endpoint definitions, follow-up completeness, analysis populations, and uncertainty estimates matters. Such information helps readers understand what a zero count can and cannot establish. Its absence does not erase the reported observations, but it limits how precisely they can be interpreted and how confidently they can be compared with other settings.

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 registry record gives limited detail on the primary endpoint. It states that side effects would be clinically monitored and recorded, but does not report operational criteria, severity grading, attribution rules, or how events were adjudicated. The record does not clarify whether the serious and other adverse-event categories correspond fully to the primary endpoint’s definition of side effects.

Follow-up completeness is also unclear. The primary endpoint timeframe is one year, but the adverse-event entries do not report each participant’s observation duration, losses to follow-up, withdrawals, or missing outcome status. The number listed at risk does not by itself establish complete one-year observation.

The nonrandomized design, absence of masking, and unequal group sizes constrain comparisons between dose groups. The registry record does not report how treatment assignment was determined, whether groups differed at baseline, or whether any adjustment was planned or conducted. It also does not describe a separate control group, and explicitly notes that the arms are not a like-for-like control comparison.

Although the study lists secondary outcomes concerning infusion-related toxicity, anti-HCC effect, and anti-HCV effect if applicable, no analyses are listed. A statistical analysis plan, multiplicity approach, missing-data strategy, analysis populations, interim analyses, and stopping rules are not reported in the registry record. These omissions limit interpretation of the intended analyses and the completeness of the reported findings.

The adverse-event counts are small-group descriptive results, with no reported confidence intervals or formal primary endpoint analysis. They cannot establish a precise underlying event frequency or a causal difference between doses. The registry record also does not provide enough information to assess how well the findings apply beyond the participants enrolled.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Safety Study of Liver Natural Killer Cell Therapy for Hepatoma Liver Transplantation — NCT01147380.
• Registry record last updated 2016-03-10; status last verified 2016-02.
• Results first posted 2016-03-10.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The registry record describes a completed phase 1, nonrandomized, unmasked, parallel-group study of IL-2-activated donor liver natural killer cell infusion for liver transplant recipients with hepatocellular carcinoma. Enrollment is listed as 18. The dose groups are described as small dose and large dose, with different cell-dose ranges.

The primary outcome is side effects of cadaveric donor liver natural killer cell infusion, measured as participants over one year. The adverse-event entries list six participants at risk in the small-dose group and 12 in the large-dose group, with zero serious and zero other affected participants in each group. No analyses are listed.

Secondary outcomes concern infusion-related toxicity, anti-HCC effect, and anti-HCV effect if applicable. The registry record does not report analyses for these outcomes, a detailed statistical plan, or additional information on endpoint definitions and follow-up completeness. The study is listed as starting in 2010-06 and reaching primary completion in 2014-12.