Pembrolizumab With Liver-Directed or Peptide Receptor Radionuclide Therapy for Neuroendocrine Tumors and Liver Metastases: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pembrolizumab With Liver-Directed or Peptide Receptor Radionuclide Therapy for Neuroendocrine Tumors and Liver Metastases (NCT03457948), a non-randomized phase 2 study. The primary outcome is Proportion of Participants With an Overall Response.

2. Clinical Question

This pilot phase II study investigates response and safety when pembrolizumab is combined with a treatment directed at liver metastases or with peptide receptor radionuclide therapy. The clinical question is about the observed response within each treatment group, rather than a randomized estimate of the difference between treatments. The registry record specifies that overall response rate is the proportion of the analysis population with a complete response or partial response. It calls for a separate estimate and 90% confidence interval for each of the three groups.

The study description gives a rationale for combining pembrolizumab with liver-directed treatment or peptide receptor radionuclide therapy, but that rationale is not evidence that one combination is more effective than another. Because treatment assignment was non-randomized, observed group outcomes may reflect differences in who entered each group as well as treatment. The registry record does not report the primary response estimates, so the clinical question cannot be answered from the recorded outcome information.

3. Trial Design

This was a phase II pilot study with parallel groups and non-randomized allocation. The registry lists an enrollment of 32 and a completed status. The recorded start date is 2018-08-27, and the primary completion date is 2025-07-31. The study concerned well-differentiated neuroendocrine tumors with symptomatic and/or progressive liver metastases.

All groups included pembrolizumab alongside a second treatment approach. Group I combined pembrolizumab with intravenous peptide receptor radionuclide therapy using 177Lu-DOTATATE. Group II combined pembrolizumab with transarterial embolization, and Group III combined pembrolizumab with yttrium-90 microsphere radioembolization. The registry marks Groups II and III as closed. It does not report a randomized assignment process, and the group descriptions indicate different treatment pathways and eligibility features. The fuller Group I description is truncated in the registry record, so its complete eligibility criteria are not reported there.

The study was unmasked. In a non-randomized design, clinical decisions and eligibility can influence which group a participant enters. Differences in outcomes between these groups therefore cannot be interpreted as if treatment were randomly assigned. The registry does not report a like-for-like control comparison or a pembrolizumab-alone group in the listed arms.

The primary endpoint was the proportion with an overall response by investigator-reported RECIST 1.1, assessed up to 25 months. Secondary outcomes listed include median duration of response, treatment-related adverse events by maximum grade for Groups II and III, median progression-free survival, and immune-related progression-free survival. Although six outcome measures are recorded, the registry record does not provide results for all of them. Its completed status describes the trial record status; it does not by itself establish that every planned analysis or outcome result is available.

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 Proportion of Participants With an Overall Response, measured over Up to 25 months.

The primary endpoint is the proportion of participants with an overall response. Under the registry definition, a participant is counted as a responder if the best observed response is a complete response or partial response according to investigator-reported RECIST 1.1. The endpoint is assessed up to 25 months, and the analysis is specified for the All Subjects as Treated population, also described in the record as ITT. The plan is to report a proportion and a 90% confidence interval separately for each group.

As a general statistical concept, an overall response proportion summarizes how often a defined response category occurs in the population being analyzed. It is a binary outcome at the participant level: a participant either meets the response definition or does not. A confidence interval expresses uncertainty around the estimated proportion; it does not show the range of individual participant outcomes.

Response is often used as an endpoint because it gives a direct, prespecified way to describe tumor shrinkage under stated assessment criteria. In this study, the endpoint is investigator-reported, so it should be understood as the response assessment recorded by investigators under RECIST 1.1. The registry record does not report the observed proportions or their confidence intervals. It therefore does not allow the precision or size of the primary response estimates to be evaluated.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary response measure: each participant is classified according to whether their best observed response met the complete-response or partial-response definition. The group-level statistic is a proportion. The registry specifies investigator-reported RECIST 1.1 and asks for separate estimates for the three groups, rather than one pooled estimate.

What does a confidence interval communicate?

As a general statistical explanation, a confidence interval describes uncertainty in an estimate under the method used to calculate it. A narrower interval indicates greater precision under that method, while a wider interval indicates less precision. The registry specifies a 90% confidence interval for each response proportion. The record does not include the actual response proportions or intervals, so their widths and the uncertainty around the observed responses are not reported in the registry record.

A confidence interval is not a range expected to contain every participant's outcome, nor does it eliminate bias in the estimate. It reflects sampling uncertainty under the model and method; it cannot correct for non-randomized group assignment, incomplete outcome recording, or differences in participant characteristics. The particular interval calculation method is not reported in the registry record.

Why does the design matter for interpretation?

In a randomized comparison, assignment by chance can help balance measured and unmeasured participant characteristics between groups, on average. This study was non-randomized, so that protection is absent. The groups also represent different treatment combinations and eligibility pathways. A difference between their response proportions, if reported, could be associated with treatment, participant selection, disease features, or other factors. The record does not report a statistical adjustment strategy for such differences. Accordingly, these groups should not be treated as randomized alternatives or as a direct estimate of the relative effect of the additional therapies.

The study was unmasked. As a general consideration, knowledge of treatment can influence some aspects of clinical assessment or reporting, depending on the outcome and procedures. The endpoint is investigator-reported, and the registry record does not describe independent blinded review. It does not report whether any other safeguards were used to reduce assessment bias.

How should the safety counts be read?

The safety information gives the number of participants at risk and the number with serious or other adverse events for each group. These are participant counts. They are not event totals, and they do not convey how many events occurred per participant. A participant may experience more than one event; the record does not provide event-level detail for interpreting frequency, recurrence, timing, duration, or severity beyond the listed categories. The counts also do not establish that an event was caused by treatment unless attribution is specifically reported.

The registry lists serious-event terms with percentages, including fatigue at 66.7%, abdominal pain at 33.3%, alkaline phosphatase increased at 33.3%, aspartate aminotransferase increased at 33.3%, bile duct stenosis at 33.3%, and blood bilirubin increased at 33.3%. The group and denominator associated with these percentages are not reported in the registry record. They should not be assigned to a particular group or interpreted as risks for an individual participant without that context.

What does over-dispersion mean in general?

As a general statistical concept, over-dispersion occurs when observed variation is greater than a statistical model assumes. It matters because a model that underestimates variation can produce confidence intervals that are too narrow and tests that appear more conclusive than warranted. The primary endpoint is a proportion, and the registry record does not report an over-dispersion assessment or a model-based adjustment. No conclusion about over-dispersion can be drawn from the available record.

The record specifies a sample-size planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. Such planning calculations are not observed power and do not establish that the completed study had a particular probability of detecting an effect. The described assumptions do not make the non-randomized groups comparable, and the record does not report the trial's own sample-size justification or design calculations. The specific planning values are not reported in the registry record.

The statistical analysis plan is not described in enough detail to determine how the response interval was calculated, how missing assessments were handled, or whether any covariate adjustment was planned. These omissions limit reproducibility and interpretation. The record describes the intended endpoint and analysis population, but it does not provide the primary numerical results needed to assess the response 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("Group I [Pembrolizumab, 177Lu DOTATATE]", "Group II [Pembrolizumab, TAE]", "Group III [Pembrolizumab, Yttrium-90 Microsphere R")
n      <- c(26, 3, 3)
events <- c(0, 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  Group I [Pembrolizumab, 177Lu DOTATATE]  26  0  0.0  0.0  13.2  0.0  12.9  0.1  12.8
ARM  Group II [Pembrolizumab, TAE]  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Group III [Pembrolizumab, Yttrium-90 Microsphere R  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2

11. Multiplicity, Missing Data and Analysis Populations

The primary endpoint plan calls for a separate response proportion and 90% confidence interval in each of the three groups. The registry record does not report a formal hypothesis-testing strategy for comparing those groups, nor does it report an adjustment for making multiple comparisons or assessing multiple outcomes. It is therefore not reported in the registry record whether multiplicity was addressed in the trial's analysis plan.

Multiplicity matters as a general statistical issue when several hypotheses or outcomes are examined: the chance of at least one apparently unusual result can increase as more tests are considered. Separate descriptive estimates and confidence intervals can still be useful, but they should not be treated as formal evidence of between-group differences without a stated comparison method and interpretation framework. The registry record does not report such a framework.

The primary description identifies the All Subjects as Treated population and labels it ITT. The record does not provide participant flow by analysis population, the number with evaluable response assessments, or the handling of participants without an assessment. It does not report whether missing response data were excluded, counted as non-response, imputed, or addressed in another way. The impact of missing outcome data therefore cannot be assessed from the record.

The listed secondary outcomes include median duration of response, treatment-related adverse events by maximum grade for Groups II and III, median progression-free survival, and immune-related progression-free survival. Six outcome measures are recorded, but the record does not provide all results or a detailed hierarchy of confirmatory and exploratory analyses. The statistical analysis plan, interim analysis procedures, and missing-data methods are not reported in the registry record. No inference about those procedures should be made from the trial's completed status.

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: fatigue (66.7%), abdominal pain (33.3%), alkaline phosphatase increased (33.3%), aspartate aminotransferase increased (33.3%), bile duct stenosis (33.3%).

The registry records serious adverse events among 14 of 26 participants at risk in Group I, 3 of 3 in Group II, and 2 of 3 in Group III. For other adverse events, it records 26 of 26 in Group I, 3 of 3 in Group II, and 3 of 3 in Group III. These figures describe participants with at least one event in the recorded categories; they are not counts of all events or a severity distribution.

The record also lists deaths: 4 in Group I, 0 in Group II, and 0 in Group III. Whether any death was considered treatment-related is not reported in the registry record. The listed serious-event terms include fatigue, abdominal pain, increased alkaline phosphatase, increased aspartate aminotransferase, bile duct stenosis, and increased blood bilirubin. The percentages attached to those terms are 66.7% and 33.3% entries, but their group and denominator are not reported in the registry record.

These figures are descriptive. The smaller groups make their counts particularly sensitive to individual outcomes, and the groups were not randomized. The registry record does not provide enough information to compare safety causally or to establish event timing, recurrence, or attribution.

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

The study addresses a clinically specific question: response and safety when pembrolizumab is combined with peptide receptor radionuclide therapy or a liver-directed procedure for well-differentiated neuroendocrine tumors with liver metastases. Its non-randomized, parallel-group design allows the registry to describe the treatment pathways and record outcomes within groups, but does not support treating those groups as randomized comparators.

The available registry information does not report the primary response proportions or their 90% confidence intervals. It therefore does not establish the frequency or precision of response for any group. The description of the endpoint is not itself a result, and the study's completed status does not fill that gap. The record also lists several secondary outcomes without reporting their numerical results in the information available.

The safety counts document participants recorded with serious and other adverse events, as well as deaths, by group. They do not, on their own, show that treatment caused an event or death, explain how long events lasted, or permit a reliable comparative safety estimate. The group sizes differ, and the smaller groups have limited information for estimating event frequency precisely. The serious-event term percentages lack a reported group and denominator in the registry record.

In general, a pilot study can provide descriptive information and help frame questions for further evaluation. That general role does not establish effectiveness or comparative benefit. Here, the registry record does not report the primary efficacy results, the statistical comparison plan, or enough information to evaluate missing response assessments. The responsible interpretation is therefore limited to the specified design, outcome definitions, and recorded safety counts—not a conclusion that any treatment combination is superior, effective, or ineffective.

14. Why This Trial Matters Statistically

This trial is statistically notable because its treatment groups are non-randomized and represent distinct combinations of pembrolizumab with peptide receptor radionuclide therapy or liver-directed procedures. That structure makes the distinction between within-group description and between-group comparison essential. Separate response proportions, even if available, would not by themselves establish that one treatment combination produced a different response from another.

The primary outcome is clearly defined in the registry as a complete or partial response by investigator-reported RECIST 1.1, with a separate 90% confidence interval planned for each group. This helps identify what the intended endpoint means. However, the registry record does not report the estimated proportions or intervals, so the intended statistical summary cannot be evaluated from the recorded results.

The safety information also illustrates why denominators and context matter. The group-level counts are reported alongside the number at risk, but the smaller groups contain limited information for estimating event frequency. The listed serious-event percentages do not have a reported group or denominator in the registry record. These details affect how confidently a reader can interpret the numbers.

Finally, the record identifies several secondary outcomes but does not report their numerical results or a multiplicity strategy. This makes it important to distinguish a study's planned questions from analyses that can actually be interpreted using the registry record. The design and reporting gaps do not negate the value of the clinical question; they constrain the strength and type of statistical conclusions that can be drawn from the posted information.

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 primary response estimates and their confidence intervals are not reported in the registry record. This is the main limitation for interpreting the intended efficacy endpoint: the endpoint is defined, but its observed value and precision cannot be assessed. Numerical results for the listed secondary outcomes are also not reported in the information available.

The study used non-randomized allocation, no masking, and parallel groups with different treatment pathways. The record does not describe a strategy for controlling confounding or adjusting for differences between groups. As a result, comparisons between groups would be vulnerable to selection and other systematic differences, even if complete outcome estimates were available.

Group II and Group III are marked closed, and their recorded safety groups are small. The registry record does not provide a full participant-flow account, group-specific enrollment history, or the number with evaluable primary outcomes. The fuller Group I arm description is truncated, so complete eligibility details are not reported in the registry record.

The registry does not report the method used to calculate confidence intervals, missing-data handling, a multiplicity plan, or the statistical analysis plan in sufficient detail. It also does not report the group and denominator for the percentages attached to the serious-event terms. These omissions limit the ability to reproduce analyses, judge uncertainty, or interpret safety terms in context. The reported deaths are not identified as treatment-related or otherwise attributed in the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Pembrolizumab With Liver-Directed or Peptide Receptor Radionuclide Therapy for Neuroendocrine Tumors and Liver Metastases — NCT03457948.
• Registry record last updated 2026-08-28; status last verified 2026-08.
• Results first posted 2026-08-28.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

This completed phase II pilot study enrolled 32 participants and used non-randomized, unmasked, parallel groups. The recorded start date is 2018-08-27, and the primary completion date is 2025-07-31. The study concerned well-differentiated neuroendocrine tumors with symptomatic and/or progressive liver metastases.

The groups paired pembrolizumab with 177Lu-DOTATATE, arterial embolization, or yttrium-90 microsphere radioembolization. The primary outcome was the proportion with complete or partial response by investigator-reported RECIST 1.1, assessed up to 25 months, with a 90% confidence interval planned separately for each group.

The registry records serious and other adverse-event counts and deaths by group, but does not report the primary response estimates or the numerical results of the listed secondary outcomes. Because allocation was non-randomized and group pathways differed, the groups are not like-for-like controls. The registry record does not report the missing-data, multiplicity, or detailed analysis procedures needed to resolve these limitations.