Drug-Eluting Bead, Irinotecan Therapy for Unresectable Intrahepatic Cholangiocarcinoma w/Concomitant Gemcitabine and Cisplatin or Carboplatin: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Drug-Eluting Bead, Irinotecan Therapy for Unresectable Intrahepatic Cholangiocarcinoma w/Concomitant Gemcitabine and Cisplatin or Carboplatin (NCT01648023), a randomized phase 2 study. The primary outcome is Tumor Response (Overall Response = Complete Response + Partial Response + Progressive Disease + Stable Disease).

2. Clinical Question

The trial was designed to compare the addition of irinotecan-loaded beads to infusional chemotherapy with infusional chemotherapy alone. The registry describes it as a prospective, open-label, randomized phase II study. Its stated purpose included assessing safety and whether the combination was more effective, but the primary outcome posted is a count of tumor-status classifications at 6 months.

The key interpretive issue is how that outcome is defined. In ordinary clinical usage, an overall response rate generally refers to complete and partial responses. In this record, the parenthetical definition also includes progressive disease and stable disease. Because those categories cover distinct clinical states, a count that adds all four cannot by itself tell readers how many participants had tumor shrinkage, stable disease, or progression.

The registry lists overall survival as a secondary outcome, but a numerical secondary analysis is not reported in the registry record. Thus, the primary result supports a narrow description of the reported counts and their recorded comparison; it does not resolve the broader question of how the treatment groups differed in response category or survival.

3. Trial Design

This was a randomized, parallel-group, phase II study with no masking. The official title describes a prospective, open-label comparison. One group received transarterial chemoembolization with LC or ONCOZENE irinotecan-loaded beads alongside gemcitabine and cisplatin or carboplatin. The other received gemcitabine with cisplatin or carboplatin alone. The choice between cisplatin and carboplatin is part of the recorded regimens; further details about how that choice was made are not reported in the registry record.

The record lists total enrollment as 49. For the primary outcome, the arm entries list 24 participants for the bead-plus-chemotherapy group and 22 for the chemotherapy-alone group. These outcome counts do not account for the full enrollment figure. Reasons for the difference, including any exclusions, missing outcome assessments, or differences in analysis populations, are not reported in the registry record. The posted values should therefore be read as outcome-specific counts, not as a complete accounting of enrollment.

Tumor response was assessed at 2, 4 and 6 months, with the 6-month result reported. The registry states that assessment used modified Response Evaluation Criteria in Solid Tumors (mRECIST). The record describes complete response as disappearance of all lesions; partial response as a 30% decrease in the sum of the longest diameter of target lesions from baseline or a 30% reduction in arterial enhancement; and progressive disease as a 20% increase from the smallest recorded target-lesion sum or appearance of one or more new lesions greater than 1cm. Stable disease is defined as neither sufficient shrinkage for partial response nor sufficient increase for progressive disease.

The record marks the trial as completed and lists a primary completion date of 2019-01. It lists overall survival as a secondary outcome, but the registry record does not report a numerical secondary outcome analysis. A statistical analysis plan, details of allocation concealment, the handling of missing observations, and the analysis population are not reported in the registry record. The posted arm counts should be interpreted with these reporting gaps in mind.

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

Reasons for not completing this period:

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 Tumor Response (Overall Response = Complete Response + Partial Response + Progressive Disease + Stable Disease), measured over Assessed at 2, 4 and 6 months. 6 months reported..

The primary endpoint is titled tumor response and is expressed as a number of participants. Its timeframe says assessments occurred at 2, 4 and 6 months, with 6 months reported. The registry specifies mRECIST criteria for classifying tumor status. These criteria use changes in lesion size or arterial enhancement, as well as the appearance of new lesions, to assign categories such as complete response, partial response, progressive disease, or stable disease.

The endpoint’s parenthetical definition is important: it defines overall response as the sum of complete response, partial response, progressive disease, and stable disease. That is broader than a response measure limited to complete and partial response. It appears to count participants across all four listed classifications, rather than isolating tumor shrinkage. The record does not give separate counts for each category, so the posted aggregate cannot show how many participants were in each state.

In general, tumor-response endpoints can provide a structured assessment of tumor status at a specified time. A category-based measure can be clinically informative when the category counts and rules for assigning them are clear. Here, however, combining all four categories means that the aggregate count does not distinguish favorable shrinkage from stable disease or progression. The reason for choosing this particular aggregate as the primary endpoint is not reported in the registry record. Consequently, the result is best read as the number recorded under the stated definition, not as a conventional complete-or-partial overall response rate.

7. Results Posted in the Registry

• Randomization to Gem-Cis or Gem-Carbo: 100.0% (22/22) versus 100.0% (24/24) on control; risk difference 0.0 percentage points (95% CI 0.0 to 0.0), risk ratio 1.00 (1.00–1.00); unadjusted two-proportion z-test p = n/a.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary in the statistical summary: each participant either contributes to the recorded outcome count or does not. For such data, a common comparison is the difference between the proportions in two groups. A risk difference of 0.0 means that the posted comparison estimates no difference on that scale. The record reports 1.0 for each of the two comparison values and a risk difference of 0.0. These are the stated comparison results; they do not correct or clarify the unusual endpoint definition.

In general, a risk difference expresses an absolute contrast between group proportions. It is often easier to interpret than a relative measure because it describes the difference on the proportion scale. Its meaning depends on what counts as an event. In this trial, the event is tied to an aggregate that includes complete response, partial response, progressive disease, and stable disease. Thus, the comparison cannot be interpreted as a difference in tumor shrinkage alone. Clear event definitions are essential: the same numerical contrast can answer very different questions depending on which outcome categories count as events.

Confidence intervals are used in general to represent uncertainty around an estimated effect. A narrow interval indicates that the estimate is more precisely determined under the analysis assumptions; a wide interval indicates less precision. The registry comparison lists lower and upper bounds of 0.0 and 0.0. The confidence level and method for producing those bounds are not reported in the registry record. The bounds therefore can be quoted as posted, but the record does not support a fuller account of their construction or coverage.

A test of a null hypothesis and an interval estimate address related but different questions. A test evaluates compatibility with a specified null value under its assumptions, while an interval describes a range of effect values supported by an estimation procedure. The registry record does not post a primary p-value or a detailed testing procedure for this outcome. The comparison’s 1.0 values are not identified as p-values in the record and should not be described as such.

The study is described as randomized and parallel-group. In general, random allocation is intended to reduce systematic differences between groups at assignment, while a parallel design follows the assigned groups under their respective regimens. The record does not report details of sequence generation or allocation concealment. It also states that the study was open-label, meaning treatment assignment was not masked as described in the registry. Whether knowledge of assignment affected assessment or reporting cannot be determined from the posted record.

mRECIST is the named method for tumor classification. The record describes thresholds based on lesion measurements and arterial enhancement, and criteria involving new lesions. The primary result, however, does not report each mRECIST category separately. It also does not describe the assessment process in further detail, including who performed assessments, how disagreements were handled, or whether central review was used. Those details are not reported in the registry record.

The sample-size material associated with the results is a planning illustration based on the observed effect, not the trial’s original design calculation. The stated illustration assumptions are a two-sided significance level of 0.05, equal allocation, and a normal approximation, with the target effect expressed as a fraction of the observed effect. Such calculations explore how a future study might be planned under those assumptions. They are not observed power and do not show that the completed study had a particular probability of detecting an effect. The registry does not report the original sample-size justification or the trial’s own design assumptions.

In general, over-dispersion matters when the variability of observed outcomes exceeds what a chosen statistical model expects. It can make standard errors too small if the model assumes less variability than is present. For a simple binary endpoint, the chosen model and variance method still matter, especially with limited data. The registry does not report a model specification or a method for variance estimation, so no claim about over-dispersion or its handling can be made for this analysis.

The registry record lists no posted analysis entries and does not describe adjustment for covariates, the intention-to-treat principle, per-protocol analysis, or any other analysis population. In general, these choices affect which participants contribute to an estimate and how it should be interpreted. Without them, readers cannot determine whether the reported counts reflect all randomized participants, participants with an available assessment, or another defined group. The difference between total enrollment and the outcome arm counts reinforces the need for that information, but does not by itself identify what happened to the remaining enrolled participants.

10. Confidence Intervals and Precision

The comparison record reports a risk difference of 0.0 and lower and upper bounds of 0.0 and 0.0. It does not state the confidence level or interval method in the registry record. These posted bounds can be described as the recorded interval limits, but it would be inappropriate to assign a conventional confidence level without further information.

In general, an interval communicates uncertainty as well as the direction and scale of an estimated effect. Precision depends on factors such as the amount of information in the data, the outcome frequency, and the analysis method. An interval with identical reported limits may look exact, but readers need to know how it was constructed and what assumptions were used before treating it as a precise statement about the underlying treatment effect. The registry record does not provide those details.

The primary outcome entries show 24 of 24 and 22 of 22, with comparison values of 1.0 in both groups. The risk difference and interval bounds are therefore reported as equal in this record. Yet the outcome’s aggregate definition combines four clinically distinct response categories, including progressive disease. This creates a more fundamental interpretive limitation than numerical precision alone: even a precisely estimated contrast in the aggregate would not reveal the distribution across those categories.

Separate category counts, assessment completeness, and details of the interval calculation are not reported in the registry record. Without them, the precision of clinically specific response contrasts cannot be assessed. The posted comparison should be treated as the reported estimate for the registered aggregate, not as a precise estimate of differences in complete or partial response.

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists tumor response as the primary outcome and overall survival as a secondary outcome. It does not report a numerical secondary analysis for overall survival. A statistical analysis plan identifying the full set of planned outcomes, hypotheses, or testing order is not reported in the registry record. It is therefore not possible to determine whether or how multiplicity was addressed.

In general, when multiple outcomes or comparisons are examined, the chance of at least one apparently noteworthy result can increase if each is tested independently without an appropriate plan. Adjustment, hierarchical testing, or a clearly designated primary analysis can address different aspects of that issue. No such approach is described in the record, so none should be assumed. The absence of a posted secondary analysis also means the secondary outcome cannot be interpreted from the numerical evidence available in this record.

The primary outcome entries list 24 and 22 participants, while the registry lists total enrollment as 49. The reason the arm-specific outcome counts do not account for the enrollment figure is not reported in the registry record. The record does not state whether the difference reflects unavailable assessments, exclusions, a different analysis population, or another cause. Missing-data methods and analysis-population definitions are also not reported.

In general, missing outcomes can affect both the estimated group contrast and its uncertainty, particularly if the likelihood of missingness is related to treatment assignment or tumor status. No assumptions about missingness or imputation can be inferred. The posted counts should therefore be read as the registered outcome entries, with completeness and population membership unresolved.

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: nausea (16.7%), abdominal pain (9.1%), pelvic fracture (9.1%), sepsis (9.1%), fever (8.3%).

The adverse-event record gives arm-specific counts for serious adverse events, other adverse events, and deaths. In the bead-plus-chemotherapy group, 21 of 24 participants are listed as affected by serious adverse events, 24 of 24 by other adverse events, and 17 deaths are recorded. In the chemotherapy-alone group, the corresponding entries are 19 of 22, 19 of 22, and 16 deaths. These are descriptive registry counts and do not establish that the treatment caused an event.

The record also lists serious-event terms with percentages: nausea, 16.7%; abdominal pain, 9.1%; pelvic fracture, 9.1%; sepsis, 9.1%; fever, 8.3%; and hypertension, 8.3%. It does not identify the arm-specific denominators for these term percentages in the information recorded. They should not be assigned to a particular arm or treated as a direct comparison between arms.

In general, adverse-event counts depend on how events are defined, collected, and attributed, and the number affected is not the same as the number of event episodes. The registry does not provide further detail about event severity grading, timing, exposure, or attribution. These figures are best read as descriptive counts within the recorded groups, not as a complete comparative safety analysis.

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

The result establishes what the registry recorded for its defined primary outcome: arm entries of 24 and 22, comparison values of 1.0 in each group, and a risk difference of 0.0 with bounds of 0.0 and 0.0. The endpoint is defined as the sum of complete response, partial response, progressive disease, and stable disease. Consequently, the aggregate does not indicate how many participants experienced tumor shrinkage, how many had stable disease, or how many had progression.

The result does not establish equivalence between the treatment strategies. A reported estimate of zero is not, by itself, evidence that clinically meaningful differences have been ruled out. That judgment would require a suitable endpoint, a prespecified equivalence or non-inferiority framework if relevant, and interpretable uncertainty estimates. Those elements are not reported in the registry record. In particular, the interval confidence level and method are unspecified.

The posted primary result also does not answer the secondary overall-survival question. Overall survival is listed as a secondary outcome, but a numerical result is not reported in the registry record. The adverse-event counts provide descriptive information, but without further methodological detail they do not establish a causal safety difference between groups.

In general, trial results are interpreted in light of the endpoint definition, study design, analysis population, missing data, and uncertainty. Here, the randomized design offers a framework for group comparison, but important details about allocation, analysis methods, and participant accounting are not reported. The difference between total enrollment and the primary outcome arm counts remains unexplained.

The most defensible reading is therefore narrow: the registry reports equal aggregate primary-outcome values for the two groups under a definition that includes every listed tumor-status category. It does not provide category-specific response counts or a numerical survival analysis. The result should not be translated into a claim that the regimens have the same effect, that all counted participants benefited, or that one regimen is preferable.

14. Why This Trial Matters Statistically

This trial illustrates how statistical interpretation depends on the operational definition of an endpoint, not only on the reported estimate. The primary outcome is called tumor response, but its stated aggregate includes complete response, partial response, progressive disease, and stable disease. A reader who assumes that overall response means only tumor shrinkage could draw a substantially different conclusion from the one supported by the actual definition. Clear category-level reporting is necessary to understand what the aggregate represents.

The record also highlights the distinction between an estimated effect and the clinical question readers want answered. A risk difference of 0.0 describes the posted aggregate comparison. It cannot tell whether the groups differ in the distribution of response categories, and the registry does not supply separate counts to answer that. Statistical equality on one aggregate is not automatically equality on other outcomes or a demonstration of equivalence.

Another important issue is participant accounting. Total enrollment is recorded as 49, whereas the primary outcome lists 24 and 22 participants in the arms. The reason for this gap is not reported in the registry record. Understanding which participants contributed to the result is central to evaluating possible missing-data and analysis-population concerns.

Finally, the registry lists overall survival as a secondary outcome but does not report a numerical result for it. The record therefore demonstrates why a trial’s stated aims, its selected primary endpoint, and its posted statistical evidence should be considered separately. Readers need a clear endpoint definition, transparent population accounting, category-specific results where relevant, and an interpretable uncertainty analysis to connect a numerical comparison with a clinical question.

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 trial-specific reporting gaps limit interpretation. The primary outcome definition adds complete response, partial response, progressive disease, and stable disease together. The record does not provide separate counts for those categories, so the aggregate cannot be used to identify how many participants had response, stability, or progression.

The registry lists enrollment of 49, but the primary outcome arm entries list 24 and 22 participants. The reason for the difference is not reported in the registry record. It also does not state whether the primary outcome analysis included all randomized participants, how missing assessments were handled, or whether exclusions occurred. These omissions make it difficult to evaluate completeness and potential selection effects.

The comparison record gives a risk difference of 0.0 and bounds of 0.0 and 0.0, but does not report the confidence level or method for calculating those bounds. The statistical analysis plan, primary testing procedure, and any adjustment for multiple outcomes are not reported in the registry record. No numerical secondary analysis is posted for the listed overall-survival outcome.

The study is described as randomized and open-label, but details of sequence generation, allocation concealment, and response-assessment review are not reported. The registry also does not provide further detail on how the choice of cisplatin or carboplatin was determined. The serious-event term percentages are not linked to arm-specific denominators in the record. These limitations do not negate the posted values, but they constrain how far those values can be interpreted beyond the registered descriptions.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Drug-Eluting Bead, Irinotecan Therapy for Unresectable Intrahepatic Cholangiocarcinoma w/Concomitant Gemcitabine and Cisplatin or Carboplatin — NCT01648023.
• Registry record last updated 2025-02-27; status last verified 2021-12.
• Results first posted 2022-01-13.
• 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 completed, randomized, open-label, parallel-group phase II study in patients with unresectable intrahepatic cholangiocarcinoma. It compared irinotecan-loaded LC or ONCOZENE beads with gemcitabine plus cisplatin or carboplatin against gemcitabine with cisplatin or carboplatin alone.

The primary outcome was tumor response assessed at 2, 4 and 6 months, with the 6-month result reported using mRECIST. Its stated definition combines complete response, partial response, progressive disease, and stable disease. The outcome entries list 24 participants in the bead group and 22 in the chemotherapy-alone group. The comparison reports a risk difference of 0.0 and bounds of 0.0 and 0.0.

Overall survival is listed as a secondary outcome, but a numerical secondary analysis is not reported in the registry record. The record does not explain the difference between total enrollment and the primary outcome arm counts, nor does it report the analysis population or missing-data methods. The aggregate primary result does not distinguish response categories.