This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Study of Nivolumab in Patients With Advanced Refractory Biliary Tract Cancers (NCT02829918), a na phase 2 study. The primary outcome is Overall Response Rate (ORR) After 4 Cycles of Treatment.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02829918 |
| Phase | Phase 2 |
| Condition | Biliary Tract Cancer, Biliary Tract Neoplasms |
| Allocation | Not applicable (single group) |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 54 |
| Arms in the primary outcome | n/a |
| Lead sponsor | H. Lee Moffitt Cancer Center and Research Institute |
| Results posted | Yes |
2. Clinical Question
The study asked whether nivolumab could produce an objective tumor response in participants with advanced refractory biliary tract cancers. Its primary endpoint was overall response rate (ORR) after 4 cycles of treatment, assessed at 16 weeks. The registry defines ORR as the proportion with a complete response (CR) or partial response (PR), according to Response Evaluation Criteria in Solid Tumors (RECIST) 1.1.
Population
Participants with advanced biliary tract cancers who had failed or were intolerant to at least one line of therapy and no more than 2 lines of therapy. The primary endpoint lists 46 participants as evaluable for response; the registry enrollment is 54.
Intervention
Nivolumab as a single agent. The registry describes a single-arm study and lists nivolumab as its drug intervention.
Comparator
No concurrent comparator is listed. Allocation is recorded as not applicable, and the study model is single group.
Primary question
What was the ORR after 4 cycles, at the 16-week assessment, among participants evaluable for response? The numerical ORR is not reported in the registry record.
This is a descriptive question about response in a treated group, not a randomized estimate of the difference between nivolumab and another treatment. Without a comparator, outcomes cannot establish how the group would have fared under a different treatment or without nivolumab. The registry also lists overall survival (OS) and progression-free survival (PFS) as secondary outcomes, but does not report their numerical results in the information available for this analysis.
3. Trial Design
This was a phase II, single-group study of nivolumab for advanced refractory biliary tract cancers. The registry records allocation as not applicable and masking as none. There is no randomized control group listed, so the study is not a like-for-like comparison of treatment arms. Its enrollment is recorded as 54, and its status is completed. The registry lists a start date of 2016-10-05 and a primary completion date of 2019-03-27.
The arm description calls the study a two-stage design. It says that the first stage would accrue 18 participants, but the description ends before giving the stage decision rule. The continuation of the accrual plan and any criteria for proceeding between stages are not reported in the registry record. The label “two-stage” therefore cannot, by itself, establish how the study would stop early or what response threshold would guide that decision.
Participants were described as having failed or being intolerant to at least one line of therapy and no more than 2 lines of therapy. The registry does not provide a complete eligibility description in the information recorded for this analysis. It identifies H. Lee Moffitt Cancer Center and Research Institute as the sponsor.
The primary endpoint is ORR after 4 cycles, with a timeframe of 16 weeks. The primary outcome description identifies 46 participants as evaluable for response, distinct from the enrollment figure. The registry does not explain why the evaluable response population differs from the enrollment figure, or describe any exclusions from the response analysis. OS and PFS are listed as secondary outcomes. Their assessment definitions, analysis methods, and numerical results are not reported in the registry record.
The single-group design can describe response and safety observations among participants in the study, but it cannot isolate a treatment effect from other influences through a randomized comparison. Any causal interpretation would require assumptions that cannot be assessed from the registry information alone.
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
| Stage | Nivolumab Treatment |
|---|---|
| STARTED | 54 |
| COMPLETED | 46 |
| NOT COMPLETED | 8 |
Reasons for not completing this period:
| Reason | Nivolumab Treatment |
|---|---|
| Death | 2 |
| Withdrawal by Subject | 4 |
| Adverse Event | 1 |
| Progression | 1 |
Analysis population for the primary outcome, as defined in the registry: 46 participants were evaluable for response
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.
Population: 54 participants at baseline, 46 participants were evaluable for outcome measures
| Characteristic | Unit | Nivolumab Treatment |
|---|---|---|
| Participants analysed | 54 | |
| Age, Categorical: <=18 years | Participants | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 24 |
| Age, Categorical: >=65 years | Participants | 30 |
| Sex: Female, Male: Female | Participants | 27 |
| Sex: Female, Male: Male | Participants | 27 |
| Ethnicity (NIH/OMB): Hispanic or Latino | Participants | 8 |
| Ethnicity (NIH/OMB): Not Hispanic or Latino | Participants | 46 |
| Ethnicity (NIH/OMB): Unknown or Not Reported | Participants | 0 |
| Race (NIH/OMB): American Indian or Alaska Native | Participants | 0 |
| Race (NIH/OMB): Asian | Participants | 3 |
| Race (NIH/OMB): Native Hawaiian or Other Pacific Islander | Participants | 0 |
| Race (NIH/OMB): Black or African American | Participants | 8 |
| Race (NIH/OMB): White | Participants | 42 |
| Race (NIH/OMB): More than one race | Participants | 0 |
| Race (NIH/OMB): Unknown or Not Reported | Participants | 1 |
| Region of Enrollment: United States | participants | 54 |
6. Primary Endpoint
The registry lists the primary outcome as Overall Response Rate (ORR) After 4 Cycles of Treatment, measured over 16 weeks.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | percent |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint, overall response rate, measures the share of response-evaluable participants whose tumors meet criteria for complete or partial response. The registry specifies RECIST 1.1 as the response criteria and places the assessment after 4 cycles of treatment, at 16 weeks. The response-evaluable population is listed as 46 participants.
ORR is used in a study of this type to describe tumor shrinkage meeting a predefined response threshold during a specified assessment period. It is a proportion-based endpoint: the numerator would be participants classified as having CR or PR, and the denominator would be the response-evaluable population under the stated analysis definition. The registry record does not provide the number with CR, the number with PR, or the resulting ORR, so those values cannot be stated. It also does not report a confidence interval for the endpoint.
ORR focuses on response status at the specified assessment rather than how long a response lasts. The registry lists OS and PFS as secondary outcomes, which address different aspects of clinical outcome, but their definitions and results are not reported in the registry record. A response endpoint can be clinically informative as a measure of observed tumor response, but a single-group response rate does not provide a randomized estimate of benefit relative to another treatment. Its interpretation also depends on how response evaluations, missing assessments, and the evaluable population are handled; those details are not reported in the registry record.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry describes a completed phase II study of nivolumab in a single group of participants with advanced refractory biliary tract cancers. Its primary question concerned ORR after 4 cycles, with the outcome timeframe recorded as 16 weeks. The primary endpoint description gives 46 participants as evaluable for response, while enrollment is recorded as 54. No numerical ORR, response count, confidence interval, or p value is reported in the registry record. Consequently, the magnitude and precision of the primary response result cannot be assessed from this record.
The distinction between enrollment and the response-evaluable population matters. An outcome rate depends on both the response count and the denominator used. The registry identifies the denominator but does not report the response count or explain why the response-evaluable population differs from enrollment. It is therefore not possible to reconstruct the response rate or determine how the evaluable population was formed. No inference should be made about outcomes among participants not included in that population.
The design limits what the results could establish even if a response estimate were available. With no concurrent randomized comparator, observed outcomes in the nivolumab group cannot be interpreted as a randomized difference from another treatment. The study can describe what happened in its participants, but other factors may also influence outcomes. The registry lists a two-stage design, yet the stage continuation or stopping rule is not reported in the registry record. That missing detail prevents assessment of how the design’s decision process may have shaped enrollment or interpretation.
OS and PFS appear as secondary outcomes, but their definitions, estimates, uncertainty measures, and analyses are not reported in the registry record. The record lists 3 outcome measures and no analysis records. Thus, the outcome labels alone do not show what was observed or how results were analyzed. Statistical significance, precision, and consistency across outcomes cannot be evaluated.
The planning illustration uses assumptions of a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the detectable effect expressed as a stated fraction of the observed effect. These are planning assumptions, not the study’s single-group design and not evidence of observed power. The illustration is described as being based on the observed effect, but the numerical response result needed to inspect that effect is not reported in the registry record. Sample-size illustrations should therefore be read as hypothetical planning scenarios, not as a substitute for the missing endpoint estimate or its uncertainty.
The safety record lists 32 serious adverse events affected in a group with 54 at risk, 52 other adverse events affected in a group with 54 at risk, and deaths of 2. These are descriptive counts recorded for the treatment group. They do not establish that nivolumab caused any event or death. The registry information does not provide comparative safety data, event timing, attribution, or the details needed to evaluate individual risks.
9. Statistical Methods Explained
The registry describes an ORR endpoint, a single-group design, and a two-stage study label, but does not report the statistical analysis plan or numerical outcome analysis. The following concepts explain how such methods are generally interpreted; they should not be read as confirmation that a particular method was used in this trial.
What does ORR measure statistically?
ORR is generally treated as a proportion: the response count divided by the population included in the response analysis. Here, the response definition combines CR and PR under RECIST 1.1, and the registry identifies 46 participants as evaluable. The response count and resulting proportion are not reported in the registry record. A proportion describes the observed frequency in the analyzed group; it does not, by itself, measure how much better the outcome is than it would have been under a comparator.
A confidence interval expresses the uncertainty around an estimated quantity under a statistical model. For a response proportion, an interval can help show how much uncertainty remains because the estimate is based on a finite group of participants. The registry record does not report an ORR confidence interval, so the precision of the primary endpoint cannot be judged. Different interval procedures can have different properties, especially when the response count is small or near the boundaries of the possible range. The analysis procedure is not reported, so a specific interval method cannot be attributed to this study.
What does the single-group structure allow?
In a single-group study, the observed response proportion describes outcomes in participants receiving the intervention. A randomized comparison instead estimates a contrast between groups assigned to different strategies. Since allocation is recorded as not applicable and no comparator is listed, this study does not provide that randomized contrast. Comparisons with historical experience, if attempted, would require assumptions about comparability of populations, assessment, eligibility, and follow-up. Such a comparison is not reported in the registry record.
The absence of masking is also recorded. Masking can reduce the potential for expectations to influence some study procedures or assessments, depending on the outcome and how it is measured. The registry does not describe safeguards for response assessment or whether any independent review was used. No additional procedures should be assumed.
How should the two-stage label be read?
A two-stage design generally allows a planned evaluation after an initial stage before continuing to a later stage. Depending on the prespecified rule, the design may stop for an insufficient early signal or continue to gather more information. The arm description states that the first stage would accrue 18 participants, but does not give the decision threshold or the subsequent-stage details. Those features are essential to evaluating the design’s operating characteristics, including the chance of stopping or continuing under different response patterns. They are not reported in the registry record.
The registry’s planning assumptions should be distinguished from the study design. The planning illustration is described as a calculation based on the observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. It considers detecting a stated fraction of the observed effect. These assumptions describe hypothetical planning, not the single-group design recorded for the trial. The numerical illustrations and the observed effect are not reported in the registry record for this analysis. A planning sample size is not observed power, and it does not show the probability that the trial detected a true effect.
Why can event counts need different statistical treatment?
For count outcomes, a basic model may assume that events follow a particular probability distribution. Over-dispersion means that the observed variability is greater than that model expects. When over-dispersion is present, a model that ignores it can understate uncertainty and make results appear more precise than they are. More flexible models or robust uncertainty estimates may be considered in general, but the registry record does not report an over-dispersion assessment or a count-outcome model for this study. The primary endpoint is described as a response rate, not as an event count.
Time-to-event outcomes such as OS and PFS raise additional issues. Their analysis generally depends on event times, censoring, follow-up definitions, and the way participants with incomplete follow-up are handled. The registry lists OS and PFS as secondary outcomes but does not report those definitions, analysis methods, estimates, or event information. No survival model or treatment-effect estimate can therefore be described.
What can analysis records establish?
The registry lists no analysis records. This does not supply an estimate, a test, or an uncertainty measure for ORR, OS, or PFS. A complete statistical interpretation would ordinarily consider the endpoint definition, analysis population, missing outcome handling, estimate, uncertainty interval, and any prespecified decision rule. Several of those details are not reported in the registry record. The design description and endpoint definition can be discussed, but they cannot fill in the absent results or analytic methods.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists 3 outcome measures: the primary ORR outcome and the secondary outcomes OS and PFS. When several outcomes are examined, the chance of at least one apparently unusual result can increase if each is tested separately without adjustment. Statistical methods can address multiplicity by prioritizing outcomes, adjusting testing procedures, or emphasizing estimation and uncertainty rather than isolated significance tests. The registry record does not report whether any multiplicity strategy was planned or used. It also lists no analysis records, so no outcome-specific hypothesis tests or adjustments can be evaluated.
The primary endpoint description identifies 46 participants as evaluable for response, while the enrollment is 54. The reason for that difference and the rules defining evaluability are not reported in the registry record. It does not say how participants without an evaluable response assessment were handled, whether missing response status was classified in any particular way, or whether sensitivity analyses were conducted. No missing-data method or imputation procedure is reported.
The analysis population for OS and PFS is also not described. The registry does not give the follow-up or censoring rules, nor does it say how incomplete outcome information was handled. Those details matter because the analysis population and missing-data approach can affect both the estimate and its uncertainty. The record supports description of the stated endpoints and the response-evaluable population, but not conclusions about the robustness of analyses to missing information or multiplicity.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Nivolumab Treatment | 32 | 54 | 59.3% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Nivolumab Treatment | 2 of 54 |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Nivolumab Treatment | 52 of 54 |
Most frequent serious adverse event terms
| Event term | Organ system | Nivolumab Treatment |
|---|---|---|
| Blood bilirubin increased | Investigations | 6 of 54 |
| Vomiting | Gastrointestinal disorders | 5 of 54 |
| Gastrointestinal disorders - Other | Gastrointestinal disorders | 3 of 54 |
| Nausea | Gastrointestinal disorders | 3 of 54 |
| Sepsis | Infections and infestations | 3 of 54 |
| Acute kidney injury | Renal and urinary disorders | 3 of 54 |
| Anemia | Blood and lymphatic system disorders | 2 of 54 |
| Death NOS | General disorders | 2 of 54 |
| Infections and infestations - Other | Infections and infestations | 2 of 54 |
| Alkaline phosphatase increased | Investigations | 2 of 54 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Nivolumab Treatment |
|---|---|---|
| Abdominal pain | Gastrointestinal disorders | 25 of 54 |
| Fatigue | General disorders | 21 of 54 |
| Weight loss | Investigations | 18 of 54 |
| Vomiting | Gastrointestinal disorders | 13 of 54 |
| Constipation | Gastrointestinal disorders | 12 of 54 |
| Anorexia | Metabolism and nutrition disorders | 12 of 54 |
| Aspartate aminotransferase increased | Investigations | 10 of 54 |
| Pruritus | Skin and subcutaneous tissue disorders | 10 of 54 |
| Anemia | Blood and lymphatic system disorders | 10 of 54 |
| Nausea | Gastrointestinal disorders | 10 of 54 |
Serious events with the highest arm-level rate: blood bilirubin increased (11.1%), vomiting (9.3%), gastrointestinal disorders - other (5.6%), nausea (5.6%), sepsis (5.6%).
The registry reports serious adverse events for the Nivolumab Treatment group, with 54 at risk and 32 affected. It separately reports other adverse events, with 54 at risk and 52 affected, and lists deaths of 2. These are recorded counts, not randomized comparisons. They should not be interpreted as proof that nivolumab caused the events or deaths. The registry does not report attribution, timing, duration, or participant-level overlap between serious and other adverse event categories.
The serious adverse event terms listed most often include Blood bilirubin increased at 11.1%, Vomiting at 9.3%, and Gastrointestinal disorders - Other, Nausea, Sepsis, and Acute kidney injury, each at 5.6%. These percentages are reported as event-category figures; the registry record does not provide additional detail about how repeated events or participants appearing in multiple categories were counted. The categories should not be added together to infer an overall rate, since the record does not say whether categories overlap.
Safety counts are descriptive and do not establish comparative risk without a control group and compatible follow-up. The registry record does not report a detailed adverse-event analysis or a causal assessment. These figures should be read as recorded safety observations in the study group, with those limitations in mind.
12. What the Result Does, and Does Not, Mean
The record establishes that a phase II, single-group study evaluated nivolumab in participants with advanced refractory biliary tract cancers, and that its primary endpoint was ORR after 4 cycles at 16 weeks. It identifies 46 participants as evaluable for response and defines response as CR or PR under RECIST 1.1. However, the numerical ORR and its uncertainty are not reported in the registry record. The record therefore does not support a statement about the size or precision of the observed response result.
The study design also determines the scope of interpretation. With no comparator and no randomized allocation, an outcome observed in the nivolumab group is not a randomized estimate of the difference between nivolumab and another treatment. It cannot separate the effect of the intervention from other possible influences. The listed OS and PFS outcomes likewise cannot be interpreted numerically because their results and analysis details are not reported.
The adverse-event figures describe recorded events in the treatment group, including the reported serious and other adverse-event counts and the listed serious terms. They do not establish causation, comparative risk, or the timing and course of events. The deaths recorded in the adverse-event information should not be attributed to treatment on the basis of these counts alone.
The registry’s two-stage design label and planning illustration do not replace outcome results. The stage decision rule is not reported, and the planning assumptions describe a hypothetical sample-size calculation rather than observed power. The registry also does not report missing-data handling, an analysis plan, or multiplicity procedures. These omissions restrict assessment of how the endpoint was analyzed and how robust any result might be.
In short, the record describes the study question, design, endpoint definition, response-evaluable population, and safety observations. It does not report the numerical primary endpoint result or the statistical details needed to quantify its uncertainty. Those limits preclude conclusions about the magnitude, precision, or comparative effect of the response outcome.
13. Why This Trial Matters Statistically
This study raises several important statistical distinctions. First, a defined clinical endpoint is not the same as a reported result. The registry clearly specifies ORR after 4 cycles at 16 weeks and defines response using CR plus PR under RECIST 1.1. Yet it does not report the response count or numerical ORR. Without those values, readers cannot evaluate the observed response frequency or its precision. The presence of an endpoint definition alone does not answer the study’s primary question.
Second, the response-evaluable population and overall enrollment are not interchangeable. The registry records enrollment of 54 and identifies 46 participants as evaluable for response, but does not explain the difference. Transparent reporting of analysis populations and exclusions is important because an estimate depends on who is included and how missing assessments are handled. The record does not provide enough information to assess those choices.
Third, a single-group study answers a different statistical question from a randomized comparison. It can describe observed outcomes under the study conditions, but it does not provide a controlled estimate of the difference between treatments. This distinction is essential when interpreting response and safety counts; neither should be presented as comparative evidence when no comparator is listed.
Finally, a two-stage design and a planning illustration are meaningful only in context. The record identifies a two-stage design but omits the decision rule. The planning assumptions use equal allocation between two arms even though the recorded trial is single-group, so that illustration is a hypothetical planning scenario rather than a description of the actual allocation. Together, these details show why study design, analysis populations, uncertainty, and complete outcome reporting are all needed for a sound statistical reading.
14. 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 does not report a numerical ORR, response count, confidence interval, or p value. Although 46 participants are identified as evaluable for response, the reason this population differs from the enrollment of 54 is not reported. The evaluability criteria, exclusions, and handling of participants without a response assessment are also not reported in the registry record.
The design is single group, with allocation recorded as not applicable and no comparator listed. Consequently, the study does not provide a randomized comparison. Masking is recorded as none, while the registry does not describe any independent response review or other assessment safeguards. The description of the two-stage design is incomplete: the first stage is described as accruing 18 participants, but the continuation rule, stopping threshold, and later-stage details are not reported.
The record lists OS and PFS as secondary outcomes but does not give their operational definitions, results, analysis methods, or follow-up and censoring information. It lists 3 outcome measures and no analysis records. The statistical analysis plan, uncertainty methods, multiplicity approach, and missing-data procedures are not reported in the registry record.
Safety information is presented as group-level counts and event-category percentages. Attribution, event timing, duration, participant overlap among categories, and comparative safety information are not reported. The record also does not explain the context or attribution of the deaths listed in the adverse-event information. These limitations constrain interpretation of both efficacy and safety; they do not establish that an unreported analysis was or was not conducted outside the registry record.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Study of Nivolumab in Patients With Advanced Refractory Biliary Tract Cancers — NCT02829918.
- Registry record last updated 2026-03-11; status last verified 2026-02.
- Results first posted 2020-03-30.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT02829918 is a completed phase II, single-group study of nivolumab for advanced refractory biliary tract cancers, sponsored by H. Lee Moffitt Cancer Center and Research Institute. The registry records enrollment of 54, a start date of 2016-10-05, and a primary completion date of 2019-03-27. Allocation is not applicable and masking is none.
The primary outcome is ORR after 4 cycles, assessed at 16 weeks and defined as CR plus PR under RECIST 1.1. The response-evaluable population is 46 participants. The record also lists OS and PFS as secondary outcomes. Numerical efficacy results and analysis records are not reported in the registry record.
Safety information lists 32 participants affected by serious adverse events and 52 affected by other adverse events, each among 54 at risk, as well as deaths of 2. The registry describes a two-stage design, but its decision rule is not reported. The planning illustration is hypothetical and does not describe the single-group trial allocation.