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 Pembrolizumab in Patients With Poor-Prognosis Carcinoma of Unknown Primary Site (CUP) (NCT03391973), a na phase 2 study. The primary outcome is Number of Participants That Develop an Objective Response to Treatment (Objective Response Rate, ORR)..
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT03391973 |
| Phase | Phase 2 |
| Condition | Carcinoma of Unknown Primary |
| Allocation | Not applicable (single group) |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 33 |
| Arms in the primary outcome | n/a |
| Lead sponsor | AHS Cancer Control Alberta |
| Results posted | Yes |
2. Clinical Question
Population
The trial concerned treatment-naïve patients with poor-prognosis carcinoma of unknown primary site (CUP). The registry record says 33 patients were enrolled and 27 were evaluable for response. It defines evaluability as receiving study treatment and not discontinuing before completing a response evaluation period at 9 weeks.
Intervention
The intervention was pembrolizumab injection, given intravenously. The arm description specifies a dose of 200 mg on Day 1 of each 3-week cycle. The brief summary describes treatment for up to 24 months.
Comparator
There was no comparator arm. The registry classifies the trial as single-group, with allocation and masking not applicable. Consequently, the trial record does not provide a randomized contrast between pembrolizumab and another treatment or control.
Primary question
The primary endpoint asks how many participants developed an objective response to treatment, assessed using RECIST 1.1. Response assessment used MRI or CT imaging. The primary outcome timeframe is within 3 years. The registry record does not report a numerical response count or response rate, so the result of that question cannot be read from the endpoint information alone.
The record also lists overall survival and progression-free survival as secondary outcomes. It does not provide numerical results for those outcomes in the stated outcome information. The central statistical question is therefore about the observed response experience in the treated, evaluable population—not about a treatment effect relative to a control group. Without a comparator, a response count would describe what occurred in this trial but would not, by itself, estimate a randomized difference attributable to pembrolizumab.
3. Trial Design
This was a phase 2, multicentre, single-arm trial sponsored by AHS Cancer Control Alberta. The registry lists the condition as carcinoma of unknown primary and the intervention as pembrolizumab injection. The trial status is completed. Its recorded start date is 2018-08-24, and its primary completion date is 2023-06-16.
The arm description specifies pembrolizumab 200 mg by intravenous infusion on Day 1 of each 3-week cycle. The brief summary describes treatment every 3 weeks for up to 24 months. There was one treatment group, no applicable control, and no masking. These design features mean that the trial can describe outcomes among participants receiving the intervention, but it cannot make a randomized comparison with an untreated or differently treated group.
The enrollment information is not fully consistent across registry fields. The main enrollment field and the primary endpoint population description state that 33 patients were enrolled; the brief summary states a total number of trial subjects of 25. The primary endpoint description says 27 were evaluable. These are not interchangeable counts: enrollment and response evaluability are described separately, and the discrepancy between the registry entries is not reconciled in the record.
The primary outcome is the number of participants developing an objective response, with a timeframe of within 3 years. The record also names overall survival and progression-free survival as secondary outcomes. It lists five outcome measures, but numerical results for these outcomes are not reported in the registry record. The protocol’s detailed statistical analysis plan, including any planned analysis populations or procedures beyond the endpoint descriptions, is not reported in the registry record.
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 | Arm 1 - Pembrolizumab |
|---|---|
| STARTED | 33 |
| COMPLETED | 27 |
| NOT COMPLETED | 6 |
Reasons for not completing this period:
| Reason | Arm 1 - Pembrolizumab |
|---|---|
| Lack of Efficacy | 2 |
| Physician Decision | 2 |
| Adverse Event | 1 |
| Development intercurrent illness preventing ongoing treatment | 1 |
Analysis population for the primary outcome, as defined in the registry: 33 patients were enrolled and 27 were evaluable. Subjects enrolled were considered evaluable for response if they received study treatment and did not discontinue from study treatment before completion of a response evaluation period (at 9 weeks)
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: Eligible patients were identified and treated at 3 academic cancer centers in Canada, (Tom Baker Cancer Centre now the Arthur J.E. Child Comprehensive Cancer Centre, Calgary, Alberta; Cross Cancer Institute, Edmonton, Alberta; Ottawa Cancer Centre, Ottawa, Ontario). Eligible patients were 18 years of age or older, and have histologically documented, metastatic carcinoma with no primary site identified.
| Characteristic | Unit | Arm 1 - Pembrolizumab |
|---|---|---|
| Participants analysed | 33 | |
| Age, Customized: Age | years | 64 (21 to 80) |
| Sex: Female, Male: Female | Participants | 14 |
| Sex: Female, Male: Male | Participants | 19 |
| Race and Ethnicity Not Collected | Participants | — |
| ECOG: 0 | Participants | 5 |
| ECOG: 1 | Participants | 28 |
| Histology: Adenocarcinoma | Participants | 22 |
| Histology: Squamous Cell Carcinoma | Participants | 2 |
| Histology: Poorly Differentiated | Participants | 8 |
| Histology: Other | Participants | 1 |
| Sites of Metastatic Disease: Lung | Participants | 14 |
| Sites of Metastatic Disease: Liver | Participants | 23 |
| Sites of Metastatic Disease: Bone | Participants | 8 |
| Sites of Metastatic Disease: Lymph Nodes | Participants | 20 |
| Sites of Metastatic Disease: Skin/soft tissues | Participants | 1 |
| Sites of Metastatic Disease: Genitourinary | Participants | 1 |
| Sites of Metastatic Disease: Other | Participants | 8 |
| Number of Metastatic Sites: 1 | Participants | 5 |
| Number of Metastatic Sites: 2 | Participants | 12 |
| Number of Metastatic Sites: 3 or more | Participants | 16 |
| Baseline Endoscopy: Baseline upper and lower endoscopy | Participants | 22 |
| Baseline Endoscopy: Baseline upper and no lower endoscopy | Participants | 2 |
| Baseline Endoscopy: Baseline lower and no upper endoscopy | Participants | 2 |
| Baseline Endoscopy: No endoscopy | Participants | 7 |
| MMR: MMR Proficient | Participants | 6 |
| MMR: MMR Deficient | Participants | 2 |
| MMR: MMR status unknown/not done | Participants | 25 |
| Elevated LDH at baseline | Participants | 10 |
| Baseline PET: Baseline PET | Participants | 3 |
| Baseline PET: no PET | Participants | 30 |
6. Primary Endpoint
The registry lists the primary outcome as Number of Participants That Develop an Objective Response to Treatment (Objective Response Rate, ORR)., measured over Within 3 years.
| Element | Registry entry |
|---|---|
| Result type | Count Of Participants |
| Measure of spread | not applicable |
| Unit | Participants |
| Endpoint classification | Binary Endpoint |
The primary endpoint is the number of participants who develop an objective response to treatment, also described as the objective response rate (ORR). It is assessed using RECIST 1.1, with tumor response evaluated from MRI or CT imaging. The registry gives the outcome timeframe as within 3 years. The endpoint population description says that participants were evaluable if they received study treatment and did not discontinue before completing a response evaluation period at 9 weeks.
RECIST response categories distinguish tumor shrinkage, stability, and progression. Complete response means disappearance of all target lesions. Partial response means at least a 30% decrease in the sum of target-lesion diameters from baseline. Stable disease meets neither the response nor progression criteria, while progressive disease includes at least a 20% increase in that sum or the appearance of new lesions. These definitions specify how imaging findings are classified; they do not establish how many participants met any category in this trial.
ORR is used to describe the frequency of complete or partial tumor responses under the stated response criteria. It focuses on tumor shrinkage rather than survival or disease control more broadly. The record defines disease control rate as the proportion achieving complete response, partial response, or stable disease as best overall response, and defines median duration of response from the first date of complete or partial response to progression or death. Those definitions are distinct from the primary ORR endpoint. The registry record does not report a numerical ORR result, response-category counts, or a confidence interval.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry describes a single-arm phase 2 trial in which response was to be assessed using RECIST 1.1. The primary outcome is framed as the number of participants who developed an objective response, with a stated assessment timeframe within 3 years. The endpoint description gives 33 enrolled and 27 evaluable participants, while a separate brief summary states a total of 25 trial subjects. The record does not explain this disagreement. It also does not state a numerical primary response result. Therefore, the posted registry information does not support a quantitative conclusion about the observed response rate or its precision.
A response count and an ORR are descriptive measures of outcomes in the population assessed. In general, a confidence interval communicates the range of values reasonably compatible with the observed data under the chosen statistical method; it helps show the uncertainty around an estimate. No response estimate or corresponding confidence interval is reported in the registry outcome information. Without those results, it is not possible to judge the precision of the primary endpoint estimate from the registry record.
The absence of a control arm also matters for interpretation. In general, a single-arm study can document outcomes among people receiving a treatment, but it cannot separate treatment effects from other explanations using a randomized between-group contrast. The record provides no like-for-like control group, and the trial is explicitly classified as single-group. Any response result, if reported elsewhere, would need to be interpreted as an outcome observed in this trial population rather than as a randomized estimate of benefit over an alternative.
The record lists overall survival and progression-free survival as secondary outcomes but does not report their numerical results in the outcome information described. It gives no numerical response-category counts, duration-of-response result, or statistical uncertainty for the primary endpoint. The registry also lists five outcome measures and no analysis entries. That does not establish which statistical analyses were conducted outside the registry; the analysis plan and any additional analyses are not reported in the registry record.
Safety information is also descriptive. The adverse-event fields report 27 at risk and 2 affected for serious events, 27 at risk and 27 affected for other events, and 26 deaths. The record lists fatigue/generalized weakness and fever, nausea, vomiting among the leading serious-event terms, each with a reported percentage of 3.7%. These entries do not establish that an event was caused by treatment. The registry record does not report the relationship assessment, timing, or detailed context needed to attribute events or deaths to pembrolizumab.
Overall, the registry establishes the design, treatment schedule, endpoint definitions, and some population and safety counts. It does not provide the numerical efficacy results needed to quantify the primary endpoint. The completed status should not be confused with complete reporting of results: status indicates the recorded trial state, while endpoint estimates and their precision are not reported in the registry record.
9. Statistical Methods Explained
The statistical information in the registry should be read in light of the study design and the specific outcome definitions. The trial is classified as phase 2 and single-group, with no applicable allocation or masking. The primary outcome is a participant count for objective response, assessed under RECIST 1.1. The record does not provide a numerical response estimate, confidence interval, or inferential analysis for that endpoint. The statistical explanations below are general unless explicitly tied to a registry detail.
What does a single-arm result describe?
In general, a single-arm study estimates or describes outcomes among people who receive the study intervention. It does not supply a randomized comparison against another treatment or a control condition. A response proportion can summarize how often a defined response was observed in the assessed group, but a comparison with an external expectation would require additional assumptions and information. The registry record does not identify a comparator or report a formal comparison. The design therefore supports descriptive interpretation of the treated group, not a randomized treatment contrast.
For a binary outcome such as whether a participant had an objective response, the numerator is the number meeting the response definition and the denominator is the relevant analysis population. The primary endpoint description identifies 27 participants as evaluable and gives the enrollment count as 33. It does not give the response numerator or a reported response proportion. The separate brief summary says 25 trial subjects, and the record does not reconcile that with the enrollment and evaluability entries. The analysis population and outcome count should not be inferred from the safety fields.
How should uncertainty be understood?
In general, a confidence interval gives a range of estimates compatible with the observed data under a specified method and its assumptions. For a response proportion, the interval’s width conveys how much uncertainty remains around the estimated proportion. Different interval procedures can behave differently, particularly with small samples or proportions near the boundaries. The statistical method used for an ORR confidence interval is not reported in the registry record, and no interval is included with the primary outcome information. Thus, neither a numerical estimate nor its precision can be assessed from that information.
In general, a p-value summarizes how compatible data are with a specified null model, assuming that model and the analysis procedure. It is not a measure of the size or clinical importance of an effect. The registry outcome information contains no p-value or hypothesis test for ORR. It also does not report a prespecified null response rate or other benchmark. No hypothesis-testing conclusion should be inferred from the endpoint definition alone.
What do the planning illustrations mean?
The described planning approach is an illustration based on an observed effect, not the trial’s own design. It assumes a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is described as a stated fraction of the observed effect. Those assumptions do not match the recorded single-arm design, which has no two-arm allocation. Accordingly, such planning values are not a reconstruction of the trial’s enrollment plan and are not observed power. In general, power is a design property calculated under specified assumptions; a post hoc calculation based on an observed estimate cannot establish the probability that the completed study detected a true effect. The numerical planning outputs and the observed effect used as their basis are not reported in the registry record.
Why do endpoint definitions matter?
RECIST 1.1 supplies the rules for assigning imaging-based response categories. Complete response, partial response, stable disease, and progressive disease are not interchangeable outcomes. ORR concerns complete and partial response, whereas the registry’s disease control definition also includes stable disease. The defined median duration of response concerns the interval from the first complete or partial response to progression or death. A result for one of these measures cannot stand in for another. Although the record describes all these definitions, it does not provide the category counts or numerical results needed to evaluate them.
Over-dispersion is a general issue for count outcomes when observed variability exceeds what a simple model assumes. It is not the primary modeling question for a binary response indicator, and the registry does not report a model or a dispersion assessment. For time-to-event outcomes such as overall survival or progression-free survival, censoring and event timing are important; the registry lists these outcomes but does not provide the corresponding analyses or methods. The statistical analysis plan, handling of censoring, and any model diagnostics are not reported in the registry record.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists objective response as the primary outcome and overall survival and progression-free survival as secondary outcomes. It also states that there are five outcome measures, but the outcome information described does not include numerical results for those measures. The record does not report how testing across multiple outcomes, time points, or response categories was handled. Any multiplicity adjustment, hierarchy of tests, or distinction between confirmatory and exploratory analyses is not reported in the registry record.
The primary endpoint population description says that participants were considered evaluable if they received study treatment and did not discontinue before completing the response evaluation period at 9 weeks. It reports 33 enrolled and 27 evaluable. This is an evaluability rule, not a complete account of all analysis populations. The record does not report a full intention-to-treat analysis, a per-protocol population, or the treatment of participants with missing or indeterminate response assessments. Reasons for non-evaluability, missing imaging, and the methods used to address missing data are not reported in the registry record.
The enrollment entries also differ: the main field and endpoint description state 33 enrolled, while the brief summary states a total of 25 trial subjects. The registry does not reconcile these figures. The safety fields use 27 as the at-risk count for both serious and other adverse events. These different counts refer to different record fields, and the registry does not explain their relationship to the response-evaluable population.
No analysis entries are listed, and no numerical primary endpoint result or confidence interval is reported in the outcome information. This does not prove that no analyses were performed; details of the statistical analysis plan, interim analyses, multiplicity procedures, and missing-data methods are not reported in the registry record.
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 |
|---|---|---|---|
| Arm 1 - Pembrolizumab | 2 | 27 | 7.4% |
Deaths
| Arm | Deaths (participants affected of at risk) |
|---|---|
| Arm 1 - Pembrolizumab | 26 of 27 |
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 |
|---|---|
| Arm 1 - Pembrolizumab | 27 of 27 |
Most frequent serious adverse event terms
| Event term | Organ system | Arm 1 - Pembrolizumab |
|---|---|---|
| Fatigue, generalized weakness | General disorders | 1 of 27 |
| Fever, nausea, vomiting | Gastrointestinal disorders | 1 of 27 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Arm 1 - Pembrolizumab |
|---|---|---|
| Fatigue/malaise/weakness | General disorders | 7 of 27 |
| Joint Pain | Musculoskeletal and connective tissue disorders | 6 of 27 |
| Nausea | General disorders | 5 of 27 |
| Rash | Skin and subcutaneous tissue disorders | 4 of 27 |
| Pruritis | Skin and subcutaneous tissue disorders | 4 of 27 |
| Cough/Dyspnea | Respiratory, thoracic and mediastinal disorders | 4 of 27 |
| Vomiting | Gastrointestinal disorders | 3 of 27 |
| Anorexia/Decreased food intake | General disorders | 2 of 27 |
| Diarrhea | Gastrointestinal disorders | 2 of 27 |
| Fever | General disorders | 2 of 27 |
Serious events with the highest arm-level rate: fatigue, generalized weakness (3.7%), fever, nausea, vomiting (3.7%).
The adverse-event record lists one arm, Pembrolizumab. For serious adverse events, it reports 27 at risk and 2 affected. For other adverse events, it reports 27 at risk and 27 affected. The record also lists 26 deaths. These are registry counts; the entries do not establish that pembrolizumab caused an event or death, and the relationship of the reported deaths to treatment is not reported in the registry record.
The leading serious-event terms listed are “Fatigue, generalized weakness” and “Fever, nausea, vomiting,” each reported as 3.7%. These percentages are registry entries, not a substitute for the underlying clinical context. The record does not provide the event-level timing, duration, severity grading, recurrence, or causal assessment in the information described. It also does not explain how the combined terms are grouped or how event counts relate to individual participants.
At-risk and affected counts should be read within their respective adverse-event fields; they are not efficacy outcomes and should not be used to infer response. The listed death count does not specify cause or treatment attribution. Without fuller detail, these figures describe what the registry records but cannot establish comparative safety or causation.
12. What the Result Does, and Does Not, Mean
The registry record documents a completed, single-arm phase 2 study of pembrolizumab in treatment-naïve patients with poor-prognosis carcinoma of unknown primary site. It identifies objective response assessed by RECIST 1.1 as the primary endpoint and gives the response-evaluable population definition. It also records the treatment schedule, secondary outcomes, and selected adverse-event counts. These details help establish what was studied and how the primary outcome was meant to be classified.
The record does not report the numerical objective response result, response-category counts, or a confidence interval. Therefore, it does not allow readers to quantify the observed ORR or the uncertainty around it from the primary outcome information. Numerical results for the listed overall survival and progression-free survival outcomes are likewise not reported in the registry outcome information described. The trial’s completed status does not fill in those missing estimates.
Even if a response estimate were available, the single-arm design would mean it described outcomes in the treated study group rather than a randomized difference against a comparator. In general, outcomes in a single group may reflect multiple influences, and the design cannot isolate a treatment effect through randomization. The registry record does not provide a control group, and its classification makes clear that there was no like-for-like randomized comparison.
The safety counts also require restraint. The record lists affected participants, at-risk counts, event terms, and deaths, but does not provide the causal assessments and clinical context needed to attribute those events to the intervention. The enrollment figures are not fully consistent across registry fields, and the record does not reconcile them.
In short, the registry is informative about the trial’s design and definitions but does not report enough numerical efficacy results to support a quantitative conclusion about response, survival, or precision. It does not establish comparative efficacy or causal safety findings.
13. Why This Trial Matters Statistically
This trial is statistically notable for the contrast between a clearly defined response endpoint and the limited numerical outcome reporting in the registry record. The primary endpoint is an imaging-based objective response under RECIST 1.1, with explicit criteria for complete response, partial response, stable disease, and progression. Such definitions matter because they make the intended outcome classification more reproducible and distinguish response from broader disease control.
The single-arm design also shapes what can be learned. It can describe response and safety experience among participants who received pembrolizumab, but it does not provide a randomized comparator. That distinction is central when interpreting any observed outcome: descriptive results and comparative treatment effects answer different questions. No numerical ORR result is reported in the registry outcome information, so the response question cannot be quantified from that record.
The population entries illustrate why denominators need careful attention. The record says 33 patients were enrolled and 27 were evaluable, while the brief summary states a total of 25 trial subjects. These figures are not reconciled. A clear analysis population is essential for interpreting a response proportion, and the registry’s evaluability rule alone does not explain every participant’s outcome status.
The record also lists survival outcomes and adverse-event information without supplying a complete statistical analysis or causal account. This distinction between outcome definitions, observed results, and analysis methods is useful in clinical research: a registered endpoint does not itself constitute a reported finding, and event counts do not by themselves establish causation. The record supports a careful description of design and reporting gaps, rather than an estimate of treatment effect.
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.
Several trial-specific limitations constrain interpretation of the registry record. First, the single-arm design has no randomized comparator, so it cannot estimate a treatment difference against another intervention or control condition. Second, the primary endpoint is defined, but no numerical objective response result, response-category count, or confidence interval is reported in the outcome information. The listed secondary outcomes, overall survival and progression-free survival, also lack numerical results in the record information described.
The population counts are not consistent across fields. The enrollment field and primary endpoint description state 33 enrolled, the endpoint description says 27 evaluable, and the brief summary states a total of 25 trial subjects. The record does not reconcile these values. Although it supplies an evaluability definition involving treatment receipt and completion of response assessment at 9 weeks, it does not report reasons for missing or unevaluable outcomes, the handling of missing assessments, or other analysis population definitions.
The registry does not report the statistical analysis plan, confidence-interval method, missing-data procedure, multiplicity strategy, or any interim-analysis details. It lists no analysis entries, but this does not establish what analyses may have been conducted outside the registry. The planning assumptions described for two-arm illustrations do not reproduce the recorded single-arm design, and the planning outputs are not observed power.
Safety entries provide counts and selected event terms, but the information described does not establish event timing, severity details, or treatment attribution. The death count is not accompanied by cause or relationship to treatment. These reporting limitations prevent a complete assessment of efficacy precision, comparative effects, and safety causality from the registry record alone.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Pembrolizumab in Patients With Poor-Prognosis Carcinoma of Unknown Primary Site (CUP) — NCT03391973.
- Registry record last updated 2026-07-02; status last verified 2026-06.
- Results first posted 2026-07-02.
- 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
The registry record describes a completed, multicentre, single-arm phase 2 trial sponsored by AHS Cancer Control Alberta. It lists pembrolizumab 200 mg by intravenous infusion on Day 1 of each 3-week cycle, with treatment described for up to 24 months. The primary outcome is the number of participants developing an objective response under RECIST 1.1, within 3 years.
The record states 33 enrolled and 27 evaluable in the primary endpoint description, while the brief summary states a total of 25 trial subjects. It lists overall survival and progression-free survival as secondary outcomes. Numerical results for the primary and secondary outcomes are not reported in the registry outcome information described. Adverse-event fields list 2 affected by serious events among 27 at risk, 27 affected by other events among 27 at risk, and 26 deaths; causal attribution is not reported.