This page separates reported registry information from statistical interpretation. The ClinicalTrials.gov record identifies a binary primary endpoint and confirm that results have been posted, but does not provide formal statistical-analysis results for that endpoint.
1. Trial at a Glance
ZUMA-5 is a completed phase 2 clinical trial evaluating axicabtagene ciloleucel in participants with relapsed/refractory indolent non-Hodgkin lymphoma. The registry describes a non-randomized, unmasked, parallel-group treatment study with 159 enrolled participants and two arms.
| Feature | ZUMA-5 |
|---|---|
| Trial name | ZUMA-5 |
| NCT identifier | NCT03105336 |
| Phase | Phase 2 |
| Status | Completed |
| Therapeutic area | Hematology |
| Conditions | Follicular Lymphoma; Marginal Zone Lymphoma; Indolent Non-Hodgkin Lymphoma |
| Enrollment | 159 |
| Allocation | Non-randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Primary endpoint type | Binary |
| Primary endpoints registered | 1 |
| Outcome measures posted | 21 |
| Lead sponsor | Kite, A Gilead Company |
| Sponsor type | Industry |
2. Clinical Question
The registry describes ZUMA-5 as a study of axicabtagene ciloleucel in participants with relapsed/refractory indolent non-Hodgkin lymphoma. The principal statistical question represented by the registered primary endpoint is how frequently participants achieved an objective response according to the specified Lugano Classification assessment by central assessment.
Population
Participants with relapsed/refractory indolent non-Hodgkin lymphoma, with follicular lymphoma and marginal zone lymphoma among the listed conditions.
Intervention
Axicabtagene ciloleucel is the principal biological intervention identified in the ClinicalTrials.gov record. Cyclophosphamide and fludarabine are also listed as interventions.
Comparator
The registry design is non-randomized with two parallel arms. The ClinicalTrials.gov record does not identify a randomized control arm.
Primary question
What percentage of participants achieved an objective response per the Lugano Classification by central assessment?
3. Trial Design
The registry profile classifies ZUMA-5 as a phase 2, non-randomized, parallel study with no masking and a primary purpose of treatment. This design structure is important when interpreting the primary binary endpoint because a non-randomized study does not provide the same randomized between-group causal comparison as a conventional controlled trial.
Interventions recorded in the registry
Axicabtagene ciloleucel
- Listed as a biological intervention.
- Associated with the primary study treatment in the ClinicalTrials.gov record.
Cyclophosphamide and fludarabine
- Cyclophosphamide is listed as a drug intervention.
- Fludarabine is listed as a drug intervention.
4. Primary Endpoint
| Endpoint | Registry definition / time frame | Endpoint type |
|---|---|---|
| Objective Response Rate (ORR) | Percentage of Participants With Objective Response Per the Lugano Classification by Central Assessment | Binary |
| Time frame | Up to 85.6 months | Registry-reported time frame |
Registry definition
The registry definition describes objective response as OR: CR (complete metabolic response [CMR] + complete radiological response [CRR]) + PR (partial MR response [PMR] + partial RR [PRR]). The definition further describes CMR using the positron emission tomography 5-point scale, including scores of 1, 2, and 3, and identifies response assessment according to the Lugano Classification by central assessment.
5. Planned Analysis
ClinicalTrials.gov indicates that results have been posted for ZUMA-5, but the ClinicalTrials.gov record contains no formal statistical analyses. Therefore, there is no reported formal effect estimate, confidence interval, or p-value for the primary ORR endpoint in the ClinicalTrials.gov record.
For a binary response endpoint such as ORR, the basic descriptive quantity is the observed response proportion:
The proportion is commonly multiplied by 100 to express the result as a percentage. The relevant denominator should be the prespecified response-analysis population rather than automatically assuming that the overall enrollment is the denominator.
A confidence interval for a single response proportion would normally quantify uncertainty around the estimated population response rate. Depending on the statistical analysis plan and the sample size, an exact binomial interval or another prespecified binomial confidence interval could be used. The ClinicalTrials.gov record does not identify which interval method, if any, was used for the posted ORR result.
What can be described
The primary endpoint is a binary objective-response measure assessed by central assessment under the Lugano Classification, with a time frame of up to 85.6 months.
What the registry record does not include
The ClinicalTrials.gov record does not provide the response count, analysis denominator, confidence interval, hypothesis-test result, or formal statistical-analysis method.
The absence of a registry-reported statistical-analysis entry is important. A posted outcome measure is not the same thing as a documented formal statistical comparison. Without the response numerator and the applicable analysis denominator, a numerical ORR cannot be calculated from the ClinicalTrials.gov record without introducing information from another source.
Likewise, a confidence interval cannot legitimately be reverse-engineered from the overall enrollment of 159 participants because the relevant response-evaluable population may differ. The correct denominator is a statistical-analysis detail, not a quantity that should be assumed.
6. Statistical Methodology
Binary endpoint analysis
The primary endpoint is explicitly classified as binary. In a binary endpoint analysis, each participant contributes a response status according to the prespecified assessment rule. The principal descriptive estimate is therefore a proportion.
Here, x is the number of participants classified as responders and n is the relevant analysis population. Neither x nor n is reported in the ClinicalTrials.gov record.
Confidence intervals for a response rate
A confidence interval adds uncertainty information to the point estimate. For a binary endpoint, the interval is based on the binomial sampling distribution. The choice between exact and approximate methods should be driven by the prespecified statistical analysis plan and the characteristics of the observed data.
This distinction matters because a response percentage without a measure of precision does not communicate how much statistical uncertainty surrounds the estimate.
Central assessment
The primary endpoint specifies assessment by central assessment. Central assessment is relevant to statistical interpretation because response classification can otherwise be affected by differences in assessment procedures or reviewer judgment. The ClinicalTrials.gov record does not provide enough detail to characterize the central-review process beyond its identification in the endpoint definition.
Non-randomized allocation
Non-randomized allocation changes the inferential question. In a randomized trial, differences between treatment groups can be interpreted within the framework created by random assignment. In a non-randomized study, observed response may reflect the treatment as well as patient selection, disease characteristics, prior treatment history, or other factors associated with enrollment.
Time frame
The registered primary endpoint has a time frame of up to 85.6 months. For a response endpoint, the time frame establishes the period during which the endpoint can be assessed. It should not automatically be interpreted as a median follow-up time or as the duration of response; those are different statistical quantities and are not reported here.
7. Statistical Methods Explained
What does ORR measure?
ORR measures the percentage of participants who meet the prespecified definition of objective response. It is a response endpoint, not a direct measure of overall survival or progression-free survival. A participant can respond and later experience disease progression, while another participant may not meet the response definition but still have stable disease for a period of time.
Why is ORR treated as a binary endpoint?
The underlying response assessment contains clinically defined categories, but the primary endpoint reduces those categories to the binary outcome of response versus no objective response. This makes a binomial proportion a natural summary of the endpoint.
Why does the analysis denominator matter?
The percentage depends on both the number of responders and the number of participants included in the response analysis. Using the overall enrollment as the denominator without confirming that it is the prespecified response population can produce a different and potentially incorrect estimate.
What does a confidence interval add?
A confidence interval communicates statistical precision around the observed response proportion. A narrow interval indicates greater precision under the chosen statistical model and sampling framework, whereas a wider interval indicates more uncertainty. The interval does not describe the range of responses that an individual patient can experience.
Why does non-randomization matter?
Without randomization, differences in patient characteristics can be associated with treatment assignment or study participation. Consequently, an observed response rate is primarily descriptive unless additional design and analytical methods establish a causal comparison. The ClinicalTrials.gov record does not report such a formal comparative analysis.
Why should central assessment be distinguished from local assessment?
Central assessment provides a standardized response classification framework across the study population. When the endpoint explicitly specifies central assessment, the response result should be interpreted according to that assessment rather than silently substituting another source of response determination.
Why should the 85.6-month time frame not be called a median follow-up?
The registry identifies up to 85.6 months as the endpoint time frame. “Up to” describes the maximum registered assessment period. A median follow-up is a separate summary statistic requiring a distribution of follow-up times, which is not provided in the ClinicalTrials.gov record.
8. Safety Results
the ClinicalTrials.gov record reports serious adverse events by arm and disease/treatment subgroup. These figures are reported as affected participants divided by participants at risk. They should be read as subgroup-specific serious-adverse-event counts rather than as a single overall safety percentage.
| Reported subgroup | Serious adverse events |
|---|---|
| Axicabtagene ciloleucel: Follicular Lymp | 65/124 |
| Axicabtagene ciloleucel: Marginal Zone L | 19/28 |
| Axicabtagene ciloleucel: Retreatment (FL | 5/14 |
| Axicabtagene ciloleucel: Retreatment (MZ | 1/2 |
This is a useful example of why denominator discipline is essential in clinical-trial statistics. A value such as 65/124 is interpretable as the reported number affected among the reported number at risk for that subgroup. It does not establish the serious-adverse-event rate for all 159 enrolled participants.
9. Interpreting a Non-Randomized Response Study
The primary endpoint is an objective response proportion. In a non-randomized study, that proportion answers a descriptive question about the observed study population: how frequently participants met the prespecified response definition.
It does not, by itself, estimate the causal difference between axicabtagene ciloleucel and an alternative treatment. A causal comparative interpretation would require an appropriate comparator and design or analytical framework capable of addressing confounding.
A response percentage is only one part of the statistical description. The numerator, denominator, confidence interval, assessment method, analysis population, and timing of assessment all affect interpretation. The ClinicalTrials.gov record provides the endpoint definition and follow-up frame but do not provide the numerical response estimate or its uncertainty.
If a formal hypothesis test had been prespecified, its p-value would quantify compatibility of the observed data with the tested null hypothesis under the specified statistical model. It would not measure the size of the treatment effect, clinical importance, or the probability that the treatment is effective. No such p-value is reported in the ClinicalTrials.gov record for this page.
10. Limitations
- No formal statistical analyses reported: the ClinicalTrials.gov record states that results are posted, but no formal statistical analyses were posted. Therefore, numerical effect estimates, confidence intervals, and p-values cannot be reported from the ClinicalTrials.gov record.
- Non-randomized allocation: The registry classifies allocation as non-randomized. Observed response therefore should not automatically be interpreted as a randomized causal treatment effect.
- Unknown response denominator: Overall enrollment is 159, but the ClinicalTrials.gov record does not establish that all 159 participants comprise the response-analysis population.
- Safety subgroup denominators: Serious adverse-event data are provided for specific subgroups, and the registry-reported extract does not establish whether these subgroups are mutually exclusive.
- No reported subgroup estimates: The ClinicalTrials.gov record does not provide response estimates by disease subtype, retreatment status, or other baseline characteristics.
- No reported inferential framework: The ClinicalTrials.gov record does not identify a formal hypothesis-testing procedure, confidence-interval method, multiplicity strategy, missing-data method, interim-analysis method, or Bayesian framework for the primary endpoint.
11. Why This Trial Matters Statistically
ZUMA-5 is a useful statistical teaching case because the study combines a clearly defined binary clinical endpoint with a non-randomized design. That combination forces an important distinction between estimating an observed response proportion and estimating a causal treatment effect.
| Concept | How it appears in ZUMA-5 |
|---|---|
| Binary endpoint | Objective Response Rate is the single registered primary endpoint and is classified as binary. |
| Response proportion | ORR is naturally summarized as the proportion of participants meeting the objective-response definition. |
| Central assessment | The registered primary endpoint specifies assessment by central assessment. |
| Lugano Classification | Objective response is defined using the Lugano Classification framework reported in the ClinicalTrials.gov record. |
| Non-randomized design | Allocation is explicitly recorded as non-randomized. |
| Parallel design | The registry identifies a parallel design model with two arms. |
| Precision | A confidence interval would ordinarily accompany a response proportion to quantify statistical uncertainty. |
| Denominator discipline | The response-analysis denominator must be distinguished from overall enrollment unless the registry explicitly equates them. |
| Safety analysis | Serious adverse events are reported as affected participants divided by participants at risk in specified subgroups. |
12. Design Features Not Reported in the ClinicalTrials.gov record
Several statistical-design features that are important in clinical-trial analysis are not identified in the ClinicalTrials.gov record. They should therefore not be inferred from the trial phase or endpoint type.
| Design topic | What the ClinicalTrials.gov record establishes |
|---|---|
| Non-inferiority margin | Not reported. |
| Crossover | Not reported. |
| Factorial design | Not reported; the registered design model is parallel. |
| Multiplicity adjustment | Not reported. |
| Interim analysis | Not reported. |
| Missing-data / imputation method | Not reported. |
| Stratification factors | Not reported in the ClinicalTrials.gov record. |
| Bayesian methods | Not reported. |
13. Trial Timeline
Study start
The registry lists June 6, 2017 as the study start date.
Primary completion
The registry lists December 20, 2024 as the primary completion date.
Current trial status in the ClinicalTrials.gov record
the ClinicalTrials.gov record identifies the study status as completed.
14. Statistical Interpretation of the Primary Endpoint
What the endpoint asks
Whether each participant meets the prespecified objective-response definition under the Lugano Classification by central assessment.
What the summary would be
The primary descriptive result would normally be the proportion of participants classified as responders within the relevant response-analysis population.
What precision requires
A confidence interval around the response proportion provides information about statistical uncertainty.
What the design limits
Because allocation is non-randomized, the observed response proportion is not automatically a causal comparison against an alternative therapy.
This formula describes the statistical structure of the endpoint only. The registry-reported ZUMA-5 data do not provide the values of x and n required to calculate its numerical result.
This distinction is especially important for educational interpretation. It would be easy to take the enrollment figure of 159 and assume that it is the response denominator, but doing so would introduce an unsupported assumption. The statistically correct approach is to use the prespecified response-analysis population once it is known.
15. Related Tutorials
Learn more about the methods used in this trial:
16. Related Calculators
17. Sources
- ClinicalTrials.gov: NCT03105336.
- Linked publication: PubMed PMID 42745632.
- Linked publication: PubMed PMID 41100801.
- Linked publication: PubMed PMID 40536814.
- Linked publication: PubMed PMID 39240498.
Continue with Clinical Biostats statistical methods
Explore the statistical concepts behind binary endpoints, response-rate estimation, confidence intervals, and clinical-trial analysis.
18. Record Summary
ZUMA-5 is a completed phase 2, non-randomized, parallel study with 159 enrolled participants and one registered binary primary endpoint: Objective Response Rate (ORR): Percentage of Participants With Objective Response Per the Lugano Classification by Central Assessment, with a time frame of up to 85.6 months.
The statistical structure of the primary endpoint is straightforward—a response proportion with an appropriate measure of uncertainty—but its interpretation depends critically on the response-analysis population, the central-assessment definition, and the non-randomized study design. The ClinicalTrials.gov record does not provide the response numerator, response denominator, confidence interval, or formal hypothesis-test result, so those quantities are intentionally not reconstructed here.