A Study to Evaluate the Safety, Tolerability, Pharmacokinetics, Immunogenicity, and Preliminary Efficacy of Atezolizumab (Anti-Programmed Death-Ligand 1 [PD-L1] Antibody) in Pediatric and Young Adult Participants With Solid Tumors: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate the Safety, Tolerability, Pharmacokinetics, Immunogenicity, and Preliminary Efficacy of Atezolizumab (Anti-Programmed Death-Ligand 1 [PD-L1] Antibody) in Pediatric and Young Adult Participants With Solid Tumors (NCT02541604), a na phase 1, phase 2 study. The primary outcome is Percentage of Participants With an Objective Response (Complete Response [CR] or Partial Response [PR]) as Determined by the Investigator Using Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST v1.1) in Participants With Solid Tumors.

2. Clinical Question

This question concerns an observed response proportion in a treated population. It is different from asking whether atezolizumab improves outcomes compared with another approach: without a control group, the study design does not provide that direct contrast. The study’s early-phase scope also included safety and pharmacokinetic questions, while its response endpoint was described as preliminary efficacy.

The population and response criteria require care in interpretation. The study enrolled pediatric and young adult participants with solid tumors after prior treatment had been ineffective, and the primary endpoint used an investigator assessment under RECIST v1.1. Other listed response outcomes use different criteria for particular diagnoses, including neuroblastoma, lymphoma, and ATRT. Those endpoints are not interchangeable with the primary RECIST v1.1 endpoint. The registry record also notes a cohort assignment issue for a response in Cohort 5; that detail should be kept in mind when interpreting cohort-specific descriptions.

3. Trial Design

This was an early-phase, multicenter, open-label, single-group study listed as Phase 1 and Phase 2. The registry lists Hoffmann-La Roche as sponsor and records an enrollment of 87. The study started on 2015-11-30, and primary completion is recorded as 2019-06-06. Its current status is TERMINATED; the reason for termination is not reported in the registry record.

Participants received atezolizumab by intravenous infusion on Day 1 of each 21-day cycle, with a maximum dose of 1200 milligrams. The study description identifies pediatric and young adult participants with previously treated solid tumors for which prior treatment had been proven ineffective. The registry record does not give further dosing details, such as treatment duration or dose adjustments.

Open-label means masking was not used, and the registry lists masking as NONE. The study is also recorded as single-group, with allocation listed as NA. Thus, the design does not include randomized assignment to atezolizumab and a concurrent comparator arm. Cohort-level safety records are present, but they do not create a like-for-like control comparison. Differences across cohorts may reflect differences in the participants or diagnoses rather than an intervention contrast.

The primary endpoint was the percentage of participants with a complete or partial response according to investigator-assessed RECIST v1.1. The stated assessment period began at baseline and continued until disease progression or death from any cause, whichever occurred first, up to approximately 42 months. The primary endpoint population is described as safety-evaluable, defined as patients who received any amount of study drug. The registry record does not report a numerical primary endpoint result or describe the analysis procedures used to produce one.

Additional outcomes included duration of response and objective response assessed with diagnosis-specific criteria, as well as overall survival. The record lists 31 outcome measures and 0 analyses. This means that the endpoint definitions and some safety counts are available, but a reported analysis of the efficacy outcomes is not included 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.

Only the first 6 of 12 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Included safety-evaluable population, defined as patients who received any amount of study drug.

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.

Only the first 6 of 13 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With an Objective Response (Complete Response [CR] or Partial Response [PR]) as Determined by the Investigator Using Response Evaluation Criteria in Solid Tumors Version 1.1 (RECIST v1.1) in Participants With Solid Tumors, measured over Baseline until disease progression, or death from any cause, whichever occurs first (up to approximately 42 months).

The primary endpoint is the percentage of participants with a solid tumor who had an objective response: a complete response or partial response, as assessed by the investigator using RECIST v1.1. It is a binary outcome at the participant level—response or no recorded response under the specified criteria—then expressed as a percentage in the defined population. The registry record defines that population as safety-evaluable, meaning patients who received any amount of study drug.

Objective response is used in oncology studies to describe measurable tumor shrinkage under prespecified response criteria. RECIST v1.1 supplies a standardized framework for classifying tumor response, which can make assessments more consistent than an informal description. The endpoint addresses whether a response was observed; by itself, it does not establish how long that response lasted, whether survival changed, or what would have happened under a different treatment. Those are separate questions, reflected in other listed outcomes such as duration of response and overall survival.

The endpoint’s stated timeframe runs from baseline until disease progression or death from any cause, whichever occurs first, up to approximately 42 months. The registry record does not report the numerical response percentage, its uncertainty, or the analysis details. It also notes that a response observed in a participant in Cohort 5 involved a rhabdoid tumor, although that participant was erroneously enrolled in the Non-rhabdomyosarcoma soft tissue sarcoma cohort. This is a relevant qualification when reading the cohort description, not a substitute for the missing overall endpoint result.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods help connect a clinical question to the data used to answer it. In this study, the primary question was expressed as a percentage: the fraction of the safety-evaluable population classified as having a complete or partial response under investigator-assessed RECIST v1.1. The registry record provides that definition, but does not report a numerical estimate or a statistical analysis for it.

What does a response percentage describe?

In general, a response percentage is a sample proportion. Each participant is classified according to whether the specified response criterion was met, and the proportion is reported as a percentage. The estimate depends on both the response classifications and the population included in the analysis. Here, the population is defined as patients who received any amount of study drug. The response count and percentage are not reported in the registry record, so the estimate cannot be evaluated from this record.

A confidence interval is a statistical range intended to convey uncertainty around an estimate. In general, an interval for a proportion reflects that the observed participants are one sample from a wider population of interest; it does not remove study-design limitations or guarantee that the underlying value lies within the interval. No confidence interval for the primary response endpoint is reported in the registry record. It is therefore not possible to judge the precision of the response estimate from the registered information.

Why does the design matter?

Randomization is a method for assigning participants to study groups in a way intended to make groups comparable on average. A concurrent control group can help separate changes associated with treatment from other explanations. This study is listed as single-group, with allocation NA and masking NONE. It therefore has no randomized control comparison. General statistical methods cannot recover a treatment contrast that the design did not create.

Because the study is open-label, investigators and participants were not masked to treatment assignment. In general, awareness of treatment can affect subjective decisions or assessments. The primary outcome uses investigator assessment, so the lack of masking is relevant context. RECIST v1.1 provides standardized response criteria, but standard criteria do not make the study blinded and do not eliminate every possible assessment influence.

How should the response criteria be understood?

RECIST v1.1 is the specified framework for the primary response endpoint in participants with solid tumors. The registry also lists other response outcomes using immune-modified RECIST v1.1 or irRC for particular participant groups, and duration-of-response outcomes assessed with criteria including mINRC, revised lymphoma response criteria, and RANO. In general, different criteria can classify response differently because they use distinct rules and are intended for different clinical contexts. These outcomes should therefore be interpreted according to their own definitions rather than pooled conceptually into one response measure.

The primary endpoint timeframe is baseline until disease progression or death from any cause, whichever occurs first, up to approximately 42 months. In general, time-to-event methods are used when the timing of an event and incomplete observation are central to the analysis. The primary endpoint is a response percentage, not a time-to-event endpoint, although the observation window is defined in relation to progression and death. The statistical rules for handling participants without adequate follow-up or assessment are not reported in the registry record.

What can the analysis record establish?

The registry lists 31 outcome measures and 0 analyses. An outcome measure describes what is to be assessed; an analysis supplies an estimate and may describe how that estimate was obtained. Since the record does not report an analysis for the primary response endpoint, readers cannot verify the calculation, assess uncertainty, or compare the estimate with a prespecified benchmark. The statistical analysis plan and protocol details are not reported in the registry record.

The sample-size material is described as a planning illustration computed from an observed effect, not the trial’s own design. Its stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. General power calculations estimate how often a specified analysis might identify an effect under assumed conditions. They are not a guarantee of a result and do not measure the probability that a treatment is effective. In this case, an equal-allocation, two-arm planning scenario is not the design recorded for the study. Any such illustration should be read as hypothetical planning information, not as observed power or evidence from a randomized comparison.

In general, statistical methods for count data may need to account for over-dispersion when variability exceeds what a simple model assumes. Over-dispersion can make standard errors too small if ignored. That is a general concept only; the registry record does not describe a count-data model or indicate that over-dispersion was assessed in this study.

Multiplicity refers, in general, to the increased chance of at least one false-positive finding when many statistical tests are considered. The record lists many outcome measures, but it does not report a multiplicity adjustment, a hierarchy of confirmatory endpoints, or the testing strategy. The existence of multiple outcomes alone does not reveal whether or how statistical error was controlled.

Finally, the adverse-event counts are descriptive. General comparisons of event frequencies require attention to the populations at risk, outcome definitions, follow-up, and whether groups are comparable. Cohort counts from an uncontrolled, open-label study do not by themselves permit a causal comparison. The registry record does not provide a randomized safety contrast or report an analysis that would support one.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists 31 outcome measures, including the primary objective-response endpoint and several secondary endpoints for response duration, overall survival, and diagnosis-specific response assessments. It lists 0 analyses. This records a broad set of planned or registered measures, but it does not show which analyses were conducted or provide their estimates.

Multiplicity is a general statistical concern when many hypotheses are tested: the chance of at least one false-positive result can increase as tests accumulate. Analysts may address this through prespecified testing sequences, adjusted significance thresholds, or other strategies. The multiplicity strategy, any endpoint hierarchy, and any adjustment are not reported in the registry record. No claim about whether such procedures were used is warranted.

The record defines the primary endpoint population as safety-evaluable patients who received any amount of study drug. Other analysis populations, such as an intention-to-treat or response-evaluable population, are not reported in the registry record. The actual number assessed for the primary response endpoint and the number with missing or unevaluable assessments are also not reported.

In general, missing outcome data can affect an estimate when participants without an assessment differ from those with an assessment. Analysts may use different approaches depending on the reasons for missingness and the endpoint. The handling of missing response assessments, withdrawals, and incomplete follow-up is not reported in the registry record. Although the endpoint timeframe is described as ending at progression or death, the detailed rules for censoring or other incomplete observations are not reported.

The statistical analysis plan, sample-size justification for the study itself, and any interim analysis procedures are not reported in the registry record. The separate planning illustration is not evidence of the trial’s original design or analysis plan.

11. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: lung infection (50.0%), hydrocephalus (33.3%), papilloedema (25.0%), upper gastrointestinal haemorrhage (25.0%), diabetic ketoacidosis (25.0%).

The safety information is reported by cohort. For each cohort, the record gives participants at risk and participants affected by serious adverse events, participants at risk and participants affected by other adverse events, and a death count. These categories should be read separately: serious adverse events and other adverse events are not the same outcome, and the death count is not identified as a count of deaths caused by treatment.

Serious adverse events were recorded for 4 of 11 participants in Cohort 1, 3 of 9 in Cohort 2, 2 of 11 in Cohort 3, 0 of 3 in Cohort 4, 3 of 10 in Cohort 5, 6 of 10 in Cohort 6, 3 of 10 in Cohort 7, 5 of 10 in Cohort 8, 2 of 4 in Cohort 9, 3 of 4 in Cohort 10, 1 of 2 in Cohort 11, and 1 of 3 in Cohort 12. These are descriptive counts within the listed cohorts, not a comparison against a control group.

The registry lists lung infection at 50.0%, hydrocephalus at 33.3%, and papillœdema, upper gastrointestinal haemorrhage, diabetic ketoacidosis, and headache each at 25.0% among the top serious terms. The denominator and calculation details for these term percentages are not reported in the registry record. The listed cohort death counts likewise do not establish cause or treatment attribution. General interpretation of adverse-event counts depends on the event definitions, observation period, and population at risk; further details are not reported in the registry record.

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

The record describes an early-phase study of atezolizumab in pediatric and young adult participants with previously treated solid tumors. Its primary efficacy measure was the percentage with an investigator-assessed complete or partial response under RECIST v1.1. That definition explains what the study set out to measure, but the numerical response result is not reported in the registry record. Consequently, the record does not establish the observed response percentage or how precisely it was estimated.

Even if an observed response estimate were available, a single-group study would describe outcomes among treated participants rather than provide a randomized treatment comparison. Without a concurrent control group, the study cannot by itself determine how outcomes would have differed under another treatment or no treatment. Open-label assessment is another design feature relevant to interpretation, particularly when the primary endpoint depends on investigator assessment.

The endpoint also has a specific scope. It concerns the safety-evaluable population, defined as patients who received any amount of study drug, and uses RECIST v1.1 during the stated assessment period. It does not automatically answer questions about response duration, overall survival, all enrolled participants, or other tumor types and response criteria. Those questions require their own outcome results and analysis details, which are not reported in the registry record.

The safety counts show descriptive event records by cohort. They should not be interpreted as proof that treatment caused an event, nor do death counts alone establish why deaths occurred. The record’s TERMINATED status is factual, but the reason for termination is not reported and cannot be inferred from the status alone.

In practical statistical terms, the registry record is informative about study structure, endpoint definitions, and some safety counts, but incomplete for evaluating the primary efficacy result. It does not report the response estimate, its uncertainty, the analysis methods, or key details about missing data and multiplicity. Those absences limit what can be concluded from the registered information without changing the meaning of the stated endpoint.

13. Why This Trial Matters Statistically

This study raises several important statistical issues for clinical research. First, an endpoint definition and a result are different things. The record specifies an objective-response endpoint and the criteria used to assess it, but does not report the numerical estimate. Naming an outcome cannot substitute for reporting its estimate, uncertainty, and analysis population.

Second, the single-group design shapes the strength of the inference. A treated cohort can provide descriptive information about observed outcomes and safety events, but it does not supply the randomized contrast needed to estimate a treatment difference. This distinction is important whenever readers encounter efficacy language in an early-phase, uncontrolled study: an observed outcome and a causal treatment effect are not interchangeable statistical quantities.

Third, the study includes diverse diagnoses and multiple response frameworks. The primary endpoint uses RECIST v1.1, while listed secondary measures use criteria tailored to other conditions. Such distinctions matter because an endpoint’s definition determines what counts as a response. They also make clear why results from different criteria should not be treated as if they were directly equivalent.

Finally, the registry lists 31 outcome measures but 0 analyses. This makes the separation between measurement planning and statistical reporting especially visible. Multiplicity procedures, missing-data handling, and analysis populations beyond the stated safety-evaluable population are not reported in the registry record. The study therefore illustrates why transparent reporting of the statistical plan and completed analyses is necessary to judge precision, interpret multiple outcomes, and understand the population represented by an estimate.

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 principal limitation for interpreting efficacy is that the numerical result for the primary endpoint is not reported in the registry record. The record gives no response percentage, responder count, confidence interval, or analysis output for that endpoint. Without those details, neither the size of the observed response nor its precision can be assessed.

The study was open-label and single-group, with allocation listed as NA and masking as NONE. There is no randomized comparator, so differences from outcomes under another approach cannot be estimated directly. Investigator assessment under a standardized response framework does not remove the design’s lack of masking or control group.

The primary endpoint population is defined as safety-evaluable patients who received any amount of study drug, but the record does not report how many participants contributed to the primary endpoint analysis or how missing and unevaluable assessments were handled. The statistical analysis plan, missing-data procedures, multiplicity strategy, and any interim analysis procedures are not reported in the registry record. The reason for the TERMINATED status is also not reported.

There are multiple outcome measures, including response and duration outcomes that use different assessment criteria for different diagnoses. No analysis results or multiplicity procedures are reported, so the relationships among the outcomes and any protection against false-positive findings cannot be assessed. The registry also notes a Cohort 5 response observed in a rhabdoid tumor for a participant erroneously enrolled in the Non-rhabdomyosarcoma soft tissue sarcoma cohort. This cohort assignment detail is a specific interpretive qualification.

Safety counts are reported by cohort, but they are not a controlled comparison, and the registry record does not provide enough detail to establish event causation or interpret the listed death counts as treatment-related. The top serious-event term percentages are reported without denominator details. These limitations concern interpretation of the registry record; they do not establish what the full study dataset or any unreported analysis might show.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: A Study to Evaluate the Safety, Tolerability, Pharmacokinetics, Immunogenicity, and Preliminary Efficacy of Atezolizumab (Anti-Programmed Death-Ligand 1 [PD-L1] Antibody) in Pediatric and Young Adult Participants With Solid Tumors — NCT02541604.
• Registry record last updated 2020-02-25; status last verified 2020-02.
• Results first posted 2020-02-25.
• 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.

18. Record Summary

The registry describes NCT02541604, an early-phase, multicenter, open-label, single-group study of atezolizumab in pediatric and young adult participants with previously treated solid tumors. It is listed as Phase 1 and Phase 2, with enrollment of 87 and sponsor Hoffmann-La Roche. The recorded start date is 2015-11-30, primary completion is 2019-06-06, and status is TERMINATED.

Atezolizumab was given by intravenous infusion on Day 1 of each 21-day cycle, with a maximum dose of 1200 milligrams. The primary endpoint was the percentage with a complete or partial response by investigator-assessed RECIST v1.1 in the safety-evaluable population. The assessment period was baseline until progression or death, whichever occurred first, up to approximately 42 months.

The registry lists 31 outcome measures and 0 analyses. A numerical result for the primary endpoint, the reason for termination, and detailed analysis procedures are not reported in the registry record. Safety counts are recorded by cohort and are descriptive, not randomized comparisons.