← Clinical Trials
Breast Cancer Phase 3 Time-to-Event NCT02513394

PALLAS: Complete Statistical Analysis of Palbociclib in Early Breast Cancer

An independent statistical review of the randomized phase 3 PALLAS trial evaluating 2-year palbociclib treatment added to at least 5 years of endocrine therapy versus at least 5 years of endocrine therapy alone in patients with histologically confirmed HR+/HER2- invasive early breast cancer.

PALLAS  ·  Phase 3  ·  Enrollment 5796.0  ·  Primary completion 2020-11
Scope of this record

This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.

1. Trial at a Glance

PALLAS is a randomized, parallel-group phase 3 treatment trial in breast cancer. The registered primary endpoint is invasive disease-free survival (iDFS) at 4 years, a time-to-event endpoint comparing palbociclib plus endocrine therapy with endocrine therapy alone.

5796.0
Enrollment
Phase 3 trial
2
Arms
Randomized parallel design
0.96
iDFS HR
95% CI 0.81–1.14
0.65
iDFS P-value
Two-sided
FeaturePALLAS
Trial namePALLAS — PALbociclib CoLlaborative Adjuvant Study
PhasePhase 3
ConditionBreast Cancer
PopulationPatients with histologically confirmed HR+/HER2- invasive early breast cancer
DesignRandomized, parallel-group
MaskingNone
Primary purposeTreatment
Enrollment5796.0
Primary endpointInvasive Disease Free Survival (iDFS) at 4 years
Primary endpoint typeTime-to-event
Registered statistical methodLog-rank test
Effect measureHazard ratio
ClinicalTrials.govNCT02513394
StatusACTIVE_NOT_RECRUITING
Start2015-08
Primary completion2020-11

2. Clinical Question

The central statistical question is whether adding 2 years of palbociclib to a planned course of at least 5 years of endocrine therapy changes invasive disease-free survival at 4 years compared with at least 5 years of endocrine therapy alone.

Population

Patients with histologically confirmed HR+/HER2- invasive early breast cancer.

Intervention

Palbociclib plus standard adjuvant endocrine therapy, with the registered endpoint describing 2-year palbociclib treatment and at least 5 years of endocrine therapy.

Comparator

Standard adjuvant endocrine therapy alone, with the registered endpoint describing at least 5 years of endocrine therapy.

Primary question

Does the randomized treatment comparison produce a different time-to-invasive-disease-event profile at 4 years?

3. Trial Design

01
Randomize5796.0 enrolled
02
Arm APalbociclib + endocrine therapy
03
Arm BEndocrine therapy alone
04
FollowTime-to-event outcome
05
AssessiDFS at 4 years
Allocation
Randomized allocation was used to create the treatment comparison.
Design model
Parallel-group design with 2 arms.
Masking
The registry describes the trial as having no masking.
Primary purpose
Treatment.
ARM A

Palbociclib Plus Endocrine Therapy

  • Palbociclib
  • Standard adjuvant endocrine therapy
  • The registered primary endpoint describes 2-year palbociclib treatment with at least 5 years of endocrine therapy.
ARM B

Endocrine Therapy Alone

  • Standard adjuvant endocrine therapy
  • The registered primary endpoint describes at least 5 years of endocrine therapy.

4. Primary Endpoint

EndpointRegistered definition / frameEndpoint type
Invasive Disease Free Survival (iDFS) Invasive disease-free survival (iDFS) for the combination of at least 5 years endocrine therapy and 2-year palbociclib treatment versus at least 5 years endocrine therapy alone in patients with histologically confirmed HR+/HER2- invasive early breast cancer (EBC) at 4 years. iDFS is defined as the time from randomization to the date of the first event: local/regional invasive ipsilateral recurrence, contralateral invasive breast cancer, distant recurrence, second primary invasive cancer of non-breast origin or death from any cause. Direct comparison between arms used time to iDFS events and Kaplan-Meier Log-rank analysis. Due to the medians not yet achieved, the percentage of patients considered iDFS at 4 years is reported. Time-to-event

The registry definition is the controlling description for this page. The ClinicalTrials.gov record ends the definition at “local/regional invasive ipsilateral recurrence, contralateral invasive breast cancer, distant recurrence, second primary invasive cancer of non-breast origin or death from any cause. Direct comparison between arms used time to iDFS events and Kaplan-Meier Log-rank analysis. Due to the medians not yet achieved, the percentage of patients considered iDFS at 4 years is reported.,” so this analysis does not extend or reconstruct the remainder of the event definition.

Why iDFS is a time-to-event endpoint: the endpoint is defined from randomization to the occurrence of a specified first event. This structure means that both the timing of events and observations that remain event-free during follow-up are relevant to the analysis.

5. Primary Results: Invasive Disease-Free Survival

The ClinicalTrials.gov results contain a formal statistical analysis for the primary endpoint. The comparison is between palbociclib plus endocrine therapy and endocrine therapy alone at 4 years.

Hazard ratio for invasive disease-free survival

0.96

95% CI: 0.81–1.14   ·   P = 0.65

Two-sided log-rank analysis; the reported analysis was stratified by (neo)adjuvant chemotherapy (yes vs no) and age (≤50 vs >50).

Primary endpointComparisonMethodHR95% CIP-value
Invasive Disease Free Survival (iDFS), 4 years Palbociclib + endocrine therapy vs endocrine therapy alone Log Rank 0.96 0.81–1.14 0.65
Clinical Biostats interpretation

The hazard ratio of 0.96 is a relative time-to-event measure comparing the estimated instantaneous event rates between the two randomized groups under the analysis framework. An HR of 0.96 corresponds to an estimated hazard approximately 4% lower in Arm A than Arm B; this is a model-based relative comparison, not a statement that 4% of patients avoided an event.

The 95% confidence interval of 0.81–1.14 describes uncertainty around the estimated hazard ratio. Because the interval spans 1, the data are compatible with both a lower and a higher estimated hazard under the model. The interval does not describe the range of effects experienced by individual patients.

The P-value of 0.65 is a measure of compatibility between the observed data and the statistical testing framework; it is not a measure of the size or clinical importance of the treatment effect. A p-value should therefore not be read as a probability that the treatment works or does not work.

The registry specifies a two-sided analysis and states that the p-value was stratified by (neo)adjuvant chemotherapy and age. Because iDFS is a time-to-event endpoint, interpretation also depends on the event and censoring rules defined for the endpoint. The ClinicalTrials.gov record does not provide a proportional-hazards diagnostic, so the hazard-ratio summary should not be interpreted as proving that proportional hazards held throughout follow-up.

Hazard-ratio framework
HR = estimated instantaneous event rate in Arm A ÷ estimated instantaneous event rate in Arm B

For this trial, the reported HR of 0.96 compares palbociclib plus endocrine therapy with endocrine therapy alone. The hazard ratio is not an absolute risk difference, a risk ratio, or a percentage of patients benefiting.

6. Secondary Time-to-Event Results

The registry contains four secondary time-to-event outcome analyses with hazard-ratio estimates and 95% confidence intervals. The ClinicalTrials.gov record does not report a formal statistical method or p-value for these secondary analyses.

Secondary endpointTime frameHR95% CIMethod in the ClinicalTrials.gov record
Invasive Disease Free Survival (iDFS) Excluding Second Primary Invasive Cancers of Non-breast Origin. 4 years 0.99 0.82–1.19 Not reported
Distant Recurrence-free Survival (DRFS) 4 years 1.05 0.87–1.28 Not reported
Overall Survival (OS) 4 years 1.32 0.98–1.78 Not reported
Locoregional Recurrences-free Survival (LRRFS) 4 years 0.84 0.57–1.23 Not reported

How to read these estimates

All four secondary endpoints are time-to-event measures, and the reported effect measure is the hazard ratio. An HR below 1 describes a lower estimated instantaneous event rate in the palbociclib-plus-endocrine-therapy group; an HR above 1 describes a higher estimated instantaneous event rate under the same comparison direction.

The confidence intervals show why the point estimate alone should not be treated as the complete result. For example, the DRFS estimate is 1.05 with a 95% CI of 0.87–1.28. The interval spans 1, so the ClinicalTrials.gov record does not establish a direction of effect with high precision. Likewise, the OS estimate of 1.32 has a 95% CI of 0.98–1.78, which also spans 1.

Registry-method limitation: the ClinicalTrials.gov record identifies the formal method for the primary iDFS analysis as Log Rank, but mark the statistical method for each of the four secondary analyses as “Not reported.” It would therefore be inappropriate to assign a specific secondary analysis method or p-value that is not contained in the ClinicalTrials.gov record.

7. Statistical Methodology

Log-rank test

The primary analysis uses a log-rank test, a standard method for comparing time-to-event distributions between randomized groups. Rather than reducing follow-up to a single event/no-event indicator, the log-rank framework uses the ordering of event times and the numbers at risk over follow-up.

Conceptual comparison
H0: survival experience in Arm A = survival experience in Arm B

The log-rank test evaluates whether the observed pattern of events over time differs between the treatment groups under its testing framework.

Hazard ratio

The primary result is reported as a hazard ratio. This provides a relative comparison of the estimated instantaneous event rates between the two treatment groups. It is especially useful for summarizing a time-to-event comparison, but it does not directly give an absolute probability of being event-free at a specified time.

Interpretation of the primary estimate
HR = 0.96  →  approximately 4% lower estimated instantaneous event rate under the fitted comparison

The derived “4%” is simply 1 − 0.96. It should not be interpreted as a 4-percentage-point difference in 4-year iDFS, nor as a 4% absolute reduction in the number of patients experiencing an event.

Stratified analysis

The primary analysis notes that its two-sided p-value was stratified by (neo)adjuvant chemotherapy (yes vs no) and age (≤50 vs >50). Stratification allows the comparison to account for these specified factors rather than treating every patient as belonging to one undifferentiated risk set.

Conceptually, stratification asks whether the treatment comparison is consistent after organizing the risk sets according to prespecified characteristics. It does not mean that separate treatment effects must be estimated for every stratum.

Kaplan-Meier estimation

Kaplan-Meier estimation is the natural descriptive companion to a time-to-event endpoint such as iDFS. It estimates the probability of remaining event-free over time while accommodating right-censored observations.

Kaplan-Meier concept
S(t) = ∏ti ≤ t (1 − di/ni)

Here, di represents events at an event time and ni represents participants at risk immediately before that time. The registry-reported PALLAS data do not provide the underlying event and censoring records needed to reconstruct a trial-specific curve.

Censoring and follow-up

Time-to-event analyses can use information from participants who have not yet experienced the endpoint, provided their follow-up remains available. Such observations are censored according to the applicable endpoint rules. Correct censoring is important because incorrectly treating a censored participant as an event, or vice versa, can change the estimated survival experience.

Analysis population and randomization

The ClinicalTrials.gov record identifies the trial allocation as randomized but do not provide a separate analysis-population definition such as an explicit intention-to-treat or as-treated rule. Accordingly, this page does not assign an analysis population beyond what is explicitly contained in the ClinicalTrials.gov record.

8. Statistical Methods Explained

Why was a log-rank test used?

The primary endpoint is a time-to-event outcome. A log-rank test is designed for comparing event-time distributions between groups while accounting for the timing of events and censoring. This is more informative for a time-to-event endpoint than simply comparing the proportion of participants with an event at one arbitrary follow-up point.

What does an HR of 0.96 mean?

An HR of 0.96 means the estimated instantaneous event rate in the palbociclib-plus-endocrine-therapy group was approximately 96% of that in the endocrine-therapy-alone group under the reported analysis. Equivalently, it is approximately a 4% lower estimated hazard. It does not mean that 4% fewer patients had an event, and it does not imply an identical effect for every participant.

Why is the confidence interval important?

The point estimate is only one summary of the treatment comparison. The 95% CI of 0.81–1.14 gives a range describing uncertainty around the estimated hazard ratio. An interval that spans 1 indicates that the data are compatible with hazard ratios on either side of the no-difference value under the specified statistical framework.

What does the P-value of 0.65 tell us?

The p-value summarizes the degree of compatibility between the observed data and the null hypothesis within the testing framework. It does not quantify the size of the treatment effect, the probability that the null hypothesis is true, or the probability that a treatment is ineffective.

Why was the analysis stratified?

The primary analysis states that its two-sided p-value was stratified by (neo)adjuvant chemotherapy and age. Stratification can improve alignment between the statistical comparison and the structure of the randomized design by accounting for specified prognostic or design factors within the comparison.

Why does time-to-event analysis need censoring?

Not every participant necessarily experiences an endpoint during the period in which they are observed. A censored participant still contributes information up to the censoring time. Kaplan-Meier and log-rank methods are built to use that partial follow-up rather than discarding the participant entirely.

9. What the Primary Result Does — and Does Not — Mean

Statistical interpretation

The primary hazard ratio of 0.96 is a relative estimate comparing the instantaneous event rates for iDFS between the two randomized groups. The estimate is close to the no-difference value of 1, and its 95% CI of 0.81–1.14 includes 1.

This does not mean that 96% of patients remained disease-free, that 4% of patients benefited, or that the treatment changed absolute 4-year iDFS by 4 percentage points. None of those quantities is represented by the hazard ratio.

Why the confidence interval matters

The 95% CI of 0.81–1.14 conveys statistical precision around the reported point estimate. Its width indicates that the point estimate should not be interpreted as an exact treatment effect. The interval also includes the null value of 1.

Why the p-value is not an effect size

The reported P = 0.65 is associated with the two-sided primary test. A p-value does not tell us how large the treatment effect is. Effect magnitude and uncertainty are better conveyed by the hazard ratio together with its confidence interval.

Relative versus absolute measures

A hazard ratio and a time-specific percentage answer different statistical questions. The registry-reported PALLAS data provide a hazard ratio for the primary endpoint but do not provide the underlying 4-year iDFS percentages needed for an absolute-risk presentation.

10. Safety Results

The ClinicalTrials.gov record provides serious adverse-event counts by randomized arm as affected participants divided by participants at risk. No additional safety categories are contained in the ClinicalTrials.gov record.

Safety measureArmAffected / at risk
Serious adverse events Palbociclib Plus Endocrine Therapy (Arm A) 369/2841
Serious adverse events Endocrine Therapy Alone (Arm B) 229/2902

These figures should be interpreted as a safety summary rather than as a time-to-event efficacy result. The ClinicalTrials.gov record does not provide a statistical comparison, confidence interval, p-value, severity breakdown, or exposure-adjusted analysis for serious adverse events.

Safety interpretation: the affected/at-risk figures should not be converted into a treatment effect estimate without an explicit statistical framework. In particular, they do not by themselves establish a causal difference in serious-adverse-event risk because the ClinicalTrials.gov record does not provide the additional definitions, exposure information, follow-up structure, or analysis method needed for such a comparison.

11. Multiplicity and Secondary Endpoints

The registry identifies one registered primary endpoint and four additional posted outcome analyses. The primary endpoint is iDFS at 4 years. The four secondary outcomes are additional time-to-event measures at 4 years.

EndpointRoleStatistical result reported
Invasive Disease Free Survival (iDFS)PrimaryHR 0.96; 95% CI 0.81–1.14; P = 0.65
iDFS excluding second primary invasive cancers of non-breast originSecondaryHR 0.99; 95% CI 0.82–1.19
Distant Recurrence-free Survival (DRFS)SecondaryHR 1.05; 95% CI 0.87–1.28
Overall Survival (OS)SecondaryHR 1.32; 95% CI 0.98–1.78
Locoregional Recurrences-free Survival (LRRFS)SecondaryHR 0.84; 95% CI 0.57–1.23

The ClinicalTrials.gov record does not state a multiplicity-adjustment procedure, hierarchical testing strategy, alpha allocation, or multiplicity-adjusted p-values for the secondary endpoints. Those features therefore should not be inferred from the presence of multiple posted outcomes.

12. Stratification and the Meaning of the Primary Test

The primary analysis specifically states that the two-sided p-value was stratified by (neo)adjuvant chemotherapy (yes vs no) and age (≤50 vs >50). This is an important detail because the reported p-value is not described as coming from an unstratified comparison.

Stratification factor

(Neo)adjuvant chemotherapy: yes vs no.

Age factor

≤50 vs >50.

Test direction

The registry reports a two-sided p-value.

Reported test

Log Rank for the primary iDFS analysis.

Stratification does not change the randomized treatment assignment. Instead, it changes how the treatment comparison is evaluated across the specified strata. This can be particularly useful when a prognostic factor affects event timing.

13. Limitations and Interpretation Issues

Data discipline: these limitations are deliberate. This page uses the ClinicalTrials.gov record rather than importing additional numerical results from the linked publications or other external sources.

14. Why This Trial Matters Statistically

PALLAS is useful for learning how a large randomized trial can be reduced to a precise time-to-event question. The primary endpoint is not simply whether an event occurred; it is invasive disease-free survival measured from randomization. That structure naturally leads to survival-analysis methods such as the log-rank test and hazard ratios.

The primary result also illustrates why statistical interpretation should not stop at a p-value. The reported HR of 0.96, 95% CI of 0.81–1.14, and P = 0.65 tell different parts of the statistical story. The HR describes the estimated relative event rate, the confidence interval describes uncertainty around that estimate, and the p-value describes evidence within the specified testing framework.

The stratified nature of the primary analysis provides another important lesson. The reported p-value was stratified by (neo)adjuvant chemotherapy and age, showing that the treatment comparison was not evaluated as a completely unstructured pooled comparison.

Finally, the four secondary endpoints demonstrate why a trial's statistical hierarchy matters. Multiple time-to-event outcomes can provide different perspectives on disease recurrence and survival, but their estimates should not automatically be treated as independent confirmatory conclusions when the ClinicalTrials.gov record does not specify the multiplicity framework.

15. A Practical Statistical Reading of PALLAS

QuestionWhat the ClinicalTrials.gov record showsHow to interpret it
What was the primary endpoint? iDFS at 4 years A time-to-event endpoint comparing the randomized treatment groups.
What was the primary method? Log Rank A survival-analysis method for comparing event-time distributions.
What was the effect measure? HR 0.96 A relative measure of estimated instantaneous event rates.
How precise was the estimate? 95% CI 0.81–1.14 The interval includes the no-difference value of 1.
What was the reported p-value? 0.65 A two-sided p-value from the stratified primary analysis.
Which factors were used for stratification? (Neo)adjuvant chemotherapy and age The p-value was stratified by these specified categories.

16. Longitudinal Trial Context

Trial start

2015-08

Primary completion

2020-11

Current registry status

ACTIVE_NOT_RECRUITING

Lead sponsor

Alliance Foundation Trials, LLC.

The registry identifies PALLAS as a phase 3 randomized treatment study with 5796.0 participants enrolled. The ClinicalTrials.gov record establishes the trial's start and primary-completion dates, but do not provide a detailed sequence of interim analyses, protocol amendments, data cutoffs, or later statistical updates.

17. Sources

The statistical content on this page is based on the ClinicalTrials.gov record. The following links are the registry and PubMed records identified in that source material.

Source hierarchy: ClinicalTrials.gov is the official registry record for the trial. The numerical trial results presented here are restricted to the ClinicalTrials.gov record; the linked PubMed records are provided as source references identified in that data.