A Study of Zanubrutinib (BGB-3111) Versus Ibrutinib in Participants With Relapsed/Refractory Chronic Lymphocytic Leukemia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Zanubrutinib (BGB-3111) Versus Ibrutinib in Participants With Relapsed/Refractory Chronic Lymphocytic Leukemia (NCT03734016), a randomized phase 3 study. The primary outcome is Overall Response Rate (ORR) Assessed by the Investigator.

2. Clinical Question

3. Trial Design

This was a randomized, open-label, parallel-group Phase 3 study sponsored by BeiGene. The registry record describes a comparison of zanubrutinib with ibrutinib among participants with relapsed or refractory CLL or SLL. Masking was listed as none, so treatment assignment was not masked according to the record. Enrollment was 652.

The zanubrutinib group received 160 mg orally twice daily, and the ibrutinib group received 420 mg orally once daily. Treatment could continue until disease progression, intolerable toxicity, initiation of alternative anticancer therapy, an investigator or sponsor decision, need for prohibited medication, study withdrawal, or pregnancy. The registry record lists the trial as completed. Its start date was 2018-11-01, and primary completion was 2022-08-08.

The primary outcome was investigator-assessed overall response in the ITT Analysis Set. The reported timeframe ran from randomization to the final efficacy analysis cutoff date of 08 August 2022. Median follow-up was 29.6 months, with a maximum of 45.2 months. The record also lists progression-free survival assessed by investigators and an independent review committee, atrial fibrillation or atrial flutter, response duration, treatment failure, and response-related outcomes among the secondary measures. The registry record does not report the original design assumptions or the full statistical analysis plan in the information available for this summary.

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.

Before assignment: Participants were randomly assigned to one of two treatment groups. Randomization was stratified according to age (\< 65 versus ≥ 65 years), geographic region (China vs non-China), refractory status (yes or no), and chromosome 17p deletion or TP53 mutation status (present or absent).

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent To Treat (ITT) Analysis Set

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: Intent-to-Treat (ITT) Analysis Set includes all randomized participants

6. Primary Endpoint

The registry lists the primary outcome as Overall Response Rate (ORR) Assessed by the Investigator, measured over From randomization to the final efficacy analysis cutoff date of 08 August 2022, median time on follow-up was 29.6 months (maximum of 45.2 months)..

Overall response rate (ORR) is the percentage of participants meeting a specified response category. For this study, the investigator’s assessment counted complete response (CR), complete response with incomplete bone marrow recovery (CRi), nodular partial response (nPR), or partial response (PR). Thus, the endpoint combines several response categories into one binary outcome: each participant either met the response definition or did not.

The registry record says response was assessed using the 2008 criteria of the International Workshop on CLL, modified for treatment-related lymphocytosis in participants with CLL, and the Lugano classification for participants with SLL. These criteria provide a structured basis for assigning response status. An overall response endpoint can make a treatment comparison easier to interpret than separate estimates for every response category, while still depending on how and when response is assessed.

Investigator assessment is the source specified for the primary endpoint. It is not the same as an independent review committee assessment: the record lists independent review separately for some secondary outcomes. ORR describes response status over the stated analysis period; by itself, it does not describe how long a response lasted or how long participants remained free of progression. Those questions are addressed by other types of outcomes.

7. Results Posted in the Registry

• Ibrutinib: 74.2% (241/325) versus 83.5% (273/327) on control; risk difference -9.3 percentage points (95% CI -15.6 to -3.1), risk ratio 0.89 (0.82–0.96); unadjusted two-proportion z-test p = 0.004.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Progression-free Survival (PFS) Assessed by the Investigator

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Stratified Wald test); it was not recalculated.
• Result:
  • Zanubrutinib vs Ibrutinib: hazard ratio 0.65 (95% CI 0.49–0.86), reported p < 0.001 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Zanubrutinib vs Ibrutinib than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Progression-free Survival Assessed by the Independent Review Committee

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Stratified Wald test); it was not recalculated.
• Result:
  • Zanubrutinib vs Ibrutinib: hazard ratio 0.65 (95% CI 0.49–0.86), reported p < 0.001 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Zanubrutinib vs Ibrutinib than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants With Atrial Fibrillation or Atrial Flutter

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Ibrutinib: 13.3% (43/324) versus 5.2% (17/324) on control; risk difference 8.0 percentage points (95% CI 3.6 to 12.4), risk ratio 2.53 (1.47–4.34); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Ibrutinib than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The secondary outcomes extend the primary response question to disease control over time, response durability, treatment failure, and selected adverse outcomes. The broad pattern in the posted results is compatible with the primary ORR comparison: the registry reports favorable PFS comparisons for zanubrutinib using both investigator assessment and independent review, and reports a lower recorded percentage with atrial fibrillation or atrial flutter for zanubrutinib. These outcomes address different questions and should not be treated as repetitions of ORR.

The registry lists 13 posted secondary outcomes, while the outcome record describes 15 outcome measures and 25 analyses. The named secondary measures include progression-free survival, response duration, time to treatment failure, atrial fibrillation or atrial flutter, and response categories. The available result details vary by outcome. For the zanubrutinib PFS median, a median value is not reported in the registry record, although a lower confidence limit is listed. This kind of incomplete reporting makes a simple across-outcome ranking inappropriate.

Secondary outcomes can help assess whether findings extend beyond the primary endpoint, but they generally carry less confirmatory weight when many outcomes or analyses are examined. Testing many questions increases the chance that some results appear statistically notable by chance unless the analysis plan accounts for multiplicity. The registry record does not report a multiplicity adjustment strategy for the secondary analyses. Their p-values should therefore be read as evidence for their particular comparisons, not as if each were the sole prespecified test.

Differences in assessors and outcome definitions also matter. Investigator assessment and independent review may not classify events identically, while a time-to-event measure and a binary response measure have different interpretations. The secondary results provide context for the primary finding, but do not establish a single overall effect across all outcomes or resolve the unreported details of the analysis plan.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Progression-free Survival (PFS) Assessed by the Investigator
• Progression-free Survival Assessed by the Independent Review Committee
• Percentage of Participants With Atrial Fibrillation or Atrial Flutter
• Duration of Response Assessed by the Independent Review Committee
• Duration of Response (DOR) Assessed by the Investigator
• Time to Treatment Failure
• Rate of Partial Response With Lymphocytosis (PR-L) or Higher Assessed by the Independent Review Committee
• Rate of Partial Response With Lymphocytosis (PR-L) or Higher Assessed by the Investigator

9. Statistical Interpretation

10. Statistical Methods Explained

Several statistical ideas help distinguish the posted arm results from the formal comparisons. The primary measure is binary: a participant either met the response definition or did not. The arm-level percentage estimates the proportion meeting that definition. A comparison then expresses how those proportions relate, while a confidence interval describes uncertainty around the estimate.

What does the response ratio represent?

The registry reports a response ratio of 1.12 for zanubrutinib relative to ibrutinib. A ratio compares the response proportions by dividing one by the other; a value of 1 would represent equal proportions. The 95% confidence interval, 1.04 to 1.22, gives a range of values compatible with the reported analysis under its statistical assumptions. The registry identifies the method as a stratified Wald test. Stratification means the analysis accounts for specified groups or factors when forming the comparison. The stratification factors themselves are not reported in the registry record, so their influence cannot be assessed from these results alone.

The record labels this response-ratio analysis as non-inferiority. A non-inferiority assessment asks whether the result rules out an unacceptable loss relative to a comparator, according to a prespecified margin. The margin and its justification are not reported in the registry record. Therefore, the label and reported analysis can be described, but the margin-based decision process cannot be independently reconstructed from the information available.

Why is there also a superiority test?

The registry separately reports a Cochran-Mantel-Haenszel superiority test with p = 0.004. This test assesses evidence of a difference between groups while accounting for stratification. A p-value is a measure of how incompatible the observed data are with a specified no-difference hypothesis under the model and test assumptions. It is not the probability that the no-difference hypothesis is true, and it does not measure the size or clinical importance of an effect. The registry does not report the full test specification or all analysis details in this record summary.

The unadjusted re-calculation uses the posted arm-level response summaries to estimate a risk difference and an unadjusted p-value. Risk difference expresses the contrast between proportions on an absolute scale. Its sign depends on which group is subtracted from which; the reported re-calculation is oriented as ibrutinib relative to zanubrutinib. This simple comparison does not incorporate the stratified analysis method used in the sponsor’s reported tests. It is a complementary check, not a replacement for the reported trial analysis.

What does a confidence interval add?

A point estimate gives one best estimate under the chosen method, while a confidence interval communicates its sampling uncertainty. A narrower interval indicates greater precision under that method; a wider interval indicates less precision. The interval does not account automatically for every possible source of bias, such as differences in assessment, missing data, or deviations from the analysis plan. Nor does it identify which individual participants experienced a response.

For time-to-event outcomes, the analysis differs from a binary response comparison. Progression-free survival accounts for the timing of progression or other defined events and for follow-up that may end before an event is observed. The registry reports hazard ratios and log-rank tests for the PFS comparisons. A hazard ratio summarizes a relative event-rate contrast over follow-up under the model’s assumptions; it is not the ratio of the reported medians. The registry does not provide enough information to evaluate all assumptions or censoring procedures.

What information is missing for a fuller audit?

The registry record does not report the original sample-size justification, the non-inferiority margin, detailed stratification factors, interim-analysis procedures, or the complete statistical analysis plan in the information. It also does not give sufficient detail to reconstruct all handling of missing outcomes or analysis-population exclusions. These omissions limit independent evaluation of whether each reported test followed a prespecified procedure. They do not change the numerical results reported, but they constrain how completely the analysis can be reproduced or assessed.

11. Confidence Intervals and Precision

The primary arm estimates are accompanied by 95% confidence intervals: 79.0% to 87.3% for zanubrutinib and 69.0% to 78.8% for ibrutinib. These intervals show that the observed percentages are estimates rather than exact population values. Their widths reflect sampling uncertainty under the methods used, but they do not capture every possible uncertainty, including measurement differences or bias from missing outcomes.

For the sponsor-reported response ratio, the 95% confidence interval is 1.04 to 1.22 around the estimate of 1.12. Because the interval is above 1, it is consistent with a higher response proportion in the zanubrutinib group under that ratio’s direction. The reported stratified superiority test also gives p = 0.004. The unadjusted re-calculation has a risk-difference interval from -0.156 to -0.031, expressed in the direction of ibrutinib relative to zanubrutinib. These intervals use different comparison scales and methods, so their numerical endpoints should not be compared as if they described the same quantity.

A confidence interval is not a guarantee that the true effect lies inside its particular bounds. In repeated studies using the same procedure, intervals would vary; the confidence level describes the long-run behavior of that procedure. It also does not mean that 95% of participants had outcomes inside the interval. The arm intervals alone are not a substitute for an interval around the treatment contrast, and apparent overlap or non-overlap of arm intervals is not the formal comparison reported by the trial.

Precision should be considered alongside design and analysis details. The registry identifies randomization and the ITT Analysis Set for the primary endpoint, but the record does not report all details needed to assess missing response assessments, stratification, or the original analysis plan. Statistical precision therefore describes uncertainty under the stated analysis, not certainty about every feature of the evidence.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size estimates are planning illustrations based on the observed effect, not the trial’s original design assumptions. For binary outcomes such as ORR, the figures are participants per arm. The planning assumptions listed in the registry-derived analysis are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect.

The 80% power column gives the illustrative number of participants per arm for a design with an 80% chance of detecting the specified effect under those assumptions, if that effect and the model assumptions were correct. The 90% power column gives the corresponding planning number for a 90% chance. These are prospective design calculations, not measures of how likely the completed study was to produce a significant result after observing its data.

The rows consider planning for 100%, 75%, and 50% of the observed effect. Smaller effects need more participants to detect under otherwise unchanged assumptions, which is why the illustrative requirements increase as the target effect becomes smaller. The reported calculations are not observed power; they are not a retrospective assessment of the trial’s success or a substitute for interpreting the estimate and its confidence interval.

For time-to-event outcomes, the planning figures are total events rather than participants per arm. They use the stated Schoenfeld approach. Event-based calculations are distinct from the binary-outcome illustrations because follow-up duration and event occurrence contribute to the information available for a time-to-event comparison. The registry record does not report the original sample-size justification or establish that these illustrations reproduce the sponsor’s original planning process.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.835; treatment proportion p₁ = 0.742; average p̄ = 0.788.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5778 = 1.1325.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5740 = 0.4831.
4. n = (1.1325 + 0.4831)² / (-0.0933)² = 299.7, rounded up to 300 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.834862385321101                          # control proportion
p1 <- 0.741538461538462                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(3734016)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 300 (80% power), 401 (90% power)
 75% of observed effect: n per arm = 512 (80% power), 685 (90% power)
 50% of observed effect: n per arm = 1,102 (80% power), 1,475 (90% power)
power.prop.test: 299.7 
Simulated power at n = 300 per arm: 0.798 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary analysis as a stratified Wald test of the response ratio, with estimate 1.12, 95% confidence interval 1.04 to 1.22, and p < 0.001 under the non-inferiority label. It also reports a Cochran-Mantel-Haenszel superiority test with p = 0.004. These are the trial’s posted analyses and should be distinguished from the unadjusted re-calculation based on the arm summaries.

The re-calculation gives a risk difference of -0.093, with a 95% confidence interval from -0.156 to -0.031 and p = 0.004, in the direction of ibrutinib relative to zanubrutinib. It uses a different effect scale and does not reproduce the stratified methods. The re-calculated p-value is close to the reported superiority p-value, but that numerical proximity does not make the procedures interchangeable.

The registry record does not flag a discrepancy between the posted results and the simple re-calculation. The reported sponsor analysis therefore remains the primary basis for describing the trial’s inferential results; the unadjusted estimate is best read as a transparent, simpler comparison of the posted arm summaries. The non-inferiority margin and full analysis plan are not reported in the registry record, limiting independent reconstruction of the sponsor’s decision framework.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study tests several outcomes, comparisons, or analysis types. Even if each test uses a conventional significance threshold, examining many tests can increase the chance that at least one result appears notable by chance. This is why prespecified outcome hierarchies, family-wise error control, or other strategies may matter for confirmatory interpretation.

The registry lists a primary endpoint and multiple secondary outcomes and analyses. It reports a non-inferiority analysis and a superiority analysis for the primary response question, and it lists further analyses for secondary outcomes. The multiplicity strategy, testing hierarchy, and any adjustment to significance thresholds are not reported in the registry record. It is therefore not possible from this record alone to determine how the different tests were coordinated or which secondary findings were intended to be confirmatory.

The primary endpoint is identified as using the ITT Analysis Set. The record does not provide detailed counts or rules for missing primary assessments, exclusions from analysis, or sensitivity analyses addressing missing outcomes. It also does not describe how missing data were classified for the secondary endpoints in the information. These details can affect the interpretation of an endpoint that depends on participant assessments.

Randomization supports a structured comparison of assigned groups, but the analysis population and handling of post-randomization events are relevant to how that comparison is interpreted. The listed treatment stopping conditions include several events that may affect exposure or follow-up. The registry record does not report all analysis rules for those circumstances. Nor does it provide the original statistical plan, interim-analysis details, or the non-inferiority margin. These are not grounds to assume a particular procedure; they are limits on what can be verified from the registry record.

Consequently, the reported estimates and tests can be described as posted, while claims about multiplicity control, missing-data robustness, or adherence to a detailed prespecified plan cannot be independently established from this record.

16. 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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: covid-19 pneumonia (10.2%), pneumonia (10.2%), covid-19 (8.6%), atrial fibrillation (3.4%), urinary tract infection (3.1%).

The registry reports serious adverse events among participants at risk in each group: 172 of 324 for zanubrutinib and 196 of 324 for ibrutinib. It also reports other adverse events among 324 at risk in each group: 310 for zanubrutinib and 314 for ibrutinib. These are posted counts, not calculated percentages. The serious-event counts and other-event counts refer to different categories and should not be combined.

The record lists deaths of 69 in the zanubrutinib group and 83 in the ibrutinib group. It does not establish in these summary figures whether deaths were caused by treatment, nor do the counts alone describe timing, severity patterns, or individual circumstances. The most frequently listed serious adverse event terms include COVID-19 pneumonia, pneumonia, COVID-19, atrial fibrillation, urinary tract infection, and anaemia, with percentages reported for each term in the registry.

These safety summaries are descriptive. They do not by themselves establish a causal difference between groups, and interpretation depends on event definitions, duration of exposure, follow-up, and completeness of reporting. Further details on those matters are not reported in the registry record.

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

The posted primary results show a higher investigator-assessed response percentage in the zanubrutinib group than in the ibrutinib group. The sponsor’s reported stratified analyses include a response-ratio estimate with a confidence interval and a superiority test, alongside an analysis labeled non-inferiority. The unadjusted re-calculation points in the same direction, while using a simpler method and a different effect scale. Together, these results describe the observed comparison in this randomized trial; they do not guarantee the same magnitude in other populations or settings.

ORR is a measure of whether a participant met one of the specified response categories during the assessment period. It does not show how long a response lasted, whether disease later progressed, or how each participant experienced treatment. The registry lists PFS and response duration as separate outcomes because these address different aspects of disease course. Likewise, safety counts require their own interpretation and cannot be inferred from efficacy results.

A p-value does not express the probability that one treatment is better, and a confidence interval does not cover all sources of uncertainty. The registry record does not report the non-inferiority margin, full statistical plan, multiplicity strategy, or detailed missing-data procedures in the information. Those omissions limit independent assessment of some design and analysis choices.

These findings are a statistical description of a specific randomized comparison among participants with relapsed or refractory CLL or SLL. They are not individualized medical guidance, and the registry results alone do not determine which treatment is appropriate for any particular person. Interpretation in practice would require broader clinical context beyond the statistical information recorded for this trial.

18. Why This Trial Matters Statistically

This trial illustrates why a clinical result should be read through several statistical lenses rather than through a single p-value. The primary endpoint is a binary response measure, but the reported analysis includes a stratified response ratio and separate tests labeled for non-inferiority and superiority. The unadjusted risk difference offers an intuitive absolute-scale contrast, yet it answers the comparison using a simpler method. Naming the method and effect scale prevents unlike estimates from being mistaken for direct duplicates.

The trial also shows how complementary outcomes can broaden interpretation. ORR records whether a response criterion was met; PFS concerns time until a defined event; atrial fibrillation or atrial flutter is a safety outcome. Agreement in the broad direction across some outcomes can add context, but each endpoint has its own definition, assessment process, and uncertainty. Secondary outcomes also raise multiplicity concerns when the adjustment strategy is not reported.

Precision matters alongside statistical significance. The response estimates have confidence intervals, and the reported ratio has its own interval. These show uncertainty around group-level estimates, not certainty for individual outcomes. Missing information about the non-inferiority margin, analysis plan, and missing-data procedures limits how completely the inferential framework can be audited from the registry alone.

Finally, the sample-size illustrations demonstrate the distinction between planning and observation. They use the observed effect as a planning input, whereas the original design assumptions are not reported. They should not be confused with observed power or treated as evidence that the trial was designed around those calculations. Taken together, these points make the record useful for understanding what was measured and reported while keeping the limits of statistical interpretation visible.

19. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

Several trial-specific limits affect interpretation of the registry results. The study was open-label, with masking listed as none. Investigator assessment was used for the primary ORR endpoint, so the primary measure was not an independent review committee assessment. The registry describes disease-specific response criteria, but the summarized record does not provide detailed information on assessment schedules or adjudication procedures.

The record reports a stratified non-inferiority analysis but does not report the non-inferiority margin or its justification. It also does not provide the complete statistical analysis plan, detailed stratification factors, or interim-analysis procedures. This restricts independent evaluation of how the reported tests relate to the planned decision rules. The simple unadjusted re-calculation is not a substitute for those details.

For missing data, the primary population is identified as ITT, but the record does not describe the handling of missing response assessments, exclusions, or sensitivity analyses in the summarized information. The multiplicity strategy for the several outcomes and analyses is also not reported. These gaps matter particularly when interpreting secondary p-values as confirmatory evidence.

Some secondary outcome details are incomplete in the record; for example, a median value for zanubrutinib on the investigator-assessed PFS outcome is not reported, although a lower confidence limit is listed. Safety counts are presented descriptively, without sufficient detail to determine causality, exposure-adjusted event rates, or the circumstances of deaths. These limitations do not erase the posted results, but they constrain their scope and independent reproducibility.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Study of Zanubrutinib (BGB-3111) Versus Ibrutinib in Participants With Relapsed/Refractory Chronic Lymphocytic Leukemia — NCT03734016.
• Registry record last updated 2025-03-30; status last verified 2025-03.
• Results first posted 2023-07-27.
• 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.

23. Record Summary

The registry record describes a completed, randomized, open-label, parallel-group Phase 3 comparison of zanubrutinib with ibrutinib in relapsed or refractory CLL or SLL. Enrollment was 652. The primary endpoint was investigator-assessed ORR in the ITT Analysis Set, assessed through the final efficacy cutoff date of 08 August 2022.

The posted response percentages were 83.5% for zanubrutinib and 74.2% for ibrutinib. The registry reports stratified non-inferiority and superiority analyses, and an unadjusted re-calculation provides a complementary risk-difference comparison. Secondary outcomes include PFS, response duration, treatment failure, and selected safety measures.

The record does not report several details needed for a complete statistical audit, including the original sample-size justification, non-inferiority margin, multiplicity strategy, and detailed missing-data procedures. Estimates should therefore be read with their methods and limitations in view.