This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical results on this page are restricted to the ClinicalTrials.gov record.
1. Trial at a Glance
GOYA was a randomized, open-label, parallel-group phase 3 trial evaluating obinutuzumab in combination with chemotherapy versus rituximab in combination with chemotherapy in participants with CD20-positive diffuse large B-cell lymphoma. The registry reports 1418 participants enrolled and a primary time-to-event endpoint of investigator-assessed progression-free survival.
| Feature | GOYA |
|---|---|
| Trial name | GOYA |
| Phase | Phase 3 |
| Therapeutic area | Hematology |
| Condition | Diffuse Large B-Cell Lymphoma |
| Population described in title | Participants with CD20-positive diffuse large B-cell lymphoma |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 1418 |
| Primary endpoint type | Time-to-event |
| Hypothesis type | Superiority |
| Primary analysis method | Log-rank test |
| Primary effect measure | Hazard ratio |
| ClinicalTrials.gov | NCT01287741 |
| Status | Terminated |
| Start | 2011-07-26 |
| Primary completion | 2016-04-29 |
| Lead sponsor | Hoffmann-La Roche |
| Sponsor type | Industry |
2. Clinical Question
The trial asks whether treatment with obinutuzumab plus chemotherapy differs from rituximab plus chemotherapy with respect to investigator-assessed progression-free survival in participants with CD20-positive diffuse large B-cell lymphoma.
Population
Participants with CD20-positive diffuse large B-cell lymphoma, as specified in the trial's brief title.
Intervention
Obinutuzumab in combination with chemotherapy. The registry also identifies cyclophosphamide, doxorubicin, vincristine, and prednisone among the trial interventions.
Comparator
Rituximab in combination with chemotherapy. The registry also identifies cyclophosphamide, doxorubicin, vincristine, and prednisone among the trial interventions.
Primary question
Does the randomized comparison provide evidence of a difference in investigator-assessed progression-free survival under the prespecified superiority framework?
3. Trial Design
Rituximab + Chemotherapy
- Rituximab
- Cyclophosphamide
- Doxorubicin
- Vincristine
- Prednisone
Obinutuzumab + Chemotherapy
- Obinutuzumab
- Cyclophosphamide
- Doxorubicin
- Vincristine
- Prednisone
The registry classifies GOYA as a randomized, parallel-group, unmasked phase 3 treatment study. The statistical structure is therefore fundamentally a comparison of time-to-event outcomes between randomized groups rather than a comparison of a continuous measurement at a single time point.
4. Endpoints
| Endpoint | Time frame | Type | Registry definition / analysis |
|---|---|---|---|
| Median Time to Progression-Free Survival (PFS), Investigator-Assessed | Baseline up to approximately 6.5 years (up to 31 January 2018) | Time-to-event | Kaplan-Meier estimate of the median PFS was defined as the time at which half of the participants have progressed (progressive disease [PD]). Progression-free survival was defined as the time from randomization until the first documented day of disease progression or relapse, using a modified version of the Revised Response Criteria for Malignant Lymphoma, or death from any cause. |
| Median Time to Progression-Free Survival (PFS), Independent Review Committee (IRC)-Assessed | Baseline up to approximately 4 years and 9 months (up to 29 April 2016) | Time-to-event | Secondary PFS endpoint assessed by an Independent Review Committee. The posted statistical analysis used a log-rank test with hazard ratio as the effect measure. |
The ClinicalTrials.gov record identifies one registered primary endpoint, investigator-assessed PFS, and classify it as a time-to-event endpoint. The registry also contains an IRC-assessed PFS analysis as a secondary endpoint.
5. Analysis Populations and Stratification
The primary posted analysis used an intention-to-treat (ITT) population. The registry describes the ITT population as including all randomized participants, while the analysis notes a serious Good Clinical Practice non-compliance issue at a single study site in China involving all 4 participants enrolled there.
| Analysis feature | Registry-supported description |
|---|---|
| Primary analysis population | Intent-to-treat population including all randomized participants, with the posted analysis noting the serious Good Clinical Practice non-compliance issue at a single study site in China involving all 4 participants enrolled there. |
| Secondary PFS analysis population | Intent-to-treat population including all randomized participants. |
| Primary stratification | International Prognostic Index (IPI) score: low/low-intermediate, excluding participants having an IPI score 0 without bulky disease. |
| Secondary stratification | The IRC-assessed PFS analysis used the same stated IPI-score stratification. |
Using the ITT principle keeps the primary treatment comparison anchored to randomized assignment. This is important because the causal contrast created by randomization can be weakened if participants are selectively moved between treatment groups during analysis.
The registry-reported analysis text also shows that the log-rank comparison was stratified by IPI score. Stratification allows the time-to-event comparison to account for the prespecified prognostic grouping rather than treating all participants as if they had identical baseline prognostic structure.
6. Statistical Methodology
Kaplan-Meier estimation
The registered primary endpoint is a time-to-event outcome, and the endpoint definition explicitly describes the Kaplan-Meier estimate of median PFS. Kaplan-Meier estimation is designed for settings in which some participants may not experience the event before the end of their observed follow-up.
Here, di represents events at time ti, while ni is the number at risk immediately before that time. The resulting curve estimates the probability of remaining event-free through time t.
The registry's primary endpoint is specifically framed around the median PFS, which is the time at which the Kaplan-Meier estimate of PFS reaches 0.50. The registry-reported statistical analysis, however, reports a hazard ratio rather than a median PFS estimate. Accordingly, this page does not substitute an unreported median value.
Stratified log-rank test
The posted primary and secondary analyses both use a log-rank test. The analysis notes specify stratification by IPI score. A log-rank procedure compares the observed and expected pattern of events between treatment groups over follow-up, with the stratified version accounting for the specified prognostic strata.
The log-rank framework evaluates whether the event experience differs systematically between randomized groups across the observed follow-up period.
Hazard ratio
The registry reports the treatment effect as a hazard ratio (HR). A hazard ratio compares the estimated instantaneous event rates between the groups under the time-to-event analysis.
HR < 1 → lower estimated hazard in the numerator group
HR > 1 → higher estimated hazard in the numerator group
For GOYA, the posted comparison is expressed as rituximab + chemotherapy versus obinutuzumab + chemotherapy. The direction of interpretation therefore depends on that stated group ordering.
Intention-to-treat analysis
The ITT population includes all randomized participants according to the registry's analysis description. ITT analysis is particularly important in randomized trials because it preserves the treatment assignment generated by the randomization process rather than redefining groups according to treatment actually received or later events.
Stratified analysis
Both posted PFS analyses were stratified by the specified IPI-score category. Stratification is useful when a baseline prognostic variable is expected to influence event timing and was incorporated into the trial's analytical framework.
7. Primary Result: Investigator-Assessed PFS
The primary endpoint was investigator-assessed progression-free survival from baseline through approximately 6.5 years, with the registry specifying follow-up through 31 January 2018. The posted formal comparison used a stratified log-rank test and reported the treatment effect as a hazard ratio.
Hazard ratio for investigator-assessed PFS
95% CI: 0.78–1.12 · P = 0.4753
Comparison: Rituximab + chemotherapy vs Obinutuzumab + chemotherapy
| Primary endpoint | Rituximab + chemotherapy | Obinutuzumab + chemotherapy | Effect estimate |
|---|---|---|---|
| Median Time to Progression-Free Survival (PFS), Investigator-Assessed | Randomized comparison | Randomized comparison | HR 0.94 (95% CI 0.78–1.12); P = 0.4753 |
The reported HR of 0.94 means that, under the posted time-to-event comparison and stated group ordering, the estimated hazard of progression or death in the rituximab-plus-chemotherapy group was approximately 94% of that in the obinutuzumab-plus-chemotherapy group. Expressed descriptively, this corresponds to an estimated hazard that is about 6% lower for the rituximab group relative to the obinutuzumab group.
The HR does not mean that 6% fewer participants progressed, that individual patients experienced a 6% reduction in risk, or that the probability of remaining progression-free was exactly 6% different. A hazard ratio is a relative time-to-event measure, not an absolute risk difference.
The 95% CI of 0.78–1.12 describes the statistical uncertainty around the estimated hazard ratio. Because the interval includes 1, values corresponding to a lower hazard for rituximab as well as values corresponding to a higher hazard for rituximab remain compatible with the interval.
The P-value of 0.4753 is evidence about the statistical compatibility of the observed comparison with the null hypothesis under the specified testing framework. It is not a measure of the magnitude or clinical importance of the treatment effect. A P-value cannot by itself tell us how large a treatment difference is.
The analysis is based on time-to-event data and therefore depends on how progression, death, and censoring are handled. In addition, a single hazard ratio summarizes the relative event experience through the analyzed follow-up and should not be interpreted as an absolute survival probability.
What is not reported in the registry-reported primary analysis
The registered endpoint is explicitly defined in terms of the median PFS, but the ClinicalTrials.gov record reports the hazard ratio, 95% confidence interval, and P-value rather than a numerical median PFS for either treatment group. ClinicalTrials.gov therefore does not provide a median PFS value in the ClinicalTrials.gov record, and no median is added here.
8. Secondary Result: IRC-Assessed PFS
The registry also reports an Independent Review Committee (IRC)-assessed PFS analysis using a time frame from baseline through approximately 4 years and 9 months, up to 29 April 2016.
Hazard ratio for IRC-assessed PFS
95% CI: 0.72–1.10 · P = 0.2736
Comparison: Rituximab + chemotherapy vs Obinutuzumab + chemotherapy
| Secondary endpoint | Analysis method | Effect estimate | P-value |
|---|---|---|---|
| Median Time to Progression-Free Survival (PFS), IRC-Assessed | Stratified log-rank test | HR 0.89 (95% CI 0.72–1.10) | 0.2736 |
The reported HR of 0.89 means that, under the stated group ordering, the estimated hazard of progression or death in the rituximab-plus-chemotherapy group was approximately 89% of that in the obinutuzumab-plus-chemotherapy group. Descriptively, this corresponds to an estimated hazard about 11% lower for rituximab relative to obinutuzumab.
The HR does not mean that 11% fewer participants experienced progression or death, nor does it describe an individual patient's probability of remaining progression-free.
The 95% CI of 0.72–1.10 spans 1. It therefore includes values compatible with a lower hazard as well as values compatible with a higher hazard for the rituximab group relative to the obinutuzumab group.
The P-value of 0.2736 describes the statistical evidence under the specified hypothesis-testing framework; it is not an effect-size measure. The numerical size of the HR and the width and location of the confidence interval provide information that the P-value alone does not.
This is a secondary analysis rather than the registered primary endpoint analysis. The analysis is also based on a different assessment framework and a different time frame from the primary investigator-assessed PFS analysis, so the two estimates should not be treated as if they were measurements from an identical analysis dataset.
9. Comparing the Two PFS Analyses
The two posted PFS analyses provide a useful statistical teaching example because both address the same broad time-to-event construct while differing in assessment framework and time frame.
| Feature | Primary PFS | Secondary PFS |
|---|---|---|
| Assessment | Investigator-assessed | Independent Review Committee-assessed |
| Time frame | Baseline up to approximately 6.5 years (up to 31 January 2018) | Baseline up to approximately 4 years and 9 months (up to 29 April 2016) |
| Population | ITT | ITT |
| Method | Stratified log-rank test | Stratified log-rank test |
| Stratification | IPI score | IPI score |
| Effect measure | Hazard ratio | Hazard ratio |
| HR | 0.94 | 0.89 |
| 95% CI | 0.78–1.12 | 0.72–1.10 |
| P-value | 0.4753 | 0.2736 |
The estimates are numerically different, but that difference should not be interpreted as a treatment-effect change without accounting for the different endpoint assessment and cutoff dates. Statistical comparisons across analyses require attention to the populations, data cutoffs, endpoint definitions, and analysis procedures used in each comparison.
10. Safety Results
The ClinicalTrials.gov record reports serious adverse events by randomized arm using affected participants over participants at risk. These are the only arm-specific safety counts provided in the trial data.
| Safety measure | Rituximab + chemotherapy | Obinutuzumab + chemotherapy |
|---|---|---|
| Serious adverse events, affected / at risk | 269 / 701 | 312 / 702 |
These counts should be interpreted as serious adverse events, not as overall adverse-event incidence, grade 3 or higher adverse events, treatment discontinuations, or deaths. The ClinicalTrials.gov record does not provide additional arm-specific safety categories, so those outcomes are not added.
11. Statistical Methods Explained
Why was a log-rank test used?
Progression-free survival is a time-to-event endpoint. Participants can be followed for different lengths of time, and some may not experience progression or death during observed follow-up. The log-rank test is designed to compare event-time distributions between groups while incorporating the timing of events and censoring rather than reducing each participant to a simple yes/no outcome.
What does an HR of 0.94 mean?
For the posted primary analysis, the comparison is rituximab plus chemotherapy versus obinutuzumab plus chemotherapy. An HR of 0.94 means the estimated instantaneous event rate in the rituximab group was approximately 94% of that in the obinutuzumab group under the reported time-to-event analysis. It is not an absolute probability and should not be read as "6% of patients benefited."
Why does the confidence interval include 1 matter?
For a hazard ratio, 1 represents equal estimated hazard between the comparison groups. The primary 95% CI is 0.78–1.12, so it spans that reference value. The interval therefore expresses uncertainty that includes both directions of relative difference.
Why doesn't the P-value measure the size of the treatment effect?
The P-value addresses evidence against a specified null hypothesis under the statistical model and testing framework. It does not directly quantify how large or clinically important the treatment difference is. The HR gives the estimated relative effect, while the confidence interval communicates its statistical precision.
Why was the analysis stratified by IPI score?
The registry states that the primary and secondary PFS analyses were stratified by International Prognostic Index score using the specified low/low-intermediate category definition. Stratification incorporates this prespecified prognostic structure into the time-to-event comparison rather than ignoring it.
Why is the ITT population important?
The ITT principle analyzes participants according to randomized assignment. Because randomization creates the basis for a comparable treatment contrast, preserving randomized assignment helps maintain that basis when participants discontinue treatment, deviate from treatment, or otherwise have differing experiences after randomization.
12. Reading the Hazard Ratio Correctly
The primary HR is below 1, so its point estimate corresponds to a lower estimated hazard for the rituximab-plus-chemotherapy group under the stated group ordering.
The 95% CI extends from 0.78 to 1.12. That interval is not a range of individual patient outcomes. It is a statistical interval around the estimated relative hazard under the analysis framework.
A hazard ratio and a median PFS answer different questions. The registry endpoint is defined using a Kaplan-Meier median, while the posted formal comparison reports an HR from the log-rank analysis. One should not be substituted for the other.
The primary P-value of 0.4753 provides a hypothesis-testing result. It does not quantify the magnitude of the HR, the absolute difference in PFS, or the clinical importance of the comparison.
13. Primary Analysis in Statistical Context
The GOYA primary analysis combines several standard clinical-trial statistical principles in a single time-to-event comparison.
| Component | GOYA implementation | Why it matters |
|---|---|---|
| Randomization | Randomized, parallel-group phase 3 design | Creates the basis for comparing outcomes according to treatment assignment. |
| Primary endpoint | Investigator-assessed PFS | Requires methods that account for event timing and censoring. |
| Estimation | Kaplan-Meier framework | Allows estimation of time-to-event distributions with censored observations. |
| Comparison | Stratified log-rank test | Compares treatment groups across event times while incorporating the stated IPI stratification. |
| Effect measure | Hazard ratio | Provides a relative summary of the event hazard between groups. |
| Analysis population | ITT | Preserves randomized treatment assignment as the basis of the efficacy comparison. |
| Hypothesis type | Superiority | The statistical question is whether the randomized groups differ under a superiority framework. |
14. What the Registry Data Allow Us to Conclude Statistically
The ClinicalTrials.gov record provides a formal primary endpoint analysis with an HR of 0.94, a two-sided 95% CI of 0.78–1.12, and a P-value of 0.4753. The analysis was based on the ITT population and used a stratified log-rank test with IPI-score stratification.
These values describe the observed statistical comparison. They do not by themselves provide an absolute PFS difference, a numerical median PFS for either group, or a complete description of the individual event and censoring histories. Those quantities require additional information not contained in the ClinicalTrials.gov record.
The secondary IRC-assessed PFS analysis similarly reports an HR of 0.89, 95% CI 0.72–1.10, and P-value 0.2736. Because this analysis has a different assessment framework and cutoff date, it should be interpreted as its own statistical analysis rather than treated as a duplicate measurement of the primary analysis.
15. Limitations
- Limited reported endpoint detail: the registry analysis reports hazard ratios, confidence intervals, and P-values but does not provide numerical median PFS estimates for the treatment groups.
- Time-to-event interpretation: hazard ratios summarize relative event hazards and are not equivalent to absolute risk differences or median differences.
- Confidence interval: the primary 95% CI of 0.78–1.12 spans 1, so the point estimate should not be interpreted without its uncertainty interval.
- Assessment framework: the primary analysis was investigator-assessed, while the secondary analysis was IRC-assessed. These are distinct analyses.
- Different follow-up periods: the primary analysis extends to 31 January 2018, whereas the IRC analysis extends to 29 April 2016.
- Stratification: the analysis was stratified by IPI score using the specific category definition reported by the registry. The hazard ratio should therefore be understood in the context of that stratified analysis.
- Analysis-population issue: the primary analysis notes a serious Good Clinical Practice non-compliance issue at a single study site in China involving all 4 participants enrolled there. The ClinicalTrials.gov record does not provide enough additional detail to characterize its statistical consequences beyond the registry's stated analysis-population description.
- Safety denominator: serious adverse-event counts are reported over participants at risk and should not be treated as if their denominator were the full enrollment of 1418.
- Incomplete methodological detail: the ClinicalTrials.gov record identifies the log-rank test, hazard ratio, ITT population, and IPI stratification but do not provide a complete statistical analysis plan. Additional unreported modeling or missing-data procedures should not be inferred.
- Multiplicity: the ClinicalTrials.gov record does not specify a multiplicity-adjustment procedure or endpoint hierarchy beyond identifying the primary endpoint and a secondary PFS analysis. No such procedure is added here.
16. Why This Trial Matters Statistically
GOYA is a useful clinical-trial statistics teaching case because its registry record connects a randomized phase 3 comparison directly to a classic survival-analysis workflow. The endpoint is time-to-event, the endpoint definition explicitly invokes Kaplan-Meier estimation, the formal comparison uses a stratified log-rank test, and the treatment effect is expressed as a hazard ratio.
| Statistical concept | How it appears in GOYA |
|---|---|
| Randomization | Participants were randomized in a parallel-group phase 3 treatment study. |
| Intention-to-treat analysis | The posted primary analysis used an ITT population. |
| Time-to-event endpoints | The primary endpoint is investigator-assessed PFS. |
| Kaplan-Meier estimation | The registered endpoint defines median PFS using the Kaplan-Meier estimate. |
| Log-rank testing | The primary and secondary formal comparisons use a log-rank test. |
| Stratified analysis | The analyses are stratified by the specified IPI-score category. |
| Hazard ratio | The primary and secondary treatment effects are reported as HRs. |
| Confidence intervals | The primary HR is accompanied by a two-sided 95% CI of 0.78–1.12. |
| P-values | The primary comparison reports P = 0.4753. |
| Independent assessment | A secondary PFS analysis was IRC-assessed. |
| Safety analysis | Serious adverse events are reported by treatment arm using affected/at-risk counts. |
17. Trial Timeline
Trial start
The GOYA trial began on 26 July 2011 according to the registry profile.
Primary completion
The registry lists 29 April 2016 as the primary completion date and uses this date as the upper time point for the IRC-assessed PFS analysis.
Primary PFS analysis cutoff
The investigator-assessed primary PFS endpoint has a time frame through approximately 6.5 years, up to 31 January 2018.
Terminated
The ClinicalTrials.gov profile classifies the study status as terminated.
18. Interpreting the Primary and Secondary Estimates Together
The primary HR of 0.94 and secondary HR of 0.89 are both below 1 when expressed as rituximab plus chemotherapy relative to obinutuzumab plus chemotherapy. However, the corresponding confidence intervals are 0.78–1.12 and 0.72–1.10, respectively, and both intervals include 1.
The statistical lesson is that the point estimate alone is not sufficient. A complete interpretation considers the direction of the effect, magnitude of the estimate, confidence interval, P-value, analysis population, endpoint definition, stratification, and assessment framework.
Point estimate
The HR summarizes the estimated relative event hazard between the randomized groups under the stated comparison.
Precision
The 95% confidence interval shows the statistical uncertainty around the point estimate.
Hypothesis test
The P-value evaluates the statistical evidence under the specified superiority testing framework; it does not measure treatment magnitude.
Analysis context
The interpretation depends on the ITT population, IPI stratification, time frame, and investigator versus IRC assessment.
19. Statistical Interpretation vs Clinical Interpretation
Statistical interpretation
The primary investigator-assessed PFS analysis reported an HR of 0.94 with a two-sided 95% CI of 0.78–1.12 and P = 0.4753 using a stratified log-rank test in the ITT population. The secondary IRC-assessed analysis reported an HR of 0.89 with a 95% CI of 0.72–1.10 and P = 0.2736.
Clinical interpretation
The ClinicalTrials.gov record does not provide enough numerical information to characterize median PFS by treatment group or an absolute difference in PFS. Accordingly, this page does not convert the hazard-ratio results into an unreported absolute treatment effect.
This distinction is important. Statistical evidence can describe the relative treatment comparison, while clinical interpretation often requires absolute event probabilities, median event times, duration of benefit, adverse-event profiles, and other information. The ClinicalTrials.gov record provides only some of those components.
20. Serious Adverse Events in Context
The visualization above preserves the registry's reported affected/at-risk counts rather than converting them into newly calculated percentages. The important statistical point is that the safety denominator is explicitly provided as the number at risk for each arm.
21. What Is and Is Not Being Inferred
| Information | Status in this analysis |
|---|---|
| Enrollment | Reported directly: 1418 |
| Primary PFS HR | Reported directly: 0.94 |
| Primary 95% CI | Reported directly: 0.78–1.12 |
| Primary P-value | Reported directly: 0.4753 |
| IRC PFS HR | Reported directly: 0.89 |
| IRC 95% CI | Reported directly: 0.72–1.10 |
| IRC P-value | Reported directly: 0.2736 |
| Median PFS by arm | Not numerically reported in the ClinicalTrials.gov record |
| Subgroup HRs | Not reported |
| Long-term survival rates | Not reported |
| Additional safety categories | Not reported |
| Multiplicity procedure | Not reported |
| Interim-analysis procedure | Not reported |
| Missing-data or imputation procedure | Not reported |
| Bayesian analysis | Not reported in the ClinicalTrials.gov record |
This distinction prevents a common trial-analysis error: filling gaps in a registry record with results remembered from an external publication. For this page, the ClinicalTrials.gov record is the numerical boundary for the analysis.
22. Related Tutorials
Learn more about the methods used in this trial:
23. Related Calculators
24. Sources
- ClinicalTrials.gov: GOYA, NCT01287741. Official trial registry record and source for the trial data presented on this page.
- PubMed: PubMed record for PMID 42699559.
- PubMed: PubMed record for PMID 39962481.
- PubMed: PubMed record for PMID 38843483.
- PubMed: PubMed record for PMID 38498084.
- PubMed: PubMed record for PMID 37702406.
The numerical results and trial-specific statistical statements on this page are restricted to the ClinicalTrials.gov record. The linked PubMed records are provided as the publication records associated with the trial and are not used to introduce additional numerical results.
Continue through the Clinical Biostats statistical pathway
Use the trial's time-to-event methods as a starting point for deeper study of hazard ratios, Kaplan-Meier estimation, log-rank testing, confidence intervals, randomization, and intention-to-treat analysis.
25. Record Summary
GOYA provides a focused example of randomized clinical-trial survival analysis. The registry identifies a phase 3 randomized parallel design with 1418 enrolled participants, a primary investigator-assessed PFS endpoint, an ITT analysis population, IPI-score stratification, a stratified log-rank test, and hazard ratio as the treatment-effect measure.
The primary analysis reports an HR of 0.94 with a two-sided 95% CI of 0.78–1.12 and P = 0.4753. The secondary IRC-assessed PFS analysis reports an HR of 0.89 with a 95% CI of 0.72–1.10 and P = 0.2736. Serious adverse events are reported as 269/701 for rituximab plus chemotherapy and 312/702 for obinutuzumab plus chemotherapy.
The most important statistical lesson is that these results should be read as a complete set: effect estimate, confidence interval, P-value, endpoint definition, analysis population, stratification, and assessment framework. The hazard ratio supplies a relative time-to-event measure; the confidence interval communicates its statistical uncertainty; and the P-value addresses the hypothesis test. None of these quantities, alone or together, should be converted into unreported median PFS values or individual-level probabilities.