This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the MAGNITUDE ClinicalTrials.gov record.
1. Trial at a Glance
MAGNITUDE is a phase 3 randomized parallel-group trial evaluating niraparib in combination with abiraterone acetate and prednisone versus abiraterone acetate and prednisone for participants with castration-resistant prostatic cancer. The ClinicalTrials.gov record identifies two primary time-to-event endpoints, both analyzed with log-rank testing and hazard ratios.
| Feature | MAGNITUDE |
|---|---|
| Trial name | MAGNITUDE |
| NCT ID | NCT03748641 |
| Phase | Phase 3 |
| Condition | Castration-Resistant Prostatic Cancer |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Quadruple |
| Primary purpose | Treatment |
| Enrollment | 765 |
| Start | 2019-01-25 |
| Primary completion | 2021-10-08 |
| Status | Active, not recruiting |
| Lead sponsor | Janssen Research & Development, LLC |
| Sponsor type | Industry |
2. Clinical Question
The central statistical question is whether adding niraparib to abiraterone acetate and prednisone improves radiographic progression-free survival compared with abiraterone acetate and prednisone alone in the prespecified randomized comparisons reported for MAGNITUDE.
Population
Participants with castration-resistant prostatic cancer enrolled in the MAGNITUDE phase 3 trial.
Intervention
Niraparib in combination with abiraterone acetate and prednisone.
Comparator
Abiraterone acetate and prednisone with placebo in the reported comparisons.
Primary question
Does the niraparib-containing regimen change the time to radiographic progression or death relative to the comparator regimen?
3. Trial Design
Niraparib + abiraterone
- Niraparib 200 mg
- Abiraterone acetate 1000 mg
- Prednisone 10 mg
Placebo + abiraterone
- Placebo
- Abiraterone acetate 1000 mg
- Prednisone 10 mg
4. Trial Timing and Registry Status
Trial start
the ClinicalTrials.gov record lists January 25, 2019 as the study start date.
Primary completion
the ClinicalTrials.gov record lists October 8, 2021 as the primary completion date.
Active, not recruiting
The ClinicalTrials.gov profile identifies the study status as ACTIVE_NOT_RECRUITING.
5. Primary Endpoints
| Endpoint | Time frame | Registry definition reported | Analysis |
|---|---|---|---|
| Cohort 1: Radiographic Progression-Free Survival (rPFS) as Assessed by Blinded Independent Central Review (BICR) | Up to 32 months | As per BICR, rPFS is time interval from the date of randomization to radiographic progression or death, whichever occurred first. Radiographic progression was determined by progression of soft tissue lesions measured by CT or MRI as per RECIST 1.1 and progression of bone lesions observed by bone scan based on PCWG3 criteria. | Log-rank test; hazard ratio; superiority hypothesis |
| Cohort 1 Breast Cancer Gene (BRCA) Subgroup: Radiographic Progression-Free Survival (rPFS) as Assessed by Blinded Independent Central Review (BICR) | Up to 32 months | As per BICR, rPFS is time interval from the date of randomization to radiographic progression or death, whichever occurred first. Radiographic progression was determined by progression of soft tissue lesions measured by CT or MRI as per RECIST 1.1 and progression of bone lesions observed by bone scan based on PCWG3 criteria. | Log-rank test; hazard ratio; superiority hypothesis |
Both posted primary endpoints are therefore time-to-event outcomes. The event definition combines radiographic progression and death: whichever occurs first defines the endpoint event. This is important statistically because a participant who dies before documented radiographic progression is counted as having experienced the rPFS event.
6. Analysis Populations and Statistical Comparisons
| Analysis | Population | Comparison | Method | Effect measure |
|---|---|---|---|---|
| Cohort 1 rPFS | Randomized analysis set for Cohort 1, including all randomized participants in Cohort 1 | Niraparib 200 mg + abiraterone acetate 1000 mg + prednisone 10 mg vs placebo + abiraterone acetate 1000 mg + prednisone 10 mg | Log-rank test | Hazard ratio |
| Cohort 1 BRCA-subgroup rPFS | Randomized analysis set for Cohort 1 BRCA subgroup, including all randomized participants in that subgroup | Niraparib 200 mg + abiraterone acetate 1000 mg + prednisone 10 mg vs placebo + abiraterone acetate 1000 mg + prednisone 10 mg | Log-rank test | Hazard ratio |
The registry explicitly identifies the randomized analysis sets for these analyses. That matters because treatment assignment remains the basis for the comparison rather than redefining groups according to treatment received after randomization.
7. Statistical Methodology
Time-to-event analysis
rPFS is not an ordinary continuous outcome because patients can be followed for different lengths of time and may experience the event at different points. The analysis therefore uses survival-analysis methods that incorporate both observed events and right-censored follow-up.
For the registered rPFS endpoint, the event occurs at radiographic progression or death, whichever occurs first. Participants without an observed event during available follow-up contribute information up to their censoring time.
Log-rank test
The registry method for both primary analyses is the log-rank test. The log-rank test compares the survival experience of two groups over follow-up rather than comparing only a single time point or a simple proportion.
Conceptually, the test repeatedly compares the observed number of events in each treatment group with the number expected under the null hypothesis of no difference in the time-to-event distributions. Because the calculation is built around event times and the numbers at risk, it can use information from participants who have different follow-up durations.
Hazard ratio
The effect measure reported for both primary analyses is the hazard ratio (HR). A hazard ratio below 1 indicates a lower estimated instantaneous event rate in the niraparib-containing group relative to the comparator, under the model and follow-up represented by the analysis.
For example, an HR of 0.729 corresponds to a treatment-group hazard that is 72.9% of the comparator hazard under the estimated relative-hazard interpretation. The complementary arithmetic interpretation is a 27.1% lower estimated hazard, calculated as 1 − 0.729.
Superiority hypothesis
The registry identifies the hypothesis type for both analyses as superiority. This means the statistical question is whether the treatment comparison provides evidence of a difference in favor of the niraparib-containing regimen, rather than whether it merely remains within a prespecified non-inferiority margin.
8. Primary Result: Cohort 1 rPFS
The first posted primary analysis evaluated radiographic progression-free survival in Cohort 1 using the randomized analysis set. The comparison was niraparib 200 mg + abiraterone acetate 1000 mg + prednisone 10 mg versus placebo + abiraterone acetate 1000 mg + prednisone 10 mg.
Hazard ratio for rPFS
95% CI: 0.556–0.956 · P = 0.0217
Two-sided 95% confidence interval · Log-rank test · Superiority hypothesis
| Measure | Cohort 1 rPFS analysis |
|---|---|
| Endpoint | Radiographic Progression-Free Survival as assessed by BICR |
| Time frame | Up to 32 months |
| Analysis population | Randomized analysis set for Cohort 1 |
| Comparison | Niraparib 200 mg + abiraterone acetate 1000 mg + prednisone 10 mg vs placebo + abiraterone acetate 1000 mg + prednisone 10 mg |
| Method | Log Rank |
| Effect measure | Hazard Ratio |
| Estimate | 0.729 |
| 95% CI | 0.556–0.956 |
| P-value | 0.0217 |
| Hypothesis | Superiority |
The estimated HR of 0.729 means that, within the statistical framework represented by this analysis, the estimated instantaneous rate of the rPFS event was 72.9% of the corresponding rate in the comparator group. Expressed as a complementary relative quantity, this is a 27.1% lower estimated hazard.
The HR does not mean that 27.1% of participants avoided progression, that each participant experienced exactly a 27.1% reduction in risk, or that median rPFS differed by 27.1%. A hazard ratio is a relative time-to-event measure, not an absolute probability or an individual-level prediction.
The two-sided 95% CI of 0.556–0.956 describes the uncertainty around the estimated relative hazard under the analysis framework. Because the interval remains below 1, the interval is consistent with a lower estimated event hazard for the niraparib-containing group throughout the interval represented by the reported confidence limits.
The P-value of 0.0217 addresses statistical evidence against the null hypothesis under the specified testing framework. It does not measure the size of the treatment effect, the probability that the treatment works, or the probability that the null hypothesis is true. The effect estimate and its confidence interval are therefore essential parts of the interpretation.
Finally, rPFS is subject to the usual time-to-event considerations, including censoring and the way progression is assessed. Here, progression was assessed by BICR, which provides a prespecified central assessment rather than relying solely on local investigator assessment.
9. Primary Result: Cohort 1 BRCA Subgroup rPFS
The second posted primary analysis evaluated rPFS in the Cohort 1 Breast Cancer Gene (BRCA) subgroup. The analysis population was the randomized analysis set for the Cohort 1 BRCA subgroup.
Hazard ratio for BRCA-subgroup rPFS
95% CI: 0.361–0.789 · P = 0.0014
Two-sided 95% confidence interval · Log-rank test · Superiority hypothesis
| Measure | Cohort 1 BRCA subgroup analysis |
|---|---|
| Endpoint | Breast Cancer Gene (BRCA) Subgroup: Radiographic Progression-Free Survival as assessed by BICR |
| Time frame | Up to 32 months |
| Analysis population | Randomized analysis set for Cohort 1 BRCA subgroup |
| Comparison | Niraparib 200 mg + abiraterone acetate 1000 mg + prednisone 10 mg vs placebo + abiraterone acetate 1000 mg + prednisone 10 mg |
| Method | Log Rank |
| Effect measure | Hazard Ratio |
| Estimate | 0.533 |
| 95% CI | 0.361–0.789 |
| P-value | 0.0014 |
| Hypothesis | Superiority |
The estimated HR of 0.533 corresponds to an estimated instantaneous rPFS event rate that is 53.3% of the comparator rate under the hazard-ratio interpretation. The complementary arithmetic interpretation is a 46.7% lower estimated hazard.
Again, this is not equivalent to saying that 46.7% of participants were protected from progression or death. It is a relative time-to-event estimate that summarizes the treatment comparison over the analyzed follow-up.
The two-sided 95% CI of 0.361–0.789 is substantially below 1. The width of the interval also matters: the data support a range of plausible relative hazard estimates rather than a single exact treatment effect.
The P-value of 0.0014 provides evidence against the relevant null hypothesis within the reported testing framework. It should not be interpreted as an effect-size metric. A smaller P-value does not automatically mean a larger clinical effect, and the HR with its confidence interval remains the primary quantitative description of the relative effect reported by the registry.
Because this is a subgroup analysis, the estimate also needs to be interpreted in the context of the subgroup's analysis population. A subgroup-specific HR describes the treatment comparison within that subgroup; it does not by itself establish that the treatment effect differs from the effect in participants outside the subgroup. Demonstrating effect modification requires a formal comparison of treatment effects across groups.
10. Comparing the Two Reported Primary Estimates
| Primary analysis | HR | 95% CI | P-value | Complementary interpretation |
|---|---|---|---|---|
| Cohort 1 rPFS | 0.729 | 0.556–0.956 | 0.0217 | 27.1% lower estimated hazard |
| Cohort 1 BRCA-subgroup rPFS | 0.533 | 0.361–0.789 | 0.0014 | 46.7% lower estimated hazard |
The two estimates are both below 1, but they should not be converted into a ranking of treatment effects. The second estimate describes a particular subgroup, whereas the first describes the broader Cohort 1 randomized analysis set. Differences between estimates can arise from the populations being analyzed, the number of events contributing information, and ordinary sampling variation.
11. Why the Kaplan-Meier Framework Matters
Although the statistical analyses posted on ClinicalTrials.gov report hazard ratios and log-rank tests rather than reconstructed Kaplan-Meier curves, the endpoint itself is a survival-analysis outcome. Kaplan-Meier estimation is a natural descriptive framework for displaying the probability of remaining event-free over time.
Here, di represents the number of events at event time ti, while ni represents the number at risk immediately before that time.
The important educational point is that a time-to-event analysis does not require every participant to experience the event. Participants can contribute follow-up until progression, death, or censoring. This makes survival methods particularly useful when follow-up duration varies between participants.
The registry-reported MAGNITUDE data do not provide median rPFS values, time-specific rPFS percentages, or patient-level event and censoring times. Those quantities therefore cannot be reconstructed from the ClinicalTrials.gov record without introducing information that is outside the requested trial dataset.
12. Why the Log-Rank Test Was Used
The log-rank test is specifically designed for comparing time-to-event distributions between groups. It differs from a simple comparison of the proportion of patients who progressed because it incorporates when events occurred and the number of participants still at risk at each event time.
Uses event timing
A participant who experiences progression early contributes information differently from a participant who remains event-free for a longer period.
Handles censoring
Participants who do not experience the event during available follow-up can still contribute information before censoring.
Matches the endpoint
rPFS is explicitly defined as a time interval from randomization to progression or death, making a survival-analysis framework appropriate.
Complements the HR
The log-rank test supplies a hypothesis test, while the hazard ratio supplies a quantitative estimate of relative event hazard.
13. Statistical Methods Explained
Why was a log-rank test used?
Because rPFS is a time-to-event endpoint. A log-rank test compares the event-time experience of the randomized groups while accounting for the changing number of participants at risk over follow-up. It is therefore more appropriate for rPFS than a simple chi-square comparison of whether an event ever occurred.
What does an HR of 0.729 mean?
An HR of 0.729 means the estimated instantaneous rPFS event rate in the niraparib-containing group was 72.9% of the comparator rate under the hazard-ratio interpretation. The complementary arithmetic statement is a 27.1% lower estimated hazard. It does not mean a 27.1% absolute reduction in the probability of progression or death.
Why does the confidence interval matter?
The point estimate is only one estimate of the treatment effect. The 95% CI of 0.556–0.956 communicates uncertainty around the Cohort 1 estimate. It shows that the analysis does not identify a single exact hazard ratio; instead, it provides an interval of values compatible with the statistical model and data under the stated confidence procedure.
Why does the P-value not measure effect size?
A P-value measures how compatible the observed data are with a specified null hypothesis under the testing framework. It does not tell us how large the treatment effect is. Two studies can produce different P-values for effects of similar size because sample size, event count, variability, and information content differ.
Why is the BRCA subgroup analysis different from the overall Cohort 1 analysis?
The BRCA analysis is restricted to the Cohort 1 BRCA subgroup, while the first analysis uses the randomized analysis set for Cohort 1. A subgroup HR therefore answers a narrower question. Its estimate should not automatically be interpreted as proof that the treatment works differently according to BRCA status without a formal treatment-by-subgroup interaction analysis.
Why is BICR important for rPFS?
The endpoint was assessed by Blinded Independent Central Review. Central review can provide a standardized assessment of radiographic progression and helps separate the endpoint definition from treatment-group knowledge at the individual assessment level. The registry definition also specifies radiographic progression using CT or MRI for soft-tissue lesions according to RECIST 1.1 and bone-scan assessment according to PCWG3 criteria.
What does the rPFS event actually represent?
The registered definition makes rPFS the interval from randomization to radiographic progression or death, whichever occurs first. Therefore, death is an rPFS event even if radiographic progression was not documented first. This composite event definition is an important part of interpreting the resulting hazard ratio.
14. Confidence Intervals and Precision
Confidence intervals are particularly important when interpreting hazard ratios because the point estimate alone can create a false impression of exactness. The registry-reported Cohort 1 analysis estimates an HR of 0.729, but the 95% CI extends from 0.556 to 0.956.
The intervals are not probability statements about where the true treatment effect "has a 95% chance" of being. In frequentist inference, the confidence procedure describes the long-run coverage behavior of intervals constructed in this way. For an individual analysis, the interval communicates statistical uncertainty conditional on the analysis framework.
It is also useful to distinguish precision from clinical importance. A narrow confidence interval can indicate a precisely estimated effect that is clinically modest, while a wide interval can surround effects that range from small to substantial. The HR, confidence interval, and clinical context therefore answer different questions.
15. P-Values and Hypothesis Testing
The analyses posted on ClinicalTrials.gov report two-sided confidence intervals and P-values of 0.0217 and 0.0014. Both analyses are identified as superiority tests.
| Analysis | Two-sided 95% CI | P-value | Hypothesis type |
|---|---|---|---|
| Cohort 1 rPFS | 0.556–0.956 | 0.0217 | Superiority |
| Cohort 1 BRCA subgroup rPFS | 0.361–0.789 | 0.0014 | Superiority |
For a hazard ratio, the null value is generally 1, because an HR of 1 represents equal instantaneous event rates under the hazard-ratio interpretation. The reported confidence intervals do not include 1. This is consistent with the corresponding superiority-test results.
16. Censoring and Time-to-Event Interpretation
Time-to-event analysis is necessary partly because clinical follow-up is rarely synchronized across all participants. A participant may remain free of progression at the time the available observation ends. That participant has not reported a known event time, but the participant has the ClinicalTrials.gov record about being event-free up to the censoring time.
This distinction is central to interpreting rPFS. A censored participant is not automatically equivalent to a participant who never experienced progression. Censoring means that the event status is not observed after the available follow-up under the analysis framework.
Observed event
Radiographic progression or death occurs during observed follow-up.
Censored observation
No qualifying rPFS event is observed through the participant's available follow-up.
Risk set
The participants still considered at risk contribute to comparisons as event times are evaluated.
Why timing matters
The statistical analysis uses the timing of events rather than reducing follow-up to a simple yes/no outcome.
17. Blinding and Central Review
the ClinicalTrials.gov record identifies the study as quadruple masked. The primary rPFS endpoint was assessed by Blinded Independent Central Review.
Blinding is statistically relevant because knowledge of treatment assignment can influence assessment and other post-randomization decisions. For a radiographic endpoint, an independent blinded assessment provides a structured mechanism for reducing the opportunity for treatment assignment to influence progression classification.
This does not eliminate every possible source of bias. It does, however, establish an important separation between treatment assignment and the central radiographic assessment used for the reported primary endpoint.
18. Safety Results
The ClinicalTrials.gov record reports serious adverse events by arm and cohort. These are presented as affected participants divided by participants at risk.
| Cohort | Treatment group | Serious adverse events | Percentage of at-risk participants |
|---|---|---|---|
| Cohort 1 | Niraparib 200 mg + Abiraterone | 76 / 212 | 35.8% |
| Cohort 1 | Placebo + Abiraterone Acetate | 52 / 211 | 24.6% |
| Cohort 2 | Niraparib 200 mg + Abiraterone | 66 / 123 | 53.7% |
| Cohort 2 | Placebo + Abiraterone Acetate | 45 / 123 | 36.6% |
| Cohort 3 | Niraparib 200 mg and Abiratero | 21 / 95 | 22.1% |
The percentages in the final column are simple arithmetic descriptions of the registry-reported affected/at-risk counts and are included to make the reported fractions easier to interpret. They are not statistical effect estimates and should not be confused with a formally adjusted safety comparison.
19. What the Hazard Ratios Do — and Do Not — Mean
The HR of 0.729 describes the estimated relative rate of the rPFS event between the two randomized Cohort 1 treatment groups. The complementary arithmetic interpretation, 1 − 0.729, is a 27.1% lower estimated hazard.
It does not mean a 27.1% absolute increase in the probability of remaining progression-free, nor does it specify how long an individual patient will remain progression-free.
The HR of 0.533 describes the corresponding relative event rate within the Cohort 1 BRCA subgroup. Its complementary arithmetic interpretation is a 46.7% lower estimated hazard.
It does not establish that the treatment effect is 46.7% larger in BRCA-positive participants than in participants outside that subgroup. A formal interaction analysis would be required to make that comparison.
A hazard ratio is relative. To understand absolute benefit, analysts generally also examine quantities such as survival probabilities at prespecified times, median event times, or absolute risk differences when those data are available. The registry-reported MAGNITUDE dataset does not provide median rPFS or time-specific rPFS estimates, so those measures are not added here.
20. Statistical Interpretation of the Confidence Intervals
The Cohort 1 95% CI extends from 0.556 to 0.956. The BRCA-subgroup 95% CI extends from 0.361 to 0.789. Both intervals are entirely below the null value of 1.
The intervals should not be read as the range of treatment effects that individual patients will experience. They quantify uncertainty in the estimated population-level treatment effect under the statistical model and confidence procedure.
The difference between the two interval widths is also informative. The subgroup estimate is based on a narrower population than the full Cohort 1 analysis, so subgroup estimates generally require particular attention to precision. The ClinicalTrials.gov record does not provide the number of BRCA-subgroup participants or the number of events, so no additional precision calculation is appropriate here.
21. Multiplicity and Multiple Primary Analyses
The registry profile identifies 2 primary endpoints and 2 posted primary-endpoint analyses. The two analyses are Cohort 1 rPFS and Cohort 1 BRCA-subgroup rPFS.
| Primary analysis | Role | Reported statistical evidence |
|---|---|---|
| Cohort 1 rPFS | Primary | HR 0.729; 95% CI 0.556–0.956; P = 0.0217 |
| Cohort 1 BRCA-subgroup rPFS | Primary | HR 0.533; 95% CI 0.361–0.789; P = 0.0014 |
When multiple hypotheses are formally tested within a clinical trial, the interpretation of each P-value depends on the prespecified multiplicity strategy. For example, a trial may allocate a familywise type I error rate across endpoints or use a hierarchical testing strategy. The registry-reported MAGNITUDE data identify the primary analyses and their results but do not provide an alpha-allocation or multiplicity procedure.
22. Interim Analysis, Missing Data, and Other Design Features
The ClinicalTrials.gov record does not report an interim-analysis strategy, alpha-spending method, missing-data or imputation procedure, non-inferiority margin, crossover procedure, Bayesian analysis, or treatment-stratification factors.
| Design topic | Information posted on ClinicalTrials.gov for MAGNITUDE |
|---|---|
| Non-inferiority margin | Not reported in the ClinicalTrials.gov record; the registered hypothesis type is superiority. |
| Crossover | Not reported in the ClinicalTrials.gov record. |
| Factorial design | Not reported; the design model registry-reported is parallel. |
| Multiplicity procedure | Not reported in the ClinicalTrials.gov record. |
| Interim analysis | Not reported in the ClinicalTrials.gov record. |
| Missing-data/imputation method | Not reported in the ClinicalTrials.gov record. |
| Stratification factors | Not reported in the ClinicalTrials.gov record. |
| Bayesian methods | Not reported; the analyses posted on ClinicalTrials.gov use log-rank testing and hazard ratios. |
This distinction is important. A statistical analysis page should not fill gaps in a registry extract by assuming that a familiar oncology trial used a particular interim boundary, stratification scheme, imputation method, or multiplicity adjustment. Those are protocol-specific features and should be reported only when documented.
23. Limitations
- Limited reported efficacy measures: the ClinicalTrials.gov record contains two primary rPFS analyses but do not provide median rPFS, time-specific rPFS estimates, or reconstructed patient-level survival information.
- Subgroup interpretation: the BRCA analysis is a subgroup analysis. Its HR should not be interpreted as evidence of treatment-effect heterogeneity without a formal interaction assessment.
- Multiplicity information: two primary analyses are reported, but the ClinicalTrials.gov record does not specify the alpha-allocation or other multiplicity strategy.
- Censoring: rPFS is a time-to-event endpoint and therefore depends on the handling and assumptions associated with censored observations.
- Hazard-ratio interpretation: a single HR summarizes relative event hazard and should not be treated as an absolute risk difference or individual-level prediction.
- Proportional-hazards considerations: hazard-ratio interpretation is most straightforward when the relative hazard is reasonably stable over time. The ClinicalTrials.gov record does not report a proportional-hazards assessment.
- Safety denominators differ by cohort: the registry-reported serious-adverse-event counts are reported separately for cohorts and should not be pooled without an appropriate prespecified analysis.
- Incomplete statistical design detail: the ClinicalTrials.gov recordset does not report interim monitoring, missing-data methods, stratification factors, or a detailed multiplicity procedure.
- Three-arm design: although the overall trial has 3 arms, the registry-reported primary analyses are specific two-group comparisons. The reported HRs should not be generalized to the unreported third-arm comparison.
24. Why This Trial Matters Statistically
MAGNITUDE is a useful teaching case because the ClinicalTrials.gov record illustrate a complete core time-to-event analysis without requiring a large collection of secondary endpoints. The central statistical structure is clear: randomized comparison, blinded central assessment, a defined composite time-to-event endpoint, log-rank testing, hazard ratios, two-sided confidence intervals, and a subgroup analysis.
| Concept | How it appears in MAGNITUDE |
|---|---|
| Randomization | The trial allocation is randomized. |
| Parallel-group design | The design model is parallel. |
| Quadruple masking | the ClinicalTrials.gov record identifies the masking as quadruple. |
| Time-to-event endpoint | Both registered primary endpoints are rPFS outcomes. |
| BICR | Both primary rPFS endpoints are assessed by Blinded Independent Central Review. |
| Kaplan-Meier framework | rPFS is a time-to-event endpoint for which Kaplan-Meier estimation is a standard descriptive method. |
| Log-rank testing | The registry reports Log Rank for both posted primary analyses. |
| Hazard ratio | Both primary analyses use HR as the effect measure. |
| Confidence interval | Both reported HRs include two-sided 95% confidence intervals. |
| Superiority testing | Both analyses are identified as superiority hypotheses. |
| Subgroup analysis | A primary analysis is specifically reported for the Cohort 1 BRCA subgroup. |
| Safety analysis | Serious adverse events are reported as affected participants divided by those at risk across cohorts. |
25. A Practical Statistical Reading of MAGNITUDE
A useful way to read the primary results is to separate the analysis into four questions.
1. What was measured?
rPFS, defined as the time from randomization to radiographic progression or death, whichever occurred first.
2. How was it compared?
The randomized groups were compared using the log-rank test, with hazard ratio as the effect measure.
3. How large was the estimate?
The Cohort 1 HR was 0.729, while the Cohort 1 BRCA-subgroup HR was 0.533.
4. How certain is the estimate?
The reported 95% CIs were 0.556–0.956 and 0.361–0.789, respectively.
This framework prevents a common statistical mistake: treating the P-value as the entire result. A complete interpretation combines the endpoint definition, analysis population, effect estimate, confidence interval, testing method, and relevant design features.
26. What Cannot Be Concluded From These Results Alone
- The HR does not establish an absolute probability of remaining progression-free at a particular time.
- The HR does not indicate the percentage of individual patients who benefit.
- The P-value does not quantify the magnitude of treatment benefit.
- The BRCA-subgroup HR does not, by itself, prove that BRCA status modifies treatment effect.
- The serious-adverse-event fractions do not constitute a formal comparative safety model.
- The reported two-sided 95% CIs do not describe the distribution of treatment effects among individual participants.
- The ClinicalTrials.gov record does not support a median-rPFS calculation, a Kaplan-Meier reconstruction, or an absolute rPFS difference.
- The ClinicalTrials.gov record does not support conclusions about unreported interim-monitoring, missing-data, multiplicity, or stratification procedures.
27. Related Tutorials
Learn more about the methods used in this trial:
28. Related Calculators
29. Sources
- ClinicalTrials.gov: NCT03748641 — MAGNITUDE.
- Linked PubMed record: PMID 42015879.
- Linked PubMed record: PMID 39317633.
- Linked PubMed record: PMID 39111209.
- Linked PubMed record: PMID 38958846.
- Linked PubMed record: PMID 36952634.
Continue through the Clinical Biostats statistical library
Use the related tutorials and calculators to examine the survival-analysis concepts that underlie randomized time-to-event trials.
30. Record Summary
MAGNITUDE provides a focused example of randomized clinical-trial time-to-event methodology. The ClinicalTrials.gov record identifies a phase 3 randomized, parallel-group, quadruple-masked study with 765 participants and two posted primary rPFS analyses. Both analyses used the log-rank test and hazard ratio as the effect measure, with two-sided 95% confidence intervals and superiority hypotheses. The Cohort 1 analysis reported an HR of 0.729 (95% CI 0.556–0.956; P = 0.0217), while the Cohort 1 BRCA-subgroup analysis reported an HR of 0.533 (95% CI 0.361–0.789; P = 0.0014).
The central statistical lesson is that a clinical-trial result should be read as a combination of endpoint definition, analysis population, effect estimate, confidence interval, and hypothesis test. For time-to-event endpoints, the timing of events and censoring structure are essential. The hazard ratio supplies a relative measure, while the confidence interval communicates precision and the P-value addresses the specified null hypothesis. Subgroup estimates require additional caution because a subgroup-specific estimate does not automatically establish treatment-effect heterogeneity.