This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
MIRAS is a real-world study evaluating mirvetuximab soravtansine (MIRV) in patients with ovarian cancer, fallopian tube cancers, and primary peritoneal cancer with expression of folate receptor α (FRα). The registry describes a single-arm study with planned enrollment of 400 participants and a primary safety endpoint focused on grade 3 or higher treatment-related adverse events.
| Feature | MIRAS |
|---|---|
| Study acronym | MIRAS |
| ClinicalTrials.gov identifier | NCT07690332 |
| Registry status | Not yet recruiting |
| Study type | Real-world study |
| Population | Ovarian cancer; fallopian tube cancers; primary peritoneal cancer |
| Intervention | Mirvetuximab Soravtansine (MIRV) |
| Number of arms | 1 |
| Planned enrollment | 400 |
| Lead sponsor | NING LI-GYN |
| Sponsor type | Other |
| Primary endpoint | Incidence of grade ≥ 3 treatment-related adverse events (TRAEs) |
2. Clinical Question
The central statistical question is descriptive rather than a randomized treatment comparison: among patients receiving mirvetuximab soravtansine in this real-world study, how frequently do grade 3 or higher treatment-related adverse events occur during the prespecified safety-observation period?
Population
Patients with ovarian cancer, fallopian tube cancers, or primary peritoneal cancer in the MIRAS real-world study population, with expression of folate receptor α (FRα).
Intervention
Mirvetuximab soravtansine (MIRV).
Comparator
The registry describes 1 study arm; therefore, no randomized comparator arm is identified.
Primary question
What is the incidence of grade 3 or higher treatment-related adverse events assessed by the investigator?
3. Trial Design
The single-arm structure is statistically important. In a randomized comparative trial, a treatment effect can be expressed through a contrast between randomized groups. Here, the registered primary endpoint is an incidence measure within the MIRAS treatment population. The principal estimand is therefore naturally expressed as a proportion or incidence rather than as a treatment-versus-control effect.
4. Endpoints
| Endpoint | Registry definition | Time frame |
|---|---|---|
| Incidence of grade ≥ 3 treatment-related adverse events (TRAEs) | Incidence of grade 3 or higher treatment-related adverse events (TRAEs) assessed by the investigator. | From first dose to 30 days after last dose of MIRV |
The endpoint combines three important concepts: severity, relationship to treatment, and time at risk. The severity threshold is grade 3 or higher; treatment relatedness is determined by the investigator; and the observation period extends from the first dose through 30 days after the last MIRV dose.
Why the time frame matters
An adverse-event incidence is not fully defined by its numerator and denominator alone. The period during which events can qualify is part of the endpoint definition. Here, an event occurring within the specified window can contribute to the primary endpoint, while an event outside that window would not be part of this particular registered endpoint simply because it occurred at a different time.
5. Statistical Methodology
No formal statistical analyses are posted on the ClinicalTrials.gov record. The registered primary endpoint is nevertheless well suited to a straightforward incidence analysis once outcome data become available.
Incidence as a binary safety endpoint
At the participant level, the primary endpoint can be represented as whether each participant experienced at least one investigator-assessed grade 3 or higher treatment-related adverse event during the registered observation period.
The resulting proportion describes the fraction of the analyzed population experiencing at least one qualifying event during the prespecified safety window.
This participant-level approach is different from simply counting every adverse-event occurrence. A single participant can experience multiple adverse events, but the incidence of participants with at least one qualifying event counts that participant once for the binary endpoint.
Confidence interval for the incidence
For a single-arm binary safety endpoint, an interval estimate around the observed proportion is important because the observed percentage is only an estimate of the underlying event probability in the population represented by the study. A binomial confidence interval, with the specific interval construction selected in the statistical analysis plan, would typically accompany the incidence estimate.
The statistical objective is to estimate p in the MIRAS study population over the registered observation window.
Denominator selection
The denominator is especially important in a real-world safety analysis. It should correspond to the analysis population defined for the endpoint rather than automatically assuming that every person ever enrolled is evaluable for every safety observation. The registry record identifies planned enrollment but does not provide a posted analysis population or observed participant counts.
Event-level versus participant-level summaries
The primary endpoint is an incidence measure, which naturally focuses on participants experiencing at least one qualifying event. An event-level table could provide additional clinical detail, but it would answer a different question. For example, counting repeated events from the same participant would measure event burden rather than the incidence of participants with at least one grade 3 or higher TRAE.
6. Statistical Methods Explained
What does "incidence" mean for this endpoint?
Here, incidence is the proportion of participants who experience at least one grade 3 or higher treatment-related adverse event during the specified period. It is therefore a participant-level probability estimate rather than a count of all adverse-event episodes.
Why is this a proportion rather than a hazard ratio?
The registered endpoint is defined as an incidence over a specified treatment and follow-up window, and the study has 1 arm. A hazard ratio requires a comparison of event hazards between groups or another explicitly defined reference process. No such comparator is registered for MIRAS.
Why does the 30-day post-treatment period matter?
The endpoint explicitly extends from the first dose to 30 days after the last dose. This creates a defined observation window. Without that time frame, an incidence estimate would be ambiguous because the opportunity to observe treatment-related toxicity would differ depending on how long participants were followed.
Why is investigator assessment important?
The registry defines the endpoint as treatment-related adverse events assessed by the investigator. Thus, the primary endpoint depends not only on the recorded severity grade but also on the investigator's assessment of whether the event is treatment related.
Why should a confidence interval accompany the incidence?
An observed proportion is subject to sampling variability. A confidence interval communicates how precisely the study estimates the underlying event probability. A narrow interval indicates greater statistical precision than a wide interval, although precision does not determine whether an adverse event is clinically serious or important.
Why is there no treatment-effect estimate?
A single-arm incidence estimate describes what occurred among participants receiving MIRV. It does not provide a randomized estimate of how MIRV changes risk relative to another treatment. Any comparison with an external population would require additional assumptions about comparability that are not part of the registered primary endpoint.
7. Planned Analysis
The registry identifies the primary endpoint as the incidence of grade ≥ 3 treatment-related adverse events (TRAEs), assessed by the investigator from first dose through 30 days after the last dose of MIRV.
For an endpoint of this type, the primary statistical summary would typically consist of the number and proportion of participants experiencing at least one qualifying event, accompanied by an appropriate confidence interval. Additional safety summaries could distinguish individual adverse-event terms, grades, and treatment-relatedness when those data are available.
Primary estimand
The probability that a participant experiences at least one investigator-assessed grade 3 or higher treatment-related adverse event during the registered observation window.
Statistical summary
A participant-level incidence proportion with an appropriate confidence interval would directly describe the registered endpoint.
Study structure
Because the registry describes 1 arm, the primary analysis is naturally descriptive rather than a randomized between-arm hypothesis test.
Current evidence state
Results have not been posted on ClinicalTrials.gov.
8. Interpreting a Future Incidence Estimate
Suppose a future analysis reports an incidence proportion for the primary endpoint. The most direct interpretation would be: among the analyzed MIRAS participants, that percentage experienced at least one investigator-assessed grade 3 or higher treatment-related adverse event during the specified observation window.
That interpretation should remain distinct from several stronger claims. An incidence percentage would not by itself establish comparative safety, because the registry describes only 1 arm. It would also not mean that every participant had the same underlying probability of toxicity, because individual characteristics and treatment exposure can differ within a real-world population.
An observed incidence proportion would summarize the frequency of participants experiencing at least one qualifying grade 3 or higher TRAE during the registered period.
It would not by itself establish a causal comparison with another therapy, quantify the number of adverse-event episodes per participant, or describe the probability of every possible adverse event.
A confidence interval would show the statistical uncertainty around the estimated incidence. Two studies can report similar percentages while having different levels of precision because the amount of information contributing to the estimates can differ.
9. Safety Endpoint: Statistical Interpretation
The MIRAS primary endpoint is a safety endpoint with a severity threshold. That structure makes the endpoint clinically focused: it does not count every adverse event equally. Only events meeting the grade threshold and judged treatment related by the investigator qualify for the registered primary endpoint.
| Component | Statistical role |
|---|---|
| Grade ≥ 3 | Defines the severity threshold for events entering the primary endpoint. |
| Treatment related | Restricts the endpoint to adverse events assessed by the investigator as related to MIRV. |
| First dose | Defines the beginning of the registered observation period. |
| 30 days after last dose | Defines the end of the registered observation period. |
| Participant-level incidence | Captures whether a participant experienced at least one qualifying event. |
This distinction is important when interpreting future safety tables. A participant who experiences multiple qualifying events should not automatically be interpreted as representing multiple participants with toxicity when the primary endpoint is participant-level incidence. Conversely, a participant-level incidence percentage does not convey the complete clinical burden of recurrent or multiple events.
10. Study Timing and Statistical Information Maturity
Study start date
The registry lists 2026-07-20 as the start date for MIRAS.
Not yet recruiting
The ClinicalTrials.gov record currently lists the study status as NOT_YET_RECRUITING.
Primary completion date
The registry lists 2028-09-30 as the primary completion date.
The timing matters statistically because the registered endpoint requires participants to contribute safety information across a defined treatment period and through 30 days after the last dose. Until participant-level outcome data are available, there is no observed incidence estimate to interpret.
11. What Can and Cannot Be Inferred From a Single-Arm Safety Study?
What an incidence estimate can describe
The frequency of participants meeting the registered grade, relatedness, and time-window criteria for the primary safety endpoint.
What it cannot establish alone
Whether MIRV has a higher or lower risk than another treatment in a randomized comparative sense.
What confidence intervals add
Information about statistical precision around the estimated incidence.
What the primary endpoint does not summarize
All aspects of treatment tolerability, because the registered endpoint specifically focuses on grade ≥ 3 treatment-related adverse events.
These distinctions are particularly important for real-world evidence. A real-world study can characterize safety in the population it observes, but a causal comparative interpretation depends on the design, comparator, confounding control, and analysis framework actually used.
12. Limitations
- No randomized comparator: the registry describes 1 arm, so the primary endpoint does not provide a randomized comparison with another therapy.
- No posted results: the ClinicalTrials.gov record does not currently report statistical analysis results for the registered primary endpoint.
- Investigator-assessed relatedness: treatment-relatedness is an assessment component of the endpoint and therefore affects which adverse events qualify.
- Severity threshold: the primary endpoint is restricted to grade 3 or higher events and therefore does not represent the full frequency of adverse events of all grades.
- Defined observation window: the endpoint covers the period from first dose through 30 days after the last dose, so its incidence estimate is specific to that time frame.
- Planned enrollment versus analyzed population: 400 is the planned enrollment; it is not an observed analysis population or an observed number of participants experiencing an event.
- Real-world design: treatment exposure and participant characteristics can vary in real-world settings, which can affect the interpretation and generalizability of observed safety rates.
13. Why This Trial Matters Statistically
MIRAS provides a useful teaching example of how the statistical structure of a clinical trial follows directly from its endpoint and design. Unlike a randomized two-arm efficacy study, the registered primary question is centered on the incidence of a clinically defined safety outcome within a single treatment population.
| Concept | How it appears in MIRAS |
|---|---|
| Single-arm design | The registry lists 1 study arm. |
| Real-world evidence | The study is described as a real-world study of MIRV. |
| Binary safety endpoint | The primary endpoint can be represented at the participant level as whether at least one qualifying TRAE occurred. |
| Incidence proportion | The primary endpoint is the incidence of grade ≥ 3 treatment-related adverse events. |
| Severity classification | Only grade 3 or higher treatment-related adverse events qualify for the registered primary endpoint. |
| Treatment-relatedness | The registry specifies investigator assessment of whether adverse events are treatment related. |
| Time-at-risk definition | The endpoint spans first dose through 30 days after last dose of MIRV. |
| Confidence intervals | An interval around a future incidence estimate would quantify statistical uncertainty. |
| Denominator definition | The appropriate analysis population must be distinguished from planned enrollment. |
| Comparative inference | A single-arm incidence does not provide a randomized treatment effect. |
14. Statistical Concepts in This Trial
The MIRAS design illustrates several fundamental ideas in clinical-trial safety analysis: defining a participant-level endpoint, separating event incidence from event counts, specifying a time window, distinguishing planned enrollment from an evaluable analysis population, and interpreting uncertainty around a single-arm proportion.
Endpoint definition
The statistical endpoint is not simply "adverse events." It specifically requires grade ≥ 3 severity and investigator-assessed treatment relatedness.
Analysis population
Planned enrollment of 400 describes the intended study size, while a future safety analysis would require a defined evaluable population.
Absolute frequency
An incidence proportion provides an absolute description of how frequently participants meet the primary endpoint.
Statistical uncertainty
Confidence intervals help distinguish the observed estimate from the underlying population parameter it is intended to estimate.
15. Sources
- ClinicalTrials.gov: MIRAS (NCT07690332).
Continue through Clinical Biostats
Explore additional clinical trial analyses, statistical tutorials, and practical biostatistical methods.
16. Record Summary
MIRAS is registered as a real-world, single-arm study of mirvetuximab soravtansine in patients with ovarian cancer, fallopian tube cancers, and primary peritoneal cancer with expression of folate receptor α. The planned enrollment is 400 participants, and the primary endpoint is the incidence of grade ≥ 3 treatment-related adverse events assessed by the investigator from first dose through 30 days after the last dose of MIRV.
Statistically, the study is centered on estimating a clinically defined safety incidence rather than comparing randomized treatment effects. The key analytical elements are therefore the definition of the participant-level endpoint, the denominator used for the analysis, the prespecified observation window, and the uncertainty surrounding the resulting incidence estimate. Because the registry currently lists the study as not yet recruiting and does not post statistical analysis results, no observed safety incidence or comparative treatment effect is available for interpretation.