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
VISSIT was a randomized, parallel-group phase 2/3 treatment study evaluating the Pharos Vitesse Neurovascular Stent System with medical therapy versus medical therapy alone in patients with ischemic stroke or transient ischemic attack. The registry reports an enrollment of 125 and a terminated study status.
| Feature | VISSIT |
|---|---|
| Study | VISSIT Intracranial Stent Study for Ischemic Therapy |
| NCT identifier | NCT00816166 |
| Phase | Phase 2/3 |
| Status | Terminated |
| Conditions | Ischemic Stroke; Transient Ischemic Attack |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 125 |
| Number of arms | 2 |
| Lead sponsor | Codman & Shurtleff |
| Sponsor type | Industry |
2. Clinical Question
The central question was whether adding implantation of the Pharos Vitesse Neurovascular Stent System to medical therapy could produce a better 12-month effectiveness outcome than medical therapy alone in the study population.
Population
Patients enrolled in a study of ischemic stroke or transient ischemic attack.
Intervention
Pharos Vitesse Neurovascular Stent System with stent implantation plus medical therapy consisting of aspirin and clopidogrel.
Comparator
Medical therapy consisting of aspirin and clopidogrel.
Primary question
Does the stent-plus-medical-therapy strategy improve the registry-defined 12-month successful outcome compared with medical therapy alone?
3. Trial Design
Pharos Vitesse Neurovascular Stent System
- Stent implantation
- Aspirin
- Clopidogrel
Medical Therapy
- Aspirin
- Clopidogrel
4. Trial Timeline and Status
Study start
The registry lists October 2008 as the study start date.
Primary completion
The registry lists April 2013 as the primary completion date.
Early termination of enrollment
The registry states that early termination of enrollment led to small numbers of subjects analyzed, representing less than 50% of the prospectively planned sample size.
5. Primary Endpoint
| Endpoint | Time frame | Registry definition and analysis |
|---|---|---|
| Successful Outcome: No Stroke or Hard TIA in the Same Territory Within 12 Months | One Year | The primary effectiveness endpoint was a composite of stroke in the same territory, distal to the target lesion, as the presenting event within 12 months of randomization, and hard transient ischemic attack in the same territory, distal to the target lesion, as the presenting event from day 2 through month 12 post-randomization. A subject was deemed a primary endpoint success if neither outcome occurred. The Kaplan-Meier success rate at 12 months post-operatively was calculated with Kaplan-Meier time-to-event methodology. Patients who were successful were censored at the time of last follow-up, while patients who were not successful were censored at the time of the first qualifying event. |
Understanding the composite endpoint
The endpoint is framed as a success outcome: a patient succeeds if neither of the two specified events occurs during the defined observation periods. Statistically, this is different from counting only strokes or only TIAs. Both components can cause a patient to be classified as a primary endpoint failure.
The definition also distinguishes the timing of the two components. Stroke in the same territory is evaluated within 12 months, whereas hard TIA in the same territory is evaluated from day 2 through month 12. Those timing rules are part of the endpoint definition and determine when an event qualifies for the composite.
6. Statistical Methodology
Kaplan-Meier estimation
The registry specifies Kaplan-Meier time-to-event methodology for calculating the 12-month success rate. This is appropriate for an endpoint in which participants may have different amounts of observed follow-up and some participants may reach the end of their available observation period without experiencing a qualifying event.
Here, di represents events at time ti, while ni represents the number at risk immediately before that time.
Success as a time-to-event outcome
Although the endpoint is worded as a successful outcome, the registry's specified analysis treats the occurrence of either qualifying event as the time-to-event event. A patient without such an event remains event-free until the last available follow-up and is then censored.
Successful patient
No qualifying stroke or hard TIA occurs during the applicable period. The time variable is censored at the last follow-up.
Unsuccessful patient
A qualifying stroke or hard TIA occurs. The time variable is censored at the first qualifying event.
Censoring
Censoring is important because the registry does not define a successful patient simply by whether an event was observed before a fixed administrative date. Instead, the specified Kaplan-Meier method uses the available follow-up time. A censored participant contributes information through the time at which follow-up ends, without being counted as having experienced the event.
What the analysis estimates
The registry specifies a Kaplan-Meier success rate at 12 months. In statistical terms, this is a time-to-event estimate based on the probability of remaining free from a qualifying primary endpoint event through the relevant time point.
The estimate is model-free in the sense that Kaplan-Meier estimation does not require a proportional-hazards assumption. It does, however, depend on the validity of the censoring framework and the observed follow-up information.
7. Results
The registry reports that results were posted, but no formal statistical analyses are posted in the record. The primary endpoint is defined and its intended Kaplan-Meier analysis is specified, but the record summarized here does not provide a numerical estimate, confidence interval, or p-value for the primary endpoint.
Primary effectiveness endpoint
Successful Outcome: No Stroke or Hard TIA in the Same Territory Within 12 Months
Time frame: One Year
The registry specifies Kaplan-Meier time-to-event methodology, with censoring at last follow-up for patients without a qualifying event and at the first qualifying event for patients with an event.
For this type of endpoint, the usual statistical presentation would include the Kaplan-Meier estimate at 12 months together with a measure of uncertainty, such as a confidence interval. A treatment-group comparison could additionally be presented using a time-to-event hypothesis test or a model-based effect measure when such methods are prespecified. The record does not provide those numerical statistical results here.
The primary endpoint is designed so that higher success means fewer qualifying events over follow-up. A Kaplan-Meier success estimate therefore describes the estimated proportion remaining free of the specified same-territory stroke or hard TIA event through 12 months.
That estimate does not mean that every participant had exactly the same follow-up, nor does it imply that a censored participant was known to remain event-free indefinitely. Censoring means that the participant's observed information ends at the censoring time.
Because the registry does not report a numerical estimate, confidence interval, or p-value for this endpoint, there is no reported effect size from which to assess the magnitude or statistical precision of a treatment difference. A p-value, if reported, would address evidence against a specified null hypothesis; it would not itself measure the size or clinical importance of an effect.
8. Safety Results
The registry reports serious adverse events by treatment group as affected participants divided by participants at risk.
| Safety measure | Stent Group | Medical Therapy Group |
|---|---|---|
| Serious adverse events | 30/58 | 20/53 |
The affected/at-risk figures correspond to the reported serious-adverse-event counts and denominators. The registry does not provide a formal statistical comparison of these safety figures in the information summarized here.
The numerator identifies participants affected by serious adverse events and the denominator identifies participants at risk for the reported safety assessment. These figures describe observed safety experience within the reported groups.
A difference in observed counts or proportions is not, by itself, a formal statistical test. A comparative safety analysis would normally specify the analysis population, exposure period, event definitions, handling of multiple events, and an appropriate statistical method.
9. Statistical Methods Explained
Why was Kaplan-Meier used for the primary endpoint?
The registry explicitly specifies Kaplan-Meier time-to-event methodology. The method is useful when participants can have different follow-up times and when some participants are censored before experiencing the event. Rather than treating every participant as though exactly the same amount of follow-up were available, Kaplan-Meier estimation uses the timing of events and censoring.
What does "successful outcome" mean statistically?
Success is defined by the absence of either qualifying event. Stroke and hard TIA are therefore treated as components of a composite time-to-event outcome. Once the first qualifying event occurs, the participant is no longer a primary endpoint success.
Why does the registry distinguish day 2 from the rest of the 12-month period?
The endpoint definition specifies different time windows for its two components. Stroke in the same territory is considered within 12 months of randomization, while hard TIA in the same territory is considered from day 2 through month 12. Those rules determine which observed events enter the primary endpoint.
What does censoring at last follow-up mean?
For a participant who remains free of a qualifying event, censoring at last follow-up means the analysis uses the participant's observed event-free time up to that point. It does not treat the participant as having experienced the event after follow-up ended, and it does not establish what happened after the final observation.
What would a 12-month Kaplan-Meier success rate represent?
It would represent the estimated probability of remaining free from the specified qualifying primary endpoint event through 12 months, under the Kaplan-Meier framework. Because censoring is incorporated into the estimator, the result is not simply the percentage of enrolled participants who happened to have no event recorded.
Why is an unreported p-value not a measure of no effect?
The absence of a reported p-value is not itself evidence for or against a treatment difference. A p-value requires a specified statistical test and an observed test statistic. Here, the registry specifies the endpoint and Kaplan-Meier methodology but does not provide a formal statistical analysis with a reported p-value in the record summarized here.
10. Analysis Population and Early Termination
The most important design limitation identified by the registry is early termination of enrollment. The record states that this resulted in small numbers of subjects analyzed, less than 50% of the prospectively planned sample size.
The registry further states that various endpoints, subgroup analyses, and sensitivity analyses were not conducted because the study was not fully enrolled. This is an important distinction between an endpoint being part of a study's design and having enough observed information to support the intended analysis.
Why sample size matters beyond power
Sample size affects more than the probability of detecting a prespecified effect. With fewer observations, estimates can be less stable, confidence intervals can be wider, and random variation can have a greater influence on observed treatment-group differences. Small numbers can also make secondary analyses difficult to interpret because there may be too few events to support reliable estimation.
11. Design Features That Affect Statistical Interpretation
| Feature | VISSIT | Statistical implication |
|---|---|---|
| Randomization | Yes | Provides the framework for comparing outcomes according to randomized treatment assignment. |
| Parallel design | Yes | Participants are compared across concurrent treatment groups rather than within a crossover sequence. |
| Masking | None | The registry describes the study as unmasked, so treatment assignment was not masked. |
| Early termination | Yes | The registry states that enrollment ended early and that fewer than 50% of the prospectively planned sample size was analyzed. |
| Subgroup analyses | Not conducted | The registry states that various subgroup analyses were not conducted because the study was not fully enrolled. |
| Sensitivity analyses | Not conducted | The registry states that various sensitivity analyses were not conducted because the study was not fully enrolled. |
The registry information does not identify a crossover design, factorial design, Bayesian analysis, interim-analysis procedure, multiplicity-adjustment strategy, or stratification factors. Those methods therefore do not form part of the statistical interpretation presented here.
12. Why Censoring Is Central to This Endpoint
The primary endpoint is not simply a binary assessment recorded at exactly 12 months for every participant. The registry explicitly defines the endpoint through time-to-event methodology. That makes the treatment of follow-up especially important.
Event timing
The first qualifying stroke or hard TIA determines when a participant experiences the primary endpoint event.
Follow-up timing
A participant without a qualifying event contributes information until the time of last follow-up.
12-month estimate
The Kaplan-Meier method combines event and censoring times to estimate event-free success through one year.
Interpretation
The estimate depends on the observed follow-up and the assumptions underlying the censoring framework.
This distinction is important in clinical-trial statistics because a participant lost to follow-up is not equivalent to a participant who experiences the endpoint. Kaplan-Meier estimation preserves the information available before censoring rather than automatically classifying the participant as an event.
13. What a Formal Treatment Comparison Would Normally Report
For a randomized trial with a Kaplan-Meier primary endpoint, a complete statistical presentation would typically describe the estimated event-free or success probability at the prespecified time point and quantify uncertainty around that estimate. A comparison between randomized groups could additionally use a time-to-event hypothesis test and, when prespecified, a model-based relative effect estimate.
| Statistical quantity | Purpose |
|---|---|
| Kaplan-Meier success estimate | Describes the estimated probability of remaining free of a qualifying event through the specified time point. |
| Confidence interval | Describes statistical uncertainty around the estimated success probability or treatment effect. |
| Hypothesis test | Assesses the evidence against a prespecified null hypothesis about the treatment comparison. |
| Effect estimate | Quantifies the magnitude and direction of the treatment difference using a specified statistical scale. |
These quantities answer different questions. A confidence interval addresses precision, while a p-value addresses compatibility with a null hypothesis under a specified testing framework. Neither one is a substitute for describing the clinical magnitude of an effect.
14. Interpretation of the Primary Endpoint
The VISSIT primary endpoint is a time-to-event success outcome. The registry defines success as having neither a same-territory stroke within 12 months nor a same-territory hard TIA from day 2 through month 12. The specified analysis is Kaplan-Meier estimation of the 12-month success rate.
A 12-month success rate, considered alone, would not describe the severity of an event, distinguish the clinical consequences of stroke from hard TIA, or establish that one treatment is superior without a corresponding comparative analysis. The composite endpoint combines its qualifying components into one time-to-event outcome.
Because the registry states that fewer than 50% of the prospectively planned sample size was analyzed, the amount of information available for statistical estimation was substantially smaller than originally planned. Precision and the ability to perform additional analyses are therefore important considerations when interpreting the study record.
15. Limitations
- Early termination: the registry states that enrollment ended early and that the number of subjects analyzed was less than 50% of the prospectively planned sample size.
- Small analyzed numbers: the reduced enrollment limits the amount of statistical information available for treatment-effect estimation.
- Unperformed analyses: the registry states that various endpoints, subgroup analyses, and sensitivity analyses were not conducted because the study was not fully enrolled.
- Composite endpoint: the primary endpoint combines same-territory stroke and hard TIA, so a single success estimate does not separately quantify each component.
- Different event windows: the two components use different registry-defined time windows, with stroke assessed within 12 months and hard TIA from day 2 through month 12.
- Unmasked design: the registry describes the trial as having no masking.
- Safety denominator: the reported serious-adverse-event figures use denominators of 58 and 53, which are distinct from the overall enrollment of 125.
- Limited formal statistical reporting: the registry specifies the primary endpoint and Kaplan-Meier methodology but does not provide a formal statistical analysis with a reported estimate, confidence interval, or p-value for the primary endpoint in the record summarized here.
16. Why This Trial Matters Statistically
VISSIT is a useful teaching case because it illustrates an important principle in clinical-trial statistics: the quality of an endpoint definition and the appropriateness of an analysis method do not eliminate the consequences of an under-enrolled study.
| Concept | How it appears in VISSIT |
|---|---|
| Randomization | The study used randomized allocation to two parallel treatment groups. |
| Time-to-event analysis | The primary effectiveness endpoint was analyzed using Kaplan-Meier time-to-event methodology. |
| Composite endpoint | Primary success required absence of both specified same-territory stroke and hard TIA outcomes. |
| Censoring | Successful patients were censored at last follow-up; patients with a qualifying event were censored at the first event. |
| Fixed-time estimation | The registry specified a Kaplan-Meier success rate at 12 months. |
| Safety analysis | Serious adverse events were reported as affected participants divided by participants at risk in each group. |
| Early termination | Enrollment ended early, resulting in less than 50% of the prospectively planned sample size being analyzed. |
| Analysis feasibility | The registry states that various endpoints, subgroup analyses, and sensitivity analyses were not conducted because the study was not fully enrolled. |
The study therefore demonstrates why statistical interpretation should consider the entire analysis chain: the clinical question, endpoint definition, event timing, censoring, analysis method, available sample size, and completeness of the planned analyses.
17. A Practical Statistical Reading of VISSIT
A useful way to read the VISSIT record is to separate what was designed from what was ultimately available for analysis.
Designed question
Compare a stent-plus-medical-therapy strategy with medical therapy alone using a predefined 12-month effectiveness outcome.
Designed analysis
Estimate the primary endpoint using Kaplan-Meier time-to-event methodology.
Observed study context
The registry lists the study as terminated and states that early termination produced fewer than 50% of the prospectively planned sample size.
Analytic consequence
Various endpoints, subgroup analyses, and sensitivity analyses were not conducted because the study was not fully enrolled.
This framework prevents an important statistical mistake: treating a planned analysis as though it were equivalent to a completed and fully powered analysis. The registry's design information establishes how the endpoint was intended to be evaluated, while the termination statement establishes the limitations on the resulting evidence base.
18. Serious Adverse Events: Statistical Context
The safety figures provide a second illustration of the importance of denominators. The registry reports 30 affected participants among 58 at risk in the Stent Group and 20 among 53 at risk in the Medical Therapy Group.
| Group | Affected | At risk | Reported format |
|---|---|---|---|
| Stent Group | 30 | 58 | 30/58 |
| Medical Therapy Group | 20 | 53 | 20/53 |
The denominators show why raw event counts alone are insufficient for comparing groups. A count must be interpreted relative to the number of participants at risk. At the same time, a formal comparative safety conclusion requires more information than these two affected/at-risk figures alone, including the precise safety-analysis framework and treatment exposure context.
19. What the Registry Does and Does Not Establish
| Question | Registry information |
|---|---|
| Was the study randomized? | Yes. |
| Was the design parallel? | Yes. |
| Was the study masked? | No. |
| What was the primary endpoint? | No stroke or hard TIA in the same territory within 12 months, defined as a Kaplan-Meier success outcome. |
| What was the primary time frame? | One Year. |
| What statistical method was specified? | Kaplan-Meier time-to-event methodology. |
| Was the study fully enrolled? | No. The registry states that fewer than 50% of the prospectively planned sample size was analyzed. |
| Were subgroup and sensitivity analyses completed? | The registry states that various subgroup and sensitivity analyses were not conducted because the study was not fully enrolled. |
| Are numerical primary-effect estimates and p-values reported in the summarized record? | No formal statistical analyses were posted in the record summarized here. |
20. Sources
- ClinicalTrials.gov: VISSIT — NCT00816166.
- PubMed: VISSIT publication record.
Continue through Clinical Biostats
Explore the broader collection of biostatistics tutorials, statistical calculators, and clinical trial analyses.
21. Record Summary
VISSIT was a randomized phase 2/3 parallel-group treatment study of the Pharos Vitesse Neurovascular Stent System plus aspirin and clopidogrel versus aspirin and clopidogrel alone in ischemic stroke and transient ischemic attack. Its primary effectiveness endpoint was a one-year composite success outcome requiring no same-territory stroke within 12 months and no same-territory hard TIA from day 2 through month 12. The registry specifies Kaplan-Meier time-to-event methodology, including explicit rules for censoring successful participants at last follow-up and unsuccessful participants at their first qualifying event.
The principal statistical limitation is the study's early termination. The registry states that fewer than 50% of the prospectively planned sample size was analyzed and that various endpoints, subgroup analyses, and sensitivity analyses were not conducted because the study was not fully enrolled. The safety record reports serious adverse events affecting 30 of 58 participants in the Stent Group and 20 of 53 in the Medical Therapy Group.