This page separates reported registry information from statistical interpretation. The ClinicalTrials.gov record reports results for the registered primary endpoints, but the ClinicalTrials.gov record contains no formal statistical-analysis estimates, confidence intervals, or p-values.
1. Trial at a Glance
PATENT-2 was a phase 3 long-term extension study evaluating riociguat (BAY63-2521) in patients with pulmonary arterial hypertension. The registry describes a single-group, unmasked treatment study with an enrollment of 396 participants and two registered primary safety-related endpoints.
| Feature | PATENT-2 |
|---|---|
| Trial name | PATENT-2 |
| Brief title | BAY63-2521:Long-term Extension Study in Patients With Pulmonary Arterial Hypertension |
| Phase | Phase 3 |
| Condition | Hypertension, Pulmonary |
| Allocation | NA |
| Design model | Single-group |
| Masking | None |
| Primary purpose | Treatment |
| Enrollment | 396.0 |
| Arms | 1 |
| Intervention | Riociguat (BAY63-2521) |
| Lead sponsor | Bayer |
| Sponsor type | Industry |
| Status | Completed |
| Study start | 2009-03-12 |
| Primary completion | 2019-08-19 |
| ClinicalTrials.gov | NCT00863681 |
2. Clinical Question
The registry describes PATENT-2 as a long-term extension study of riociguat in patients with pulmonary arterial hypertension. Because the study is a single-group extension without masking and without a randomized comparator within PATENT-2, the principal statistical question is different from that of a conventional parallel-group efficacy trial.
Population
Patients with pulmonary arterial hypertension enrolled in the PATENT-2 long-term extension study.
Intervention
Riociguat (BAY63-2521).
Comparator
PATENT-2 is registered as a single-group study; no concurrent comparator arm is specified in the ClinicalTrials.gov record.
Primary question
How many participants experienced treatment-emergent adverse events, and how many participants died during the prespecified safety-follow-up period?
3. Trial Design
The distinction between single-group extension and randomized controlled trial is central to interpreting PATENT-2. A randomized trial can estimate a between-group treatment effect under its randomization scheme. A single-group extension instead primarily describes outcomes observed during continued treatment and follow-up. Without a concurrent randomized control group in the extension itself, an observed event rate cannot be interpreted as the causal effect of riociguat relative to an untreated or alternative-treatment population.
4. Study Population and Exposure Groups
The ClinicalTrials.gov record reports an overall enrollment of 396.0 participants. The safety information is additionally organized according to prior treatment-group labels, producing three reported groups in the serious-adverse-event summary.
Former Riociguat 1.0- 2.5 mg
- Serious adverse events: 161 affected participants
- At-risk denominator: 231
- Reported as 161/231
Former Riociguat 1.0 - 1.5 mg
- Serious adverse events: 39 affected participants
- At-risk denominator: 56
- Reported as 39/56
Former Placebo
- Serious adverse events: 76 affected participants
- At-risk denominator: 109
- Reported as 76/109
Prior assignment is not a randomized PATENT-2 arm
- PATENT-2 itself is registered as single-group.
- The labels above describe the safety groups reported in the ClinicalTrials.gov record.
- They should not automatically be treated as contemporaneous randomized treatment arms.
This structure is statistically important. The phrase former indicates that these groups are defined by prior treatment history rather than by a new randomized allocation within the extension study. Comparisons among them therefore do not have the same causal interpretation as comparisons generated by randomization.
5. Primary Endpoints
| Registered primary endpoint | Time frame | Registry definition / analysis information |
|---|---|---|
| Number of Participants With Treatment-emergent Adverse Events (TEAE) | From administration of first dose of study medication in PATENT-2 up to 2 days after end of treatment with study medication, up to 10 years and 5 months. | Analyses of drug-related TEAEs were based on the assessment of causal relationship to study medication. |
| Number of Participant With Death | From baseline to end of safety follow-up visit, up to 10 years and 6 months (1 month more than End of study visit) | Analyses of deaths were based on the assessment of causal relationship to study medication. The safety follow-up visit was to be performed 30 days after the last dose of riociguat. |
Both registered primary endpoints are safety-oriented. The first is a binary participant-level event endpoint: whether a participant experienced a treatment-emergent adverse event during the specified observation period. The second records whether a participant died during the specified follow-up period.
6. Results Posted by the Registry
ClinicalTrials.gov reports that results are posted for both registered primary endpoints. However, the ClinicalTrials.gov record contains no formal statistical analyses. Therefore, there is no reported hazard ratio, risk ratio, odds ratio, confidence interval, or p-value available in the ClinicalTrials.gov record for either primary endpoint.
Primary Endpoint 1: Treatment-emergent Adverse Events
Registry result type
Results are posted for this primary endpoint, but no formal statistical-analysis estimate is reported.
The endpoint is defined as the number of participants with treatment-emergent adverse events during the registered treatment-emergent adverse-event time frame.
The important statistical feature is that this endpoint is a participant-level binary outcome. For each participant, the relevant question is whether at least one treatment-emergent adverse event occurred during the specified observation window. A registry count can therefore describe the frequency of participants experiencing TEAEs.
What it does not establish by itself is a causal treatment effect. PATENT-2 is a single-group extension, so a TEAE count does not provide a randomized estimate of how much more or less frequently events would have occurred under an alternative treatment.
Because the ClinicalTrials.gov record contains no effect estimate or confidence interval, there is no precision interval to interpret and no p-value that can be used to assess a formal hypothesis test. A p-value, if one had been reported, would address evidence against a specified null hypothesis; it would not itself measure the magnitude or clinical importance of an adverse-event rate.
Primary Endpoint 2: Death
Registry result type
Results are posted for this primary endpoint, but no formal statistical-analysis estimate is reported.
The endpoint records the number of participants with death from baseline through the end of safety follow-up.
Death is a time-sensitive outcome, but the registered primary endpoint is expressed as the number of participants with death over a defined follow-up period rather than as a registry-reported hazard ratio or survival curve. A descriptive death count can therefore summarize how many participants experienced the event during the registry-defined period.
It does not establish that riociguat caused a particular increase or decrease in mortality. In a single-group extension, there is no randomized concurrent comparator against which an extension-period mortality hazard can be directly contrasted.
No confidence interval or p-value for this endpoint is present in the ClinicalTrials.gov record. Consequently, the available evidence supports description of the posted outcome rather than a formal between-treatment inference.
7. Serious Adverse Events by Reported Group
The ClinicalTrials.gov record reports serious adverse events using the following affected-participant and at-risk counts.
| Reported group | Affected | At risk | Registry value |
|---|---|---|---|
| Former Riociguat 1.0- 2.5 mg | 161 | 231 | 161/231 |
| Former Riociguat 1.0 - 1.5 mg | 39 | 56 | 39/56 |
| Former Placebo | 76 | 109 | 76/109 |
The graphic above displays the reported affected-to-at-risk counts as ratios; it does not introduce a new statistical comparison. The three reported ratios are close in magnitude, but the denominators differ and the groups represent prior-treatment categories rather than newly randomized PATENT-2 arms.
An adverse-event count cannot be interpreted without knowing how many participants were exposed or otherwise at risk. The registry therefore supplies both the number affected and the denominator for each reported group.
Even when two groups have similar affected-to-at-risk proportions, the underlying number of participants and exposure history can differ. Conversely, a numerical difference between two percentages would not automatically establish a treatment effect without accounting for the study design, timing, censoring, and source population.
8. Statistical Methodology
Binary participant-level safety endpoints
The TEAE endpoint is naturally represented as a binary outcome at the participant level: each participant either experienced at least one treatment-emergent adverse event during the specified window or did not. The corresponding descriptive quantity is the number of participants with the event, usually accompanied by the number evaluated or at risk.
A descriptive event proportion is formed from the number of participants with the event relative to the relevant analysis denominator. The ClinicalTrials.gov record reports affected/at-risk counts for serious adverse events but do not supply formal inferential estimates for the registered primary endpoints.
Descriptive safety analysis
For a single-group extension study, descriptive summaries are particularly important. Counts and proportions can characterize how frequently events occurred during treatment exposure and follow-up. These summaries should be tied to the correct safety population and observation period rather than treated as if every participant had identical follow-up.
Time-to-event analysis for death
Death is inherently a time-to-event outcome because both the occurrence of the event and the time at which it occurs can be informative. In a conventional survival analysis, Kaplan-Meier methods can estimate the survivor function and Cox models can estimate relative hazards when a comparison group exists. However, the registry-reported PATENT-2 statistical-analysis data do not provide a Kaplan-Meier estimate, hazard ratio, confidence interval, or p-value.
The survival function represents the probability of remaining free of the event beyond time t. In a single-group study it can describe observed survival, but it does not by itself provide a causal treatment comparison.
Causal interpretation and randomization
Randomization is what makes treatment-group comparisons particularly useful for causal inference in a controlled trial. PATENT-2 is registered as a single-group extension. Consequently, the extension can provide important information about outcomes during longer-term treatment exposure, but its within-study observations should not be interpreted as a randomized comparison between riociguat and placebo.
Censoring and follow-up
For a mortality endpoint, participants may have different amounts of observed follow-up. Survival methods can accommodate right censoring when their assumptions and censoring rules are appropriate. The registry-defined mortality time frame extends from baseline to the end of the safety follow-up visit, up to 10 years and 6 months, with the safety follow-up visit specified as 30 days after the last dose of riociguat.
Causal relationship assessment
The registry specifically states that analyses of drug-related TEAEs and deaths were based on assessment of causal relationship to study medication. This is a clinical classification rather than a statistical adjustment. A determination that an event is considered related to study medication does not turn an uncontrolled event rate into a randomized treatment effect.
9. What the Serious-Adverse-Event Counts Do — and Do Not — Mean
The reported values 161/231, 39/56, and 76/109 describe affected participants relative to the corresponding at-risk denominators for the three reported groups. They provide a direct descriptive summary of serious-adverse-event occurrence.
They do not establish that one prior-treatment group has a statistically or clinically different safety profile from another. PATENT-2 is registered as a single-group study, and the ClinicalTrials.gov record does not provide a formal comparative analysis.
For an event proportion, a confidence interval can quantify uncertainty caused by finite sample size. For a comparative measure, such as a risk ratio or odds ratio, a confidence interval would additionally quantify uncertainty around the between-group effect estimate.
No such interval is reported in the ClinicalTrials.gov record for PATENT-2. It would therefore be inappropriate to create one and present it as a registry-reported result.
A p-value evaluates compatibility with a specified null hypothesis under the chosen statistical model. It does not quantify the magnitude of an adverse-event difference, its clinical importance, or the probability that a treatment caused an event.
Because no formal statistical-analysis p-values are reported here, the appropriate emphasis is on the observed descriptive safety information and the design limitations that govern its interpretation.
10. Statistical Methods Explained
What is a treatment-emergent adverse event?
A treatment-emergent adverse event is an event assessed during the registered period beginning with the first dose of study medication and extending through the specified post-treatment window. In this registry, drug-related TEAE analyses were based on an assessment of causal relationship to study medication.
Why is PATENT-2 not analysed like a randomized parallel-group trial?
The registry identifies the design model as single-group and masking as none. There is therefore no concurrent randomized PATENT-2 control arm in the registry-reported design information. Statistical methods that estimate a treatment effect by directly comparing randomized groups do not have the same role here.
Why are the former-treatment groups still shown?
The serious-adverse-event data are reported separately for Former Riociguat 1.0- 2.5 mg, Former Riociguat 1.0 - 1.5 mg, and Former Placebo. These labels preserve the registry's registry-reported grouping. They are useful descriptively, but they should not be relabeled as randomized PATENT-2 arms because the study itself is registered as single-group.
Why can death require survival analysis rather than a simple percentage?
Death has both an event indicator and an event time. Participants can therefore contribute different amounts of follow-up. Kaplan-Meier methods are designed to estimate survival while accounting for right-censored observations. A simple cumulative proportion can answer a different question and may not use the timing information.
What would a hazard ratio mean if one were reported?
A hazard ratio compares estimated instantaneous event rates between two groups under a time-to-event model. An HR below 1 would indicate a lower estimated instantaneous event rate in the numerator group relative to the reference group. It would not mean that the same percentage of individual participants experienced a reduction in risk.
For PATENT-2, however, the ClinicalTrials.gov record contains no hazard ratio. A hazard ratio should therefore be treated here as a methodological concept, not as a reported PATENT-2 result.
Why does the safety-follow-up definition matter?
The death endpoint is defined from baseline through the end of the safety follow-up visit, up to 10 years and 6 months. The registry further states that the safety follow-up visit was to occur 30 days after the last dose of riociguat. Changing the observation window can change the number of events observed, so the time frame is part of the endpoint definition rather than incidental metadata.
11. Follow-Up and Study Duration
Study start
The PATENT-2 study began on 2009-03-12 according to the ClinicalTrials.gov record.
Primary completion
The registry lists 2019-08-19 as the primary completion date.
Mortality endpoint window
The registered death endpoint is assessed from baseline to the end of safety follow-up, up to 10 years and 6 months, defined as 1 month more than the End of study visit.
Safety follow-up visit
The registry states that the safety follow-up visit was to be performed 30 days after the last dose of riociguat.
The long observation window is one of the most statistically consequential features of an extension study. Longer follow-up can reveal events that would not be observable during a short controlled treatment period. At the same time, participants may have unequal follow-up durations, making time-to-event methods and clearly defined observation windows important for mortality analyses.
12. Design Features That Are Not Supported by the Supplied Data
The ClinicalTrials.gov record does not provide information on several statistical design features commonly reported for randomized confirmatory trials. They should therefore not be inferred from the existence of posted outcomes.
| Design topic | What can be stated from the ClinicalTrials.gov record |
|---|---|
| Non-inferiority margin | Not provided in the ClinicalTrials.gov record. |
| Crossover | Not provided in the ClinicalTrials.gov record. |
| Factorial design | Not provided; the registered design model is single-group. |
| Multiplicity adjustment | Not provided in the ClinicalTrials.gov record. |
| Interim analysis | Not provided in the ClinicalTrials.gov record. |
| Missing-data / imputation method | Not provided in the ClinicalTrials.gov record. |
| Stratification factors | Not provided in the ClinicalTrials.gov record. |
| Bayesian methods | Not provided in the ClinicalTrials.gov record. |
| Formal statistical analyses | No entries are in the posted results. |
13. Interpreting a Single-Group Extension Study
The statistical role of a long-term extension is different from the role of the randomized parent trial. The extension can accumulate substantially more treatment exposure and longer observation, allowing investigators to describe safety and other outcomes over an extended period.
What the design supports
Describing adverse events and deaths observed during the specified long-term treatment and follow-up period.
What the design does not automatically support
A causal estimate of riociguat versus an alternative treatment based solely on the extension cohort.
Why prior treatment matters
The serious-adverse-event groups are defined using prior treatment labels, so treatment history is part of the interpretation.
Why time matters
Long-term observation changes the opportunity for events to occur and can create unequal follow-up across participants.
This distinction is particularly important when interpreting safety. A high cumulative number of adverse events over a long period does not necessarily imply a high short-term event rate, and a lower observed count can reflect shorter exposure rather than a safer treatment. Exposure time, event timing, and the population contributing observations all matter.
14. Safety Interpretation
The serious-adverse-event data posted on ClinicalTrials.gov for PATENT-2 are substantial descriptive information: 161 of 231 participants in the Former Riociguat 1.0- 2.5 mg group, 39 of 56 in the Former Riociguat 1.0 - 1.5 mg group, and 76 of 109 in the Former Placebo group are reported as affected.
These values should be read as observed safety outcomes within the reported groups, not as a randomized efficacy or safety comparison. The denominators identify the populations against which the affected counts are being expressed, but the ClinicalTrials.gov record does not specify a formal statistical test comparing those groups.
15. Results Reporting: What Is Missing From the Formal Analysis Field?
The registry data indicate that results have been posted and that both primary endpoints have results posted for them. Yet no formal statistical analyses were posted.
| Element | Supplied PATENT-2 data |
|---|---|
| Results posted | Yes |
| Outcome measures posted | 26 |
| Primary endpoints | 2 |
| Results posted for TEAE endpoint | Yes |
| Results posted for death endpoint | Yes |
| Formal statistical analyses reported | None |
| Effect estimates reported | None |
| Confidence intervals reported | None |
| P-values reported | None |
This distinction is valuable when reading a clinical-trial registry. A result being posted does not necessarily mean that a formal inferential analysis is available in every extracted results field. For PATENT-2, the ClinicalTrials.gov record supports a descriptive review but do not support reconstruction of a reported hypothesis test or effect estimate.
16. Why This Trial Matters Statistically
PATENT-2 is a useful teaching example because it illustrates a statistical situation that differs from the conventional randomized phase 3 comparison. Its central challenge is not choosing between two treatment effects; it is understanding what can legitimately be inferred from a long-term, single-group extension with safety-focused endpoints.
| Concept | How it appears in PATENT-2 |
|---|---|
| Single-group design | The registered design model is SINGLE_GROUP. |
| Long-term extension | The study is described as a long-term extension study in patients with pulmonary arterial hypertension. |
| Safety endpoint | Number of participants with treatment-emergent adverse events is a registered primary endpoint. |
| Mortality endpoint | Number of participants with death is a second registered primary endpoint. |
| Binary outcomes | The registered TEAE endpoint is a participant-level event indicator. |
| Time-to-event reasoning | Death is assessed over a potentially long follow-up period. |
| Exposure time | Long-term treatment and follow-up make the observation window central to interpretation. |
| Prior-treatment groups | Serious adverse events are reported for former riociguat and former placebo groups. |
| Causal inference | Absence of a randomized concurrent PATENT-2 comparator limits causal treatment comparisons. |
| Statistical reporting | Results are posted, but no formal statistical analyses were posted. |
17. Limitations
- Single-group design: PATENT-2 is registered as a single-group study, so its extension-period observations do not provide the same causal treatment comparison as a randomized controlled trial.
- No formal statistical analyses reported: no formal statistical analyses were posted, so effect estimates, confidence intervals, and p-values cannot be reported as registry-derived results.
- Prior-treatment grouping: the serious-adverse-event categories are labeled according to former treatment groups. These should not automatically be interpreted as randomized PATENT-2 treatment arms.
- Long observation window: the death endpoint can extend up to 10 years and 6 months, making follow-up duration and event timing important to interpretation.
- Exposure differences: participants in a long-term extension may contribute different amounts of treatment exposure and follow-up. A simple cumulative event count does not account for exposure time.
- Safety versus causality: an observed adverse event during treatment is not automatically caused by treatment. The registry separately describes assessment of causal relationship to study medication.
- Limited inferential information: without the formal statistical-analysis results, the ClinicalTrials.gov record does not establish statistical significance or quantify uncertainty around a treatment effect.
- Registry scope: the analysis on this page is constrained to the trial data posted on ClinicalTrials.gov for PATENT-2 and does not import numerical results from external publications.
18. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
PATENT-2 provides descriptive long-term safety information in a single-group extension. The ClinicalTrials.gov record does not provide a formal comparative treatment-effect estimate for either registered primary endpoint.
Clinical interpretation
The registry documents treatment-emergent adverse-event and mortality endpoints and reports serious-adverse-event counts by former-treatment category. Their clinical meaning depends on exposure, follow-up, event definitions, and the distinction between observed association and treatment causality.
The two perspectives should remain separate. Statistical interpretation asks what the study design and available data allow investigators to estimate. Clinical interpretation asks what the observed events mean in the context of treatment exposure and patient follow-up. Neither perspective should turn a descriptive single-group result into a randomized treatment comparison that the design did not generate.
19. Primary Endpoint Analysis Framework
| Endpoint | Data structure | Typical analytical framework | What PATENT-2 reported here |
|---|---|---|---|
| Number of Participants With TEAE | Binary participant-level event | Descriptive counts and proportions; comparative methods require an appropriate reference population | Endpoint definition and posted-result status; no formal statistical-analysis estimate reported |
| Number of Participant With Death | Time-to-event / event count | Kaplan-Meier or other survival methods when timing and censoring information are available | Endpoint definition and posted-result status; no formal statistical-analysis estimate reported |
This table illustrates why statistical methodology must follow the endpoint and design. A binary adverse-event endpoint and a mortality endpoint may both be expressed as participant counts, but death also has an event time that can be analysed directly. At the same time, neither endpoint automatically produces a causal treatment effect in a single-group extension.
20. Statistical Reasoning About Long-Term Safety
Long-term safety analysis has a fundamental denominator problem: participants can be observed for different lengths of time. Suppose two groups each contain the same number of participants with serious adverse events. If one group contributes substantially more treatment exposure, the interpretation of the crude cumulative count is different from that of the group with less exposure.
The registry-reported PATENT-2 data partly address this issue by providing affected / at-risk counts for the reported serious-adverse-event groups. They do not, however, provide the exposure-time information needed to construct an incidence-rate analysis from the ClinicalTrials.gov record.
This framework is distinct from a cumulative proportion. It becomes particularly relevant when participants contribute different amounts of follow-up. No exposure-time denominator is reported here, so no incidence rate should be attributed to PATENT-2 on the basis of the ClinicalTrials.gov record.
This is also why a long-term extension should not be reduced to a single adverse-event percentage. The statistical meaning of a safety outcome depends on the event definition, treatment exposure, observation window, censoring, competing events where relevant, and analysis population.
21. Sources
- ClinicalTrials.gov: PATENT-2, NCT00863681.
- Linked PubMed record: PMID 27067479.
- Linked PubMed record: PMID 27263466.
- Linked PubMed record: PMID 34353714.
- Linked PubMed record: PMID 27457511.
- Linked PubMed record: PMID 27162632.
Continue through the Clinical Biostats statistical library
Explore tutorials on survival analysis, binary safety endpoints, clinical-trial design, and interpretation of long-term follow-up, then apply the concepts with statistical calculators.
22. Record Summary
PATENT-2 is a phase 3, single-group, unmasked long-term extension study of riociguat in patients with pulmonary arterial hypertension, with enrollment of 396.0 and two registered primary endpoints: the number of participants with treatment-emergent adverse events and the number of participants with death. The registry reports serious-adverse-event counts for Former Riociguat 1.0- 2.5 mg, Former Riociguat 1.0 - 1.5 mg, and Former Placebo groups, but the registry-reported statistical-analysis field contains no formal comparative estimates, confidence intervals, or p-values.
The key statistical lesson is therefore one of design-aware interpretation. A long-term extension can provide valuable descriptive information about safety and mortality during extended follow-up, but those observations should not be converted into randomized treatment effects when the extension is registered as single-group. The appropriate analysis depends on the endpoint, exposure time, censoring, prior-treatment structure, and the inferential information actually reported by the registry.