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
MDP 301 was a randomized, double-masked, parallel phase 3 prevention trial evaluating PRO 2000/5 gel against placebo in a population enrolled for assessment of vaginally acquired HIV infection. The registry reports an enrollment of 9404 participants and identifies HIV infection and serious clinical or laboratory adverse events as the two primary endpoints, both assessed through 12 months.
| Feature | MDP 301 |
|---|---|
| Phase | Phase 3 |
| Status | Completed |
| Primary purpose | Prevention |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Double |
| Enrollment | 9404 |
| Number of arms | 2 |
| Interventions | PRO 2000/5 (drug); Placebo (drug) |
| Lead sponsor | Endo USA Inc., a Keenova Therapeutics Company |
| Sponsor type | Industry |
| Conditions | HIV Infections; Gonorrhea; Chlamydial Infections; Genital Herpes |
| ClinicalTrials.gov | NCT00262106 |
2. Clinical Question
The central statistical question was whether assignment to PRO 2000/5 rather than placebo affected the occurrence of confirmed HIV infection before or at the 12 month time point among participants confirmed to be HIV negative at enrollment. The registry also designated severe or life-threatening clinical or laboratory adverse events as a co-primary safety endpoint at 12 months.
Population
Participants enrolled in the MDP 301 prevention trial who were confirmed to be HIV negative at enrollment for the primary HIV acquisition endpoint.
Intervention
PRO 2000/5, identified in the registry as a drug intervention.
Comparator
Placebo, identified in the registry as the comparator drug intervention.
Primary questions
How frequently was confirmed HIV infection acquired by 12 months, and how frequently did Grade 3 or Grade 4 clinical or laboratory adverse events occur by 12 months?
3. Trial Design
The registry describes MDP 301 as a randomized, double-masked, parallel phase 3 trial with a primary purpose of prevention. A parallel design means that participants are assigned to one study arm and remain associated with that randomized comparison rather than moving between treatment groups as part of the core design.
PRO 2000/5
- Drug intervention
- Randomized parallel-group assignment
- Double-masked trial design
- Primary HIV endpoint assessed through 12 months
Placebo
- Drug comparator
- Randomized parallel-group assignment
- Double-masked trial design
- Primary HIV endpoint assessed through 12 months
4. Trial Timeline
Study start
The ClinicalTrials.gov record lists October 2005 as the study start date.
Primary completion
The registry lists August 2008 as the primary completion date.
5. Endpoints
The registry identifies two primary endpoints. Both have a 12-month time frame, but they measure fundamentally different types of outcomes: acquisition of an infectious disease and occurrence of severe or life-threatening adverse events.
| Endpoint | Registry definition | Time frame | Statistical outcome type |
|---|---|---|---|
| Acquisition of HIV infection | Acquisition of HIV infection before or at the 12 month time point, confirmed in a central laboratory, in participants confirmed to be HIV negative at enrollment | 12 months | Binary event by a fixed time point, with potential time-to-event structure if dates of infection are available |
| Grade 3 or 4 adverse event | Grade 3 (severe) or 4 (life-threatening) clinical or laboratory adverse event confirmed on examination or repeat testing, respectively | 12 months | Binary safety event by a fixed time point, with potential repeated-event or time-to-first-event structure depending on the underlying data |
6. Planned Analysis
The ClinicalTrials.gov record does not contain posted statistical analyses or numerical outcome results for the primary endpoints. Accordingly, the appropriate statistical discussion is centered on how the registered endpoints would ordinarily be analyzed rather than on an observed treatment effect.
HIV acquisition by 12 months
The primary HIV endpoint can be represented as whether each eligible participant experienced confirmed HIV acquisition before or at 12 months. A straightforward fixed-time analysis would compare the proportion of participants with confirmed infection between the PRO 2000/5 and placebo groups.
A between-group analysis could report an absolute risk difference, a relative risk, or another prespecified effect measure with an appropriate confidence interval. The exact choice should be determined from the statistical analysis plan rather than inferred from the registry endpoint definition alone.
If the exact timing of HIV acquisition is available and censoring before 12 months is material, a time-to-event analysis can use the infection date rather than reducing all observations to a simple 12-month indicator. Kaplan-Meier methods or another appropriate survival-analysis framework can then account for differing follow-up times. The registry endpoint itself, however, is stated as acquisition before or at the 12 month time point.
Grade 3 or 4 clinical or laboratory adverse events
The safety endpoint can similarly be analyzed as the occurrence of at least one qualifying Grade 3 or Grade 4 event during the 12-month assessment period. A basic comparison would summarize the number and percentage of participants experiencing such an event in each randomized arm.
Risk difference and relative risk are natural descriptive measures for a binary safety endpoint. A time-to-first-event analysis could also be appropriate if event dates and censoring information are available and if that approach was prespecified.
7. Statistical Methodology
Randomized comparison
The primary comparison in a randomized parallel-group prevention trial is the difference between outcomes under the two assigned interventions. Randomization provides the structural basis for causal interpretation because treatment assignment is determined independently of the participant's eventual outcome.
For the HIV endpoint, the most direct estimand is the difference in the probability of confirmed infection by 12 months between participants assigned to PRO 2000/5 and those assigned to placebo. For the safety endpoint, the analogous estimand concerns the probability of experiencing at least one qualifying Grade 3 or Grade 4 event during the same period.
Absolute risk difference
An absolute risk difference is expressed in percentage points. For a prevention endpoint such as HIV acquisition, a negative value would indicate fewer observed infections in the intervention group, while a positive value would indicate more.
Absolute measures are especially useful for prevention trials because they translate a relative comparison into the actual difference in event probability. They also make the baseline event frequency visible, which a relative measure alone can obscure.
Relative risk
A relative risk below 1 would indicate a lower observed event probability in the PRO 2000/5 group, whereas a relative risk above 1 would indicate a higher event probability.
The relative risk and absolute risk difference answer different questions. A small relative difference can correspond to a very small absolute difference when the event is uncommon, while the same relative effect can correspond to a larger absolute difference when baseline risk is higher.
Time-to-event analysis
Because HIV acquisition occurs at a point in time rather than only as a status observed at the end of follow-up, the underlying data can contain more information than a simple 12-month yes/no indicator. If dates of infection and censoring are available, survival analysis can incorporate the timing of infection and differing observation periods.
Here, di represents the number of events at time ti, while ni is the number at risk immediately before that time. The same mathematical framework can describe the probability of remaining event-free over time.
Hazard ratios
A Cox proportional-hazards model could be used if a time-to-event analysis were prespecified and the necessary event and censoring dates were available. The resulting hazard ratio would compare the estimated instantaneous event rates between randomized groups under the model.
A hazard ratio is not the same as a risk ratio and does not directly represent the percentage of participants who avoid infection. Its interpretation also depends on the proportional-hazards assumption when a conventional Cox model is used.
Confidence intervals
A confidence interval should accompany the primary treatment-effect estimate. For an absolute risk difference, it quantifies uncertainty around the estimated difference in event probabilities. For a relative risk or hazard ratio, it quantifies uncertainty on the corresponding relative-effect scale.
A confidence interval that includes the null value does not by itself establish that the intervention and placebo are identical. It indicates that the observed data are compatible with a range of effects that includes the null under the statistical framework used. Conversely, exclusion of the null is evidence against the null hypothesis at the corresponding confidence level, but it does not by itself establish clinical importance.
8. Statistical Methods Explained
Why is a fixed 12-month endpoint useful?
A common follow-up horizon gives every primary analysis a clearly defined time window. The registry's HIV endpoint specifically asks whether confirmed infection occurred before or at 12 months, which creates a common reference point for comparing randomized groups.
Why might a time-to-event analysis be preferable to a simple percentage?
A fixed-time percentage treats all participants who remain event-free at 12 months similarly, even though some participants may have been observed for shorter periods because of censoring or loss to follow-up. A time-to-event analysis can use the timing of events and censoring information more efficiently, provided those data are available and the method is prespecified.
What does an absolute risk difference tell us?
The risk difference describes the difference in event probability between the randomized groups. It is directly interpretable in percentage points and can be translated into the number of additional or fewer events expected per specified number of participants, provided the underlying event risks are known.
What does a relative risk tell us?
The relative risk compares the probability of an event in the PRO 2000/5 group with the probability in the placebo group. It describes relative rather than absolute change, so it should generally be interpreted alongside the underlying event risks.
Why are confidence intervals important?
An effect estimate is only one summary of the observed data. The confidence interval communicates how precisely that estimate has been determined under the statistical model and sampling framework. Wide intervals indicate greater uncertainty, while narrow intervals indicate greater precision.
Why must the HIV endpoint be confirmed centrally?
The registry specifies central laboratory confirmation as part of the primary endpoint definition. This makes confirmation method part of the endpoint itself rather than a separate optional quality-control feature. Consistent confirmation helps ensure that participants are classified according to the same endpoint definition across randomized groups.
Why does the Grade 3 or 4 safety definition matter?
Safety analyses can change substantially depending on whether they count every adverse event, only the first event, or events above a specified severity threshold. MDP 301 defines the primary safety endpoint specifically as Grade 3 or Grade 4 clinical or laboratory adverse events, with confirmation by examination or repeat testing, respectively. That definition determines which events qualify for the primary safety comparison.
9. Interpreting the Two Primary Endpoints Together
The two primary endpoints represent complementary dimensions of a prevention trial. The HIV endpoint addresses the intended preventive outcome, while the Grade 3 or Grade 4 endpoint addresses severe or life-threatening clinical and laboratory safety outcomes.
Efficacy endpoint
Confirmed acquisition of HIV before or at 12 months among participants confirmed HIV negative at enrollment.
Safety endpoint
Grade 3 or Grade 4 clinical or laboratory adverse event confirmed on examination or repeat testing.
Same time frame
Both registered primary endpoints use a 12-month time frame.
Different estimands
The endpoints concern different event types and therefore should not be collapsed into a single numerical outcome.
A prevention intervention can only be interpreted comprehensively by considering both outcomes. A lower HIV acquisition rate would address the preventive objective, while the safety endpoint provides information about severe or life-threatening clinical and laboratory events. The statistical evidence for one endpoint does not automatically establish the result for the other.
10. Multiplicity and Two Primary Endpoints
MDP 301 has two registered primary endpoints. When a trial has more than one primary endpoint, the statistical interpretation depends on the prespecified hypothesis-testing strategy. The analysis plan needs to define whether success requires evidence on both endpoints, either endpoint, or another specified combination, and how the overall type I error rate is controlled.
| Issue | Statistical question |
|---|---|
| Multiple primary endpoints | How should the overall type I error rate be controlled across the two primary hypotheses? |
| Success criterion | Does the protocol require both endpoints to support the intervention, either endpoint, or another prespecified rule? |
| Confidence intervals | Are the reported intervals marginal confidence intervals or adjusted for the multiplicity structure? |
| Interpretation | Are efficacy and safety conclusions evaluated separately or within a combined decision framework? |
The registry information identifies the two primary endpoints but does not provide a posted statistical analysis describing the multiplicity procedure. Consequently, the existence of two primary endpoints is a design fact, while the precise confirmatory error-control strategy would ordinarily come from the protocol or statistical analysis plan.
11. Analysis Populations and Eligibility for the HIV Endpoint
The wording of the HIV endpoint contains an important population restriction: participants must be confirmed to be HIV negative at enrollment. That condition is part of the endpoint definition and prevents participants already infected at enrollment from being counted as incident HIV acquisitions during follow-up.
| Population feature | Statistical significance |
|---|---|
| Confirmed HIV negative at enrollment | Defines eligibility for the primary HIV acquisition endpoint. |
| Randomized assignment | Defines the intervention comparison in the parallel trial. |
| 12-month assessment | Defines the primary endpoint time window. |
| Central laboratory confirmation | Defines how HIV acquisition is confirmed for the primary endpoint. |
This distinction is important because a prevention trial is concerned with incident infection after enrollment. Baseline infection status therefore affects whether an individual can contribute to the primary HIV acquisition analysis.
12. Missing Data and Censoring
The registry information does not provide a posted missing-data or censoring strategy. These issues nevertheless matter for interpreting a 12-month prevention endpoint.
If a participant's HIV status is unknown at 12 months, the analysis must specify how that participant is handled. A complete-case approach, a censoring-based survival approach, or another prespecified method can produce different assumptions and potentially different estimates.
For time-to-event analysis, censoring is appropriate only under assumptions about why follow-up ends. Informative loss to follow-up can bias the estimated event probability if the likelihood of leaving observation is related to subsequent HIV acquisition or safety outcomes in a way not accounted for by the analysis.
13. Blinding and Bias Control
The trial is registered as double-masked. Blinding is particularly relevant when participant behavior, clinical assessment, laboratory testing, or reporting could otherwise differ according to knowledge of treatment assignment.
For the primary HIV endpoint, central laboratory confirmation provides an additional standardized endpoint definition. For the safety endpoint, the registry specifies confirmation on examination or repeat testing, respectively, for the clinical or laboratory adverse-event components.
Randomization
Reduces systematic differences in treatment assignment and establishes the basis for the randomized comparison.
Double masking
Reduces the opportunity for knowledge of assignment to influence participant or investigator behavior and assessment.
Central confirmation
Applies a standardized confirmation mechanism to the HIV endpoint.
Repeat testing
Provides confirmation for the laboratory component of the registered severe or life-threatening safety endpoint.
14. What a Primary HIV Effect Estimate Would Mean
If the analysis reported a risk difference for HIV acquisition, that value would describe the estimated difference in the probability of confirmed HIV infection by 12 months between participants assigned to PRO 2000/5 and those assigned to placebo.
If the analysis reported a relative risk, it would describe the ratio of the 12-month HIV acquisition probability in the PRO 2000/5 group to that in the placebo group. A value below 1 would correspond to a lower observed event probability in the PRO 2000/5 group.
If a hazard ratio were reported from a Cox model, it would compare estimated instantaneous HIV acquisition rates over follow-up under the proportional-hazards model. It would not mean that the same percentage reduction applied to every participant or that the hazard ratio was equivalent to the 12-month risk ratio.
A p-value would quantify the compatibility of the observed data with a specified null hypothesis under the chosen statistical model. It would not measure the size, clinical importance, or practical usefulness of an effect.
15. Safety Endpoint Interpretation
The registered safety endpoint is narrower than overall adverse-event reporting. It specifically targets Grade 3 (severe) or Grade 4 (life-threatening) clinical or laboratory adverse events and requires confirmation on examination or repeat testing, respectively.
That distinction matters statistically because an analysis of all adverse events would answer a different question from the registered primary safety endpoint. Similarly, the number of adverse-event episodes is not automatically equivalent to the number of participants experiencing at least one qualifying event.
| Question | Primary safety endpoint implication |
|---|---|
| Which severity levels qualify? | Grade 3 and Grade 4. |
| Which event types qualify? | Clinical or laboratory adverse events. |
| How are clinical events confirmed? | On examination. |
| How are laboratory events confirmed? | By repeat testing. |
| What is the assessment window? | 12 months. |
Because arm-specific safety event counts are not reported in the registry information, a treatment-group safety rate cannot be calculated from the record.
16. Limitations
- No posted statistical analyses: the ClinicalTrials.gov record does not provide numerical effect estimates, confidence intervals, p-values, or arm-specific event counts for the registered primary endpoints.
- No arm-specific enrollment counts: the registry reports total enrollment of 9404 and two arms, but does not report the number assigned to each arm in the information available here.
- No posted analysis plan: the registry information does not specify the exact hypothesis-testing framework, effect measure, multiplicity procedure, or confidence-interval construction used for the primary endpoints.
- Fixed-time versus time-to-event analysis: HIV acquisition is defined through 12 months, but the optimal statistical representation depends on the availability of event dates, censoring information, and the prespecified analysis method.
- Missing follow-up: the treatment of participants without complete 12-month outcome information can affect fixed-time estimates and should be prespecified.
- Endpoint definition: the HIV endpoint is restricted to participants confirmed HIV negative at enrollment and requires central laboratory confirmation, so broader measures of HIV status would not necessarily correspond to the registered primary endpoint.
- Safety interpretation: the Grade 3 or Grade 4 endpoint is not equivalent to all adverse events, and participant-level event rates should not be inferred from event counts without knowing the underlying analysis population.
- Generalizability: the trial includes the population enrolled under its specific eligibility criteria and study procedures; the ClinicalTrials.gov record does not describe the full baseline characteristics needed for a detailed assessment of external validity.
17. Why This Trial Matters Statistically
MDP 301 is a useful teaching example because it illustrates how prevention trials translate a clinical question into a precisely defined event endpoint. The statistical challenge is not simply to compare two percentages; it is to define who is eligible for the endpoint, establish how infection is confirmed, preserve the randomized comparison, account for incomplete follow-up, and distinguish efficacy from severe safety outcomes.
| Concept | How it appears in MDP 301 |
|---|---|
| Randomization | Participants were randomly allocated to two parallel study arms. |
| Blinding | The trial is registered as double-masked. |
| Prevention estimand | The primary HIV endpoint concerns confirmed acquisition through 12 months. |
| Eligibility for incident events | Participants must be confirmed HIV negative at enrollment for the primary HIV endpoint. |
| Laboratory confirmation | HIV acquisition is confirmed in a central laboratory. |
| Binary risk comparison | The 12-month HIV endpoint can naturally be summarized as the proportion with confirmed acquisition. |
| Time-to-event methods | If event timing and censoring are available, survival methods can incorporate when HIV acquisition occurs. |
| Safety endpoint definition | Grade 3 or Grade 4 clinical or laboratory adverse events form the registered primary safety endpoint. |
| Multiple primary endpoints | HIV acquisition and severe or life-threatening adverse events are both designated primary endpoints. |
| Missing data | Incomplete follow-up requires an explicit statistical strategy rather than automatic classification as event-free. |
The trial therefore connects several fundamental biostatistical ideas: randomized comparisons, fixed-time risk, time-to-event analysis, censoring, endpoint confirmation, confidence intervals, hypothesis testing, multiplicity, and safety-event definitions.
18. Clinical Interpretation vs Statistical Interpretation
Statistical interpretation
The registered design defines a randomized two-arm comparison with two primary 12-month endpoints. The appropriate statistical analysis would compare the randomized groups using an effect measure suited to the endpoint and prespecified analysis population.
Clinical interpretation
The clinical question concerns prevention of vaginally acquired HIV infection while monitoring severe or life-threatening clinical and laboratory adverse events. The efficacy and safety endpoints address different aspects of the intervention's clinical profile.
Neither endpoint should be interpreted in isolation. A prevention analysis needs an explicit denominator and follow-up window, while a safety analysis needs a clear definition of event severity and confirmation. The same randomized comparison can therefore generate distinct efficacy and safety estimands.
19. Key Statistical Takeaways
| Question | Answer |
|---|---|
| What was randomized? | Participants were allocated between PRO 2000/5 and placebo in a parallel-group design. |
| How large was the trial? | The registry reports enrollment of 9404. |
| What was the main efficacy endpoint? | Confirmed acquisition of HIV before or at 12 months among participants confirmed HIV negative at enrollment. |
| What was the main safety endpoint? | Grade 3 or Grade 4 clinical or laboratory adverse events, confirmed on examination or repeat testing, respectively. |
| What was the endpoint time frame? | 12 months for both registered primary endpoints. |
| Were numerical primary results posted? | No formal statistical analyses were posted to ClinicalTrials.gov. |
| What is the natural efficacy effect measure? | A 12-month risk difference or relative risk can directly describe the fixed-time endpoint; a time-to-event measure may also be appropriate when event timing is available. |
| What statistical issue is especially important? | The two primary endpoints require a prespecified strategy for interpreting multiple primary hypotheses and controlling type I error. |
20. Sources
- ClinicalTrials.gov: NCT00262106 — MDP 301.
Continue through the Clinical Biostats trial library
Explore additional clinical trial analyses and statistical methods across the Clinical Biostats site.
21. Record Summary
MDP 301 was a randomized, double-masked, parallel phase 3 prevention trial with 9404 enrolled participants and two study arms, PRO 2000/5 and placebo. The registry identifies two primary endpoints, both assessed over 12 months: confirmed acquisition of HIV among participants confirmed HIV negative at enrollment, and Grade 3 or Grade 4 clinical or laboratory adverse events confirmed on examination or repeat testing. The statistical structure therefore centers on randomized-group comparison of a defined prevention event and a defined severe safety event, with careful attention to endpoint eligibility, follow-up, censoring, effect measures, confidence intervals, and multiplicity. The ClinicalTrials.gov record does not contain posted statistical analyses or numerical results for these primary endpoints.