A Placebo- and Active-Controlled Study of Preladenant in Early Parkinson's Disease (PD) (P05664): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Placebo- and Active-Controlled Study of Preladenant in Early Parkinson's Disease (PD) (P05664) (NCT01155479), a randomized phase 3 study. The primary outcome is Change From Baseline in the Sum of Unified Parkinson's Disease Rating Scale Parts 2 and 3 Scores (UPDRS2+3).

2. Clinical Question

The record describes a two-part, one-year study. Part 1 was intended to evaluate efficacy during the first 26 weeks; Part 2 was intended to evaluate safety and tolerability during the second 26 weeks. The posted primary endpoint concerns Part 1 at Week 26, not the full one-year treatment period.

The registry identifies Parkinson disease as the condition and describes preladenant as an adenosine type 2a receptor antagonist. The clinical question is framed as a comparison of changes in a clinician-rated measure, rather than as a direct measure of an individual participant’s experience or a long-term disease outcome. The results therefore address the stated scale and time point; other questions require other outcomes and follow-up.

3. Trial Design

This was a randomized, parallel-group Phase 3 study with quadruple masking. The official title describes it as double-blind and double-dummy, with placebo and active controls. The registry lists Merck Sharp & Dohme LLC as sponsor. The recorded study start was 2010-07-06, and the primary completion date was 2013-07-16. The current status is terminated; the reason for termination is not reported in the registry record.

The study had two 26-week parts. In Part 1, participants were assigned to preladenant 2 mg, 5 mg, or 10 mg, placebo, or rasagiline 1 mg. The placebo and active-control design allowed comparisons with both placebo and an active drug. Matching placebo treatments were described for the preladenant and rasagiline regimens. Part 2 continued treatment for another 26 weeks in the preladenant and rasagiline groups. Participants assigned to placebo in Part 1 received preladenant 5 mg twice daily in Part 2.

The primary outcome was assessed at baseline and Week 26, making it a Part 1 efficacy comparison. The stated primary analysis population was the Full Analysis Set, defined in the record as all randomized and treated participants with at least one post-baseline value. The registry reports different group counts for endpoint analyses and safety summaries; these are analysis-specific populations, and the record does not establish that the denominators should be treated as interchangeable.

The original statistical plan, including the study’s initial sample-size justification and any planned interim analyses, is not reported in the registry record. The record also does not state why the trial was terminated. Those omissions limit what can be concluded about planning decisions and the context of the final results.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: In Part 1, participants were randomized to one of five treatment groups and treated for 26 weeks. In Part 2, which was conducted for an additional 26 weeks, participants continued taking the same study treatment from Part 1, except for placebo participants who were re-assigned to receive preladenant 5 mg twice daily.

Part I

Reasons for not completing this period:

Part II

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Full Analysis Set (FAS): All randomized and treated participants with at least one post-baseline value.

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

Population: All Participants as Randomized

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in the Sum of Unified Parkinson's Disease Rating Scale Parts 2 and 3 Scores (UPDRS2+3), measured over Baseline and Week 26.

The primary endpoint was change from baseline to Week 26 in the sum of Unified Parkinson’s Disease Rating Scale Parts 2 and 3. Part 2 assesses activities of daily living, and Part 3 is the motor examination. The registry describes the combined scale as ranging from 0 to 160, with a higher score indicating worse condition. It is a clinician-based measure drawing on interview and examination, used to capture motor impairment and disability.

Combining the two parts gives the analysis a single continuous outcome that reflects both daily activities and motor examination. Change from baseline compares each group’s Week 26 status with its starting measurement; the treatment comparison then considers the difference between groups’ estimated changes. In this trial, the primary question was whether at least the 10 mg twice-daily dose was superior to placebo on this outcome.

Interpretation depends on both the direction and magnitude of the change and the uncertainty around the between-group contrast. Because higher scores indicate worse condition, a more negative change represents movement toward a lower score within the scale. However, the comparison is not a stand-alone description of every aspect of Parkinson’s disease, and the registry does not report a threshold for a clinically important difference. The endpoint is evidence about the specified score and time point, not a complete account of symptoms, safety, or longer-term outcomes.

7. Results Posted in the Registry

• Preladenant 2 mg (Part 1): mean 0.30 versus -2.20 on control; difference 2.50 (95% CI 0.78 to 4.22), standardized effect d = 0.29; z-test p = 0.004.
• Preladenant 5 mg (Part 1): mean -1.00 versus -2.20 on control; difference 1.20 (95% CI -0.48 to 2.88), standardized effect d = 0.14; z-test p = 0.161.
• Preladenant 10 mg (Part 1): mean -1.80 versus -2.20 on control; difference 0.40 (95% CI -1.29 to 2.09), standardized effect d = 0.05; z-test p = 0.643.
• Rasagiline (Part 1): mean -1.90 versus -2.20 on control; difference 0.30 (95% CI -1.39 to 1.99), standardized effect d = 0.04; z-test p = 0.728.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Percentage of Responders (Participants With a ≥20% Improvement in UPDRS2+3)

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Preladenant 2 mg (Part 1): 26.2% (51/195) versus 35.4% (69/195) on control; risk difference -9.2 percentage points (95% CI -18.3 to -0.1), risk ratio 0.74 (0.55–1.00); unadjusted two-proportion z-test p = 0.048.
  • Preladenant 5 mg (Part 1): 29.7% (60/202) versus 35.4% (69/195) on control; risk difference -5.7 percentage points (95% CI -14.9 to 3.5), risk ratio 0.84 (0.63–1.12); unadjusted two-proportion z-test p = 0.227.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Preladenant 2 mg (Part 1) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Preladenant 5 mg (Part 1) and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline in the UPDRS Part 2 Score (Activities of Daily Living [ADL])

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • Preladenant 2 mg (Part 1): mean 0.30 versus -0.40 on control; difference 0.70 (95% CI 0.10 to 1.30), standardized effect d = 0.23; z-test p = 0.021.
  • Preladenant 5 mg (Part 1): mean 0.10 versus -0.40 on control; difference 0.50 (95% CI -0.08 to 1.08), standardized effect d = 0.17; z-test p = 0.092.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Preladenant 2 mg (Part 1) than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Preladenant 5 mg (Part 1) and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcomes add two kinds of information: a responder definition based on at least 20% improvement in the combined UPDRS Parts 2 and 3 score, and change in the Part 2 activities-of-daily-living score. The responder outcome presents the primary scale as a thresholded yes-or-no result; the Part 2 outcome narrows attention to daily activities rather than combining daily activities and motor examination.

Across these outcomes, the pattern does not provide consistent, clear corroboration of the primary endpoint hypothesis. Some unadjusted recalculated comparisons have intervals that exclude zero, while the sponsor-reported responder contrasts have intervals that include zero. The Part 2 sponsor analysis is mixed across the reported preladenant comparisons. Because the sponsor analyses differ from the recalculations, they are the appropriate reference for interpreting the trial’s reported tests.

These secondary outcomes may help describe whether the primary finding is reflected in a related threshold or a component scale, but they cannot independently establish the primary claim. They are also not interchangeable: a responder percentage and a mean change answer different statistical questions and may respond differently to the distribution of participant outcomes.

Unadjusted secondary results carry less weight when multiple outcomes and comparisons are examined. If many tests are considered without a reported multiplicity strategy, some apparently notable results can arise by chance. The registry does not report whether multiplicity adjustments were planned or applied. The secondary findings should therefore be treated as contextual evidence, not as a replacement for the sponsor-reported primary analysis or as a definitive set of confirmatory tests.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Percentage of Responders (Participants With a ≥20% Improvement in UPDRS2+3)
• Change From Baseline in the UPDRS Part 2 Score (Activities of Daily Living [ADL])

9. Statistical Interpretation

10. Statistical Methods Explained

What does the primary model estimate?

The registry says that change from baseline was analyzed using constrained longitudinal analysis, or cLDA. It lists treatment, time, strata, and treatment-by-time interaction as fixed effects, with participant as a random effect. In general, a longitudinal model uses measurements across time rather than treating each time point as unrelated. A treatment-by-time interaction allows the pattern over time to differ among treatment groups. A participant-level random effect accounts for the fact that repeated measurements from the same person are related.

The posted primary comparison is expressed as a difference in estimated means. This is a model-based contrast between groups, rather than simply a subtraction of the two posted group means. In general, model-based estimates can account for the structure of repeated measurements and the terms included in the model. The registry does not report further details about covariance structure, model diagnostics, or how missing values were handled within the model.

What do the confidence interval and p-value say?

In general, a confidence interval represents a range of values compatible with the data and the model at a stated confidence level. A narrower interval indicates greater precision than a wider interval, but precision is not the same as importance. For a difference, an interval that includes zero is compatible with no difference as well as with differences in either direction. A p-value describes how unusual data at least as incompatible with a null hypothesis would be under the assumptions of the test; it does not measure the probability that the hypothesis is true.

The registry’s primary contrasts report 95% confidence intervals and p-values. These are the sponsor’s analyses and should be distinguished from the separate, simple unadjusted recalculations. The latter use posted group summary values and yield estimates that differ somewhat from the sponsor’s model results. Because the record marks a discrepancy, the sponsor’s own analysis takes precedence.

Why are there group means and model contrasts?

The posted means and standard errors describe each arm’s endpoint summary. A standard error describes uncertainty in an estimated mean, rather than the spread of individual participant measurements. In general, it tends to decrease as the amount of information increases and is distinct from a standard deviation, which describes variation among individual observations.

The recalculated comparisons are simple unadjusted comparisons based on the posted summaries. They do not reproduce all features of the cLDA model. For example, the registry’s model includes time, strata, treatment, and their specified interaction structure, while an unadjusted summary comparison does not. Differences between these approaches can occur without implying an error in either calculation. In this record, the sponsor’s reported analyses are the relevant source for the trial’s inferential conclusions.

What is the role of the active control?

Rasagiline was included as an active-control arm, alongside the placebo comparison. The posted primary outcome includes a placebo-versus-rasagiline analysis. An active-control comparison can describe how two active regimens compare on the measured outcome, but its presence does not by itself validate the sensitivity of the trial or establish a particular effect. The registry does not report a formal assay-sensitivity assessment or a noninferiority framework.

What can be inferred from secondary analyses?

The registry also reports responder status and change in the UPDRS Part 2 score. These analyses use different outcome forms: responder status is binary, while the Part 2 change is continuous. In general, a binary threshold discards some detail about the underlying continuous measurements, while a continuous score preserves gradations but depends on the scale’s interpretation. The secondary outcomes can add context, but they do not replace the primary endpoint or automatically confirm it.

The record contains multiple outcome measures and reported analyses. General statistical practice treats a collection of comparisons differently from a single prespecified test because repeated testing can increase the chance of apparently noteworthy results. The registry does not report whether or how multiplicity was controlled, nor does it provide a complete statistical analysis plan. Unadjusted secondary comparisons should therefore be read as exploratory or supportive rather than as independent confirmation.

General explanation: over-dispersion is relevant when variation in observed outcomes exceeds what a particular statistical model expects. Models for binary or count outcomes may need to account for this to avoid misleading standard errors. The registry identifies a generalized linear mixed model for responder analyses, but it does not report a dispersion assessment or a special over-dispersion adjustment. No conclusion about such diagnostics can be drawn from the posted record.

The record defines the primary analysis population, but it does not fully describe missing-data assumptions, sensitivity analyses, or the handling of withdrawals. In general, these choices matter because incomplete follow-up can affect estimated group comparisons, particularly when missingness is related to outcome or treatment. The extent and implications of any such issue are not reported in the registry record.

11. Confidence Intervals and Precision

Precision is best read from the confidence intervals around the sponsor-reported contrasts. For preladenant 2 mg versus placebo, the 95% interval is 0.86 to 4.30. Its endpoints are both above zero, so the reported model is inconsistent with a zero contrast at the conventional 95% interval level. The estimate is still uncertain across the span of values in that interval, and the interval does not by itself determine clinical importance.

For 5 mg versus placebo, the 95% interval is -0.41 to 2.94; for 10 mg versus placebo, it is -1.29 to 2.11. Both include zero and cover values on either side of no difference. These intervals therefore leave uncertainty about the direction and size of the contrast. The 10 mg estimate is close to zero relative to the width of its interval, but that observation should not be converted into a claim that the true difference is exactly zero.

The placebo-versus-rasagiline interval is -1.35 to 2.03, also including zero. For all comparisons, the interval’s width communicates how much uncertainty remains under the reported model. A p-value alone does not communicate that range, and a confidence interval alone does not establish whether a difference matters to patients.

The independent recalculations have somewhat different interval estimates from the sponsor’s analyses. The record flags this discrepancy, so the sponsor’s reported results take precedence for interpreting the trial. The posted group means and standard errors are useful descriptive summaries, but they are not substitutes for the model-based intervals. The registry does not report a minimum clinically important difference for this scale, so the clinical meaning of the interval endpoints cannot be classified from this record alone.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size illustrations are planning estimates, not descriptions of the original trial’s design. For these continuous outcomes, the figures refer to participants per arm. The assumptions listed for the illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set as the stated fraction of the observed effect.

The 80% power column describes the estimated number of participants per arm for a hypothetical study to have an 80% chance of detecting the specified effect under those assumptions. The 90% power column uses a higher target chance of detecting it and generally requires more participants. These are planning illustrations based on the observed effect, not observed power.

Each fraction is a scenario for how large an effect a future study aims to detect relative to the effect observed in this trial. Planning to detect a smaller effect requires more participants, because smaller differences are harder to distinguish from random variation. The estimates can change if the target effect, significance level, allocation, outcome variability, or analysis differs from the listed assumptions.

The original design assumptions and sample-size justification are not reported in the registry record. Consequently, these planning figures should not be read as evidence about what the trial was designed to detect or whether its original enrollment target was adequate. They are illustrative for future planning under the stated simplified assumptions, not a reconstruction of the trial’s formal design.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 2.500; pooled standard deviation σ = 8.659.
2. n = 2 × (1.96 + 0.842)² × 8.659² / 2.500² = 188.3, rounded up to 189 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 2.5                      # observed difference in means
sigma <- 8.65895490229623                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1155479)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 189 (80% power), 253 (90% power)
 75% of observed effect: n per arm = 335 (80% power), 449 (90% power)
 50% of observed effect: n per arm = 754 (80% power), 1,009 (90% power)
power.t.test: 189.3 
Simulated power at n = 189 per arm: 0.801 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The sponsor’s own analysis uses constrained longitudinal analysis for the primary endpoint and reports model-based differences in estimated means, 95% confidence intervals, and p-values. The recalculated comparisons are simple unadjusted estimates from the posted summary statistics. They are not the same analysis, and their estimates and intervals differ from the sponsor’s reported values.

The record explicitly flags a discrepancy. For that reason, the sponsor’s own analysis takes precedence when describing the trial’s reported inferential results. The recalculations can serve as a transparent check of the direction and rough scale of the comparisons, but they should not replace the sponsor’s model or be presented as the trial’s official test.

This distinction matters especially when an estimate is near a decision boundary or when a confidence interval differs across methods. A difference between a simple summary calculation and a longitudinal model can arise because the model incorporates its stated fixed and random effects, whereas the recalculation does not reproduce the complete model. The registry does not report enough detail to recreate every aspect of the analysis independently.

Thus, interpretation should quote the sponsor-reported primary contrasts when drawing conclusions about the posted trial analyses. The recalculated results remain unadjusted illustrations, not an alternative sponsor result and not a basis for overriding it.

15. Multiplicity, Missing Data and Analysis Populations

The registry describes a primary efficacy hypothesis: at least the 10 mg twice-daily preladenant dose would be superior to placebo on Week 26 change in the combined UPDRS Parts 2 and 3 score. It also reports multiple secondary outcomes and analyses. The record does not report the full statistical analysis plan, the ordering of hypotheses, or whether a multiplicity adjustment was applied across doses, outcomes, or contrasts.

In general, testing multiple hypotheses can increase the chance of at least one apparently unusual result, even when no true effect is present. Adjustment methods or a prespecified testing sequence can address this issue, but the registry record does not identify such a procedure. The reported p-values should therefore be interpreted as posted for their respective analyses, without assuming they control a family-wide error rate.

The primary population is defined as the Full Analysis Set: all randomized and treated participants with at least one post-baseline value. The record does not give enough detail to determine how participants without post-baseline measurements were handled, how missing observations were modeled, or whether sensitivity analyses examined alternative missing-data assumptions. It also does not report reasons for missing measurements or withdrawals in a way that would permit evaluation of their relationship to outcome.

These omissions do not invalidate the posted estimates, but they limit assessment of robustness. In general, the interpretation of a longitudinal analysis depends on assumptions about repeated observations and missingness. The registry identifies the cLDA model components but not all implementation details, diagnostics, or sensitivity checks. Any claims about unreported multiplicity control, interim analyses, or missing-data procedures would be speculative; those details are not reported in the registry record.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: atrial fibrillation (0.6%), appendicitis (0.6%), pneumonia (0.6%), femur fracture (0.6%), uterine leiomyoma (0.6%).

The registry gives serious adverse event counts and counts for other adverse events separately for Part 1 and Part 2. In Part 1, serious events are reported for each of the five arms, with affected-participant counts ranging from 3 to 9 in the listed groups. In Part 2, the listed serious-event counts range from 1 to 8. These counts are paired with their arm-specific at-risk denominators in the record; they should be read within the corresponding study part rather than combined across parts.

The posted counts for other adverse events also vary by arm and part. They describe participants with events, not the number of all events, and they do not by themselves establish a causal relationship to study treatment. The record lists atrial fibrillation, appendicitis, pneumonia, femur fracture, uterine leiomyoma, and suicide attempt among the most frequent serious-event terms, each reported as 0.6%.

Deaths are not reported in the registry record. The safety figures are descriptive and do not replace a full assessment of event timing, severity, exposure, or attribution. Part 2 followed a different treatment arrangement for the original placebo group, so Part 1 and Part 2 counts represent different treatment periods and should not be treated as a single like-for-like comparison.

17. What the Result Does, and Does Not, Mean

The posted results answer a narrow question: how the randomized groups compared on change in the combined UPDRS Parts 2 and 3 score at Week 26, under the sponsor’s stated longitudinal analysis. The reported estimate for preladenant 10 mg versus placebo has a 95% confidence interval that includes zero, and the reported p-value does not establish superiority. The trial’s stated primary efficacy hypothesis is therefore not supported by that posted comparison.

This does not prove that the true effect is exactly zero, nor does it establish that every dose, outcome, or treatment duration has the same result. Confidence intervals express uncertainty, and the record does not report a clinically important difference threshold for the scale. The data also do not answer broader questions about long-term outcomes, effects in other populations, or outcomes not included in the reported measures.

The 2 mg primary comparison has a sponsor-reported interval above zero, while the higher-dose comparisons do not establish a difference from placebo. This dose pattern is not the hypothesized finding that at least the 10 mg dose would outperform placebo. The distinction between an observed group contrast and a treatment claim is important: the trial result is about estimated scale changes in this study population and analysis, not a universal statement about treatment effects.

Safety counts should likewise be interpreted descriptively. The record lists events by study part and treatment group, but these numbers alone do not provide a complete causal or comparative safety assessment. Nothing in these statistical results is medical advice; individual treatment decisions require clinical judgment beyond the scope of the registry data.

18. Why This Trial Matters Statistically

This study is statistically informative because it combines randomization, masking, placebo control, and an active-control arm with a continuous longitudinal primary outcome. Those features define a structured comparison and provide multiple reference groups for the Week 26 result. The two-part design also separates the stated efficacy focus in Part 1 from the stated safety and tolerability focus in Part 2.

The record illustrates why a prespecified primary question matters. The stated hypothesis concerned superiority of at least the 10 mg twice-daily dose to placebo, while the most clearly separated sponsor-reported primary contrast is for the 2 mg arm. Interpreting only the most favorable-looking comparison would not answer the primary question. Dose-specific estimates, their uncertainty, and their relation to the stated hypothesis all matter.

It also demonstrates the importance of distinguishing model-based sponsor analyses from simple recalculations. The sponsor’s cLDA uses repeated measurements and specified model terms; recalculations from summary data are useful for orientation but do not reproduce that analysis. Because the record flags a discrepancy, the sponsor’s reported analysis is the proper basis for the trial’s formal inference.

Finally, the registry makes visible several interpretive limits common to trial records: incomplete detail on multiplicity, missing-data procedures, and design assumptions. These gaps do not erase the reported results, but they affect how confidently readers can assess the robustness and planning behind them. Statistical interpretation depends not only on point estimates and p-values, but also on the question asked, the analysis population, precision, and what the record leaves undocumented.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry record is marked terminated, but the reason is not reported. Termination status alone does not explain whether the cause was administrative, operational, safety-related, or efficacy-related, so no conclusion about the reason is warranted.

The record gives the primary model’s broad structure but not a full analysis plan. Details about covariance assumptions, model diagnostics, handling of missing data, sensitivity analyses, and multiplicity procedures are not reported in the registry record. These omissions restrict independent evaluation of the robustness of the sponsor’s estimates and the interpretation of multiple tests.

The primary endpoint is measured at Week 26, although the study had a further 26-week part. The primary result therefore should not be treated as a report of the entire one-year period. Treatment assignment also changed for the original placebo group in Part 2, which limits direct interpretation of later comparisons as continued placebo control.

The registry reports the scale’s possible range and direction but not a clinically important difference threshold. This makes it difficult to translate the estimated contrasts into a judgment about clinical importance from the record alone. The reported arm summaries and analysis populations also vary by outcome and safety period; counts should not be combined or assumed to have identical denominators.

Finally, the record’s discrepancy between simple recalculations and sponsor analyses underscores that summary statistics do not reproduce a longitudinal model. The sponsor’s analyses take precedence, but limited model detail makes an exact independent reconstruction unavailable.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Placebo- and Active-Controlled Study of Preladenant in Early Parkinson's Disease (PD) (P05664) — NCT01155479.
• Registry record last updated 2018-11-07; status last verified 2018-10.
• Results first posted 2016-07-28.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

NCT01155479 is a randomized, quadruple-masked, parallel-group Phase 3 study of preladenant in early Parkinson’s disease, with placebo and rasagiline active-control arms. It was designed as a two-part study, with the primary efficacy endpoint measured at Week 26.

The primary outcome was change from baseline in the combined UPDRS Parts 2 and 3 score. The sponsor-reported analysis did not establish superiority of preladenant 10 mg over placebo. The record also reports secondary outcomes and safety counts, but multiplicity procedures, missing-data details, and the reason for termination are not reported in the registry record.