Open- Label Trial of Sipuleucel-T Administered to Active Surveillance Patients for Newly Diagnosed Prostate Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Open- Label Trial of Sipuleucel-T Administered to Active Surveillance Patients for Newly Diagnosed Prostate Cancer (NCT03686683), a randomized phase 3 study. The primary outcome is Efficacy of Sipuleucel-T Measured as the Percentage of Subjects Without Histological Reclassification (Gleason Group Upgrade)..

2. Clinical Question

The study was designed to assess whether sipuleucel-T could reduce progression of lower-risk, non-metastatic prostate cancer compared with active surveillance. For the primary endpoint, “progression” is operationalized as a specified change in Gleason group, rather than as a broad clinical judgment. A participant starting in Grade Group 1 met the upgrade definition upon reaching Grade Group 2 or higher; a participant starting in Grade Group 2 met it upon reaching Grade Group 3 or higher.

The registry reports 356 participants in the sipuleucel-T group and 176 in the active-surveillance group for this outcome. The outcome is defined as the percentage without reclassification; the posted group counts are 108 and 48, respectively. The question is therefore about the observed group proportions and the uncertainty around their contrast, not about an individual participant’s prognosis or a clinical recommendation.

3. Trial Design

This was a randomized, phase 3, open-label, parallel-group trial. Randomization assigned participants to sipuleucel-T or active surveillance. The registry record reports no masking, so participants and study personnel were not blinded to assigned group. However, the primary endpoint was determined by blinded independent central review (BICR), a feature intended to have the histological assessment made independently of the treatment assignment.

The registry lists enrollment of 532 and describes the trial as completed. The recorded start date is 2018-10-18, and the primary completion date is 2023-03-10. These record dates do not, by themselves, describe how long each participant was followed. The primary outcome timeframe is defined as within 36 months of randomization, with assessment once all participants had completed at least 3 years following randomization.

Participants assigned to sipuleucel-T were scheduled to receive 3 intravenous infusions at approximately 2-week intervals. Participants assigned to the control group were followed on active surveillance. The registry describes the study as comparing sipuleucel-T with active surveillance as standard of care for reducing progression of lower-risk, non-metastatic prostate cancer.

The posted primary outcome is binary at the participant level: each participant is classified according to whether histological reclassification occurred within the defined timeframe. The reported measure is the percentage without reclassification. The outcome count and analyzed group sizes are listed, but the registry record does not report a detailed account of participant flow, exclusions, adherence, or missing endpoint assessments. It also does not report the trial’s original sample-size justification or the detailed procedures used to implement randomization.

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.

Overall Study

Reasons for not completing this period:

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.

6. Primary Endpoint

The registry lists the primary outcome as Efficacy of Sipuleucel-T Measured as the Percentage of Subjects Without Histological Reclassification (Gleason Group Upgrade)., measured over Once all participants have completed at least 3 years following randomization.

The primary endpoint is the percentage of participants without histological reclassification within 36 months of randomization, based on blinded independent central review. Reclassification is defined by an upgrade in Gleason group: a participant starting in ISUP Grade Group 1 (Gleason 3+3) is considered upgraded at Grade Group 2 (Gleason 3+4) or higher, while a participant starting in Grade Group 2 is considered upgraded at Grade Group 3 (Gleason 4+3) or higher.

This definition makes the endpoint a binary outcome for each participant: no qualifying upgrade or a qualifying upgrade within the specified period. The registry reports counts without reclassification of 108 among 356 participants assigned to sipuleucel-T and 48 among 176 assigned to active surveillance. The group percentages are 30.34% and 27.27%, respectively.

A histological endpoint offers a defined assessment of change in tumor grade, rather than relying on an unspecified impression of progression. Central review can support consistent classification across participants, and the registry explicitly states that the review was blinded. These design features explain why this endpoint was chosen to assess the trial’s stated question; they do not establish that the endpoint captures every clinically relevant outcome. Other primary or secondary outcomes and their results are not reported in the registry record.

7. Results Posted in the Registry

• Treatment Group: Sipuleucel-T: 30.3% (108/356) versus 27.3% (48/176) on control; risk difference 3.1 percentage points (95% CI -5.1 to 11.2), risk ratio 1.11 (0.83–1.48); unadjusted two-proportion z-test p = 0.465.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: a participant either had no qualifying histological reclassification within 36 months or did not. For a binary outcome, a group proportion describes the fraction of participants in that group with the specified outcome. The registry reports the counts and analyzed group sizes, as well as percentages without reclassification. These quantities describe the observed data; they do not by themselves account for uncertainty in the comparison.

What does the risk difference express?

The recalculated contrast is an unadjusted risk difference, which compares the two group proportions by subtraction. Its positive direction indicates that the observed proportion without reclassification was higher in the sipuleucel-T group. The estimate is 0.031, and its confidence interval runs from -0.051 to 0.112. Because the interval includes zero, it includes no difference between the group proportions. The interval also contains values on either side of zero, reflecting uncertainty about both the direction and size of the contrast.

“Unadjusted” means this recalculation uses the posted group counts and denominators without accounting for additional covariates or strata. It is a simple comparison of the displayed proportions. It should not be treated as the same estimand or analysis procedure as the sponsor’s reported Cochran-Mantel-Haenszel odds ratio.

What does an odds ratio express?

An odds ratio compares the odds of the specified outcome between groups. Odds and proportions are related but are not interchangeable measures. An odds ratio of 1 represents equal odds under the comparison; values on either side represent different directions of association. The sponsor reports an odds ratio of 1.16, with a 95% confidence interval from 0.78 to 1.74. This interval includes 1, the no-association value.

The sponsor names the Cochran-Mantel-Haenszel method for this odds-ratio analysis. In general, this method can combine comparisons across strata. The registry record does not identify the strata used or provide further details of the model, so the exact adjustment structure cannot be described. The reported estimate should be understood as the sponsor’s stated analysis, not as an unadjusted calculation from the overall counts.

How should the p-values be read?

A p-value measures how compatible observed data are with a specified null hypothesis under the assumptions of the statistical test. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The unadjusted recalculation has p = 0.465. The sponsor’s reported p-value is greater than 0.05. Both are consistent with the associated intervals including their respective null values, but the two analyses use different effect measures and methods.

The registry does not report whether the p-value was adjusted for multiple comparisons. Only one primary outcome measure and one reported analysis are listed, and no secondary outcomes are posted. This does not establish what other analyses may have been planned or performed. Multiplicity procedures and the trial’s full statistical analysis plan are not reported in the registry record.

Why does the confidence interval matter?

A confidence interval gives a range of estimates compatible with the data and the method used, at the stated confidence level. In repeated use of the same method under its assumptions, intervals at a given confidence level would cover the underlying parameter at that rate. It is not a statement that a particular parameter has a specified probability of lying in the observed interval. For the risk difference, the interval’s inclusion of zero indicates that the data do not distinguish the observed contrast from no difference at the stated confidence level. Its width also shows that the estimate is not highly precise.

What limits the recalculation?

The recalculation uses only the posted counts and denominators and is unadjusted. It cannot reproduce any analysis that depends on unreported stratification, covariates, participant-level information, or missing-data procedures. The registry does not report those details for the posted primary outcome. The recalculated values therefore serve as a transparent summary-statistics comparison, while the sponsor’s reported analysis remains the registry’s own named inferential analysis.

10. Confidence Intervals and Precision

Precision concerns how narrowly the data locate an effect estimate. The recalculated risk difference is 0.031, with a 95% confidence interval from -0.051 to 0.112. The interval includes zero, the value corresponding to no difference in group proportions, and extends to both negative and positive values. This means the posted data do not pin down the direction of the underlying difference with high precision.

The interval also conveys that estimates materially different from the point estimate remain compatible with the data under the calculation’s assumptions. It should not be read as a list of equally likely effects, nor as a range that guarantees the true effect is contained in this particular interval. It is a measure of sampling uncertainty based on the selected method and the available summary data.

The sponsor’s odds-ratio interval is 0.78 to 1.74 at the 95% confidence level. It includes 1, the null value for an odds ratio, and allows for associations in either direction. This interval is for a different effect measure and analysis from the unadjusted risk-difference calculation; the intervals should not be compared as if they were measuring the same quantity.

The open-label design and the use of blinded central review are both relevant to interpretation, but neither changes what the intervals mean. The registry record does not provide enough detail to assess how missing outcome assessments, analysis-population definitions, or other modeling choices may affect precision. The available intervals therefore describe uncertainty under the reported and recalculated analyses, not every possible source of uncertainty in the trial.

11. 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 planning figures are participants per arm for 80% power and participants per arm for 90% power. They are illustrations based on the observed effect, not the trial’s original design assumptions. The calculation assumptions listed for these 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 planning display gives figures for planning for 100%, 75%, and 50% of the observed effect.

The 80% power column gives the planned number of participants per arm for a design with an 80% chance of rejecting the null hypothesis if the specified effect were the true effect and the calculation assumptions held. The 90% power column gives the corresponding per-arm planning number for a design with a 90% chance under those conditions. These are prospective planning interpretations, not descriptions of the chance that the completed trial detected an effect.

For the binary outcome, the display reports participants per arm, not total events. The planning values increase as the assumed effect becomes smaller: planning for 75% or 50% of the observed effect requires more participants per arm than planning for 100% of the observed effect. This reflects a general statistical principle: smaller differences are harder to distinguish from random variation, so a larger study is generally needed to detect them with a specified power.

The original trial’s sample-size justification and design assumptions are not reported in the registry record. These planning figures are not observed power, and they should not be used to infer what the completed study was designed to achieve or whether its result is valid.

12. 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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.273; treatment proportion p₁ = 0.303; average p̄ = 0.288.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6404 = 1.2552.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6401 = 0.5387.
4. n = (1.2552 + 0.5387)² / (0.0306)² = 3,427.1, rounded up to 3,428 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
p0 <- 0.272727272727273                          # control proportion
p1 <- 0.303370786516854                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(3686683)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 3,428 (80% power), 4,588 (90% power)
 75% of observed effect: n per arm = 6,045 (80% power), 8,092 (90% power)
 50% of observed effect: n per arm = 13,491 (80% power), 18,060 (90% power)
power.prop.test: 3427.1 
Simulated power at n = 3,428 per arm: 0.801 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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.

13. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary analysis as a Cochran-Mantel-Haenszel odds ratio of 1.16, with a 95% confidence interval from 0.78 to 1.74 and a p-value greater than 0.05. The odds ratio is based on the sponsor’s stated analysis method. Details of any strata or other analysis specifications are not reported in the registry record.

The recalculation uses the posted group counts to estimate an unadjusted risk difference. Its estimate is positive, its confidence interval includes zero, and its unadjusted p-value is 0.465. This is not the same calculation as the sponsor’s odds ratio: the effect measure differs, and the recalculation does not account for any stratification or other adjustments.

The recalculated odds ratio from the summary statistics is close to the sponsor’s reported rounded estimate, and the registry record does not flag a discrepancy. The sponsor’s result and the unadjusted recalculation point to uncertainty that includes the respective null value. The recalculation is a check using summary data; it does not replace the sponsor’s named analysis or establish that the analyses are identical.

14. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary outcome measure and one reported analysis, and it lists no posted secondary outcomes. This limits what can be assessed about multiplicity from the record. It does not establish that no other outcomes or analyses were planned or conducted. The statistical analysis plan, any hierarchy of outcomes, and any adjustment for multiple testing are not reported in the registry record.

Multiplicity arises when several hypotheses are tested: the chance of at least one apparently positive result can increase as more tests are considered. In general, a prespecified testing strategy or adjustment can address this issue. Whether such procedures were used in this trial is not reported. The single posted outcome and analysis should therefore be interpreted as the available registry results, not as a complete account of all analyses.

The analyzed population is not described in sufficient detail to determine whether it was intention-to-treat, per-protocol, or another analysis set. The registry lists group sizes for the outcome, but does not explain how these relate to all randomized participants or whether exclusions occurred. Missing endpoint data, reasons for missingness, and the method used to handle any missing assessments are also not reported in the registry record.

These omissions matter because analysis-population choices and missing-data procedures can affect the estimate and its uncertainty. Without participant-level data and a stated analysis plan, it is not possible to evaluate the assumptions used for incomplete outcomes or to reproduce the sponsor’s analysis fully. The blinded independent central review is reported for the endpoint assessment, but the registry does not give further operational details about review procedures or adjudication.

15. Safety Results

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

Deaths

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: death (1.4%), haematuria (1.2%), pulmonary embolism (1.2%), coronary artery disease (0.9%), atrial fibrillation (0.9%).

The registry lists serious adverse events among participants at risk: 74 affected among 345 in the sipuleucel-T group and 22 among 176 in the active-surveillance group. It also reports other adverse events affecting 321 among 345 and 133 among 176, respectively. Death counts are listed as 10 in the sipuleucel-T group and 5 in the active-surveillance group.

The listed serious-event terms and percentages include Death at 1.4%, Haematuria at 1.2%, Pulmonary embolism at 1.2%, Coronary artery disease at 0.9%, Atrial fibrillation at 0.9%, and Sepsis at 0.9%. These are registry-listed event terms and percentages; the record does not provide enough detail to interpret them as comparative risks or to establish whether events were caused by treatment.

These figures should be read as descriptive safety reporting, not as a randomized statistical comparison. The registry does not report a safety-analysis method, confidence intervals for these counts, or further detail about event definitions and follow-up in the safety record. Differences in counts alone do not establish a causal effect or support conclusions about comparative safety.

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

The results describe a randomized comparison of sipuleucel-T with active surveillance for a defined histological outcome. The observed percentages without reclassification were 30.34% and 27.27%, respectively. The unadjusted risk-difference interval includes no difference, and the sponsor’s odds-ratio interval includes its null value. The evidence in the registry record is therefore not a precise demonstration of a difference in this endpoint.

That interpretation is not equivalent to proving that the groups are the same. An interval that includes a null value can also include effects in either direction, and the posted intervals are not narrow enough to exclude every potentially meaningful contrast. The data support uncertainty, rather than a definite claim of equivalence or a definite claim of benefit or harm.

The scope of the conclusion is limited to absence of the specified Gleason group upgrade within 36 months of randomization, assessed by blinded independent central review. It does not cover every definition of prostate cancer progression, outcomes after the stated period, or unreported clinical outcomes. The registry’s posted results also do not describe how missing assessments or analysis-population decisions were handled.

Statistical uncertainty and clinical importance are separate considerations. The record does not give a threshold for a clinically important difference on this endpoint, so the estimates cannot be classified against such a threshold. The reported findings are a summary of the available outcome data and their statistical precision; they do not provide medical advice or establish what an individual participant should do.

17. Why This Trial Matters Statistically

This trial illustrates how a randomized comparison can be statistically informative without producing a precise estimate of the treatment contrast. The primary endpoint is clearly defined around histological reclassification, and the registry reports group counts, percentages, and a sponsor analysis. Together, these permit readers to distinguish the observed proportions from the uncertainty around their comparison.

It also illustrates why an endpoint’s definition matters. “Without histological reclassification” is not a general measure of all possible disease changes; it is tied to specified upgrades in Gleason group and a defined follow-up timeframe. The blinded independent central review is an important reported assessment feature, while the open-label design remains part of the trial context.

The contrast between the sponsor’s odds ratio and the unadjusted risk difference is a useful reminder that different effect measures answer related but distinct questions. Their estimates should be interpreted alongside their methods and intervals rather than treated as interchangeable. The registry does not report enough detail to reconstruct the sponsor’s full analysis or assess its stratification.

Finally, the record demonstrates the value of separating a study’s observed findings from planning illustrations. The sample-size figures use the observed effect under stated assumptions and do not represent the trial’s original design rationale. The completed trial’s result is best understood from its posted estimates and uncertainty, with missing-data, multiplicity, and analysis-population details recognized as unreported.

18. 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 leaves several trial-specific details unavailable. It does not report the original sample-size justification, the assumptions used in the trial’s own design, or a full statistical analysis plan. As a result, the planning illustrations cannot be used to reconstruct the rationale for the enrollment target or the analysis strategy that was prespecified.

The endpoint results are reported with group counts and denominators, but the record does not explain participant flow through randomization and analysis, the analysis-population definition, or exclusions. It also does not describe missing outcome assessments or how any were handled. These omissions limit assessment of whether the posted comparison includes all randomized participants and whether incomplete assessments could affect the estimate.

The sponsor names the Cochran-Mantel-Haenszel method, but the registry does not identify the strata or provide detailed model specifications. The unadjusted recalculation from aggregate counts cannot reproduce an analysis that uses participant-level information or stratification. The record also does not report adjustment for multiple testing, any interim analyses, or the full set of outcomes that may have been considered.

The trial was open-label, although the primary histological assessment is described as blinded independent central review. The registry does not provide further details on how review was conducted or how disagreements were handled. Finally, the posted primary endpoint is limited to a specified histological upgrade within 36 months; the registry results do not establish effects on outcomes outside that definition or timeframe.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Open- Label Trial of Sipuleucel-T Administered to Active Surveillance Patients for Newly Diagnosed Prostate Cancer — NCT03686683.
• Registry record last updated 2024-10-08; status last verified 2023-10.
• Results first posted 2024-10-08.
• 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.

22. Record Summary

NCT03686683 is a completed, randomized, phase 3, open-label, parallel-group trial comparing sipuleucel-T with active surveillance in newly diagnosed, lower-risk, non-metastatic prostate cancer. The registry lists 532 enrolled participants and records a start date of 2018-10-18 and a primary completion date of 2023-03-10.

The primary endpoint is the percentage without histological reclassification within 36 months of randomization, determined by blinded independent central review. The posted percentages are 30.34% for sipuleucel-T and 27.27% for active surveillance. The unadjusted risk-difference interval includes zero; the sponsor reports an odds ratio of 1.16 with a 95% confidence interval from 0.78 to 1.74.

The record does not report the original sample-size rationale, analysis-population details, or missing-data procedures. Safety counts are descriptive and should not be interpreted as causal comparisons.