A Study to Determine the Bioequivalence of Oraxol in Cancer Patients Treated With Intravenous Paclitaxel: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Determine the Bioequivalence of Oraxol in Cancer Patients Treated With Intravenous Paclitaxel (NCT04035473), a randomized phase 1 study. The primary outcome is Area Under the Concentration-Time Curve Zero Time Extrapolated to Infinite Time (AUC0-∞).

2. Clinical Question

3. Trial Design

This was a randomized, open-label, crossover study in Phase 1. Participants received both treatments in different periods and were assigned to a treatment order: Oraxol followed by intravenous paclitaxel, or intravenous paclitaxel followed by Oraxol. Oraxol was given on Days 1, 2, and 3 of its treatment period; intravenous paclitaxel was given on Day 1 of its period. Because participants served in both treatment conditions, the design sought to compare exposure under the two regimens within the trial population, though the record does not report the statistical model used to account for that pairing.

The brief description calls the study multicenter and describes a two-stage process. Stage 1 began with an initial cohort of up to 6 evaluable participants. An interim pharmacokinetic assessment was planned to inform whether the regimen appeared likely to meet the AUC0-∞ bioequivalence criterion in a larger Stage 2. If it appeared unlikely, a second Stage 1 cohort of up to 6 evaluable participants could be enrolled, with a possible paclitaxel dose adjustment of no more than +/- 25%. The record describes a consensus decision involving the Data Safety and Monitoring Board, Kinex, Zenith Technology, and the Principal Investigator. The available registry description does not report the detailed Stage 2 allocation or the results of those interim assessments.

The study is listed as completed, with enrollment of 42. The primary pharmacokinetic analysis set is described as 35 randomized subjects who completed the study. The registry record does not report a detailed account of participant flow or explain the difference between those figures.

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: 45 subjects signed ICF. 3 subjects were screening failures and 42 were randomized. All of the screen failure subjects failed to meet inclusion or exclusion criteria

Screening/Baseline

Reasons for not completing this period:

Treatment Period 1

Treatment Period 2

Follow-up Period

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The 35 randomized subjects completed the study and made up the PK Analysis Set

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: Of the 42 randomized subjects, 2 discontinued study participation due to an AE prior to receiving study drug

6. Primary Endpoint

The registry lists the primary outcome as Area Under the Concentration-Time Curve Zero Time Extrapolated to Infinite Time (AUC0-∞), measured over Pharmacokinetic Sampling for IV Paclitaxel - Predose to 96 hours after infusion (Day 1-5). Pharmacokinetic Sampling for Oraxol- predose of Day1 to 144 hours after third dose of day 3 (Day 1-9).

The primary endpoint was the area under the plasma concentration-time curve extrapolated from time zero to infinite time, AUC0-∞, reported as a mean with standard deviation and measured in ng*h/mL. AUC represents overall drug exposure across the concentration-time profile. Extending the area to infinite time includes an estimate of exposure after the last measured concentration, rather than limiting the measure to the observed sampling window.

Exposure is relevant to a bioequivalence question because the comparison concerns whether drug exposure after one formulation is sufficiently similar to exposure after another. The record says paclitaxel concentrations were normalized to 615 mg/m2 for Oraxol and 80 mg/m2 for intravenous paclitaxel before analysis. It also describes noncompartmental analysis using concentration-time data from oral and intravenous paclitaxel. Sampling was described as predose to 96 hours after the intravenous infusion, and from predose on Day 1 to 144 hours after the third Oraxol dose on Day 3.

The posted summaries give the mean and standard deviation for each treatment and an unadjusted mean difference. Bioequivalence is ordinarily assessed against prespecified acceptance limits, not solely by asking whether two means are exactly equal. Those limits and a formal bioequivalence analysis are not reported in the registry record.

7. Results Posted in the Registry

• IV Paclitaxel: mean 5,596 versus 5,034 on control; difference 562.40 (95% CI 90.05 to 1,035), standardized effect d = 0.56; z-test p = 0.020.

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.

Maximum Observed Concentration (Cmax)

• 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 deviations.
• Result:
  • IV Paclitaxel: mean 2,733 versus 397.20 on control; difference 2,336 (95% CI 2,121 to 2,551), standardized effect d = 5.09; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for IV Paclitaxel than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Area Under the Concentration-time Curve From Time 0 to Time of Last Quantifiable Concentration (AUC0-t)

• 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 deviations.
• Result:
  • IV Paclitaxel: mean 5,432 versus 4,829 on control; difference 602.40 (95% CI -2.30 to 1,207), standardized effect d = 0.47; z-test p = 0.051.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between IV Paclitaxel and the control. The interval gives the range of effects that remain compatible with the data.

Time at Which the Highest Drug Concentration Occurs (Tmax)

• 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 deviations.
• Result:
  • IV Paclitaxel: mean 1.00 versus 1.40 on control; difference -0.40 (95% CI -0.61 to -0.19), standardized effect d = -0.89; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for IV Paclitaxel than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary pharmacokinetic results cover maximum observed concentration, AUC from time zero to the last quantifiable concentration, and the time at which the highest concentration occurs. Their overall pattern is that the treatment summaries differ across these aspects of the concentration-time profile, rather than matching on every measure. These measures complement the primary AUC0-∞ endpoint: maximum concentration focuses on the observed peak, AUC0-t captures observed exposure through the last quantifiable sample, and Tmax describes the timing of that peak.

Secondary measures can help characterize how exposure profiles differ, but they do not substitute for the primary bioequivalence assessment. The AUC0-t and AUC0-∞ measures address related but not identical windows, and the peak and timing measures describe different features of the profile. The registry record does not report a formal analysis connecting the secondary outcomes to the primary estimand or explaining how discrepant patterns would be interpreted.

The numerical comparisons for the posted secondary outcomes are unadjusted recalculations. Multiple secondary outcomes create multiple opportunities to observe a nominally notable result by chance. Without a reported multiplicity strategy, unadjusted p-values should carry less evidential weight than a prespecified primary analysis. The registry record lists additional outcomes, including terminal elimination half-life and safety and tolerability, but numerical results for all listed measures are not reported in the record summarized for these results. The absence of posted analyses makes it especially important to distinguish descriptive patterns from confirmatory evidence.

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

• Maximum Observed Concentration (Cmax)
• Area Under the Concentration-time Curve From Time 0 to Time of Last Quantifiable Concentration (AUC0-t)
• Time at Which the Highest Drug Concentration Occurs (Tmax)
• Terminal Elimination Phase Half-life (t½)
• Safety and Tolerability of Oraxol Compared With IV Paclitaxel

9. Statistical Interpretation

10. Statistical Methods Explained

The posted endpoint is continuous: each participant contributes a measured concentration-time summary, and the registry reports group means and standard deviations. A mean describes the central value of a set of measurements. A standard deviation describes how dispersed individual measurements are around that mean. Neither statistic alone describes the full distribution, and the mean can be affected by unusually high or low values.

What does the mean difference describe?

The comparison is expressed as intravenous paclitaxel minus Oraxol. A positive estimate therefore means the posted intravenous mean is larger than the posted Oraxol mean on the endpoint's stated scale. The primary estimate is a simple unadjusted recalculation from the posted summary statistics. It should not be mistaken for an adjusted crossover estimate or a model-based result unless the registry record documents that method. The record lists no reported analyses, and it does not provide a model specification.

In a crossover trial, the same participant receives both treatments in different periods. This can allow treatment comparisons to use within-participant information, but analysis must reflect the paired observations and the order and timing of treatment. Period effects, treatment sequence, and possible carryover are statistical considerations in crossover studies generally. The registry record does not report whether or how these factors were modeled, or whether a paired analysis was used. Summary statistics separated by treatment do not retain all the within-participant information needed to reproduce such an analysis.

What do a confidence interval and p-value mean?

A confidence interval describes uncertainty in an estimate under the assumptions of the method used to construct it. A narrower interval indicates greater precision on that scale; a wider interval indicates more uncertainty. It is not a range containing a fixed proportion of individual participant values. The primary interval is for a mean difference, not for a ratio of geometric means or a bioequivalence margin.

The unadjusted p-value summarizes how unusual data at least as incompatible with a specified null hypothesis would be if that hypothesis and the calculation assumptions were true. In the recalculation, the comparison is a test of a zero mean difference. A p-value does not measure the probability that the null hypothesis is true, the size or clinical importance of a difference, or whether bioequivalence criteria have been satisfied. The registry record does not report a trial-level inferential analysis that would define a different null hypothesis or account for the crossover design.

Why does variability matter?

When measurements vary substantially between participants, an estimated mean may be less precise unless the analysis has adequate information. The primary standard deviations are notably different in scale, which makes it important not to read the two means without their accompanying spread. The simple recalculation pools variability for its standardized effect-size illustration, but that pooled summary is not a replacement for an analysis appropriate to the randomized crossover design. In general, unequal variability and correlated repeated measurements can affect the uncertainty calculation; the record does not provide enough detail to assess the trial's own handling of these issues.

Over-dispersion is a general concern when observed variability exceeds what a chosen statistical model expects. It can make model-based standard errors too small if the model fails to represent the extra variation. The posted summaries do not establish that over-dispersion occurred or that a particular model was used; this is a general reason to check model assumptions, not a finding about this study.

How does the design connect to the endpoint?

Pharmacokinetic sampling was scheduled over different windows for the intravenous and oral regimens. AUC0-∞ uses the measured concentration-time profile and an extrapolated tail, so its precision can depend on sampling adequacy and the estimated terminal portion of that profile. The record gives the sampling windows and says noncompartmental analysis was used, but it does not report the proportion of AUC extrapolated, individual profile quality checks, or handling of values that could not be estimated.

The concentrations were dose-normalized before analysis. This permits a comparison on the stated normalized dose basis, but it does not make the regimens identical in administration or schedule. Oraxol involved oral dosing over three consecutive days, while intravenous paclitaxel was administered on one day in its period. Interpretation should therefore remain tied to the normalization and sampling procedures described in the record. Details of the statistical analysis plan, including any prespecified transformations, model terms, or estimands, are not reported in the registry record.

11. Confidence Intervals and Precision

The primary unadjusted recalculated mean difference was 562.4 ng*h/mL, with a 95% confidence interval from 90.05 to 1035.0. The interval does not include zero, but its breadth shows that the estimated size of the difference is not highly precise. Values toward either end of the interval represent meaningfully different magnitudes of the mean difference, even though they are on the same positive side of zero.

A confidence interval is an uncertainty range for an estimated quantity under the calculation's assumptions. It does not show the range of participant-level AUC measurements, and it does not give the probability that the true value lies within this particular interval. The standard deviations describe participant-level spread; the interval describes uncertainty in the estimated mean difference. These are related but distinct ideas.

For this trial, the interval comes from an unadjusted recalculation based on posted summary statistics. It does not incorporate an explicitly reported paired crossover model, and the registry record does not report a sponsor analysis or a confidence interval tied to bioequivalence limits. As a result, the interval is useful for describing uncertainty in the simple mean comparison, but not for deciding whether the intended bioequivalence criteria were met.

Precision should also be considered alongside the observed variation. The standard deviation was 1401.0 for Oraxol and 264.1 for intravenous paclitaxel. The larger Oraxol spread signals substantial variability in that treatment summary, but group-level statistics do not show whether this arose from a few extreme measurements, broadly variable responses, or another pattern. Participant-level distributions and the trial's detailed analysis are not reported in the registry record.

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 planning figures are stated as participants per arm, not total events. They are illustrations based on the observed effect and assume a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set in turn to the observed effect or a stated fraction of it. These are planning assumptions for the illustration; the original design assumptions for this trial are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm under which the calculation would have an 80% chance of detecting the specified effect if the assumptions hold. The 90% power column describes the corresponding target for a higher chance of detection. These are not guarantees: realized performance depends on the true effect, variability, design, and analysis. This is not observed power.

The rows describe planning for the full observed effect, 75% of the observed effect, and 50% of the observed effect. Smaller effects are harder to distinguish from random variation and therefore require more participants under the same assumptions. A count such as 1 participant per arm in an illustration should not be treated as a practical recommendation; it results from the particular normal-approximation calculation and the unusually large standardized effect in those posted summaries. The crossover structure and the absence of a documented original sample-size plan limit how directly these generic two-arm illustrations can be applied to the study.

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 = 562.400; pooled standard deviation σ = 1008.174.
2. n = 2 × (1.96 + 0.842)² × 1008.174² / 562.400² = 50.4, rounded up to 51 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 <- 562.4                      # observed difference in means
sigma <- 1008.17409706856                     # 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(4035473)
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 = 51 (80% power), 68 (90% power)
 75% of observed effect: n per arm = 90 (80% power), 121 (90% power)
 50% of observed effect: n per arm = 202 (80% power), 271 (90% power)
power.t.test: 51.4 
Simulated power at n = 51 per arm: 0.798 (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. Multiplicity, Missing Data and Analysis Populations

The registry record lists a primary endpoint and multiple secondary outcomes, but it does not report a statistical analysis plan, a multiplicity adjustment, or a hierarchy for testing outcomes. When several outcomes are tested, the chance of at least one apparently notable result can increase. This is a general statistical concern; it does not show that any particular result in this study is false. Because the secondary comparisons are unadjusted, they should be treated as exploratory rather than as a set of independent confirmations.

The record lists zero reported analyses. The posted mean comparisons and intervals are therefore best understood as simple recalculations from the reported summary statistics, not as a documented sponsor analysis of the crossover trial. The registry record does not report the planned statistical model, treatment of period or sequence, handling of within-participant correlation, or a prespecified bioequivalence decision rule. It also does not report the acceptance limits needed to interpret the primary endpoint as a formal bioequivalence test.

The primary population is identified as the 35 randomized subjects who completed the study and made up the pharmacokinetic analysis set. The record does not report an intention-to-treat analysis, a separate per-protocol analysis, or criteria for inclusion in the pharmacokinetic analysis set. It does not provide participant-level missingness, reasons for exclusions, or a method for handling missing concentration measurements. These details matter because complete-case summaries can represent a selected subset if missingness is related to treatment or participant characteristics.

The record also does not report whether any interim analyses were actually conducted, despite describing interim pharmacokinetic assessments in the study summary. Nor does it state how such assessments, if conducted, related to later inference. Without these details, the posted estimates cannot be evaluated against a complete prespecified analysis framework.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: septic shock (2.6%), vomiting (2.6%), altered state of consciousness (2.6%), atrial fibrillation (2.6%), cardiac tamponade (2.6%).

The registry reports serious adverse events affecting 4 participants at risk in the Oraxol group and 2 participants at risk in the intravenous paclitaxel group. The corresponding groups at risk were 39 and 38. These are group-specific registry counts; the record does not explain the relationship between these safety denominators and the pharmacokinetic analysis set. The listed serious-event terms each have a reported frequency of 2.6%: septic shock, vomiting, altered state of consciousness, atrial fibrillation, cardiac tamponade, and tachycardia.

Other adverse events were reported for 36 participants at risk in the Oraxol group and 29 participants at risk in the intravenous paclitaxel group. The death counts were 0 and 1, respectively. These figures describe what the registry reports during the study; they do not establish that treatment caused an event or that one regimen is safer. The crossover design, differing group denominators, and lack of event-level timing and attribution details limit direct interpretation. The registry record does not report a statistical comparison of adverse-event rates or provide enough detail to determine whether events occurred during particular treatment periods.

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

The posted primary summaries describe dose-normalized paclitaxel exposure in the pharmacokinetic analysis set. The observed means differ, and the unadjusted recalculated interval for their mean difference lies above zero. This is evidence of a difference in the posted mean summaries under that simple calculation. It is not, by itself, a formal finding that Oraxol and intravenous paclitaxel are or are not bioequivalent.

Bioequivalence is judged against defined acceptance limits using an appropriate analysis. The registry record does not report those limits, a formal decision, or the trial's statistical model. A test of whether a mean difference equals zero answers a different question from whether exposure falls within an acceptable bioequivalence range. Readers should not infer a bioequivalence conclusion from the unadjusted p-value or confidence interval alone.

The results are also specific to the dosing and sampling conditions described. Oraxol was administered orally across three consecutive days, while intravenous paclitaxel was administered on one day in its treatment period. Concentrations were normalized to the stated dose levels. The posted summaries therefore characterize the measured exposure under those conditions and analysis choices; they do not establish results for other regimens, populations, or clinical outcomes.

This was a pharmacokinetic study, and the primary endpoint concerns drug exposure rather than tumor response, survival, or another clinical benefit. The registry record does not report evidence that permits conclusions about those outcomes. Nor do the adverse-event summaries establish comparative safety or causation. The record supports a careful description of the posted exposure and event counts, while important elements needed for confirmatory interpretation are not reported.

17. Why This Trial Matters Statistically

This study illustrates how the design and the endpoint need to align with the clinical question. A crossover design compares treatments received by the same participants, which can be informative for pharmacokinetic measurements. But the analysis needs to account for paired observations and treatment order. The registry record describes the randomized crossover structure but does not report the analysis details needed to assess that alignment.

It also highlights the distinction between a difference test and a bioequivalence test. The posted unadjusted comparison assesses a mean difference against zero. Bioequivalence instead depends on prespecified limits and a suitable analysis of exposure. Without the limits and trial analysis, a statistically notable mean difference does not settle the intended question.

Finally, the record demonstrates why estimates should be read with their uncertainty and variability. The primary interval spans a range of positive differences, while the treatment standard deviations are notably different. Neither a single p-value nor an observed mean captures the full structure of the data. The secondary measures add profile information, but unadjusted comparisons across multiple endpoints do not carry the same weight as a clearly specified primary analysis.

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 most important limitation for interpretation is the absence of a reported trial analysis. The registry record provides arm-level means, standard deviations, and simple recalculated comparisons, but not a model specification for the crossover design. The posted comparison therefore cannot show how the trial's own analysis handled paired measurements, treatment periods, sequence, or possible carryover.

The bioequivalence acceptance limits and the formal decision rule are not reported in the registry record. Without them, the primary mean-difference comparison cannot establish whether the intended criterion was met. The record also does not report the original sample-size assumptions or the rationale for the final analysis-set size.

Although the study summary describes interim pharmacokinetic assessments and possible dose adjustment between cohorts, the registry record does not report whether those steps occurred or how their results affected the final analysis. That missing detail limits assessment of the relationship between the described two-stage process and the posted results.

Participant flow and missing-data handling are not described in enough detail to explain the difference between enrollment of 42 and the 35 randomized subjects in the pharmacokinetic analysis set. The record does not report individual concentration profiles, the extent of AUC extrapolation, or detailed secondary analyses. These limitations constrain assessment of robustness, precision, and generalizability beyond the studied dosing conditions.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Study to Determine the Bioequivalence of Oraxol in Cancer Patients Treated With Intravenous Paclitaxel — NCT04035473.
• Registry record last updated 2022-04-13; status last verified 2022-02.
• Results first posted 2022-04-13.
• 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

The registry record describes a completed Phase 1, randomized, open-label crossover study of oral Oraxol and intravenous paclitaxel in cancer patients. Enrollment is listed as 42; the primary pharmacokinetic analysis set comprised 35 randomized subjects who completed the study. The primary endpoint was dose-normalized AUC0-∞, reported as a mean and standard deviation. The posted unadjusted recalculation estimates a positive mean difference for intravenous paclitaxel relative to Oraxol, but the registry record does not report a formal bioequivalence analysis or its acceptance limits. Secondary pharmacokinetic summaries and adverse-event counts are also posted, with no reported analysis plan, multiplicity approach, or detailed crossover model.