Safety and Efficacy of Nitrogen Mustard in Treatment of Mycosis Fungoides: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety and Efficacy of Nitrogen Mustard in Treatment of Mycosis Fungoides (NCT00168064), a randomized phase 2 study. The primary outcome is Ratio of Response Rates Based on CAILS.

2. Clinical Question

3. Trial Design

This was a phase II, randomized, double-masked, parallel-group trial sponsored by Yaupon Therapeutics. The registry lists an enrollment of 260 and two treatment arms, each with 130.0 participants in the posted primary results. One arm received mechlorethamine 0.02% in the PG gel formulation; the other received mechlorethamine 0.02% compounded in Aquaphor. The trial is marked completed.

The population named in the official title was patients with stage I or IIA mycosis fungoides. The primary outcome population is identified as ITT. The registry gives the primary assessment timeframe as Day 1 and every subsequent visit during treatment, with treatment continuing for up to 12 months. It does not give further details about visit scheduling or the operational definition of the ITT population.

The registry records a start date of 2006-05 and a primary completion date of 2010-08. These dates provide context for the trial record, but do not establish how long any individual participant was followed. The record lists two secondary outcomes: SWAT assessment within up to 12 months using at least two consecutive observations over at least 4 weeks, and the percentage achieving at least 50% improvement in SWAT from baseline to the end of therapy. Further protocol details, including the planned sample-size justification, are not reported in the registry record.

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:

Analysis population for the primary outcome, as defined in the registry: ITT

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 Ratio of Response Rates Based on CAILS, measured over Assessment made at Day 1 and every subsequent visit during treatment.

The primary outcome is the ratio of response rates based on CAILS. Response means a reduction of at least 50% from baseline in the Composite Assessment of Index Lesion Severity after up to 12 months of treatment. The registry states that assessments were made at Day 1 and every subsequent visit during treatment. The primary analysis population is identified as ITT.

CAILS is used to define whether a participant met a specified skin-response threshold. Reducing the endpoint to a response or no-response outcome makes it possible to compare the proportions meeting that threshold in the two treatment groups. It also means that the outcome does not describe the full range of changes in CAILS: participants with different degrees of improvement can be counted alike if they meet the response criterion.

The registry reports the response counts and percentages for each arm, as well as a sponsor-reported ANCOVA analysis of the ratio of proportions. It does not report the ANCOVA model specification, any covariates, or a non-inferiority margin. Without those details, the ratio and its confidence interval can be described, but the record alone does not establish whether a formal non-inferiority criterion was met.

7. Results Posted in the Registry

• AP- Aquaphor Formulation Mechlorethamine-MCH (NM) 0.02%: 62.3% (81/130) versus 76.2% (99/130) on control; risk difference -13.8 percentage points (95% CI -24.9 to -2.8), risk ratio 0.82 (0.69–0.96); unadjusted two-proportion z-test p = 0.016.

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.

Percent of Participants Achieving at Least 50% Improvement of Severity Weighted Assessment Tool (SWAT)

• 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:
  • AP- Mechlorethamine-MCH (NM) 0.02% Compounded in Aquaphor: 60.0% (78/130) versus 60.8% (79/130) on control; risk difference -0.8 percentage points (95% CI -12.7 to 11.1), risk ratio 0.99 (0.81–1.20); unadjusted two-proportion z-test p = 0.899.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between AP- Mechlorethamine-MCH (NM) 0.02% Compounded in Aquaphor and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes broaden the assessment beyond the primary CAILS response definition. The posted SWAT responder comparison is close between arms in the reported group percentages, and its unadjusted comparison is not distinguishable from no difference on the recalculated analysis. This pattern does not reproduce the direction of the primary CAILS comparison as a clear statistical signal.

The registry also lists a SWAT outcome assessed within up to 12 months using at least two consecutive observations over at least 4 weeks. A numerical result for that outcome is not reported in the registry record, so it cannot add a direct quantitative comparison to the primary result. The other posted secondary outcome concerns achieving at least 50% improvement in SWAT from baseline to the end of therapy; it uses a different measure and should not be assumed to be interchangeable with CAILS response.

Secondary analyses are generally interpreted as supportive or exploratory unless the trial’s analysis plan establishes otherwise. When multiple outcomes are assessed, the chance of at least one apparently notable result can increase. The registry does not report whether the secondary outcomes were prespecified in a hierarchy, whether multiplicity was controlled, or how missing measurements were treated. The unadjusted secondary findings therefore carry less weight than a primary result interpreted under a clearly specified analysis plan. They should not be used alone to confirm or overturn the primary comparison.

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

• Severity-weighted Assessment Tool (SWAT) Within up to 12 Months by 2 or More Consecutive Observations Over at Least 4 Weeks
• Percent of Participants Achieving at Least 50% Improvement of Severity Weighted Assessment Tool (SWAT)

9. Statistical Interpretation

10. Statistical Methods Explained

Several statistical ideas help distinguish what the posted results say from what they cannot establish. The explanations below are general; the trial-specific analysis details are described only where the registry record reports them.

What does a response-rate comparison estimate?

A response rate is the proportion of participants in a group who meet the endpoint definition. Here, the definition is at least a 50% reduction from baseline in CAILS. Comparing response rates focuses on the chance of crossing that threshold. It does not show the distribution of all CAILS changes, and it does not distinguish degrees of improvement among participants who all qualify as responders.

The registry reports an arm-level response rate of 76.0% for PG and 62.0% for AP. A risk difference expresses the contrast between proportions on an absolute scale. A ratio of response rates expresses the contrast on a relative scale. These measures answer related but different questions, so their numerical values should not be compared as if they were the same quantity.

What does the sponsor ANCOVA result mean?

The registry identifies the sponsor’s method as ANCOVA and its parameter as a ratio of proportions. ANCOVA is a modeling approach that can compare outcomes while accounting for covariates, but the registry does not report which covariates were used, how the model was specified, or how missing observations were handled. Those details cannot be inferred from the method label alone.

The sponsor-reported ratio is 1.226, with a 95% confidence interval from 0.974 to 1.552. For a ratio, 1 represents equal proportions. The interval includes that value, so the estimate is compatible with equality as well as with values on either side of equality. The registry does not report a p-value for this analysis. Because the trial record labels the analysis as non-inferiority or equivalence but does not report the margin or decision rule, the label alone is insufficient to assess whether a non-inferiority or equivalence criterion was satisfied.

Why show an unadjusted recalculation?

A recalculated comparison uses the posted summary statistics to make a simple, unadjusted comparison between groups. It can help readers understand the observed data, but it is not a substitute for an analysis specified and conducted by the sponsor. It does not incorporate covariate adjustment or any other model features that are not contained in the group summaries. The recalculated primary comparison has an unadjusted p-value of 0.016. That p-value describes how unusual the observed or more extreme data would be under a particular no-difference model; it does not measure the size or clinical importance of the difference.

Confidence intervals add information that a p-value does not provide by displaying a range of estimates compatible with the data and analysis assumptions. They are not guarantees that the true value lies within the interval. Their width reflects statistical precision, while the interval’s location indicates the direction and scale of the estimated contrast.

Why be cautious with secondary outcomes?

Secondary outcomes can provide supporting or complementary information, but testing several outcomes creates more opportunities for apparently notable results to occur by chance. A multiplicity procedure can account for that issue, but whether one was planned or used is not reported in the registry record. The SWAT responder outcome has an unadjusted p-value of 0.899, and its recalculated interval includes no difference. The repeated-observation SWAT outcome is listed, but a numerical comparative analysis is not posted. Neither should be treated as an independent confirmation of the primary endpoint.

What do planning calculations represent?

Sample-size planning calculations illustrate how many participants per arm might be needed under stated assumptions to detect selected fractions of an observed effect at specified power levels. They are not the trial’s original design justification. The assumptions listed for the planning illustration are a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. The trial’s own original design assumptions, including its sample-size justification, are not reported in the registry record.

Power is the probability, under a specified model and effect, of detecting that effect using a chosen statistical test and threshold. A larger planned sample can increase the chance of detecting a given effect, while a smaller effect generally requires more participants to detect with the same power. These calculations depend on the assumptions being appropriate; they do not establish that the trial was adequately powered for its actual analyses.

11. Confidence Intervals and Precision

The sponsor-reported ANCOVA estimate is a ratio of proportions of 1.226. Its 95% confidence interval extends from 0.974 to 1.552. On the ratio scale, 1 corresponds to equal proportions. Because the interval includes 1, the sponsor estimate is not precise enough to rule out equal proportions under this analysis. It also spans values above 1, so the point estimate should not be read as a certain or exact description of the underlying difference.

A confidence interval gives a range of values compatible with the observed data and the assumptions of the analysis. It describes statistical uncertainty, not the range of individual participant outcomes, and it does not guarantee that the true effect lies inside the interval. A wider interval indicates less precision than a narrower interval, all else being equal. The registry does not provide a sponsor p-value for the ANCOVA result, so the reported interval is the key posted measure of its precision.

The recalculated primary comparison is unadjusted and has a p-value of 0.016. Its confidence interval for the risk difference excludes no difference, but the registry’s recalculated risk-difference endpoints have more than three decimal places and are not reproduced. That interval is based on posted summaries and does not reflect the sponsor’s ANCOVA analysis. The two intervals therefore should not be treated as interchangeable.

The secondary SWAT responder comparison is less conclusive: its recalculated interval includes no difference, and its unadjusted p-value is 0.899. The available precision should be interpreted alongside the endpoint definitions, the analysis methods, and the fact that multiple outcomes are listed. Details needed to assess uncertainty from missing data or the sponsor model specification 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 sample-size planning figures are expressed as participants per arm, not as total events. The listed planning assumptions 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 trial’s original design assumptions and sample-size justification are not reported in the registry record.

For the primary CAILS comparison, planning for 100% of the observed effect gives 174 participants per arm for 80% power and 232 participants per arm for 90% power. Planning for 75% of the observed effect gives 299 participants per arm for 80% power and 400 participants per arm for 90% power. Planning for 50% of the observed effect gives 649 participants per arm for 80% power and 869 participants per arm for 90% power.

The 80% power column describes the planned number per arm under which the calculation gives an 80% chance of detecting the selected effect, assuming the listed model and testing conditions. The 90% power column uses a higher chance of detection under those assumptions. For a fixed effect, aiming for higher power generally requires more participants. The rows also illustrate that smaller target effects need more participants: detecting a fraction of the observed effect requires a larger planned sample than detecting the full observed effect.

These are planning illustrations based on the observed effect, not observed power. They do not describe the trial’s original sample-size calculation, establish what the trial was designed to detect, or adjust for design and analysis details that are not reported. They should be read as conditional calculations rather than as evidence that the trial did or did not have adequate power.

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

Worked calculation for this trial

1. Control proportion p₀ = 0.762; treatment proportion p₁ = 0.623; average p̄ = 0.692.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6527 = 1.2793.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6453 = 0.5431.
4. n = (1.2793 + 0.5431)² / (-0.1385)² = 173.2, rounded up to 174 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 <- 0.761538461538461                          # control proportion
p1 <- 0.623076923076923                          # 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(168064)
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 = 174 (80% power), 232 (90% power)
 75% of observed effect: n per arm = 299 (80% power), 400 (90% power)
 50% of observed effect: n per arm = 649 (80% power), 869 (90% power)
power.prop.test: 173.2 
Simulated power at n = 174 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.

14. Analyses Reported by the Sponsor

The sponsor-reported analysis uses ANCOVA and estimates a ratio of proportions of 1.226, with a 95% confidence interval from 0.974 to 1.552. The registry does not report a p-value for this analysis. The interval includes 1, so the sponsor estimate does not rule out equal response proportions at the level of precision reported.

The recalculated comparison is a simple, unadjusted analysis of the posted response summaries. Its unadjusted p-value is 0.016, and its confidence interval excludes no difference. It is not the same analysis as the sponsor’s ANCOVA and does not account for any model adjustments. The trial record does not provide enough detail to reproduce the sponsor model or to explain the difference between the two analyses.

The registry does not identify a discrepancy between the posted summaries and the sponsor analysis. Even so, the recalculated result should be understood as a supplementary unadjusted comparison, not as a replacement for the sponsor’s reported method. The record labels the sponsor analysis as non-inferiority or equivalence, but the margin, decision rule, and model details needed to evaluate that designation are not reported in the registry record.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and multiple secondary outcomes, but it does not report a multiplicity adjustment or a testing hierarchy. When several outcomes are examined, interpreting each unadjusted result as though it were the only test can overstate the strength of evidence. The reported unadjusted p-values should therefore be understood as individual comparisons, not as a family-wise conclusion across all outcomes.

The primary population is identified as ITT, but the registry does not define exactly how that population was formed or state how participants with missing endpoint data were handled. It does not report the amount or pattern of missing primary or secondary outcome data, imputation methods, sensitivity analyses, or reasons for missingness. Missing-data handling can affect estimated response rates and their uncertainty, so the absence of these details limits assessment of robustness.

The secondary results do not list an analysis population. The registry does not report whether the secondary outcomes used the same participants or follow-up availability as the primary analysis. Nor does it report whether the sponsor ANCOVA adjusted for baseline values or other covariates. These omissions prevent a full appraisal of how the analytic population, covariate adjustment, and missing data may have influenced the reported estimates.

Interim analyses and any associated decision rules are not reported in the registry record. The record also does not provide the protocol or a statistical analysis plan, so it is not possible to establish from the posted information whether the endpoint definitions and analyses were changed after trial initiation.

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 non-serious adverse event terms

The registry reports serious adverse events in 0 participants among 128 at risk in the PG group and 0 participants among 127 at risk in the AP group. These figures describe the posted serious-event counts for the groups and should not be read as a complete account of all possible safety outcomes or of the timing and duration of observation.

Other adverse events were reported for 108 participants among 128 at risk in the PG group and 99 participants among 127 at risk in the AP group. The registry does not provide further breakdowns of these events in the information reported for this comparison. Death counts are not reported in the registry record.

Safety figures are descriptive in this record. They do not, by themselves, establish whether the formulations differ in safety, particularly without event definitions, severity and duration details, follow-up completeness, and a comparative safety analysis. The posted serious-event counts should be kept distinct from the broader other-adverse-event counts.

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

The posted results describe a randomized comparison of two 0.02% mechlorethamine formulations in patients with stage I or IIA mycosis fungoides. For the primary CAILS response definition, the reported response rate is higher in the PG arm than in the AP arm. The recalculated unadjusted comparison has a p-value of 0.016, while the sponsor ANCOVA estimate is a ratio of proportions of 1.226 with a 95% confidence interval from 0.974 to 1.552.

Those results do not provide a single, definitive interpretation of the formulations’ comparative effect. The unadjusted recalculation and sponsor ANCOVA are different analyses, and the sponsor interval includes 1. The registry does not report the ANCOVA model details, the non-inferiority margin, or the decision rule. The non-inferiority or equivalence label therefore cannot, on its own, establish that a formal criterion was met.

The endpoint is a threshold-based measure of CAILS response, not a comprehensive measure of all changes in disease severity, quality of life, or longer-term outcomes. The secondary SWAT outcomes use a different measure, and one listed outcome has no numerical comparative result in the registry record. These findings should not be generalized beyond the population and outcomes described in the record.

The statistical results are not medical advice and do not determine the appropriate treatment for an individual. They describe the estimates and uncertainty reported for this trial. Missing-data handling, detailed analysis plans, and some protocol features are not reported in the registry record, which limits how fully the results can be evaluated.

18. Why This Trial Matters Statistically

This trial illustrates why a clinical comparison should be read through both its endpoint definition and its analysis method. The primary outcome is a threshold-based response: a participant must reach at least a 50% reduction from baseline in CAILS. That makes the response rates straightforward to describe, while also leaving out the range of changes that do not cross the threshold and the differing degrees of improvement among responders.

It also illustrates how analyses of the same trial can answer different statistical questions. The posted arm rates support an unadjusted recalculation, while the sponsor reports an ANCOVA estimate of a ratio of proportions. The sponsor interval includes 1, whereas the recalculated unadjusted comparison has a p-value of 0.016. Reading both results carefully makes clear why a model-based sponsor analysis and an unadjusted summary-data comparison should not be conflated.

Finally, the trial highlights the importance of reporting the information needed to interpret claims of non-inferiority or equivalence. The estimate and confidence interval are available, but the margin, decision rule, model specification, and missing-data methods are not reported in the registry record. Those details are essential to a complete statistical appraisal. Secondary outcomes and safety counts add context, but their interpretation also depends on analysis plans and reporting details that are not available in the record.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record leaves important details unavailable for evaluating the results. It does not report a protocol-specific sample-size justification or the original planning assumptions, so the posted planning illustrations cannot establish what effect the trial was designed to detect. It does not report the ANCOVA model specification, including covariates, or the non-inferiority margin and decision rule associated with the analysis label.

Missing-data methods and the amount or pattern of missing outcome data are not reported. Although the primary population is identified as ITT, the record does not explain how that population was defined or how participants without outcome assessments were handled. It also does not identify the analysis population for the secondary outcomes.

The registry lists multiple secondary endpoints but does not report a multiplicity adjustment or a hierarchy among them. One listed SWAT outcome has no numerical comparative result in the record. The safety information provides counts of participants with serious and other adverse events, but further details about event definitions, severity, timing, and follow-up are not reported. Death counts are also not reported.

The official title refers to ointment formulations, while the PG arm is described as a gel in the arm title and description. This difference in wording is present in the registry record and should be kept in mind when describing the intervention. The posted results address specified response outcomes in the named population; they do not answer every question about disease severity, duration of response, or other outcomes not reported in the record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Safety and Efficacy of Nitrogen Mustard in Treatment of Mycosis Fungoides — NCT00168064.
• Registry record last updated 2012-10-31; status last verified 2012-10.
• Results first posted 2012-10-31.
• 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

The registry record describes a completed, randomized, double-masked, parallel-group phase II trial sponsored by Yaupon Therapeutics. It compares mechlorethamine 0.02% in a PG gel formulation with mechlorethamine 0.02% compounded in Aquaphor in patients with stage I or IIA mycosis fungoides.

The primary endpoint is the ratio of response rates based on at least a 50% reduction from baseline in CAILS after up to 12 months of treatment. The posted arm response rates are 76.0% for PG and 62.0% for AP. The sponsor ANCOVA result is a ratio of proportions of 1.226, with a 95% confidence interval from 0.974 to 1.552. The registry also lists SWAT secondary outcomes and adverse-event counts.

Important details, including the original sample-size justification, missing-data methods, multiplicity approach, ANCOVA specification, and non-inferiority margin, are not reported in the registry record.