A Randomized Phase III Study to Assess the Effect of a Longer Duration of Consolidation Treatment With Nilotinib on TFR in CP CML.: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Randomized Phase III Study to Assess the Effect of a Longer Duration of Consolidation Treatment With Nilotinib on TFR in CP CML. (NCT01743989), a randomized phase 3 study. The primary outcome is Percentage of Participants Who Remained in Treatment Free Remission (TFR) Without Molecular Relapse 12 Months After Entering the TFR Phase.

2. Clinical Question

The trial asked whether a longer period of nilotinib consolidation was associated with a different treatment-free remission (TFR) rate after participants stopped treatment. The primary endpoint was the percentage who remained in TFR without molecular relapse 12 months after entering TFR. The comparison was between participants assigned to 24 months of consolidation and those assigned to 36 months.

The endpoint is timed from entry into TFR, rather than from the start of study treatment. Consequently, the primary assessment occurred 36 months after treatment start in the 24-month consolidation group and 48 months after treatment start in the 36-month group. The registry record describes molecular relapse using loss of major molecular response or confirmed loss of MR4.0. Participants who dropped out early during TFR were counted as unsuccessful for this endpoint.

3. Trial Design

This was a randomized, open-label, parallel-group phase III study sponsored by Novartis Pharmaceuticals. The registry record lists an enrollment of 620 and a completed study status. The recorded study start was 2013-04-15, and the primary completion date was 2020-07-08. Participants in the two randomized treatment groups received the same drug and dose, but for different consolidation durations before entering TFR.

The 24-month treatment group received nilotinib 300 mg twice daily for 24 months and then entered a 36-month TFR phase. The 36-month treatment group received the same dose for 36 months and then entered a 24-month TFR phase. The primary outcome was assessed 12 months after TFR entry, corresponding to different elapsed times from treatment start in the two groups. The registry record also describes participants who received 24 months of treatment but did not achieve a sustained molecular response after that period; they were not randomized.

Random allocation and a parallel-group model define the comparison as one between separately assigned groups followed over the study period. Open-label means the study was not masked. The registry record does not report the randomization procedure, allocation concealment, stratification factors, or the rationale for selecting the treatment durations. The original sample-size justification and the trial's own design assumptions are not reported in the registry record.

The primary endpoint denominators were 119 participants in the 24-month group and 104 in the 36-month group. These are participants who entered TFR and contributed to the stated endpoint population, rather than the full enrollment count. The registry record lists 23 outcome measures and 21 posted secondary outcomes. It describes cumulative incidence outcomes for molecular response measures during pre-randomization and post-randomization phases, in addition to the primary TFR outcome.

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: For one participant randomized to Nilotinib 36-month treatment arm, the informed consent was not obtained prior to any study specific procedure. This participant discontinued the study before entering the TFR phase.

Treatment Phase

Reasons for not completing this period:

TFR Phase

Reasons for not completing this period:

Re-treatment Phase

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All enrolled subjects who entered the TFR phase

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 Percentage of Participants Who Remained in Treatment Free Remission (TFR) Without Molecular Relapse 12 Months After Entering the TFR Phase, measured over 12 months after entering the TFR phase, which is after 36 months from study treatment start for Nilotinib 24-month treatment arm and after 48 months from study treatment start for Nilotinib 36-month treatment arm.

The primary endpoint is the percentage of participants who remained in TFR without molecular relapse 12 months after entering TFR. Its denominator is the number who entered TFR; participants who restarted nilotinib or met the registry's molecular relapse definition did not meet the endpoint. Early dropout during TFR was also classified as unsuccessful TFR. The registry record defines molecular relapse as loss of major molecular response, or confirmed loss of MR4.0 across three consecutive tests at three consecutive visits during TFR.

The endpoint captures a defined period without treatment after consolidation, and therefore addresses the study's stated question about TFR following different consolidation durations. It is a binary outcome at a specified assessment point: each participant is classified as having or not having remained in TFR without molecular relapse for the required period. The reported percentage is not the proportion of all enrolled participants; it is calculated among those who entered TFR.

The registry record defines major molecular response as at least a 3.0 log reduction in BCR-ABL transcripts compared with the standardized baseline or BCR-ABL at or below 0.1%. MR4.0 is defined as detectable disease at or below 0.01% BCR-ABL, or undetectable disease in cDNA with at least 10,000 ABL transcripts. These definitions specify the molecular thresholds underlying the endpoint rather than a broader clinical outcome.

7. Results Posted in the Registry

• Nilotinib 36-month Treatment: 37.5% (39/104) versus 31.9% (38/119) on control; risk difference 5.6 percentage points (95% CI -7.0 to 18.1), risk ratio 1.17 (0.82–1.69); unadjusted two-proportion z-test p = 0.383.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a binary percentage: participants either remained in TFR without molecular relapse for the stated 12-month period or did not. The registry describes the percentage as the number meeting that outcome divided by the number who entered TFR, multiplied by 100. This type of endpoint is commonly analyzed as a proportion within each group. General statistical explanation: a proportion estimates the frequency of an outcome in the population from which the participants were drawn, while the observed percentage is subject to sampling variability.

What does the confidence interval describe?

The registry reports 95% confidence intervals for each group's percentage and states that they were calculated using the exact Clopper-Pearson method. General statistical explanation: a confidence interval gives a range of values compatible with the observed data under a specified method and assumptions. It expresses uncertainty in an estimate; it is not a guarantee that the true value lies within the interval, nor does it describe the range of outcomes for individual participants. The exact binomial method is one approach for estimating uncertainty around a proportion, particularly when the outcome is treated as a count of participants meeting a binary definition.

The two group-specific intervals answer a different question from a confidence interval for the difference between groups. A visual comparison of separate intervals is not a substitute for a direct comparative analysis. The registry reports a chi-squared test for the randomized groups, and the unadjusted recalculation gives the same p-value. A p-value describes how unusual data at least as incompatible with a null hypothesis would be if that hypothesis and the test assumptions held. It is not the probability that the null hypothesis is true, nor a measure of clinical importance.

How should the group comparison be read?

The registry identifies the analysis as a superiority comparison and reports a chi-squared p-value of 0.383. General statistical explanation: a chi-squared test can compare observed outcome counts across groups with counts expected under a null hypothesis of no association between group and outcome. The test does not adjust for covariates unless an adjusted model is specified; no adjusted model is described in the registry record. A nonsmall p-value does not prove the null hypothesis, and it does not demonstrate equivalence. Interpretation also depends on the effect estimate and its uncertainty, the population analyzed, and the study design.

The posted comparison also contains recalculated relative measures, but the registry's reported analysis is the chi-squared test. Such recalculations are unadjusted comparisons based on the posted summary statistics; they do not replace the sponsor's analysis. The primary comparison is not described as adjusted for baseline characteristics. No covariate adjustment method, subgroup testing plan, or alternative model is reported in the registry record.

What do the endpoint rules contribute?

The molecular relapse definition specifies which events count as failure to remain in TFR. It includes loss of major molecular response or confirmed loss of MR4.0, with confirmation defined by three consecutive tests below the MR4.0 threshold at three consecutive visits during TFR. Early dropout during TFR is also counted as unsuccessful. These rules make the binary outcome more explicit, but the registry record does not describe how testing schedules were implemented in practice or how results that did not meet the confirmation rule were handled.

General statistical explanation: endpoint definitions can affect an estimate because changing which observations count as events or failures changes the outcome being measured. Treating early dropout as unsuccessful avoids counting those participants as confirmed successes, but the registry's approach to other missing measurements is not reported. The stated endpoint population is all enrolled subjects who entered TFR, and no separate intention-to-treat designation is given.

Why can over-dispersion matter?

General statistical explanation: over-dispersion occurs when observed variability exceeds what a statistical model assumes. For a binary endpoint, this can arise when outcomes are more heterogeneous than a simple model represents. If relevant, unaccounted over-dispersion can make standard errors too small and confidence intervals or tests too optimistic. The registry record does not report an over-dispersion assessment or a model that would permit one to be evaluated. The stated Clopper-Pearson intervals and chi-squared comparison should therefore be understood as the methods named in the record, without assuming additional adjustments.

The record lists one reported comparative analysis for the primary comparison and multiple secondary outcome measures. It does not report an interim-analysis plan, a multiplicity adjustment, or a hierarchy for interpreting the secondary outcomes. General statistical explanation: when many hypotheses are tested, the chance of at least one apparently notable result can increase unless the analysis plan accounts for multiple testing. Without a reported plan, the secondary outcomes should not be treated as if their interpretation were protected by an adjustment that is not documented.

10. Confidence Intervals and Precision

The registry reports 95% confidence intervals of 23.7% to 41.1% for the 24-month group and 28.2% to 47.5% for the 36-month group. These ranges show that each estimated percentage has substantial uncertainty. Their overlap is consistent with the reported comparison not showing a statistically clear difference, although comparing separate group intervals is not itself the formal test of the difference.

The direct unadjusted risk-difference interval includes zero and extends in both directions. In general, an interval for a difference that includes zero is compatible with no group difference at the confidence level used, as well as with differences in either direction. It does not establish that the groups are the same. The registry's reported p-value is 0.383, and the unadjusted recalculation gives the same value.

Precision is affected by the number of participants contributing to the endpoint and by how often the outcome occurs. The primary denominators are 119 and 104 participants. The intervals are therefore estimates based on those TFR entrants, not on the entire enrollment. The registry uses the exact Clopper-Pearson method for the group-specific intervals. No confidence interval for an adjusted effect is reported in the registry record.

These intervals quantify statistical uncertainty under the stated methods; they do not account for every possible source of uncertainty, such as limitations in endpoint ascertainment or unreported analysis choices. They also do not establish the range of effects that would be clinically important. The registry record does not report a threshold for a clinically important difference.

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 illustrations based on the observed effect, not the trial's original design calculations. They assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set at the stated fraction of the observed effect. The registry's original sample-size assumptions and justification are not reported in the registry record.

For this binary outcome, the displayed planning counts are participants per arm. The note that time-to-event calculations give total events applies to time-to-event outcomes; these figures are not total events. Under the stated assumptions, planning for 100% of the observed effect corresponds to 1,147 participants per arm for 80% power and 1,535 participants per arm for 90% power. Planning for 75% of the observed effect corresponds to 2,020 participants per arm for 80% power and 2,704 participants per arm for 90% power. Planning for 50% of the observed effect corresponds to 4,501 participants per arm for 80% power and 6,025 participants per arm for 90% power.

In general, the 80% power column means the planned study would have an 80% chance of rejecting the specified null hypothesis if the assumed effect and other planning assumptions were true. The 90% column expresses the same idea at a higher target probability. Power is a planning property tied to a particular assumed effect, significance level, allocation, and model; it is not the probability that the observed result is correct. Smaller effects require more participants to detect under otherwise comparable assumptions, which is why the planning counts increase as the assumed fraction of the observed effect gets smaller. These figures are not observed power.

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 = 0.319; treatment proportion p₁ = 0.375; average p̄ = 0.347.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6733 = 1.3196.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6721 = 0.5657.
4. n = (1.3196 + 0.5657)² / (0.0557)² = 1,146.7, rounded up to 1,147 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 <- 0.319327731092437                          # control proportion
p1 <- 0.375                          # 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(1743989)
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 = 1,147 (80% power), 1,535 (90% power)
 75% of observed effect: n per arm = 2,020 (80% power), 2,704 (90% power)
 50% of observed effect: n per arm = 4,501 (80% power), 6,025 (90% power)
power.prop.test: 1146.7 
Simulated power at n = 1,147 per arm: 0.804 (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 a chi-squared superiority analysis for the two randomized groups, with p = 0.383. The unadjusted recalculation from the posted group summaries also gives p = 0.383. This correspondence means the recalculated test result matches the reported p-value at the precision shown; the recalculation is not a replacement for the sponsor's analysis.

The registry record does not report an adjusted comparative model for this endpoint. The recalculation is a simple unadjusted comparison based on the posted summary statistics, so it should not be interpreted as accounting for covariates or other design features not included in those summaries. The record's named method remains the chi-squared test. No discrepancy between the reported analysis and recalculation is indicated in the record.

Both analyses should be read alongside the group percentages and their confidence intervals. The p-value does not quantify the size or clinical importance of a difference, and it does not show that the treatment durations are equivalent. The registry record does not report an equivalence or non-inferiority margin.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists 23 outcome measures and 21 posted secondary outcomes, including cumulative incidence outcomes for MMR, MR4.0, and MR4.5 during different study phases. It reports one comparative analysis for the primary comparison. A multiplicity adjustment, testing hierarchy, or strategy for controlling error across the secondary outcomes is not reported in the registry record. Without such information, multiple secondary findings should be interpreted cautiously rather than treated as a single confirmatory family with documented error control.

The primary endpoint population is described as all enrolled subjects who entered the TFR phase. This is a defined population for the endpoint, but the registry record does not label the analysis as intention-to-treat or provide further details about the analysis populations. It does not report how participants who were randomized but did not enter TFR were handled in the primary analysis, beyond defining the endpoint denominator as those who entered TFR.

The endpoint definition explicitly treats early dropout during TFR as unsuccessful. That rule addresses one circumstance, but the approach to other missing observations, missed molecular tests, or incomplete follow-up is not reported in the registry record. No imputation method, sensitivity analysis for missing data, or per-protocol analysis is described. As a result, the available record does not permit a fuller assessment of how incomplete data may have affected the estimate.

The registry record also does not report interim analyses, stopping rules, or the prespecified statistical analysis plan. These omissions do not establish that such procedures were absent; they mean their details are not reported in the registry record. The posted primary comparison can be interpreted using the reported endpoint definition and named test, while conclusions about unreported analysis choices remain limited.

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: atrial fibrillation (2.5%), cellulitis (1.7%), acute coronary syndrome (1.7%), myocardial infarction (1.7%), osteoarthritis (1.7%).

For the 24-month treatment group, the registry lists 33 participants with a serious adverse event among 120 at risk. For the 36-month group, it lists 27 among 118 at risk. The not-randomized group has 76 among 381 at risk. Across the total safety population, 136 participants with a serious adverse event are listed among 619 at risk. These safety denominators differ from the primary endpoint denominators, which include participants who entered TFR.

The record lists one death in each randomized group, three in the not-randomized group, and five in total. The most frequently listed serious adverse-event terms include atrial fibrillation at 2.5%, followed by cellulitis, acute coronary syndrome, myocardial infarction, osteoarthritis, and dizziness, each at 1.7%. These are registry-reported event frequencies; they do not establish that an event was caused by treatment.

Safety counts should be read with their corresponding at-risk denominators and the study's group definitions. The not-randomized participants are a distinct group and should not be combined conceptually with the randomized comparison when interpreting treatment-duration effects. The registry record does not provide additional context in these figures about event timing, severity beyond the serious-event classification, or causal attribution.

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

The posted primary results describe the proportion of TFR entrants who remained in TFR without molecular relapse 12 months after entry. The estimated percentages were 31.9% after 24 months of consolidation and 37.5% after 36 months. The registry's comparative chi-squared test reports p = 0.383, and the unadjusted recalculation gives the same value. The comparison does not establish a statistically clear difference between the two groups.

This result does not demonstrate that the two durations are equivalent. A failure to establish a difference is not evidence that differences of meaningful size have been ruled out. The record does not report an equivalence or non-inferiority framework or a clinical threshold for interpreting a difference. The confidence intervals are broad, and the comparative interval includes zero as well as values on either side.

The finding is specific to the endpoint definition, its 12-month assessment after TFR entry, and the participants who entered TFR. It does not describe all enrolled participants, those who did not enter TFR, longer-term TFR outcomes, or outcomes under other consolidation schedules. The endpoint also classifies early TFR dropout as unsuccessful, as specified in the registry record.

These statistics do not determine which treatment duration is appropriate for an individual. They describe the posted study comparison and its precision. Interpretation beyond that comparison would require evidence and context not reported in the registry record, including the handling of missing measurements, the rationale for the study's assumptions, and whether differences of particular sizes would be clinically important.

17. Why This Trial Matters Statistically

This trial presents a useful statistical distinction between a numerical difference and evidence that groups differ. The reported TFR percentages are not identical, but the between-group test has p = 0.383, and the unadjusted comparative interval includes zero. The estimates therefore do not support a statistically clear separation, while still leaving uncertainty about the size and direction of the underlying difference.

The endpoint also illustrates why population and time origin matter. The comparison concerns participants who entered TFR, and the 12-month outcome is measured from TFR entry. Because the consolidation durations differed, that assessment occurred 36 months after treatment start in one randomized group and 48 months after treatment start in the other. Reading the percentages without these definitions could lead to an incorrect understanding of who was counted and when the outcome was assessed.

The record's sample-size illustrations connect effect assumptions to the number of participants needed for planning. They show that aiming to detect a smaller fraction of the observed effect requires more participants under the stated assumptions. These are planning calculations rather than a retrospective measure of trial success, and they do not disclose the original design rationale.

Finally, the endpoint's rules for molecular relapse and early dropout demonstrate how operational definitions shape a binary outcome. The primary comparison is interpretable as reported, but missing-data procedures, multiplicity handling, and other analysis-plan details are not reported in the registry record. Recognizing those limits helps distinguish what the posted statistics answer from questions they do not resolve.

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

The primary comparison is restricted to participants who entered TFR, with denominators of 119 and 104. It is not a comparison of all enrolled participants. The registry record does not give further detail on the characteristics of those who did or did not enter TFR, so the posted endpoint results alone cannot describe how entry into TFR may relate to the randomized comparison.

The study was open-label. The primary endpoint is based on molecular response criteria and includes a specific early-dropout rule, but the registry record does not report assessment masking or additional safeguards against outcome-assessment bias. It also does not describe the timing and completeness of molecular testing beyond the stated relapse definition.

The registry record does not report the original sample-size justification, an interim-analysis plan, an adjustment for multiple secondary outcomes, a prespecified hierarchy, or detailed missing-data methods. It does not report an equivalence or non-inferiority margin. These absences limit assessment of the original design rationale and the interpretation of results beyond the reported superiority test; they do not establish that the procedures were absent.

The primary endpoint is assessed at 12 months after TFR entry. It does not itself characterize later outcomes. The record also includes a group of participants who were not randomized because they did not achieve a sustained molecular response after 24 months of treatment. Their existence is important when considering the endpoint's selected population, but their results should not be substituted for the randomized comparison.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Randomized Phase III Study to Assess the Effect of a Longer Duration of Consolidation Treatment With Nilotinib on TFR in CP CML. — NCT01743989.
• Registry record last updated 2021-07-15; status last verified 2021-06.
• Results first posted 2021-07-15.
• 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, randomized, open-label, parallel-group phase III study of different nilotinib consolidation durations in Philadelphia chromosome-positive CML in chronic phase. The randomized groups received 24 or 36 months of nilotinib 300 mg twice daily before entering TFR. The primary outcome was the percentage remaining in TFR without molecular relapse 12 months after TFR entry.

The reported percentages were 31.9% after 24 months of treatment and 37.5% after 36 months, with a reported chi-squared p-value of 0.383. The unadjusted recalculation gives the same p-value. The registry record also lists secondary molecular-response outcomes and safety results. Details about the original sample-size rationale, multiplicity, interim analyses, and several missing-data procedures are not reported in the registry record.