A Study of Oral Asciminib Versus Other TKIs in Adult Patients With Newly Diagnosed Ph+ CML-CP: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Oral Asciminib Versus Other TKIs in Adult Patients With Newly Diagnosed Ph+ CML-CP (NCT04971226), a randomized phase 3 study. The primary outcome is Percentage of Participants With Major Molecular Response (MMR) at Week 48 - Ascimimib vs. Investigator Selected TKI.

2. Clinical Question

The study addresses an efficacy question in a specific setting: newly diagnosed, previously untreated Ph+ CML-CP. It does not compare asciminib with just one alternative medicine. The comparator combines participants assigned to investigator-selected TKIs, so the overall contrast is between asciminib and that selection strategy rather than a separate primary comparison with each listed TKI.

The registry reports 201 participants in the asciminib group and 204 in the investigator-selected TKI group for the primary outcome. The posted outcome counts are 136 and 100 participants with MMR, respectively. Those counts and the reported risk-difference estimates describe the Week 48 result; they do not by themselves explain how treatment choices within the comparator group were made or how results might differ for an individual TKI. The registry record does not report those details in the primary comparison.

3. Trial Design

This is a Phase III, multicenter, open-label, randomized, parallel-group study. Participants were randomized 1:1 to asciminib or investigator-selected TKIs. The registry lists an enrollment of 405 and a study start date of 2021-10-06. Its status is active, not recruiting; the recorded primary completion date is 2023-11-28. The registry record does not report the original sample-size justification or the trial's detailed design assumptions.

Asciminib is described as 80 mg once daily under fasting conditions. The investigator-selected options are imatinib 400 mg once daily with food, nilotinib 300 mg twice daily under fasting conditions, dasatinib 100 mg once daily with or without a meal, and bosutinib 400 mg once daily. The ongoing treatment period applies to randomized participants. The study also includes an optional treatment-free remission period for participants meeting eligibility criteria and a treatment re-initiation period for participants who relapse after a treatment-free remission attempt.

Open-label means the study is not masked, according to the registry. This feature describes the treatment-awareness arrangement, not the direction or reliability of the outcome result. Randomization supports a planned comparison between assigned groups, while the registry's primary population includes all participants to whom study treatment was assigned by randomization. The registry record does not report further details about allocation concealment or the operational conduct of randomization.

The primary outcome is MMR at Week 48, and the record lists 33 outcome measures and 2 reported analyses. It also lists 31 posted secondary outcomes. The registry record does not report whether all secondary outcomes were prespecified in the same form, how their analyses were prioritized, or whether the reported study status indicates that all follow-up and analyses are complete.

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: Prior to randomization, the Investigator, in consultation with the participant, considering current treatment paradigm and participant characteristics - and comorbidities, made a selection of imatinib or 2G TKI (nilotinib, dasatinib, or bosutinib) to be used if the participant is randomized to the comparator arm. The enrolment into the strata of imatinib versus 2G TKI was managed by IRT to be approximately 50% versus 50%.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Full Analysis Set (FAS) comprised of all participants to whom study treatment has been assigned by randomization.

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 With Major Molecular Response (MMR) at Week 48 - Ascimimib vs. Investigator Selected TKI, measured over At 48 weeks.

The primary endpoint is the percentage of participants with major molecular response (MMR) at Week 48. Molecular response is assessed from BCR-ABL1 transcript levels measured by real-time quantitative polymerase chain reaction. MMR is defined as a BCR-ABL1/ABL ratio at or below 0.1% on the international scale, corresponding to at least a 3 log reduction from a standardized baseline value.

This endpoint turns a molecular measurement into a binary outcome: each participant either meets the stated MMR threshold at the specified assessment or does not. A binary endpoint can be summarized as the number meeting the threshold relative to the number analyzed, then compared between randomized groups. The registry reports 136 of 201 participants with MMR in the asciminib group and 100 of 204 in the investigator-selected TKI group.

Using a defined molecular threshold makes the outcome interpretable as attainment of a specified response level at a fixed time point. It is not the same as measuring the full trajectory of transcript levels, the time taken to reach response, or outcomes beyond Week 48. The registry lists other molecular response measures among the secondary outcomes, including MMR at Week 96 and MR4.0 and MR4.5 at scheduled assessments. The primary endpoint alone does not describe those other outcomes or establish the longer-term pattern of response.

7. Results Posted in the Registry

• Asciminib (All Asciminib ): 67.7% (136/201) versus 49.0% (100/204) on control; risk difference 18.6 percentage points (95% CI 9.2 to 28.1), risk ratio 1.38 (1.17–1.64); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: participants are classified according to whether they met the MMR definition at Week 48. For such an endpoint, the most direct summaries are the event count and group size, followed by the event proportion in each group. The posted counts are 136 of 201 for asciminib and 100 of 204 for investigator-selected TKIs. A risk difference compares the two group proportions by subtraction. It is expressed on the proportion scale in the unadjusted re-calculation, while the registry's reported Mantel-Haenszel estimate is presented as 18.88 with its interval limits.

The unadjusted re-calculation uses the posted arm-level summaries without adjustment for other variables. It gives a risk difference of 0.186 and a 95% confidence interval from 0.092 to 0.281. This calculation is a simple summary comparison; it is not a replacement for the sponsor's reported analysis. The registry's reported method is Mantel-Haenszel, with a risk difference of 18.88, a 95% confidence interval from 9.59 to 28.17, and p < 0.001. The registry record does not give enough detail to describe further features of the sponsor's analysis beyond that method and the reported result.

What does a risk difference convey?

A risk difference compares the probability of the outcome between groups on an absolute scale. Here, the outcome is meeting the MMR threshold at Week 48. The estimate describes the observed group contrast; it does not explain the biological process, establish the cause of an individual participant's outcome, or capture the timing of response. It is also distinct from a relative measure, which would express the comparison as a ratio rather than an absolute difference.

What does the confidence interval add?

A confidence interval describes uncertainty around an estimated effect under the statistical method's assumptions. In repeated applications of a procedure, intervals constructed in the stated way would cover the underlying parameter at the stated rate, assuming the model and sampling conditions apply. For this result, the reported intervals communicate a range of effect estimates compatible with the respective analysis. Their width reflects precision: narrower intervals indicate greater precision, while wider intervals indicate more uncertainty. Neither interval represents the range of individual participant outcomes.

How should the p-value be read?

The reported p-value is <0.001. In general, a p-value assesses how unusual the data, or more extreme data, would be under a null hypothesis and the statistical model used. It does not quantify the probability that the treatment groups are truly equal, and it does not measure clinical importance. Its interpretation depends on the analysis plan and on which comparisons were considered, including any multiplicity control. The registry record does not report a multiplicity procedure for the primary and secondary outcomes.

Why might analysis methods matter?

Different analysis procedures can yield estimates or uncertainty limits that are not numerically identical, even when they address the same outcome and groups. The registry presents an unadjusted re-calculation from the posted counts and a Mantel-Haenszel result. The reason for their numerical difference is not reported in the registry record, so no specific explanation should be assigned to it. The sponsor's reported analysis is the study's reported inferential result; the unadjusted calculation offers a transparent comparison derived from the posted summaries.

In general, binary-outcome analyses can also be affected by missing outcome assessments, how the analysis population is defined, and the handling of multiple planned comparisons. The Full Analysis Set is defined in the record as all randomized participants assigned study treatment. The registry record does not report the missing-data method or enough detail to determine how participants without an observed Week 48 assessment were handled. The observed counts should therefore be read together with that reporting limitation, not as evidence about unreported participant outcomes.

10. Confidence Intervals and Precision

Precision is reflected in the range of values around an effect estimate, commonly expressed as a confidence interval. The unadjusted re-calculation gives a risk difference of 0.186, with a 95% confidence interval from 0.092 to 0.281. The sponsor-reported Mantel-Haenszel risk difference is 18.88, with a 95% confidence interval from 9.59 to 28.17. Each interval belongs to its corresponding analysis; the two presentations should not be combined or treated as interchangeable.

Both intervals exclude the no-difference value on their respective presentation scales. This indicates that the reported estimates are not close to that reference value within the intervals, under the methods used. It does not mean that the true effect is known exactly, nor does it capture uncertainty from aspects not represented in the calculation, such as unreported missing-data assumptions or uncertainty about generalizing to other populations.

Confidence intervals are useful because they show both the direction and the range of estimates compatible with the analysis. They are not prediction intervals for individual participants, and they do not state where a future study result must fall. The registry record does not report a prespecified threshold for a clinically important difference, so the intervals alone cannot classify the observed contrast as clinically important or unimportant.

The posted group sizes and event counts determine the information available for the binary comparison, but precision also depends on the analysis method and its assumptions. Since the sponsor result and the unadjusted re-calculation are presented differently, each should be interpreted in its own form. The registry record does not give an explanation for the numerical differences between those presentations.

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 illustrative calculations based on the observed effect, not the trial's original design assumptions. They use a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. For these binary-outcome calculations, the figures are participants per arm. In time-to-event calculations, figures are total events, but that does not apply to this primary binary outcome.

The column for 80% power gives the number of participants per arm required by the planning calculation for an 80% chance of detecting the specified effect under its assumptions, if that effect is the true one. The 90% power column gives the corresponding planning number for a 90% chance. These are design concepts: power is the probability of obtaining a statistically significant result under a specified effect and model, not the probability that a result is true.

The scenarios use 100%, 75%, and 50% of the observed effect. The corresponding participant figures per arm are 109 and 145 for the full observed effect, 197 and 263 for 75% of the observed effect, and 449 and 600 for 50% of the observed effect, for the 80% and 90% power columns respectively. Smaller effects require more participants to detect under the same planning assumptions, which is why the listed figures increase as the assumed effect becomes smaller.

These figures are not observed power. The trial's original sample-size justification and design assumptions are not reported in the registry record. The planning calculations do not establish that the study was designed around the observed effect, and they do not show what effect the original design was intended to detect. A planning curve should be read in the same way: as an illustration of how required sample size changes with the assumed effect, not as a description of the trial's observed performance.

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 = 0.490; treatment proportion p₁ = 0.677; average p̄ = 0.583.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6972 = 1.3665.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6846 = 0.5762.
4. n = (1.3665 + 0.5762)² / (0.1864)² = 108.6, rounded up to 109 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 <- 0.490196078431373                          # control proportion
p1 <- 0.676616915422886                          # 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(4971226)
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 = 109 (80% power), 145 (90% power)
 75% of observed effect: n per arm = 197 (80% power), 263 (90% power)
 50% of observed effect: n per arm = 449 (80% power), 600 (90% power)
power.prop.test: 108.6 
Simulated power at n = 109 per arm: 0.798 (20000 replicates)

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

13. Analyses Reported by the Sponsor

The registry reports the sponsor's primary analysis as a Mantel-Haenszel risk difference of 18.88, with a 95% confidence interval from 9.59 to 28.17 and p < 0.001. The analysis is labelled superiority. That is the reported study analysis and should be distinguished from the unadjusted re-calculation based directly on the posted group counts.

The re-calculation gives a risk difference of 0.186, with a 95% confidence interval from 0.092 to 0.281 and p < 0.001. It is a simple unadjusted check from the two group summaries, not an alternative reconstruction of the sponsor's full analysis. The numerical formats and interval limits differ. The registry record does not explain the reason for that difference, so it should not be attributed to a particular adjustment or modeling choice beyond the named Mantel-Haenszel method.

Both reported analyses have estimates in the same direction and both report p < 0.001. This does not make their numerical results interchangeable. The sponsor's reported analysis remains the primary account of the study's inferential comparison, while the re-calculation helps readers connect the posted counts with a straightforward unadjusted contrast. No discrepancy flag is reported in the registry record.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists 33 outcome measures, 31 posted secondary outcomes, and 2 reported analyses. The primary endpoint is MMR at Week 48. Secondary outcomes include MMR at Week 96, time to discontinuation of study treatment due to adverse events, MMR at scheduled data-collection time points, and MR4.0 and MR4.5 outcomes at scheduled time points. This breadth matters statistically because examining multiple outcomes or time points can increase the chance that at least one comparison appears statistically significant by chance when no multiplicity strategy is used.

The registry record does not report whether or how the trial controlled multiplicity, which outcomes were included in any testing hierarchy, or how the two reported analyses relate to a prespecified testing plan. It also does not report whether interim analyses were planned or conducted. These details should not be inferred from the number of outcomes or analyses listed. The reported primary p-value can be described as posted, but the registry record does not establish from the information available whether it was adjusted for multiple testing.

The Full Analysis Set comprises all participants to whom study treatment was assigned by randomization. This definition identifies the population named for the primary outcome. The record does not report the detailed analysis population for each secondary outcome, nor whether every randomized participant had an observed Week 48 assessment. It also does not report the approach to missing outcome data, such as assumptions, imputation, or sensitivity analyses. As a result, the effect of missing assessments on the reported estimate cannot be evaluated from the registry record.

In general, randomization and an analysis population that retains randomized participants can help preserve the planned comparison, but the interpretation depends on how outcomes after randomization are handled. The registry record's population definition alone does not answer how discontinuations, incomplete assessments, or other post-randomization events were treated in the efficacy analysis. The statistical analysis plan and protocol details needed to assess those choices are not reported in the registry record.

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: pneumonia (2.9%), covid-19 (2.0%), pleural effusion (2.0%), atrial fibrillation (1.0%), abdominal pain upper (1.0%).

The serious adverse event figures report affected participants and participants at risk for each listed group. For asciminib, 22 of 200 participants were recorded as affected by a serious adverse event. For all comparators, the corresponding figures are 32 of 201. The registry also lists imatinib separately, with 12 of 99 affected, and the second-generation TKI group, with 20 of 102 affected. These are group-level counts, not a complete description of event timing, duration, severity beyond the serious classification, or causal attribution.

The listed serious adverse event terms and percentages are pneumonia at 2.9%, COVID-19 at 2.0%, pleural effusion at 2.0%, atrial fibrillation at 1.0%, abdominal pain upper at 1.0%, and melaena at 1.0%. The registry reports zero deaths in each listed safety group. Those entries describe the posted safety record; they do not by themselves establish whether an event was caused by treatment or permit a full benefit-risk assessment.

Safety comparisons require care because group-level counts do not show the distribution of exposure time or all relevant participant characteristics. The registry record does not report those details in the safety summary, and the figures should not be read as a formal comparative test of risk. They are descriptive counts and percentages in the listed groups.

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

The primary result describes MMR status at Week 48 among participants in the Full Analysis Set. The posted counts are 136 of 201 in the asciminib group and 100 of 204 in the investigator-selected TKI group. The unadjusted re-calculation gives a positive risk difference, and the sponsor-reported Mantel-Haenszel analysis also reports a positive risk difference with p < 0.001. These statistics describe the observed group comparison for the defined endpoint and assessment time.

The result does not establish the effect for each comparator drug separately. The investigator-selected arm pools imatinib, nilotinib, dasatinib, and bosutinib, and the primary comparison is against that group as a whole. Nor does the Week 48 endpoint describe time to response, response durability, outcomes during an optional treatment-free remission period, or results after treatment re-initiation. Those questions require their own outcome data and analyses.

A statistically significant result does not, by itself, indicate the clinical importance of an effect. The registry record does not report a threshold for a clinically meaningful difference, and the reported estimates do not describe individual-level benefit or predict any one participant's outcome. Similarly, the serious adverse event counts are descriptive and do not establish causal attribution or a complete comparative safety profile.

Interpretation is also bounded by reporting gaps. Missing-outcome handling, multiplicity procedures, interim analyses, and detailed analysis-plan choices are not reported in the registry record. The primary result should therefore be understood as the posted Week 48 comparison, not as a complete account of all possible outcomes, analyses, or long-term effects.

17. Why This Trial Matters Statistically

This trial illustrates how an efficacy question can be framed around a defined binary endpoint and a randomized group comparison. MMR at Week 48 is operationalized using a molecular threshold, so the analysis can report how many participants in each group met that criterion. The result can then be expressed as a risk difference, with an interval conveying uncertainty rather than relying on a p-value alone.

The registry also demonstrates why analysis labels and scales matter. It reports a Mantel-Haenszel risk difference and a separate unadjusted re-calculation based on the posted counts. Readers need to distinguish the sponsor's reported analysis from a straightforward recalculation, and should not treat numerical presentations as interchangeable when the registry does not explain their differences.

The broad comparator strategy has an important interpretive consequence: the primary contrast is against a group of investigator-selected TKIs, not a series of individually powered comparisons with each medicine. The trial's posted primary result can address the pooled comparison, while questions about individual comparator treatments require specific analyses and adequate reporting.

Finally, this record is a useful reminder that statistical interpretation depends on more than the headline estimate. Confidence intervals, the analysis population, missing-data handling, multiplicity, and safety reporting all affect how fully a result can be understood. Several of those methodological details are not reported in the registry record, so conclusions should remain tied to the stated endpoint, groups, and follow-up time.

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 primary comparison pools investigator-selected TKIs. The registry record does not report separate primary efficacy estimates for imatinib, nilotinib, dasatinib, and bosutinib, so the overall comparison cannot answer whether the contrast is similar for each option. The choice of comparator was investigator-selected, and the registry record does not provide further detail about the distribution of those selections in the primary efficacy groups.

The study is open-label and unmasked according to the registry. The primary endpoint is based on a laboratory measurement, but the registry record does not report operational details about assessment blinding, central review, or laboratory procedures beyond the use of real-time quantitative polymerase chain reaction. No further claim about the possible influence of study awareness on outcome assessment can be made from the record.

The registry lists a primary completion date and an active, not recruiting status, but does not report whether all follow-up is complete or whether the posted results represent the final analysis. It does not report missing-data handling, multiplicity control, interim analysis details, or the original sample-size justification. These omissions make it difficult to assess the robustness of the full analysis plan from the registry record alone.

The primary endpoint is restricted to MMR at Week 48. It does not summarize the full course of molecular response or establish outcomes in the optional treatment-free remission and treatment re-initiation periods. The registry lists later and other molecular outcomes, but those require separate interpretation. Safety counts are descriptive and do not include, in the posted summary, all details needed to compare exposure-adjusted event rates or establish treatment causality.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Study of Oral Asciminib Versus Other TKIs in Adult Patients With Newly Diagnosed Ph+ CML-CP — NCT04971226.
• Registry record last updated 2026-08-28; status last verified 2026-08.
• Results first posted 2025-03-10.
• 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

NCT04971226 is a Phase III, multicenter, open-label, randomized, parallel-group study of asciminib versus investigator-selected TKIs in adults with newly diagnosed, previously untreated Ph+ CML-CP. The registry lists an enrollment of 405, a study start date of 2021-10-06, primary completion on 2023-11-28, and status active, not recruiting.

The primary endpoint is MMR at Week 48, defined using the BCR-ABL1/ABL ratio on the international scale. The posted counts are 136 of 201 for asciminib and 100 of 204 for investigator-selected TKIs. The sponsor reports a Mantel-Haenszel risk difference of 18.88, with a 95% confidence interval from 9.59 to 28.17 and p < 0.001. The unadjusted re-calculation gives a risk difference of 0.186, with a 95% confidence interval from 0.092 to 0.281 and p < 0.001.

The primary comparison is against the pooled investigator-selected TKI group. Important details on missing data, multiplicity, and the trial's original design assumptions are not reported in the registry record.