LDK378 Versus Chemotherapy in ALK Rearranged (ALK Positive) Patients Previously Treated With Chemotherapy (Platinum Doublet) and Crizotinib: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for LDK378 Versus Chemotherapy in ALK Rearranged (ALK Positive) Patients Previously Treated With Chemotherapy (Platinum Doublet) and Crizotinib (NCT01828112), a randomized phase 3 study. The primary outcome is Progression Free Survival (PFS) Per Blinded Independent Review Committee (BIRC).

2. Clinical Question

The clinical question was whether progression-free survival differed between ceritinib and standard chemotherapy in adults with advanced ALK-rearranged, or ALK-positive, non-small cell lung cancer who had previously received platinum-doublet chemotherapy and crizotinib. The primary outcome was assessed by a blinded independent review committee, reducing reliance on the treating investigator’s judgment for the main comparison.

The trial record describes a superiority-or-other-legacy analysis and reports a hazard ratio for ceritinib versus chemotherapy. The comparison concerns the randomized treatment assignment and the primary endpoint; it does not establish that every patient would experience the same course. The permitted switch from chemotherapy to ceritinib after confirmed progression is important context, particularly when considering outcomes beyond the primary progression-free survival comparison. Whether any adjustment for switching was used in other analyses is not reported in the registry record.

3. Trial Design

This was a phase III, multicenter, randomized, open-label, parallel-group study. The registry reports enrollment of 231 participants, with 115 assigned to ceritinib and 116 assigned to chemotherapy for the primary outcome. Because masking was not used, participants and treating teams knew the assigned treatment. The primary endpoint nevertheless relied on blinded independent review of disease assessments.

The study compared oral ceritinib, listed as LDK378 in the title, with standard chemotherapy in adults with advanced ALK-positive non-small cell lung cancer previously treated with platinum-doublet chemotherapy and crizotinib. Chemotherapy was determined by the blinded independent review committee; pemetrexed and docetaxel are listed among the interventions. The chemotherapy group could switch to ceritinib after confirmed progressive disease, while also having the choice to continue pemetrexed. This crossover option means that later outcomes may reflect treatment received after progression as well as initial assignment.

The registry lists the study as completed. The start date was 2013-06-28, and the primary completion date was 2016-01-26. The primary population was the Full Analysis Set, defined as all patients to whom treatment had been assigned by randomization. The primary endpoint timeframe extended from randomization to first radiologically documented progression or death from any cause, up to approximately 24 months. Details of the randomization procedure, allocation concealment, treatment adherence, and the original protocol’s analysis plan 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.

Before assignment: In the treatment phase, patients were randomized 1:1 to one of the treatment arms (ceritinib or chemotherapy). In the extension treatment phase, only patients randomized to the chemotherapy arm were allowed to crossover to receive ceritinib therapy after BIRC-confirmed, RECIST-defined disease progression.

Treatment Phase

Reasons for not completing this period:

Extension Treatment Phase

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The Full Analysis Set (FAS) consisted of all patients to whom study treatment had 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.

Population: The Full Analysis Set (FAS) consisted of all patients to whom study treatment had been assigned by randomization.

6. Primary Endpoint

The registry lists the primary outcome as Progression Free Survival (PFS) Per Blinded Independent Review Committee (BIRC), measured over From the date of randomization to the date of first radiologically documented disease progression or death due to any cause up to approximately 24 months.

Progression-free survival (PFS) measures the time from randomization until the first radiologically documented disease progression or death from any cause. In this study, progression was assessed by a blinded independent review committee. Combining progression and death into one endpoint means that either event ends the time counted as progression-free survival; it is not a measure of survival alone.

The registry reports the median PFS in months for each randomized group, along with 95% confidence intervals. The median is the time by which half of the relevant group has experienced progression or death, as estimated from the time-to-event data. It does not describe the experience of every participant, and it does not convey the full shape of the event-time distribution.

PFS is often used in cancer trials because it captures time until worsening documented by imaging or death, without requiring death to occur before an outcome is observed. Independent review can help limit the influence of knowledge of assigned treatment on progression assessment. Open-label treatment remains relevant to care and other outcomes, but the primary endpoint’s blinded review is a design feature of this comparison. The registry does not report the assessment schedule or detailed rules for handling missed or unevaluable scans.

7. Results Posted in the Registry

• Ceritinib vs Chemotherapy: hazard ratio 0.49 (95% CI 0.36–0.67), reported p < 0.001 (as reported in the registry).

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.

Overall Response Rate (ORR) Per BIRC

• 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:
  • Chemotherapy: 6.9% (8/116) versus 40.9% (47/115) on control; risk difference -34.0 percentage points (95% CI -44.1 to -23.9), risk ratio 0.17 (0.08–0.34); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Chemotherapy than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Overall Response Rate (ORR) Per Investigator Assessment

• 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:
  • Chemotherapy: 6.9% (8/116) versus 44.3% (51/115) on control; risk difference -37.5 percentage points (95% CI -47.6 to -27.3), risk ratio 0.16 (0.08–0.31); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Chemotherapy than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Disease Control Rate (DCR) Per BIRC

• 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:
  • Chemotherapy: 37.9% (44/116) versus 76.5% (88/115) on control; risk difference -38.6 percentage points (95% CI -50.3 to -26.8), risk ratio 0.50 (0.38–0.64); unadjusted two-proportion z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for Chemotherapy than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The posted secondary results cover response and disease control assessed by both a blinded independent review committee and investigators. Across these outcome types, the reported percentages favor ceritinib over chemotherapy. The agreement in direction across assessment approaches offers supporting context for the primary PFS comparison, while response and disease control remain distinct outcomes rather than alternative measures of PFS.

The registry also names overall survival, investigator-assessed PFS, and duration of response among the secondary outcomes. Their numerical results are not included in the secondary summaries described in the record and are therefore not reported. The permitted switch from chemotherapy to ceritinib after confirmed progression is especially relevant when interpreting outcomes that continue after initial progression, including overall survival. The record does not report how switching was addressed in those analyses.

The posted binary comparisons are unadjusted. Multiple secondary outcomes and assessment methods create multiple opportunities for nominally small p-values, so an isolated unadjusted result carries less confirmatory weight than a prespecified primary comparison with a defined analysis. The registry does not report a multiplicity-control strategy. The secondary findings are best treated as supportive descriptions of the observed pattern, not as independent proof of several separate treatment effects.

These outcomes also differ in what they capture: response concerns tumor shrinkage meeting endpoint criteria, disease control includes a broader outcome definition, duration of response concerns how long a response lasts, and overall survival concerns death. Their assessment rules and analysis populations may differ. The available record does not provide enough detail to assume that every secondary endpoint was measured or analyzed in an identical way.

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

• Overall Survival (OS)
• Progression Free Survival (PFS) Per Investigator Assessment
• Overall Response Rate (ORR) Per BIRC
• Overall Response Rate (ORR) Per Investigator Assessment
• Duration of Response (DOR) Per BIRC
• Duration of Response (DOR) Per Investigator Assessment
• Disease Control Rate (DCR) Per BIRC
• Disease Control Rate (DCR) Per Investigator Assessment

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a time-to-event outcome: each participant contributes time from randomization until progression or death, or until the participant’s outcome is censored under the analysis rules. The registry defines progression-free survival as the time to first radiologically documented progression or death from any cause. It identifies the median as the reported summary and specifies assessment by a blinded independent review committee.

What does the median describe?

The median is the estimated point in the event-time distribution at which half of participants have had the event. It is a compact summary, but it does not show when all individual events occurred or whether the separation between groups changed over time. It also differs from a mean: the median is not an average duration across all participants. A confidence interval around a median expresses uncertainty in its estimated location.

What does the hazard ratio describe?

A hazard is the instantaneous event rate among participants who have not yet had the event at a given time. A hazard ratio compares those rates between groups in a time-to-event analysis. The reported value of 0.49 is for ceritinib versus chemotherapy. It is not a probability, an absolute risk difference, or a comparison of the two medians. A single hazard ratio is easiest to interpret when the relative event rates are reasonably stable over time; whether that condition was assessed is not reported in the registry record.

The registry identifies the log-rank method for the primary comparison and reports a hazard ratio with a 95% confidence interval and p-value. The log-rank test compares the event-time experience across groups while accounting for the timing of events and censoring. The registry does not provide enough detail to identify every modeling choice, the method used to construct the hazard-ratio interval, or any adjustment variables. Those details should not be inferred from the listed test alone.

Why use blinded review?

In an open-label trial, participants and clinical teams know the assigned treatment. A blinded independent review committee assesses the primary progression endpoint without relying solely on investigator judgment. This can reduce the influence of treatment knowledge on the classification of radiologic progression. It does not eliminate every possible source of bias: scan timing, missing scans, and rules for censoring can also matter, and those details are not reported in the registry record.

How should the p-value be read?

The reported p-value is the probability, under the null comparison and the assumptions of the test, of seeing results at least as incompatible with that null as those observed. It is not the probability that the null is true, nor does it measure the size or clinical importance of a difference. The p-value is reported as p < 0.001. It should be interpreted alongside the estimated hazard ratio, its confidence interval, the endpoint definition, and the design.

The posted recalculations for the binary secondary outcomes are unadjusted comparisons based on the arm-level summaries. They are not substitutes for a sponsor analysis where one is reported. For the primary time-to-event endpoint, the record supplies a sponsor-reported hazard ratio, so that result takes precedence over a simple reconstruction from summary statistics. The primary comparison has no discrepancy flagged in the posted record.

For binary outcomes, a response or disease-control percentage is the proportion of participants counted as having the outcome under that endpoint’s definition and assessment method. A risk difference expresses an absolute contrast between proportions; risk ratios and odds ratios express relative contrasts on different scales. These measures can yield different numerical impressions because they answer different questions. The secondary results include reported percentages and confidence intervals, but the registry does not supply sponsor-reported analyses for those posted binary outcomes.

Over-dispersion is a general concern when observed variability exceeds what a statistical model assumes. It can make standard errors too small and confidence intervals too narrow if ignored. It is more directly relevant to models for counts or repeated observations than to interpreting the posted median PFS and hazard ratio. There is no indication in the registry record that over-dispersion was assessed or was an issue in this trial; it should not be presumed.

The record lists a Full Analysis Set consisting of all patients assigned treatment by randomization. This defines who was intended for the primary analysis, but does not disclose how missing outcome information, censoring, protocol deviations, or any later treatment switching were handled in each analysis. General statistical practice distinguishes the population definition from the specific procedures used to handle incomplete follow-up; the latter are not reported in the registry record.

11. Confidence Intervals and Precision

The primary PFS estimate for ceritinib is a median of 5.4 months, with a 95% confidence interval from 4.1 to 6.9 months. For chemotherapy, the median is 1.6 months, with a 95% confidence interval from 1.4 to 2.8 months. The intervals communicate uncertainty in the estimated medians; they are not ranges containing 95% of individual participants’ PFS times.

The hazard ratio is 0.49, with a 95% confidence interval from 0.36 to 0.67. The interval conveys uncertainty in the relative time-to-event comparison under the reported analysis. Its limits are below 1, consistent with the direction of the point estimate. Confidence intervals are not guarantees that the true value lies within their limits, and they do not account for every source of uncertainty, including limitations in the registry’s reporting.

Precision is affected by the amount and timing of follow-up and by the number of events, not simply by the number enrolled. The registry gives group sizes but does not report the event count used for the primary analysis or detailed censoring information. The planned observation timeframe is up to approximately 24 months; this is not the same as the actual follow-up received by each participant. The interval and point estimate are therefore best read together, with the analysis method and endpoint definition in view.

The reported p-value, p < 0.001, addresses compatibility with a null comparison under the test assumptions, while the confidence interval describes a range of values compatible with the data and method. Neither alone explains the effect’s clinical relevance or the degree of variation among participants. The record does not report a prespecified threshold for a clinically important PFS difference.

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 time-to-event planning figures are total numbers of events, not participants per arm. For planning at 100% of the observed effect, the illustration lists 62 total events for 80% power and 83 for 90% power. At 75% of the observed effect, it lists 110 and 147 total events, respectively. At 50% of the observed effect, it lists 247 and 331 total events, respectively.

The 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 the stated fraction of the observed effect. The event totals use the Schoenfeld approach. These are planning illustrations based on the observed effect, not a reconstruction of the trial’s original sample-size justification; the original design assumptions are not reported in the registry record.

The 80% power column describes the event total that the illustration estimates would give an 80% chance of detecting the specified effect under its assumptions, if that effect were the true one. The 90% column describes the corresponding target for a 90% chance. These are not statements about the probability that the trial result is correct. Smaller effects require more events, which is why the listed totals increase as the assumed fraction of the observed effect becomes smaller.

Power figures are prospective planning quantities, not a measure of the probability that the observed result is true or a retrospective score for the completed study. The values are not observed 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

d is the total number of events needed (Schoenfeld's formula) with 1:1 allocation and proportional hazards.

Worked calculation for this trial

1. Hazard ratio HR = 0.490, so ln(HR) = -0.7133.
2. d = 4 × (1.96 + 0.842)² / [ln(0.490)]² = 61.7, rounded up to 62 events in total 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
hr <- 0.49                              # observed hazard ratio
events_needed <- function(power, frac = 1) ceiling(round(4 * (z_a + qnorm(power))^2 / log(hr^frac)^2, 6))   # Schoenfeld, 1:1 allocation
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: events = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(events_needed(0.80, f), big.mark = ","), format(events_needed(0.90, f), big.mark = ",")))

# Simulation: given d total events under proportional hazards, events in the treatment arm are Binomial(d, hr / (1 + hr)).
# The score (log-rank type) statistic is (D1 - d / 2) / sqrt(d / 4).
set.seed(1828112)
d    <- events_needed(0.80)
reps <- 20000
D1 <- rbinom(reps, d, hr / (1 + hr))
z  <- (D1 - d / 2) / sqrt(d / 4)
cat(sprintf("Simulated power with %s events: %.3f (%d replicates)\n", format(d, big.mark = ","), mean(abs(z) > z_a), reps))

Output from running the code

100% of observed effect: events = 62 (80% power), 83 (90% power)
 75% of observed effect: events = 110 (80% power), 147 (90% power)
 50% of observed effect: events = 247 (80% power), 331 (90% power)
Simulated power with 62 events: 0.795 (20000 replicates)

The simulation uses the fact that, with d events in total and proportional hazards, the number of events in the treatment arm is binomial. It is an approximation that conditions on the total number of events, and it should land close to the target power.

14. Group-Sequential Design Analysis

Many time-to-event trials, especially in oncology, stop or report early when interim data are overwhelming, so a single fixed-sample number understates the real design problem. The calculation below is an illustrative group-sequential design built from this trial's posted control median and hazard ratio. The registry record does not state the planned number of looks or the spending function, so the design is a planning illustration, not a reconstruction of the protocol. Assumed: one-sided α = 0.025, 90% power, three equally spaced looks on information, Lan-DeMets O'Brien-Fleming spending, 24 months of accrual, 60 months in total and a 1% monthly dropout hazard.

Under these assumptions the design needs about 87 participants in total, with the boundaries below.

Fixed design versus group-sequential alternatives

Group-sequential designs cost a small increase in maximum sample size and events compared with the fixed design, in exchange for the chance to stop early for overwhelming efficacy. O'Brien-Fleming spending keeps early boundaries very conservative and the final boundary close to the fixed-design value, whereas Pocock spending spends more alpha early.

Sensitivity to a smaller true effect

Planning for exactly the observed effect is optimistic, because observed effects in published trials tend to be somewhat larger than the true effect. The table repeats the main design (three looks, O'Brien-Fleming spending, 90% power) for smaller true hazard ratios.

For context, this trial enrolled 231 participants.

Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993. Agreement between the two packages confirms the boundary calculation.

R code and output

suppressWarnings(suppressMessages({ library(gsDesign); library(rpact) }))
med_control <- 1.6          # control-arm median, months (converted from the posted unit)
hr          <- 0.49          # hazard ratio used for planning
lambdaC     <- log(2) / med_control
eta         <- 0.01               # dropout hazard per month (assumption)
acc_months  <- 24                 # accrual duration (assumption)
study_months<- 60                 # total study duration (assumption)

design <- function(k, sfu, beta, h = hr) gsSurv(k = k, test.type = 1, alpha = 0.025, beta = beta, sfu = sfu,
  lambdaC = lambdaC, hr = h, hr0 = 1, eta = eta, gamma = 1, R = acc_months, T = study_months, minfup = study_months - acc_months)

# Main design: 3 looks, Lan-DeMets O'Brien-Fleming spending, one-sided alpha 0.025, 90% power
x <- design(3, sfLDOF, 0.10)
cat(sprintf("N_TOTAL|%s\n", format(ceiling(sum(x$gamma * x$R)), big.mark = ",")))
for (i in 1:3) cat(sprintf("GS|%d|%.1f|%.0f|%.3f|%.5f|%.5f\n", i, x$T[i], x$n.I[i], x$upper$bound[i], pnorm(-x$upper$bound[i]), cumsum(x$upper$spend)[i]))
cat("\n"); print(gsBoundSummary(x))

# Comparison of designs
cmp <- list(list("Fixed design, one analysis", 1, sfLDOF), list("Two looks, O'Brien-Fleming", 2, sfLDOF),
            list("Three looks, O'Brien-Fleming", 3, sfLDOF), list("Three looks, Pocock", 3, sfLDPocock))
cat("\n")
for (b in c(0.20, 0.10)) for (d in cmp) {
  y <- design(d[[2]], d[[3]], b)
  cat(sprintf("DES|%s|%d|%.0f|%s|%.1f\n", d[[1]], round(100 * (1 - b)), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ","), tail(y$T, 1)))
}

# Sensitivity: what if the true effect is smaller than the observed one? (3 looks, O'Brien-Fleming, 90% power)
cat("\n")
for (fr in c(1, 0.75, 0.5)) {
  y <- design(3, sfLDOF, 0.10, exp(log(hr) * fr))
  cat(sprintf("SENS|%.2f|%.3f|%.0f|%s\n", fr, exp(log(hr) * fr), tail(y$n.I, 1), format(ceiling(sum(y$gamma * y$R)), big.mark = ",")))
}

# Independent cross-check of the boundaries with rpact
r <- getDesignGroupSequential(kMax = 3, alpha = 0.025, beta = 0.10, sided = 1, typeOfDesign = "asOF", informationRates = x$timing)
cat(sprintf("\nCross-check, upper boundaries: gsDesign %s | rpact %s\n", paste(round(x$upper$bound, 3), collapse = ", "), paste(round(r$criticalValues, 3), collapse = ", ")))

Method: LachinFoulkes 
   Analysis               Value Efficacy
  IA 1: 33%                   Z   3.7103
      N: 42         p (1-sided)   0.0001
 Events: 28        ~HR at bound   0.2447
  Month: 11    P(Cross) if HR=1   0.0001
            P(Cross) if HR=0.49   0.0338
  IA 2: 67%                   Z   2.5114
      N: 70         p (1-sided)   0.0060
 Events: 56        ~HR at bound   0.5098
  Month: 19    P(Cross) if HR=1   0.0060
            P(Cross) if HR=0.49   0.5603
      Final                   Z   1.9930
      N: 88         p (1-sided)   0.0231
 Events: 84        ~HR at bound   0.6463
  Month: 60    P(Cross) if HR=1   0.0250
            P(Cross) if HR=0.49   0.9000



Cross-check, upper boundaries: gsDesign 3.71, 2.511, 1.993 | rpact 3.71, 2.511, 1.993

15. Analyses Reported by the Sponsor

The sponsor-reported primary analysis gives a hazard ratio of 0.49 for ceritinib versus chemotherapy, with a 95% confidence interval of 0.36 to 0.67 and p < 0.001. The reported method is the log-rank test. The posted record also includes the arm-level median PFS estimates and their confidence intervals, which describe the endpoint on an absolute time scale.

For a time-to-event outcome, a hazard ratio cannot be recovered exactly from the two arm medians alone. The sponsor’s reported analysis uses the trial’s event-time information, whereas a reconstruction from summary medians would discard timing and censoring details. The sponsor result is therefore the appropriate inferential comparison to cite; the record does not flag a discrepancy between it and the posted summary.

The primary analysis population is the Full Analysis Set: all participants assigned treatment by randomization. This population definition is relevant to the randomized comparison, but the registry does not detail all implementation choices, including censoring or handling of missing assessments. The sponsor-reported result should be interpreted with those reporting limits in mind. It does not resolve whether the effect varied over time or how the optional switch to ceritinib affected later outcomes.

16. Multiplicity, Missing Data and Analysis Populations

The registry identifies PFS assessed by a blinded independent review committee as the primary outcome and lists numerous secondary outcomes, including overall survival, investigator-assessed PFS, response, duration of response, and disease control. It reports 26 outcome measures and 25 posted secondary outcomes. Testing multiple outcomes can increase the chance of obtaining at least one apparently notable result by chance when each is evaluated separately.

A multiplicity strategy can address this by specifying which outcomes are confirmatory, how testing is ordered, or how the significance threshold is adjusted. Whether the trial used such a strategy is not reported in the registry record. The secondary binary comparisons are unadjusted, so their p-values should not be read as though they were protected against all the possible comparisons. The primary sponsor-reported analysis remains distinct from those secondary calculations.

The primary analysis population is defined as the Full Analysis Set, consisting of all patients assigned treatment by randomization. This is useful information about who was included in the intended analysis, but the registry does not report detailed procedures for missing scans, incomplete follow-up, censoring, or protocol deviations. The number of participants with evaluable outcomes at each assessment is also not given for all endpoints. These omissions limit independent assessment of how missing information could have influenced estimates.

The registry does not report whether interim analyses were planned or conducted, whether stopping boundaries were used, or the original sample-size justification. It also does not report detailed methods for handling the optional post-progression switch from chemotherapy to ceritinib in analyses of later outcomes. These items should not be inferred from the reported p-value, the completed status, or the planning illustrations. Without the protocol and a fuller statistical analysis plan in the registry record, the extent of prespecification and the handling of all analysis decisions remain uncertain.

17. 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 (7.7%), nausea (7.0%), dyspnoea (7.0%), vomiting (6.4%), pleural effusion (6.4%).

The safety record separates on-treatment groups from survival follow-up and also identifies a chemotherapy/ceritinib group. In the ceritinib on-treatment group, 57 of 115 participants were listed as affected by serious adverse events; in the chemotherapy on-treatment group, 35 of 113 were listed. In the chemotherapy/ceritinib on-treatment group, 45 of 78 were listed. These are group-specific registry counts and should be read in the context of the recorded treatment periods.

The registry lists pneumonia at 7.7%, nausea at 7.0%, dyspnoea at 7.0%, vomiting at 6.4%, pleural effusion at 6.4%, and general physical health deterioration at 6.1% among the leading serious adverse event terms. The record does not give enough detail in these summaries to establish the relationship of each event to treatment or to compare like-for-like exposure across groups.

Deaths are also reported separately for on-treatment and survival-follow-up groups. Those counts are not equivalent to deaths caused by treatment. The crossover group and unequal treatment exposure further complicate simple comparisons. These registry summaries describe recorded safety events; they do not by themselves establish comparative safety or individual risk.

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

The posted primary results show a higher estimated median PFS in the ceritinib group than in the chemotherapy group, together with a sponsor-reported hazard ratio below 1 and a confidence interval that excludes 1. This is evidence of a difference in the time-to-progression-or-death outcome under the reported analysis. The result applies to the trial population and comparison as recorded, not automatically to every patient or clinical setting.

The endpoint combines radiologic progression and death, so it should not be interpreted as overall survival. The registry names overall survival as a secondary outcome, but its numerical result is not included in the posted secondary summaries. The chemotherapy group’s option to switch to ceritinib after confirmed progression also matters when considering outcomes after the primary endpoint event.

The result does not show that every participant benefited, nor does a group median reveal any individual’s expected time without progression. The hazard ratio is a relative time-to-event measure, not a probability or a direct comparison of median durations. Statistical significance does not itself specify clinical importance; a threshold for a clinically meaningful PFS difference is not reported in the registry record.

The trial was open-label, although the primary progression assessment was conducted by a blinded independent review committee. Random assignment supports a comparison based on assigned groups, but the available registry entries do not fully describe missing-data handling, censoring, or other analysis decisions. The secondary outcomes are supportive, with unadjusted comparisons and no multiplicity strategy reported. The evidence should therefore be read as a specific randomized comparison with clearly stated limits, not as a complete account of benefit, harms, or outcomes in all populations.

19. Why This Trial Matters Statistically

Statistically, the trial pairs randomization with a time-to-event primary endpoint and blinded independent review of progression. That combination is informative: randomization supports comparing assigned groups, while independent review reduces reliance on unblinded investigator assessment for the main endpoint. The registry also reports both group medians and a hazard ratio, allowing the result to be considered on absolute and relative time-to-event scales.

The case illustrates why those summaries should not be treated as interchangeable. Median PFS gives a time point within each group’s event-time distribution; the hazard ratio compares relative event rates over time. The latter uses event-time information not captured by the medians alone. Confidence intervals add information about precision, while the p-value addresses compatibility with a null comparison under the test assumptions.

Study design context is equally important. The chemotherapy group could switch to ceritinib after confirmed progression, and the study was open-label. The primary assessment was blinded, but later outcomes and safety summaries may reflect differing treatment exposure. Recognizing these features prevents an overly simple reading of the primary result as a description of every outcome or every participant.

Finally, the record demonstrates the distinction between a sponsor’s reported analysis and unadjusted secondary comparisons. The primary hazard ratio is the sponsor-reported inferential result; secondary response and disease-control comparisons are supportive and should be interpreted with awareness of multiple outcomes and the absence of a reported multiplicity strategy. This is a useful example of reading estimates, uncertainty, analysis populations, and reporting gaps together rather than relying on a single p-value.

20. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry record does not report the original sample-size justification, detailed randomization procedures, allocation concealment, interim-analysis plans, or a full statistical analysis plan. The planning illustrations based on the observed effect do not fill those gaps. Detailed information about censoring, missing scans, protocol deviations, and the number of progression events used in the primary analysis is also not reported.

Although the primary endpoint used blinded independent review, the study itself was open-label. The registry does not describe the imaging schedule or all adjudication procedures. The chemotherapy group’s option to switch to ceritinib following confirmed progression also complicates interpretation of outcomes after progression. How switching was handled in secondary analyses is not reported.

Secondary outcomes include several endpoints and assessment methods. Their posted binary comparisons are unadjusted, and a multiplicity strategy is not reported. Numerical results for all named secondary outcomes are not present in the secondary summaries. Consequently, the available results cannot provide a complete account of the full set of outcomes specified in the record.

The safety summaries separate treatment periods and include a crossover group, but the information presented does not support simple comparisons that account for exposure duration, treatment sequence, or event attribution. The listed deaths should not be interpreted as treatment-caused deaths. These limitations concern what can be concluded from the registry reporting; they do not establish that a particular analysis or trial procedure was absent.

21. Related Tutorials

22. Related Statistical Calculators

23. Sources

• ClinicalTrials.gov: LDK378 Versus Chemotherapy in ALK Rearranged (ALK Positive) Patients Previously Treated With Chemotherapy (Platinum Doublet) and Crizotinib — NCT01828112.
• Registry record last updated 2025-02-07; status last verified 2025-01.
• Results first posted 2017-07-27.
• 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.

24. Record Summary

NCT01828112 is a completed phase III, randomized, open-label, parallel-group study sponsored by Novartis Pharmaceuticals. It enrolled 231 adults with advanced ALK-positive non-small cell lung cancer previously treated with platinum-doublet chemotherapy and crizotinib. Participants were assigned to ceritinib or chemotherapy.

The primary endpoint was PFS assessed by a blinded independent review committee. The sponsor reported a hazard ratio of 0.49, with a 95% confidence interval of 0.36 to 0.67 and p < 0.001. Secondary results include response and disease-control outcomes; several other named outcomes do not have numerical summaries in the record.

The chemotherapy group could switch to ceritinib after confirmed progression or continue pemetrexed. The registry does not report the original sample-size justification, detailed missing-data procedures, interim analyses, or a multiplicity strategy. These reporting limits are relevant when interpreting secondary outcomes, later follow-up, and safety summaries.