A Dose Escalation Study of Carfilzomib Taken With Thalidomide and Dexamethasone in Relapsed AL Amyloidosis: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Dose Escalation Study of Carfilzomib Taken With Thalidomide and Dexamethasone in Relapsed AL Amyloidosis (NCT02545907), a na phase 1, phase 2 study. The primary outcome is Number of Participants With Dose-Limiting Toxicities as Assessed by Reported Data.

2. Clinical Question

In statistical terms, this is principally a dose-finding question: what toxicity was observed at the escalating dose levels, and what dose can be recommended for further evaluation? The stated study aims also include assessing combination efficacy in an expansion phase. However, the available record does not give an efficacy result or detail the expansion-phase findings. The study enrolled 10 participants overall; the safety groups list 3 participants at Dose Level 0 and 7 at Dose Level 1. These group counts describe the posted safety information, not a randomized treatment comparison. The trial was completed, but completion alone does not establish the size or precision of any treatment effect. The page therefore distinguishes the study's intended questions from the results actually reported.

3. Trial Design

The registry describes a single-arm, open-label, multicentre phase 1b dose-escalation study, with phase classifications of PHASE1 and PHASE2. It is listed as completed, with enrollment of 10. The recorded study start was 2017-09-14 and primary completion was 2019-09-26. The sponsor is University College, London. No masking was used, and allocation is listed as not applicable.

All participants were assigned to the combination described as KTD: carfilzomib, thalidomide, and dexamethasone. The registry summary says the trial begins with dose escalation to determine the maximum tolerated and recommended dose, followed by an expansion phase to assess combination efficacy. The arm description specifies thalidomide at 50mg/day and dexamethasone at 20mg on Day 1, 8, and 15. It states that carfilzomib is administered on Day 1, 8, and 15, but the dose-level information is not reported in the registry record. The record does not describe the number of participants in each intended phase.

The displayed safety data are separated into Dose Level 0 and Dose Level 1. They should be read as dose-level groups within a dose-escalation study, not as randomized arms. Escalation studies allocate participants according to dose-finding procedures, so groups may differ in enrollment timing, dose exposure, and opportunities to observe events. The registry record does not report those procedures in enough detail to reconstruct how participants moved between dose levels or how the escalation decision was made.

The primary endpoint is the number of participants with dose-limiting toxicities after one cycle, for the purpose of establishing the maximum tolerated dose. The outcome description says the assessment spans from the first registered carfilzomib dose in cycle 1 through day 1 of cycle 2, and that completion was to occur within 1 year. Other listed outcomes address clonal response, organ response, time to response, deaths, progression-free status, maximum response, and treatment withdrawal. Numerical results for these outcomes are not reported in the registry record.

A concurrent control group, randomization, and masking are not part of the recorded design. Consequently, the trial can describe observed outcomes under the combination and across its dose levels, but it cannot by itself separate treatment effects from other explanations in the way a controlled comparison might. The displayed tables and plots should be read within that design limitation.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Analysis population for the primary outcome, as defined in the registry: Evaluable set- patient who received at least one cycle of the study treatment. KTD Patients who do not receive one complete cycle due to experiencing a DLT will be included in the analysis; patients who do not receive at least one complete cycle for reasons other than toxicity, without experiencing a DLT, and who miss a dose of Carfilzomib, more than 14 doses Thalidomide or 2 doses of Dexamethasone in the first cycle, will be replaced.

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 Number of Participants With Dose-Limiting Toxicities as Assessed by Reported Data, measured over After 1 cycle of treatment; to be completed within 1 year..

The primary endpoint counts participants with dose-limiting toxicities, assessed using reported data. The registry frames this as a safety endpoint for the dose-escalation phase, covering the interval from receiving the first registered dose of carfilzomib in cycle 1 to day 1 of cycle 2. The endpoint is intended to inform the maximum tolerated dose of carfilzomib when combined with thalidomide and dexamethasone. The record gives the timeframe as after 1 cycle of treatment, to be completed within 1 year.

Counting participants with a dose-limiting toxicity is useful in early dose-finding because the key concern is whether clinically important toxicity appears at a dose level during the defined observation window. It is a participant-level count: it describes how many people met the endpoint, rather than the total number of episodes or the duration of toxicity. General statistical context: the count is easiest to interpret when the number of evaluable participants and the endpoint definition are clear, because a count alone does not convey the precision of an estimated underlying risk. The registry gives a partial evaluable-set description: participants who received at least one cycle are included, and those who did not complete a cycle because of a dose-limiting toxicity are included. The rest of the population definition is not reported in the registry record.

The registry record does not report a numerical result for this primary endpoint. The safety tables elsewhere report serious and other adverse-event information, but those summaries should not be treated as a substitute for the specified dose-limiting-toxicity result. The endpoint's result, detailed adjudication rules, and any uncertainty interval are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

This discussion explains how to read the statistical structure of the study and its posted results. It separates general statistical principles from details that the registry record does and does not describe.

What does a dose-escalation design estimate?

General explanation: a dose-escalation study examines safety as participants receive treatment at different dose levels. The aim is usually to identify a dose suitable for further evaluation, using toxicity observations collected under prespecified rules. Here, the registry says dose escalation was intended to determine the maximum tolerated and recommended dose. The specific escalation algorithm, decision thresholds, and rules for moving between dose levels are not reported in the registry record. Without those details, readers cannot reconstruct how the observed events informed dose selection.

The record gives two dose-level safety groups, but they are not randomized groups. In a dose-escalation study, dose groups can reflect the sequence in which treatment was evaluated and the decisions made during the trial. Differences in counts therefore cannot automatically be attributed to dose. The registry record does not state whether other factors, such as participant characteristics or exposure duration, differed by dose level.

What does a participant count tell us?

The primary endpoint is a count of participants with dose-limiting toxicities during a specified window. A count is direct and interpretable as an observed number of people meeting the endpoint. It does not, by itself, give a stable estimate of the underlying frequency of toxicity. General explanation: when the number of observed participants is limited, one or a few events can have a large effect on the apparent pattern. A confidence interval can communicate uncertainty around an estimated frequency, but no primary endpoint estimate or confidence interval is reported in the registry record.

The adverse-event summary distinguishes participants at risk from participants affected. For example, the serious-event data list 3 at risk and 1 affected at Dose Level 0, and 7 at risk and 2 affected at Dose Level 1. The other-event data list 3 at risk and 3 affected at Dose Level 0, and 7 at risk and 7 affected at Dose Level 1. These are descriptive group summaries. The record does not provide a formal comparison or uncertainty estimates for them. The serious-event and other-event summaries are not interchangeable with the primary dose-limiting-toxicity endpoint, which has its own definition and observation window.

Why does the observation window matter?

The registry defines the primary toxicity period as the time from the first registered carfilzomib dose in cycle 1 through day 1 of cycle 2. An endpoint tied to a fixed observation window has a clear scope: it describes events considered within that interval, not all toxicity that might occur at any later time. The registry also lists other safety information, including deaths and adverse events, but the exact observation windows for those posted group summaries are not described in the registry record. The primary endpoint result and the handling of participants with incomplete observation are not reported in the registry record beyond the partial evaluable-set description.

What does open-label and single-group mean for interpretation?

General explanation: in an open-label study, treatment assignment is not masked. In a single-group study, there is no concurrent control group against which to estimate a comparative treatment effect. These features do not prevent the study from documenting observed toxicity, but they limit the conclusions that can be drawn about whether outcomes differ from what would have happened under another treatment or no study treatment. The registry lists no masking, a single-group model, and allocation as not applicable. It does not report a randomized comparison.

Because participants received a combination of carfilzomib, thalidomide, and dexamethasone, a recorded adverse event cannot be assigned to one component from these summaries alone. The registry entries do not report detailed attribution results. Similarly, efficacy outcomes listed in the record cannot establish a comparative effect without results and a suitable comparator.

What can the analysis record support?

The registry lists 12 outcome measures and 0 analyses. The primary outcome is specified as a count of participants, but its numerical result is not reported in the registry record. The other listed outcomes include response measures, time-to-response measures, deaths, progression-free status, and treatment withdrawals; numerical results for these are not reported. General explanation: outcome listings describe what a study planned or intended to assess, whereas an analysis result requires an estimate, count, or other reported value. The distinction matters because an outcome's presence in a registry is not evidence that a result is available or that an analysis was completed.

The page also contains calculated tables, plots, and a sample-size section. The planning illustrations described for that section use a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation, with the effect to detect expressed as a fraction of the observed effect. They are planning illustrations, not observed power and not the trial's own design justification. This trial itself is single-group, so the assumptions in a two-arm illustration should not be mistaken for the trial's actual allocation. No sample-size justification or statistical plan is reported in the registry record. The values in those generated displays should be read as planning calculations under their stated assumptions, not as additional trial observations.

In short, the statistical record supports description of the design and the posted safety counts, but does not supply the primary endpoint estimate, inferential analyses, or detail needed to reproduce the dose-escalation decisions. General statistical concepts can clarify what such data mean, but they cannot fill those reporting gaps.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Dose Level 0", "Dose Level 1")
n      <- c(3, 6)
events <- c(0, 1)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Dose Level 0  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Dose Level 1  6  1  16.7  0.4  64.1  3.0  56.4  3.7  57.9

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 12 outcome measures. These include clonal response at several time points, amyloidotic organ response, time to organ response, deaths, progression-free status, maximum response, time to maximum response, and treatment withdrawal, in addition to the primary dose-limiting-toxicity endpoint. Listing multiple outcomes raises a general statistical issue: if many outcomes or time points are examined, the chance of finding an apparently notable result by chance can increase. This is the general reason that trial reports often specify which outcome is primary and how secondary analyses will be interpreted.

The record identifies the primary endpoint, but the statistical analysis plan, any prespecified hierarchy among secondary outcomes, and any multiplicity adjustment are not reported in the registry record. The record lists 0 analyses, and numerical results for the response outcomes are not reported in the registry record. It is therefore not possible to assess how multiplicity was handled or whether all intended analyses were completed. No statistical plan should be inferred from the outcome list alone.

The primary population is described as an evaluable set of patients who received at least one cycle of study treatment. Participants who did not receive one complete cycle because of a dose-limiting toxicity are included. The remainder of the population description is not reported in the registry record. A complete account of missing data would explain who lacked endpoint assessment, why information was unavailable, and how such participants were handled in each analysis. Those details are not reported in the registry record.

General explanation: excluding participants without complete follow-up can affect interpretation if their missingness is related to toxicity or outcome. Including participants who stop early because of toxicity, as the record says for the primary endpoint, addresses one specific situation, but it does not describe all possible reasons for incomplete data. The record does not give a missing-data method, sensitivity analysis, or imputation approach. Because the primary result itself is not reported, the effect of missing observations cannot be evaluated.

12. 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: abdominal pain (33.3%), acute kidney injury on ckd (14.3%), fever (14.3%).

The serious adverse-event summary is organized by dose level and shows participants at risk, participants affected, and deaths. At Dose Level 0, the record lists 3 participants at risk and 1 affected by a serious adverse event; at Dose Level 1, it lists 7 at risk and 2 affected. Deaths are listed as 0 in both groups. The separate other-adverse-event summary lists 3 at risk and 3 affected at Dose Level 0, and 7 at risk and 7 affected at Dose Level 1.

These figures are counts, not a randomized comparison. The dose levels belong to a dose-escalation study, and the registry does not report enough allocation or follow-up detail to interpret differences between them as effects of dose. The summaries also do not indicate, by themselves, severity beyond the serious-versus-other classification, duration, recurrence, or attribution to a particular drug. Such details are not reported in the registry record.

The top serious terms are abdominal pain at 33.3%, acute kidney injury on CKD at 14.3%, and fever at 14.3%. The denominator and term-level counting details are not reported in the registry record, so those percentages should be interpreted as listed rather than converted into participant counts. The serious adverse-event summaries are not the same as the primary dose-limiting-toxicity endpoint; the registry does not report the numerical result for that endpoint.

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

The record can establish that this was a completed, open-label, single-group dose-escalation study of carfilzomib with thalidomide and dexamethasone in relapsed AL amyloidosis, with 10 participants enrolled. It also shows that the registry separated safety information by Dose Level 0 and Dose Level 1. In those groups, the record lists serious adverse events affecting 1 participant among 3 at risk and 2 among 7 at risk, respectively; other adverse events affected 3 among 3 and 7 among 7. Deaths are listed as 0 in each group. These are the observations in the posted safety summary.

Those observations do not establish a causal comparison between dose levels. There was no randomized control group, and the groups are part of a dose-escalation study. The counts also do not establish the frequency of every possible adverse event, the cause of an event, or how outcomes would compare with another treatment. The record does not provide the detail needed to answer those questions.

The primary endpoint is the number of participants with dose-limiting toxicities within the defined first-cycle window. Its numerical result is not reported in the registry record. Consequently, the posted information does not show how many participants met the primary endpoint or what maximum tolerated or recommended dose was selected. The registry's description of the study's aims should not be mistaken for evidence that a particular dose was established.

The registry also lists response and disease-status outcomes, but their numerical results are not reported in the registry record. The stated plan to assess efficacy in an expansion phase does not, by itself, show what response was observed. The available safety counts therefore have a narrow descriptive meaning. They should not be taken as a measure of treatment efficacy or as proof that a treatment is beneficial or ineffective.

General statistical context: small descriptive groups provide limited precision, and a count can be sensitive to a small number of observations. The record does not report confidence intervals or formal analyses for these safety summaries. The most defensible reading is to report the observed counts as recorded, acknowledge the missing primary result and analysis details, and avoid broader conclusions that the registry information cannot support.

14. Why This Trial Matters Statistically

This study illustrates why the design and the endpoint must be read together. A dose-escalation trial is built to evaluate safety across dose levels and inform a dose choice, rather than to provide a randomized estimate of comparative benefit. The registry describes that purpose directly: the escalation phase was intended to determine the maximum tolerated and recommended dose, while an expansion phase was intended to assess combination efficacy. Understanding the aim prevents descriptive dose-level safety counts from being misread as a controlled comparison.

The record also shows the value of separating endpoint definitions from general adverse-event summaries. The primary endpoint has a specific participant-level definition and observation window, from the first registered carfilzomib dose in cycle 1 through day 1 of cycle 2. The safety groups report serious and other adverse events, but those summaries are not the same outcome. Treating all safety counts as interchangeable would blur the question the trial set out to answer.

At the same time, this record highlights the importance of complete result reporting. It lists the primary outcome and several secondary outcomes, but the primary numerical result and the secondary numerical results are not reported in the registry record. The absence of those results limits what readers can infer, regardless of the study's completion status or its stated goals. Reporting the endpoint estimate, analysis population, uncertainty, and dose-selection rationale would be important to understanding the evidence.

Finally, the record is a useful example of a general principle in clinical biostatistics: the apparent amount of detail in a table does not automatically imply strong inferential evidence. Participant counts, event terms, and percentages are informative descriptions, but their interpretation depends on denominators, follow-up, design, and analysis methods. Here, the group counts help describe what was recorded, while the single-group design and reporting gaps limit comparisons and precision. Statistical interpretation means retaining both parts of that account: what the record says and what it cannot answer.

15. 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 principal limitation for interpretation is that the registry record does not report a numerical result for the primary dose-limiting-toxicity endpoint. Without that result, readers cannot determine from this record how many participants met the endpoint in its specified observation window. The maximum tolerated dose and recommended dose are stated aims, but the dose-selection result is not reported in the registry record.

The study is single-group, open-label, and not randomized. The safety information is split by dose level, but the groups are not a like-for-like control comparison. The registry does not provide the dose-escalation algorithm, the decision rules for moving between levels, or detail on how the dose groups were formed. These gaps limit interpretation of any apparent differences between the groups.

The overall enrollment was 10, and the displayed safety groups list 3 and 7 participants at risk. Counts from groups of this size are descriptive and provide limited precision. The registry does not report confidence intervals or formal comparisons for the adverse-event summaries. It also does not give the denominator or counting details for the percentages attached to the listed serious-event terms.

The arm description gives the thalidomide and dexamethasone schedules and says carfilzomib was administered on Day 1, 8, and 15, but the carfilzomib dose levels are not reported in the registry record. The full eligibility criteria, complete evaluable-set definition, missing-data approach, adverse-event attribution details, and statistical analysis plan are also not reported in the registry record. Although secondary outcomes are listed, their numerical results are not reported. These limitations mean the posted information can describe the design and selected safety counts, but cannot support a full assessment of efficacy, uncertainty, or dose selection.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Dose Escalation Study of Carfilzomib Taken With Thalidomide and Dexamethasone in Relapsed AL Amyloidosis — NCT02545907.
• Registry record last updated 2021-04-15; status last verified 2020-11.
• Results first posted 2021-04-15.
• 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.

19. Record Summary

This completed trial enrolled 10 participants and evaluated carfilzomib with thalidomide and dexamethasone in relapsed AL amyloidosis. The registry classifies it as a single-group, open-label phase 1b study with PHASE1 and PHASE2 designations. Its stated plan was to use dose escalation to determine a maximum tolerated and recommended dose, followed by an expansion phase to assess efficacy.

The primary endpoint counts participants with dose-limiting toxicities after 1 cycle, during the interval from the first registered carfilzomib dose in cycle 1 through day 1 of cycle 2. The primary numerical result is not reported in the registry record. Safety summaries list Dose Level 0 and Dose Level 1, with serious-event counts, other-event counts, and 0 deaths in each group. These dose levels are not randomized comparison arms.

The registry lists 12 outcome measures and 0 analyses. Numerical results for the listed response outcomes are not reported in the registry record. The available information supports a descriptive account of the study design and posted safety figures, but does not provide the primary endpoint result, statistical analysis details, or a basis for a controlled treatment-effect conclusion.