← Clinical Trials
Advanced Solid Tumors Phase 1/2 Single Group NCT03157128

LIBRETTO-001: Complete Statistical Analysis of Selpercatinib in Advanced Solid Tumors

An independent statistical analysis of LIBRETTO-001, a phase 1/2 single-group study of selpercatinib (LOXO-292) in participants with advanced solid tumors, RET fusion-positive solid tumors, and medullary thyroid cancer.

ClinicalTrials.gov NCT03157128  ·  Start: 2017-05-02  ·  Primary completion: 2025-02-14
Scope of this record

This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.

1. Trial at a Glance

LIBRETTO-001 is a phase 1/2, single-group study evaluating selpercatinib (LOXO-292) in participants with advanced solid tumors, RET fusion-positive solid tumors, and medullary thyroid cancer. The registry identifies three primary endpoints spanning dose determination and objective tumor response.

857
Enrollment
Participants
1/2
Phase
Phase 1/2
16
Arms
Registry design
3
Primary endpoints
Registered
FeatureLIBRETTO-001
Trial nameLIBRETTO-001
NCT IDNCT03157128
Brief titleA Study of Selpercatinib (LOXO-292) in Participants With Advanced Solid Tumors, RET Fusion-Positive Solid Tumors, and Medullary Thyroid Cancer (LIBRETTO-001)
Phase1/2
StatusACTIVE_NOT_RECRUITING
Start2017-05-02
Primary completion2025-02-14
Lead sponsorEli Lilly and Company
Sponsor typeINDUSTRY
Therapeutic areaOncology
AllocationNA
Design modelSINGLE_GROUP
MaskingNONE
Primary purposeTREATMENT
Enrollment857.0
InterventionLOXO-292 (drug)
Registry interpretation: The single-group design is fundamentally different from a randomized comparative trial. The primary statistical questions therefore center on dose selection, response estimation, and the precision and interpretation of observed outcomes rather than a randomized treatment-versus-control effect estimate.

2. Clinical Question

The clinical question changes across the two phases of the study. In Phase 1, the principal statistical task is dose escalation and identification of a maximum tolerated dose and recommended Phase 2 dose. In Phase 2, the registered primary question is whether selpercatinib produces an objective response, based on Independent Review Committee assessment, in the studied participant population.

Population

Participants with advanced solid tumors, RET fusion-positive solid tumors, and medullary thyroid cancer. The registered conditions include Non-Small Cell Lung Cancer; Medullary Thyroid Cancer; Colon Cancer; Any Solid Tumor.

Intervention

LOXO-292, also identified in the registry as selpercatinib.

Comparator

No concurrent comparator arm is identified in the registry design. Allocation is listed as NA and the design model is SINGLE_GROUP.

Primary questions

What dose level satisfies the registered MTD definition, what dose is selected as the RP2D, and what is the Phase 2 objective response rate based on IRC assessment?

3. Trial Design

01
Enroll857.0 participants
02
Phase 1Dose evaluation
03
MTD / RP2DCycle 1
04
Phase 2Response assessment
05
IRCObjective response
Design
Phase 1/2, single-group, unmasked treatment study.
Allocation
NA. The ClinicalTrials.gov record does not describe randomized allocation.
Masking
NONE.
Primary purpose
TREATMENT.
Arms
16 registered arms.
Intervention
LOXO-292 (drug).

the ClinicalTrials.gov record identifies 16 arms. The serious-adverse-event data reported with the registry record identify several Phase 1 dose groups, including 20 mg Selpercatinib QD, 20 mg Selpercatinib BID, 40 mg Selpercatinib BID, 60 mg Selpercatinib BID, 160 mg Selpercatinib QD, 80 mg Selpercatinib BID, and 120 mg Selpercatinib BID. The registry-reported safety string ends during the description of another 160 mg Selpercatinib group and therefore does not provide a complete arm-by-arm safety inventory.

4. Registered Primary Endpoints

EndpointTime frameRegistry definition
Phase 1: Maximum Tolerated Dose (MTD) Cycle 1 (cycle length = 28 days) The MTD is defined as the highest dose level at which none of the first 3 treated patients, or not more than 1 of the first 6 treated patients, experiences a DLT. A DLT is any adverse events that starts on or after first administration of study drug, as defined by National Cancer Institute's Common Terminology Criteria for Adverse Events (NCI CTCAE) v5.0. * Any Grade(G) ≥3 nonhematologic toxicity, excluding * G3 AST, ALT, and/or total bilirubin elevation for \<7 days. * G3 neutropenia \<7 days * G3 thrombocytopenia without clinically significant bleeding * G3 or G4 lymphopenia. * First occurrence of G3 or G4 electrolyte abnormalities * G3 fatigue, weakness, nausea; other manageable constitutional symptom * G3 or G4 vomiting or diarrhea that lasts for \<48hours with antiemetic/antidiarrheal medication in case of G3 and \<24 hours in case of G4 * G4 manageable constitutional symptom.
Phase 1: Recommended Phase 2 Dose (RP2D) Cycle 1 (cycle length = 28 days) Phase 1: RP2D
Phase 2: Objective Response Rate (ORR) Based on Independent Review Committee (IRC) Assessment Approximately for up to 7 years 8 months Objective Response Rate was defined as the percentage of participants with best overall response of confirmed response (CR), or Partial response (PR). Response was confirmed by a repeat assessment no less than 28 days. * CR is defined as disappearance of all target lesions. * PR is defined as at least a 30% decrease in the sum of diameter.

5. Statistical Structure of the Primary Questions

LIBRETTO-001 illustrates why the phrase "primary endpoint" does not necessarily imply a single hypothesis test comparing two randomized groups. Its three registered primary endpoints address different statistical tasks.

PhaseEndpointStatistical role
Phase 1Maximum Tolerated DoseDose-escalation decision based on dose-limiting toxicities during Cycle 1.
Phase 1Recommended Phase 2 DoseDose-selection decision for subsequent development.
Phase 2Objective Response RateEstimation of the proportion of participants with confirmed CR or PR based on IRC assessment.

There is no randomized comparator in the registry-reported design, so an effect estimate such as a treatment hazard ratio versus control is not the natural primary quantity for this trial. The principal inferential quantities are instead dose-toxicity information and the response proportion, together with its statistical uncertainty.

6. Planned Analysis

ClinicalTrials.gov reports that results have been posted for all three registered primary endpoints, but the ClinicalTrials.gov record contains no formal statistical analyses. Consequently, the available data do not provide formal analysis estimates, confidence intervals, or p-values for the primary endpoints in the structured statistical-analysis field.

What would normally be analysed?
ORR = number of participants with confirmed CR or PR ÷ evaluable participants

For the Phase 2 endpoint, the primary descriptive quantity would ordinarily be the observed proportion of participants with a confirmed response. An exact or otherwise prespecified confidence interval would normally accompany the response estimate to quantify its precision.

Phase 1: Maximum Tolerated Dose

The registered MTD rule is explicitly dose- and cohort-based. The highest dose level satisfying the stated Cycle 1 DLT rule would be identified as the MTD. The registry-reported definition specifies that none of the first 3 treated patients, or not more than 1 of the first 6 treated patients, may experience a DLT at that dose level.

Phase 1: Recommended Phase 2 Dose

The registry identifies RP2D as a primary endpoint but supplies only the endpoint label and its Cycle 1 time frame. The statistical interpretation is therefore a dose-selection decision rather than a conventional comparative hypothesis test.

Phase 2: Objective Response Rate

The registered endpoint is based on Independent Review Committee assessment and requires a confirmed CR or PR. Confirmation requires a repeat assessment no less than 28 days after the initial response assessment. The appropriate analysis would therefore distinguish confirmed responses from unconfirmed tumor changes and express the endpoint as a proportion.

No formal outcome claim is made here. Although the registry record indicates that results are posted, the registry-reported structured data do not include the statistical-analysis estimates needed to state a numerical ORR, MTD, or RP2D result.

7. Phase 1 Dose-Finding Logic

The MTD definition provides an unusually clear example of how an early-phase trial converts observed toxicity into a dose-selection rule. The decision is made using the first Cycle 1 experience at a dose level rather than by comparing long-term outcomes between randomized groups.

First 3 treated patients

The registered rule allows a dose level to satisfy the MTD definition when none of the first 3 treated patients experiences a DLT.

First 6 treated patients

The registered rule also allows a dose level when not more than 1 of the first 6 treated patients experiences a DLT.

DLT window

The primary endpoint is evaluated during Cycle 1, with a cycle length of 28 days.

Highest qualifying dose

The MTD is defined as the highest dose level satisfying the registered DLT rule.

8. Serious Adverse Events by Phase 1 Dose Group

The ClinicalTrials.gov record provides affected/at-risk counts for several Phase 1 dose groups. These counts are safety descriptions and should not be interpreted as efficacy results or as randomized treatment effects.

Phase 1 dose groupParticipants with serious AEs / participants at risk
20 mg Selpercatinib QD4/6
20 mg Selpercatinib BID8/10
40 mg Selpercatinib BID9/16
60 mg Selpercatinib BID7/12
160 mg Selpercatinib QD1/1
80 mg Selpercatinib BID13/20
120 mg Selpercatinib BID11/19
Safety interpretation: these counts should not be compared as if they were randomized-arm event rates. The groups have different dose levels and different denominators, and the ClinicalTrials.gov record does not establish that the listed groups represent mutually exclusive populations or a controlled comparison.

9. Statistical Methods Explained

What is the statistical purpose of an MTD?

The MTD is a dose-selection endpoint. Instead of asking whether one treatment is better than another, the analysis asks how observed dose-limiting toxicity changes as dose level increases. The registered rule makes that decision explicit by defining the highest dose at which the specified number of early participants experience DLTs.

Why is Cycle 1 important for the MTD endpoint?

The registered MTD time frame is Cycle 1, with a cycle length of 28 days. Restricting the primary dose-limiting-toxicity decision to a defined observation window makes the escalation rule operational and reproducible. It also means that the MTD endpoint is not equivalent to cumulative toxicity over the entire study.

What does "not more than 1 of the first 6" mean?

It is a decision threshold, not a statement that 1 of 6 participants represents a universal acceptable toxicity rate. In the context of this specific registered rule, a dose level can meet the MTD definition when no more than 1 of the first 6 treated participants experiences a DLT.

Why is RP2D different from MTD?

MTD is explicitly defined in the registry as a toxicity-based endpoint. RP2D is a separate registered endpoint intended to identify the dose selected for Phase 2. The ClinicalTrials.gov record provides the RP2D label but do not provide its decision algorithm, so the two endpoints should not be treated as interchangeable.

How should ORR be interpreted in a single-group study?

ORR is the percentage of participants with a confirmed CR or PR according to the registered IRC-based definition. In a single-group design, the response proportion describes the observed activity in the studied population. Without a concurrent randomized control group, it does not by itself quantify a treatment effect relative to another intervention.

Why does response confirmation matter?

The registered ORR endpoint requires a confirmed response, with a repeat assessment no less than 28 days after the initial response assessment. Confirmation reduces the chance that a transient or uncertain imaging finding is classified as a definitive response under the primary endpoint.

Why should a confidence interval accompany ORR?

A response proportion calculated from a finite sample is subject to sampling variability. A confidence interval provides information about the precision of the estimated response proportion. The interval does not mean that individual participants have a probability equal to the interval endpoints of responding; it describes uncertainty about the population parameter under the chosen statistical framework.

10. Objective Response Rate: Statistical Interpretation

What ORR measures

The registered Phase 2 ORR endpoint measures the percentage of participants whose best overall response is a confirmed response, specifically CR or PR, based on Independent Review Committee assessment.

What ORR does not measure

ORR does not measure overall survival, progression-free survival, duration of treatment, or the probability that an individual participant will experience long-term clinical benefit. A response endpoint is a tumor-response measure, not a complete summary of all possible clinical outcomes.

Why the analysis population matters

The denominator used for an ORR calculation must follow the prespecified analysis population and evaluability rules. Using a different denominator can materially change the reported percentage. The ClinicalTrials.gov record does not provide a statistical-analysis record defining the denominator for the posted result, so no numerical ORR is reconstructed here.

11. Independent Review Committee Assessment

The Phase 2 primary endpoint is explicitly based on Independent Review Committee (IRC) Assessment. This is statistically important because the primary response classification is not simply an investigator-reported judgment.

Assessment source
Independent Review Committee (IRC)
Response categories
Confirmed response defined as CR or PR
Confirmation
Repeat assessment no less than 28 days after the initial response assessment
Time frame
Approximately for up to 7 years 8 months

Independent review can reduce the influence of investigator expectations on the primary tumor-response classification. For statistical interpretation, the important point is that the endpoint is defined by a prespecified assessment framework rather than by an informal statement that a tumor "responded."

12. What the Single-Group Design Changes Statistically

A single-group design changes the meaning of nearly every familiar clinical-trial statistic. In a randomized trial, a treatment effect is usually defined by the difference between concurrent treatment groups. In LIBRETTO-001, the primary Phase 2 response endpoint is instead a response proportion within the treated study population.

QuestionRandomized comparative trialLIBRETTO-001 framework
Primary comparisonUsually treatment versus comparatorNo concurrent comparator identified in registry-reported design
Typical effect measureRisk ratio, risk difference, odds ratio, hazard ratio, etc.Observed response proportion is central to the registered Phase 2 endpoint
RandomizationBalances prognostic factors in expectationNot present in the registry-reported single-group design
Phase 1 objectiveUsually not a comparative efficacy questionMTD and RP2D dose-selection endpoints
Reference populationComparator arm provides a concurrent referenceExternal context would be needed for comparative interpretation
Key statistical caution: an observed response proportion can be precisely estimated and still leave substantial uncertainty about comparative treatment benefit. Precision of an estimate and causal comparison are separate statistical concepts.

13. Statistical Methodology

Binary endpoint estimation

The Phase 2 ORR endpoint is classified in the ClinicalTrials.gov record as a Binary primary endpoint. The fundamental quantity is the proportion of participants meeting the confirmed-response definition.

Core response estimator
p̂ = R / N

where R is the number of participants meeting the registered confirmed-response definition and N is the prespecified analysis denominator.

For a binary endpoint, an interval estimate should ordinarily accompany the point estimate. Exact binomial methods are one common approach when the sample size or observed proportion makes a normal approximation inappropriate, although the specific method should be taken from the prespecified statistical analysis plan when available.

Dose-escalation decision rules

The Phase 1 MTD endpoint uses an explicit rule based on observed DLTs among the first 3 or first 6 treated participants at a dose level. This is a form of rule-based early-phase dose finding: the statistical decision is driven by toxicity observations rather than a conventional two-arm hypothesis test.

Descriptive safety analysis

Serious adverse events are naturally summarized using counts and denominators, often accompanied by percentages. For this trial, the ClinicalTrials.gov record provides affected/at-risk counts for several dose groups. Those counts should be retained as descriptive safety information rather than converted into a comparative effect estimate.

Independent response assessment

For the Phase 2 endpoint, IRC assessment provides the basis for response classification. The response must also be confirmed by repeat assessment no less than 28 days after the initial assessment. Statistically, the confirmation rule defines which observations enter the binary response endpoint.

14. Multiplicity and Interim Analysis

The ClinicalTrials.gov record identifies three registered primary endpoints but do not provide a statistical-analysis record describing multiplicity adjustment, alpha allocation, interim-analysis boundaries, or a formal hypothesis-testing hierarchy.

Three primary endpoints

The registry identifies MTD, RP2D, and Phase 2 ORR as primary endpoints. They represent different development decisions rather than three interchangeable versions of the same outcome.

Multiplicity

No multiplicity procedure is reported in the ClinicalTrials.gov record. Therefore, no claim about familywise type I error control is made here.

Interim analysis

No interim-analysis method is reported in the ClinicalTrials.gov record. The Phase 1 dose rule itself should not automatically be labeled an interim efficacy analysis.

Formal p-values

No formal statistical analyses were posted. No p-value is therefore reported or reconstructed for the primary endpoints.

For a single-group binary endpoint, statistical significance may be assessed against an external or historical benchmark if a protocol specifies one. No such benchmark or null response rate is included in the ClinicalTrials.gov record, so a formal one-sample hypothesis test cannot be reconstructed from the available information.

15. Missing Data and Analysis Population

The ClinicalTrials.gov record does not specify a missing-data or imputation strategy for the Phase 2 ORR endpoint. This matters because response assessment requires adequate tumor evaluation, and the definition of the analysis denominator can influence the resulting response proportion.

IssueStatistical implication
Missing response assessmentCan affect which participants contribute to the ORR denominator.
Unconfirmed responseDoes not satisfy the registered confirmed-response definition.
Repeat assessmentRequired no less than 28 days after the initial response assessment for confirmation.
Analysis denominatorMust follow the prespecified endpoint population; the ClinicalTrials.gov record does not provide that denominator.

No imputation method is stated in the ClinicalTrials.gov record. In particular, this page does not assume that missing assessments were counted as responses, nonresponses, or censored observations.

16. Bayesian Methods

No Bayesian method is identified in the ClinicalTrials.gov record. The registered endpoints and registry-reported design information do not specify a Bayesian prior, posterior probability criterion, Bayesian dose-escalation algorithm, or Bayesian response analysis.

That distinction is important because early-phase oncology studies can use Bayesian dose-finding methods, but the presence of a dose-escalation endpoint alone is not evidence that a Bayesian model was used. The registry-reported MTD definition is expressed as an explicit rule involving the first 3 or first 6 treated participants.

17. Stratification and Covariate Adjustment

The ClinicalTrials.gov record does not identify stratification factors, covariate-adjusted models, or prespecified subgroup-adjusted analyses. Because this is a single-group design, there is no randomized treatment comparison for which stratified treatment-effect estimation would ordinarily be required.

Interpretive principle
Single-group response estimate ≠ randomized treatment effect

A response proportion summarizes observed activity in the studied population. It does not automatically establish how the intervention compares with an alternative treatment, because the study lacks a concurrent randomized comparator in the registry-reported design.

18. Trial Timeline

2017-05-02

Study start

The registry identifies 2017-05-02 as the study start date.

Phase 1

Dose evaluation

The registered primary endpoints include MTD and RP2D, both evaluated over Cycle 1, with a cycle length of 28 days.

Phase 2

Objective response assessment

The registered Phase 2 primary endpoint is ORR based on Independent Review Committee assessment, with a time frame of approximately up to 7 years 8 months.

2025-02-14

Primary completion

The registry identifies 2025-02-14 as the primary completion date.

Current registry status

ACTIVE_NOT_RECRUITING

the ClinicalTrials.gov record identifies the current status as ACTIVE_NOT_RECRUITING.

19. How to Read the Phase 1 Safety Data

The serious-adverse-event counts illustrate a central issue in early-phase dose-finding: the denominator can be very small. For example, the ClinicalTrials.gov record reports 1/1 for the 160 mg Selpercatinib QD group. A proportion calculated from 1 participant is necessarily unstable as an estimate of a broader dose-level safety rate.

Small denominators

Affected/at-risk counts should always be read together with the denominator. A count of 1/1 means that the affected count equals the registry-reported at-risk count for that group; it does not establish a population-level serious-adverse-event probability.

Dose cannot be separated from sample size

The listed Phase 1 groups differ in both dose and number of participants. Simple comparison of their affected/at-risk fractions would not by itself provide a controlled estimate of a dose-response relationship.

Serious AE versus DLT

The registry-reported serious-adverse-event field and the registered MTD endpoint are not interchangeable. MTD is defined using DLT criteria during Cycle 1, whereas the registry-reported safety field reports serious adverse events by arm.

20. What This Trial Does Not Establish From the Supplied Statistical Record

21. Limitations

22. Why This Trial Matters Statistically

LIBRETTO-001 is useful as a statistical teaching case because it illustrates a different branch of clinical-trial methodology from the conventional randomized phase 3 model. The study combines early-phase dose selection with a later binary response endpoint in a single-group framework.

ConceptHow it appears in LIBRETTO-001
Early-phase dose findingMTD and RP2D are registered primary endpoints.
Rule-based toxicity decisionMTD uses the specified first-3 / first-6 DLT rule.
Cycle-specific safety windowMTD and RP2D use Cycle 1, with a cycle length of 28 days.
Binary endpointPhase 2 ORR is classified as Binary in the ClinicalTrials.gov record.
Confirmed responseCR or PR must be confirmed by repeat assessment no less than 28 days later.
Independent assessmentORR is based on Independent Review Committee assessment.
Single-group inferenceObserved response proportions do not automatically represent comparative treatment effects.
Small-sample safety dataSeveral listed dose groups have small denominators.
MultiplicityThree registered primary endpoints are present, but no registry-reported multiplicity method is identified.
Missing dataThe ClinicalTrials.gov record does not identify an imputation or missing-response strategy.

23. Related Tutorials

Learn more about the methods used in this trial:

24. Related Calculators

25. Sources

The numerical and endpoint information presented on this page is restricted to the registry-reported LIBRETTO-001 trial data. The linked PubMed records are provided as registry-linked publication resources; no additional numerical trial results from those publications are incorporated into this page.

Continue through the Clinical Biostats statistical library

Connect this trial's dose-finding, binary-response, and single-arm design concepts to deeper statistical tutorials and calculation tools.

26. Record Summary

LIBRETTO-001 provides a useful example of how statistical methodology changes across clinical-development phases. Its registered primary endpoints combine a Cycle 1 dose-toxicity decision for MTD, a Phase 1 dose-selection endpoint for RP2D, and a Phase 2 confirmed objective-response endpoint based on Independent Review Committee assessment. Because the registry-reported design is single-group and the structured statistical-analysis array contains no entries, the available record supports a detailed explanation of the statistical framework but does not support reconstruction of numerical primary endpoint estimates, confidence intervals, or p-values.

Clinical Biostats methodology: A trial-results page should distinguish clearly between what the registry reports, what can be derived directly from those reported data, and what would require additional statistical-analysis records. For LIBRETTO-001, that distinction is particularly important because dose-finding and single-arm response estimation do not have the same interpretation as randomized comparative efficacy analysis.