A Study of LY3039478 in Participants With Advanced Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of LY3039478 in Participants With Advanced Cancer (NCT01695005), a non-randomized phase 1 study. The primary outcome is Part A and F: Number of Participants With Dose Limiting Toxicities (DLTs).

2. Clinical Question

This was a Phase 1 study of LY3039478 in participants with advanced or metastatic cancer. Its central early-development question was dose-related: the study evaluated escalating doses and then cohorts receiving fixed doses, with the primary endpoint counting dose-limiting toxicities (DLTs) in Parts A and F. The DLT definition focused on specified adverse events during Cycle 1 that were considered related to LY3039478, along with certain other toxicities judged dose limiting by the investigators and Lilly clinical research personnel.

The registry also describes pharmacokinetic objectives, including measures of drug concentration over time, and says that changes in blood-cell and tissue markers and tumor activity would be explored. These are objectives, not results. The registry record does not report a count of participants with DLTs, a recommended dose determined by the study, or a result demonstrating tumor activity. It also does not report a comparative treatment-effect estimate. Because participants were not randomized to a control group, differences between the study’s dose and cohort groups cannot be read as randomized comparisons.

3. Trial Design

The registry classifies the study as Phase 1, non-randomized, unmasked, and sequential. It records enrollment of 247 participants and a completed status. The recorded conditions include neoplasms, neoplasm metastasis, and lymphoma. The study began in 2012-10, and primary completion is recorded as 2018-06-26.

The study included several parts rather than a single two-arm comparison. Part A used escalating oral doses of LY3039478 administered three times per week. The listed doses were 2.5 milligrams, 5 mg, 10 mg, 20 mg, 30 mg, 45 mg, 60 mg, 75 mg, and 100 mg, for two 28 day cycles; participants receiving benefit could continue until disease progression. The registry describes Parts B, C, D, and E as cohort-expansion parts using fixed doses determined in Part A, although the descriptions for some cohorts are not complete in the registry record.

Parts F1 and F2 included LY3039478 with prednisone. The F1 description lists a 75 mg or 100 mg loading dose and a 50 mg LY3039478 dose, with prednisone co-administered for the first 2 weeks in cycle 1. The F2 description lists loading doses of 75 mg, 100 mg, or 125 mg and a 50 mg LY3039478 dose, also with prednisone during the first 2 weeks in cycle 1. The registry identifies the study drugs as LY3039478 and prednisone.

Sequential dose assignment and cohort expansion serve a different purpose from randomized allocation: groups may differ in dose, cancer type, study part, or enrollment period. The registry record does not report a concurrent randomized comparator or masking. Accordingly, the study design supports describing experience within cohorts, but does not establish a like-for-like causal comparison between those cohorts.

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: Completers of the study were participants that are off treatment that did not withdraw consent and were not lost to follow up. One participant assigned to the Part A: 45 mg group LY3039478 arm received 75 mg in error on Day 1 and had evaluable PK data. Therefore, this participant is included in the Part A: 75 mg LY3039478 arm PK analysis population (N=6), though only N=5 is shown in the Participant Flow module.

Only the first 6 of 29 groups are shown.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All Part A and F participants who received at least one dose of study drug.

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: All participants who received at least one dose of study drug.

Only the first 6 of 30 groups are shown.

6. Primary Endpoint

The registry lists the primary outcome as Part A and F: Number of Participants With Dose Limiting Toxicities (DLTs), measured over Baseline through the end of Cycle 1, Up to 28 Days.

The primary endpoint is the number of participants with dose-limiting toxicities in Parts A and F. The registry defines its population as participants in those parts who received at least one dose of study drug. The timeframe is baseline through the end of Cycle 1, up to 28 Days.

A DLT was defined as an adverse event during Cycle 1 that was related to LY3039478 and met a listed criterion under the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0. The criteria include specified Grade 3 non-hematological toxicity, Grade 4 hematological toxicity lasting greater than 5 days, febrile neutropenia, specified thrombocytopenia, or another significant toxicity judged dose limiting by the primary investigator and Lilly clinical research personnel. The registry includes withdrawal during Cycle 1 for a possibly related toxicity as an example of the last criterion.

In general, a DLT count is useful in early-phase dose evaluation because it records how many participants meet a prespecified toxicity definition in a defined observation period. It can inform dose-selection decisions when considered alongside the dose levels, cohort sizes, and other relevant evidence. A count by itself does not describe the full range, timing, or severity of all adverse events, and it does not establish that a particular dose is safe. The registry record does not report the primary endpoint count for this study, so the observed DLT frequency cannot be described from this record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the primary count represent?

The primary outcome is a count of participants meeting the DLT definition, rather than a continuous measurement or a time-to-event estimate. The definition sets both an observation window and clinical criteria: Cycle 1, up to 28 Days, and toxicity considered related to LY3039478 that meets one of the listed criteria. In general, defining the window and event criteria in advance helps clarify what is included in the endpoint. It does not remove the need to know how many participants were assessed or how missing assessments were treated. Those details are not reported in the registry record.

The population stated for this endpoint is not all enrolled participants. It is limited to Part A and F participants who received at least one dose of study drug. This matters because a count must be interpreted alongside its denominator. The registry record does not report the DLT count or the denominator for this endpoint, so neither an event proportion nor its uncertainty can be described.

Why does the sequential design matter?

In general, dose-escalation studies assess successive dose levels and may expand selected cohorts. The sequence can be useful for examining tolerability and collecting pharmacokinetic information across dose groups. But assignment is not random, and the groups need not be comparable in all respects. A participant’s dose group may be tied to the study part, cohort, cancer type, or decisions made during the study. Differences in observed counts between such groups are therefore descriptive unless a suitable analysis and assumptions support a stronger interpretation. No such comparative analysis is reported in the registry record.

Randomization is commonly used in comparative trials to make treatment groups comparable, on average, with respect to measured and unmeasured factors. This study is classified as non-randomized and has no reported control group. It would therefore be inappropriate to interpret cohort differences as if participants had been randomly assigned to competing treatments. The registry’s description of a sequential model does not by itself provide the dose-selection algorithm or other protocol details; these are not reported in the registry record.

How should the safety counts be read?

The registry reports serious adverse-event figures by group, including counts of participants at risk and counts affected, as well as death counts. These are event summaries, not estimates of causal effects. The number of participants affected by a serious event is distinct from the total number of serious events, because one participant may experience more than one event; the registry record does not clarify every counting detail. The posted term percentages should be read with particular care because their denominators are not stated alongside the terms.

In general, a rate or proportion is more interpretable when its numerator, denominator, observation period, and event definition are clear. A count from a small cohort may change noticeably with the experience of an individual participant. Comparisons also need to account for how long participants were observed and whether groups had similar opportunities for events to be recorded. The registry record does not supply enough information to make an adjusted comparison of safety across the cohorts.

What can inferential statistics add?

In general, a confidence interval communicates uncertainty around an estimate, under the assumptions of the analysis. A p-value summarizes how compatible the observed data are with a specified statistical null model; it is not a measure of clinical importance or the probability that a hypothesis is true. Neither type of result is reported for the primary endpoint in this record. The number of listed analyses is zero, so there is no posted inferential result to interpret.

For a binary outcome such as having a DLT, an analysis might describe the count and proportion in each group and, where appropriate, estimate uncertainty. A comparative model would require a defined comparison, adequate outcome information, and assumptions appropriate to the design. Because this study was non-randomized and no primary endpoint counts are reported, there is no basis in the registry record for an adjusted treatment contrast or a precision estimate. Any sample-size calculation used for planning would be a design illustration, not observed power; no such calculated values are reported in this record.

What are the other statistical objectives?

The registry lists pharmacokinetic objectives, including maximum concentration, time to maximum concentration, and area under the concentration-time curve for specified study parts. In general, these measures describe different features of drug concentration over time. They do not substitute for the DLT endpoint, and their interpretation depends on the sampling and analysis methods. The registry record does not report pharmacokinetic results or enough detail to explain the estimation procedures.

The study description also identifies exploration of blood-cell and tissue markers and documentation of tumor activity as objectives. These aims do not, by themselves, establish a prespecified confirmatory hypothesis or a statistical analysis plan. The analysis methods for these objectives, including any model choices or handling of multiple endpoints, are not reported in the registry record.

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("Part A: 2.5 mg LY3039478", "Part A: 5 mg LY3039478", "Part A: 10 mg LY3039478", "Part A: 20 mg LY3039478", "Part A: 30 mg LY3039478", "Part A: 45 mg LY3039478")
n      <- c(4, 3, 4, 8, 7, 5)
events <- c(0, 0, 0, 1, 1, 0)
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  Part A: 2.5 mg LY3039478  4  0  0.0  0.0  60.2  0.0  49.0  0.5  52.2
ARM  Part A: 5 mg LY3039478  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Part A: 10 mg LY3039478  4  0  0.0  0.0  60.2  0.0  49.0  0.5  52.2
ARM  Part A: 20 mg LY3039478  8  1  12.5  0.3  52.7  2.2  47.1  2.8  48.2
ARM  Part A: 30 mg LY3039478  7  1  14.3  0.4  57.9  2.6  51.3  3.2  52.7
ARM  Part A: 45 mg LY3039478  5  0  0.0  0.0  52.2  0.0  43.4  0.4  45.9

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 17 outcome measures and includes multiple pharmacokinetic measures across study parts, alongside the primary DLT outcome and the broader goals of exploring markers and tumor activity. When many outcomes are examined, general statistical practice recognizes a multiplicity issue: the chance of at least one apparently notable finding can increase as the number of tests or analyses grows. Methods for addressing this can include identifying a primary analysis in advance, adjusting inference, or treating exploratory findings as hypothesis-generating. Which approach was planned for this study is not reported in the registry record.

The record lists zero analyses. It does not provide a statistical analysis plan, p-values, confidence intervals, or a description of any multiplicity adjustment. Consequently, the registry record does not establish which outcomes were analyzed as confirmatory, how any analyses were prioritized, or how interpretation accounted for multiple outcomes and cohorts.

Missing-data handling is also not reported in the registry record. The endpoint definition identifies participants who received at least one dose in Parts A and F, but the record does not say how incomplete Cycle 1 observation, withdrawal, or unavailable endpoint assessments were treated. It does not report the number of participants with missing primary endpoint data.

The registry does not describe additional analysis populations for the primary endpoint, such as a population based on protocol adherence, nor does it give counts for any such populations. It also does not explain whether participants were analyzed by assigned dose, dose received, or another rule. These details matter because choices about eligibility and missing observations can affect a reported count or estimate. Without those methods and results, the primary endpoint cannot be independently reconstructed from the registry record.

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: colitis (100.0%), diarrhoea (100.0%), superinfection (100.0%), hypokalaemia (100.0%), myocardial infarction (50.0%).

The serious adverse-event figures are organized by study part and dose or cohort. For example, the Part A 2.5 mg group lists 4 participants at risk and 1 affected; the Part A 10 mg group lists 4 at risk and 2 affected. The Part B groups list 14 affected among 25 at risk at 75 mg and 14 affected among 30 at risk at 50 mg. These figures describe the recorded groups; they do not make the groups equivalent or establish a dose-related pattern.

The registry also lists deaths by group. A death count is not, by itself, evidence that the study drug caused a death. The record does not state that the deaths were attributed to treatment, and it does not provide a causal analysis. Serious events and deaths should therefore be distinguished from the primary DLT endpoint, which has its own criteria and Cycle 1 timeframe.

The listed serious-event terms include Colitis, Diarrhoea, Superinfection, and Hypokalaemia at 100.0%, and Myocardial infarction and Pyrexia at 50.0%. The registry record does not state the denominators alongside these percentages, so the percentages cannot be interpreted as proportions of the total enrollment. In general, safety counts are most informative when read with the event definitions, observation periods, denominators, and attribution rules. Several of those details are not reported in the registry record.

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

The registry record establishes that this was a completed Phase 1, non-randomized study of LY3039478 in participants with advanced or metastatic cancer. It documents sequential dose-escalation and cohort-expansion parts, including Parts F1 and F2 with prednisone, and defines a primary endpoint focused on DLTs in Parts A and F. It also records pharmacokinetic objectives and broader exploratory aims concerning markers and tumor activity.

However, the record does not report the primary DLT count, a recommended dose, pharmacokinetic findings, marker changes, or tumor-activity results. It therefore cannot answer from the posted results how frequently participants met the primary endpoint or how precisely any dose-specific DLT estimate was determined. The serious adverse-event figures provide descriptive counts for listed cohorts, but they do not establish a causal relationship with study treatment or support a randomized comparison.

In general, a completed study record and an enrollment count do not by themselves demonstrate an effect or establish safety. Interpretation depends on the observed endpoint results, their denominators, follow-up, missing-data methods, and analysis plan. Those results and methods are not reported in the registry record for the primary endpoint.

The findings should also not be generalized automatically beyond the participants and cohorts described. The study included different parts and doses, and its non-randomized design does not support treating those cohorts as equivalent groups. The record’s objectives are not the same as evidence that those objectives were achieved. The appropriate reading is therefore limited: the registry describes the study structure, endpoint definitions, and some cohort-level safety counts, but it does not report the analyses needed to quantify the primary endpoint or establish comparative conclusions.

14. Why This Trial Matters Statistically

This study illustrates a central statistical distinction in early-phase research: defining an endpoint and collecting cohort-level observations are not the same as reporting an estimate with interpretable precision. The record specifies a DLT definition and a Cycle 1 assessment window. Those details give the endpoint a clinical and temporal scope, but the missing result prevents readers from assessing how many participants met that definition.

The sequential, non-randomized design also makes careful language important. Dose-escalation and expansion cohorts can describe observed experience at different doses and in different study parts. They do not automatically form a controlled comparison. Cohort composition, study part, and follow-up can differ, so a numerical contrast between groups would not alone isolate a dose effect.

There is also a useful lesson in separating the primary endpoint from other objectives. The registry lists pharmacokinetic measures and exploratory goals involving markers and tumor activity. These address different questions and would require their own results and methods. Without a reported analysis plan or outcomes, the registry record cannot establish which findings were intended to carry confirmatory weight or how multiplicity was handled.

Finally, the record shows why safety summaries need denominators, event definitions, and attribution context. Cohort counts and term percentages may appear straightforward, yet they can be misleading if readers mistake them for comparative risks or treatment-caused events. The record’s structure is informative about the planned evaluation, while the absence of primary results and inferential analyses limits what can be concluded statistically.

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.

A central limitation of the registry record is that it does not report the primary endpoint result, despite defining the DLT criteria, timeframe, and population. The number of participants with DLTs and the corresponding endpoint denominator are not reported in the registry record. As a result, the primary endpoint’s observed frequency and precision cannot be assessed.

The study is classified as non-randomized and has no reported control group. Its cohorts span multiple study parts, doses, and cancer-group labels. They are not a like-for-like control comparison, and the registry does not report an adjusted analysis that would address differences between cohorts. The sequential model is recorded, but the detailed dose-selection procedure and decision rules are not reported in the registry record.

The adverse-event information is presented by group, but the registry record does not provide the primary endpoint result or a complete account of follow-up, event attribution, and missing observations. The denominators for the serious-event term percentages are not stated alongside those terms. Death counts do not establish treatment causation.

The record lists zero analyses and does not report confidence intervals, p-values, a statistical analysis plan, multiplicity handling, or missing-data methods. It also does not report pharmacokinetic, marker, or tumor-activity results. Some cohort descriptions are incomplete in the registry record, which limits detailed reconstruction of study procedures. These limitations restrict interpretation to the design, endpoint definitions, and descriptive safety information that are recorded; they do not justify filling in unreported results or methods.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A Study of LY3039478 in Participants With Advanced Cancer — NCT01695005.
• Registry record last updated 2025-12-26; status last verified 2025-12.
• Results first posted 2025-12-26.
• 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

NCT01695005 is a completed Phase 1 study of LY3039478 in participants with advanced or metastatic cancer, including lymphoma. It was non-randomized, unmasked, and sequential, with enrollment of 247 participants. The study included escalating and fixed-dose cohorts and Parts F1 and F2 combining LY3039478 with prednisone.

The primary outcome was the number of participants with DLTs in Parts A and F during Cycle 1, up to 28 Days. The registry defines the endpoint and its assessment population but does not report the DLT count. It also records pharmacokinetic objectives and exploratory aims concerning markers and tumor activity, without reporting those results. Serious adverse-event counts and deaths are listed by cohort, but they are descriptive and do not establish treatment causality or a randomized comparison. The record lists zero analyses; statistical methods for multiplicity and missing data are not reported in the registry record.