Study of Paclitaxel, Carboplatin, and PF-05212384 in Advanced or Metastatic NSCLC (UF-STO-LUNG-002): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Paclitaxel, Carboplatin, and PF-05212384 in Advanced or Metastatic NSCLC (UF-STO-LUNG-002) (NCT02920450), a non-randomized phase 1, phase 2 study. The primary outcome is Dose Tolerability.

2. Clinical Question

The study asked whether a maximum tolerated dose of PF-05212384, also called gedatolisib, could be identified when it was combined with paclitaxel and carboplatin in people with non-small cell lung cancer. The registry describes the study population as having advanced or metastatic disease; the brief summary refers to unresectable disease. The primary measure was dose tolerability, with a stated timeframe of 1 year.

This is a dose-finding question rather than a direct test of whether the combination improves a clinical outcome relative to another treatment. Although the study summary describes an interest in whether the combination would work against the cancer, the primary outcome reported is dose tolerability. Efficacy results are not reported in the registry record.

3. Trial Design

The registry classifies the study as Phase 1 and Phase 2, non-randomized, unmasked and sequential. Enrollment is listed as 3, and the study status is terminated. The start date is 2017-09-25 and the primary completion date is 2019-01-02. The University of Florida is listed as sponsor.

The treatment arms are described as three Phase 1b gedatolisib dose levels—110 mg, 150 mg and 180 mg—and a Phase 2 arm using the maximum tolerated dose from Phase Ib. All are described in the context of treatment with paclitaxel and carboplatin. The registry does not report arm-specific descriptions beyond these labels, nor does it report the sequence or decision rules used to move between dose levels.

Because allocation is non-randomized and the study is sequential, the listed arms are not equivalent to randomly assigned groups in a controlled comparison. Differences between arms could reflect their place in the study process as well as dose, and the registry does not provide the information needed to disentangle those factors. The record also does not describe a separate control treatment.

General statistical context: early-phase dose-finding studies commonly focus on tolerability and dose selection, while later-phase studies may examine outcomes such as disease response. That general purpose does not establish which specific escalation algorithm, monitoring schedule or analysis rule was used in this study. Those details are not reported in the registry record.

The registry lists enrollment as 3, while its safety records show participants at risk only in the 110 mg Phase 1b arm. The other listed arms have zero participants at risk in the posted safety data. Those safety figures do not establish how the study proceeded through its intended stages, and the registry does not report a maximum tolerated dose or a Phase 2 efficacy result.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Data are not reported for this outcome measure because an insufficient number of participants were evaluable to determine the MTD.

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 Dose Tolerability, measured over 1 year.

The primary endpoint is titled “Dose Tolerability.” Its description states that the study aimed to identify the maximum tolerated dose of PF-05212384 in combination with paclitaxel and carboplatin in participants with non-small cell lung cancer. The stated timeframe is 1 year. No measurement unit, numerical endpoint value or estimate is reported.

The registry explicitly says that data are not reported for this outcome because an insufficient number of participants were evaluable to determine the maximum tolerated dose. Accordingly, the record does not establish which dose, if any, met the study’s dose-selection objective.

General statistical context: dose-finding endpoints are used to guide selection of a dose for further study by relating dose levels to tolerability under a study’s defined rules. A maximum tolerated dose is not simply the largest dose listed; it depends on how tolerability is defined, what observation period and events count, and what decision procedure is used. The registry record does not report those operational details, so they cannot be reconstructed from the dose labels or safety counts alone.

This endpoint is distinct from an efficacy measure. It does not directly quantify tumor response, survival or a comparison with another treatment. The record lists 2 outcome measures, but details of the other measure are not included in the registry entries. Efficacy outcomes and their results are therefore not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

General statistical context: clinical trial methods determine how observations are collected, summarized and used to answer a study question. For a dose-finding question, the important elements include the definition of a tolerability event, the observation window, the rule for considering a dose acceptable and the method used to choose a dose for further study. The registry’s primary outcome description identifies the goal—finding the maximum tolerated dose of PF-05212384 with paclitaxel and carboplatin—but does not report those operational definitions or a dose-selection algorithm.

What does the design label mean?

The registry describes allocation as non-randomized, masking as none and the model as sequential. In general, randomization uses chance to allocate participants to groups, helping make baseline differences less systematically related to assigned treatment. This study was not randomized, so its arms should not be treated as if they were balanced control groups. Sequential describes the listed study model, but the record does not explain the actual sequence of enrollment, escalation decisions or transition criteria. No masking means the registry does not describe blinding; further details about who knew treatment assignments are not reported.

General statistical context: sequential dose-finding designs can use accumulating tolerability information to inform later dose decisions. Different designs use different rules, and their operating characteristics depend on those rules and on assumptions about dose-related risk. The presence of three named dose levels does not reveal which rule was used, whether every dose was administered, or whether a particular dose was selected. The registry reports insufficient evaluable participants to determine the maximum tolerated dose.

What can the safety counts tell a reader?

Counts describe the number of events or affected participants recorded in a category, while “at risk” gives the corresponding group denominator used in the registry’s safety display. They are not interchangeable with an adjusted risk estimate. The posted safety figures list serious adverse events, other adverse events and deaths separately. These categories should not be combined into a single total: the record does not say whether events overlap, and no addition or deduplication is justified.

General statistical context: a denominator is essential for interpreting a count, and a small denominator makes each observation influential. A zero count does not establish that an event cannot occur; it only means no event was recorded in that listed group in the posted data. Conversely, a nonzero count does not by itself show that treatment caused the event. Attribution, timing and definitions matter, and those details are not reported in the registry record.

How should uncertainty be understood?

General statistical context: a confidence interval is a range of estimates compatible with the data and a chosen method under its assumptions. It communicates uncertainty around an estimate, but its interpretation depends on the design, measurement and analysis. No confidence interval is reported for the primary endpoint or the safety figures in the registry record for this record. No p-value or model-based result is reported either.

General statistical context: over-dispersion occurs when variation in event counts exceeds what a simple model assumes. Accounting for it can avoid portraying estimates as more precise than the data justify. However, the registry does not report a fitted count model or statistical analysis for this trial, so over-dispersion adjustment cannot be attributed to the study. More generally, sparse safety data are not made precise merely by choosing a more elaborate model.

What do the planning assumptions mean?

The record describes planning illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation; the effect to detect is expressed as a stated fraction of the observed effect. For time-to-event illustrations, the stated quantity is total events; other illustrations refer to participants per arm. These are planning assumptions, not a description of the trial’s own design or its observed power. Sample-size illustrations are not observed power.

The registry does not report an effect estimate for the primary outcome, and no analysis results are reported. Consequently, the reported details do not support interpreting an effect-based planning illustration as evidence that the study was adequately powered or that a given effect was detected. A planning calculation describes requirements under assumptions; it does not replace observations or remedy the lack of an evaluable dose-finding result.

There are 0 analyses reported in the record. The statistical analysis plan, any interim analyses, the sample-size justification for this trial and any adjustment for multiple testing are not reported in the registry record. Those omissions prevent a reconstruction of the sponsor’s intended inferential procedure. Where a sponsor analysis exists, it takes precedence over a simple unadjusted recalculation; no sponsor analysis is listed in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 2 outcome measures, but the available record describes only the primary outcome, dose tolerability. The other outcome’s title, definition, measurement method, time window and results are not reported in the registry record. The record lists 0 analyses. It therefore does not provide an account of which outcomes were analyzed or how any outcome findings were evaluated.

Multiplicity arises when several outcomes, comparisons, time points or analyses are considered. General statistical context: examining multiple results can increase the chance of finding an apparently notable result by chance if the analyses are treated as independent opportunities without an appropriate plan. Methods such as prespecifying a primary outcome or adjusting inference can address particular multiplicity concerns. The statistical analysis plan and any multiplicity strategy for this study are not reported, so no such approach can be attributed to the trial.

The analysis populations are also not described. The record does not specify which participants were considered enrolled, treated, evaluable for dose tolerability or included in any safety summary beyond the displayed at-risk counts. The primary endpoint description does state that too few participants were evaluable to determine the maximum tolerated dose, but it does not define evaluability or give a count for that population.

Missing-data handling is not reported in the registry record. There is no description of how incomplete follow-up, missing measurements or participants without an evaluable endpoint were handled, nor any sensitivity analysis. General statistical context: missingness can affect interpretation when the likelihood or reason for missing information is related to a participant’s outcome or treatment experience. Without documentation of how missing observations arose and were handled, the direction and extent of any resulting bias cannot be assessed.

11. 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

Serious events with the highest arm-level rate: sepsis (33.3%).

The safety data list the 110 mg Phase 1b arm with 3 participants at risk for serious adverse events and 1 affected. For other adverse events, that arm has 3 at risk and 0 affected. The registry lists 1 death in this arm. The 150 mg Phase 1b, 180 mg Phase 1b and Phase 2 arms each have 0 participants at risk and zeros for the corresponding reported counts.

Sepsis is listed among the serious adverse event terms at 33.3%. The record does not give further detail about the term’s denominator, timing, severity criteria, attribution or outcome in the registry record, so the percentage should not be expanded into a causal interpretation.

General statistical context: the at-risk count gives context for an event count, but it does not by itself explain follow-up duration or exposure. A zero in an arm with zero participants at risk is not evidence that no event could occur at that dose; it indicates that the posted record has no at-risk participants listed for that arm. Counts across categories should not be added, since overlap is not described. The registry does not establish whether the death was related to treatment.

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

The record establishes that this was a non-randomized, sequential Phase Ib-II study of gedatolisib with paclitaxel and carboplatin, with a primary objective of identifying a maximum tolerated dose. It also establishes that the registry did not report a dose result because there were insufficient evaluable participants to determine the maximum tolerated dose. These are the clearest conclusions supported by the posted information.

The record does not establish that a dose was safe, unsafe, effective or ineffective. It does not report a numerical primary endpoint result, an efficacy estimate, a comparison with a control treatment, or a statistical test. The brief summary expresses the research purpose in terms of whether the drug combination would work against unresectable non-small cell lung cancer, but that purpose is not an efficacy finding. The primary outcome described is dose tolerability.

Nor do the safety figures establish causation. They report counts for serious adverse events, other adverse events and deaths in the listed arms. The registry does not provide the details needed to determine whether an event or death was related to treatment. The zero counts in arms with zero participants at risk do not provide evidence about the safety of those doses in an exposed population.

General statistical context: the distinction between “no evidence of an effect” and “evidence of no effect” matters. When an outcome is not evaluable or results are not reported, there is no estimate with which to assess direction, magnitude or precision. The absence of a reported result cannot be converted into a null effect. Similarly, the planned use of Phase 2 does not establish that a Phase 2 efficacy assessment was completed or analyzed; the record lists the arm but gives no such results.

The termination status documents that the study ended, but the reason is not reported. It cannot be used by itself to infer a safety concern, lack of efficacy, administrative decision or any other explanation. The data support a limited account of design, enrollment, endpoint availability and posted safety counts—not a clinical recommendation or a comparative treatment conclusion.

13. Why This Trial Matters Statistically

This study illustrates a central statistical distinction in early-phase research: defining a dose-finding objective does not guarantee that the collected data will answer it. The registry names maximum tolerated dose as the primary objective, but reports that too few participants were evaluable to determine that dose. The key statistical issue is therefore not the magnitude of an estimated effect; no such primary estimate is reported. It is that the intended endpoint could not be established from the available evaluable information.

The non-randomized sequential structure also matters. Dose levels in a dose-finding study are not automatically independent comparison groups. Enrollment and treatment decisions may follow the study’s sequence, and the registry does not report the rule governing those decisions. Treating the arms as a conventional controlled comparison would overstate what the design and posted data support.

The safety counts show why denominators and exposure context matter. The posted safety record includes participants at risk in the 110 mg arm but none in the other listed arms. Counts from arms with no participants at risk cannot answer how those doses compare in safety. The displayed figures are useful as a record of what was reported, but they do not supply a stable basis for comparative risk estimates.

General statistical context: transparent reporting of endpoint definitions, analysis populations, missing-data methods and dose-selection rules helps readers evaluate what a trial can establish. Those details are not reported in this record. Their absence limits reproducibility and makes it difficult to separate a design limitation from an implementation or reporting limitation. The case is therefore statistically informative as an example of why results and methods must be interpreted together, especially when a dose-finding endpoint is not evaluable.

14. Limitations

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

The primary limitation is explicit: the registry reports that an insufficient number of participants were evaluable to determine the maximum tolerated dose. Although enrollment is listed as 3, the registry does not provide a definition of evaluability or a separate count of participants included in the primary endpoint assessment. There is no reported dose estimate or uncertainty interval.

The record is marked terminated, but the reason for termination is not reported in the registry record. The listed Phase 1b dose arms and Phase 2 arm do not establish that each stage or dose was carried out; the posted safety data show zero participants at risk in all arms other than the 110 mg Phase 1b arm. No details about the duration or completeness of exposure are included in the registry entries.

Additional limitations are that allocation was non-randomized, masking was none and the study was sequential. The registry does not describe a control arm, decision rules for dose escalation, a definition of dose-limiting tolerability, eligibility criteria beyond the disease description, or the analysis populations. Thus, the arm labels cannot be used to make a controlled dose comparison or to reconstruct a dose-selection procedure.

The record lists 2 outcome measures but gives details only for the primary outcome. It lists 0 analyses, and does not report efficacy results, missing-data methods, a statistical analysis plan, interim analyses, multiplicity handling or sample-size justification. Safety reporting is limited to the listed counts and serious adverse event term; treatment attribution and further event context are not reported. These gaps restrict interpretation to the design description and the observed registry entries.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Study of Paclitaxel, Carboplatin, and PF-05212384 in Advanced or Metastatic NSCLC (UF-STO-LUNG-002) — NCT02920450.
• Registry record last updated 2019-10-04; status last verified 2019-10.
• Results first posted 2019-10-03.
• 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.

18. Record Summary

NCT02920450 is a University of Florida-sponsored, non-randomized, unmasked, sequential Phase Ib-II study of gedatolisib with paclitaxel and carboplatin for non-small cell lung cancer. Enrollment is listed as 3. The study start date is 2017-09-25, the primary completion date is 2019-01-02, and the status is terminated.

The primary outcome was dose tolerability, intended to identify the maximum tolerated dose over a stated timeframe of 1 year. The registry says that insufficient participants were evaluable to determine the maximum tolerated dose and reports no result for that outcome. Safety data list events only for the 110 mg Phase 1b arm; the other named arms have zero participants at risk in those data. The registry lists 2 outcome measures and 0 analyses, but efficacy results, details for the other outcome and the reason for termination are not reported.