A Dose Escalation Safety Study of Locally-Delivered Gemcitabine in Pancreatic Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Dose Escalation Safety Study of Locally-Delivered Gemcitabine in Pancreatic Cancer (NCT02237157), a non-randomized phase 1 study. The primary outcome is Determine the Maximum Tolerated Dose of Gemcitabine to be Delivered Locally to the Pancreas.

2. Clinical Question

This is a dose-escalation safety question, not a randomized test of treatment benefit. The registry also lists CA19-9 tumor reduction as a secondary outcome, but a result for that outcome is not reported in the registry record. The record therefore establishes the study’s intended safety question and dose levels, but does not establish a reported maximum tolerated dose or a tumor-reduction finding.

3. Trial Design

The registry classifies the study as phase 1, non-randomized, unmasked, and sequential. Enrollment was 20, and the study status is completed. Its stated purpose was to assess safety during dose escalation of locally delivered gemcitabine and determine the maximum tolerated dose. The registry record lists pancreatic cancer as the condition and describes the intended population as people with unresectable pancreatic adenocarcinoma.

Four dose arms are listed: 250mg/m2, 500mg/m2, 750mg/m2, and 1000mg/m2. Each arm description specifies one cycle and two doses per cycle. The intervention is intra-arterial gemcitabine delivered to the pancreas using the RenovoCath catheter. Although the arms represent different dose levels, the registry explicitly identifies the allocation as non-randomized. They should not be treated as randomized groups or as equivalent groups differing only by dose.

The primary outcome was the maximum tolerated dose, assessed through toxicity evaluation one week after treatment. The description refers to predefined dose-limiting toxicity as the basis for stopping or holding a dose. The specific toxicity criteria, dose-escalation decision rules, and rules for moving between dose levels are not reported in the registry record. A detailed account of participant assignment across the listed arms is also not reported in the registry record.

The registry lists two outcome measures and zero analyses. It identifies CA19-9 tumor reduction as a secondary outcome, but does not report an outcome result for it. A primary endpoint value, inferential analysis, or statistical analysis plan is not reported in the registry record. The reported design information therefore supports a description of the intended sequential safety evaluation, not a statistical comparison of dose arms.

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: Subjects who complete a cycle ("arm"/"period") then proceed to receive the next higher dose.

Gemcitabine, Local Delivery (Dose 1)

Reasons for not completing this period:

Gemcitabine, Local Delivery (Dose 2)

Gemcitabine, Local Delivery (Dose 3)

Gemcitabine, Local Delivery (Dose 4)

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 Determine the Maximum Tolerated Dose of Gemcitabine to be Delivered Locally to the Pancreas, measured over 1 week post treatment.

The primary endpoint was the maximum tolerated dose of gemcitabine delivered intra-arterially to pancreatic tumors, expressed in mg/m^2. The registry specifies an assessment one week after treatment. Its description says toxicities would be assessed to determine whether a predefined dose-limiting toxicity necessitated stopping or holding a dose. The numerical maximum tolerated dose and the observed toxicity decision for each dose level are not reported in the registry record.

In general, a maximum tolerated dose endpoint is used in early-phase dose-escalation research to identify a dose boundary based on tolerability under the study’s prespecified toxicity rules. It is not, by itself, a measure of comparative efficacy. Interpreting such an endpoint depends on the toxicity definition, observation window, dose-escalation rules, and handling of participants whose assessments are incomplete. Those detailed rules are not reported in the registry record, so the endpoint’s intended role can be described, but its exact operational definition cannot be reconstructed.

The registry also names CA19-9 tumor reduction as a secondary outcome. The posted record does not include its numerical result or an analysis. No conclusion about tumor reduction follows from the endpoint name alone. The endpoint information describes what the study set out to assess, rather than a reported finding.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What a sequential dose-escalation design means. In general, a sequential design evaluates dose levels in an ordered process, with safety observations informing whether escalation, holding, or stopping is appropriate under the protocol. This differs from a randomized comparison in which participants are assigned by chance to treatment groups. The registry labels this study sequential and non-randomized, and its primary outcome description refers to stopping or holding a dose when a predefined dose-limiting toxicity is present. The detailed decision rules and toxicity threshold are not reported in the registry record, so the actual escalation algorithm cannot be reconstructed.

What the dose groups can show. The record lists dose levels of 250mg/m2, 500mg/m2, 750mg/m2, and 1000mg/m2. It also reports different numbers at risk across the groups for the adverse-event summaries. General statistical interpretation of a non-randomized dose sequence requires caution: groups may differ in ways other than dose, and later dose groups are not automatically comparable with earlier ones. The registry specifically warns that these arms are not a like-for-like control comparison. No causal dose-effect estimate is reported.

How the primary endpoint is intended to work. The endpoint is a maximum tolerated dose measured in mg/m^2, with toxicity assessed one week after treatment. In general, a dose-limiting toxicity rule turns safety observations into a dose-escalation decision. The result depends on the rule’s definition, the number of participants evaluated at each level, the assessment window, and the protocol’s handling of incomplete assessment. The registry refers to a predefined toxicity but does not report its definition or decision rule. It also does not report the final maximum tolerated dose or a dose-level account of the primary endpoint.

Counts and uncertainty. The posted adverse-event information is expressed as numbers affected and numbers at risk, alongside selected term-specific percentages. A count is a direct description of how many participants were recorded in a category; it does not by itself quantify uncertainty or establish a difference between groups. In general, uncertainty measures such as confidence intervals help show how much imprecision surrounds an estimate. No confidence intervals for the primary endpoint or adverse-event comparisons are reported in the registry record. No statistical test or model is listed among the analyses.

Why denominators matter. A percentage is interpretable only when its denominator and event definition are clear. The record gives percentages for selected serious adverse-event terms, but the detailed denominator for each term is not reported in the registry record. It is also not clear from the listed information whether the same participant could contribute to multiple terms. The percentages should therefore be read as registry-reported summaries, not combined or reinterpreted as a single overall rate.

Over-dispersion and count outcomes. In general, count data can vary more than a simple statistical model expects; this is called over-dispersion. When present, over-dispersion can make standard errors too small and confidence intervals too narrow if the model does not accommodate it. This concept is relevant to some count-outcome analyses, but the registry reports no model for the primary endpoint or listed adverse-event counts. It is not possible to say whether over-dispersion was assessed or whether any particular count model was used.

Planning calculations are not trial findings. The registry’s sample-size assumptions describe planning illustrations based on a stated fraction of an observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation. For time-to-event outcomes, the stated convention is a total number of events; for other outcomes, it is participants per arm. These assumptions are not a report of this trial’s design rationale. The study is sequential and non-randomized, and no observed effect estimate or primary endpoint analysis is reported in the registry record. In general, planning sample sizes describe hypothetical design requirements under assumptions; they are not observed power and do not demonstrate that a study had adequate power after completion.

What can be concluded methodologically. The design and endpoint establish that the study was intended as a safety-focused dose escalation, not a randomized efficacy comparison. The registry gives the dose levels and selected adverse-event summaries, but not the analysis methods needed to quantify uncertainty or assess the primary endpoint. The statistical plan, model specifications, interval estimates, and inferential results are not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures: the maximum tolerated dose as the primary outcome and CA19-9 tumor reduction as a secondary outcome. It reports zero analyses. No multiplicity procedure, hierarchy for testing outcomes, or adjustment for multiple comparisons is reported in the registry record. In general, multiplicity matters when several statistical claims are tested, because the chance of at least one false-positive result can increase as the number of tests grows. Since no inferential analyses are listed, the record does not allow assessment of whether multiplicity was relevant to the actual analysis or how it may have been handled.

Missing-data methods are not reported in the registry record. The record does not describe how participants with unavailable toxicity assessments, incomplete follow-up, or missing CA19-9 measurements would be handled. It also does not identify an analysis population, such as a safety population or an evaluable population, or state whether all enrolled participants contributed to each endpoint. The adverse-event counts include different numbers at risk across the dose groups, but the reason for those denominators is not reported. They should not be interpreted as a missing-data pattern or as evidence of a particular participant flow.

In general, missing-data handling can affect the estimate and its uncertainty, especially when the reason data are unavailable is related to a participant’s condition or treatment experience. Without the endpoint-specific denominators, analysis-population definitions, and missing-data rules, the completeness and comparability of the posted summaries cannot be evaluated. The record also does not report a statistical analysis plan, prespecified analysis details, or sensitivity analyses.

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

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pain - abdominal (nos) (12.5%), sepsis (12.5%), intraoperative arterial injury / dissection (9.1%), intraoperative arterial injury - lower extremity (9.1%), respiratory failure (9.1%).

The adverse-event summaries report counts of participants affected and at risk within each listed dose group. For serious adverse events, the counts at risk are 20, 15, 11, and 8, while the counts affected are 5, 4, 3, and 2, respectively. For other adverse events, the counts at risk are 20, 15, 11, and 8, while the counts affected are 7, 7, 4, and 0. Deaths are listed as 0 for every group.

The serious adverse-event terms with reported percentages are Pain - Abdominal (NOS), 12.5%; Sepsis, 12.5%; Intraoperative Arterial Injury / Dissection, 9.1%; Intraoperative Arterial Injury - Lower Extremity, 9.1%; Respiratory Failure, 9.1%; and Dehydration, 6.7%. These are registry-reported term summaries. The record does not specify the denominator for each term or whether participants could appear under multiple terms, so the percentages should not be added or treated as mutually exclusive categories.

In general, safety counts describe events recorded in the study; they do not by themselves establish that an intervention caused an event. The dose groups were not randomized and are not a like-for-like comparison. The record does not report adjusted comparisons or uncertainty intervals for these safety summaries.

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

The record documents a phase 1, non-randomized, sequential study designed to evaluate the safety of locally delivered gemcitabine and determine a maximum tolerated dose. It identifies four dose levels, a one-week toxicity assessment timeframe, and selected adverse-event counts. These registry entries describe the study’s design and reported safety summaries, but they do not provide a complete account of the primary outcome.

The maximum tolerated dose result is not reported in the registry record. Nor is a statistical analysis of the primary endpoint, an interval estimate, or a reported result for the secondary CA19-9 tumor reduction outcome. Accordingly, the record does not establish a numerical maximum tolerated dose, the precision of such an estimate, or a tumor-reduction finding. A stated endpoint or study objective is not itself an observed result.

The adverse-event figures provide descriptive counts and selected term-specific percentages. They do not establish a comparative safety effect across doses. Non-randomized sequential groups may differ in participant composition or other factors, and the registry cautions against treating the arms as a like-for-like control comparison. The counts also lack the analysis details and uncertainty estimates needed for a formal comparison.

In general, early-phase safety studies can inform decisions about dose evaluation, but interpretation depends on the protocol’s toxicity definitions, escalation rules, and analysis populations. Those details are not reported in the registry record. The posted information should therefore be read as a partial description of the study and its recorded safety summaries, not as a complete statistical report or an efficacy comparison.

13. Why This Trial Matters Statistically

This study raises an important statistical distinction between dose escalation and treatment comparison. The registry describes a sequential, non-randomized design intended to identify a maximum tolerated dose. Its dose groups are not randomized control groups, so differences in their adverse-event counts cannot be interpreted as if dose were the only difference between otherwise equivalent groups.

The endpoint also illustrates why the operational definition of a safety outcome matters. A maximum tolerated dose depends on what counts as dose-limiting toxicity, when toxicity is assessed, and how the study uses those observations to stop, hold, or advance a dose. The record specifies assessment one week after treatment and refers to predefined dose-limiting toxicity, but the detailed criteria and escalation rules are not reported. Without them, readers cannot reproduce the decision process from the posted summary.

Statistically, a reported endpoint estimate needs context about its precision and the participants contributing to it. The registry includes adverse-event counts and selected percentages, but no analysis, confidence interval, or primary endpoint value. This limits assessment of how stable the safety summaries are and what the study concluded about dose tolerance. The varying at-risk counts also make clear why denominators and participant-flow details matter when interpreting group summaries.

Finally, a planning illustration is not a substitute for a completed analysis. The registry describes generic assumptions for sample-size planning, while the study itself is sequential and non-randomized. Keeping planning quantities separate from observed findings avoids treating a hypothetical design calculation as evidence about the study’s realized precision or power. The statistical significance of this record lies in identifying these distinctions and the specific information needed to interpret the study’s central safety question.

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 of the posted statistical record is that it does not report the numerical maximum tolerated dose or an analysis of the primary endpoint. Although the record describes a one-week toxicity assessment and a predefined dose-limiting toxicity, the toxicity criteria and dose-escalation decision rules are not reported. This prevents independent reconstruction of how the primary endpoint was determined.

The study is classified as non-randomized and sequential. The dose arms are not like-for-like control groups, so their adverse-event counts do not support a simple causal comparison of dose levels. The record does not detail participant assignment across the listed arms, and the reasons for the different at-risk counts are not reported. The group figures should therefore be interpreted descriptively.

The registry lists adverse-event counts and selected serious adverse-event terms, but does not report confidence intervals, statistical tests, adjusted analyses, or a statistical model. The denominators for the term-specific percentages and the possibility of participants appearing in more than one term are not reported. The CA19-9 tumor reduction outcome is named, but its result is not reported in the registry record.

Analysis populations, missing-data handling, and sensitivity analyses are not reported. The registry also does not report a statistical analysis plan or sample-size justification for the trial. The stated planning assumptions do not fill those gaps or establish observed power. These limitations restrict interpretation to the design description and the descriptive safety figures that are posted.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: A Dose Escalation Safety Study of Locally-Delivered Gemcitabine in Pancreatic Cancer — NCT02237157.
• Registry record last updated 2021-05-07; status last verified 2021-04.
• Results first posted 2021-05-07.
• 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

The registry describes a completed phase 1, non-randomized, unmasked, sequential safety study with enrollment of 20 participants. It evaluated locally delivered intra-arterial gemcitabine for unresectable pancreatic adenocarcinoma, with four listed dose levels: 250mg/m2, 500mg/m2, 750mg/m2, and 1000mg/m2. Each arm description specifies one cycle with two doses per cycle.

The primary outcome was the maximum tolerated dose, assessed through toxicity one week after treatment. CA19-9 tumor reduction was listed as a secondary outcome. The registry reports selected adverse-event counts and terms, but lists zero analyses. A numerical maximum tolerated dose, statistical analysis, and secondary outcome result are not reported in the registry record. The non-randomized dose groups should be read as descriptive safety summaries, not as a randomized comparison.