This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for ATX Study:A Study of Avastin (Bevacizumab), Tarceva (Erlotinib) and Xeloda (Capecitabine) in Patients With Locally Advanced and/or Metastatic Pancreatic Cancer (NCT00925769), a non-randomized phase 1 study. The primary outcome is Part 1: Maximum Tolerated Dose (MTD) of Capecitabine.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00925769 |
| Phase | Phase 1 |
| Condition | Pancreatic Cancer |
| Allocation | Non-Randomized |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 32 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Hoffmann-La Roche |
| Results posted | Yes |
2. Clinical Question
Population
Patients with locally advanced and/or metastatic pancreatic cancer. The study summary describes the combination as second-line treatment. The primary outcome population is identified as the per-protocol population.
Intervention
Triple combination treatment with bevacizumab, erlotinib and capecitabine. The first part used escalating dose cohorts; the second part was intended to evaluate disease progression using a recommended dose.
Comparator
There was no randomized comparator. The study is registered as non-randomized and single-group, with no masking. Dose cohorts are not a like-for-like control comparison.
Primary question
What was the maximum tolerated dose of capecitabine in the combination, as assessed up to Week 6 (Cycle 1-3)? The registry defines dose-limiting toxicity criteria, but the primary endpoint result is not reported in the registry record.
The record concerns a dose-finding study rather than a randomized test of the combination against another treatment. Its primary question is therefore about tolerability and dose selection: it seeks the capecitabine dose associated with the specified dose-limiting toxicity rule while participants receive the three-drug regimen. The registry describes a two-part design. In the first part, cohorts received escalating doses to determine a maximum tolerated dose. In the second part, the recommended dose was intended to inform assessment of the regimen's effect on disease progression.
The distinction between these aims matters. A dose-finding endpoint is not itself a comparative estimate of efficacy, and the registry's stated secondary outcomes for the second part do not substitute for a reported primary endpoint result. The registry record names disease-progression and other clinical outcomes, but it does not report their results in the registry record for interpretation. With no randomized control group, the record also does not support a direct treatment-versus-control comparison. The clinical question that can be assessed from this record is thus the study's stated dose-finding question and the completeness of its reported results, not whether the regimen outperformed another approach.
3. Trial Design
This was a Phase 1, open-label, non-randomized, single-group study sponsored by Hoffmann-La Roche. The registry lists the study as completed, with enrollment of 32. It was designed around a two-part treatment program for patients with locally advanced and/or metastatic pancreatic cancer. The study summary describes the regimen as second-line treatment and identifies bevacizumab, erlotinib and capecitabine as its components.
In the first part, cohorts received escalating doses of the combination. The stated aim was to identify the maximum tolerated dose, with the primary endpoint specifically naming capecitabine. The registry describes dose-limiting toxicity criteria and gives the primary endpoint timeframe as up to Week 6 (Cycle 1-3). In the second part, the recommended dose was intended to be used to assess the regimen's effect on disease progression. Listed secondary outcomes include freedom from disease progression, disease control, clinical benefit response and overall survival. The registry also lists pharmacokinetic outcomes for erlotinib, its metabolite, capecitabine and capecitabine metabolites in Part 1.
Because allocation was non-randomized and the study was single-group, the dose cohorts should not be interpreted as randomized treatment arms or as a control comparison. The registry lists six dose-level groups in its adverse-event summary, but the available record does not give an arm description for the registered study group. No masking was used, and the study is described as open label.
The registry gives the start as 2009-01 and primary completion as 2013-12. These dates describe the registered study timeline; they do not establish when each outcome was assessed or when analyses were completed. A detailed statistical analysis plan, a sample-size justification, and details of interim analyses are not reported in the registry record. The record lists 14 outcome measures and zero analyses, so the listed study aims should be distinguished from results that are actually reported.
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: Thirty two (32) participants were included however, 2 participants discontinued before reaching the end of the evaluation period of 3 cycles of the complete study regimen (of all 3 drugs) at the recommended dose due to progressive disease.
Overall Study
| Stage | ERL+BEV+CAP Dose Level-1 | ERL+BEV+CAP Dose Level-2 | ERL+BEV+CAP Dose Level-3 | ERL+BEV+CAP Dose Level-4 | ERL+BEV+CAP Dose Level-5 | ERL+BEV+CAP Dose Level-6 |
|---|---|---|---|---|---|---|
| STARTED | 6 | 6 | 6 | 6 | 3 | 3 |
| COMPLETED | 0 | 0 | 0 | 0 | 0 | 0 |
| NOT COMPLETED | 6 | 6 | 6 | 6 | 3 | 3 |
Reasons for not completing this period:
| Reason | ERL+BEV+CAP Dose Level-1 | ERL+BEV+CAP Dose Level-2 | ERL+BEV+CAP Dose Level-3 | ERL+BEV+CAP Dose Level-4 | ERL+BEV+CAP Dose Level-5 | ERL+BEV+CAP Dose Level-6 |
|---|---|---|---|---|---|---|
| Major toxicity | 1 | 0 | 1 | 1 | 0 | 0 |
| Disease progression | 4 | 5 | 3 | 4 | 2 | 3 |
| Death | 0 | 0 | 1 | 0 | 0 | 0 |
| Physician Decision | 1 | 0 | 0 | 0 | 0 | 0 |
| Missing documentation | 0 | 1 | 0 | 0 | 0 | 0 |
| End of study visit not done | 0 | 0 | 1 | 0 | 0 | 0 |
| Participant is unable to swallow drug | 0 | 0 | 0 | 1 | 0 | 0 |
| Adverse Event | 0 | 0 | 0 | 0 | 1 | 0 |
Analysis population for the primary outcome, as defined in the registry: Per-protocol population.
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: Per-protocol population: All participants who received at least 1 evaluation period of 3 cycles of the complete study regimen (of all 3 drugs) at the recommended dose unless withdrawn from therapy earlier because of toxicity.
| Characteristic | Unit | Triple Combination (Bevacizumab/Erlotinib/Capecitabine) |
|---|---|---|
| Participants analysed | 30 | |
| Age, Continuous | years | 63.9 (SD 8.4) |
| Sex: Female, Male: Female | Participants | 15 |
| Sex: Female, Male: Male | Participants | 15 |
6. Primary Endpoint
The registry lists the primary outcome as Part 1: Maximum Tolerated Dose (MTD) of Capecitabine, measured over Up to Week 6 (Cycle 1-3).
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | mg/m^2 BID |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the maximum tolerated dose of capecitabine, expressed in mg/m^2 BID, in the per-protocol population. Its timeframe is up to Week 6 (Cycle 1-3). The registry states that the maximum tolerated dose for each medication was defined as the lowest dose studied that resulted in dose-limiting toxicity in at least 33 percent of participants of the same quality category.
The listed dose-limiting toxicity criteria encompass specified severe toxicities, treatment-related non-hematological toxicities, selected blood and liver-related events, prolonged interruption of oral capecitabine and/or erlotinib, and cancellation of bevacizumab infusions due to adverse events. The registry's full definition includes exceptions and conditions, so the criteria should be read as a protocol-specific operational rule rather than as a general definition of tolerability.
In dose-finding research, a maximum tolerated dose is used to guide selection of a dose for later study. It summarizes a toxicity-based decision within the tested dose levels and observation period; it does not, by itself, establish comparative efficacy or characterize every possible harm. The registry record does not report the observed maximum tolerated dose or the dose-limiting toxicity results. It is therefore possible to describe the endpoint and its intended role, but not to say which dose met the rule or what conclusion the investigators reached.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry describes a dose-escalation study of a three-drug regimen, with the maximum tolerated dose of capecitabine as the primary outcome. That description establishes the study question and the endpoint definition, but it does not provide the endpoint result. The dose selected, the observed dose-limiting toxicities, and the investigators' conclusion about dose are not reported in the registry record. No numerical efficacy results are reported for the listed secondary outcomes in the information available for interpretation.
The study's design limits the kinds of comparisons that can be made. It is registered as non-randomized, single-group and open label. Dose levels describe successive treatment cohorts, not randomized groups with a common comparator. Differences in outcomes across dose levels, if reported elsewhere, would not automatically be attributable to dose alone; this record does not provide the information needed to assess such comparisons. There is no randomized control comparison from which to estimate a treatment effect relative to another treatment.
The registry lists enrollment of 32, but that figure is not a substitute for outcome-specific denominators. It does not show how many participants contributed to the per-protocol primary analysis, how many were evaluated at each endpoint, or whether all enrolled participants completed follow-up. Those details are not reported in the registry record. In particular, a reader cannot infer the precision of an endpoint estimate from the enrollment figure alone.
The adverse-event information gives affected counts within dose-level groups and lists serious adverse-event terms with percentages. These are safety summaries, not evidence of a causal comparison between dose levels. The registry record does not report an analysis adjusting for differences between cohorts, nor does it establish whether the listed terms occurred in the same or different participants. The terms therefore provide a descriptive account, not a complete basis for ranking doses or attributing events to treatment.
Planning calculations based on an observed effect, where displayed alongside the posted results, have a different role from trial results. They are illustrations under stated assumptions, not observed power and not evidence that the trial was designed to achieve a particular precision. Here, the registry record does not report the primary endpoint effect estimate needed to interpret such calculations as a comparison. The sample-size assumptions are explicitly planning assumptions and should not be mistaken for the study's own design justification.
Overall, the record supports a cautious account of what the investigators intended to study: dose tolerability in Part 1 and disease progression and other outcomes in Part 2. It does not support a conclusion about the selected dose, the efficacy of the regimen, or comparative safety. Those questions require results and analysis details that are not reported in the registry record.
9. Statistical Methods Explained
Several statistical ideas help distinguish the design of this study from the evidence actually reported. The explanations below are general unless they describe a feature explicitly stated in the registry record.
What does dose escalation answer?
In a dose-finding study, treatment is given at successive dose levels to learn about tolerability and to guide selection of a dose for further study. The endpoint is the maximum tolerated dose of capecitabine within a combination regimen. The registry defines this using a dose-limiting toxicity rule: the maximum tolerated dose is the lowest dose studied at which dose-limiting toxicity occurs in at least 33 percent of participants of the same quality category. This is a protocol-defined decision rule, not a general statistical property of every dose-finding study.
The endpoint depends on which dose levels were tested, how participants were assigned to those levels, how toxicity was classified, and how long participants were observed. The registry describes escalating cohorts and gives the primary timeframe as up to Week 6 (Cycle 1-3). The actual dose levels, allocation details within cohorts, and observed toxicity results are not reported in the registry record. Without those results, the endpoint definition cannot be used to identify the selected dose.
Why does the single-group design matter?
A randomized comparison is designed to make treatment groups comparable in expectation, helping separate treatment effects from other influences. This study is recorded as non-randomized and single-group, with no masking. Its dose-level groups are not a randomized control group. General differences between cohorts may reflect dose, but can also reflect other differences in who entered a cohort or how outcomes were observed. The record does not report an adjusted comparison that would resolve these issues.
Open-label treatment means participants and investigators are not masked to treatment assignment. In general, knowledge of treatment can influence reporting, assessment, and decisions about care, particularly for outcomes requiring judgment. The registry gives no details of blinded outcome assessment, so it is not possible to determine whether any assessments used safeguards against this possibility.
What does a confidence interval communicate?
In general, a confidence interval expresses the uncertainty around an estimated quantity under the assumptions of a statistical method. A narrow interval indicates greater precision than a wide interval, while both the estimate and interval must be interpreted in light of the design and measurement process. A confidence interval does not correct bias caused by a non-randomized design, incomplete follow-up, or outcome misclassification. The registry record does not report a confidence interval for the primary endpoint or the listed secondary outcomes.
For a dose-finding endpoint, uncertainty also concerns the toxicity pattern observed at the tested levels and how well that pattern informs a dose decision. A threshold-based rule can be affected by small cohort sizes and random variation. In general, a dose decision based on few observations may be unstable; it does not imply that the true toxicity rate is known precisely. The record does not report an uncertainty interval or a statistical model for the maximum tolerated dose.
How should the listed adverse-event counts be read?
The registry presents dose-level counts for participants at risk and participants affected by serious and other adverse events. These are descriptive summaries within the listed groups. A count of affected participants is not the same as a count of individual event episodes, and a participant may in principle experience more than one event. Whether the registry's terms overlap among participants is not reported in the registry record.
In general, comparing proportions across small, non-randomized cohorts can be imprecise. Over-dispersion is a general concern when observed variation is greater than a statistical model expects; in such settings, a model that assumes less variation can give uncertainty estimates that are too small. No over-dispersion model or comparative adverse-event analysis is reported in the registry record. The descriptive counts should therefore not be read as a formal test of dose-related risk.
What is the role of planning calculations?
Sample-size calculations can illustrate how much information a future study might need to detect a chosen effect under explicit assumptions. In general, such calculations depend on the target effect, significance level, allocation, outcome type and statistical approximation. The registry's stated planning assumptions describe a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect specified as a fraction of the observed effect. For time-to-event outcomes, the assumptions refer to total events; for other outcomes, they refer to participants per arm.
These assumptions are described as planning illustrations computed from an observed effect, not as the trial's own design. They are not observed power. The study itself is registered as single-group and non-randomized, so an equal-allocation, two-arm illustration does not describe the registered design. The registry record does not report a sample-size justification or a primary endpoint effect estimate in the information available for interpretation. The assumptions can explain the conditions of a planning exercise, but they cannot establish the actual design rationale or the precision of this study's results.
10. Multiplicity, Missing Data and Analysis Populations
The registry lists 14 outcome measures, including the primary maximum tolerated dose endpoint, pharmacokinetic measures, and several Part 2 clinical outcomes. When a study evaluates multiple outcomes, there is a general risk that some results may appear notable by chance if many statistical tests are conducted. Methods such as prespecifying a primary outcome, adjusting for multiple testing, or treating secondary analyses as exploratory can help frame interpretation. The registry record does not report a multiplicity strategy or a statistical analysis plan, so it is not possible to determine how multiple outcomes were handled.
The registry lists zero analyses. It does not report the primary endpoint result, results for the listed secondary outcomes, or outcome-specific estimates and uncertainty measures. This absence should not be interpreted as evidence of no effect or no events; it means that the results are not reported in the registry record being described.
Missing-data methods and participant flow are also not reported in the registry record. The primary endpoint population is identified as per-protocol, but the record does not state how that population was defined, how many participants contributed to it, or how participants with incomplete toxicity assessments were handled. The analysis populations for the secondary outcomes are not reported. There is also no reported account of missing outcome data, exclusions after enrollment, or sensitivity analyses for missingness.
In general, missing outcomes can affect both the estimate and its precision, especially when missingness is related to a participant's health or treatment experience. Statistical methods cannot resolve this concern without information about the amount and pattern of missingness and the assumptions used in analysis. Because those details are absent, the effect of missing data on any intended result cannot be assessed from the registry record.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| ERL+BEV+CAP Dose Level-1 | 3 | 6 | 50.0% |
| ERL+BEV+CAP Dose Level-2 | 2 | 6 | 33.3% |
| ERL+BEV+CAP Dose Level-3 | 4 | 6 | 66.7% |
| ERL+BEV+CAP Dose Level-4 | 5 | 6 | 83.3% |
| ERL+BEV+CAP Dose Level-5 | 3 | 3 | 100.0% |
| ERL+BEV+CAP Dose Level-6 | 3 | 3 | 100.0% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 5% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| ERL+BEV+CAP Dose Level-1 | 6 of 6 |
| ERL+BEV+CAP Dose Level-2 | 5 of 6 |
| ERL+BEV+CAP Dose Level-3 | 6 of 6 |
| ERL+BEV+CAP Dose Level-4 | 6 of 6 |
| ERL+BEV+CAP Dose Level-5 | 3 of 3 |
| ERL+BEV+CAP Dose Level-6 | 3 of 3 |
Most frequent serious adverse event terms
| Event term | Organ system | ERL+BEV+CAP Dose Level-1 | ERL+BEV+CAP Dose Level-2 | ERL+BEV+CAP Dose Level-3 | ERL+BEV+CAP Dose Level-4 | ERL+BEV+CAP Dose Level-5 | ERL+BEV+CAP Dose Level-6 |
|---|---|---|---|---|---|---|---|
| Dyspnoea | Respiratory, thoracic and mediastinal disorders | 0 of 6 | 0 of 6 | 1 of 6 | 2 of 6 | 0 of 3 | 0 of 3 |
| Bilirubinaemia | Blood and lymphatic system disorders | 1 of 6 | 0 of 6 | 0 of 6 | 1 of 6 | 0 of 3 | 0 of 3 |
| Vomiting | Gastrointestinal disorders | 1 of 6 | 0 of 6 | 0 of 6 | 0 of 6 | 1 of 3 | 0 of 3 |
| Fever | General disorders | 0 of 6 | 1 of 6 | 1 of 6 | 0 of 6 | 0 of 3 | 0 of 3 |
| Vertigo | Nervous system disorders | 0 of 6 | 0 of 6 | 0 of 6 | 0 of 6 | 1 of 3 | 1 of 3 |
| Anemia | Blood and lymphatic system disorders | 0 of 6 | 0 of 6 | 0 of 6 | 0 of 6 | 0 of 3 | 1 of 3 |
| Embolism pulmonary | Blood and lymphatic system disorders | 0 of 6 | 0 of 6 | 0 of 6 | 1 of 6 | 0 of 3 | 0 of 3 |
| Hyponatremia | Blood and lymphatic system disorders | 0 of 6 | 0 of 6 | 0 of 6 | 1 of 6 | 0 of 3 | 0 of 3 |
| Myocardial infarction | Cardiac disorders | 1 of 6 | 0 of 6 | 0 of 6 | 0 of 6 | 0 of 3 | 0 of 3 |
| Abdominal pain | Gastrointestinal disorders | 0 of 6 | 0 of 6 | 0 of 6 | 0 of 6 | 0 of 3 | 1 of 3 |
Most frequent non-serious adverse event terms
| Event term | Organ system | ERL+BEV+CAP Dose Level-1 | ERL+BEV+CAP Dose Level-2 | ERL+BEV+CAP Dose Level-3 | ERL+BEV+CAP Dose Level-4 | ERL+BEV+CAP Dose Level-5 | ERL+BEV+CAP Dose Level-6 |
|---|---|---|---|---|---|---|---|
| Rash erythematous | Skin and subcutaneous tissue disorders | 3 of 6 | 5 of 6 | 4 of 6 | 4 of 6 | 2 of 3 | 2 of 3 |
| Fatigue | General disorders | 6 of 6 | 3 of 6 | 3 of 6 | 3 of 6 | 2 of 3 | 2 of 3 |
| Nausea | Gastrointestinal disorders | 4 of 6 | 2 of 6 | 4 of 6 | 4 of 6 | 2 of 3 | 2 of 3 |
| Diarrhoea | Gastrointestinal disorders | 2 of 6 | 4 of 6 | 4 of 6 | 4 of 6 | 0 of 3 | 2 of 3 |
| Vomiting | Gastrointestinal disorders | 1 of 6 | 3 of 6 | 4 of 6 | 4 of 6 | 1 of 3 | 1 of 3 |
| Palmar-plantar erythrodyesthesia | Skin and subcutaneous tissue disorders | 2 of 6 | 5 of 6 | 3 of 6 | 3 of 6 | 1 of 3 | 0 of 3 |
| Abdominal pain | Gastrointestinal disorders | 3 of 6 | 4 of 6 | 3 of 6 | 2 of 6 | 1 of 3 | 0 of 3 |
| Appetite loss | Metabolism and nutrition disorders | 4 of 6 | 1 of 6 | 2 of 6 | 4 of 6 | 0 of 3 | 1 of 3 |
| Anemia | Blood and lymphatic system disorders | 2 of 6 | 0 of 6 | 2 of 6 | 3 of 6 | 1 of 3 | 1 of 3 |
| Dysphonia | Nervous system disorders | 1 of 6 | 0 of 6 | 2 of 6 | 2 of 6 | 1 of 3 | 1 of 3 |
Serious events with the highest arm-level rate: anemia (33.3%), abdominal pain (33.3%), gastrointestinal haemorrhage (33.3%), vomiting (33.3%), infection (33.3%).
The registry's adverse-event summary is organized by six named dose levels. For each group, it reports numbers at risk and numbers affected by serious adverse events and by other adverse events. These are counts within dose-level cohorts; they should not be added together to create an overall total. Cohort counts describe the recorded groups and do not show whether events were judged to be caused by the study treatment.
The registry lists Anemia, Abdominal pain, Gastrointestinal haemorrhage, Vomiting, Infection and Skeletal pain as the top serious terms, each at 33.3%. The record does not clarify whether these terms occurred in overlapping participants, how the percentages were calculated, or how event severity and relatedness were assessed for this summary. The terms are therefore descriptive and should not be treated as a ranking of risk or as a comparison between treatments.
In general, safety counts depend on the period of observation, the definition of an adverse event, how events are collected, and the population included in the summary. The registry record does not report enough detail to resolve those aspects for these figures. It also does not report a comparative safety analysis or an adjusted estimate across dose levels. The numbers describe the posted adverse-event summaries; they do not establish a causal treatment effect.
12. What the Result Does, and Does Not, Mean
The record establishes that investigators planned a two-part study of bevacizumab, erlotinib and capecitabine in patients with locally advanced and/or metastatic pancreatic cancer. Part 1 was intended to identify a maximum tolerated dose, and Part 2 was intended to evaluate disease progression using a recommended dose. The primary endpoint and its toxicity-based definition are specified, and the registry includes descriptive adverse-event counts by dose level.
However, the results needed to answer the primary question are not reported in the registry record. The maximum tolerated dose, dose-limiting toxicity findings, and conclusions about dose selection are absent. The registry also lists secondary outcomes for disease progression, disease control, clinical benefit response and overall survival, but their results are not reported in the information available for interpretation. Consequently, the study description alone cannot establish what dose was selected or what outcomes participants experienced.
The design also does not support a randomized treatment comparison. Non-randomized dose cohorts can describe events observed at tested dose levels, but they do not provide the same protection against confounding as randomized allocation. Without a control group, an observed outcome could not be directly contrasted with an outcome under another treatment using this registry record.
These limitations do not mean that the study had no findings. They mean that the registry record does not report the analyses and results required to evaluate those findings. The enrollment figure and study status cannot fill that gap, nor can the listed endpoint definitions substitute for results. Any broader conclusion about efficacy, comparative safety, or the suitability of a particular dose would go beyond the information reported in the registry record.
13. Why This Trial Matters Statistically
This study illustrates the distinction between defining a clinical question and reporting evidence that answers it. The registry gives a specific primary endpoint, an operational toxicity rule, and a timeframe, which make the intended dose-finding question clear. But an endpoint definition alone cannot show which dose met the rule. For statistical interpretation, the result and its uncertainty are just as important as the planned measurement.
It also shows why design labels matter. A non-randomized, single-group dose-escalation study has a different purpose from a randomized comparative trial. Cohort summaries can inform a description of tolerability at tested doses, but the cohort structure does not create a control comparison. Recognizing this prevents readers from interpreting dose-level counts as evidence that one dose or treatment is superior.
The record further demonstrates the importance of outcome reporting across a multi-part study. Part 1 focuses on dose selection and pharmacokinetics, while Part 2 lists disease-progression and other clinical outcomes. Without reported analyses, readers cannot connect the stated aims to estimates of effect, precision, or uncertainty. A registry's outcome list is useful for identifying what was intended to be measured, but it is not a substitute for the findings.
Finally, the adverse-event summaries emphasize the need to preserve the distinction between descriptive counts and comparative inference. Counts by dose level can provide context, but incomplete information about denominators, overlap, attribution and analysis limits what they can establish. Clear reporting of participant flow, analysis populations, missing data, and statistical methods would be needed to assess reliability and precision. These are general lessons for interpreting early-phase clinical research, not conclusions about unreported study results.
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 available registry information has important study-specific limitations. Most directly, it does not report the result for the primary endpoint, the maximum tolerated dose of capecitabine. It also does not report the observed dose-limiting toxicities used to determine that endpoint. The intended dose-finding question therefore cannot be answered from the registry record.
The study was non-randomized, single-group and open label. Dose cohorts are not a randomized control comparison, and the record does not provide a comparative analysis that addresses cohort differences. This limits causal interpretation of differences in outcomes across dose levels. The registry lists adverse-event counts, but details about event attribution, assessment procedures, participant overlap across terms, and outcome-specific observation are not reported.
The registry identifies a per-protocol population for the primary endpoint, but does not define that population or report its size. It does not provide participant flow, analysis populations for secondary outcomes, or methods for handling missing data. The number enrolled does not establish how many participants were evaluated for each endpoint. Statistical analysis plans, multiplicity handling, interim analyses, and a sample-size justification are not reported in the registry record.
The registry describes Part 2 outcomes but does not report their results in the information available for interpretation. It is consequently not possible to assess the precision of clinical outcome estimates or whether follow-up was complete. These reporting limitations are distinct from the design limitations: even a carefully conducted study cannot be interpreted fully without endpoint results and analysis details. No conclusion about efficacy or comparative safety can be drawn from the study description and adverse-event summaries alone.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: ATX Study:A Study of Avastin (Bevacizumab), Tarceva (Erlotinib) and Xeloda (Capecitabine) in Patients With Locally Advanced and/or Metastatic Pancreatic Cancer — NCT00925769.
- Registry record last updated 2015-08-07; status last verified 2015-07.
- Results first posted 2015-08-07.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
The registry record describes NCT00925769, a completed Phase 1, open-label, non-randomized, single-group study sponsored by Hoffmann-La Roche. It lists enrollment of 32 and evaluates bevacizumab, erlotinib and capecitabine in patients with locally advanced and/or metastatic pancreatic cancer.
The study had two parts: dose escalation to determine a maximum tolerated dose, followed by planned evaluation of disease progression at a recommended dose. The primary endpoint was the maximum tolerated dose of capecitabine, assessed up to Week 6 (Cycle 1-3) in the per-protocol population. The registry lists 14 outcome measures and zero analyses.
The registry record does not report the primary endpoint result or the listed secondary outcome results. Adverse-event counts by dose level and serious-event terms are descriptive summaries, not a randomized comparison. The dose selected, efficacy findings, analysis methods and missing-data handling are not reported in the registry record.