S0635: Erlotinib and Bevacizumab in Stage IIIB and IV Bronchioloalveolar Carcinoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for S0635: Erlotinib and Bevacizumab in Stage IIIB and IV Bronchioloalveolar Carcinoma (NCT00436332), a na phase 2 study. The primary outcome is Overall Survival.

2. Clinical Question

The registry record describes a phase II study of erlotinib and bevacizumab for patients with stage IIIB or stage IV bronchioloalveolar carcinoma (BAC) and adenocarcinoma with BAC features (ADENOBAC). Its primary outcome is overall survival, measured from registration and followed for a maximum of 3 years. The record does not report the observed median overall survival or its confidence interval, so the primary result cannot be interpreted from the registry entry.

That is a descriptive question about outcomes in the study group, not a direct estimate of how outcomes compare with an untreated group or another regimen. Without a concurrent comparator, the record cannot by itself separate treatment effects from differences in patient characteristics, disease course, other care or follow-up. The registry rationale describes reasons for studying the combination; it is not a report of evidence that the combination changed survival.

3. Trial Design

NCT00436332 is listed as a completed phase II trial sponsored by SWOG Cancer Research Network. The registry lists enrollment of 84 and identifies lung cancer as the condition. Its official title specifies stage IIIB and IV bronchioloalveolar carcinoma and adenocarcinoma with BAC features. The brief summary describes stage III or stage IV non-small cell lung cancer. Detailed eligibility criteria are not reported in the registry record, so the exact enrolled disease definition cannot be reconstructed from the title and summary alone.

The design is listed as single group, with one group named Erlotinib and Bevacizumab. The interventions are bevacizumab, classified as biological, and erlotinib hydrochloride, classified as a drug. Allocation is listed as NA and masking as none. The record does not describe random assignment, blinding, or a concurrent comparison group. Thus, the study's structure supports describing the outcomes recorded for its treatment group, but not a randomized estimate of the combination's effect relative to another intervention.

The primary outcome is overall survival in months, summarized as a median with a 95% confidence interval. The timeframe is from date of registration to a maximum of 3 years. Listed secondary outcomes include progression-free survival, response assessed by RECIST criteria versus a central computer-assisted image-analysis system in patients with measurable disease, and frequency and severity of toxicities. The registry record does not provide the outcome values or detailed analysis methods for these outcomes.

The record lists enrollment of 84, while the safety entries list 79 at risk for serious adverse events and 79 at risk for other adverse events. These are distinct registry fields with different stated contexts; the record does not explain their relationship. Trial start is listed as 2007-07, primary completion as 2019-07-20, and status as completed. Further details about recruitment, treatment delivery, follow-up completeness and protocol deviations are not reported in the registry record.

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:

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 Overall Survival, measured over From date of registration to maximum of 3 years.

The primary endpoint is overall survival, reported in the registry as a median measured in months, with a 95% confidence interval. Its stated timeframe begins at registration and extends to a maximum of 3 years. The registry names the endpoint but does not report a median value or confidence limits. It also does not give a detailed operational definition of the event or describe how follow-up ended for participants without a recorded event.

In general, overall survival is used in clinical research because it captures a patient-important time-to-event outcome and does not depend on a judgment that a tumor has responded on imaging. A median is the point in the estimated survival-time distribution at which half of the distribution lies on either side; it is not the survival time of every participant. A confidence interval conveys statistical uncertainty around an estimate. Its width and limits help readers assess how precisely the median is estimated, but neither the estimate nor its precision can be evaluated without the reported values.

In general, time-to-event analyses can include participants whose event has not been observed by the end of follow-up by accounting for their available follow-up through censoring. The registry record does not state how censoring, loss to follow-up, or the maximum follow-up limit was handled analytically. It therefore supports identifying the intended endpoint and timeframe, but not assessing the endpoint calculation or the resulting survival estimate.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Overall survival is a time-to-event outcome: the quantity of interest is elapsed time from a stated starting point until an event. The registry specifies registration as the starting point and names a maximum follow-up of 3 years. It does not state the detailed event definition or the statistical procedure used. In general, survival methods are designed to use information from participants with different follow-up lengths, including those whose event status is not observed by the end of follow-up.

What does a median survival estimate express?

In general, the median is a location summary for a time-to-event distribution. It describes the time at which the estimated survival function reaches the halfway point, rather than an average time or a guarantee about an individual participant. The registry's chosen summary is a median in months. No median value is reported in the registry record, so the estimate itself cannot be interpreted.

What does the confidence interval add?

A confidence interval describes uncertainty around an estimate under the assumptions of the analysis. In repeated use of a method, intervals constructed at a stated confidence level are designed to cover the underlying parameter at that rate. For a reader, the interval's endpoints and width help indicate which values are compatible with the data and method. They do not eliminate uncertainty, and they are not a range containing a specified proportion of individual participants. The registry names a 95% confidence interval for the primary median but does not report its limits. Its precision therefore cannot be judged.

Why does the single-group design constrain inference?

In general, a concurrent control group helps distinguish changes associated with an intervention from changes that might have occurred over time or under other care. A single-group study can describe outcomes among its participants, but the observed course alone does not supply the counterfactual outcome those same participants would have experienced without the intervention or with another intervention. Differences from a historical benchmark would also require careful consideration of whether populations and methods were comparable. No benchmark comparison is reported in the registry record.

The registry marks the design as single group, allocation as NA, and masking as none. Those entries do not describe randomization or blinded outcome assessment. The record does not report adjustment methods, a model for the primary endpoint, or a prespecified statistical analysis plan. It would therefore be inappropriate to infer a particular survival model, covariate adjustment, stratification, or testing procedure.

How should a hazard ratio be understood?

In general, a hazard ratio compares event rates over time between groups, subject to the model and assumptions used. It is not the same as a ratio of median survival times, nor does it state the probability that an individual benefits. A hazard ratio requires a comparison; the registry record describes only one group and does not report a sponsor-reported hazard ratio. The general interpretation cannot be turned into a study-specific estimate in the absence of such a result.

Why can over-dispersion matter?

In general, over-dispersion occurs when observed variation in count data exceeds what a statistical model assumes. If ignored, standard errors and confidence intervals can be too optimistic, depending on the model and setting. That consideration may matter for some count outcomes, but the registry identifies overall survival as a time-to-event endpoint and does not report a count model or an analysis involving over-dispersion. It is not possible to say whether this issue arose in the study's analyses.

The registry's planning assumptions describe illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two hypothetical arms and a normal approximation. For time-to-event planning, the stated approach concerns a total number of events; other planning illustrations concern participants per arm. These are planning illustrations, not observed power and not a description of this single-group trial's actual allocation. No numerical planning results are reported in the registry record. In general, a planning calculation depends on its assumptions and should not be mistaken for evidence that an effect was observed or that a study had a particular probability of success.

The record lists zero analysis records. It does not report a statistical analysis plan, interim analyses, sample-size justification for the actual trial, model diagnostics, or the handling of censoring and missing data. These details are necessary for a fuller appraisal of how the outcome estimates were produced. Statistical concepts can explain what the endpoint and stated interval would mean, but they cannot fill in unreported results or methods.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists overall survival as the primary outcome and names progression-free survival, response assessment and frequency and severity of toxicities among the secondary outcomes. It also lists four outcome measures. A multiplicity plan describing how the possibility of multiple statistical tests was addressed is not reported in the registry record. Without that plan and the results, it is not possible to assess how outcome selection, repeated testing or interpretation of secondary findings was handled.

The record does not report missing-data counts, reasons for missingness, follow-up completeness, or methods for handling missing observations. For a time-to-event endpoint, details about censoring and loss to follow-up are important because they affect how observed follow-up contributes to an estimate. The registry states a maximum follow-up of 3 years but does not specify the analytic treatment of participants whose follow-up ended earlier or whose status was unavailable.

Analysis populations are also not defined in the registry record. Enrollment is listed as 84, while the safety entries list 79 at risk for serious adverse events and 79 at risk for other adverse events. The registry does not explain the distinction among these counts or specify the population used for overall survival. The record lists zero analysis records, so it does not provide an outcome analysis population or an analytic result.

In general, transparent reporting of endpoint-specific denominators, missingness, censoring and population definitions allows readers to understand which observations support an estimate and how exclusions may affect interpretation. Those details are not reported in the registry record. Their absence does not establish that data were missing or that a particular method was used; it means the record does not enable those questions to be answered.

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: left ventricular diastolic dysfunction (1.3%), retinal detachment (1.3%), ulcer, gi - stomach (1.3%), inf w/normal anc or gr 1-2 neutrophils - lung (1.3%), carbon monoxide diffusion capacity (dl(co)) (1.3%).

The registry safety entry lists one group, Erlotinib and Bevacizumab. For serious adverse events, it reports 79 at risk and 6 affected. For other adverse events, it reports 79 at risk and 77 affected. It also lists 56 deaths. These are registry counts and should be read with their stated categories; the entry does not explain the relationship between these safety denominators and the enrollment of 84.

The listed serious adverse-event terms include left ventricular diastolic dysfunction, retinal detachment, ulcer, GI - stomach, infection with normal ANC or grade 1-2 neutrophils in lung, carbon monoxide diffusion capacity, and creatinine. Each is listed with 1.3%. The record does not provide additional context about event timing, duration, clinical course, or attribution to study treatment.

In general, an adverse event recorded during a study is not automatically caused by the intervention. Safety counts describe events under the reporting definitions and observation process; assessing causality requires clinical and methodological context. The registry entry does not provide that context or explain how repeated events were counted. The event counts should therefore not be converted into a causal claim or treated as a complete account of tolerability.

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

The registry record establishes that a completed phase II, single-group trial evaluated a combination of erlotinib and bevacizumab and named overall survival as its primary outcome. It provides the endpoint's unit, summary measure, confidence-interval type and follow-up timeframe. It also names secondary outcomes and includes safety counts for the treatment group. These details identify what the study set out to examine and what the safety entry records.

The record does not report the primary median overall survival, its 95% confidence interval, or statistical analysis results. It does not report results for the listed secondary outcomes. Therefore, the registry entry does not establish the size or precision of a survival estimate, whether an outcome differed from a comparator, or how observed outcomes relate to another treatment. The single-group design has no concurrent control group in the registry record, limiting causal interpretation even if a descriptive outcome estimate were available.

The registry rationale describes biological reasons for combining the interventions. Such rationale motivates a study but is not evidence that a clinical outcome improved. Likewise, the completion status and enrollment entry describe trial administration, not efficacy. The safety counts describe recorded events within the registry's categories, but do not establish that the intervention caused them or that all relevant safety information is included.

In general, an estimate from a clinical study is interpreted alongside its uncertainty, design, population, outcome definitions, follow-up and analysis methods. Several of those details are not reported in the registry record, and the key primary outcome value is absent. The most supportable reading is consequently narrow: the study registered a single treatment group and planned to assess overall survival and other outcomes, while the posted registry entry does not report outcome estimates sufficient to characterize the results.

13. Why This Trial Matters Statistically

This trial is statistically notable for the distinction between a registered study question and a reported result. The record clearly names a primary time-to-event endpoint and its intended summary, but it does not include the median or confidence interval. A defined endpoint is essential for interpretation, yet it does not tell readers what the study observed. Reporting the estimate and its uncertainty is necessary to assess both the outcome's magnitude and precision.

The single-group design also makes the inferential target important. A treatment-group outcome can describe the experience recorded in that group, but it does not provide a randomized comparison. Readers must distinguish descriptive evidence from a treatment-effect contrast. The registry marks the design as single group, allocation as NA, and masking as none; those features do not support interpreting this entry as a like-for-like control comparison.

The safety reporting illustrates another statistical reading issue: at-risk counts are tied to specific event categories, and the registry does not explain their relationship to the enrollment value. Keeping denominators attached to their category avoids implying that all counts arise from an identical analysis population. Likewise, event percentages in the serious-event listing should be understood as registry-reported figures, not as proof of causal attribution.

In general, confidence intervals, analysis populations, censoring rules and multiplicity methods are central to judging how robustly findings answer a clinical question. Their absence from this record means that methodological uncertainty remains. The entry is still useful for identifying the trial's design, interventions, outcome intentions and safety reporting, while also showing why registry descriptions should not be mistaken for a complete statistical report.

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.

Several limitations are specific to what is reported for this study. First, the registry does not include the primary endpoint result: the median overall survival and 95% confidence interval are not reported in the registry record. Without them, the primary outcome cannot be assessed for magnitude or precision. Results for progression-free survival, response and toxicity outcomes are also not reported in the entry.

Second, the design is single group, with no concurrent comparator described. Any outcome estimate from the treatment group would be descriptive rather than a randomized estimate against another treatment. The registry does not report an external benchmark analysis, and no comparison should be inferred from the study title or rationale.

Third, details needed to appraise endpoint analysis are absent. The registry does not report the operational event definition, censoring rules, follow-up completeness, analysis populations, covariate adjustment, statistical model or missing-data methods. It also does not report a multiplicity plan, interim analyses or a sample-size justification for the actual trial. The planning assumptions listed in the record are illustrations and do not supply those missing design or analysis details.

Fourth, the record lists enrollment of 84 and safety at-risk counts of 79 for both serious and other adverse events, without explaining their relationship. The safety entry lists 56 deaths but does not describe whether or how that count relates to the primary survival analysis. These counts cannot resolve the absence of an endpoint-specific analysis or its denominator.

Finally, the official title and brief summary describe the disease population in different terms, and detailed eligibility criteria are not reported in the registry record. This limits a precise assessment of who was eligible and to whom any results would apply. These are reporting limitations; they do not establish that the unreported procedures or information were absent from the full study documentation.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: S0635: Erlotinib and Bevacizumab in Stage IIIB and IV Bronchioloalveolar Carcinoma — NCT00436332.
• Registry record last updated 2020-04-02; status last verified 2020-03.
• Results first posted 2016-12-22.
• 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

NCT00436332 is a completed phase II, single-group study sponsored by SWOG Cancer Research Network. It evaluates erlotinib hydrochloride with bevacizumab in the disease population described in the official title as stage IIIB and IV bronchioloalveolar carcinoma and adenocarcinoma with BAC features. The registry lists enrollment of 84, allocation as NA, and masking as none.

Overall survival is the primary outcome, specified as a median in months with a 95% confidence interval from registration to a maximum of 3 years. Progression-free survival, response assessment and toxicities are also listed. The primary and secondary outcome results are not reported in the registry record, and zero analysis records are listed.

The safety entry lists 79 at risk and 6 affected for serious adverse events, 79 at risk and 77 affected for other adverse events, and 56 deaths. The registry does not explain how these counts relate to enrollment or the primary analysis. The entry supports describing the study structure and registered outcomes, but not estimating a treatment effect or drawing a comparative conclusion.