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 A Trial Comparing Safety and Efficacy of Carboplatin and Paclitaxel Plus or Minus Sorafenib (BAY43-9006) in Chemonaive Patients With Stage IIIB-IV Non-Small Cell Lung Cancer (NSCLC) (NCT00558636), a randomized phase 3 study. The primary outcome is Progression Free Survival.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT00558636 |
| Phase | Phase 3 |
| Condition | Carcinoma, Non-Small-Cell Lung |
| Allocation | Randomized |
| Design model | Parallel |
| Masking | Triple |
| Enrollment | 91 |
| Arms in the primary outcome | n/a |
| Lead sponsor | Bayer |
| Results posted | Yes |
2. Clinical Question
Population
Chemonaive patients in the Asia-Pacific study with unresectable stage IIIB non-small cell lung cancer with effusion, or stage IV disease.
Intervention
Sorafenib combined with paclitaxel and carboplatin.
Comparator
Placebo combined with paclitaxel and carboplatin.
Primary question
The registered primary endpoint was progression-free survival: the time from randomization to first documented progression or death before progression was documented. Because the study ended early and 89% of subjects’ data were censored, the record reports event and censoring counts rather than the usual time-to-event measure in days.
The trial asked whether adding sorafenib to the chemotherapy combination changed the course of disease, while also assessing safety. The randomized comparison was between two groups receiving the same named chemotherapy drugs, with sorafenib in one group and placebo in the other. The record identifies the intended analysis population as all randomized subjects, described as the intent-to-treat population.
The registry lists overall survival, best tumor response by RECIST, duration of response, lung cancer symptom scores, and health-related quality-of-life scores among the secondary outcomes. These outcomes represent distinct questions: survival, tumor response, symptom change, and quality of life are not interchangeable measures. The record does not report a single result for these secondary outcomes in the information available for this analysis.
3. Trial Design
This was a randomized, parallel-group Phase 3 trial with triple masking. The registry lists an enrollment of 91 participants. The study was conducted in the Asia-Pacific region and enrolled chemonaive patients with unresectable stage IIIB non-small cell lung cancer with effusion, or stage IV disease. Bayer is listed as the sponsor.
The experimental group received sorafenib with paclitaxel and carboplatin; the comparison group received placebo with paclitaxel and carboplatin. The registry describes oral sorafenib or placebo on Study Days 2-19 and chemotherapy on Study Day 1, with a cycle duration of 21 days in the placebo-group description. The posted intervention description for the sorafenib group is truncated, so it does not give a complete account of that regimen. The record does not specify which groups of people were masked.
The listed start was 2007-09, and primary completion was 2008-05. The study status is terminated. The registry explains that the trial was terminated prematurely after results from an earlier Phase 3 study of similar design showed an overall lack of efficacy and increased mortality in subjects with the squamous subtype. It also states that the available data are presented as descriptive analyses because implementing the statistical analysis plan was limited.
These features matter when interpreting the results. Randomization defines the intended comparison, but early termination and limited follow-up affect how much information the trial could provide. The registry does not report the actual allocation ratio, the detailed masking procedures, or a full account of the statistical analysis plan. Those details should not be inferred from the parallel-group and triple-masking labels.
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: At the time of study termination, 91 (out of 108 screened subjects) of the planned 294 subjects had been screened and randomized. All 91 randomized subjects received at least 1 dose of study drug and were included in both the intent-to-treat (ITT) and safety analysis populations.
Treatment
| Stage | Sorafenib + Paclitaxel + Carboplatin | Placebo + Paclitaxel + Carboplatin |
|---|---|---|
| STARTED | 47 | 44 |
| Start of Maintenance Monotherapy | 2 | 2 |
| COMPLETED | 0 | 0 |
| NOT COMPLETED | 47 | 44 |
Reasons for not completing this period:
| Reason | Sorafenib + Paclitaxel + Carboplatin | Placebo + Paclitaxel + Carboplatin |
|---|---|---|
| Adverse Event | 9 | 2 |
| Withdrawal by Subject | 2 | 0 |
| Premature termination of the study | 35 | 37 |
| Disease progression | 1 | 5 |
Follow-up
| Stage | Sorafenib + Paclitaxel + Carboplatin | Placebo + Paclitaxel + Carboplatin |
|---|---|---|
| STARTED | 12 | 7 |
| COMPLETED | 0 | 0 |
| NOT COMPLETED | 12 | 7 |
Reasons for not completing this period:
| Reason | Sorafenib + Paclitaxel + Carboplatin | Placebo + Paclitaxel + Carboplatin |
|---|---|---|
| Premature termination of the study | 12 | 7 |
Analysis population for the primary outcome, as defined in the registry: It was intended to include all randomized subjects (the intent to treat (ITT) population) in the analysis. Since 89% of subjects were censored, PFS could not be calculated. The number of subjects who Failed (progressed or died before progression) or were Censored are reported.
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.
| Characteristic | Unit | Sorafenib + Paclitaxel + Carboplatin | Placebo + Paclitaxel + Carboplatin | Total |
|---|---|---|---|---|
| Participants analysed | 47 | 44 | 91 | |
| Age, Categorical: <=18 years | Participants | 0 | 0 | 0 |
| Age, Categorical: Between 18 and 65 years | Participants | 38 | 32 | 70 |
| Age, Categorical: >=65 years | Participants | 9 | 12 | 21 |
| Age, Continuous | years | 56.2 (SD 9.3) | 55.4 (SD 12.6) | 55.8 (SD 11.0) |
| Sex: Female, Male: Female | Participants | 20 | 14 | 34 |
| Sex: Female, Male: Male | Participants | 27 | 30 | 57 |
| Eastern Cooperative Oncology Group (ECOG) performance status: Grade 0: Fully active | Participants | 8 | 5 | 13 |
| Eastern Cooperative Oncology Group (ECOG) performance status: Grade 1: Restricted strenuous activity, ambulatory | Participants | 39 | 39 | 78 |
| Geographic region: China | Participants | 35 | 36 | 71 |
| Geographic region: Non-China | Participants | 12 | 8 | 20 |
| Histology: Squamous cell (epidermoid) carcinoma | Participants | 13 | 14 | 27 |
| Histology: Other | Participants | 33 | 30 | 63 |
| Histology: Missing histology | Participants | 1 | 0 | 1 |
| Stage at study entry: Missing | Participants | 3 | 1 | 4 |
| Stage at study entry: Stage IIIB | Participants | 7 | 3 | 10 |
| Stage at study entry: Stage IV | Participants | 37 | 40 | 77 |
6. Primary Endpoint
The registry lists the primary outcome as Progression Free Survival, measured over Up to 5 months after randomization of the first patient.
| Element | Registry entry |
|---|---|
| Result type | Number |
| Measure of spread | not applicable |
| Unit | Participants |
| Endpoint classification | Binary Endpoint |
Progression-free survival (PFS) is defined in the registry as the time in days from randomization to the first observed disease progression, whether radiological or clinical, or death from any cause if death occurred before progression was documented. It therefore combines progression and death into a time-to-event endpoint, rather than treating tumor progression alone as the outcome.
In general, PFS is used to describe how long participants remain alive without documented progression. It can provide an endpoint based on disease course even when overall survival is not the immediate focus. Its interpretation depends on consistent outcome assessment and adequate follow-up, as well as clear handling of participants whose event has not been observed by the end of their available follow-up.
For this trial, the registry says that early termination and censoring of 89% of subjects’ data meant PFS could not be calculated in the usual way. Instead, it reports that the number of participants who failed—progressed or died before progression—or were censored should be reported. The actual counts in those categories are not reported in the registry record. Consequently, the record does not provide a median PFS, a hazard ratio, or another numerical summary of the time to progression or death.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The central interpretive point is that the registry does not provide a conventional comparative PFS result. It defines PFS and explains why the usual time-based measure could not be calculated: the study was terminated early, and 89% of subjects’ data were censored. The event and censoring counts themselves are not reported in the registry record. Without those counts or a time-to-event estimate, the direction and size of a PFS difference between treatment groups cannot be described from this record.
The record characterizes the available analyses as descriptive because of limitations in implementing the statistical analysis plan. It lists no statistical analyses. This means the registry information does not support a claim about statistical significance or a quantified treatment effect for the primary endpoint. Nor does it support a conclusion that the groups were equivalent: lack of a reported comparison is not evidence of no difference.
The enrollment was 91, and the safety groups at risk are listed as 47 for sorafenib plus chemotherapy and 44 for placebo plus chemotherapy. These group counts provide context for the posted adverse-event counts, but they do not restore the missing time-to-event information. The registry also lists several secondary outcomes, including overall survival, tumor response, duration of response, symptoms, and quality of life. Their presence as registered outcomes is not itself evidence of an observed result; results for them are not reported in the registry record.
The early termination context should be kept distinct from evidence about this trial’s own treatment comparison. The registry says the decision followed results from a similar earlier study that indicated an overall lack of efficacy and increased mortality in subjects with the squamous subtype. That explains the context for stopping, but it does not supply a numerical estimate for this trial’s outcomes or establish the result for each participant subgroup in this study.
Any sample-size calculations presented as planning illustrations are not observed power and are not the trial’s own design justification. They depend on assumptions and an observed effect, but this record does not present a usable primary-endpoint comparison from which to interpret the clinical result. Overall, the posted information is most informative about the trial’s design, early termination, endpoint limitations, and descriptive safety counts—not about the magnitude or precision of a PFS effect.
9. Statistical Methods Explained
Randomization is intended to create groups that can be compared without assigning treatment according to participant characteristics. In a randomized trial, the treatment contrast is the difference in outcomes between the assigned groups. The registry identifies this study as randomized and parallel-group, but it does not report the allocation ratio or the method used to generate and conceal the allocation sequence. The fact that the design was randomized does not by itself establish that the groups were balanced on every characteristic, especially in a trial with limited enrollment and early termination.
What does censoring mean for PFS?
In general, a participant is censored in a time-to-event analysis when the event of interest has not been observed by the last point at which outcome information is available. Censoring does not mean that the participant had no event ever; it means the event time is not known from the available follow-up. Time-to-event methods can use partial follow-up, provided their assumptions and data are adequate. A high proportion of censored observations can leave few observed events and make estimates imprecise. The registry reports censoring of 89% of subjects’ data and says PFS could not be calculated. It does not report the event and censoring counts by group, or the details needed to assess censoring patterns.
Why is a time-to-event summary usually useful?
For PFS, the endpoint is the time until the first qualifying progression or death. A time-to-event analysis generally accounts for different lengths of follow-up and incorporates censored observations rather than treating all participants as if they had the same observation period. Common summaries include a median event time or a relative event-rate measure, but neither is reported for this trial. No numerical PFS estimate should be reconstructed from the available description.
A confidence interval, in general, describes the range of values compatible with a statistical estimate under the method’s assumptions. Its width conveys precision: broader intervals indicate less precise estimates. The registry information does not report a PFS estimate or confidence interval, so there is no interval with which to assess the precision of a treatment comparison. A p-value, in general, addresses how compatible data are with a specified null hypothesis under a statistical model; it does not quantify the size or clinical importance of an effect. No inferential PFS result or p-value is reported in the registry record.
What does descriptive analysis tell the reader?
Descriptive analyses present observed data without necessarily making a formal inferential comparison. They can communicate counts and distributions, but they do not by themselves establish that differences between groups are unlikely to reflect random variation. The registry explicitly says the available data are descriptive because of limitations in implementing the statistical analysis plan. It lists no statistical analyses. The absence of a formal analysis should therefore not be replaced by an informal claim of benefit, harm, or equivalence.
The stated sample-size assumptions describe planning illustrations based on an observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation; the intended effect is a stated fraction of the observed effect. Such illustrations are hypothetical planning calculations, not observed power. Observed power is not a substitute for an estimate and its uncertainty. The registry does not report a sample-size justification for this trial, and planning illustrations should not be read as evidence that the trial was designed to detect a particular effect.
Over-dispersion is a general issue in count data: it occurs when the variability exceeds what a simple model assumes, and can make standard errors too small if ignored. The registry’s primary outcome is described as a participant count in the available reporting, but it does not report a model for these counts or an over-dispersion assessment. It would be inappropriate to infer that such a model was used.
Multiplicity refers, in general, to the increased chance of false-positive findings when many outcomes, time points, or comparisons are tested without an appropriate plan. This trial lists multiple secondary outcomes and multiple quality-of-life assessment times. The registry does not report a multiplicity-adjustment procedure or a complete statistical analysis plan. The endpoint list alone does not establish how any multiplicity was handled.
10. Multiplicity, Missing Data and Analysis Populations
The registry identifies the intended PFS analysis population as all randomized subjects, the intent-to-treat population. This approach generally retains participants in the groups to which they were assigned, regardless of treatment received, and preserves the basis of the randomized comparison. However, the record says PFS could not be calculated because the study ended early and 89% of subjects’ data were censored. The event and censoring counts are not reported in the registry record. The amount and pattern of missing outcome information by treatment group are therefore unclear.
In general, missing data and censoring are not identical. Censoring is a defined feature of time-to-event follow-up when an event has not been observed by the last available observation. Other missingness can arise when outcome information is unavailable for reasons not represented by a known follow-up time. The registry does not describe reasons for missing data, follow-up completeness by group, or sensitivity analyses to examine how missing information might affect results.
The study lists several secondary outcomes, including overall survival, tumor response, duration of response, symptom scores, and health-related quality-of-life measures at different treatment cycles. Multiple outcomes and assessment times create multiplicity concerns in general because repeated testing can increase the chance of at least one apparently positive result. A prespecified testing hierarchy or adjustment can address this issue, but such procedures are not reported in the registry record. The registry also says implementation of the statistical analysis plan was limited and describes available analyses as descriptive.
The record does not report whether interim analyses were planned or conducted, how sample size was justified, or what procedures were used for missing data beyond the PFS censoring description. These omissions limit evaluation of the planned inferential framework and of how early termination affected the analysis.
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 |
|---|---|---|---|
| Sorafenib + Paclitaxel + Carboplatin | 11 | 47 | 23.4% |
| Placebo + Paclitaxel + Carboplatin | 3 | 44 | 6.8% |
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 |
|---|---|
| Sorafenib + Paclitaxel + Carboplatin | 45 of 47 |
| Placebo + Paclitaxel + Carboplatin | 39 of 44 |
Most frequent serious adverse event terms
| Event term | Organ system | Sorafenib + Paclitaxel + Carboplatin | Placebo + Paclitaxel + Carboplatin |
|---|---|---|---|
| Febrile Neutropenia | Infections and infestations | 3 of 47 | 0 of 44 |
| Erythema multiforme | Skin and subcutaneous tissue disorders | 2 of 47 | 0 of 44 |
| Rash/Desquamation | Skin and subcutaneous tissue disorders | 2 of 47 | 0 of 44 |
| Dyspnea (shortness of breath) | Respiratory, thoracic and mediastinal disorders | 1 of 47 | 1 of 44 |
| Allergic reaction | Immune system disorders | 0 of 47 | 1 of 44 |
| Neutrophiles | Blood and lymphatic system disorders | 1 of 47 | 0 of 44 |
| Infection (documented clinically), Lung (Pneumonia) | Infections and infestations | 1 of 47 | 0 of 44 |
| Infection with normal ANC, blood | Infections and infestations | 1 of 47 | 0 of 44 |
| Pain, abdomen NOS | General disorders | 1 of 47 | 0 of 44 |
| Hypoxia | Respiratory, thoracic and mediastinal disorders | 1 of 47 | 0 of 44 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Sorafenib + Paclitaxel + Carboplatin | Placebo + Paclitaxel + Carboplatin |
|---|---|---|---|
| Neutrophils | Blood and lymphatic system disorders | 31 of 47 | 25 of 44 |
| Alopecia | Skin and subcutaneous tissue disorders | 18 of 47 | 16 of 44 |
| Rash/Desquamation | Skin and subcutaneous tissue disorders | 23 of 47 | 6 of 44 |
| Platelets | Blood and lymphatic system disorders | 18 of 47 | 7 of 44 |
| Leukocytes | Blood and lymphatic system disorders | 14 of 47 | 10 of 44 |
| Fatigue | General disorders | 11 of 47 | 12 of 44 |
| Hand-foot skin reaction | Skin and subcutaneous tissue disorders | 19 of 47 | 2 of 44 |
| Diarrhea | Gastrointestinal disorders | 14 of 47 | 7 of 44 |
| Neuropathy: sensory | Nervous system disorders | 8 of 47 | 12 of 44 |
| Hemoglobin | Blood and lymphatic system disorders | 12 of 47 | 5 of 44 |
Serious events with the highest arm-level rate: febrile neutropenia (6.4%), erythema multiforme (4.3%), rash/desquamation (4.3%), allergic reaction (2.3%), dyspnea (shortness of breath) (2.3%).
The registry lists 47 participants at risk for serious adverse events in the sorafenib plus paclitaxel and carboplatin group, with 11 affected. In the placebo plus paclitaxel and carboplatin group, 44 were at risk, with 3 affected. These are descriptive counts; they do not establish that treatment caused an event, and they should be interpreted in light of the group sizes and the limited study information.
The listed serious adverse-event terms and percentages are febrile neutropenia, 6.4%; erythema multiforme, 4.3%; rash/desquamation, 4.3%; allergic reaction, 2.3%; dyspnea (shortness of breath), 2.3%; and thrombosis/embolism (vascular access), 2.3%. The record does not provide a full group-by-group breakdown for these terms in the information, so the percentages should not be treated as a complete comparative safety analysis.
For other adverse events, the registry lists 45 affected out of 47 at risk in the sorafenib group and 39 affected out of 44 at risk in the placebo group. These figures are distinct from the serious-event counts. Death counts are not reported in the registry record. The adverse-event information is descriptive and does not supply adjusted comparisons, uncertainty intervals, or a complete account of event timing and severity.
12. What the Result Does, and Does Not, Mean
The registry establishes that this randomized Phase 3 trial compared sorafenib plus paclitaxel and carboplatin with placebo plus the same named chemotherapy drugs in chemonaive patients with advanced non-small cell lung cancer. It also records that the trial was terminated early and that the available data are described as descriptive because of limitations in implementing the statistical analysis plan. These details define what can be concluded from the posted information.
For the primary endpoint, the record defines PFS but says it could not be calculated because 89% of subjects’ data were censored. It does not report the event and censoring counts, a time-based estimate, a confidence interval, or an inferential comparison. Therefore, the posted record does not quantify how the treatment groups differed in PFS, nor how precise such a difference might be. It also does not establish that the treatments had the same effect.
The record gives some descriptive safety counts and lists serious adverse-event terms with percentages. These observations can describe what was recorded, but they are not a complete comparative safety analysis and do not by themselves establish causation. The record does not report death counts. Secondary outcomes are listed, but results for them are not reported in the registry record.
The earlier study’s results are reported as the reason for premature termination. That context should not be confused with a direct estimate of this trial’s own treatment effect. The trial’s early end and high censoring limit the information available for its primary endpoint. A reader should distinguish the registered question from the evidence the record is able to present: the question was comparative, but the available PFS information is descriptive and incomplete.
13. Why This Trial Matters Statistically
This trial illustrates the difference between a planned clinical question and the evidence that remains available after early termination. The registered primary endpoint was a time-to-event outcome, but the registry says that early termination and censoring of 89% of subjects’ data prevented calculation of PFS. The available endpoint description therefore does not provide the usual estimate of time to progression or death, or an interpretable comparative measure with uncertainty.
The record also highlights why analysis populations and missing follow-up matter. The intended PFS population was all randomized participants, but a stated intention to analyze everyone does not recover outcome information that was not observed. Knowing how many participants were censored, when follow-up ended, and whether censoring differed by group would help readers evaluate the available evidence; those details are not reported in the registry record.
It also demonstrates the importance of separating descriptive safety reporting from comparative inference. The record contains affected counts and selected serious adverse-event terms, but these data are not a substitute for a complete analysis of event rates, uncertainty, and group differences. The same principle applies to the listed secondary outcomes: their registration does not mean that results or formal comparisons are available.
Finally, the registry’s explanation of termination distinguishes the trial’s decision context from its own results. The earlier study influenced the decision to stop, while the posted information for this trial remains limited. This distinction is important when interpreting a trial record that reports a compelling rationale for termination but does not present a complete outcome analysis.
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 most consequential limitation is the lack of a calculable PFS result. The registry reports that 89% of subjects’ data were censored and that PFS could not be calculated, but it does not report the number of participants who failed or were censored in each group. No PFS estimate, confidence interval, or comparative test is reported. The amount of follow-up and the distribution of censoring by group are not described.
Early termination reduced the available information. The record states that the study was stopped prematurely after results from a similar earlier Phase 3 study raised concerns about overall efficacy and mortality in subjects with the squamous subtype. The registry does not provide a complete account of how the decision affected planned recruitment, follow-up, or outcome assessment in this trial.
The record describes the available data as descriptive because of limitations in implementing the statistical analysis plan, and it lists no statistical analyses. The full analysis plan, sample-size justification, interim-analysis procedures, and multiplicity procedures are not reported in the registry record. The secondary outcomes are listed, but corresponding numerical results are not reported in the information available for this analysis.
Safety information is also limited. The registry lists serious and other adverse-event counts and selected serious-event terms, but does not provide death counts, a complete event breakdown by group, or uncertainty estimates. The intervention description for the sorafenib group is truncated. These omissions constrain interpretation of both treatment exposure and comparative safety.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: A Trial Comparing Safety and Efficacy of Carboplatin and Paclitaxel Plus or Minus Sorafenib (BAY43-9006) in Chemonaive Patients With Stage IIIB-IV Non-Small Cell Lung Cancer (NSCLC) — NCT00558636.
- Registry record last updated 2013-12-27; status last verified 2013-12.
- Results first posted 2010-10-20.
- 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
NCT00558636 was a randomized, triple-masked, parallel-group Phase 3 trial sponsored by Bayer, with an enrollment of 91. It compared sorafenib plus paclitaxel and carboplatin with placebo plus paclitaxel and carboplatin in chemonaive patients with unresectable stage IIIB non-small cell lung cancer with effusion, or stage IV disease.
The study started in 2007-09, reached primary completion in 2008-05, and is listed as terminated. The primary endpoint was PFS. The registry says early termination and censoring of 89% of subjects’ data prevented calculation of PFS; the event and censoring counts are not reported in the registry record. The available analyses are described as descriptive because of limitations in implementing the statistical analysis plan.
The record lists secondary survival, tumor response, symptom, and quality-of-life outcomes, and reports descriptive adverse-event counts. It does not report a conventional comparative PFS estimate, statistical analysis results, or death counts.