S0500 Treatment Decision Making Based on Blood Levels of Tumor Cells for Metastatic Breast Cancer Treated With Chemo: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for S0500 Treatment Decision Making Based on Blood Levels of Tumor Cells for Metastatic Breast Cancer Treated With Chemo (NCT00382018), a randomized phase 3 study. The primary outcome is Overall Survival.

2. Clinical Question

3. Trial Design

This was a randomized phase III trial with a parallel design and no masking. The registry lists an overall enrollment of 624 and records the study as completed. The recorded start is 2006-10 and primary completion is 2015-12. The trial concerned women with metastatic breast cancer receiving chemotherapy, and investigated whether circulating tumor cell levels could inform a decision to change treatment or maintain it.

The reported primary outcome is overall survival. The registry defines it as time from registration to death from any cause; those last known to be alive are censored at their last contact. Follow-up is described as every 3 months until progression and then every 6 months for 5 years or until death. That description is the registry’s stated timeframe, not a participant-level account of how long each person was observed.

The posted survival results concern Arm C1 and Arm C2, each labeled as high risk because baseline and Day 22 circulating tumor cell levels were at least 5. The arm counts shown with the survival estimates are 64 for C1 and 59 for C2. These are the counts associated with the posted arm-level estimates, not a restatement of overall enrollment. The separate safety record uses 64 and 56 as the numbers at risk for these groups; the record does not explain the difference between the C2 counts.

The trial record also describes groups continuing usual treatment, continuing current chemotherapy, or switching to a different regimen selected by the physician. It does not explicitly map the two posted C1 and C2 survival labels onto those treatment descriptions. Randomization, parallel allocation, and the trial’s stated treatment-decision question can therefore be described, while the precise relationship between the arm labels and strategy descriptions remains uncertain 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.

Before assignment: 624 patients were registered for screening. 29 patients were excluded due to initial CTC test not completed(17) and ineligible(12). 276 did not have increased CTC levels (Arm A). 319 patients had increased CTC levels. Additional 31 patients were excluded from 319 due to no CTC values on Day 21. Thus total 288 patients in ArmB, ArmC2 and ArmC2.

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 Every 3 months until progression then every 6 months for 5 years or until death.

Overall survival measures the time from registration until death from any cause. Participants last known to be alive are censored at the date of their last contact. This endpoint captures survival regardless of the specific cause of death, as reflected in the registry’s definition.

Time-to-event outcomes are useful when participants can have different lengths of follow-up and the event of interest may occur at different times. Censoring allows information from a participant’s observed follow-up to contribute even if death has not been recorded by the last contact. In general, this approach depends on appropriate follow-up and analysis of censoring; the registry record does not report detailed assumptions or diagnostics for the analysis.

The registry reports median overall survival in months for each high-risk group, together with a 95% confidence interval, and a hazard ratio from a log-rank analysis. A median describes the time point at which the estimated survival distribution reaches its midpoint. It is a summary of the survival experience, not a description of every participant’s survival time. The hazard ratio provides a relative comparison of event rates over follow-up under the analysis model; it is not a difference in median survival. These summaries answer related but distinct questions, so they should be read together rather than treated as interchangeable measures.

7. Results Posted in the Registry

• Arm C1 (Baseline CTCs >= 5, Day 22 CTCs >= 5, High Risk) vs Arm C2 (Baseline CTCs >= 5, Day 22 CTCs >= 5, High Risk): hazard ratio 1.00 (95% CI 0.69–1.47), reported p = 0.980 (as reported in the registry).

8. Statistical Interpretation

9. Statistical Methods Explained

The registry reports an overall-survival comparison between Arm C1 and Arm C2 using a log-rank method and gives the sponsor’s hazard ratio, confidence interval, and p-value. These are time-to-event statistics: they use information about when events occur and account for participants whose event time is not observed before follow-up ends through censoring. The registry’s definition specifies death from any cause as the event and last contact as the censoring time for participants last known alive.

What does the hazard ratio describe?

In general, a hazard is the instantaneous event rate among participants who have not yet had the event, conditional on remaining under observation and event-free up to that time. A hazard ratio compares the hazards between groups according to the model or analysis convention used. A value of 1.0 is the reference for equal hazards. Values on either side can indicate a relative difference, but direction depends on which group is in the numerator. The registry result gives the comparison as C1 versus C2, but it does not state a numerator convention in the posted summary. Avoid assigning a directional interpretation beyond the reported estimate without that convention.

A hazard ratio is not a ratio of the two median survival times, and it is not a direct measure of the number of months gained or lost. It also does not describe an individual participant’s risk at a particular time. Its interpretation as a single summary is most straightforward when the relative hazards are reasonably stable over time. Whether that assumption was assessed, and what diagnostics were used, are not reported in the registry record.

What does the log-rank method contribute?

In general, the log-rank method compares observed event timing between groups over follow-up, taking account of the number of participants at risk at event times. It is commonly used for time-to-event comparisons. The registry names this method for the reported hazard-ratio analysis. It does not provide further detail about model specification, adjustment for other variables, or the handling of ties and other analytic choices. Those details are not reported in the registry record.

The log-rank test’s p-value evaluates how compatible the observed comparison is with a null hypothesis of no difference under the test’s assumptions. The posted p-value is 0.98. It should not be interpreted as the probability that the null hypothesis is true or as evidence that the groups are exactly equivalent. A large p-value can arise when the data are compatible with no difference, but it can also coexist with uncertainty broad enough to include meaningful effects in either direction. The confidence interval is important for understanding that uncertainty.

How should medians and intervals be read?

The posted median survival estimates are 10.7 months for C1 and 12.5 months for C2. Each has its own 95% confidence interval: 9.1 to 14.8 months for C1 and 8.4 to 16.3 months for C2. A median is a summary of the estimated survival distribution. In general, a confidence interval describes the range of values reasonably compatible with the data and the method, under the method’s assumptions. The intervals are not guarantees that the true value lies within them, and they do not show the spread of survival times among participants.

Separate confidence intervals for two arm medians do not constitute a confidence interval for the difference between medians. No such interval is reported. The hazard-ratio interval is the posted interval for the relative time-to-event comparison and answers a different question. The survival medians and the hazard ratio should therefore be understood as complementary descriptions, not as competing estimates of the same quantity.

Why can event-rate variation matter?

In general, time-to-event data can have changing event rates over follow-up, unequal follow-up, and censoring. If the relative hazards change over time, a single hazard ratio can compress different patterns into one summary. An analysis of survival curves or time-specific effects may help clarify such patterns, but the registry record does not report those analyses or whether the proportional-hazards assumption was assessed. General concerns about assumptions are not evidence that an assumption failed in this trial.

Over-dispersion is a concept used for some count outcomes when observed variability exceeds what a simple count model expects. It is not the defining issue for the reported overall-survival endpoint, which is time-to-event. No over-dispersion result is reported for this comparison. The relevant general considerations instead include event timing, censoring, group assignment, and the assumptions behind the chosen time-to-event summary.

The record identifies the study as randomized, phase III, parallel, and unmasked. These design features provide context for the comparison but do not supply missing details about the specific analysis population or treatment mapping. The registry does not report an adjusted analysis, an interim analysis plan, or a sample-size justification for the posted comparison. The sample-size assumptions associated with planning illustrations are not the trial’s own design justification and should not be interpreted as observed power.

10. Confidence Intervals and Precision

Precision concerns how tightly an estimate is determined by the data and method. For C1, the median overall survival estimate is 10.7 months and its 95% confidence interval is 9.1 to 14.8 months. For C2, the median is 12.5 months and its 95% confidence interval is 8.4 to 16.3 months. Each interval conveys uncertainty in its corresponding median estimate. Neither interval describes the range of individual survival times.

The hazard ratio is 1.0, with a 95% confidence interval from 0.69 to 1.47. This interval includes the reference value of 1 and extends to values on either side. Thus, the point estimate alone does not capture the uncertainty in the comparison. The interval is compatible with a range of relative hazard values, rather than pinpointing a single effect. The p-value of 0.98 is another summary of the test result, but it does not replace the interval or show the practical importance of every value that remains compatible with the data.

Confidence intervals are generally calculated using a statistical procedure with assumptions about the data and analysis. A 95% confidence level describes the long-run performance of that procedure under its assumptions; it does not mean that there is a 95% probability that this particular interval contains the true value. The registry record does not give enough methodological detail to assess all assumptions or reconstruct the intervals independently.

The number of participants associated with the survival estimates is 64 for C1 and 59 for C2. The record does not provide event counts for this comparison, and the precision of time-to-event estimates depends importantly on the observed events and censoring, not only on group counts. The safety record gives a different C2 at-risk count, 56, without explaining the difference. These details limit assessment of why the reported intervals have their particular width.

11. Analyses Reported by the Sponsor

The posted hazard ratio is the sponsor-reported estimate: 1.0, with a 95% confidence interval from 0.69 to 1.47 and p = 0.98 from a log-rank analysis. No separate recalculated hazard ratio is presented. For this time-to-event result, the interpretation therefore rests on the sponsor’s own reported analysis rather than a re-calculation from arm-level summary statistics. The registry record does not flag a discrepancy between a recalculation and the reported analysis.

The group-level medians are 10.7 months for C1 and 12.5 months for C2, with their respective confidence intervals. Those summaries do not contain enough information to independently reconstruct the sponsor’s hazard ratio or log-rank test. A comparison of the medians would not reproduce a time-to-event analysis, which uses event timing and censoring across follow-up.

The reported hazard ratio is the principal relative comparison available in the registry record. Its point estimate is at the no-difference reference value, while its interval indicates uncertainty extending to values on either side. The p-value expresses the test result under the stated method; it should not be treated as a measure of the probability that the trial’s null hypothesis is true. The registry does not report additional analysis details that would allow a fuller independent audit of the sponsor’s model or test.

12. Multiplicity, Missing Data and Analysis Populations

The registry record reports a primary endpoint of overall survival and also lists a secondary outcome concerning patients with adverse events related to study drugs. It records 6 outcome measures and 2 analyses, but it does not report a multiplicity strategy for interpreting multiple outcomes or analyses. It is therefore not possible to tell from the record whether adjustment was planned or used, how endpoints were prioritized analytically, or whether other comparisons were prespecified.

Details of interim analyses, stopping boundaries, or interim monitoring are not reported in the registry record. The record also does not provide a sample-size justification for the posted C1 versus C2 comparison, or enough information to determine the planned number of events for that analysis. Planning illustrations based on an observed effect are not observed power and are not evidence of the trial’s original design assumptions.

The analysis population for the survival comparison is not reported in the registry record. The record does not state whether the posted comparison followed an intention-to-treat principle, how participants were handled if they did not follow assigned treatment, or how eligibility for the high-risk comparison was applied in the analysis. It gives arm counts alongside the median estimates, but does not supply a participant flow or analysis-set definition.

The handling of missing data beyond censoring at last known contact is not reported in the registry record. There is no detailed account of loss to follow-up, missing circulating tumor cell measurements, or how missing outcome information was treated. Likewise, adjustment for baseline characteristics, stratification factors, and sensitivity analyses are not described. These are gaps in the available record, not evidence that any particular procedure was or was not used outside the information recorded.

Such details matter because they help readers understand whether the statistical comparison follows the randomized assignment, how incomplete observations affect the result, and how to interpret multiple tests. Without them, interpretation should remain anchored to the reported estimates and avoid assuming an analysis plan that is not documented.

13. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

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: gastrointestinal-other (1.6%), dehydration (1.6%), renal/genitourinary-other (1.6%).

The registry lists serious adverse-event counts for the risk groups, along with counts for other adverse events. For Arm C1, 1 of 64 participants at risk is recorded with a serious adverse event, and 57 of 64 with other adverse events. For Arm C2, the serious adverse-event count is 0 of 56, and the other adverse-event count is 53 of 56. The record also lists Arm B, with 0 of 161 serious adverse events and 152 of 161 other adverse events. These figures are counts as posted, not independently calculated rates.

The listed top serious adverse-event terms are Gastrointestinal-Other, Dehydration, and Renal/Genitourinary-Other, each recorded as 1.6%. The record does not specify in that listing how those terms correspond to individual groups, nor does it give event-level details. A term appearing in the list should not be assumed to represent a unique participant or to be attributable to a particular arm on the basis of the posted summary.

The safety outcome is described as adverse events related to study drugs, but the registry’s summarized counts do not provide enough detail about severity grading, timing, event attribution, or repeated events to support a broader safety comparison. Death counts for the safety groups are not reported in the registry record. These adverse-event figures should be read as recorded safety counts, not as evidence that a strategy is safe or unsafe.

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

The result establishes what the registry reports for the high-risk C1 and C2 groups: median overall survival estimates of 10.7 months and 12.5 months, respectively, and a sponsor-reported hazard ratio of 1.0, with a 95% confidence interval from 0.69 to 1.47 and p = 0.98. The reported analysis does not show a statistically significant difference by the usual interpretation of its p-value. The interval remains important because it spans relative effects on either side of the no-difference reference.

This finding does not prove that the groups have identical survival, or that the strategies under study are equivalent. Equivalence or non-inferiority requires a prespecified margin and an analysis designed for that question; such a margin is not reported in the registry record. A non-significant test alone cannot establish equivalence. Nor does the result establish that a clinically meaningful difference is impossible, given the uncertainty in the interval.

The posted arm labels identify both groups as having elevated circulating tumor cell levels at baseline and Day 22. The registry’s separate treatment descriptions include continuing or changing chemotherapy, but do not explicitly map those descriptions onto C1 and C2 in the results. Consequently, the comparison cannot be translated from these data into a definitive statement about the effect of switching versus maintaining therapy without that mapping.

Overall survival captures death from any cause from registration and is one important time-to-event measure. It does not by itself explain mechanisms, describe all aspects of disease control, or account for outcomes not reported in the comparison. The findings apply to the analyzed groups represented by the registry result; the analysis population and several details needed to judge generalizability are not reported in the registry record.

The result should be read as a specific statistical estimate with uncertainty, not as medical advice or a treatment recommendation. It does not determine what an individual should receive, and it does not answer every question about circulating tumor cell testing, chemotherapy choices, or patient outcomes.

15. Why This Trial Matters Statistically

This trial addresses a statistical challenge common in biomarker-guided research: a measurement may identify a group with a particular risk profile, while the value of changing treatment based on that measurement remains a separate question. The record describes persistent elevation of circulating tumor cell levels at baseline and Day 22 as the high-risk group definition used for the posted survival comparison. Separating risk classification from the effect of a treatment decision is essential when interpreting such results.

The endpoint is time to death, so its analysis must account for both event timing and incomplete follow-up. Reporting medians alongside a hazard ratio gives readers more than one view of the survival data, but these summaries answer different questions. The reported confidence intervals also make uncertainty visible: neither a point estimate nor a p-value alone conveys the full range of effects compatible with the data.

The registry record illustrates why analysis labels and treatment descriptions need to be linked clearly. C1 and C2 are described as high-risk groups, yet the record does not explicitly identify which label corresponds to changing chemotherapy or maintaining it. That gap limits the treatment-strategy interpretation even though a relative survival comparison is reported.

It also demonstrates the importance of transparent reporting about analysis populations, censoring, missing information, multiplicity, and protocol-level assumptions. Those details help readers distinguish what a result estimates from what remains uncertain. The record’s status as a completed randomized phase III trial and its reported survival analysis provide meaningful context, but do not fill in methodological information that is not recorded.

16. 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 immediate limitation for interpretation is that the registry record does not explicitly map Arm C1 and Arm C2 to the treatment-strategy descriptions. Both labels identify high-risk participants by circulating tumor cell levels, but the posted results cannot be confidently assigned to a switch-versus-maintain contrast based on the arm labels alone.

The registry reports 64 participants for C1 and 59 for C2 alongside the survival estimates, while the safety record lists 64 at risk for C1 and 56 for C2. The record does not explain the difference in the C2 counts. It also does not provide event counts for the survival analysis, a detailed participant flow, or the definition of the analysis population. These omissions constrain independent assessment of the survival result and its precision.

Further analytic details are not reported in the registry record, including the handling of missing observations beyond censoring at last contact, any covariate adjustment, assessment of proportional hazards, and the use of interim analyses or multiplicity procedures. Their absence from the record does not establish that the procedures were not used; it means they cannot be evaluated from the registry information.

The reported outcomes do not include a detailed account of treatment exposure, adherence, switching patterns, or the timing of subsequent therapies. The listed safety counts lack event-level information, and deaths in the safety summary are not reported in the registry record. The outcome summary therefore does not support a comprehensive assessment of benefits and harms.

Finally, the posted medians and hazard ratio are statistical summaries for the registry’s comparison. The record’s information is insufficient to establish how representative the analyzed groups are of other patients or settings. Conclusions should remain specific to the reported comparison and its documented uncertainty.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: S0500 Treatment Decision Making Based on Blood Levels of Tumor Cells for Metastatic Breast Cancer Treated With Chemo — NCT00382018.
• Registry record last updated 2017-11-06; status last verified 2017-10.
• Results first posted 2017-04-04.
• 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.

20. Record Summary

NCT00382018 is a completed, randomized, phase III, parallel, unmasked trial sponsored by SWOG Cancer Research Network. The registry lists enrollment of 624, a start date of 2006-10, and primary completion in 2015-12. It concerns treatment decision-making for women with metastatic breast cancer receiving chemotherapy, based on circulating tumor cell levels.

The primary endpoint is overall survival from registration to death from any cause, with censoring at last contact for participants last known alive. The posted high-risk comparison is Arm C1 versus Arm C2, each with baseline and Day 22 circulating tumor cell levels at least 5. Median overall survival is 10.7 months for C1 and 12.5 months for C2. The sponsor-reported hazard ratio is 1.0, with a 95% confidence interval from 0.69 to 1.47 and p = 0.98 by a log-rank method.

The uncertainty interval includes the no-difference reference and permits relative effects in either direction. The registry does not explicitly map C1 and C2 to the treatment-strategy descriptions, and several analysis details are not reported in the registry record. The result therefore supports a careful description of the posted comparison, not a definitive treatment recommendation.