Whole-Brain Radiation Therapy or Stereotactic Radiosurgery With or Without Lapatinib Ditosylate in Treating Patients With Brain Metastasis From HER2-Positive Breast Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Whole-Brain Radiation Therapy or Stereotactic Radiosurgery With or Without Lapatinib Ditosylate in Treating Patients With Brain Metastasis From HER2-Positive Breast Cancer (NCT01622868), a randomized phase 2 study. The primary outcome is Complete Response (CR) Rate in the Brain at 12 Weeks Post-radiation Therapy (RT) Using the RECIST 1.1 Criteria Based on Brain Magnetic Resonance Imaging (MRI).

2. Clinical Question

3. Trial Design

NCT01622868 was a randomized, open-label, parallel-group phase II study sponsored by the National Cancer Institute. The registry lists enrollment of 143. Its listed start date is 2012-12-06, its primary completion date is 2019-12-30, and its status is completed. The registry identifies the study as a collaborative effort of NRG Oncology and KROG.

Both groups received radiation therapy for brain metastases. The radiation options were whole-brain radiotherapy or stereotactic radiosurgery. The registry describes whole-brain radiotherapy as 15 treatments over 3 weeks and stereotactic radiosurgery as 1 treatment. The intervention group also received lapatinib ditosylate by mouth once daily for 6 weeks. The registry does not report how the choice between the radiation approaches was made or how many participants received each approach.

The primary endpoint was the brain complete-response rate at 12 weeks after radiation therapy, using RECIST 1.1 criteria assessed with brain MRI. The endpoint population description lists evaluable data for 52 eligible participants in the radiation-alone group and 64 in the lapatinib group. These are endpoint-specific analysis counts; the registry does not report a full account of how they relate to all randomized participants.

Other registered outcomes included brain response and progression measures, targeted-lesion measures, and treatment-related adverse events. The registry record does not report the randomization procedure, allocation concealment, sample-size justification, interim analyses, or a detailed statistical analysis plan. Randomization is recorded, but masking is listed as none, so the study was open label.

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:

Analysis population for the primary outcome, as defined in the registry: Evaluable data for determining the 12-week post-RT RECIST response was available for 52 eligible participants on the RT alone arm and 64 on the lapatinib arm.

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: Eligible participants

6. Primary Endpoint

The registry lists the primary outcome as Complete Response (CR) Rate in the Brain at 12 Weeks Post-radiation Therapy (RT) Using the RECIST 1.1 Criteria Based on Brain Magnetic Resonance Imaging (MRI), measured over Baseline and 12 weeks post RT (approximately 12 weeks from start of treatment if SRS and 15 if WBRT).

The primary endpoint is the percentage of analyzable participants with a complete response in the brain at 12 weeks after radiation therapy. Brain MRI was used to assess response under RECIST 1.1. The registry defines complete response as disappearance of all enhancing disease and being off all steroids; the rate is calculated using the number of participants with complete response over the analyzable participants.

This endpoint turns imaging findings into a defined binary outcome: each participant either meets the complete-response definition or does not. As a general statistical concept, a binary endpoint can be summarized as a proportion, which allows a direct description of how often the specified outcome occurred in each group. The result is specific to the RECIST definition, the assessment timing, and the evaluable population. It should not be read as a measure of every aspect of disease control or as a time-to-event outcome.

The timing matters because response can depend on when imaging is assessed. The registry specifies baseline and 12 weeks post-radiation therapy, with timing described as approximately 12 weeks from treatment start for stereotactic radiosurgery and 15 weeks for whole-brain radiotherapy. Thus, the nominal endpoint timing is tied to radiation therapy, while the stated approximate interval from treatment start differs by radiation approach. The registry record does not report details about handling scans that were unavailable or not evaluable.

7. Results Posted in the Registry

• Lapatinib and Radiation Therapy: 0.0% (0/64) versus 5.8% (3/52) on control; risk difference -5.8 percentage points (95% CI -12.1 to 0.6); unadjusted two-proportion z-test p = 0.052.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Complete Response Rate in the Brain at 4 Weeks Post-RT Using the RECIST 1.1 Criteria Based on Brain MRI

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Lapatinib and Radiation Therapy: 1.5% (1/67) versus 3.6% (2/55) on control; risk difference -2.1 percentage points (95% CI -7.9 to 3.6), risk ratio 0.41 (0.04–4.41); unadjusted two-proportion z-test p = 0.447.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Lapatinib and Radiation Therapy and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants With Progression in the Brain Outside the Targeted Measurable Disease Using the RECIST 1.1 Criteria Based on Brain MRI

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Lapatinib and Radiation Therapy: 36.8% (21/57) versus 22.0% (9/41) on control; risk difference 14.9 percentage points (95% CI -2.9 to 32.7), risk ratio 1.68 (0.86–3.28); unadjusted two-proportion z-test p = 0.115.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Lapatinib and Radiation Therapy and the control. The interval gives the range of effects that remain compatible with the data.

Overall Progression Rate in the Brain Using the RECIST 1.1 Criteria Based on Brain MRI

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Lapatinib and Radiation Therapy: 79.4% (54/68) versus 70.4% (38/54) on control; risk difference 9.0 percentage points (95% CI -6.5 to 24.6), risk ratio 1.13 (0.91–1.39); unadjusted two-proportion z-test p = 0.249.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Lapatinib and Radiation Therapy and the control. The interval gives the range of effects that remain compatible with the data.

The registered secondary outcomes cover several related but nonidentical questions: early brain complete response, response and progression under different criteria, progression beyond targeted measurable disease, lesion-specific outcomes, and adverse events. The posted outcomes therefore do not form a single interchangeable measure of treatment effect. They use different definitions, assessment windows, and analysis populations, and some describe response while others describe progression or safety.

Across the posted secondary outcomes, the estimates do not establish a consistent pattern that resolves the primary result. The secondary findings should be interpreted as additional descriptive evidence, not as a replacement for the primary endpoint or a resolution of the disagreement between the sponsor’s primary test and the unadjusted recalculation. The registry record does not report analyses for every listed outcome or a unified statistical model linking these endpoints.

Multiplicity matters because examining many outcomes creates more opportunities for apparently notable results to arise by chance. As a general statistical explanation, adjustment procedures can account for the number and structure of comparisons, but the registry record does not report a multiplicity adjustment plan for the secondary outcomes. Unadjusted secondary comparisons therefore carry less inferential weight than a prespecified primary analysis, especially when the analysis plan is not available in the registry.

These outcomes can still help describe what was measured and how the observed data varied across endpoint definitions. They should not be selectively elevated on the basis of an isolated estimate or p-value. The registry record does not report a complete analysis for all secondary measures, so the secondary evidence remains limited in both breadth and interpretability.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Complete Response Rate in the Brain at 4 Weeks Post-RT Using the RECIST 1.1 Criteria Based on Brain MRI
• Complete Response Rate in the Brain Using the World Health Organization (WHO)/Modified McDonald Criteria Based on Brain MRI
• Objective Response Rate in the Brain Using the RECIST 1.1 Criteria Based on Brain MRI
• Percentage of Participants With Progression in the Brain Outside the Targeted Measurable Disease Using the RECIST 1.1 Criteria Based on Brain MRI
• Targeted Lesion-specific Objective Response Rate Using the RECIST 1.1 Measurement Criteria Based on Brain MRI
• Targeted Lesion-specific Progression Rate Using the RECIST 1.1 Measurement Criteria Based on Brain MRI
• Frequency of Highest Treatment-related Adverse Event Per Participant
• Overall Complete Response Rate in the Brain Using the RECIST 1.1 Criteria Based on Brain MRI

9. Statistical Interpretation

10. Statistical Methods Explained

What does a response rate measure?

The primary outcome is a binary response measure. Each analyzable participant is classified according to whether the MRI findings meet the registry’s complete-response definition. The group rate is the percentage meeting that definition. In this trial, the definition requires disappearance of all enhancing disease and being off all steroids, with response assessed using RECIST 1.1. That makes the reported rate dependent on both the clinical-imaging criteria and the assessment schedule.

As a general statistical explanation, a risk difference compares two group proportions by expressing how far apart they are on the probability scale. A value at zero corresponds to no difference in proportions; a positive or negative value indicates which group has the higher observed proportion, depending on the stated group order. The recalculated comparison is unadjusted, meaning it uses the posted group counts without adjustment for other variables. It is not a substitute for the sponsor’s analysis.

How should a confidence interval be read?

As a general concept, a confidence interval gives a range of values compatible with the data under the method used, while reflecting sampling uncertainty. It is not a guarantee that the true effect lies within the range, and it does not show the probability that a particular treatment effect is true. An interval that includes zero for a risk difference is compatible with no difference as well as with effects in either direction. The recalculated primary interval includes zero. The registry does not report a confidence interval for its own primary Z-test result.

What do the p-values say?

A p-value describes how unusual data at least as discrepant as those observed would be under a specified null hypothesis and analysis model. It is not the probability that the null hypothesis is true, and it does not measure the size or clinical importance of an effect. The registry reports a sponsor Z-test p-value of 0.97. A simple unadjusted recalculation from the posted primary counts gives p = 0.052. Because the registry marks a discrepancy, the sponsor’s own analysis takes precedence. The record does not describe enough of the sponsor’s method to reconstruct or explain its result.

As a general explanation, test results can vary when analyses use different populations, definitions, statistical procedures, or adjustments. Those are possible sources of differences in statistical analyses, not documented explanations for this discrepancy. The registry record does not report which, if any, apply.

Why can sparse outcomes be difficult to estimate?

Only a small number of participants met the primary complete-response definition in the posted endpoint counts. As a general statistical matter, proportions based on few events are sensitive to the observed event count and tend to carry substantial uncertainty. This is one reason to read the point estimates alongside their uncertainty, and to avoid treating a single p-value as a complete description of the evidence. The recalculated interval communicates uncertainty for the simple unadjusted comparison; it does not resolve the discrepancy with the sponsor test.

What do the planning illustrations mean?

The registry includes planning illustrations for some secondary comparisons. They are based on the observed effect and use a two-sided significance level of 0.05, equal allocation, and a normal approximation; the illustrations consider the full observed effect and stated fractions of it. Such figures are hypothetical planning quantities, not participants actually enrolled, and not observed power. They do not show that the trial was designed around those assumptions. The registry record does not report the trial’s own sample-size justification.

As a general statistical explanation, planning calculations depend on assumptions about the effect to detect, the analysis, and the desired operating characteristics. When the observed effect is used to generate an illustration, the result should not be mistaken for independent evidence that the effect is real or reliably estimated. The registry’s planning illustrations therefore cannot establish whether the completed trial had an adequate planned sample size for its primary endpoint.

What is not recoverable from the registry entry?

The registry records randomization and an open-label design, but it does not report the allocation method, concealment procedures, detailed analysis plan, handling of missing data, or interim analyses. It also does not report whether the sponsor’s primary analysis adjusted for any factors, or precisely which analysis population and test implementation produced p = 0.97. Without those details, the summary figures support a descriptive reading and a clearly labeled unadjusted recalculation, but not a full replication of the sponsor’s analysis.

11. Confidence Intervals and Precision

The primary endpoint estimates are 5.8% for radiation therapy and 0.0% for lapatinib plus radiation therapy, based on 52 and 64 evaluable participants, respectively. The recalculated risk-difference interval includes zero. This means the recalculation does not pin down a difference with high precision: values consistent with the data include no difference and effects in either direction. The registry does not provide a confidence interval for the sponsor’s own primary analysis.

As a general statistical concept, interval width reflects both the amount of information in the data and the variability of the estimate. Sparse outcomes make it difficult to estimate a response proportion precisely, because relatively few events contribute to the comparison. The primary counts contain few complete responses, so the observed percentages should not be treated as exact estimates of the underlying response rates.

The sponsor’s reported p-value and the unadjusted recalculation also differ substantially. That conflict is a separate limitation on precision of interpretation: it is unclear from the registry record how the sponsor’s analysis relates to the simple count-based comparison. The sponsor result takes precedence, but the entry does not report its confidence interval or enough method detail to assess the uncertainty around that result directly.

Precision is also affected by the evaluable-population basis. The record gives endpoint counts but does not report the reasons other participants lacked evaluable data, the missing-data strategy, or a comparison of included and excluded participants. Consequently, the stated precision applies to the posted endpoint analysis, and the record does not establish how well it represents all randomized participants.

12. Analyses Reported by the Sponsor

The registry reports a Z-test p-value of 0.97 for the sponsor’s superiority analysis of the primary endpoint. A simple unadjusted recalculation from the posted group counts gives p = 0.052, along with a recalculated risk-difference interval that includes zero. The registry marks this as a discrepancy. Under that circumstance, the sponsor’s own analysis takes precedence; the recalculation is useful only as a transparent comparison based on the posted summary counts.

The two results should not be blended or treated as interchangeable. The sponsor’s p-value is the reported trial analysis, while the recalculated result is an unadjusted computation from the summary figures. The registry record does not report the full sponsor analysis method, the analysis population used for that test, or a confidence interval for the sponsor estimate. It also does not explain why the p-values differ. Possible methodological explanations cannot be attributed to this trial without documentation.

The discrepancy limits what can be inferred from the registry entry. The sponsor result should be identified as the official analysis, while the recalculation should be described with its limitations and not used to override it. The record’s lack of details means readers cannot independently reconcile the sponsor result with the posted counts.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and multiple secondary outcomes, including several measures of brain response and progression. It reports a limited set of analyses, but does not report a multiplicity plan explaining whether or how the analyses were intended to control the chance of false-positive findings across outcomes. A general statistical concern is that the more comparisons are examined, the more likely it is that some may appear notable by chance. Without a reported adjustment strategy or hierarchy of secondary hypotheses, isolated secondary findings should be treated cautiously.

The registry record does not report the trial’s detailed statistical analysis plan, interim analyses, or sample-size justification. The planning illustrations attached to some secondary results are based on observed effects and do not establish the study’s original design assumptions. The record also does not explain whether the primary analysis was prespecified in the same form as the posted result or whether any analyses were exploratory.

The primary endpoint population is described as eligible participants with evaluable data: 52 in the radiation-alone group and 64 in the lapatinib group. The registry does not provide participant flow from enrollment through randomization and endpoint assessment, reasons for missing or unevaluable MRI data, or a description of any imputation or other missing-data methods. It also does not report whether sensitivity analyses assessed the effect of missing outcomes.

Although the study is recorded as randomized, the entry does not report whether the primary analysis followed an intention-to-treat principle, used another defined population, or handled exclusions in a different way. The meaning of “evaluable” is not further specified. These omissions make it difficult to assess potential selection effects or to determine how closely the endpoint population reflects all randomized participants. They also prevent a detailed assessment of how missing data and multiple outcomes might affect the reported inference.

14. 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: headache (4.8%), platelet count decreased (4.2%), gait disturbance (3.2%), anemia (2.8%), alanine aminotransferase increased (2.8%).

The registry reports serious adverse events affecting 17 of 63 participants at risk in the radiation-alone group and 16 of 71 participants at risk in the lapatinib-plus-radiation group. It separately reports other adverse events affecting 49 of 63 and 63 of 71 participants at risk, respectively. These are counts of participants affected, not a measure of severity across all events or a causal comparison by themselves. The registry also records 48 deaths in the radiation-alone group and 53 in the lapatinib-plus-radiation group; the entry does not identify the causes or attribute those deaths to treatment.

The most frequent serious-event terms listed include headache (4.8%), platelet count decreased (4.2%), gait disturbance (3.2%), anemia (2.8%), alanine aminotransferase increased (2.8%), and thromboembolic event (2.8%). These terms describe reported categories; the entry does not provide enough detail to determine timing, duration, overlap among participants, or attribution for each event. The group sizes at risk for the safety reporting differ from the endpoint-specific evaluable counts, so safety counts should not be conflated with the primary response population.

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

The primary endpoint records whether an analyzable participant met a specific imaging-based complete-response definition at a specified assessment. The registry reports low complete-response percentages in both groups. Those results describe the observed endpoint among participants with evaluable data; they do not establish a general effect across all possible measures of brain disease, all treatment schedules, or all participants enrolled.

The sponsor’s reported primary test is p = 0.97. The unadjusted recalculation from the posted counts produces a different p-value, p = 0.052, and a confidence interval that includes zero. Because the registry flags a discrepancy, the sponsor’s own analysis takes precedence. The record does not explain the conflict, so the recalculation should not be used as a replacement for the sponsor’s conclusion. At the same time, the missing analysis details limit independent scrutiny of the sponsor result.

As a general statistical point, a nonsignificant result does not demonstrate that groups are equivalent, and a p-value is not a direct measure of treatment benefit, harm, or clinical importance. The registry entry does not report an equivalence or noninferiority framework for the primary question. Nor does it provide enough detail to infer that the difference between the groups is absent or precisely estimated.

The outcome is tied to RECIST 1.1, brain MRI, and the 12-week post-radiation assessment. Other outcomes use different definitions and timeframes. The findings should therefore be understood within those measurement choices and the endpoint population. The registry record does not report sufficient information about missing endpoint data, analysis populations, or the primary analysis method to extend the result beyond that narrow description.

In short, the entry documents a randomized comparison and posts a primary response result, but the primary inferential evidence is difficult to reconcile from the registry entries alone. The result does not establish whether adding lapatinib changes the complete-response rate, and it should not be interpreted as a recommendation for an individual patient.

16. Why This Trial Matters Statistically

This trial illustrates why a registry result needs to be read at the level of its endpoint definition, analysis population, and statistical method. The primary outcome is not a broad impression of response: it is a binary RECIST 1.1 complete-response measure assessed on brain MRI at a specified time. Its interpretation depends on who had evaluable data and on the criteria used to classify response.

It also illustrates the difference between a sponsor-reported analysis and a recalculation from summary counts. The registry reports a sponsor Z-test p-value of 0.97, whereas an unadjusted recalculation gives p = 0.052. The discrepancy flag makes it especially important to label the recalculation and not present it as the trial’s own analysis. The sponsor result takes precedence, but the absence of methodological detail prevents readers from reproducing or reconciling that result.

The registry’s endpoint population counts are not the same as the overall enrollment figure, and the entry does not explain the missingness or evaluability process. That gap affects interpretation of both the outcome rate and its precision. The study also includes multiple secondary outcomes, but the record does not report how multiplicity was addressed. Together, these features demonstrate why numerical results need context before being treated as confirmatory evidence.

For biostatisticians and clinical researchers, the practical statistical lesson is to separate descriptive estimates, unadjusted recalculations, and the sponsor’s inferential analysis. Each answers a different question. Clear reporting of analysis populations, missing-data handling, multiplicity, and analysis methods would make a registry record more independently interpretable.

17. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The primary outcome counts are based on evaluable data rather than the full enrollment figure. The registry record does not report participant flow, reasons for missing or unevaluable MRI assessments, or how those cases were handled. This leaves uncertainty about whether the endpoint population might differ systematically from participants without evaluable data.

The primary result has a marked discrepancy between the sponsor’s reported Z-test p-value and a simple unadjusted recalculation from the posted counts. The sponsor analysis takes precedence, but the record does not report enough information about its population, test implementation, adjustments, or confidence interval to reproduce or explain it. The recalculation is not a substitute for the sponsor’s result.

The radiation treatment could be whole-brain radiotherapy or stereotactic radiosurgery in either group. The registry does not report how that option was selected, the distribution of participants across radiation approaches, or whether results differed by approach. The primary timing is described relative to radiation therapy, with different approximate timing from treatment start depending on the radiation approach. This complicates a simple interpretation of timing across participants.

The study was open label, and masking is listed as none. The registry does not report whether outcome assessors were masked or whether image review used independent assessment. It also does not report allocation concealment, the randomization procedure, a detailed analysis plan, interim analyses, or a sample-size justification.

Finally, several secondary outcomes address related but distinct questions, and the registry record does not report a multiplicity plan or results for every listed measure. These limits narrow the strength and scope of conclusions that can be drawn from the registry entry.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Whole-Brain Radiation Therapy or Stereotactic Radiosurgery With or Without Lapatinib Ditosylate in Treating Patients With Brain Metastasis From HER2-Positive Breast Cancer — NCT01622868.
• Registry record last updated 2023-10-17; status last verified 2023-09.
• Results first posted 2021-04-13.
• 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.

21. Record Summary

NCT01622868 is a completed randomized, open-label phase II study with registry enrollment of 143. It compared whole-brain radiotherapy or stereotactic radiosurgery alone with the same radiation options plus lapatinib ditosylate in participants with HER2-positive breast cancer metastatic to the brain.

The primary outcome was brain complete response at 12 weeks after radiation therapy under RECIST 1.1, assessed by MRI. The registry reports 5.8% among 52 evaluable participants receiving radiation alone and 0.0% among 64 receiving lapatinib plus radiation. Its sponsor-reported Z-test p-value is 0.97; a simple unadjusted recalculation gives p = 0.052. The registry flags a discrepancy, so the sponsor’s analysis takes precedence. The record does not explain the discrepancy or fully report the analysis methods, missing-data handling, or multiplicity plan.