Sorafenib Tosylate in Treating Patients With Metastatic, Locally Advanced, or Recurrent Medullary Thyroid Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Sorafenib Tosylate in Treating Patients With Metastatic, Locally Advanced, or Recurrent Medullary Thyroid Cancer (NCT00390325), a na phase 2 study. The primary outcome is Objective Response Rate of Sorafenib Tosylate in Metastatic Medullary Thyroid Carcinoma in Setting of Inherited Tumor Syndromes as Well as in Setting of Sporadic Medullary Thyroid Cancer.

2. Clinical Question

The clinical question is about tumor response in patients with medullary thyroid carcinoma, rather than a comparison of sorafenib with another treatment or with no treatment. The registry record describes a phase II study and gives an overall enrollment of 21. It identifies hereditary and sporadic disease as separate settings, but does not provide the number enrolled in each group or a detailed breakdown of the enrolled population.

The primary outcome is called objective response rate. It is a count of participants meeting response criteria, with MRI assessment described in the record. The stated timeframe extends up to 4 weeks after the last dose of sorafenib tosylate. The record says a 95% confidence interval would be calculated to estimate response frequency, but it does not report the number of responders or the resulting interval. Accordingly, the registry record does not establish the observed response frequency or its precision.

The study description also names outcomes involving calcitonin, carcinoembryonic antigen, imaging, tumor biology, toxicity, and RET gene defects. These outcomes broaden the questions explored, but their presence in the record does not substitute for reported results. Without results or a concurrent comparator, the record does not support a comparative treatment-effect estimate.

3. Trial Design

This is a phase II, single-group study sponsored by the National Cancer Institute (NCI). The registry record lists 21 participants as enrolled, describes the study as completed, and gives a start date of 2006-11-03 and a primary completion date of 2017-01-30. These dates describe the registry record; they do not by themselves explain recruitment duration, follow-up completeness, or the timing of each participant’s outcome assessment.

The intervention is sorafenib tosylate. The treatment description says participants received it by mouth twice daily while disease progression or unacceptable toxicity was absent. No separate control intervention is described. The registry identifies Arm A as hereditary medullary thyroid carcinoma and Arm B as sporadic medullary thyroid carcinoma. Those disease categories are not equivalent to randomized treatment groups: both are associated with the study intervention, and the registry record explicitly does not frame them as a like-for-like control comparison.

The study is listed as having no masking. The allocation field is recorded as NA, consistent with the single-group design rather than allocation between competing treatment arms. The registry record does not report further operational detail about how participants entered the hereditary or sporadic group, nor does it describe a randomization procedure.

The primary outcome is an objective-response count assessed using MRI and Response Evaluation Criteria in Solid Tumors/World Health Organization response criteria. Its stated timeframe is up to 4 weeks after the last dose. A 95% confidence interval is described as the planned way to estimate response frequency. The actual response count and interval are not reported in the registry record. The record identifies 1 participant in Arm B as not evaluable for RECIST evaluation, but does not otherwise explain how that participant was handled in the primary analysis or provide a full accounting of evaluability.

Several additional outcomes are listed, including biomarker changes, imaging measures, tumor signaling and expression measures, toxicity, and RET gene defects. The registry record does not report results for these measures. It also does not report a detailed statistical analysis plan, a sample-size justification, interim analyses, or the analysis populations used. Those details should not be inferred from the phase designation or the outcome descriptions.

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

Analysis population for the primary outcome, as defined in the registry: 1 patient was not evaluable for RECIST evaluation in Arm B

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 Objective Response Rate of Sorafenib Tosylate in Metastatic Medullary Thyroid Carcinoma in Setting of Inherited Tumor Syndromes as Well as in Setting of Sporadic Medullary Thyroid Cancer, measured over Up to 4 weeks after last dose of sorafenib tosylate.

The primary endpoint is an objective-response count among participants with medullary thyroid carcinoma. The registry record says response was assessed using MRI and determined with Response Evaluation Criteria in Solid Tumors/World Health Organization response criteria. It describes the outcome as a count of participants and gives a timeframe of up to 4 weeks after the last dose of sorafenib tosylate. The record states that a 95% confidence interval would be calculated to estimate the frequency of response.

Objective response is used in cancer studies to summarize whether tumor measurements meet prespecified response criteria. As a binary outcome for each participant, it distinguishes participants who meet the response definition from those who do not. A count of responders is interpretable only alongside the number assessed and the rules for handling participants without an evaluable assessment. The registry record identifies 1 participant in Arm B as not evaluable for RECIST evaluation, but does not report the primary response count, the analysis denominator, or the handling of that participant in the analysis.

A confidence interval communicates the statistical uncertainty around an estimated response frequency. It does not guarantee that the underlying response frequency lies within the interval, and it does not resolve limitations arising from study design, missing assessments, or selection of participants. Although the registry record says an interval would be calculated, neither the interval nor the observed response result is reported in the registry record. The endpoint definition therefore explains what the study intended to measure, not what it found.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does an objective-response count represent?

For a binary endpoint, each participant is classified according to whether the outcome definition is met. The study’s primary outcome is described as a count of participants meeting objective-response criteria. A response count is usually interpreted in relation to the number of participants assessed under the stated criteria. The registry record does not provide that count or the denominator used for the primary result, so no response frequency can be read from the record.

The criteria named in the record combine MRI assessment with Response Evaluation Criteria in Solid Tumors/World Health Organization response criteria. Such criteria make the endpoint more explicit than an unstructured judgment by specifying how tumor response is classified. The registry record does not provide further detail about scan timing, independent review, confirmation rules, or the handling of assessments that could not be interpreted. Those procedures are not reported in the registry record and should not be assumed.

What does a confidence interval add?

The registry record says a 95% confidence interval would be calculated to estimate response frequency. In general, a confidence interval gives a range of values compatible with the observed data under the statistical method used. Its width reflects precision: less precise data generally produce a wider interval, while more precise data generally produce a narrower one. The interval does not remove uncertainty, correct bias, or show whether a result is clinically important. It also cannot be interpreted without knowing the estimate, the analysis denominator, and the method used.

The record does not report the interval or specify the method used to calculate it. Thus, the planned confidence interval can be described, but its result and precision cannot be assessed. A stated plan to calculate an interval should not be treated as evidence that the interval is present in the posted results.

Why does the single-group design shape interpretation?

In a single-group study, outcomes are observed among participants receiving the study intervention without a concurrent control group described in the record. A response count can describe what happened in that group, but there is no randomized control outcome to distinguish the intervention’s contribution from other explanations. For example, the design cannot directly estimate a treatment contrast against a concurrent alternative. This limitation applies even if a response estimate is precise: precision within a group does not create a comparison that the design did not include.

The record distinguishes hereditary and sporadic disease settings as Arm A and Arm B. These are disease categories, not like-for-like treatment assignments. Comparing them as if one were a control group would confound disease setting with the comparison and would not answer the same question as a randomized treatment comparison. The registry record does not report a comparative analysis between these groups.

How should non-evaluable assessments be considered?

The primary-outcome description notes that 1 participant in Arm B was not evaluable for RECIST evaluation. In general, a non-evaluable participant creates uncertainty about how the outcome should be represented. Excluding such a participant can change the denominator; classifying the participant as not responding makes a different assumption; and methods for handling missing outcome data depend on the analysis plan and the reasons for missingness. None of these approaches should be presumed from the phrase “not evaluable.” The registry record does not report how this participant was handled in the primary analysis.

Missing outcome information can affect both the estimated response frequency and its precision. If assessment failures are related to participants’ health, treatment exposure, or outcomes, the missingness may also affect the interpretation of the observed participants. General statistical methods can address some missing-data problems under stated assumptions, but those assumptions require information about why data are absent and how the analysis was conducted. Such information is not reported in the registry record.

What can be learned from the other outcome measures?

The registry record lists measures involving calcitonin, CEA, DCE-MRI exchange rate constant, Ras-MAPK signaling inhibition, VEGF expression, PET SUV Max, toxicity, and RET gene defects. These measures address different biological, imaging, and safety questions. Their statistical interpretation depends on the measurement scale, timing, participant availability, and analysis method. The record does not provide results for these outcomes or enough detail to reconstruct their analyses.

For any outcome involving change from baseline, interpretation requires a clear definition of baseline, follow-up timing, and the handling of participants without paired measurements. For an outcome expressed as a count of participants, the denominator and classification rules are needed. For continuous imaging or laboratory measurements, summary statistics and a method for quantifying uncertainty would be needed. The registry record’s outcome titles alone do not supply those elements.

General statistical interpretation also distinguishes a descriptive result from an inferential one. A descriptive result reports what was observed in the participants analyzed. An inferential result uses a statistical model or procedure to express uncertainty about a broader quantity. The registry record’s stated plan for a confidence interval indicates an intended inferential summary for the primary endpoint, but the resulting estimate and interval are not reported. No additional test statistics or analysis outputs are listed in the registry record.

Finally, the study’s sample size and phase label do not establish the adequacy of a particular analysis. A sample-size justification, targeted response frequency, or planned precision is not reported in the registry record. A phase II designation does not reveal the assumptions used in planning, and enrollment alone cannot establish statistical power. Those design details remain unknown in the record.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Arm A (Hereditary MTC)", "Arm B (Sporadic MTC)")
n      <- c(5, 15)
events <- c(1, 1)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Arm A (Hereditary MTC)  5  1  20.0  0.5  71.6  3.6  62.4  4.3  64.1
ARM  Arm B (Sporadic MTC)  15  1  6.7  0.2  31.9  1.2  29.8  1.6  30.2

11. Multiplicity, Missing Data and Analysis Populations

The registry record lists ten outcome measures overall and identifies a primary endpoint alongside secondary outcomes involving biomarkers, imaging, tumor biology, toxicity, and RET gene defects. When a study assesses multiple outcomes, multiplicity is a general statistical concern: testing or interpreting many outcomes can increase the chance that some findings appear notable by chance. Approaches such as prespecifying a hierarchy, adjusting significance thresholds, or treating secondary results as exploratory can address different aspects of that concern. The registry record does not report a multiplicity strategy or a detailed statistical analysis plan.

The record describes the primary endpoint as an objective-response count and says a 95% confidence interval would be calculated. It does not report a response estimate, interval, or statistical test. Nor does it explain whether secondary outcomes were intended to be confirmatory or exploratory. Therefore, the registry information does not establish how the investigators planned to control or interpret multiplicity.

For missing data, the primary-outcome description identifies 1 participant in Arm B as not evaluable for RECIST evaluation. It does not state whether that participant was excluded, counted as a non-responder, or handled using another approach. The registry record also does not provide a complete accounting of missing data across outcomes or explain reasons for non-evaluability. A missing-data method and analysis population are not reported in the registry record.

Analysis populations can include all enrolled participants, participants who received treatment, or participants with evaluable outcome assessments, among other definitions. These populations answer different questions. The record reports enrollment of 21, but does not identify the analysis population for the primary endpoint or provide denominators for the listed outcomes. The treatment of participants who stopped early, missed assessments, or had unavailable measurements is not reported in the registry record. These omissions limit assessment of how representative the analyzed participants may be of those enrolled.

12. 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: cardiac ischemia/infarction (4.8%), perforation, gi-small bowel (4.8%), infection (4.8%), hypokalemia (4.8%), renal failure (4.8%).

The registry record lists one adverse-event group covering Arm A (Hereditary MTC) and Arm B (Sporadic MTC), with 21 participants at risk for serious adverse events and 5 affected. It also records 21 participants at risk in the other adverse-event category and 21 affected. These are counts in the registry record; they should be read using the record’s event categories rather than as a comparison between treatment groups.

The listed serious-event terms are cardiac ischemia/infarction, perforation of the GI-small bowel, infection, hypokalemia, renal failure, and thrombosis/thrombus/embolism. Each term is accompanied by a reported percentage of 4.8%. These term-level entries describe reported serious events; the record does not provide enough detail to determine whether participants could contribute to more than one term or how event terms were grouped for analysis.

The registry record also lists 1 death. It does not state in this information whether the death was attributed to treatment or provide a cause. Serious adverse events and deaths are important safety observations, but counts alone do not establish causation. Interpretation also depends on definitions, follow-up duration, exposure, event adjudication, and completeness of reporting; those details are not fully reported in the registry record. No control group is available for a direct comparative safety assessment.

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

The record establishes that a phase II, single-group study evaluated sorafenib tosylate in medullary thyroid carcinoma, with hereditary and sporadic disease settings identified. It reports an enrollment of 21 and describes the primary outcome as objective response assessed by MRI using Response Evaluation Criteria in Solid Tumors/World Health Organization response criteria. It also gives a primary-outcome timeframe extending up to 4 weeks after the last dose and states that a 95% confidence interval would estimate response frequency.

What the record does not establish is the observed response result. The number of responders, the analysis denominator, and the confidence interval are not reported in the registry record. The note that 1 participant in Arm B was not evaluable does not fill those gaps, because the handling of that participant and the response status of the remaining participants are not provided. Accordingly, no response rate or degree of precision can be inferred from the registry information.

Even a reported response result from this design would need to be interpreted as an outcome observed in a single group, not as a randomized estimate of benefit relative to another treatment. The hereditary and sporadic groups are disease settings, not a concurrent control comparison. Without a control group, the design does not isolate the effect of the intervention from other factors that could contribute to observed outcomes.

The secondary outcome names and safety counts add context about the measures included in the study. However, the record does not report results for the listed biomarker, imaging, tumor biology, or other outcomes. The safety record lists serious adverse events and a death, but does not establish treatment causation. Statistical significance, clinical importance, and comparative safety cannot be determined from outcome labels and counts alone.

In general, a clinical trial record can define a question and describe how an outcome was intended to be assessed without reporting the result needed to answer that question. Here, the design and endpoint are described, but the primary outcome findings are not reported in the registry record. Any interpretation should preserve that distinction and avoid turning the study’s aims into evidence of an observed effect.

14. Why This Trial Matters Statistically

This study raises a useful statistical distinction between defining an outcome and reporting its result. The registry record specifies an objective-response endpoint, an MRI-based assessment, named response criteria, a follow-up timeframe, and a planned confidence interval. Those details identify the intended measurement and the type of uncertainty summary. They do not reveal the observed response count, the denominator, or the precision of the estimate. For statistical interpretation, a clear endpoint is necessary but is not a substitute for outcome data.

The single-group design also illustrates why within-group outcomes and comparative treatment effects are different quantities. A response count in treated participants can be descriptive, while a treatment contrast requires an appropriate comparison. The hereditary and sporadic disease groups do not supply that comparison simply because they have different labels. The registry record itself cautions that the arms are not a like-for-like control comparison.

The note about 1 participant not evaluable for RECIST highlights another central issue: analysis populations and missing outcome handling affect what a reported count means. Without the denominator and the rule used for non-evaluable assessments, readers cannot determine the precise population represented by a response estimate. Making that distinction explicit is important in small phase II studies, where each participant’s classification can matter to the interpretation of a count.

Finally, the record lists multiple secondary measures and selected safety information, but does not report the corresponding results in the registry information described. This makes it important to distinguish measured outcomes from published findings and to avoid interpreting an outcome list as evidence of a particular direction or magnitude. The study is statistically informative as an example of how design, endpoint specification, missingness, and reporting completeness jointly determine what can be concluded.

15. 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 main limitation for interpreting the study’s findings is that the primary response result is not reported in the registry record. There is no reported responder count, denominator, or confidence interval, even though the record describes a planned interval. This prevents assessment of the observed response frequency and its precision.

The single-group design has no concurrent control group. It can describe outcomes among participants in the study but cannot provide a direct randomized comparison with another treatment. The hereditary and sporadic disease categories are not like-for-like control arms. The record does not describe an adjusted comparison between those categories.

Enrollment is reported as 21, and the primary outcome description notes that 1 participant in Arm B was not evaluable for RECIST evaluation. The record does not explain the handling of this participant, identify the primary analysis population, or provide a complete account of missing outcome assessments. The numbers enrolled in each disease group are also not reported in the registry record.

The registry record does not report a sample-size justification, detailed analysis plan, multiplicity approach, or interim-analysis plan. It also does not provide results for the listed secondary outcomes. These omissions limit assessment of design assumptions and the interpretation of outcomes beyond the primary endpoint.

The safety information is limited to listed counts and event terms. Attribution, timing, event definitions, and comparative safety information are not fully reported in the registry record. The listed death cannot be attributed to treatment based on the information available in the record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Sorafenib Tosylate in Treating Patients With Metastatic, Locally Advanced, or Recurrent Medullary Thyroid Cancer — NCT00390325.
• Registry record last updated 2024-02-06; status last verified 2024-01.
• Results first posted 2018-12-21.
• 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.

19. Record Summary

NCT00390325 is a completed phase II, single-group study sponsored by the National Cancer Institute (NCI). The registry record reports enrollment of 21 and describes sorafenib tosylate taken by mouth twice daily in patients with metastatic, locally advanced, or recurrent medullary thyroid carcinoma. Hereditary and sporadic disease settings are identified, without a like-for-like control comparison.

The primary endpoint is an objective-response count assessed by MRI using Response Evaluation Criteria in Solid Tumors/World Health Organization response criteria, with a stated timeframe of up to 4 weeks after the last dose. The record says a 95% confidence interval would estimate response frequency. It notes that 1 participant in Arm B was not evaluable for RECIST evaluation, but does not report the response count, denominator, or interval.

Additional listed outcomes cover biomarkers, imaging, tumor biology, toxicity, and RET gene defects. The record lists 5 participants with serious adverse events among 21 at risk and 1 death, without establishing treatment causation. The primary outcome result and detailed analysis methods are not reported in the registry record.