Pazopanib Hydrochloride in Treating Patients With Advanced Thyroid Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Pazopanib Hydrochloride in Treating Patients With Advanced Thyroid Cancer (NCT00625846), a non-randomized phase 2 study. The primary outcome is Overall Response Rate (in Cohorts 1-3).

2. Clinical Question

The central statistical question is descriptive: how often was a qualifying response recorded in the specified treated population? This endpoint does not compare pazopanib with an untreated group or another therapy. Because treatment allocation was non-randomized and there was no control arm, the study cannot by itself separate an observed response pattern from differences in patient characteristics, disease type, or other aspects of care. The record describes the intervention and response criteria, but does not report a numerical primary endpoint result or its confidence interval.

3. Trial Design

This was a phase II, non-randomized, unmasked, parallel study. The registry lists enrollment of 152 and describes four groups: Cohort 1 for DTC, Cohort 2 for MTC, Cohort 3 for ATC, and a DTC expansion cohort. The listed primary response endpoint applies to Cohorts 1-3; the expansion cohort is not named in that primary endpoint population.

Each cohort description specifies oral pazopanib hydrochloride at 800 mg once daily on days 1-28, with courses or cycles repeated every 28 days in the absence of disease progression or unacceptable toxicity. The expansion cohort description further specifies confirmed DTC and negative thyroglobulin antibody status. The registry also lists laboratory biomarker analysis as an intervention, but does not report enough detail to characterize its role in the treatment comparison.

The study is marked completed. Its recorded start date is 2008-02-22, and its primary completion date is 2018-12-21. The record identifies the National Cancer Institute (NCI) as sponsor. Non-randomized allocation means that the disease cohorts should not be read as randomized treatment arms or as a controlled comparison. The record does not report a sample-size justification, the planned statistical analysis methods, or whether interim analyses were conducted.

The endpoint descriptions include overall response rate for Cohorts 1-3, toxicity based on grade 3+ adverse events considered possibly, probably, or definitely related to treatment, and progression-free survival at 6 months for Cohorts 1 and 2 and at 3 months for Cohort 3. The record lists 11 outcome measures but does not provide analysis results for them.

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: All patients in Cohort 1, Cohort 2, and Cohort 3 that received treatment and were eligible for response assessment were included in this analysis.

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: All patients that registered and evaluable for any of the endpoints were included in this summary.

6. Primary Endpoint

The registry lists the primary outcome as Overall Response Rate (in Cohorts 1-3), measured over Up to 3 years.

The primary endpoint is overall response rate in Cohorts 1-3, expressed as the percentage of eligible patients who received treatment and were eligible for response assessment. A patient qualifies as a responder by meeting RECIST v1.0 criteria for complete response or partial response at two consecutive assessments at least 8 weeks apart. The registry describes assessment of target lesions by MRI. Complete response means disappearance of all target lesions; partial response means a decrease of at least 30% in the sum of the longest diameters of target lesions. Overall response combines complete and partial responses.

A response endpoint is used to describe the proportion of patients meeting prespecified tumor-change criteria during the stated follow-up period, which is up to 3 years in this record. Requiring confirmation across consecutive assessments makes the endpoint depend on both the extent of change and its confirmation over time. As a general statistical point, a percentage communicates the observed frequency in the analysis population, while a confidence interval conveys uncertainty around that estimate. The registry identifies a 95% confidence interval as the dispersion measure, but the response estimate and its interval are not reported in the registry record.

This endpoint is not a direct measure of comparative benefit: there is no randomized control group in this study. Nor does the response definition, by itself, describe duration of response, symptoms, or quality of life. Those are distinct questions, and their results are not reported in the primary endpoint record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the response percentage represent?

The primary endpoint is a binary outcome at the patient level: each eligible, treated patient in Cohorts 1-3 is classified as having a complete or partial response, or not meeting either response category. The overall response rate is the percentage classified as a responder in the stated analysis population. The registry does not report the number of responders or the analysis denominator, so that percentage cannot be reconstructed from the record.

Why use a confidence interval?

The registry names a 95% confidence interval as the measure of dispersion for the response percentage. As a general statistical explanation, a confidence interval describes the uncertainty associated with an estimate: an interval that is relatively narrow indicates greater precision, while a wider interval indicates less precision. It does not represent a range containing a fixed proportion of individual patient outcomes. The estimate and interval are not reported in the registry record, and the method used to calculate the interval is also not reported there.

How should response criteria be read?

RECIST v1.0 supplies the operational rules for classifying tumor response. In this record, complete response requires disappearance of all target lesions, while partial response requires at least a 30% decrease in the sum of the longest diameters of target lesions. The registry requires the response to appear at two consecutive assessments at least 8 weeks apart. The response rate therefore depends on both the imaging classification and confirmation requirement. A patient who does not satisfy the stated complete- or partial-response criteria is not counted as an overall responder under this endpoint, even if some other change occurred.

What does non-randomized allocation imply?

Allocation is described as non-randomized, with no masking. As a general principle, randomization can help balance measured and unmeasured prognostic characteristics between treatment groups. Here, there is no randomized control group, and the cohorts are defined by cancer type. The cohort outcomes should therefore be read as separate descriptive observations rather than estimates of treatment contrasts. Differences between cohorts cannot be attributed to treatment because cohort membership itself reflects different disease groups.

What can be inferred from planning calculations?

The record's planning illustrations are described as using an observed effect, a two-sided significance level of 0.05, equal allocation, and a normal approximation; the effect considered is a stated fraction of the observed effect. For time-to-event outcomes, the illustration is based on a total number of events, while other illustrations are expressed as participants per arm. These are hypothetical planning quantities, not enrolled sample sizes, trial outcomes, or observed power. They are not the trial's own design justification; that justification is not reported in the registry record.

As a general statistical concept, power is a design property under specified assumptions about an effect, variability, sample size, and analysis. A retrospective calculation based on an observed effect does not establish the probability that the study detected a true effect. It can illustrate how a future study might be planned under stated assumptions, but it cannot replace the missing primary endpoint result or resolve the absence of a randomized comparator.

Why can over-dispersion matter?

As a general explanation, a simple binomial model assumes that each patient's outcome follows the same response probability and that outcomes are independent. If response probabilities vary across patients or outcomes are more variable than the model assumes, over-dispersion can make a simple model understate uncertainty. This is relevant as a general caution when analyzing proportions in heterogeneous clinical populations, but the registry does not report whether over-dispersion was assessed or whether a particular model was used.

The record does not report the statistical tests, model specifications, confidence interval calculation method, covariate adjustments, or handling of repeated assessments. It also does not report a prespecified analysis plan. The response criteria are described, but the analysis procedures beyond those criteria are not reported in the registry 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("Cohort 1 (DTC)", "Cohort 2 (MTC)", "Cohort 3 (ATC)")
n      <- c(37, 35, 15)
events <- c(18, 5, 0)
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  Cohort 1 (DTC)  37  18  48.6  31.9  65.6  33.4  64.1  33.4  64.2
ARM  Cohort 2 (MTC)  35  5  14.3  4.8  30.3  6.3  29.4  6.4  29.5
ARM  Cohort 3 (ATC)  15  0  0.0  0.0  21.8  0.0  20.4  0.2  20.6

11. Multiplicity, Missing Data and Analysis Populations

The registry lists 11 outcome measures, including the primary overall response endpoint, a toxicity measure based on grade 3+ adverse events considered possibly, probably, or definitely related to treatment, and progression-free survival endpoints for specified cohorts. It does not report analysis results for these endpoints. The record does not report whether the analyses were arranged hierarchically, whether multiplicity adjustments were planned, or how any such adjustment would be applied. Without that information and without the numerical results, the chance of false-positive findings across multiple outcomes cannot be assessed from the record.

The stated primary response population includes eligible patients in Cohorts 1-3 who received treatment and were eligible for response assessment. That definition is important because enrollment and analysis populations need not be identical. The number of people meeting that definition is not reported in the registry record. The DTC expansion cohort is described separately and is not included in the primary endpoint population as written.

Missing-data rules are not reported in the registry record. In particular, the record does not say how patients without adequate response assessments, patients who discontinued treatment, or assessments missing at particular time points were treated in the response analysis. It also does not report whether an intention-to-treat analysis or another population definition was used beyond the stated response-assessable eligible treated group.

These omissions limit reproducibility and interpretation. For a percentage endpoint, the numerator, denominator, eligibility rules, and handling of unassessed patients all contribute to the meaning of the estimate. Because the primary result and these analysis details are not reported, the response percentage and its sensitivity to missing data cannot be checked.

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 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: weight loss (23.7%), neutrophil count decreased (13.3%), pharyngolaryngeal pain (13.3%), hypertension (9.8%), depression (8.6%).

The registry gives serious adverse event summaries separately for Cohort 1 (DTC), Cohort 2 (MTC), Cohort 3 (ATC), and the DTC expansion cohort. In those summaries, 21 of 38 were reported as affected in Cohort 1, 18 of 35 in Cohort 2, 9 of 15 in Cohort 3, and 31 of 61 in the expansion cohort. These are cohort-level serious adverse event counts, not a randomized comparison. The registry also reports deaths of 3, 0, 1, and 4, respectively. The record does not establish from these figures whether deaths were caused by treatment.

Listed serious adverse event terms include weight loss at 23.7%, neutrophil count decreased at 13.3%, pharyngolaryngeal pain at 13.3%, hypertension at 9.8%, depression at 8.6%, and dyspnea at 8.6%. The registry excerpt does not specify the denominators for these term percentages, so they should not be converted into counts. These term summaries are distinct from the primary response result and from the separately defined toxicity endpoint concerning grade 3+ adverse events considered possibly, probably, or definitely related to treatment. The registry does not report a comparative safety analysis.

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

The study record establishes that pazopanib was evaluated in a phase II, non-randomized study involving DTC, MTC, and ATC cohorts, with a separately described DTC expansion cohort. It also specifies the daily regimen and defines the response endpoint using RECIST v1.0, including a confirmation requirement at consecutive assessments. Those details clarify what the trial was designed to observe.

They do not establish a numerical response rate. The primary endpoint result and its 95% confidence interval are not reported in the registry record. Thus, the proportion of patients meeting the response definition and the precision of that estimate cannot be stated. The listed enrollment of 152 does not supply the response denominator, because the primary analysis population is limited to eligible treated patients in Cohorts 1-3 who were eligible for response assessment.

The design also does not support a randomized treatment comparison. Since cohort assignment was not randomized and the cohorts represent different cancer types, differences in their outcomes would not by themselves measure a treatment effect. No conclusion about the effect of pazopanib relative to another treatment or to no treatment can be drawn from the described design alone.

The serious adverse event summaries and listed terms describe safety reporting in the registry, but they do not establish causation or replace a full comparative safety analysis. The registry record does not report the handling of missing assessments, a complete statistical analysis plan, or the numerical primary endpoint findings. The defensible interpretation is therefore limited to the study structure, stated endpoint definition, and posted safety counts and terms; the efficacy result remains unreported in the registry record.

14. Why This Trial Matters Statistically

This study illustrates why endpoint definitions and analysis populations matter as much as a trial's headline question. Overall response rate is not simply any recorded tumor change: the registry specifies RECIST v1.0 categories, an imaging assessment, and confirmation at two consecutive assessments at least 8 weeks apart. Those rules determine who can be counted as a responder.

It also illustrates the interpretive limits of a non-randomized, parallel cohort design. DTC, MTC, and ATC cohorts are clinically distinct groups, rather than randomized alternatives. Their results would need to be interpreted separately and descriptively; a difference between them would not demonstrate that treatment performed differently across groups. The absence of a control group also prevents a direct estimate of comparative treatment effect.

Finally, the registry record distinguishes planning illustrations from trial outcomes. Planning calculations based on an observed effect and stated assumptions can help frame future study size, but they are not observed power and cannot substitute for a reported endpoint estimate. Here, the primary response percentage and its confidence interval are not reported in the registry record. That absence is statistically consequential: without the estimate, denominator, interval, and missing-data rules, readers cannot assess either the observed response frequency or its precision. Clear reporting of these elements is essential for interpreting a clinical study responsibly.

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.

A central limitation of the registry record is the absence of a numerical primary response result and confidence interval. Although the response definition and analysis population are described, the responder count and denominator are not reported. The registry lists enrollment of 152, but does not establish that this is the number included in the response analysis.

The non-randomized design and lack of a control group limit causal interpretation. The disease-specific cohorts are not like-for-like comparison arms, so comparisons across them could reflect differences in the underlying cancer groups. The record does not describe baseline characteristics in a way that would allow an assessment of comparability.

The record does not report how missing or unevaluable response assessments were handled, whether all participants were followed through the stated timeframe, or how treatment discontinuation affected endpoint classification. The statistical analysis methods for the primary endpoint, including the confidence interval method and any covariate adjustment, are not reported. Multiplicity procedures for the listed outcomes are also not reported.

Safety interpretation has additional limits. The serious adverse event summaries provide cohort counts and deaths, but the record does not establish causality for deaths or provide a randomized comparator for safety. Denominators for the listed serious adverse event term percentages are not stated in the excerpt. The available information therefore supports a limited description of reported safety items, not a comparative estimate of risk.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Pazopanib Hydrochloride in Treating Patients With Advanced Thyroid Cancer — NCT00625846.
• Registry record last updated 2020-02-26; status last verified 2020-02.
• Results first posted 2020-02-26.
• 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

The registry describes a completed phase II, non-randomized, unmasked, parallel study of pazopanib hydrochloride in advanced thyroid cancer. Cohorts included DTC, MTC, and ATC, plus a DTC expansion cohort. The primary endpoint was overall response in treated, eligible, response-assessable patients in Cohorts 1-3, using confirmed RECIST v1.0 complete or partial response. The prescribed regimen was 800 mg by mouth once daily on days 1-28, with 28-day courses or cycles continuing while disease had not progressed and toxicity remained acceptable.

The registry lists enrollment of 152 and 11 outcome measures. It reports selected serious adverse event counts, deaths, and serious adverse event terms. A numerical primary response result, its confidence interval, detailed statistical methods, missing-data procedures, and multiplicity handling are not reported in the registry record. The non-randomized design does not provide a controlled treatment comparison.