Cabozantinib in Advanced Pancreatic Neuroendocrine and Carcinoid Tumors: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Cabozantinib in Advanced Pancreatic Neuroendocrine and Carcinoid Tumors (NCT01466036), a non-randomized phase 2 study. The primary outcome is Objective Response Rate (ORR).

2. Clinical Question

The study addresses whether tumor responses were observed in these clinical groups during treatment. Objective response rate (ORR) focuses on documented tumor shrinkage that meets specified imaging criteria; it does not, by itself, describe how long a response lasted or whether patients lived longer. The registry lists progression-free survival and overall survival as secondary outcomes, but their results are not reported in the registry record.

The population is described broadly as patients with pancreatic neuroendocrine and carcinoid tumors. The registry does not report further eligibility characteristics in the record. Accordingly, the clinical question is limited to the groups and response definition explicitly described, rather than a broader claim about all patients with these diagnoses.

3. Trial Design

This was an open-label, phase II study of cabozantinib, with nonrandomized allocation and a parallel model. The registry lists enrollment as 61 and status as completed. The study began in 2012-07, with primary completion in 2021-08. These dates describe the registry record; they do not establish the duration of follow-up for any individual participant.

The study included a metastatic or unresectable PNET cohort and a metastatic or unresectable carcinoid cohort. The arm descriptions call for an anticipated 35 patients in each cohort. The registry does not report the final number enrolled in each cohort in the registry record for this account, so the overall enrollment should not be interpreted as the actual cohort sizes.

All participants received cabozantinib. Because allocation was nonrandomized and the cohorts represent different tumor types, they are not interchangeable treatment and control groups. Differences between cohort results, if reported, could not be interpreted as a randomized estimate of a treatment effect. The open-label design also means that masking was absent; the registry lists masking as none.

Imaging was scheduled at cycles 2, 4, and 6 every 8 weeks for the first 24 weeks, then every 3rd cycle every 12 weeks until progression or end of treatment. A cycle was 28 days. Median treatment duration was 8 cycles for the carcinoid cohort and 12.5 cycles for the PNET cohort. These treatment-duration summaries do not substitute for the unreported response results or establish how long any individual response lasted.

The primary outcome was objective response rate. Progression-free survival and overall survival were listed as secondary outcomes. The registry record does not report a comparator regimen, randomization procedure, allocation concealment, or a detailed statistical analysis plan.

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: Response rate by RECIST 1.1 was calculated by including all patients who started study therapy.

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 (ORR), measured over Imaging was performed cycles 2,4,6 every 8 weeks for the first 24 weeks, then every 3rd cycle every 12 weeks until progression or EOT. Median treatment duration was 8 cycles for the Carcinoid cohort and 12.5 cycles for the PNET cohort. Cycle=28 days..

The primary endpoint was objective response rate (ORR), assessed according to RECIST 1.1. The registry describes disease assessment using CT and/or MRI scans. Under the stated criteria, complete response means disappearance of all target lesions. Partial response requires a 30% or more decrease in the sum of the longest diameter of target lesions. Stable disease is a decrease of less than 30% and no more than a 20% increase in that sum. ORR consists of complete responses plus partial responses.

In general, an imaging-based response endpoint offers a defined way to summarize tumor shrinkage during treatment. It is useful for describing antitumor activity because it applies explicit criteria to measured lesions rather than relying only on a broad clinical impression. The definition also separates response from stable disease: stable disease is not counted as an objective response under the stated ORR definition.

The response population was all patients who started study therapy. The endpoint is a participant count, rather than a continuous measure, according to the registry outcome description. However, the posted results available in the registry record do not give the number of complete responses, partial responses, or participants included in the reported ORR calculation. Thus, the endpoint definition can be explained, but the observed response rate and its precision cannot be evaluated from the reported result information.

ORR does not capture the duration of response, progression-free survival, overall survival, symptoms, or quality of life. Those are distinct outcomes. The registry lists progression-free survival and overall survival as secondary outcomes, but does not report their results in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry record describes the primary endpoint and some design features, but it does not report a full statistical analysis plan or inferential results. The following concepts explain how the stated methods are generally used; they should not be mistaken for analyses confirmed to have been conducted in this study.

What does an objective response rate measure?

In general, an objective response rate is the proportion of a defined analysis population whose tumors meet prespecified criteria for complete or partial response. The registry uses RECIST 1.1 and defines ORR as complete response plus partial response. Stable disease is a separate category and is not included in that endpoint. This makes the endpoint a binary outcome for each participant: a participant either meets the stated response definition or does not.

The registry says the response rate was calculated among all patients who started study therapy. That is the stated population rule. The record does not report the number who started therapy in each cohort, how missing or unevaluable imaging was handled, or whether any additional analysis populations were used. Those details affect the denominator and interpretation of a response estimate, so they should not be inferred.

How would uncertainty in a response estimate be described?

In general, a confidence interval gives a range of values reasonably compatible with the observed data under a specified statistical method. For a binary endpoint, the interval helps readers judge how precisely the response proportion has been estimated. A narrow interval indicates greater precision under that method; a wide interval indicates more uncertainty. The registry record does not report an ORR estimate or confidence interval, so no precision assessment is possible for the primary endpoint.

In general, a p-value depends on a specified null hypothesis and statistical model. It does not measure the probability that a treatment works, nor does it describe the size or clinical importance of an effect. No primary-outcome p-value or test is reported in the registry record. A p-value should not be derived from the design description or safety counts.

What does the design imply for comparisons?

In general, random allocation can help balance measured and unmeasured characteristics between groups, supporting a causal treatment comparison. This study is listed as nonrandomized, and both cohorts received cabozantinib. The PNET and carcinoid cohorts are distinct disease groups, not randomized alternatives. Comparing their outcomes would therefore not estimate the effect of cabozantinib relative to a control treatment.

The study is also open-label, with no masking. In general, lack of masking can influence decisions or assessments that allow judgment, although imaging criteria provide a structured basis for classifying response. The registry does not report whether image review was centralized or blinded, so no further claim about assessment procedures is warranted.

How should the planning illustration be understood?

The registry describes a planning illustration based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The intended effect to detect is expressed as a stated fraction of the observed effect. For time-to-event outcomes, the illustration is based on the total number of events; for other outcomes, it is expressed as participants per arm.

In general, such calculations explore the sample size associated with chosen assumptions. They are planning illustrations, not observed power and not evidence that the original study was designed using those assumptions. The registry record does not report the trial’s own sample-size justification or a prospective statistical design based on this illustration. No numerical sample-size illustration is stated in the record information, so no sample-size value can be interpreted.

Why can over-dispersion matter?

In general, over-dispersion occurs when observed variation in count data exceeds what a chosen statistical model expects. If a model assumes less variability than the data contain, standard errors can be too small and uncertainty can be understated. The primary outcome is described as a participant count for a binary response classification, but the registry does not report a model for the primary outcome or any adjustment for over-dispersion. This concept therefore provides general statistical context, not a claim about the study’s analysis.

What is reported about secondary outcomes?

Progression-free survival and overall survival are listed as secondary outcomes. In general, time-to-event outcomes account for the timing of an event and may involve participants whose event has not been observed by the end of follow-up. Their analysis depends on event definitions, follow-up, censoring rules, and the statistical method. Those details and the secondary-outcome results are not reported in the registry record. The record lists 0 analyses, which means there are no registered analysis results to interpret in this account; it does not establish whether investigators conducted analyses outside the registry.

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("Metastatic or Unresectable PNET", "Metastatic or Unresectable Carcinoid")
n      <- c(20, 41)
events <- c(3, 7)
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  Metastatic or Unresectable PNET  20  3  15.0  3.2  37.9  5.2  36.0  5.4  36.3
ARM  Metastatic or Unresectable Carcinoid  41  7  17.1  7.2  32.1  8.5  31.3  8.6  31.4

11. Multiplicity, Missing Data and Analysis Populations

The registry lists objective response rate as the primary outcome and progression-free survival and overall survival as secondary outcomes. It does not report how testing across these outcomes was handled. In general, multiplicity arises when several hypotheses or endpoints are examined: the chance of at least one apparently positive result can increase if no adjustment or prespecified testing strategy is used. Whether any multiplicity adjustment was planned or performed is not reported in the registry record.

The record does not provide a statistical analysis plan, a testing hierarchy, or details of interim analyses. Those methods should not be inferred from the phase II designation or from the presence of multiple outcomes. The registry lists 0 analyses, and no inferential results are available in the registry record for the primary or secondary outcomes.

The stated response population includes all patients who started study therapy. However, the record does not describe how missing scans, incomplete assessments, withdrawal, or unevaluable response classifications were handled. It also does not report the number of participants who started treatment in each cohort or the denominator used for any outcome calculation. No imputation or sensitivity-analysis approach is reported.

In general, missing outcome data can affect an estimate when the availability of an assessment is related to a participant’s health or response. The direction and extent of any such effect depend on why data are missing and on the analysis method. Since the registry record does not report missingness or its handling, the effect of missing assessments cannot be assessed.

The planned enrollment descriptions are anticipated cohort targets; the record gives overall enrollment as 61 but does not state final cohort-specific counts in the information. The registry also does not report analysis populations beyond the stated response population. Consequently, no conclusion can be drawn about whether all enrolled participants contributed to each outcome analysis.

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: hypertension (19.5%), diarrhea (15.0%), hypophosphatemia (15.0%), lipase increased (9.8%), lymphocyte count decreased (9.8%).

The safety information is reported separately for the metastatic or unresectable PNET cohort and the metastatic or unresectable carcinoid cohort. For serious adverse events, the record lists 15 affected participants among 20 at risk in the PNET cohort, and 24 affected participants among 41 at risk in the carcinoid cohort. For other adverse events, it lists 20 affected among 20 at risk in the PNET cohort and 41 affected among 41 at risk in the carcinoid cohort. Deaths are recorded as 1 and 3, respectively.

These counts describe what the registry reports; they do not establish that cabozantinib caused an event or death. The registry record does not report event attribution, timing, duration, recurrence, or the criteria used to define seriousness. The listed deaths should not be interpreted as treatment-related without attribution information.

The record lists hypertension at 19.5%, diarrhea at 15.0%, hypophosphatemia at 15.0%, lipase increased at 9.8%, lymphocyte count decreased at 9.8%, and fatigue at 7.3% among selected serious-event terms. The record does not state the denominator for these percentages in the information. They are not a complete account of all possible adverse events, and they should not be directly compared with the cohort-level affected-participant counts without additional definitions.

In general, safety counts are most informative when considered alongside event definitions, severity, attribution, exposure time, and the method used to collect events. Those details are not reported in the registry record, so these figures support a limited descriptive reading only.

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

The registry record documents a phase II, open-label study of cabozantinib in separate cohorts with advanced pancreatic neuroendocrine or carcinoid tumors. It specifies objective response rate as the primary endpoint and describes how RECIST 1.1 response categories were defined. It also provides treatment-assessment timing and selected safety counts.

What the posted information does not provide is the primary outcome result. There is no reported response count, ORR estimate, confidence interval, or p-value. As a result, the record does not permit a quantitative judgment about how frequently complete or partial responses occurred, or how precisely any response estimate was measured. The registry also lists progression-free survival and overall survival but does not report their results.

The study has no randomized control group. Both cohorts received cabozantinib, and they represent different tumor types. The design therefore cannot establish a treatment effect relative to another therapy or no treatment. Any comparison between these cohorts would also be difficult to interpret because the groups differ by disease category and were not formed through random assignment.

The safety counts and percentages are descriptive. They do not establish that treatment caused the reported events, and the registry does not give attribution or detailed event definitions in the information. They also cannot replace a complete assessment of tolerability or comparative safety.

In general, a trial’s endpoint definition, observed result, uncertainty, and design all contribute to interpretation. Here, the endpoint definition and broad design are available, but the response and survival estimates are not reported in the registry record. The conclusions should therefore remain limited to what the registry describes: a study intended to evaluate response and other outcomes, with the specified imaging approach and reported safety observations. No claim that cabozantinib works or does not work follows from the missing outcome results.

14. Why This Trial Matters Statistically

This trial illustrates why an endpoint definition is not the same as an endpoint result. The registry explains that ORR comprises complete and partial responses under RECIST 1.1 and describes the imaging schedule. These details make the intended measurement understandable, but they cannot tell readers the observed response frequency or its uncertainty. Without a reported result, the statistical question remains unanswered in the registry record.

It also demonstrates how design determines the kinds of conclusions available. Nonrandomized parallel cohorts receiving the same intervention can describe outcomes within different disease groups, but they do not provide a randomized control comparison. Cohort distinctions should not be mistaken for treatment allocation, and outcome differences between distinct tumor types would not by themselves quantify a drug effect.

The record also separates efficacy outcomes from safety observations. Response, progression-free survival, and overall survival address different questions, while adverse-event counts describe reported harms. Each requires its own definitions, analysis population, and interpretation. The registry includes selected safety figures but does not provide a complete inferential analysis or the detail needed to attribute events.

Finally, the case underscores the value of reporting denominators, missing-data handling, analysis methods, and uncertainty alongside outcome estimates. The response population is described, but cohort-specific treatment-start counts and response results are absent. Statistical interpretation depends not only on what was intended to be measured, but also on what was observed and how completely the analysis is documented.

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 principal limitation for interpreting efficacy is that the registry record does not report the primary ORR result. Response counts, the estimate, its denominator, and a confidence interval are absent. Although the response population is described as all patients who started study therapy, the number who started therapy in each cohort is not reported in the information.

The study is nonrandomized and has no control group. Both cohorts received cabozantinib, and the cohorts represent different tumor types. Consequently, the design does not support a randomized comparison of treatment against a control or a causal interpretation of differences between cohorts.

The registry describes imaging assessment using CT and/or MRI and gives the timing of assessments, but it does not report details such as independent image review, handling of unevaluable scans, or missing-assessment methods. The record also does not report how response classifications were adjudicated or whether analyses beyond the stated response population were used.

Progression-free survival and overall survival are listed as secondary outcomes, but outcome definitions, analysis methods, follow-up summaries, and results are not reported in the registry record. The statistical analysis plan, sample-size justification, interim analyses, and multiplicity procedures are also not reported. A planning illustration is described, but its assumptions should not be treated as the original trial design or as observed power.

Safety interpretation is limited by the absence of event attribution, timing, duration, and detailed seriousness definitions in the information. The denominator for the listed serious-event percentages is not reported. The overall enrollment of 61 does not identify final cohort sizes, and the anticipated 35 participants per cohort are targets rather than confirmed cohort counts.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Cabozantinib in Advanced Pancreatic Neuroendocrine and Carcinoid Tumors — NCT01466036.
• Registry record last updated 2024-03-05; status last verified 2024-03.
• Results first posted 2023-07-20.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

19. Record Summary

The record describes an open-label, nonrandomized phase II study of cabozantinib in metastatic or unresectable PNET and carcinoid tumor cohorts. Overall enrollment was 61, and the study status is completed. The primary endpoint was ORR under RECIST 1.1, defined as complete plus partial response among patients who started study therapy. Imaging was scheduled during treatment, and progression-free survival and overall survival were listed as secondary outcomes.

The registry record does not report the ORR result, survival results, confidence intervals, or inferential analyses. It reports serious and other adverse-event counts by cohort, deaths, and selected serious-event percentages, but does not provide the detail needed to establish causation. Because the study was nonrandomized and had no control group, it does not provide a like-for-like treatment comparison. The available record supports a description of the design and measurement plan, but not a quantitative conclusion about efficacy.