Erlotinib, Celecoxib and Reirradiation for Recurrent Head and Neck Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Erlotinib, Celecoxib and Reirradiation for Recurrent Head and Neck Cancer (NCT00970502), a na phase 1, phase 2 study. The primary outcome is Toxicity.

2. Clinical Question

3. Trial Design

The registry describes a phase I/II trial of erlotinib and celecoxib with radiation therapy for poor-prognosis head and neck cancer, including reirradiation. The study is marked completed and lists enrollment of 15 participants. Its start is recorded as 2007-02 and its primary completion as 2010-11. These dates describe the record and do not, by themselves, specify when each participant began or completed treatment.

The design is single-group, with allocation recorded as not applicable and masking recorded as none. The listed intervention is erlotinib plus celecoxib; the official title describes use with radiation therapy. No separate comparison arm is described, and the registry explicitly notes that the study is not a randomized comparison. As a result, any outcome description is about participants in this treatment group, not a randomized estimate of the effect relative to a control regimen.

The primary outcome is toxicity, defined in the record as the number of participants with acute and late toxicity, with a timeframe of 30 days. The secondary outcomes include clinical response, locoregional progression, locoregional control, progression-free survival, overall survival and late toxicity. The record lists four outcome measures but does not report analyses for them. The doses, dose-escalation rules, treatment sequence, radiation details, eligibility criteria and endpoint assessment procedures are not reported in the registry record.

The sponsor is listed as Johnny Kao. The record's description explains the clinical context as recurrent head and neck cancer after previous radiation, where treatment options may be limited and reirradiation may involve toxicity. Those statements provide the study rationale, not outcome evidence from this trial. The record does not report a sample-size justification, so the enrollment number should not be interpreted as evidence that a particular statistical precision or power was achieved.

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: Erlotinib and celecoxib were administered orally

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: 15 subjects consented, one withdrew before receiving study treatment and was excluded from analysis

6. Primary Endpoint

The registry lists the primary outcome as Toxicity, measured over 30 DAYS.

The primary endpoint is recorded as toxicity, measured by the number of participants with acute and late toxicity. This is a count-based, binary outcome at the participant level: a participant is counted as having an event if they meet the study's toxicity definition, rather than contributing a continuous measurement. The record states that erlotinib and celecoxib were administered orally and gives a timeframe of 30 days.

Toxicity is relevant in a trial that combines drug treatment with radiation, particularly when the study rationale emphasizes the potential for treatment-related complications. Recording how many participants experience toxicity can describe the observed safety burden in the treated group. It does not, without a comparison group and clearly specified ascertainment rules, establish whether toxicity is more or less common than with another regimen. Nor does a participant count alone convey the severity, duration or clinical course of each event.

The record describes the endpoint as covering acute and late toxicity, but the relationship between that wording and the 30-day timeframe is not explained. The definitions of acute and late toxicity, grading criteria, assessment schedule, and rules for counting repeated or multiple events are not reported in the registry record. No primary toxicity result is reported. Accordingly, the endpoint can be described from its registry definition, but its observed value and the precision of any estimate cannot be assessed from the record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the single-group design allow?

A single-group study observes outcomes in participants receiving the study regimen. It can describe what was recorded in that group, such as participant counts with toxicity or the occurrence of listed adverse-event terms. It does not create a concurrent control group against which to estimate a randomized treatment difference. Comparisons with outcomes from other settings would require additional information and assumptions not described in the registry record.

Because the design is not randomized, differences between this group and another group cannot be attributed to treatment on the basis of this record alone. Participants may differ in disease characteristics, prior treatment or other factors, and the registry does not provide enough detail to assess or adjust for such differences. The listed allocation of not applicable and the single-group model are consistent with an observational description of outcomes within the treated group, not a randomized comparison.

How should a participant count be understood?

The primary outcome is defined as the number of participants with acute and late toxicity. A participant-level count is useful for conveying how many people met an outcome definition, but it depends on how the outcome is defined and assessed. For example, a participant may experience multiple events, while a participant-level endpoint may count that person once. The registry does not explain its counting rules, toxicity grades, or how repeated events were handled.

For a binary outcome, a proportion can describe the observed share of participants who experienced an event. In general, estimates based on small groups can be uncertain, and a confidence interval helps show that uncertainty. No primary estimate or confidence interval is reported in the registry record, so neither the observed toxicity proportion nor its precision can be assessed from the posted information. No new proportion or event count should be inferred from the listed fields or percentages.

Why does the lack of a comparator matter?

A treatment effect is a contrast between outcomes under different treatment strategies. In a randomized trial, random allocation is intended to make the groups comparable in expectation, allowing a treatment contrast to be interpreted under the trial's assumptions. Here, there is no randomized comparison arm. The safety observations therefore describe the treated group but do not estimate how its outcomes differ from a control regimen.

The regimen includes more than one component in a radiation context. Without groups that vary the treatment components, the registry cannot attribute an observed event or outcome separately to erlotinib, celecoxib or radiation. This is not a flaw in every single-group study; it is a limit on the questions that its design can answer. The design may describe feasibility or observed toxicity, but the registry record does not report a formal analysis addressing those questions.

What can the planning assumptions tell readers?

The record describes planning illustrations based on an observed effect, using a two-sided significance level of 0.05, equal allocation between two arms and a normal approximation; the target effect is stated as a fraction of the observed effect. Such calculations are hypothetical planning exercises. They do not represent the trial's observed power, and they do not change the fact that this study is recorded as single-group.

In general, sample-size calculations depend on assumptions about the outcome, effect to detect, variability, allocation and analysis method. A planning estimate is only as informative as those assumptions, and it is not a substitute for a trial's design justification. The sample-size justification and a statistical analysis plan are not reported in the registry record. The enrollment of 15 participants is the recorded enrollment, not proof that a prespecified target was reached or that a particular level of precision was obtained.

How should uncertainty and analysis be assessed?

In general, an estimate describes the value calculated from the observed data, while uncertainty measures express how sensitive that estimate may be to sampling variation under a statistical model. A p-value, if reported, relates the observed data to a specified null model; it does not measure clinical importance or prove that a treatment caused an outcome. The registry record reports no analyses, no p-values and no confidence intervals for the listed outcomes. No significance claim can therefore be assessed from this record.

Methods for time-to-event outcomes such as progression-free or overall survival generally require information about follow-up, event definitions and censoring. Those details and outcome analyses are not reported in the registry record. Similarly, interpreting toxicity requires definitions, assessment procedures and handling of missing observations. The record does not report those details, interim analyses, the analysis populations or how missing data were handled. Statistical interpretation must remain descriptive and limited to the material actually recorded.

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("Celecoxib 200mg", "Celecoxib 400mg", "Celecoxib 600mg")
n      <- c(3, 8, 3)
events <- c(0, 1, 2)
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  Celecoxib 200mg  3  0  0.0  0.0  70.8  0.0  56.1  0.6  60.2
ARM  Celecoxib 400mg  8  1  12.5  0.3  52.7  2.2  47.1  2.8  48.2
ARM  Celecoxib 600mg  3  2  66.7  9.4  99.2  20.8  93.9  19.4  93.2

11. Multiplicity, Missing Data and Analysis Populations

The registry lists several secondary outcomes: clinical response, locoregional progression, locoregional control, progression-free survival, overall survival and late toxicity. Considering multiple outcomes raises the general issue of multiplicity: the more hypotheses or endpoints examined, the greater the chance that an apparently notable finding may arise by chance if results are interpreted without a prespecified strategy. A multiplicity adjustment or hierarchy of outcomes may address that issue in some analyses, but such a strategy is not reported in the registry record.

The primary endpoint is named as toxicity, but the record does not provide a statistical analysis plan or report analyses for the primary or secondary outcomes. Interim analyses, stopping rules and any procedure for selecting which outcomes to emphasize are not reported in the registry record. Consequently, it is not possible to determine how many comparisons were planned, whether the primary endpoint was analyzed as specified, or how multiple outcomes were handled.

Missing-data methods are also not reported in the registry record. The registry does not describe the analysis populations, participant-level exclusions, completeness of follow-up, or how missing outcome assessments were treated. These details matter because an analysis limited to participants with available assessments may differ from one that accounts for all enrolled participants. The enrollment count alone does not establish how many participants contributed to each outcome analysis.

Without reported outcome analyses or missing-data procedures, neither the robustness of the findings nor the risk of selective reporting can be evaluated. This does not establish that data were missing or that analyses were changed; those matters are simply not reported in the registry record.

12. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: mucositis (21.4%), pain (21.4%), fatigue (14.3%), pharyngocutaneous fistula (7.1%), folliculitis (7.1%).

The registry lists 14 participants at risk and 14 affected for both serious adverse events and other adverse events in the erlotinib-plus-celecoxib group. These are registry safety fields, not a narrative account of each participant's experience. The record also lists serious adverse-event terms with percentages: mucositis and Pain at 21.4% each, Fatigue at 14.3%, and pharyngocutaneous fistula, folliculitis and nausea at 7.1% each.

The term-level percentages should not be added together or interpreted as distinct groups of participants. The registry does not report the corresponding event counts, whether participants experienced more than one term, the event grades, duration, timing, or attribution to a particular component of treatment. It also does not describe how the terms were selected for listing. The denominator for each term is not stated alongside the individual percentages in the record.

These safety figures describe what the registry records; they do not establish a comparative safety profile because no control group is listed. Nor do they provide the complete information needed to evaluate the clinical course or severity of each event. No death count is reported in the safety fields, so the record does not support a conclusion about mortality.

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

The registry record establishes that a completed phase I/II, single-group study enrolled 15 participants and evaluated erlotinib plus celecoxib in a radiation-therapy context. It identifies toxicity as the primary outcome and lists several secondary outcomes. The record also includes safety fields for 14 participants and selected serious adverse-event terms with percentages. These registry entries describe the study and the safety information recorded; they do not provide a reported primary toxicity result or analyses of the listed secondary outcomes.

The design cannot answer whether the regimen produces better or worse outcomes than another treatment, because no randomized comparator is described. It also cannot isolate the effects of individual treatment components. The registry's rationale discusses the need for less toxic and more effective approaches and cites background claims, but rationale is not a finding from this trial. The available record does not establish that the regimen was effective, ineffective, safer or more toxic than an alternative.

Even a complete count of events in a treated group would need to be interpreted in light of the endpoint definitions, assessment methods, follow-up and uncertainty. Those details are not reported in the registry record. The available fields also do not explain the relation between the 30-day timeframe and the inclusion of both acute and late toxicity.

The appropriate conclusion is therefore limited: this registry record describes a small single-group study and reports selected safety information, but it does not provide enough outcome analysis to quantify the primary endpoint or support a comparative treatment conclusion. The absence of reported outcome analyses should not be read as evidence that no outcomes occurred.

14. Why This Trial Matters Statistically

This trial illustrates how design determines the scope of statistical conclusions. A single-group study can document treatment exposure and describe observed outcomes, but it does not estimate a randomized contrast against usual care, another regimen or a control group. That distinction is central when interpreting claims about toxicity or clinical response. An observed event count in treated participants is not, by itself, a measure of excess risk attributable to treatment.

The endpoint also highlights the importance of precise definitions. “Acute and late toxicity” combines categories that may involve different assessment periods or event patterns, while the record gives a timeframe of 30 days without further explanation. Without assessment rules, grading criteria and event-counting procedures, even a reported count could be difficult to interpret or compare across studies.

The safety record gives selected terms and percentages but does not provide event counts by term or clarify overlap between terms. Statistical reading therefore requires restraint: the listed percentages do not form a complete safety profile, and their interpretation depends on denominators and definitions that are not described. The record's enrollment count and safety fields also do not supply a precision estimate.

Finally, the registry names several secondary outcomes but reports no analyses. This makes the distinction between prespecified questions, collected data and reported results especially important. A rationale can explain why a study was undertaken, but statistical conclusions require outcome data and a transparent analysis. The record is useful for identifying the study's design and intended questions; its limitations show why registration details alone may not support conclusions about treatment effects.

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 registry record provides limited detail about who was eligible, how participants were treated, and how outcomes were assessed. Although the title describes induction and concomitant erlotinib and celecoxib with radiation therapy, including reirradiation, the specific doses, schedule, dose-escalation rules and radiation regimen are not reported in the registry record. Detailed eligibility criteria and participant characteristics are also not reported, limiting assessment of how broadly the findings might apply.

The primary endpoint's wording and timing are not fully resolved: it includes acute and late toxicity while listing a timeframe of 30 days. Toxicity definitions, grading, ascertainment schedule, event attribution and rules for repeated events are not reported. The registry's safety fields and listed terms cannot fill those gaps. The term-level denominators and event counts are not reported, and the handling of multiple events within a participant is unclear.

There is no reported primary endpoint result, and the record reports no analyses for the listed outcomes. It does not give confidence intervals, p-values, follow-up duration for secondary outcomes, or analysis populations. The methods for handling missing observations, censoring, multiplicity or interim analyses are not reported in the registry record. The sample-size justification and detailed statistical plan are likewise not reported.

The single-group design is an additional constraint for comparative inference. Outcomes in this treated group cannot establish how the regimen compares with a different treatment, and the contributions of the drug and radiation components cannot be separated. The listed enrollment of 15 participants describes the study size but does not establish statistical precision or power.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Erlotinib, Celecoxib and Reirradiation for Recurrent Head and Neck Cancer — NCT00970502.
• Registry record last updated 2017-04-10; status last verified 2017-02.
• Results first posted 2017-04-10.
• 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 I/II single-group study of erlotinib plus celecoxib with radiation therapy for poor-prognosis head and neck cancer, including reirradiation. Enrollment was 15 participants; the record lists no masking and no randomized allocation. The primary endpoint is the number of participants with acute and late toxicity, with a timeframe of 30 days.

Secondary outcomes include clinical response, locoregional progression, locoregional control, progression-free survival, overall survival and late toxicity. The record reports no analyses for these outcomes and does not report a primary toxicity result. Safety fields list 14 participants at risk and 14 affected in both serious and other adverse-event categories, along with selected serious-event terms and percentages.

The absence of a control group limits the study to descriptions of the treated group rather than randomized comparisons. Detailed endpoint methods, the statistical plan, missing-data procedures and outcome analyses are not reported in the registry record.