Erlotinib in Treating Patients With Recurrent or Metastatic Skin Squamous Cell Carcinoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Erlotinib in Treating Patients With Recurrent or Metastatic Skin Squamous Cell Carcinoma (NCT01198028), a na phase 2 study. The primary outcome is Overall Response Rate.

2. Clinical Question

3. Trial Design

This was a phase II, single-group trial sponsored by M.D. Anderson Cancer Center. The registry lists an enrollment of 42 and a completed status. It records a start date of 2011-03-10 and a primary completion date of 2019-05-01. These registry entries describe the study record; they do not establish how many participants were included in an outcome analysis.

All participants were assigned to the treatment group and received erlotinib by mouth once daily while disease progression or unacceptable toxicity was absent. The registry records no masking and allocation as not applicable. Because there was no control arm, the design was not a randomized comparison. Differences between the observed outcome and outcomes in other populations or studies cannot be attributed to erlotinib using this trial alone.

The primary outcome was overall response rate, defined as the percentage of evaluable participants with complete or partial response. The registry describes assessment using Response Evaluation Criteria in Solid Tumors 1.1A and gives a timeframe of up to 6 years. Secondary outcomes include duration of response, duration of stable disease, progression-free survival, overall survival, and safety and tolerability. The registry lists 6 outcome measures, but detailed results for these outcomes are not included in the record.

The primary endpoint description says that a Bayesian design based on predictive probability would be implemented. The registry does not report the design’s decision thresholds, prior distributions, stopping rules, or actual Bayesian analysis. Such details cannot be reconstructed from the brief design statement. The record also does not report a sample-size justification, interim analyses, or a 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.

Before assignment: A total of 42 participants enrolled,39 participants started 1 withdrew, 1 came off study the same day consented and 1 participant was ineligible.

Overall Study

Reasons for not completing this period:

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 Overall Response Rate, measured over up to 6 years.

Overall response rate is intended to describe the proportion of evaluable participants who achieve either complete response or partial response. Complete response and partial response are response categories defined by the stated tumor-assessment criteria. The endpoint combines these categories into one measure, rather than describing the depth or duration of response separately. The registry names Response Evaluation Criteria in Solid Tumors 1.1A and specifies an assessment timeframe of up to 6 years.

In general, a response-rate endpoint can provide a direct summary of tumor response among participants who can be evaluated. The evaluable population is important: a percentage depends on both the number meeting the response definition and the denominator used. Here, neither the response count nor the evaluable denominator is reported in the registry record. The registry enrollment of 42 should not be substituted for the evaluable denominator because the record does not say that every enrolled participant was evaluable.

The endpoint does not by itself describe how long a response lasts, how many participants have stable disease, or how long participants live without progression or overall. Those topics are listed as secondary outcomes. Their results are not reported in the registry record. Nor does a response rate from a single-group study establish what would have occurred under another treatment or without erlotinib.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a single-group design allow?

In general, a single-group study describes outcomes in participants receiving the intervention under study. It can report how many participants met a defined outcome and how those outcomes were distributed, if the results are available. It does not directly estimate a contrast between the intervention and a concurrent control group. A numerical outcome in one group can be influenced by the characteristics of the participants, follow-up, assessment practices, and other care as well as by the intervention. Without a comparator, those influences cannot be disentangled in the same way as in a randomized comparison.

For this trial, the record identifies a single-group model, no masking, and allocation as not applicable. This is why the response question is descriptive rather than a treatment-versus-control contrast. The registry does not report a comparative estimate or a randomized analysis.

How is overall response rate defined?

The registry defines overall response rate as the percentage of evaluable participants with complete response or partial response. In general, a proportion is interpreted using its numerator and denominator: the numerator identifies participants meeting the response definition, while the denominator identifies the population assessed. If either is missing, the percentage cannot be checked or fully interpreted. The response count, evaluable denominator, and percentage are not reported in the registry record.

The registry names Response Evaluation Criteria in Solid Tumors 1.1A as the assessment criteria. General response criteria standardize how tumor changes are classified, but the registry description does not provide further information about assessment schedules, confirmation requirements, or independent review. Those details are not reported in the registry record and should not be assumed.

What does a confidence interval communicate?

In general, a confidence interval expresses statistical uncertainty around an estimate under a specified method and its assumptions. A narrower interval indicates greater precision under those assumptions; a wider interval indicates less precision. An interval does not remove possible bias, establish causation, or account automatically for limitations such as missing outcome assessments. No confidence interval for overall response rate is reported in the registry record, so its precision cannot be assessed from the primary endpoint information.

What is the stated Bayesian design?

The endpoint description says a Bayesian design based on predictive probability would be implemented. In general, Bayesian methods combine a statistical model with prior assumptions to describe uncertainty about quantities of interest. Predictive probability can be used to assess how likely a future or final result is to meet a specified goal, given the data available at a decision point. The practical meaning depends on the model, prior distribution, decision threshold, timing of assessments, and rules for acting on the calculation.

Those details are not reported in the registry record. The statement alone does not show which priors or thresholds were used, whether the trial followed the planned design, whether any interim analysis occurred, or what the final Bayesian result was. No Bayesian estimate or decision outcome is reported.

How should the planning illustrations be understood?

In general, a sample-size calculation describes how many participants might be needed under a specified design, effect, and statistical assumptions. The listed planning assumptions describe two-sided significance level 0.05, equal allocation between two arms, and a normal approximation, with an effect to detect expressed as a fraction of an observed effect. For time-to-event calculations, the stated basis is a total number of events; other calculations are described as participants per arm.

These are planning illustrations based on an observed effect, not the trial’s own design. The trial is single-group, whereas the illustration assumptions describe a two-arm allocation. The observed primary effect needed to interpret the illustrations is not reported in the registry record. The illustrations therefore cannot establish the study’s actual sample-size rationale, reproduce its design, or be interpreted as observed power. Sample-size calculations are planning quantities, not proof that a study had a particular probability of detecting an effect after the fact.

Why can over-dispersion matter?

In general, over-dispersion occurs when variability in observed data exceeds what a chosen statistical model assumes. If a model understates variability, uncertainty estimates may be too small and inferential conclusions may appear more precise than warranted. The appropriate model depends on the outcome and design. The primary endpoint is a binary response summarized as a count or proportion. The registry record does not report a fitted model, a dispersion assessment, or an analysis that addresses over-dispersion, so no trial-specific conclusion on this issue is available.

More broadly, statistical interpretation depends on the analysis population, outcome definition, missing-data handling, and uncertainty estimate as well as the point estimate. For this trial, several of those details are absent from the registry record, limiting independent assessment of the planned primary analysis.

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("Erlotinib")
n      <- c(29)
events <- c(3)
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  Erlotinib  29  3  10.3  2.2  27.4  3.6  26.4  3.8  26.5

11. Multiplicity, Missing Data and Analysis Populations

The registry lists overall response rate as the primary outcome and also lists duration of response, duration of stable disease, progression-free survival, overall survival, and safety and tolerability as secondary outcomes. It lists 6 outcome measures. When a study evaluates multiple outcomes or makes multiple statistical comparisons, the chance of at least one seemingly notable finding can increase. In general, a prespecified multiplicity strategy can address that issue, but the registry record does not report whether one was planned or used. It also does not report a statistical analysis plan or the results of analyses for the secondary outcomes.

The registry lists 0 analyses. This means no analyses are listed in the registry record; it does not establish which analyses were conducted outside the record. Interim analyses, stopping rules, and any adjustments for multiple looks are not reported in the registry record. The mention of a Bayesian predictive-probability design does not, by itself, provide these details.

The primary endpoint applies to evaluable participants, but the record does not report the size or definition of that population, the number excluded from assessment, reasons for missing assessments, or how missing outcome data were handled. It is also not reported whether the analysis was based on all enrolled participants, treated participants, or another analysis population. These details matter because the denominator and handling of unavailable assessments can affect a response-rate estimate.

General missing-data methods rely on assumptions about why outcomes are unavailable. Without counts, reasons, and analysis rules, those assumptions cannot be evaluated for this study. No missing-data sensitivity analysis is reported in the registry record.

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: alanine aminotransferase increased (2.6%), lung infection (2.6%), bronchopulmonary hemorrhage (2.6%), anemia (grade 3) (2.6%), nausea (grade 3) (2.6%).

The registry’s erlotinib safety figures list 39 participants at risk for serious adverse events and 7 affected by at least one serious adverse event. For other adverse events, it lists 39 at risk and 22 affected. The record lists 7 deaths. These are registry counts; the record does not establish that the serious events, other events, or deaths were caused by erlotinib.

The listed serious-event terms include Alanine aminotransferase increased, Lung Infection, Bronchopulmonary hemorrhage, Anemia (grade 3), Nausea (grade 3), and Diarrhea (grade 3), each recorded as 2.6%. These term-specific figures should be read as registry reports, not as a complete account of all clinical circumstances. The record does not provide further event narratives, timing, duration, or an assessment of relatedness.

In general, safety counts describe events reported within a defined study population; they do not by themselves establish comparative risk. Here, there is no control group for comparison. The record also does not describe whether participants could experience multiple events or how event severity and attribution were adjudicated beyond the labels listed. The safety figures should not be combined with the missing primary response result to infer an overall benefit–risk conclusion.

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

The registry record establishes that a phase II, single-group study of erlotinib was conducted in participants with recurrent or metastatic squamous cell carcinoma of the skin. It records an enrollment of 42, a completed status, and a primary endpoint intended to measure the percentage of evaluable participants with complete or partial response. It also provides a brief treatment description and selected adverse-event counts.

The record does not report the overall response rate, its numerator or denominator, or an uncertainty interval. It therefore does not show how frequently participants achieved the defined response endpoint or how precisely that frequency was estimated. The listed registry enrollment cannot stand in for the evaluable population, which is not reported.

The absence of a comparison group also matters. Even if an outcome estimate were available, a single-group result would describe participants in the study and would not, on its own, establish what would have happened under a different treatment or without the study treatment. It would not isolate a causal effect from differences in disease course, participant characteristics, or other care.

The record’s safety figures describe reported event counts, but do not establish treatment causation or provide a controlled comparison. Secondary outcomes are named, but their results are not reported in the registry record. The trial’s completed status does not imply any particular efficacy or safety conclusion.

Accordingly, the registry information supports a description of the study design, endpoint, and selected safety counts, but not a conclusion about the primary response result. It also does not provide the analysis details needed to judge the study’s statistical precision, handling of missing outcomes, or implementation of its stated Bayesian design.

14. Why This Trial Matters Statistically

This trial is statistically informative as an example of how a clearly named endpoint can still be difficult to interpret when its result and analysis details are absent. Overall response rate is defined among evaluable participants, making the denominator especially important. The record gives an enrollment figure but does not report the evaluable count or the response count. Without those details, neither the response proportion nor its precision can be assessed.

The design also illustrates the distinction between describing an outcome and estimating a comparative effect. A single-group study can organize response and safety observations, but it does not contain a concurrent control group. Statistical methods cannot create the missing randomized contrast. This distinction is essential when readers encounter results from non-comparative phase II studies.

The endpoint description refers to a Bayesian predictive-probability design, while the registry record does not provide model specifications or decision rules. A design label is not enough to reproduce or evaluate an analysis. Transparent reporting of assumptions, analysis populations, uncertainty, and decision thresholds is needed to understand how a planned method connects to the reported result.

Finally, the planning assumptions describe a two-arm calculation even though the trial is listed as single-group. This underscores why planning illustrations should be read in light of the actual study design and not treated as evidence of observed power or as a retrospective justification. The record offers useful design and safety context, but leaves the primary statistical result unreported.

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 trial-specific limitation in the registry record is that the primary outcome result is not reported. There is no response count, evaluable denominator, response percentage, or uncertainty interval. Thus the endpoint cannot be independently assessed from the registry information.

The study has no comparator and is not randomized. As a result, it cannot provide a direct treatment-versus-control estimate. The absence of masking is also recorded, but the registry does not describe how response assessments were conducted or whether any independent review was used. Those assessment details are not reported in the registry record.

Although enrollment is listed as 42, the record does not state how many participants received treatment, completed assessment, or contributed to each outcome. It does not report baseline participant characteristics, reasons for non-evaluation, missing-data rules, or sensitivity analyses. This restricts assessment of who is represented by any outcome population.

The registry mentions a Bayesian predictive-probability design but omits the prior assumptions, model, thresholds, decision rules, and analysis results. It does not report interim analyses or a sample-size justification. Secondary outcomes are named without results. Safety figures are available, but further information on event timing, attribution, and clinical context is not reported.

The record’s completion status and enrollment entry do not resolve these limitations. They describe administrative study information, not the completeness or validity of the outcome analysis. The planning illustrations described with the record use assumptions for a two-arm design, while this trial is listed as single-group; their connection to the study’s actual design is therefore not established by the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Erlotinib in Treating Patients With Recurrent or Metastatic Skin Squamous Cell Carcinoma — NCT01198028.
• Registry record last updated 2020-06-11; status last verified 2020-06.
• Results first posted 2020-06-11.
• 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

NCT01198028 is a completed phase II, single-group study sponsored by M.D. Anderson Cancer Center. It enrolled 42 participants with recurrent or metastatic squamous cell carcinoma of the skin. Participants received erlotinib by mouth once daily while disease progression or unacceptable toxicity was absent. The registry records no masking and allocation as not applicable.

The primary endpoint was overall response rate among evaluable participants, defined as complete or partial response using Response Evaluation Criteria in Solid Tumors 1.1A, with a timeframe of up to 6 years. The response result and evaluable denominator are not reported in the registry record. Listed secondary outcomes include response duration, stable disease duration, progression-free survival, overall survival, and safety and tolerability.

The record lists serious and other adverse-event counts and 7 deaths, without establishing causation. It does not report the analysis details needed to evaluate missing outcomes, the Bayesian design, or statistical precision.