Alpelisib in Treating Participants With Transorally Resectable HPV-Associated Stage I-IVA Oropharyngeal Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Alpelisib in Treating Participants With Transorally Resectable HPV-Associated Stage I-IVA Oropharyngeal Cancer (NCT03601507), a na phase 2 study. The primary outcome is Quantitative Change in the Sum of Response Evaluation Criteria in Solid Tumors (RECIST) -.

2. Clinical Question

3. Trial Design

This was a phase II, open-label, single-group window trial sponsored by the University of Arizona. The registry lists no masking and no allocation between groups. The planned sequence was a short course of alpelisib before conventional surgery, allowing investigation of tumor and biomarker measurements over the preoperative interval. The registry lists laboratory biomarker analysis and a pharmacodynamic study as additional interventions.

Enrollment is listed as 9. The study started on 2019-03-11, and primary completion is listed as 2022-05-24. Its status is TERMINATED. The reason for termination is not reported in the registry record, so the status alone does not establish why the study stopped or whether the planned enrollment and assessments were completed.

The treatment description specifies oral alpelisib once daily for 14-21 days when there was no disease progression or unacceptable toxicity, followed by surgery. Treatment could continue for up to 28 days if surgery was delayed. This brief treatment window is distinct from a long-term treatment comparison: the listed primary timeframe is baseline up to 28 days, and the treatment was followed by surgery.

With no control group, any observed change would describe participants’ measurements over time rather than isolate the effect of alpelisib from other influences. The registry record does not report a protocol-level explanation of the sample-size choice, detailed eligibility criteria beyond the listed conditions, 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.

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 Quantitative Change in the Sum of Response Evaluation Criteria in Solid Tumors (RECIST) -, measured over Baseline up to 28 days.

The primary endpoint is titled Quantitative Change in the Sum of Response Evaluation Criteria in Solid Tumors (RECIST) -, with the title ending after a dash in the registry record. Its description concerns measurable index lesions assessed on paired pre-treatment and post-treatment imaging, with size treated as a continuous variable. The timeframe is baseline up to 28 days. The listed unit is participants and the parameter is a count of participants, which does not make the endpoint’s description of continuous lesion-size change fully clear.

The description refers to paired computed tomography scans and then says assessment is by MRI. This difference in the stated imaging modality is not reconciled in the registry record. The endpoint description also gives RECIST v1.0 categories: complete response is disappearance of all target lesions; partial response is a >=30% decrease in the sum of longest diameters; progressive disease is an increase of >=20% and >=5 mm from nadir; and stable disease meets neither the partial-response nor progressive-disease definition.

In general, measuring lesions continuously can describe the magnitude and direction of short-term change rather than only assigning a response category. RECIST categories provide a framework for interpreting tumor measurements, but the posted record contains no primary-endpoint estimate or analysis results. The endpoint is therefore useful for understanding what the trial set out to measure, not for concluding what change occurred.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical interpretation depends on what was measured, which participants contributed data, and how the comparison was constructed. In this trial, the central design feature is the single group. Participants received alpelisib before surgery, but there was no randomized or concurrent control group. A before-and-after measurement can describe change within participants; by itself, it cannot establish how much of that change is attributable to treatment.

What a continuous lesion measure represents

The primary description treats lesion size as a continuous variable and refers to change in the sum of target-lesion measurements. In general, continuous measurements retain information about the size and direction of change. That differs from categorizing each participant as having complete response, partial response, progressive disease, or stable disease. Categorization can make clinical patterns easier to communicate, but it compresses measurement detail and depends on specified thresholds.

The registry description includes RECIST v1.0 criteria for target lesions: complete response means disappearance of all target lesions; partial response is a >=30% decrease in the sum of longest diameters; progressive disease is an increase of >=20% and >=5 mm from nadir; and stable disease meets neither the partial-response nor the progressive-disease definition. These criteria are part of the endpoint description, not reported trial outcomes. The record does not clarify how the continuous change and the stated count-of-participants parameter were reconciled in the analysis.

Why paired measurements need careful interpretation

Paired imaging means measurements are intended at baseline and after treatment for the same participant. In general, analyzing within-participant change can account for stable differences in baseline lesion size. It still depends on having comparable measurements at both times and a clear rule for handling participants without a post-treatment scan. The registry record does not report how missing or unevaluable paired scans were handled, how lesions were selected, or whether measurement review was blinded. The record refers to computed tomography scans and also says assessment was by MRI, so the imaging modality is not consistently described.

Measurement variation matters because small apparent changes can arise from image acquisition, lesion selection, or reader assessment as well as biological change. General statistical practice can address this with standardized measurement procedures and uncertainty estimates, but those procedures and estimates are not reported in the registry record. No numerical primary result is posted, so the amount of observed change and its precision cannot be evaluated.

What a control group contributes

In general, a control group helps distinguish changes associated with an intervention from changes that might have occurred without it. Here, the single-group design offers no such direct comparison. It can provide descriptive information about measurements over the treatment window, but causal conclusions require caution. The record does not report a planned external control or a method for adjusting for the absence of one.

The open-label classification means masking is listed as NONE. In general, awareness of treatment can influence subjective assessments or decisions about measurement and care. Imaging measurements may be more objective than subjective outcomes, but they are not automatically free of assessment bias. The registry record does not report reader masking or independent imaging review.

Planning illustrations are not observed power

The registry’s sample-size assumptions describe planning illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the target effect is a stated fraction of the observed effect. These assumptions do not describe the actual single-group design, which has no two-arm allocation. Such illustrations are not the trial’s original sample-size justification, and they are not observed power. The registry record does not report a sample-size justification or a calculated primary-endpoint effect to place those hypothetical planning assumptions in context.

In general, a sample-size calculation describes the amount of information a hypothetical design might need under chosen assumptions. It cannot substitute for a reported estimate and uncertainty interval from the completed analysis. Here, the record reports 0 analyses, so no result-based precision assessment can be made.

Over-dispersion and outcome distributions

In general, over-dispersion occurs when outcome variability is greater than a statistical model assumes. If it is ignored, standard errors may be too small and uncertainty may appear narrower than warranted. This concept is relevant when count or categorical outcomes are modeled, but the registry record does not report a statistical model, variance estimate, or over-dispersion assessment for this trial. It would therefore be inappropriate to claim that over-dispersion was present or that a particular correction was used.

Likewise, the record does not report confidence intervals, a hypothesis-testing procedure, covariate adjustment, missing-data methods, or prespecified subgroup analyses. General methods can explain how such choices affect interpretation, but they cannot fill in methods that are absent from the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists 9 outcome measures and 0 analyses. Several secondary outcomes address distinct domains: adverse events, surgical complications, length of hospital stay, and pre/post-treatment tumor measurements involving HPV mRNA, E6 and E7 oncoproteins, phospho-HER3, and HER3/PI3K dimers. Listing multiple outcomes raises a general multiplicity issue: when many outcomes are tested, the chance of at least one apparently noteworthy result can increase. The registry record does not report whether any multiplicity adjustment or hierarchy among outcomes was planned.

The record does not report whether outcomes were analyzed according to a prespecified analysis population, such as all enrolled participants or only those with evaluable paired scans. It also does not report how missing imaging, biomarker samples, or follow-up measurements were handled. These omissions matter because participants with complete measurements can differ from those without them, and the resulting analyzed group may not represent everyone enrolled. The safety summary’s risk denominators are 6, whereas enrollment is 9; the reason for that distinction is not reported in the registry record.

In general, a transparent analysis plan identifies the primary outcome, defines the analysis population, describes missing-data handling, and explains how multiple outcomes are interpreted. No such plan is reported in the registry record. Since there are no reported analysis entries, it is not possible to assess whether the listed outcomes were analyzed as planned, whether any results were omitted, or how uncertainty and multiple testing were addressed.

11. 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 non-serious adverse event terms

The serious adverse-event summary lists 6 participants at risk and 0 affected, with 0 deaths. The summary for other adverse events lists 6 at risk and 6 affected. These figures describe the counts in the registry record; they are not estimates of risk for all 9 enrolled participants because the stated safety denominators are 6. The reason for the denominator is not reported in the registry record.

No top serious-event terms are listed. A count of 0 affected in the reported serious-event group means no serious events are recorded for that group in this summary. It does not establish that serious events were impossible, nor does it resolve whether every enrolled participant was included in safety follow-up. The record does not report event timing, severity grading beyond the serious/non-serious distinction, duration, relatedness assessments, or participant-level details.

The other-event count indicates that all 6 participants listed as at risk were recorded as affected by at least one other adverse event. The registry record does not identify the events in this summary. These counts should be read as a limited descriptive safety report, not as a complete account of tolerability or as a comparison with another treatment.

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

The record establishes that a phase II, open-label, single-group trial examined alpelisib in a short preoperative window before surgery for HPV-associated, transorally resectable oropharyngeal cancer. It identifies quantitative change in RECIST target-lesion measurements as the primary endpoint and lists several biomarker and perioperative outcomes. It also reports limited adverse-event counts for 6 participants at risk.

It does not report analyzed primary-endpoint results. Therefore, the record does not show the direction or magnitude of lesion change, the precision of any estimate, or whether measurements differed across participants. It also does not report outcome estimates for the listed biomarker measures, surgical complications, or hospital stay. The absence of those results should not be interpreted as evidence of no change or no events.

Even if within-participant changes were reported, a single-group before-and-after study would not provide a direct treatment comparison. In general, changes over time can arise from the intervention, measurement variability, disease course, or other factors. Without a control group, those contributions cannot be separated reliably by design alone. The short follow-up window also addresses a different question from long-term outcomes; the registry record does not report long-term clinical outcomes in the listed primary timeframe.

The safety counts are similarly bounded: they describe the participants included in the registry’s stated risk groups, not necessarily every enrolled participant. The enrollment count is 9, while both safety groups list 6 at risk. The reason for the difference is not reported in the registry record. Accordingly, the strongest interpretation is descriptive: the registry identifies the study question, treatment sequence, endpoint framework, and some safety counts, but it does not provide analyzed efficacy results or enough detail to draw broader conclusions.

13. Why This Trial Matters Statistically

This trial is statistically informative as an example of a window study: treatment is given before planned surgery, creating an opportunity to compare tumor measurements and biomarkers over a defined short interval. Such designs can support investigation of biological and imaging changes without replacing the planned surgical treatment. The registry title emphasizes biomarker modulation, while the listed primary endpoint concerns lesion measurements, making it important to distinguish the trial’s rationale from its formal primary outcome.

The design also illustrates why endpoint definitions and reporting consistency matter. The primary description combines a continuous change in lesion size with a participant-count parameter, and it refers to computed tomography as well as MRI. Those features make it harder to determine from the registry alone exactly how the primary endpoint was operationalized. A clear specification of the measurement, imaging modality, analysis population, and summary statistic would support more reproducible interpretation.

A single-group study can be useful for describing short-term observations, but it cannot provide the same causal contrast as a controlled comparison. This matters especially when interpreting pre/post changes: a measurement may change during treatment without the design demonstrating that treatment caused the change. The record’s lack of primary analysis results further limits conclusions. The example therefore highlights that registration of an endpoint and completion of a study are not substitutes for posting an estimate with its uncertainty and analysis details.

Finally, enrollment and safety denominators should be distinguished. The registry lists enrollment of 9 but safety groups of 6 at risk. Without an explanation of that difference or analysis-population details, readers should avoid assuming all counts share the same denominator. These are practical issues in interpreting small, exploratory clinical studies.

14. 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.

Several trial-specific limitations constrain interpretation. The design is single group, with allocation listed as NA and masking listed as NONE. No concurrent comparator is described, so the registry cannot support a direct contrast between alpelisib and another treatment. The enrollment is listed as 9, and the record reports 0 analyses, leaving no posted primary-endpoint result or measure of its precision.

The registry record does not explain why the study was terminated. It does not report whether enrollment, treatment, surgery, imaging, or planned biomarker assessments were completed for all participants. The safety summaries list 6 participants at risk, fewer than the enrollment count, and the reason is not reported. No top serious adverse-event terms are listed, and the record does not give event-level details for the other adverse events.

The endpoint description has interpretive ambiguities: it refers to continuous lesion-size change but lists a participant-count parameter and unit, and it mentions paired computed tomography scans while also stating assessment by MRI. The endpoint title itself ends after a dash. The record does not report how these issues were resolved in the analysis, how paired measurements were handled when incomplete, or whether imaging readers were masked.

Statistical methods are also insufficiently documented for evaluating results. The registry record does not report a sample-size justification, analysis plan, missing-data method, confidence intervals, hypothesis-testing method, or multiplicity procedure. It lists multiple secondary outcomes but no analyses. These limitations describe gaps in what is reported; they do not establish what investigators did outside the registry record.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Alpelisib in Treating Participants With Transorally Resectable HPV-Associated Stage I-IVA Oropharyngeal Cancer — NCT03601507.
• Registry record last updated 2024-04-17; status last verified 2024-04.
• Results first posted 2024-04-17.
• 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.

18. Record Summary

NCT03601507 is a phase II, open-label, single-group trial of preoperative alpelisib for HPV-associated, transorally resectable oropharyngeal cancer. The treatment description specifies oral dosing for 14-21 days before surgery, with treatment up to 28 days if surgery is delayed. Enrollment is listed as 9, and the status is TERMINATED.

The primary endpoint concerns quantitative change in the sum of RECIST target-lesion measurements from baseline up to 28 days. The record lists 0 analyses, so no analyzed primary result is reported. Safety summaries list 6 participants at risk for serious adverse events, 0 affected, and 0 deaths; for other adverse events, 6 were at risk and 6 were affected. The reason for the safety-denominator difference and the reason for termination are not reported in the registry record.