This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record. Numerical trial results on this page are restricted to the results posted in the ClinicalTrials.gov record; calculations added by Clinical Biostats are labelled as such.
1. Trial at a Glance
This page covers the registry results for Soy Isoflavones in Preventing Head and Neck Cancer Recurrence in Patients With Stage I-IV Head and Neck Cancer Undergoing Surgery (NCT02007200), a na phase 2 study. The primary outcome is Mean Percent Change in p16 Methylation (% CpG Sites Methylated) in Tumor Tissue After Soy Isoflavone.
| Feature | Registry value |
|---|---|
| ClinicalTrials.gov identifier | NCT02007200 |
| Phase | Phase 2 |
| Condition | Recurrent Hypopharyngeal Squamous Cell Carcinoma, Recurrent Laryngeal Squamous Cell Carcinoma, Recurrent Laryngeal Verrucous Carcinoma |
| Allocation | Not applicable (single group) |
| Design model | Single Group |
| Masking | None (open label) |
| Enrollment | 55 |
| Arms in the primary outcome | n/a |
| Lead sponsor | National Cancer Institute (NCI) |
| Results posted | Yes |
2. Clinical Question
Population
Patients with stage I-IV or recurrent head and neck cancers who were undergoing surgery. The registry lists cancers of the hypopharynx, larynx, lip and oral cavity, oropharynx, and tongue, including squamous cell and verrucous carcinomas. The primary analysis included 39 evaluable patients.
Intervention
Patients received soy isoflavones by mouth for approximately 14 days before surgery. Laboratory biomarker analysis and survey administration were also listed as interventions. The record does not report an administered dose.
Comparator
There was no separate comparator group. The trial is recorded as single-group, with allocation listed as not applicable and masking listed as none. The primary measurement compares tissue before treatment with tissue after treatment, rather than comparing treatment with a control.
Primary question
What was the mean percent change in p16 methylation in tumor tissue between baseline and surgery after soy isoflavone treatment? The planned timeframe was from baseline to surgery, up to 42 days. The registry record does not report the measured change or its uncertainty.
3. Trial Design
This phase II trial examined preoperative soy isoflavones in patients with head and neck cancer undergoing surgery. Its stated aim was to study molecular markers in connection with prevention of head and neck squamous carcinoma. The registry record describes a single-group study: participants received soy isoflavones, and the primary endpoint was based on measurements before and after treatment. No randomized or untreated comparison group is listed. Allocation is recorded as not applicable, and masking as none.
The intervention description says patients received soy isoflavones by mouth for approximately 14 days before surgery. The primary endpoint timeframe extends from baseline to surgery, up to 42 days. These descriptions do not establish that every participant had the same interval between measurements. The administered dose and further details of treatment adherence are not reported in the registry record.
The registry records 55 patients enrolled. It states that 3 patients did not undergo treatment and 13 had insufficient tissue and were not evaluable. The primary analysis included the 39 evaluable patients. These are distinct enrollment and analysis descriptions; the registry does not report further detail about the reasons for non-evaluability or how missing measurements were handled.
The trial is recorded as completed. Its start is listed as 2009-07, and primary completion as 2015-09. The sponsor is the National Cancer Institute. The registry lists 4 outcome measures and 1 analysis. Further information about the protocol’s sample-size justification, interim analyses, and statistical analysis plan is not reported in the registry record.
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
| Stage | Treatment (Soy Isoflavones) |
|---|---|
| STARTED | 55 |
| COMPLETED | 55 |
| NOT COMPLETED | 0 |
Analysis population for the primary outcome, as defined in the registry: 55 patients were enrolled. 3 patients did not undergo treatment. 13 patients had insufficient tissue and were therefore not evaluable. Only the 39 evaluable patients were included in the 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: 55 patients were enrolled. 3 patients did not undergo treatment. 13 patients had insufficient tissue and were therefore not evaluable. Only the 39 evaluable patients were included in the analysis.
| Characteristic | Unit | Treatment (Soy Isoflavones) |
|---|---|---|
| Participants analysed | 39 | |
| Age, Continuous | Years | 60 (26 to 78) |
| Sex: Female, Male: Female | Participants | 12 |
| Sex: Female, Male: Male | Participants | 27 |
6. Primary Endpoint
The registry lists the primary outcome as Mean Percent Change in p16 Methylation (% CpG Sites Methylated) in Tumor Tissue After Soy Isoflavone, measured over From baseline to surgery, up to 42 days.
| Element | Registry entry |
|---|---|
| Result type | Mean |
| Measure of spread | Full Range |
| Unit | Percent change |
| Endpoint classification | not reduced to a single comparison |
The primary endpoint is the mean percent change in p16 methylation, expressed as the percentage of CpG sites methylated in tumor tissue after soy isoflavone treatment. The planned measurement runs from baseline to surgery, up to 42 days. The endpoint is a molecular measurement in tissue, not a count of cancer recurrences or a direct measure of survival.
A pre-treatment and post-treatment biomarker measurement can describe whether the measured marker changed over the treatment interval. The mean summarizes the average change for the analyzed group; the registry also identifies the full range as its dispersion measure. A mean alone would not show how variable individual changes were, so the range is relevant context. However, the result values, range, and number of paired observations used for this endpoint are not reported in the registry record.
General context: molecular markers can be studied as intermediate biological outcomes because they may be measured over a shorter period than clinical outcomes. A change in a marker does not by itself establish a change in recurrence risk, survival, or other clinical outcomes. The registry’s stated primary endpoint therefore answers a narrower question about p16 methylation in tumor tissue over the preoperative interval.
7. Why No Re-analysis Is Shown
8. Statistical Interpretation
The registry record describes the question and intended analysis, but it does not report the primary endpoint estimate. No mean percent change, full range, confidence interval, or p-value for the change in p16 methylation is reported in the registry record. The direction and precision of the observed result therefore cannot be assessed from the posted results.
The design is single-group and uses measurements before and after treatment. Such a comparison can describe change over time among participants with evaluable tissue, but it does not provide a concurrent control group for separating the treatment interval from other influences. For example, changes over time or differences in tissue sampling could also contribute to a pre/post difference. This is a general limitation of interpreting an uncontrolled before-and-after study, not a finding about what occurred in this trial.
The planned model included time, current smoking status, the interaction between time and smoking status, and tissue type. The registry also describes linear contrasts to assess differential effects by tissue type and smoking status. These are elements of the analysis plan, not evidence that any subgroup difference was observed. Results for these terms and contrasts are not reported in the registry record.
The primary analysis population was restricted to the 39 evaluable patients. The record reports that 3 patients did not undergo treatment and 13 had insufficient tissue and were not evaluable. Without endpoint values or further information about missing measurements, it is not possible to determine how the analyzed patients’ results compare with those who were not evaluable, or how much uncertainty is associated with the estimate.
The registry lists secondary outcomes concerning participants alive at follow-up and alive without relapse at last follow-up. Their results are not reported in the registry record. Consequently, the posted results do not establish whether the biomarker endpoint was associated with later clinical outcomes.
General context: statistical precision depends on the observed variation and the amount of usable outcome information, not enrollment alone. Since the registry does not report the primary estimate, its dispersion, or a confidence interval, the precision of the biomarker result cannot be characterized. No conclusion about the size or direction of a treatment effect follows from the design description alone.
9. Statistical Methods Explained
The primary analysis is described as a linear repeated measures model. Repeated measures means that observations from the same participant, such as a baseline measurement and a later measurement, are analyzed with recognition that they are related. Treating those observations as independent would generally fail to represent that within-participant relationship. The registry does not report the model’s covariance structure or other implementation details.
What does the model compare?
The planned model includes time as a fixed effect, with pre-treatment and post-treatment as its levels. In general, a time effect represents an average difference across the specified measurement times, subject to the rest of the model. Here, the endpoint description concerns change from baseline to surgery. The actual estimated time effect is not reported in the registry record.
Current smoking status is also listed as a fixed effect, with yes or no as its categories. The model includes an interaction between time and smoking status. In general, an interaction asks whether the time-related change differs across smoking-status categories. Including this term is not proof that the change differed; the estimated interaction and its uncertainty are not reported in the registry record.
Tissue type is another listed fixed effect, distinguishing tumor from non-tumor tissue. The record describes linear contrasts to assess the differential effect of soy isoflavone on tumor and non-tumor tissues between smokers and non-smokers. A linear contrast is a specified combination of model estimates used to address a focused comparison. The registry does not report the contrast estimates, their uncertainty, or the exact contrast coefficients.
Why use an adjustment for degrees of freedom?
The description specifies Satterthwaite’s adjustment to the degrees of freedom to account for heteroscedasticity. General context: heteroscedasticity means that variability is not constant across groups or conditions. When model assumptions about common variability are not appropriate, standard errors and inferential tests can be affected. Satterthwaite-type adjustments approximate the degrees of freedom used in inference to reflect estimated uncertainty. This addresses an aspect of inference; it does not remove possible bias from the single-group design or missing tissue measurements.
The endpoint is reported as a mean percent change, with the full range identified as the dispersion measure. These summaries answer different questions. The mean describes the group’s average change, while the range describes the spread from the lowest to highest observed values. A range can be sensitive to extreme observations and does not show the distribution’s overall shape. The registry does not provide the values needed to interpret either summary.
What can be inferred from the planned analysis?
A statistical model specifies how measurements and covariates are to be related; it does not, by itself, provide a result. The registry’s description supports understanding which factors were planned for the primary analysis and which comparisons were intended. It does not reveal whether the estimated average change was positive or negative, how large it was, whether it was precise, or whether any contrast met a statistical threshold. Those results are not reported in the registry record.
General context: confidence intervals, when reported, communicate a range of values compatible with the data and model under the stated method. A p-value addresses compatibility with a specified null hypothesis under model assumptions; it is not a measure of clinical importance. Neither can be interpreted for this primary endpoint because the relevant estimate and inferential results are not reported in the registry record. The registry’s method description also does not state the confidence level or significance threshold used for the primary endpoint.
The model can account for specified relationships among the measured factors, but it cannot create a control group that the design did not include. Nor does the listed adjustment for heteroscedasticity address every possible source of bias, measurement variation, or missingness. The registry record does not report additional sensitivity analyses or diagnostics that would clarify the robustness of the model’s assumptions.
10. Multiplicity, Missing Data and Analysis Populations
The primary analysis includes multiple model terms: time, current smoking status, their interaction, and tissue type. It also describes linear contrasts for differences by tissue type and smoking status. When several hypotheses are examined, the chance of at least one false-positive finding can increase if each is tested separately without adjustment. General context: multiplicity procedures can control or otherwise address that issue, but the appropriate approach depends on which comparisons are considered confirmatory and how they are organized. A multiplicity adjustment is not reported in the registry record.
The registry identifies one primary endpoint, but the analysis description also includes subgroup-related terms and contrasts. It does not report whether these analyses were prespecified as confirmatory or exploratory, how many contrasts were planned, or how any multiplicity was handled. No inference about statistical significance across the analyses can be made without their results and analysis details.
The analysis population is described as the 39 evaluable patients. The record reports 55 enrolled, 3 who did not undergo treatment, and 13 with insufficient tissue who were not evaluable. It does not report an imputation method, a missing-data assumption, or a sensitivity analysis for participants without evaluable tissue or outcome measurements. It also does not report whether every otherwise evaluable patient had all measurements required by the repeated measures model.
General context: excluding participants with unavailable outcomes can affect interpretation if the likelihood of missingness is related to the outcome or to participant characteristics. The registry record does not provide enough information to assess that possibility. The stated primary analysis population should therefore be understood as the evaluable group, not as a result for every enrolled participant.
11. Safety Results
The registry reports serious adverse events as affected participants divided by participants at risk.
| Arm | Participants with serious adverse events | Participants at risk | Percentage |
|---|---|---|---|
| Treatment (Soy Isoflavones) | 7 | 52 | 13.5% |
Other (non-serious) adverse events
Participants with at least one non-serious event at or above the registry's 0% reporting threshold.
| Arm | Participants affected of at risk |
|---|---|
| Treatment (Soy Isoflavones) | 25 of 52 |
Most frequent serious adverse event terms
| Event term | Organ system | Treatment (Soy Isoflavones) |
|---|---|---|
| Cardiopulmonary arrest, cause unknown (non-fatal) | Cardiac disorders | 1 of 52 |
| Nausea | Gastrointestinal disorders | 1 of 52 |
| Infection, Lung | Infections and infestations | 1 of 52 |
| Infection, Wound | Infections and infestations | 1 of 52 |
| Pain, Bone | Musculoskeletal and connective tissue disorders | 1 of 52 |
| Pulmonary/Upper Respiratory - Other | Respiratory, thoracic and mediastinal disorders | 1 of 52 |
| Acute Respiratory Distress Syndrome | Respiratory, thoracic and mediastinal disorders | 1 of 52 |
Most frequent non-serious adverse event terms
| Event term | Organ system | Treatment (Soy Isoflavones) |
|---|---|---|
| Glucose, serum-high (hyperglycemia) | Metabolism and nutrition disorders | 10 of 52 |
| Triglyceride, serum-high (hypertriglyceridemia) | Metabolism and nutrition disorders | 8 of 52 |
| Calcium, serum-low (hypocalcemia) | Metabolism and nutrition disorders | 6 of 52 |
| Fatigue (asthenia, lethargy, malaise) | General disorders | 5 of 52 |
| Albumin, serum-low (hypoalbuminemia) | Metabolism and nutrition disorders | 5 of 52 |
| Lipase | Metabolism and nutrition disorders | 5 of 52 |
| Magnesium, serum-low (hypomagnesemia) | Metabolism and nutrition disorders | 5 of 52 |
| Weight gain | General disorders | 4 of 52 |
| Nausea | Gastrointestinal disorders | 4 of 52 |
| Sodium, serum-low (hyponatremia) | Metabolism and nutrition disorders | 4 of 52 |
Serious events with the highest arm-level rate: cardiopulmonary arrest, cause unknown (non-fatal) (1.9%), nausea (1.9%), infection, lung (1.9%), infection, wound (1.9%), pain, bone (1.9%).
For the treatment group, the registry lists 52 participants at risk for serious adverse events and 7 affected. It also lists 52 at risk for other adverse events and 25 affected. These are registry counts for the reported adverse-event categories; the record does not provide enough detail to infer the timing, duration, severity distribution, or relationship of each event to soy isoflavones.
The listed serious adverse-event terms include cardiopulmonary arrest, cause unknown (non-fatal), nausea, lung infection, wound infection, bone pain, and pulmonary or upper respiratory events. Each term is reported as 1.9%. These term-specific percentages should be read as the registry’s reported figures, not as proof that every event was caused by treatment. The record does not provide a comparison group for evaluating whether the event frequency differs from another treatment or from usual care.
Deaths are not reported in the registry record. The adverse-event summaries alone do not establish the overall safety profile of the intervention, and no medical conclusion should be drawn from them without fuller clinical context.
12. What the Result Does, and Does Not, Mean
The record documents a phase II, single-group investigation of preoperative soy isoflavones and p16 methylation in tumor tissue. It identifies a planned mean percent change endpoint and describes a repeated measures analysis that includes smoking status and tissue type. The posted results do not report the endpoint estimate or its uncertainty, so the size, direction, and precision of the measured change cannot be described.
The design can, in principle, characterize a before-and-after biomarker pattern among participants with evaluable tissue. Without a separate comparison group, it cannot by itself distinguish a treatment-related change from changes associated with time, surgery-related sampling, or other influences. This is a general interpretation limit for the design, not evidence that any particular alternative explanation occurred.
The primary outcome is a molecular measurement rather than recurrence, survival, or another clinical outcome. The registry lists secondary outcomes related to being alive at follow-up and alive without relapse at last follow-up, but their results are not reported in the registry record. No relationship between p16 methylation change and those clinical outcomes can therefore be assessed from the posted results.
Enrollment and the evaluable analysis population are not interchangeable: the registry records 55 enrolled and says that 39 evaluable patients were included in the primary analysis. The reasons stated for non-evaluability include not undergoing treatment and insufficient tissue. The record does not report enough detail to assess whether the analyzed group differed systematically from those not included.
Overall, the record supports understanding what the investigators set out to measure and how they planned to analyze it. It does not support a conclusion about whether soy isoflavones changed the biomarker, how precisely any change was estimated, or whether a biomarker change corresponded to clinical benefit.
13. Why This Trial Matters Statistically
This trial is statistically informative as an example of a biomarker study in which the primary question concerns change over a short preoperative interval. The planned outcome is a mean percent change in p16 methylation in tumor tissue, and the analysis uses repeated measures rather than treating the observations at different times as unrelated. That distinction matters because repeated measurements from the same person are generally correlated.
The analysis plan also illustrates how researchers may investigate whether a measured change varies by a participant characteristic or tissue type. Smoking status, its interaction with time, and tissue type appear in the described model, with linear contrasts intended to examine differential patterns. Such planned analyses can add detail, but they also require clear reporting of estimates, uncertainty, and how multiple comparisons are handled. Those results and multiplicity details are not reported in the registry record.
The single-group design is central to interpreting any pre/post pattern. A change within participants can be described, but a control group would ordinarily help distinguish treatment-associated change from changes that might have occurred over the same interval for other reasons. The record does not contain such a comparison.
Finally, the difference between enrollment and the evaluable analysis population highlights why analysis populations and missing tissue matter in biomarker research. A stated analysis population tells readers whose measurements contribute to the primary analysis, while details about missingness help readers judge how representative that population may be. The registry reports the counts and some reasons for non-evaluability, but not the additional information needed to assess potential selection effects or the precision of the primary result.
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.
The main limitation for interpreting the trial is that the registry record does not report the primary endpoint result. Without the mean percent change, full range, confidence interval, or p-value, readers cannot assess direction, magnitude, variability, or statistical precision. The record also does not report the results of the planned smoking-status, tissue-type, or contrast analyses.
The single-group, unmasked design has no concurrent control for the primary comparison. A pre/post difference, if observed, would not by itself establish that the intervention caused it. The registry’s model description includes several measured factors, but it cannot replace a control group or account for unmeasured influences.
The primary analysis included 39 evaluable patients, while the record notes 3 patients who did not undergo treatment and 13 with insufficient tissue. Further reasons for missingness, participant characteristics by analysis status, and methods for handling missing measurements are not reported in the registry record. This limits assessment of how exclusion from the primary analysis may affect interpretation.
The intervention is described as approximately 14 days of oral soy isoflavones before surgery, while the primary endpoint timeframe extends up to 42 days. The record does not report dose, adherence, individual measurement intervals, or details of tissue collection and assay procedures. The reported analysis description also does not specify a covariance structure, diagnostic checks, or a multiplicity approach.
These limitations concern what can be concluded from the registry record; they do not establish that the underlying study lacked additional procedures or analyses. Those details are not reported in the registry record.
15. Related Tutorials
16. Related Statistical Calculators
17. Sources
- ClinicalTrials.gov: Soy Isoflavones in Preventing Head and Neck Cancer Recurrence in Patients With Stage I-IV Head and Neck Cancer Undergoing Surgery — NCT02007200.
- Registry record last updated 2016-10-24; status last verified 2016-09.
- Results first posted 2016-10-24.
- 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.
Continue through the Clinical Biostats statistical pathway
Use this trial's endpoint as a starting point for deeper study of effect measures, confidence intervals, and sample-size planning.
18. Record Summary
NCT02007200 is a completed phase II, single-group trial sponsored by the National Cancer Institute. It studied soy isoflavones before surgery in patients with stage I-IV or recurrent head and neck cancer. The intervention description specifies oral treatment for approximately 14 days before surgery.
The primary endpoint was mean percent change in p16 methylation in tumor tissue from baseline to surgery, up to 42 days. The planned analysis was a linear repeated measures model with time, current smoking status, their interaction, and tissue type; Satterthwaite’s adjustment was specified for heteroscedasticity.
The record reports 55 enrolled and 39 evaluable patients included in the primary analysis. The primary endpoint result, its uncertainty, and results for the listed secondary outcomes are not reported in the registry record. The single-group design has no separate control comparison.