GM-CSF Mouthwash for Preventing and Treating Mucositis in Patients Undergoing Radiation Therapy for Head and Neck Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for GM-CSF Mouthwash for Preventing and Treating Mucositis in Patients Undergoing Radiation Therapy for Head and Neck Cancer (NCT00293462), a randomized phase 3 study. The primary outcome is Comparison of Number of Participants With Grade 1 or 2 Oral Mucositis (GM-CSF and SS Groups Only).

2. Clinical Question

This is a question about the recorded occurrence of specified mucositis grades in the registry-defined analysis population, not about every possible effect of mouthwash during radiation therapy. The endpoint description identifies clinician-assessed oral mucosal changes under the RTOG Acute Radiation Morbidity Scoring Criteria, including anatomical findings and associated symptoms. The registry states that the study was randomized, double masked, and used a crossover model, but the full sequence and timing details are not reported in the registry record.

3. Trial Design

The registry lists this as a randomized, double-masked Phase 3 clinical trial with a crossover model. Enrollment is recorded as 91, and the trial status is completed. The listed start date is 2005-06-07 and the primary completion date is 2009-12-31. These registry-level details establish the stated design and administrative timeline, but do not provide a full account of how randomization was generated, how masking was maintained, or how the crossover was organized.

Three arm descriptions are listed. The GM-CSF group received sargramostim mouthwash as prevention. The salt-and-soda group received salt-and-soda mouthwash as prevention. The third group began with salt-and-soda mouthwash and was described as switching to GM-CSF if mucositis developed. The descriptions say patients held and swallowed mouthwash at intervals over an hour once daily. Some continuation text is incomplete in the registry record, so the exact treatment duration, crossover timing, and subsequent management are not reported in the registry record.

The primary outcome is explicitly framed as a comparison of the GM-CSF and salt-and-soda groups only. The switched group appears in the trial arms and in the secondary outcome postings, but is not part of the stated primary comparison. The registry defines the primary-outcome population as randomized subjects who completed the baseline questionnaire and had at least one mouthwash dose. The separate analysis populations for the secondary outcomes are reflected in their posted group counts, but further inclusion rules are not reported in the registry record.

The posted enrollment and outcome group counts should not be treated as interchangeable. The record reports enrollment of 91, while the primary outcome and secondary endpoints have their own arm-specific analysis counts. The registry does not explain the flow from enrollment through each outcome analysis. Randomization details, crossover period structure, any washout interval, and the statistical plan are 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.

Before assignment: Of 532 potential subjects that were screened, 374 were excluded as they did not meet the inclusion criteria, 45 refused to participate and 22 were no-shows (declined radiation therapy), thereby reducing our original target sample population to 91.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent to treat population consisted of randomized subjects who completed baseline questionnaire and had at least one dose of mouthwash

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 Comparison of Number of Participants With Grade 1 or 2 Oral Mucositis (GM-CSF and SS Groups Only), measured over From baseline to onset of mucositis, approximately 16 days.

The primary endpoint counts participants with recorded grade 1 or grade 2 oral mucositis. Grade 1 is described as mild irritation, possibly with slight pain not requiring analgesia. Grade 2 is described as moderate patchy mucositis that may produce inflammatory serosanguinitis discharge and may involve moderate pain requiring analgesia. The registry says the assessment uses the RTOG Acute Radiation Morbidity Scoring Criteria Oral Mucosa Assessment Scale, with grading based on a clinician’s judgment of anatomical changes, including the size and characteristics of ulceration, as well as symptoms.

The endpoint is useful for describing whether a participant experienced mucositis at the specified grades during the stated observation period. Counting participants with an event gives an incidence-style outcome for the comparison; it does not describe the duration, recurrence, or full severity course of mucositis. The registry’s timeframe is from baseline to onset of mucositis, approximately 16 days. The primary endpoint title specifies the GM-CSF and salt-and-soda groups only, and its population definition requires randomization, a completed baseline questionnaire, and at least one mouthwash dose.

Interpretation also depends on how the event was assessed and on which participants contributed data. The registry describes clinician grading but does not report assessor training, inter-rater agreement, or detailed handling of participants without a recorded assessment. Those details are not reported in the registry record.

7. Results Posted in the Registry

• Arm II: Salt & Soda Group (SS): 86.8% (46/53) versus 100.0% (28/28) on control; risk difference -13.2 percentage points (95% CI -22.3 to -4.1), risk ratio 0.87 (0.78–0.96); unadjusted two-proportion z-test p = 0.044.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Comparison of the Combined Mean Score on the Overall Quality of Life Questionnaires by Group

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Arm II: Salt & Soda Group (SS): mean 6.80 versus 6.91 on control; difference -0.11 (95% CI -0.93 to 0.71), standardized effect d = -0.07; z-test p = 0.793.
  • Arm III: Salt & Soda Switched to GM-CSF (SG): mean 6.77 versus 6.91 on control; difference -0.14 (95% CI -0.96 to 0.68), standardized effect d = -0.09; z-test p = 0.739.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Arm II: Salt & Soda Group (SS) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Arm III: Salt & Soda Switched to GM-CSF (SG) and the control. The interval gives the range of effects that remain compatible with the data.

Comparison of Combined Mean Score on the Karnofsky Performance Status Scale (KPS) by Group

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Arm II: Salt & Soda Group (SS): mean 82.93 versus 84.52 on control; difference -1.59 (95% CI -7.93 to 4.75), standardized effect d = -0.13; z-test p = 0.623.
  • Arm III: Salt & Soda Switched to GM-CSF (SG): mean 83.49 versus 84.52 on control; difference -1.03 (95% CI -7.73 to 5.67), standardized effect d = -0.08; z-test p = 0.763.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Arm II: Salt & Soda Group (SS) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Arm III: Salt & Soda Switched to GM-CSF (SG) and the control. The interval gives the range of effects that remain compatible with the data.

Comparison of Combined Mean Score on the Pain Questionnaire by Group

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Arm II: Salt & Soda Group (SS): mean 1.72 versus 2.15 on control; difference -0.43 (95% CI -1.70 to 0.84), standardized effect d = -0.18; z-test p = 0.506.
  • Arm III: Salt & Soda Switched to GM-CSF (SG): mean 2.15 versus 2.15 on control; difference 0.00 (95% CI -1.35 to 1.35), standardized effect d = 0.00; z-test p = 1.000.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Arm II: Salt & Soda Group (SS) and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Arm III: Salt & Soda Switched to GM-CSF (SG) and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover overall quality of life, Karnofsky performance status, and pain, with measurements described up to 3 months. The posted analyses compare the GM-CSF group with the salt-and-soda group and with the group that switched from salt-and-soda to GM-CSF. Across these outcomes, the unadjusted recalculations do not establish precise nonzero mean differences: the recalculated confidence intervals include zero. The registry’s own mixed models analyses are reported separately and remain the sponsor’s analyses of record.

These measures provide a broader view of patient-reported experience and performance than the primary mucositis event count, but they do not change what the primary endpoint measures. They also do not directly resolve the discrepancy between the primary recalculation and the sponsor-reported Fisher Exact analysis. The secondary results should be considered complementary context rather than confirmation or refutation of the primary comparison.

Several outcomes and group contrasts create opportunities for chance findings. An unadjusted secondary p-value does not account for the full set of tests, so it can overstate the strength of evidence when interpreted in isolation. The registry does not report whether multiplicity was addressed or whether secondary outcomes were arranged in a prespecified testing hierarchy. The published result pattern therefore warrants cautious, descriptive interpretation, without elevating an individual unadjusted comparison into a definitive claim.

Different measures also have different analysis counts, and the registry does not explain all reasons for those variations. Without fuller information on missing observations and model specifications, differences among secondary estimates should not be treated as evidence of a distinct pattern beyond what their uncertainty supports.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Comparison of the Combined Mean Score on the Overall Quality of Life Questionnaires by Group
• Comparison of Combined Mean Score on the Karnofsky Performance Status Scale (KPS) by Group
• Comparison of Combined Mean Score on the Pain Questionnaire by Group

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is a binary outcome: a participant either had a recorded grade 1 or grade 2 oral mucositis event during the stated period or did not. For a binary outcome, a risk difference compares the event proportions in two groups by subtracting one group’s proportion from the other. A value below zero for salt-and-soda relative to GM-CSF indicates a lower observed event proportion in the salt-and-soda group under that direction of comparison. The recalculated risk difference is -0.132. This is an unadjusted calculation from the posted counts, not a model-based adjusted estimate.

The confidence interval for the recalculated risk difference describes a range of values compatible with the data under the method’s assumptions. Its 95% interval is -0.223 to -0.041. An interval expresses uncertainty about the estimated difference; it does not provide the probability that a particular treatment effect is true. The unadjusted p-value of 0.044 evaluates compatibility with a null comparison under the recalculation’s assumptions. It should be read alongside the estimate, its interval, and the sponsor’s reported Fisher Exact result of 0.09. Because the record flags a discrepancy, the sponsor’s own analysis takes precedence.

Fisher’s exact test is a method for comparing categorical outcomes in a two-group setting. The registry reports that method for the primary comparison, but does not report further details such as the exact test options or whether the result was one-sided or two-sided. The recalculation and the reported analysis need not match if their methods, data handling, or analysis populations differ. The registry does not explain the reason for the discrepancy, so its cause is not reported in the registry record.

For the continuous secondary outcomes, the registry posts group means and standard deviations. A mean difference compares the average score in one group with the average in the reference group. The recalculated intervals describe uncertainty around those differences. The registry reports mixed models analyses for the secondary outcomes, which are the sponsor’s analyses of record. Mixed models can account for structured or repeated observations when the model and data support that approach; the precise model specification, covariance assumptions, and handling of incomplete repeated measurements are not reported in the registry record. A simple comparison of posted means does not reproduce all features of a mixed model.

The reported secondary analyses are group-level tests, while the displayed recalculations are simple unadjusted comparisons. An unadjusted p-value addresses one comparison at a time and does not account for the fact that several outcomes or contrasts may be examined. When many tests are considered, the chance of observing a small p-value by chance can increase. Whether the trial used a multiplicity adjustment or a hierarchy of outcomes is not reported in the registry record. Therefore, the secondary comparisons should be treated as supportive descriptions rather than as a set of confirmatory claims.

Sample-size illustrations are planning calculations based on the observed effects, not observed power and not the trial’s original design justification. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect to detect defined as a stated fraction of the observed effect. The illustrations show how planning requirements change when a smaller effect is targeted; they do not establish that the trial was designed around those assumptions. The trial’s own sample-size justification is not reported in the registry record.

The crossover label is relevant because observations may depend on treatment sequence and period as well as assigned group. For crossover data, within-participant relationships and possible period or carryover effects can matter. The registry does not report enough detail about sequence, period definitions, washout, or modeling to determine how those issues were handled. The primary endpoint’s stated comparison is between the GM-CSF and salt-and-soda groups, and it is not clear from the registry record whether that analysis incorporated crossover features.

Finally, analysis populations matter. The primary endpoint defines an intent-to-treat population using randomization, baseline questionnaire completion, and at least one mouthwash dose. That definition differs from a strict analysis of every randomized participant, because the listed requirements condition inclusion on post-randomization information. The registry does not report how participants with missing baseline questionnaires, no dose, or missing outcome assessments were handled. The reported arm counts should therefore be interpreted with the endpoint’s stated population definition in mind.

11. Confidence Intervals and Precision

The primary recalculated estimate is a risk difference of -0.132, with a 95% confidence interval from -0.223 to -0.041. The interval conveys uncertainty around the unadjusted difference derived from the posted counts. It is not a prediction interval for future participants and does not describe the range of individual outcomes. Its interpretation depends on the comparison and method used; it does not resolve the discrepancy with the sponsor-reported Fisher Exact p-value of 0.09.

Confidence intervals are useful because they show both the direction and the range of uncertainty around an estimate. A narrow interval would indicate greater precision under the method’s assumptions, while a wider interval would indicate less precision. The interval excludes zero for the recalculated risk difference, but the sponsor’s reported analysis differs. The sponsor’s analysis takes precedence in light of the recorded discrepancy, and no conclusion should be based on the recalculation alone.

For the continuous secondary comparisons, the recalculated confidence intervals include zero. That pattern indicates that the posted summary statistics do not pin down a nonzero unadjusted mean difference with precision under those calculations. The sponsor’s mixed models analyses are reported separately, and their results should not be conflated with these simpler recalculations. The registry does not provide confidence intervals for the reported mixed models analyses.

Precision is also affected by the available analysis counts and the endpoint definitions. The primary comparison has arm-specific counts of 28 and 53, and secondary outcome counts vary by measure and group. The registry does not report detailed missing-data procedures or explain the differences between enrollment and each analysis population. Those gaps make it difficult to assess whether uncertainty from missing observations or selection into the analysis is fully reflected in the posted intervals.

12. Analyses Reported by the Sponsor

The registry reports a Fisher Exact analysis for the primary comparison, with p = 0.09. A simple unadjusted recalculation from the posted counts gives a risk difference of -0.132, a 95% confidence interval from -0.223 to -0.041, and p = 0.044. These are not the same inferential results. The registry marks this comparison as discrepant, so the sponsor’s own analysis takes precedence.

The recalculation uses the posted arm counts and compares the two event proportions without adjustment. It does not reproduce any unreported features of the sponsor’s analysis, and the registry does not explain why the results differ. Possible details that could affect results, such as exact analysis-set construction or test options, are not reported in the registry record; no specific cause can be assigned from the available information.

The difference matters because the p-value from the sponsor’s analysis and the recalculated estimate answer related but not necessarily identical analytic questions. The recalculated confidence interval is useful as a transparent description of a basic comparison of the posted summaries. It should not replace the registry-reported test or be presented as the trial’s definitive analysis. The result is best read by keeping the reported analysis, the descriptive counts, and the recalculation distinct.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and several secondary outcomes, with multiple group comparisons for the continuous outcomes. When researchers assess multiple outcomes or contrasts, the chance of at least one apparently unusual result can rise even if no underlying group difference exists. Adjustments, prespecified outcome hierarchies, or other strategies can address this issue, but whether any such strategy was planned or used is not reported in the registry record. The unadjusted recalculations should consequently be interpreted as individual comparisons, not as multiplicity-controlled evidence.

The registry does not report interim analyses, stopping rules, or a detailed statistical analysis plan. It also does not report the trial’s sample-size justification. The posted sample-size illustrations are planning calculations based on observed effects and stated assumptions; they are not observed power and do not document the original design rationale.

The primary population definition includes randomized participants who completed a baseline questionnaire and received at least one dose of mouthwash. The record does not explain how participants failing those conditions were handled, nor how missing outcome assessments were treated. The secondary analysis counts vary by outcome and arm, but the reasons for those differences are not reported in the registry record. No imputation approach, sensitivity analysis, or missing-data assumption is described.

The trial is labeled as crossover, yet the registry does not give sufficient detail to assess treatment sequence, period effects, carryover, or how repeated observations were incorporated into the analysis. The secondary analyses are identified as mixed models, but model details are not reported. These omissions limit assessment of whether the analyses accounted for the design and missingness in the intended way.

The registry’s stated primary population is not identical in wording to an analysis of all randomized participants, because it requires baseline completion and at least one dose. The available record does not establish how that operational definition was applied in practice. These limitations affect interpretation of both the point estimates and the reach of any inference beyond participants contributing to the posted analyses.

14. 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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: other (3.6%), lower gastrointestinal hemorrhage (3.6%), thromboembolic event (3.6%), allergic reaction (3.6%), mucositis oral (3.4%).

The safety record lists serious adverse events and other adverse events separately for the three groups. Serious events are reported for 1 participant in the GM-CSF group, 2 in the salt-and-soda group, and 2 in the group that switched to GM-CSF. The corresponding serious-event populations at risk are recorded as 29, 28, and 28. The record also lists 1 death in the GM-CSF group, none in the salt-and-soda group, and none in the switched group.

These figures are descriptive counts. The registry does not provide a statistical comparison, event timing, attribution to study treatment, or detailed definitions for the listed categories. The serious-event terms include “Other,” lower gastrointestinal hemorrhage, thromboembolic event, allergic reaction, and oral mucositis; each is listed at 3.6%, except oral mucositis at 3.4%. The record does not specify the denominators or explain whether terms overlap, so those percentages should not be converted into additional counts or treated as mutually exclusive categories.

Other adverse events were recorded in 18 of 29 at risk in the GM-CSF group, 26 of 28 in the salt-and-soda group, and 27 of 28 in the switched group. These totals do not establish comparative safety by themselves. Differences in follow-up, exposure, event attribution, and reporting are not described sufficiently in the registry record to support causal interpretation.

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

The posted primary counts describe recorded grade 1 or grade 2 oral mucositis in the specified analysis population and timeframe. The simple recalculation suggests a lower event proportion in the salt-and-soda group than in the GM-CSF group, but the registry-reported Fisher Exact result differs. Because the discrepancy is flagged, the sponsor’s analysis takes precedence; the recalculated estimate is supplementary and should be interpreted as an unadjusted comparison of posted summaries.

The result does not establish that either mouthwash prevents or treats mucositis. The registry does not provide a complete account of the analysis plan, crossover details, missing-data handling, or the reason for the discrepancy. Nor does the primary endpoint capture every aspect of mucositis, such as duration or the full course of severity. The secondary measures provide additional information but have their own uncertainty, analysis counts, and multiplicity concerns.

A confidence interval expresses uncertainty in an estimated comparison under specified assumptions. It is not a guarantee that the true effect lies in the interval, and a p-value is not a measure of the size or importance of an effect. The posted results should therefore be read by considering the event counts, the sponsor’s reported analysis, the recalculated estimate, and the limits of the registry record together.

The conclusions apply most directly to the participants represented in each analysis and to the recorded outcome definitions. The primary analysis population is defined by randomization, baseline questionnaire completion, and at least one mouthwash dose. The registry does not report whether that population differs meaningfully from all randomized participants or how missing assessments were addressed. Generalization beyond the analyzed groups and stated setting requires information not reported in the registry record.

16. Why This Trial Matters Statistically

This trial illustrates why statistical interpretation should distinguish a raw event count, an estimated difference, a confidence interval, and a reported hypothesis test. Those quantities describe related aspects of the data, but they are not interchangeable. A simple recalculation from arm summaries can be informative and transparent while still differing from the sponsor’s analysis. When the registry flags such a discrepancy, the sponsor’s analysis takes precedence rather than allowing the recalculation to stand as the definitive result.

The record also illustrates how trial design labels and endpoint populations shape interpretation. A crossover label raises questions about sequence, period, and repeated measurements, while the primary outcome is presented as a comparison of two groups. Without additional design and model details, it is not possible to evaluate how those features were handled. Explicit population definitions and arm-specific analysis counts are essential context, particularly when outcome counts differ from enrollment.

Finally, the secondary outcomes and safety figures show why a collection of endpoints should not be read as a collection of independent confirmations. Multiple comparisons, variable analysis counts, and incomplete reporting can complicate interpretation. The registry record supports careful description of what was recorded and how precisely the comparisons were estimated, while leaving important elements of the planned analysis and data handling unreported.

17. 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 details are not reported in the registry record. The crossover sequence, period structure, washout arrangements, and handling of possible carryover or period effects are not described. The arm descriptions also end with incomplete treatment-continuation text, leaving the exact duration and subsequent treatment details unclear. These omissions matter when interpreting comparisons in a study identified as crossover.

The primary outcome is limited to grade 1 or grade 2 mucositis, assessed by clinician judgment under the stated RTOG scale. The registry does not report assessor training, agreement between assessors, or procedures for resolving uncertain grades. The endpoint timeframe is described as baseline to onset, approximately 16 days, but further detail about assessment schedules and outcome handling is not reported.

The reported primary population requires baseline questionnaire completion and at least one dose, and the registry does not explain exclusions or missing assessments. Outcome analysis counts vary, with no detailed account of participant flow for each endpoint. The reasons for those variations and the approach to incomplete data are not reported in the registry record.

There is a discrepancy between the sponsor-reported primary Fisher Exact p-value and the simple unadjusted recalculation. The registry does not explain its source. Secondary analyses are reported as mixed models, but model specification and multiplicity handling are not described. The trial’s sample-size justification and interim analysis plan are also not reported. These limitations prevent a full audit of the analysis strategy and constrain conclusions to cautious interpretation of the posted results.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: GM-CSF Mouthwash for Preventing and Treating Mucositis in Patients Undergoing Radiation Therapy for Head and Neck Cancer — NCT00293462.
• Registry record last updated 2021-07-30; status last verified 2021-07.
• Results first posted 2013-05-16.
• 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.

21. Record Summary

The registry describes a completed randomized, double-masked Phase 3 trial with a crossover model and enrollment of 91. The study compared GM-CSF mouthwash and salt-and-soda mouthwash for mucositis during radiation therapy for head and neck cancer, and also listed a group that switched from salt-and-soda to GM-CSF. The primary endpoint counted participants with grade 1 or grade 2 oral mucositis in the GM-CSF and salt-and-soda groups.

The primary posted counts were 28 of 28 in the GM-CSF group and 46 of 53 in the salt-and-soda group. The registry reports a Fisher Exact p-value of 0.09, while the unadjusted recalculation differs; the sponsor’s analysis takes precedence. Secondary outcomes cover quality of life, Karnofsky performance status, and pain. The record also lists adverse-event counts. Important details about the crossover, missing data, multiplicity, and analysis plan are not reported in the registry record.