Doxepin Hydrochloride in Treating Esophageal Pain in Patients With Thoracic Cancer Receiving Radiation Therapy to the Thorax With or Without Chemotherapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Doxepin Hydrochloride in Treating Esophageal Pain in Patients With Thoracic Cancer Receiving Radiation Therapy to the Thorax With or Without Chemotherapy (NCT02062632), a randomized phase 2 study. The primary outcome is Change in Mouth Pain as Measured by Average Area Under the Curve Per Assessment.

2. Clinical Question

The study was designed as a randomized, double-blind pilot, with doxepin compared with placebo in patients receiving thoracic radiotherapy. Its stated clinical focus was pain associated with radiation-related esophageal irritation, although the primary outcome title refers to mouth pain. The record describes doxepin as a tricyclic antidepressant and notes the rationale that it had been studied for mouth pain during radiation therapy. Those statements describe the study rationale, not evidence of benefit in this trial.

3. Trial Design

This was a randomized, double-blind, phase 2 pilot study sponsored by Mayo Clinic. The official title describes a two-arm crossover study of doxepin versus placebo. The arm descriptions also specify treatment on day 1 followed by crossover to the other arm on day 3. However, the registry’s model field says “parallel.” This inconsistency is not resolved in the registry record, so the design should be understood from both descriptions rather than treating the model label as a complete account of the treatment sequence.

The planned population was patients with cancer located in the chest area who were receiving radiation therapy to the thorax, with or without chemotherapy. The registry lists a range of cancer conditions, including esophageal carcinoma, lung cancers, lymphoma, mesothelioma and other thoracic or nearby cancers. The record says 5 participants were accrued, but only 3 provided data. It further states that confidentiality prevents intervention-specific reporting because only 1 patient was accrued and completed the study on 1 arm. The second crossover was optional, and none chose it.

The trial began on 2014-04-14, and primary completion is recorded as 2015-05-29. Its status is terminated. The registry lists quality-of-life assessment and questionnaire administration among the interventions, alongside doxepin and placebo. Secondary outcomes concerned grade 3 or higher adverse events using CTCAE, alternative analgesic use, and preference for continued therapy at the initial dose and crossover. No protocol explanation for termination, detailed recruitment history, eligibility criteria, or reason for the optional second crossover is reported in the registry record.

Although the crossover description suggests that participants could receive both treatments in sequence, the record says that primary results are limited to the first study period. That restriction matters: the reported primary endpoint is not presented as a complete within-participant comparison of both treatments. The registry record does not report an analysis plan explaining how randomization, treatment sequence, period effects, or possible carryover were to be handled.

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: 5 were accrued but only 3 provided data. Since only 1 patient was accrued \& completed the study on 1 arm, patient confidentiality prevents the reporting of results per intervention. 2nd crossover was optional; none chose to crossover. The study results are only reported for the 1st study period (before crossover) \& only for placebo intervention.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: 5 were accrued but only 3 provided data. Since only 1 patient was accrued \& completed the study on 1 arm, patient confidentiality prevents the reporting of results per intervention. 2nd crossover was optional; none chose to crossover. The study results are only reported for the 1st study period (before crossover) \& only for placebo intervention.

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: 5 were accrued but only 3 provided data. Since only 1 patient was accrued \& completed the study on 1 arm, patient confidentiality prevents the reporting of results per intervention. 2nd crossover was optional; none chose to crossover. The study results are only reported for the 1st study period (before crossover) \& only for placebo intervention.

6. Primary Endpoint

The registry lists the primary outcome as Change in Mouth Pain as Measured by Average Area Under the Curve Per Assessment, measured over Baseline and 5, 15, 30, 60, 120, and 240 minutes after treatment on day 1.

The primary outcome is titled “Change in Mouth Pain as Measured by Average Area Under the Curve Per Assessment.” It describes an average area-under-the-curve measure of pain across assessments at baseline and 5, 15, 30, 60, 120, and 240 minutes after treatment on day 1. Scores are on a 0–100 scale, where 100 represents a better quality-of-life outcome. The registry says the area-under-the-curve calculation uses the order of assessments rather than elapsed minutes, gives the assessment scores equal weights, and adjusts the area under the curve for baseline pain. It therefore describes the resulting scale as the same as the original pain-score scale.

In general, combining repeated assessments into an area-under-the-curve summary can describe the pattern of scores over an observation period in a single measure. Here, the stated approach gives each assessment equal weight and does not make the measure depend on the time interval between assessments. That choice makes the endpoint a summary of the recorded sequence, rather than a measure weighted by duration at each pain level.

The outcome title refers to mouth pain, while the study’s clinical rationale and summary focus on esophageal pain. The registry record does not explain this wording difference. It also does not report intervention-specific endpoint results, so the magnitude, direction and precision of any treatment contrast cannot be assessed from the posted results.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does randomization contribute?

In general, randomization assigns treatment by chance rather than by a participant’s or clinician’s choice. When a study has enough participants and follows its allocation procedure, this can reduce systematic differences between treatment groups at the outset. Randomization does not guarantee balance in a small study, and it cannot make missing outcome data disappear. This registry identifies the study as randomized, but does not report the allocation procedure, sequence-generation details, or how assignment was concealed.

What does double masking mean?

The record labels the study double-blind. In general, masking is intended to limit the influence of treatment knowledge on participant reports, clinical care or outcome assessment. That is particularly relevant when the endpoint is based on participant-reported pain. The record does not specify which roles were masked, how the placebo was matched, whether masking was assessed, or how treatment identities would have been handled during analysis. Those details are not reported in the registry record.

How should the crossover description be read?

A crossover design generally exposes a participant to more than one treatment in sequence, allowing comparisons that can account for stable differences between participants. Such designs also require attention to period effects, treatment order and possible carryover from an earlier treatment. The official title and arm descriptions describe a crossover, with treatment on day 1 and a crossover on day 3. Yet the registry model field says “parallel.” The record does not explain this discrepancy or provide an analysis plan for sequence or period effects. It states that the second crossover was optional and none chose it, and that only first-period results are reported. Thus, the posted information does not support reading the results as a completed within-participant comparison.

What does the area-under-the-curve summary represent?

In general, an area-under-the-curve measure combines repeated observations across an assessment schedule. The registry describes an average area under the curve per assessment for pain, adjusted by baseline pain. It says that assessment order, rather than the actual number of minutes after treatment, is used in the calculation, and that every assessment receives equal weight. It also says the summary remains on the original pain-score scale. This means the measure summarizes the recorded scores across the sequence of assessments, not a time-weighted total based on the duration between observations. The endpoint description identifies the scoring direction as 100 meaning a better quality-of-life outcome. Because the title says “change in mouth pain” while the description frames the scale as a quality-of-life outcome, the intended interpretation of score direction is not fully clarified beyond that description.

Why do effect estimates need uncertainty measures?

In general, an effect estimate describes the size and direction of a difference or change, while a confidence interval conveys the range of values compatible with the data under the statistical model. A narrow interval generally reflects greater precision than a wide interval. A p-value addresses compatibility between data and a specified null hypothesis under a model; it does not quantify clinical importance, and it is not a substitute for an effect estimate and interval. Here, the registry record does not report a primary-outcome estimate, confidence interval or p-value, so these aspects of inference cannot be assessed.

Why does the amount of outcome data matter?

Missing outcome data can reduce precision and may introduce bias if the participants with available measurements differ systematically from those without them. In general, the reasons for missingness and the relationship between missingness and outcomes affect how confidently results can be interpreted. The registry says that 5 were accrued but only 3 provided data. It does not report details about which measurements were missing, why participants did not provide data, or how missing data were handled. The confidentiality explanation further limits access to treatment-specific results. No particular missing-data method should be assumed.

What does the record say about the statistical plan?

The registry lists 7 outcome measures and 0 analyses. It does not report a statistical analysis plan, a primary contrast, covariate adjustment, handling of the crossover, or a method for estimating uncertainty. The sample-size assumptions describe a general planning illustration based on an observed effect, but the record does not provide a primary comparison suitable for such an illustration. No planning estimate is shown for this outcome because the posted primary outcome does not have two reported arms. Any general sample-size planning illustration would be a design aid, not an observed treatment effect or evidence of the study’s achieved power. A study’s statistical plan and justification are distinct from analyses that can be reconstructed from posted summary data; neither is available for the primary treatment contrast in this record.

In general, repeated outcomes and multiple endpoints can create multiplicity: testing many hypotheses can raise the chance of at least one apparently notable result by chance. An analysis plan may specify which comparisons are primary and how other tests are treated. The record lists secondary outcomes but does not report a multiplicity strategy or a hierarchy of tests. It is therefore not possible to determine whether any adjustment was planned or performed. These are explanations of general statistical concepts; the registry does not report that specific procedures were used in this study.

10. Multiplicity, Missing Data and Analysis Populations

The record lists a primary outcome and secondary outcomes concerning grade 3 or higher adverse events, alternative analgesic use, and patient preference for continued therapy at the initial dose and crossover. It does not report an analysis plan describing which outcomes or comparisons were primary for statistical testing, or whether any adjustment for multiple comparisons was planned. In general, testing multiple outcomes can increase the chance of at least one apparently unusual finding by chance. Without a reported strategy, the inferential status of multiple outcome tests cannot be evaluated. No primary-outcome analysis is reported, and the record lists 0 analyses.

For outcome availability, the registry states that 5 participants were accrued but only 3 provided data. It says that confidentiality prevents intervention-specific reporting because only 1 patient was accrued and completed the study on 1 arm. It also says the second crossover was optional and none chose it. These statements describe substantial limits on the outcome information available for treatment comparisons, but do not specify which individual measurements are missing or why data were unavailable. Missing-data assumptions and methods are not reported in the registry record.

The analysis population is not defined in a way that permits a clear account of who was included in a primary treatment comparison. The record does not state whether analyses were intended to follow randomized assignment, include only participants with observed outcomes, or use another population definition. It also does not describe how the optional crossover or incomplete treatment sequences would be handled. Given the lack of intervention-specific primary results, no analysis population or multiplicity adjustment can be inferred from the posted record.

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 registry lists a placebo group with 3 participants at risk, 0 serious adverse events, 2 other adverse events and 0 deaths. The record supplies no serious adverse-event terms for this group. These counts should be read as the figures recorded for the listed placebo group, not as a comparison between doxepin and placebo. No doxepin-group adverse-event figures are included in the reported data, so a relative safety assessment is not possible.

In general, an adverse-event count is easier to interpret when its definition, observation period, severity, and the number of participants assessed are clear. The record identifies the serious-event outcome as grade 3 or higher events using CTCAE among the secondary outcomes, but the adverse-event figures do not provide event descriptions or a detailed observation period. The label for the other adverse events is not further defined in the record. The available counts therefore do not establish the nature, timing or clinical significance of those events.

Zero recorded serious adverse events in the listed group does not establish that serious events could not occur, and the other-event count cannot be attributed to a particular treatment effect from these data. The record’s limited participation and lack of treatment-specific reporting further constrain interpretation. No safety conclusion comparing the treatments is supported by the posted figures.

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

This record describes a randomized, double-blind pilot study of doxepin hydrochloride versus placebo for pain during thoracic radiotherapy. It documents the treatment instructions, the planned assessment schedule and the primary outcome’s area-under-the-curve approach. It also records that the study was terminated, that 5 participants were accrued but only 3 provided data, and that intervention-specific primary results were withheld for confidentiality reasons.

What the record does not establish is the size or direction of a doxepin–placebo difference in the primary outcome. There is no reported effect estimate, uncertainty interval or p-value for that comparison. The absence of those results is not a finding of equivalence, and it does not show that either treatment was beneficial or ineffective. There is likewise no reported treatment-specific safety comparison. The placebo adverse-event counts cannot stand in for a comparison with doxepin.

The clinical focus and endpoint wording do not fully align: the study summary concerns esophageal pain, while the primary outcome title refers to mouth pain. The endpoint description gives a scale direction and explains how repeated assessments are combined, but the registry does not clarify the wording difference. The design description also contains a mismatch between the crossover title and arm instructions and the parallel model label. These ambiguities add to the limits on interpretation.

In general, pilot studies can inform future research, but this record does not report a sample-size justification or a detailed account of its feasibility objectives. The small amount of available outcome information and the lack of a reported treatment comparison mean that the posted record is descriptive rather than conclusive about comparative effects. Any claim beyond the recorded design and counts would go beyond what the registry record establishes.

13. Why This Trial Matters Statistically

This trial is statistically notable because the registry exposes a gap between the planned comparison and the results that can be interpreted. The study is described as randomized and double-blind, with doxepin and placebo, and its official title and arm descriptions indicate a crossover sequence. Yet the model field labels it parallel, and the posted primary outcome lacks results for both interventions. The record explicitly attributes the absence of intervention-specific reporting to confidentiality and describes only first-period results.

These details illustrate why design labels alone are not enough to understand an analysis. In general, a crossover comparison depends on which periods and treatments participants complete, and on how sequence and period effects are handled. This record says that none chose the optional second crossover and does not report an analysis plan. Readers therefore cannot reconstruct a within-participant treatment comparison or determine how the stated design was reflected in statistical analysis.

The record also illustrates the distinction between collecting an outcome and reporting an analyzable treatment contrast. A repeated pain score can be summarized using an area-under-the-curve measure, but the method does not by itself supply an effect estimate or indicate its precision. With no intervention-specific primary outcome results, readers cannot use the endpoint definition to determine whether the treatments differed.

Finally, the discrepancy between the esophageal-pain rationale and the mouth-pain endpoint title, along with the limited adverse-event reporting, shows why careful reading of outcome definitions and denominators matters. The statistical contribution of this record is chiefly to document what was planned and what was not available for interpretation, rather than to establish a comparative clinical effect.

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 most direct limitation is the absence of intervention-specific primary outcome results. The registry says that 5 participants were accrued but only 3 provided data, and that confidentiality prevents reporting results per intervention because only 1 patient was accrued and completed the study on 1 arm. The primary outcome is reported only for the first study period, and the second crossover was optional; none chose to crossover. These circumstances leave no reported treatment contrast for the primary endpoint.

The design is not consistently characterized. The official title and arm descriptions identify a crossover sequence, while the registry model field says parallel. No statistical plan is reported to clarify how the design was analyzed, including handling of sequence, period effects or incomplete crossover. The record also does not explain why the primary outcome title refers to mouth pain while the clinical rationale centers on esophageal pain.

Important details about missing outcomes, participant flow, analysis populations, eligibility criteria, allocation procedures, masking roles and the reason for termination are not reported in the registry record. Nor are a sample-size justification, inferential results, estimates of uncertainty or a multiplicity plan reported. The safety figures are limited to a listed placebo group and do not provide a treatment comparison or detailed event descriptions. Together, these reporting gaps prevent assessment of effect magnitude, precision, robustness and comparative safety.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Doxepin Hydrochloride in Treating Esophageal Pain in Patients With Thoracic Cancer Receiving Radiation Therapy to the Thorax With or Without Chemotherapy — NCT02062632.
• Registry record last updated 2019-10-01; status last verified 2018-11.
• Results first posted 2019-10-01.
• 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

NCT02062632 is a randomized, double-blind, phase 2 pilot study sponsored by Mayo Clinic, evaluating doxepin hydrochloride versus placebo for pain during thoracic radiotherapy. The official title and arm descriptions describe a crossover, although the registry model field says parallel. The primary endpoint is an average area-under-the-curve pain measure using assessments after treatment on day 1, with baseline adjustment and a 0–100 scale where 100 indicates a better quality-of-life outcome.

The record reports 5 participants accrued and 3 providing data. It says results are limited to the first study period and that confidentiality prevents reporting by intervention; none chose the optional second crossover. No primary treatment comparison or inferential analysis is reported. The adverse-event figures list a placebo group with 3 at risk, 0 serious events, 2 other events and 0 deaths. The study status is terminated, and the reason for termination is not reported in the registry record.