Tolerance Following Peanut Oral Immunotherapy: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Tolerance Following Peanut Oral Immunotherapy (NCT01750879), a randomized phase 1, phase 2 study. The primary outcome is Tolerance, Partial Tolerance, or Treatment Failure.

2. Clinical Question

The study’s stated unifying objective was to examine whether clinical tolerance after peanut oral immunotherapy was associated with expansion of a specific regulatory T-cell subset. Its stated hypothesis was that induction of regulatory T cells would be associated with clinical tolerance. The planned comparison was between participants classified as tolerant or partially tolerant and those classified as treatment failures, using change from baseline in induced regulatory T cells as a frequency of total CD4 T cells during active treatment.

That biological question is distinct from whether participants met the clinical categories at the challenge assessments. A registry description of an intended association does not itself establish that the association was tested or observed. The registry record lists no analyses, and outcome counts for the primary clinical classification are not reported in the registry record. It therefore does not establish how many participants met each category or whether T-cell changes differed across those categories.

3. Trial Design

The study is listed as phase 1 and phase 2, randomized, triple-masked, and parallel-group. The interventions were peanut flour and oat flour, each described as oral immunotherapy. The registry reports enrollment of 41 and lists peanut allergy as the condition. It does not give further details about eligibility, the randomization procedure, allocation concealment, or which study roles were masked. The term “triple-masked” identifies the registry’s masking classification, but the record does not specify the masked parties.

The study began in 2013-08, had primary completion on 2017-07-11, and is marked completed. The primary outcome timeframe is described as an average of 515 days from DBPCFC1 to DBPCFC3. These labels identify the study timeline and outcome window as recorded; they do not describe the number or timing of visits for each participant.

The primary outcome is a participant count divided conceptually into tolerance, partial tolerance, and treatment failure, with two failure descriptions: non-desensitized and withdrawal. Classification uses peanut protein amounts and eliciting-dose thresholds at DBPCFC1, DBPCFC2, and DBPCFC3. The registry also lists secondary clinical outcomes for tolerance, desensitization, and safety, along with mechanistic outcomes involving T-cell clonal diversity, eliciting doses, peanut allergen-specific IgG4, and gene expression.

The record does not report the planned sample-size justification, the statistical analysis plan, or details of how the challenge assessments were conducted. It also does not give the number assigned to each arm for the primary outcome. The safety record separately lists participants at risk in each arm, but those safety denominators should not be treated as the primary-outcome analysis population.

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: 113 screened, 72 excluded (3 Lost to follow-up pre-randomization; 8 Re-screened following protocol change; 21 Screen Fail no objective symptoms on DBPCF1; 29 Screen Fail prior to DBPCFC; and 11 Voluntary withdrawal pre-randomization)

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 Tolerance, Partial Tolerance, or Treatment Failure, measured over Average 515 days from DBPCFC1 to DBPCFC3.

The primary endpoint classifies participants into tolerance, partial tolerance, or treatment failure using challenge results and treatment milestones. Tolerance is defined as ingestion of 4430 mg of peanut protein at DBPCFC3 without symptoms. Partial tolerance requires an eliciting dose at DBPCFC3 below 4430 mg but at least 430 mg, and more than 10-fold above the eliciting dose at DBPCFC1. The non-desensitized treatment-failure criteria include failure to reach the minimum maintenance dose of 600 mg by 12 months, an eliciting dose below 1443 mg at DBPCFC2, or an eliciting dose at DBPCFC3 below 443 mg or less than 10-fold above the baseline eliciting dose. Treatment failure also includes withdrawal; further criteria for that category are not reported in the registry record.

These categories are intended to distinguish different levels of response according to challenge thresholds and treatment milestones rather than to report a single continuous measurement. Such categories can make a clinical outcome easier to interpret in terms of prespecified states, while making the result dependent on the cutoffs and definitions chosen. The registry record does not report why these particular thresholds were selected or how withdrawals were handled in the final classification.

The timeframe is an average of 515 days from DBPCFC1 to DBPCFC3. The record does not report the number of participants meeting each category, nor does it report a statistical comparison of those categories. Consequently, the endpoint definition is available, but its observed distribution and precision are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does the primary outcome represent?

The endpoint is a categorical classification of participants using clinical challenge results and treatment milestones. The categories are tolerance, partial tolerance, and treatment failure. In general, a categorical endpoint sorts observations into defined groups rather than preserving every measured value as a continuous score. The categories can be useful for describing clinically distinct states, but the resulting interpretation depends on how the categories are defined and on how participants near a cutoff are classified.

Here, the registry’s definitions use challenge assessments identified as DBPCFC1, DBPCFC2, and DBPCFC3. The criteria refer to peanut protein amounts, eliciting doses, symptoms, a maintenance-dose threshold, and change relative to the earlier eliciting dose. These are operational definitions: they explain how the registry says the categories are assigned. The record does not describe the rationale for selecting the thresholds, the handling of missing challenge results, or how a participant meeting more than one stated criterion was classified.

What does randomization contribute?

Random assignment is intended, in general, to make treatment groups comparable at the point of assignment with respect to measured and unmeasured characteristics, apart from chance variation. That design can support a treatment comparison when outcomes are collected and analyzed appropriately. It does not guarantee that groups are identical, remove all possible sources of bias, or resolve problems caused by missing outcomes or post-assignment exclusions. The registry lists the study as randomized and parallel-group, but does not report the sequence-generation method, concealment process, or analysis population.

The parallel-group designation means the record describes separate intervention arms rather than a crossover comparison. Triple masking is also recorded, but the record does not say which roles were masked or how masking was maintained. Those omissions limit the ability to assess potential influences on challenge assessment, treatment delivery, or outcome classification.

How should estimates and uncertainty be read?

In general, a treatment-effect estimate summarizes the contrast being studied, while a confidence interval communicates uncertainty around that estimate under the statistical model and analysis assumptions. A narrow interval indicates greater precision than a wide interval, but neither interval width nor a p-value alone determines the clinical importance of an effect. A p-value, when reported, addresses compatibility of the observed data with a specified null model; it does not measure the probability that a treatment works or that the null hypothesis is true.

No primary-outcome estimate, confidence interval, or p-value is reported in the registry record. The record therefore does not allow readers to assess the magnitude or precision of an arm comparison. Because the endpoint has several categories and the record does not present an analysis reducing them to one comparison, it would also be inappropriate to invent a single contrast from the definitions.

What do the planning assumptions mean?

The registry describes sample-size planning illustrations based on an observed effect, using a two-sided significance level of 0.05, equal allocation between arms, and a normal approximation. It says the effect to detect is a stated fraction of the observed effect. Such calculations are planning illustrations, not evidence about the results of the trial. They are not observed power and should not be interpreted as a probability that a study finding is correct.

For planning illustrations other than time-to-event outcomes, the registry’s description refers to participants per arm for 80% power or other stated power levels. For time-to-event outcomes, it describes the total number of events. These distinctions are general descriptions of how planning calculations are expressed. The primary endpoint is a categorical participant-count outcome, and the registry record does not report a protocol sample-size justification or explain how the trial’s enrollment was chosen. A sample-size illustration based on an observed effect does not substitute for that design rationale.

Why can over-dispersion matter?

In general, over-dispersion occurs when observed variability is greater than a statistical model expects. It can make standard errors too small if the model does not account for that extra variability, which in turn can make confidence intervals too narrow and tests appear more decisive than warranted. The registry does not report an over-dispersion assessment or a model for the primary endpoint. Because the outcome is described as participant counts in multiple categories and no analysis is reported, there is no basis to determine whether over-dispersion was relevant or how it was handled.

The registry lists zero analyses. That is a statement about analyses recorded in the registry record, not proof that no analysis was conducted elsewhere. No inferential method, covariate adjustment, handling of repeated measures, or approach to missing data is reported. Statistical interpretation must therefore stop at the design, outcome definition, and reported safety counts.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists several secondary outcomes, including clinical outcomes for tolerance, desensitization, and safety, as well as mechanistic measures involving T-cell clonal diversity, eliciting doses, peanut allergen-specific IgG4, and gene expression. Testing multiple outcomes can increase the chance of at least one apparently notable result by chance when many tests are performed. In general, multiplicity can be addressed by prespecifying a hierarchy, adjusting inference, or clearly treating analyses as exploratory. The registry record does not report a multiplicity strategy, a hierarchy of hypotheses, or which outcomes were intended to support confirmatory claims.

The record lists zero analyses. This does not establish that analyses were not performed; it indicates that no analyses are recorded there. Statistical methods for the primary outcome and secondary outcomes are not reported in the registry record.

Missing-data procedures are not reported. The record does not say how withdrawals, incomplete challenge assessments, or unavailable mechanistic measurements were handled in analyses. It also does not identify an intention-to-treat, per-protocol, or other analysis population for the primary endpoint. The withdrawal category appears in the outcome definition, but the record does not explain how withdrawal cases were counted or how that category was applied in the final results.

These omissions matter because the endpoint depends on assessments at multiple time points and includes thresholds relative to earlier eliciting doses. In general, missing one assessment can affect whether a participant can be assigned to a category or evaluated against a change criterion. Without details on imputation, exclusions, or denominators, readers cannot judge whether the reported population—if results were available—would represent all randomized participants or a selected subset. The registry record does not report those details or the outcome counts needed to evaluate the issue.

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 serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: suicidal ideation, unrelated (3.3%).

The safety record lists Peanut Flour with 30 participants at risk and 1 with a serious adverse event; it lists Oat Flour with 11 participants at risk and 0 with a serious adverse event. Deaths are listed as 0 in each arm. These are the registry’s arm-specific safety figures and should be read separately from the primary efficacy endpoint. The record does not provide event-level descriptions or explain the observation period or criteria used for each safety count.

The record lists other adverse events affecting 30 participants at risk in the Peanut Flour arm and 11 participants at risk in the Oat Flour arm. These figures do not show how many events occurred, whether participants had more than one event, or the event types. The record’s top serious-event term is “Suicidal Ideation, unrelated,” listed as 3.3%. The label should be retained as recorded; the record does not provide further context for that term.

Safety counts are descriptive. They do not by themselves establish that an intervention caused or prevented an event, and the arm denominators are not necessarily the analysis populations for the primary outcome. No comparative safety analysis or uncertainty estimates are reported in the registry record. The small counts and missing event-level context also limit interpretation of differences between arms.

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

The registry record is useful for identifying the study question, design, outcome definitions, and recorded safety counts. It documents a randomized, triple-masked, parallel-group comparison of peanut flour and oat flour oral immunotherapy in people with peanut allergy. It also describes a clinical classification based on challenge responses and treatment milestones, alongside a hypothesis about regulatory T-cell induction.

What the record does not establish is whether participants in one arm were more likely to meet a particular primary-outcome category, how precisely any arm difference was estimated, or whether the proposed T-cell association was observed. The primary outcome counts and analyses are not reported in the registry record. Nor are the mechanistic results, inferential estimates, or uncertainty intervals. The trial’s completed status is not itself evidence of a result.

The endpoint definitions also do not prove that a clinical state occurred in any particular participant. They specify the criteria used to classify outcomes. Without reported counts and analysis details, readers cannot determine how often each criterion was met, how withdrawals affected classification, or how missing assessments were handled.

In general, a randomized design can support comparisons when allocation and follow-up are adequately implemented and when the analysis is appropriate to the question. The registry provides only limited information about those processes. It does not report the randomization procedure, masked roles, sample-size justification, analysis populations, or missing-data methods. Those gaps do not invalidate the study, but they restrict what can be concluded from this record.

The appropriate reading is therefore descriptive and cautious: the record states what investigators aimed to assess and how the main clinical categories were defined, and it lists adverse-event counts by arm. It does not provide enough primary or mechanistic result information to draw conclusions about comparative outcomes or the proposed biological relationship.

13. Why This Trial Matters Statistically

This trial is statistically notable for linking a categorical clinical endpoint with a mechanistic hypothesis. The intended question was not limited to comparing treatment arms on a clinical classification; it also concerned whether change in an induced regulatory T-cell subset was associated with clinical tolerance. Those are related questions, but they require different outcomes and analyses. A clinical category comparison cannot, by itself, answer whether a cellular measure tracks with that category.

The endpoint combines several types of criteria: symptoms and an ingestion amount at a later challenge, eliciting-dose thresholds at different assessments, change relative to an earlier dose, a maintenance-dose milestone, and withdrawal. Such definitions can capture more than a single measurement, but they also make transparent reporting important. Category counts, denominators, assessment completeness, and rules for overlapping or missing criteria are needed for readers to understand how classifications were assigned.

The registry record identifies a randomized parallel-group design and a broad set of clinical and mechanistic outcomes. Yet it does not report the primary-outcome distribution, analyses, or uncertainty. This illustrates a central statistical distinction: a well-defined research question and outcome are not the same as evidence answering that question. Registration can make intended methods and outcomes visible, while interpretation still depends on complete results and analysis details.

The safety counts also illustrate why descriptive figures need context. Arm-specific counts should be considered alongside denominators, event definitions, follow-up, and uncertainty; they do not alone establish a causal contrast. Here, the record does not provide an inferential safety comparison or additional event-level detail.

For clinical researchers and biostatisticians, the key lesson is to separate the trial’s stated aims, operational endpoint definitions, and actually reported results. In this record, the first two are described more fully than the third. That limits statistical conclusions, but makes clear which information would be needed for a fuller assessment.

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 interpretation is that the primary outcome counts and analyses are not reported in the registry record. The record defines tolerance, partial tolerance, and treatment failure, but does not give the number of participants in each category or an arm comparison. It likewise does not report results for the proposed relationship between clinical classification and induced regulatory T-cell change.

The endpoint has multiple criteria and time points. The record does not explain the rationale for the thresholds, how participants meeting more than one criterion were classified, or how missing challenge assessments and withdrawals were handled. Although withdrawal is included as treatment failure, details of that classification are not reported in the registry record.

Important design and analysis details are also absent: the randomization procedure, masked roles, planned sample-size justification, statistical analysis methods, analysis populations, and missing-data procedures. The record lists zero analyses, but that does not establish that no analyses were conducted outside the registry.

The safety information is limited to arm-specific counts, deaths, and a listed serious-event term. Event-level descriptions, follow-up context, and comparative safety analyses are not reported. These limitations prevent a fuller assessment of outcome precision, potential bias, and the robustness of any treatment comparison. The enrollment and completion information cannot substitute for missing outcome results or analysis details.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Tolerance Following Peanut Oral Immunotherapy — NCT01750879.
• Registry record last updated 2021-07-28; status last verified 2021-07.
• Results first posted 2021-07-28.
• 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

The registry describes a completed, randomized, triple-masked, parallel-group phase 1 and phase 2 study of peanut flour versus oat flour oral immunotherapy for peanut allergy. Enrollment is listed as 41. The primary outcome is a participant classification of tolerance, partial tolerance, or treatment failure based on challenge assessments and treatment milestones, with a timeframe averaging 515 days from DBPCFC1 to DBPCFC3.

The stated scientific objective was to assess whether clinical tolerance was associated with expansion of a regulatory T-cell subset. The record lists several clinical and mechanistic secondary outcomes. It does not report primary-outcome category counts, analyses, or mechanistic results, and it lists zero analyses. Safety counts are reported by arm, including serious adverse events, other adverse events, and deaths. The registry record supports understanding the intended question and outcome definitions, but not conclusions about comparative clinical or mechanistic results.