Ultrasound and Near Infrared Imaging for Predicting and Monitoring Neoadjuvant Treatment: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Ultrasound and Near Infrared Imaging for Predicting and Monitoring Neoadjuvant Treatment (NCT02891681), a non-randomized na study. The primary outcome is Pathologic Response Based on Miller-Payne Grading System.

2. Clinical Question

The study asks whether assessment of tumor vasculature and oxygen changes using near-infrared and ultrasound imaging (NIR/US) can help predict and monitor response during neoadjuvant treatment for breast cancer. The intended comparison for assessing response is the pathological response determined from tissue before treatment and the definitive surgical specimen.

3. Trial Design

This completed study is registered as NCT02891681 and lists an enrollment of 41. Its start date is 2016-11-29 and its primary completion date is 2020-01-23. The phase is listed as NA. Washington University School of Medicine is the sponsor.

The allocation is non-randomized, the masking is listed as none, and the model is parallel. The described groups are a neoadjuvant chemotherapy cohort and a neoadjuvant endocrine cohort, with a crossover cohort from endocrine treatment to chemotherapy also identified in the adverse-event record. Because allocation was not randomized, differences between cohorts cannot be interpreted as treatment effects from a randomized comparison. The registry record does not report a like-for-like control comparison.

The intervention is a device described as optical tomography using near-infrared diffused light assisted with ultrasound. The arm descriptions state that participants were to have a baseline NIR/US scan before their first treatment, with the desired timing at least 7 days after the initial biopsy to avoid possible biopsy-related acute inflammatory effects. The arm descriptions also mention a further scan at the end of a treatment cycle, but the schedule details beyond that wording are not reported in the registry record.

The outcome timeframe is up to 6 months. For the primary outcome, the registry says the crossover participant was counted in the neoadjuvant chemotherapy cohort. It also reports that three participants in that cohort were not evaluable for the outcome; the description identifies two who developed metastases before completing neoadjuvant therapy, but the reason for the remaining participant 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.

Before assignment: An additional arm (NIR/US - Crossover from Endocrine Cohort to Chemotherapy Cohort) was added for results reporting as one participant transitioned from the Endocrine Cohort to the Chemotherapy Cohort because the participant's therapy was changed mid-treatment from Endocrine to Chemotherapy.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: For this outcome measure, the one participant who was in the crossover cohort was counted in the neoadjuvant chemotherapy cohort. Three participants in the neoadjuvant chemotherapy cohort were not evaluable for this outcome measure (2 developed metastases prior to completing neoadjuvant therapy and 1 was taken off study after first imaging appointment).

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 Pathologic Response Based on Miller-Payne Grading System, measured over Up to 6 months.

The primary endpoint is the count of participants classified by pathological response using the Miller-Payne grading system, with assessment up to 6 months. This system compares tumor cellularity in a pre-neoadjuvant core biopsy with the definitive surgical specimen. It assigns grades from 1 to 5: grade 1 indicates no overall reduction in cellularity; grades 2 and 3 describe progressively greater loss of tumor cells; grade 4 describes a marked disappearance of tumor cells; and grade 5 means no malignant cells are identifiable at the tumor site, although ductal carcinoma in situ may remain.

The grading system is used to describe pathological response on an ordered scale, including distinctions between no response, partial response, near-complete response, and complete response as defined in the registry. The endpoint is relevant to the study’s stated question because pathological findings provide the reference against which NIR/US assessment was intended to be compared. The registry record does not report the number of participants in each grade, a grouped response count, or an accuracy measure. It therefore does not show how imaging assessments corresponded to pathological response.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The study’s primary endpoint is a count of participants classified by Miller-Payne pathological response. The record gives the grading definitions but does not provide the observed counts. This distinction matters: defining an endpoint explains what investigators intended to measure, while reporting its results requires the number of participants in each outcome category and enough information to understand the analysis population. Those results are not reported in the registry record.

What does the grading system represent?

The Miller-Payne system orders pathological findings by the degree of change in tumor cellularity between the pre-treatment core biopsy and the definitive surgical specimen. In general, an ordered grading system can preserve distinctions among levels of response rather than reducing all findings to a single yes-or-no outcome. The registry describes five grades, from no overall reduction in cellularity to no identifiable malignant cells at the tumor site. It also notes that ductal carcinoma in situ may remain in grade 5. These definitions specify how the pathological response categories are intended to be interpreted; they do not reveal how many participants received each grade.

How should imaging accuracy be evaluated?

In general, an assessment of diagnostic or predictive performance requires a defined imaging result and a reference outcome, along with a clear rule for classifying each participant. Here, the stated reference is pathological response, while the stated aim is to evaluate NIR/US assessment of tumor vasculature and oxygen changes. An accuracy analysis would need posted paired information or reported performance estimates. Neither an accuracy estimate nor the underlying outcome counts are reported in the registry record. The record therefore does not support calculation or interpretation of imaging performance.

What can non-randomized allocation establish?

In general, randomization helps balance measured and unmeasured participant characteristics between treatment groups. In a non-randomized study, cohort membership may be associated with participant characteristics, treatment decisions, or other factors. This study is listed as non-randomized and has chemotherapy and endocrine cohorts, along with a crossover cohort. The registry explicitly identifies the arms as not being a like-for-like control comparison. Their results should not be read as if random assignment had made the groups comparable. The record does not report an adjusted analysis that could address differences between cohorts.

What does a confidence interval convey?

In general, a confidence interval describes statistical uncertainty around an estimate, reflecting the range of values compatible with the data under the interval’s method and assumptions. A narrower interval indicates greater precision than a wider interval, but an interval does not remove bias from study design or incomplete outcome information. No confidence intervals are reported for the primary endpoint or imaging accuracy in the registry record. Precision cannot therefore be assessed from posted estimates.

What do the planning assumptions mean?

The registry describes sample-size calculations as planning illustrations based on an observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. In general, a planning calculation estimates the sample size needed to detect a specified effect under stated assumptions. It is not a result of the trial and is not observed power. The registry does not report the observed effect or the calculated participant totals in the information available for this analysis. The study itself is listed as non-randomized, so an equal-allocation illustration should not be mistaken for the actual cohort structure.

For binary outcomes, general sample-size methods depend on the event rates anticipated in each group and the target effect. For continuous outcomes, assumptions about variability also matter; for time-to-event outcomes, the required information is commonly tied to event counts. These are general statistical principles, not descriptions of analyses reported for this study. The primary endpoint is recorded as a participant count by pathological grade, and the record lists no analyses. It does not state a statistical model, a test used to compare imaging with pathology, a confidence-interval method, or a strategy for covariate adjustment.

Over-dispersion is a general concern when count data vary more than a chosen model expects. In that situation, standard errors from a model that assumes less variability may be too small, making uncertainty appear narrower than it is. The endpoint is described as a count of participants, not as a rate with a reported model, and no model-based analysis is listed. There is therefore no basis for claiming that over-dispersion was assessed or that a particular count model was used.

Overall, the methods information supports understanding the intended endpoint and design, but not reconstructing an analysis. The absence of reported outcome counts and analysis results prevents readers from independently checking estimates, uncertainty, or assumptions.

10. Multiplicity, Missing Data and Analysis Populations

The registry record lists one outcome measure and zero analyses. It does not report multiple statistical tests, a hierarchy of hypotheses, an adjustment for multiplicity, or an interim-analysis plan. These details are not reported in the registry record, so it is not possible to determine whether multiplicity was addressed in a protocol or analysis plan.

Missing outcome information is a central limitation for interpreting the primary endpoint. The record states that three participants in the neoadjuvant chemotherapy cohort were not evaluable for this outcome. It identifies two as having developed metastases before completing neoadjuvant therapy; the reason for the third is not reported in the registry record. The record also says the crossover participant was counted in the chemotherapy cohort. It does not give grade counts, a complete analysis denominator, or an account of missing pathological assessments across all cohorts.

In general, missing outcomes can affect interpretation when the likelihood of being evaluable is related to response, disease course, or treatment. Different approaches to missing data rely on assumptions, and those assumptions should be stated alongside an analysis. No missing-data method, imputation strategy, sensitivity analysis, or analysis-population definition beyond the crossover and non-evaluable statements is reported in the registry record. The effect of non-evaluability on the primary outcome therefore cannot be assessed.

The record does not report an interim-analysis plan or stopping rules. This absence should not be taken to mean that such procedures were or were not used outside the registry record; their status is not reported there.

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.

The adverse-event record lists three groups: the neoadjuvant chemotherapy cohort, with 39 participants at risk; the neoadjuvant endocrine cohort, with 1 at risk; and the crossover cohort, with 1 at risk. For each group, the recorded number with a serious adverse event is 0, the number with another adverse event is 0, and deaths are 0. No serious adverse-event terms are listed.

These are registry counts, not comparative safety estimates. The groups are non-randomized, and the endocrine and crossover groups are small as recorded. The figures do not establish that no adverse events occurred outside the categories or reporting scope captured in this record. The registry record does not report detailed event definitions, collection procedures, observation time by participant, or a safety analysis. The counts should therefore be read as the posted adverse-event figures rather than as proof that the imaging procedure or treatment had no risk.

In general, adverse-event counts depend on how events are defined, collected, and attributed. A zero recorded count is not equivalent to evidence that the underlying risk is zero. No rates or uncertainty intervals are reported, and no between-group safety comparison is listed.

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

The study was designed to examine whether NIR/US assessment of tumor vasculature and oxygen changes could predict and monitor early response to neoadjuvant treatment, using pathological response as the reference. The registry record identifies the imaging approach, treatment cohorts, and Miller-Payne grading system. It does not report the primary outcome counts or any accuracy estimates, so it does not establish the strength or precision of a relationship between imaging findings and pathological response.

The record also does not show that one treatment cohort performed better than another. Allocation was non-randomized, and the registry notes that the cohorts are not a like-for-like control comparison. Any apparent differences between such groups would need careful interpretation, especially without reported outcome results or adjusted analyses.

In general, a study’s intended question and its result are distinct. A protocol can specify what an investigation aims to measure, but conclusions about the measured relationship require posted data and analysis. Here, the outcome definitions alone cannot tell readers whether imaging classifications agreed with surgical pathology, how often the two differed, or how uncertain any estimate would be.

The findings should not be interpreted as evidence that NIR/US does or does not predict response. Rather, the registry record documents a completed study with a relevant measurement question, while leaving the primary results unreported. The available adverse-event counts are likewise limited to the recorded categories and do not establish absence of risk. A fuller interpretation would require the outcome distribution, an account of evaluability, and methods and estimates for the imaging-pathology comparison.

13. Why This Trial Matters Statistically

This study addresses a statistical question about measurement: whether a non-invasive imaging assessment can track biological changes that correspond to pathological response. Such questions require more than a treatment comparison. They depend on defining the imaging measure, the reference outcome, the timing of assessments, and the way agreement or predictive performance will be summarized.

The registry identifies pathology as the reference and provides an ordered response scale, giving readers a clear account of the intended endpoint. It also records baseline imaging before treatment and a desired interval after biopsy. These design details are relevant because measurement timing can affect what an imaging assessment captures. The registry’s stated concern about biopsy-related acute inflammation explains why the desired interval was included in the description.

At the same time, the absence of posted outcome counts illustrates why registration details and results serve different purposes. A defined endpoint supports interpretation, but without the observed distribution and analysis, the accuracy and precision of the imaging assessment cannot be judged. Non-randomized cohort allocation also means the treatment groups are not direct randomized controls.

The statistical value of this record is therefore chiefly in making the research question and outcome definition visible. Its limitations also make clear what is needed to evaluate a prediction or monitoring claim: complete outcome reporting, transparent handling of participants who are not evaluable, and statistical estimates with uncertainty. Those elements are not reported in the registry record.

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 principal limitation is that the primary pathological response results are not reported in the registry record. The record gives the Miller-Payne definitions but no counts by grade, no grouped response results, and no estimates of imaging accuracy. It also lists zero analyses, leaving no reported inferential method or uncertainty estimate to evaluate.

The study is non-randomized, with chemotherapy and endocrine cohorts and a crossover cohort. The registry cautions that the arms are not a like-for-like control comparison. Without random allocation, cohort differences cannot be attributed to an intervention by simple comparison. No adjusted comparison is reported in the registry record.

Outcome evaluability is incompletely described. Three participants in the chemotherapy cohort were not evaluable for the primary outcome; two are reported to have developed metastases before completing neoadjuvant therapy, while the reason for the other participant is not reported in the registry record. The record does not state the complete outcome denominator or describe a missing-data strategy. It also does not report the number of participants assigned to each Miller-Payne grade.

The imaging schedule description includes a baseline scan and a desired interval of at least 7 days after biopsy, but details beyond the mention of an additional scan at the end of a treatment cycle are not reported in the registry record. The record also does not report the statistical analysis plan, interim analyses, multiplicity procedures, or full eligibility criteria. These limitations prevent a complete assessment of measurement performance, precision, and possible sources of bias.

15. Related Tutorials

16. Related Statistical Calculators

17. Sources

• ClinicalTrials.gov: Ultrasound and Near Infrared Imaging for Predicting and Monitoring Neoadjuvant Treatment — NCT02891681.
• Registry record last updated 2021-02-16; status last verified 2021-01.
• Results first posted 2021-02-16.
• 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.

18. Record Summary

NCT02891681 is a completed, non-randomized study sponsored by Washington University School of Medicine, with enrollment listed as 41. It examined near-infrared and ultrasound imaging in the context of neoadjuvant treatment for breast cancer. The study included chemotherapy and endocrine cohorts and identified a crossover cohort.

The primary outcome was participant pathological response using the Miller-Payne grading system, with a timeframe of up to 6 months. The registry describes the grading scale and notes that the crossover participant was counted in the chemotherapy cohort. It reports three participants in that cohort as not evaluable for the outcome, including two who developed metastases before completing treatment.

The primary outcome counts, imaging accuracy estimates, inferential analyses, and missing-data methods are not reported in the registry record. Adverse-event figures list zero serious events, zero other events, and zero deaths in each recorded group.