Chemotherapy Before Surgery in Treating Patients With High Grade Upper Urinary Tract Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Chemotherapy Before Surgery in Treating Patients With High Grade Upper Urinary Tract Cancer (NCT02412670), a non-randomized phase 2 study. The primary outcome is Complete Pathologic Response Rate.

2. Clinical Question

3. Trial Design

This was a prospective phase II trial of neoadjuvant systemic chemotherapy followed by extirpative surgery for high-grade upper tract urothelial carcinoma. The registry record lists the study as completed and reports enrollment of 36. It gives a start date of 2015-08-27 and a primary completion date of 2018-08-23.

The design was parallel and nonrandomized, with no masking. Participants were treated in one of two regimen groups: methotrexate, vinblastine, doxorubicin, and cisplatin in arm A, or gemcitabine and carboplatin in arm B. The descriptions specify four courses in each group, with a 14-day cycle for arm A and a 21-day cycle for arm B. The arm A description also lists pegfilgrastim. The registry lists nephroureterectomy as a procedure and describes surgery as following chemotherapy.

Because allocation was not randomized, the treatment groups should not be read as interchangeable groups created to support a controlled comparison. The registry record does not report how treatment regimen was selected, whether allocation followed a prespecified rule, or whether participant characteristics differed by arm. Those details matter when interpreting contrasts between arms: nonrandomized treatment groups may differ for reasons other than the regimen being studied.

The primary endpoint was assessed at nephroureterectomy or regional lymph node dissection, with the stated assessment window 21-60 days from completion of chemotherapy. The record describes secondary outcomes including recurrence-free survival, event-free survival, bladder cancer-free survival, cancer-specific death at 24 months, and renal insufficiency at completion of chemotherapy and surgery. Although the registry lists seven outcome measures, it reports no outcome analyses. The planned statistical methods beyond the stated endpoint and confidence-interval type 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.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Eligible and treated patients

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: Only eligible and treated patients are included in the analysis.

6. Primary Endpoint

The registry lists the primary outcome as Complete Pathologic Response Rate, measured over Assessed at nephroureterectomy or regional lymph node dissection (21-60 days from completion of chemotherapy; chemotherapy was administered for a total of 4 cycles; cycle length is 14 days and 21 days for arms A and B, respectively).

The primary endpoint, complete pathologic response rate, measures the proportion of eligible and treated patients whose surgical pathology met the definition pT0pN0. The registry defines this as no evidence of disease by pathologic evaluation of the nephrectomy or ureterectomy specimen and any identifiable regional lymph nodes. It was assessed at nephroureterectomy or regional lymph node dissection after chemotherapy, within the stated interval of 21-60 days from chemotherapy completion.

Pathologic response is a way to describe what is found in tissue examined after treatment. Unlike an outcome based only on symptoms or imaging, this endpoint uses the findings from the operation specified in the registry. The proportion expresses how common the defined response was among the stated population; it does not, on its own, describe how long a response lasts or whether later recurrence occurs. The registry lists the primary endpoint’s dispersion as an 80% confidence interval, but it does not report the response count, estimated proportion, or interval limits.

Using a response endpoint can provide a focused measure of disease status at surgery in a trial of treatment before surgery. Its interpretation depends on who is included in the analysis and whether all participants have evaluable surgical pathology. The registry record identifies eligible and treated patients as the population but does not report the analysis population actually used, the number with evaluable pathology, or how missing endpoint assessments were handled. Those omissions limit interpretation of the endpoint even before considering the absence of a reported estimate.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

Statistical methods help translate participant observations into estimates while making clear what those estimates can and cannot support. For this trial, the registry identifies a binary primary endpoint, a stated population, an assessment time, and an 80% confidence interval. It does not report a primary endpoint estimate or a detailed statistical analysis plan. The concepts below explain how such methods are generally read; they do not establish which procedures were used in this trial.

What does a response proportion represent?

A binary endpoint classifies each evaluated participant into one of two categories, such as meeting or not meeting the defined complete pathologic response criterion. A response proportion is the number meeting the criterion relative to the specified analysis population. The denominator is essential: the same response count can yield different proportions if populations differ. For this endpoint, the registry names eligible and treated patients, but the number analyzed and number with evaluable pathology are not reported. It is therefore not possible to reconstruct or verify a response proportion.

The definition also governs classification. Here, complete pathologic response is defined as pT0pN0 based on evaluation of the surgical specimen and any identifiable regional lymph nodes. If pathology is unavailable or cannot be classified, the analysis needs a prespecified way to handle that situation. The registry record does not report how such cases were treated, whether all eligible and treated participants reached surgery, or whether any endpoint values were missing.

What does a confidence interval add?

A confidence interval accompanies an estimate to communicate statistical uncertainty. Its width reflects factors such as the amount of information and the statistical method. A narrower interval generally indicates greater precision under the method’s assumptions; a wider interval indicates less precision. An interval is not a guarantee about the true value and does not resolve possible bias from a nonrandomized design, missing data, or outcome measurement.

The registry specifies an 80% confidence interval for the primary endpoint’s dispersion. It does not give the interval limits, the estimated proportion, or the interval calculation method. Without those, readers cannot assess the estimate’s precision or determine whether particular values are compatible with the reported data. The confidence level alone does not reveal the result.

Why does the nonrandomized design matter?

In a randomized comparison, assignment is determined by chance according to the study design. That approach can make groups comparable in expectation, including for factors that are difficult to measure. In a nonrandomized comparison, assignment may be associated with clinical or other participant characteristics. This can produce confounding: an apparent association between regimen and outcome may partly reflect differences in who received each regimen.

Statistical adjustment can sometimes address measured differences, but it cannot automatically eliminate unmeasured or poorly measured confounding. The registry record does not report baseline comparisons, adjustment variables, or a model for estimating an arm contrast. No causal comparison should be inferred from the arm labels alone. The record also does not describe whether the primary endpoint was intended to be analyzed separately by arm, pooled across arms, or both.

How should multiple outcomes be considered?

The record lists a primary endpoint and secondary outcomes. When a study evaluates multiple outcomes or makes multiple comparisons, the chance of at least one apparently notable result can increase if all are interpreted without regard to multiplicity. Statistical plans may address this by specifying a primary test, a hierarchy, or an adjustment procedure. The registry record does not report whether multiplicity was addressed, how secondary outcomes were prioritized, or whether any formal hypothesis tests were planned. Since no outcome analyses are reported, there are no posted p-values or estimates to assess in relation to multiplicity.

What do the planning assumptions mean?

The registry describes a planning illustration based on the observed effect, 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. These are assumptions for an illustration, not evidence that the trial was designed using those assumptions. The stated approach also does not mean that the trial had a particular probability of detecting an effect after completion. Sample-size calculations are planning tools; they are not observed power, and they do not replace an estimate with an uncertainty interval.

For time-to-event outcomes, a planning calculation may focus on the number of events because event information drives precision in many survival analyses. For other outcomes, a planning illustration may be expressed as participants per arm. The record notes that distinction but does not report the resulting planning counts. No planning count can be treated as the observed number of participants or events in an analysis.

Over-dispersion is a general statistical issue for some count outcomes: observed variability can be greater than a model assumes, making standard errors too small if the model is not adapted. It is not identified as a method issue for this binary primary endpoint in the registry record. More broadly, the analysis method, assumptions, and diagnostics for the listed outcomes are not reported, so no specific model can be attributed to the trial.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("Arm A (Methotrexate, Vinblastine, Doxorubicin, Cis", "Arm B (Gemcitabine, Carboplatin)")
n      <- c(29, 6)
events <- c(0, 0)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  Arm A (Methotrexate, Vinblastine, Doxorubicin, Cis  29  0  0.0  0.0  11.9  0.0  11.7  0.1  11.6
ARM  Arm B (Gemcitabine, Carboplatin)  6  0  0.0  0.0  45.9  0.0  39.0  0.4  41.0

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary endpoint and six named secondary outcomes: recurrence-free survival, event-free survival, bladder cancer-free survival, cancer-specific death at 24 months, renal insufficiency at completion of chemotherapy, and renal insufficiency at completion of surgery. It reports seven outcome measures in total, but no analyses. The registry record does not state whether the study used a hierarchy of outcomes, adjusted for multiple testing, or treated secondary analyses as exploratory.

Multiplicity matters because examining several outcomes can increase the chance of finding an apparently unusual result by chance. Whether an adjustment is appropriate depends on the study’s inferential goals and prespecified analysis plan. Since no results or plan are reported, there is no basis to characterize any outcome as statistically significant, confirmatory, or exploratory beyond its listing in the registry.

The record does not report missing-data counts or methods for handling missing observations. For the primary endpoint, it does not state how participants without evaluable surgical pathology were handled, whether the primary analysis included all eligible and treated patients, or whether any sensitivity analyses were performed. It also does not provide participant flow for the secondary outcomes, which may require different follow-up information or event definitions.

Analysis populations are consequential: an analysis of all eligible and treated patients may differ from one restricted to participants with observed endpoint data. The registry identifies eligible and treated patients as the primary endpoint population, but does not report the number included in the analysis or whether this description matches the analysis actually performed. No intention-to-treat population or per-protocol population is specified in the registry record.

12. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: anemia (33.3%), neutrophil count decreased (33.3%), white blood cell decreased (33.3%), platelet count decreased (16.7%), dehydration (10.0%).

The adverse-event summary reports, for arm A, 30 participants at risk, 7 with a serious adverse event, 30 with another adverse event, and 2 deaths. For arm B, it reports 6 participants at risk, 3 with a serious adverse event, 6 with another adverse event, and 1 death. These are descriptive counts in the registry record. They do not establish that an event was caused by chemotherapy or surgery, and the record does not specify whether the reported deaths were considered treatment-related.

The listed serious adverse-event terms and percentages are anemia at 33.3%, neutrophil count decreased at 33.3%, white blood cell decreased at 33.3%, platelet count decreased at 16.7%, dehydration at 10.0%, and fatigue at 3.3%. The registry’s list does not attribute these percentages to a specific arm, so they should not be read as arm-specific rates. The record does not provide the event counts behind those percentages or the duration of safety follow-up.

Participants at risk are the denominators attached to the arm-level safety summaries. The counts describe how many participants were recorded as affected, not the number of event episodes; a participant may have had more than one event. The registry record does not report event severity distributions beyond the serious-event grouping, exposure time, or a comparative safety analysis. With nonrandomized groups of different sizes, these summaries should not be treated as a controlled estimate of relative risk.

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

The record establishes that a completed phase II study evaluated chemotherapy before surgery in high-grade upper tract urothelial carcinoma. It specifies two regimens, a nonrandomized parallel design, and complete pathologic response as the primary endpoint. It also gives a precise pathologic definition and an assessment window. These details define the research question and help readers understand what an eventual endpoint estimate would describe.

The record does not report the primary response result or analyses of the secondary outcomes. It therefore cannot show how often complete pathologic response occurred, how precise an estimate was, or what the longer-term outcomes were. Completion status and enrollment do not substitute for outcome data. Nor does a listed endpoint establish that its result favored one regimen or reached a particular statistical threshold.

The two regimens were not allocated randomly. Even if arm-specific outcomes were available, a difference would be descriptive unless supported by an analysis that addresses the limitations of nonrandomized treatment assignment. Statistical adjustment might account for recorded differences, but the registry record does not report such methods or the relevant baseline information. The observed safety counts likewise do not establish causality or a comparative safety effect.

The endpoint concerns pathology at surgery. It is distinct from recurrence, survival, bladder cancer-free survival, and later renal insufficiency, which are separately listed outcomes. A pathologic response result, if reported, would not by itself answer those other questions. Similarly, the listed secondary outcomes cannot be interpreted without estimates, definitions, follow-up, and analysis details.

The most defensible reading is therefore bounded: the registry describes the trial’s design, treatment groups, endpoint definition, and selected safety summaries, but it does not provide the outcome analyses needed to judge the observed response or other clinical outcomes. This absence of reported results is not evidence of a positive or negative result; it means the statistical result is not reported in the registry record.

14. Why This Trial Matters Statistically

This trial raises a statistical question that is common in early clinical research: how should a treatment response observed after an intervention be described when the study has no randomized control group? The endpoint is clearly defined as a pathologic state at surgery, and the registry states the population and assessment timing. Those elements are central to interpreting an estimate because they specify what counts as a response, whom the estimate represents, and when status is assessed.

The nonrandomized parallel design also makes the distinction between description and causal comparison especially important. Two treatment groups can be described separately, but their contrast is not automatically an unbiased estimate of what would have happened if the same kinds of participants had received the alternative regimen. Treatment selection and differences between groups may influence observed outcomes. Without allocation details, baseline characteristics, and an analysis plan, readers cannot establish how those issues were handled.

The record’s lack of endpoint analyses illustrates another statistical point: design information and outcome evidence are not interchangeable. A completed trial can have a specified primary endpoint and a defined assessment window while the registry record still lacks the estimate and its uncertainty. Without those results, the response proportion’s magnitude and precision remain unknown.

Finally, the record lists both an endpoint at surgery and outcomes involving recurrence, survival, cancer-specific death, and renal insufficiency. These outcomes represent different aspects of follow-up and should not be collapsed into a single notion of benefit or harm. Their separate definitions, analysis populations, and results would be needed for a fuller statistical account. The trial matters as a defined phase II investigation, while the available registry information leaves its principal efficacy question unanswered.

15. 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 central limitation for statistical interpretation is that the registry record reports no primary outcome analysis. There is no complete pathologic response count, proportion, or confidence interval. It also reports no analyses for the listed secondary outcomes. As a result, the magnitude and precision of the principal endpoint and the other outcomes cannot be assessed.

The study was nonrandomized and unmasked. The registry does not report how regimen assignment was made, baseline characteristics by arm, or an adjusted comparison. This limits any interpretation of differences between groups, because treatment groups may differ for reasons beyond the regimen. The record does not identify an analysis model or explain whether any confounding adjustment was planned or conducted.

Enrollment is reported as 36, while the safety summaries give 30 participants at risk in arm A and 6 in arm B. The registry does not provide a flow accounting that explains the relationship between enrollment, eligibility, treatment, surgery, endpoint evaluation, and each analysis population. It also does not report how participants without evaluable pathology or follow-up were handled. This leaves uncertainty about the denominators that would support the primary and secondary outcome estimates.

The registry lists an 80% confidence interval for the primary endpoint but gives no limits or calculation method. It does not report a statistical analysis plan, multiplicity procedure, interim analyses, or sample-size justification. The record describes a planning illustration, but not its resulting counts; such an illustration should not be mistaken for the trial’s design justification or observed power.

The safety summary is limited to affected-participant counts and a list of serious-event terms with percentages. Arm-specific attribution for the term percentages, event counts, exposure duration, and assessment procedures are not reported. The record also does not clarify the relationship between the reported deaths and treatment. These limitations make the safety information descriptive rather than a basis for a comparative or causal conclusion.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Chemotherapy Before Surgery in Treating Patients With High Grade Upper Urinary Tract Cancer — NCT02412670.
• Registry record last updated 2023-07-10; status last verified 2023-06.
• Results first posted 2021-01-20.
• 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.

19. Record Summary

The registry record describes a completed, prospective phase II trial of chemotherapy before surgery for high-grade upper tract urothelial carcinoma. It lists enrollment of 36, a nonrandomized parallel design, and no masking. Arm A received methotrexate, vinblastine, doxorubicin, and cisplatin; arm B received gemcitabine and carboplatin.

The primary endpoint was complete pathologic response, defined as pT0pN0 and assessed at nephroureterectomy or regional lymph node dissection. The record specifies an 80% confidence interval but reports no response estimate or primary analysis. It lists secondary outcomes involving recurrence, survival, cancer-specific death, and renal insufficiency, with no analyses reported.

Safety summaries report serious and other adverse events, deaths, and selected serious-event terms. The trial’s nonrandomized design, absent outcome analyses, and unreported details about analysis populations and missing data limit statistical interpretation. The primary clinical outcome is not reported in the registry record.