Allogeneic Natural Killer (NK) Cells for Ovarian, Fallopian Tube, Peritoneal and Metastatic Breast Cancer: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Allogeneic Natural Killer (NK) Cells for Ovarian, Fallopian Tube, Peritoneal and Metastatic Breast Cancer (NCT01105650), a non-randomized phase 2 study. The primary outcome is Response Rate.

2. Clinical Question

3. Trial Design

This was a completed, single-center Phase 2 trial sponsored by Masonic Cancer Center, University of Minnesota. The registry lists a start of 2010-07 and primary completion of 2014-04. Enrollment was 13. The study used a single-group model, non-randomized allocation, and no masking. Although the regimen descriptions are divided into four arms, the record does not describe a randomized allocation process or a concurrent control group.

The common treatment platform began with fludarabine and cyclophosphamide, followed by donor-derived allogeneic NK-cell infusion. Cyclosporine and IL-2 were included in the regimens. Arm 1 used cyclosporine without a stated methylprednisolone component and scheduled IL-2 for 6 doses. Arm 2 added methylprednisolone (10mg), and Arm 3 added methylprednisolone (1 mg); both descriptions also specify 6 IL-2 doses. Arm 4 omitted methylprednisolone and reduced the IL-2 schedule to 3 doses.

The trial description frames the study as an effort to optimize a clinical platform intended to promote persistence, function, and expansion of allogeneic NK cells. That statement describes the study rationale, not an outcome demonstrated by the registry results. The record does not report why participants received particular regimens, whether the arms were sequential, or whether other protocol features differed. The precise allocation process and any criteria used to choose among regimens are not reported in the registry record.

The primary outcome was response at Month 3. Secondary outcomes were time to disease progression, the number of participants with progressive disease at one year, and overall survival. The registry lists four outcome measures, but does not provide numerical results for the primary response endpoint in the registry record. It also does not report the results for the secondary outcomes.

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: Up to 4 sequential immunosuppressive platforms (study arms) were tested to identify a platform where patients have the potential for successful Natural Killer (NK) cell expansion (defined as an absolute circulating donor derived NK cell count of \> 100 cells/μl 14 days after NK cell infusion).

Overall Study

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 Response Rate, measured over Month 3.

The primary endpoint was the number of participants with a response at Month 3. The registry defines response as complete response (CR), partial response (PR), or stable disease (SD), assessed using RECIST v.1.1 criteria for target lesions on CT or MRI. CR means disappearance of all target lesions. PR means a >=30% decrease in the sum of the longest diameter of target lesions. SD means neither enough shrinkage to qualify as PR nor enough increase to qualify as progressive disease.

Combining these categories means that the endpoint counts stable disease as well as tumor shrinkage categories. It therefore captures a broader radiographic disease-status outcome than an endpoint limited to complete or partial response. As a general statistical point, a prespecified categorical endpoint can provide a consistent rule for classifying participant outcomes at a stated assessment time. How that endpoint is interpreted depends on the definition and on complete, consistently assessed outcome data.

The registry states the assessment time and criteria, but does not report how many participants were assessed, how many met each response category, or the primary response result. It also does not report the handling of participants whose Month 3 assessment was missing or unevaluable. Consequently, the endpoint definition is available, while the observed response distribution and its precision are not reported in the registry record.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary outcome as a count of participants with an event, with the event defined as meeting the response criteria. This is a binary outcome at the participant level: a participant either meets the stated response definition at the assessment or does not. The record’s endpoint description combines CR, PR, and SD, and specifies Month 3 as the timeframe. The result itself is not reported in the registry record.

What does a confidence interval tell a reader?

General explanation: a confidence interval describes the uncertainty around an estimate under a statistical model and its assumptions. For a participant-count or response-rate outcome, it can communicate how much imprecision is associated with the observed estimate. It does not repair limitations in study design, and it does not make a non-randomized comparison equivalent to a randomized one. No confidence interval for the primary endpoint is reported in the registry record, so its precision cannot be assessed from this information.

Why does the study design matter?

General explanation: random assignment helps balance measured and unmeasured participant characteristics across treatment groups, on average. When assignment is non-randomized, differences between groups can reflect baseline characteristics, clinical decisions, or other factors as well as differences in treatment. This trial is explicitly described as non-randomized, and the registry warns that its arms are not like-for-like controls. Thus, even if arm-specific outcomes were available, a simple contrast would not by itself isolate the effect of a regimen change.

The study is also described as using a single-group model, while listing four regimen arms. The record does not explain the relationship between that model designation and the arm structure, the allocation process, or whether treatment groups were enrolled in a particular order. Those details are not reported in the registry record and should not be inferred from the arm labels.

How should the response definition be understood?

The endpoint is a composite of CR, PR, and SD. Its categories are based on target-lesion changes assessed by CT or MRI under RECIST v.1.1. CR requires disappearance of all target lesions; PR requires a >=30% decrease in the sum of their longest diameters; and SD is defined by not meeting the stated shrinkage or progression thresholds. As a general matter, combining categories can make an endpoint represent several clinically distinct states. The meaning of the combined result depends on the individual category counts and on the exact endpoint definition. Those category counts are not reported in the registry record.

The record does not state whether outcome assessment was blinded. It lists masking as none. General explanation: knowledge of treatment assignment can potentially influence subjective parts of outcome assessment, while radiographic criteria provide explicit classification rules. The registry does not describe any additional safeguards for assessing or adjudicating response.

What can planning calculations show?

The registry’s stated planning assumptions describe an illustration based on an observed effect, using a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The stated effect to detect is a fraction of the observed effect. These are planning assumptions, not a report of the trial’s original design justification. The record does not report the original sample-size justification, and no calculated planning figures are included in the registry record for this explanation.

General explanation: a sample-size calculation estimates how many participants would be needed under specified assumptions to detect a chosen effect with a chosen level of statistical power. Such calculations depend on the assumed effect, outcome variability or event frequency, allocation, and analysis method. A planning illustration based on an observed effect is retrospective and should not be described as observed power. It also cannot substitute for the trial’s prespecified design rationale, which is not reported in the registry record.

For binary outcomes, an analysis may compare participant counts or proportions, but the record does not report a primary endpoint analysis, a test, an effect estimate, or a confidence interval. It lists zero analyses. No statistical test or adjustment method should be assumed. General explanation: small datasets tend to yield imprecise estimates, and any estimate is best interpreted alongside its uncertainty and the completeness of outcome ascertainment. The registry’s enrollment count alone does not supply that uncertainty.

For time-to-event outcomes, such as time to disease progression and overall survival, analyses commonly account for the time each participant is followed and whether an event has occurred by the last observation. General explanation: censoring occurs when the event time is not observed during available follow-up; appropriate methods use the follow-up information without treating every censored participant as though an event occurred. The registry names these outcomes but does not report their analysis methods, follow-up summaries, event counts, or estimates. Whether censoring was present and how it was handled are not reported in the registry record.

Over-dispersion is a general concern when count data vary more than a chosen statistical model expects. It can make standard errors too small if the model does not account for the extra variation. This trial’s primary outcome is described as a participant count for a binary event, not as a repeated count process, and no model or dispersion estimate is reported. There is therefore no basis to claim that over-dispersion was assessed or to identify a particular model as used.

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 1: CsA", "Arm 2: CsA/Methylprednisolone (10mg)/6 Doses of In", "Arm 3: CsA/Methylprednisolone (1mg)/6 Doses of Int")
n      <- c(3, 3, 7)
events <- c(1, 2, 4)
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 1: CsA  3  1  33.3  0.8  90.6  6.1  79.2  6.8  80.6
ARM  Arm 2: CsA/Methylprednisolone (10mg)/6 Doses of In  3  2  66.7  9.4  99.2  20.8  93.9  19.4  93.2
ARM  Arm 3: CsA/Methylprednisolone (1mg)/6 Doses of Int  7  4  57.1  18.4  90.1  25.0  84.2  24.5  84.3

11. Multiplicity, Missing Data and Analysis Populations

The registry lists a primary outcome and three secondary outcomes. It does not report a multiplicity strategy, such as a hierarchy for testing endpoints or an adjustment for examining multiple outcomes. The statistical analysis plan is not reported in the registry record. Without that information, it is not possible to determine whether or how the risk of false-positive findings across multiple analyses was addressed. No endpoint-specific analysis results are listed in the registry record for this record.

Missing outcome data can affect both the estimated result and its interpretation. The record does not state how many participants had an evaluable Month 3 assessment, whether any response assessments were missing, or how missing or unevaluable outcomes were handled. It also does not describe missing-data methods for the secondary outcomes. These details are not reported in the registry record; no assumption about the completeness of follow-up or the handling of missing observations should be made.

The analysis population is also not specified. The registry does not say whether the outcome analyses included all enrolled participants, only participants who received treatment, or another defined group. It does not report exclusions or the number analyzed for any endpoint. The posted enrollment of 13 should not be taken as proof that all 13 participants contributed to every outcome analysis.

General explanation: multiplicity concerns arise when several statistical questions are examined, while missing-data concerns arise when an outcome is not observed for some participants. Prespecified analysis choices help readers understand how these issues were handled. Here, the outcome list is available, but the corresponding analysis plan and analysis populations are not reported in the registry record.

12. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

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: death nos (100.0%), abdominal pain (33.3%), dyspnea (33.3%), hearing impairmed (33.3%), vestibular disorder (33.3%).

The registry lists serious and other adverse-event counts by arm. For Arm 1, 3 participants were at risk and 3 were affected for serious events; the corresponding other-event counts are also 3 at risk and 3 affected. Arm 2 has the same listed counts. For Arm 3, 7 participants were at risk for serious events and 5 were affected; for other events, 7 were at risk and 7 were affected. Death counts are recorded as null for each listed arm.

The serious-event terms listed include Death NOS at 100.0%, and abdominal pain, dyspnea, hearing impairmed, vestibular disorder, and blurred vision, each at 33.3%. The registry does not report the denominator or other context needed to interpret these percentages. The term Death NOS appears in the serious-event listing while arm-level death counts are null; the record does not explain that relationship, so it should not be resolved by assumption.

These entries describe reported events and affected participants, not comparative risks or causal effects. The listed arm counts are small, and the arms are non-randomized, so the entries do not support a like-for-like safety comparison. The registry record does not provide additional detail on event timing, duration, severity grading beyond the serious/other grouping, or attribution to treatment.

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

The record establishes that a completed Phase 2 trial enrolled 13 participants with recurrent ovarian, fallopian tube, or primary peritoneal cancer, or advanced metastatic breast cancer. It describes a regimen platform involving preparative fludarabine and cyclophosphamide, allogeneic NK-cell infusion, cyclosporine, and IL-2, with variations in methylprednisolone and IL-2 schedules across four arms. It also defines the primary endpoint as response at Month 3 under RECIST v.1.1, counting CR, PR, and SD.

Those design and endpoint details do not establish what the trial observed. The primary response count and response rate are not reported in the registry record. Nor are the listed secondary outcomes accompanied by numerical results. Without outcome estimates and their uncertainty, the record cannot quantify response, progression, or survival, or determine how precise any such estimate would be.

The arm descriptions do not establish comparative treatment effects. Allocation was non-randomized, and the registry cautions that arms are not like-for-like controls. Differences in outcomes, if reported, could reflect differences between participants or other factors in addition to regimen differences. The posted arm structure alone does not show that one regimen is superior or inferior to another.

General statistical interpretation: a study’s rationale and treatment description are not evidence that its intended biological or clinical effects occurred. Likewise, a response definition is not a response result. The listed adverse-event counts provide descriptive safety information but do not, on their own, establish the cause of events or permit reliable comparisons between regimens.

The appropriate conclusion is therefore limited to what the registry describes: the trial design, the regimen variations, the endpoint definitions, and the listed safety counts. The efficacy and comparative findings needed to make stronger statistical claims are not reported in the registry record.

14. Why This Trial Matters Statistically

This trial raises several important statistical issues in a compact setting. It combines a small enrollment with four non-randomized regimen arms and a single primary response endpoint. The design makes clear why an arm label should not be mistaken for a randomized comparison: differences in outcomes could reflect participant selection or other factors as well as changes in treatment. The registry itself warns that the arms are not like-for-like controls.

The endpoint definition also matters. Response includes complete response, partial response, and stable disease, so the endpoint combines distinct radiographic categories. Readers need the category counts and an understanding of how missing assessments were handled to interpret a result fully. The registry provides the definitions and assessment time but does not report the outcome counts.

Another statistical lesson is the distinction between a study’s stated aim and evidence about whether that aim was achieved. The trial description says it was designed to optimize a clinical platform intended to promote NK-cell persistence, function, and expansion. Those aims do not replace reported clinical outcomes or measurements of those biological features. The registry does not report results for the primary or listed secondary outcomes in the registry record for this record.

Finally, safety summaries need context. Counts and event terms can describe what was recorded, but interpretation depends on denominators, follow-up, event definitions, and how events were collected. Some of this information is not reported in the registry record. Taken together, these details show why design, endpoint definition, analysis methods, and uncertainty all matter when reading a clinical trial, especially when the posted numerical outcome results are incomplete.

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 primary limitation for statistical interpretation is that the registry record does not report the primary response result. No number of participants with CR, PR, or SD at Month 3, response rate, uncertainty interval, or analysis result is available in the registry record for this record. The secondary outcomes are named but their numerical results are not reported either.

Enrollment was 13, and the study used non-randomized allocation. The four regimen arms are not a randomized comparison, as the registry explicitly notes. The record does not report the allocation process, arm-specific outcome results, baseline characteristics by arm, or reasons for assigning participants to a regimen. These omissions prevent an assessment of comparability between arms.

The registry does not report the number of participants assessed at Month 3, the analysis population for each outcome, or how missing or unevaluable observations were handled. It also does not report the statistical analysis plan, multiplicity procedures, endpoint-specific methods, or the original sample-size justification. These omissions limit assessment of the prespecification and robustness of any analysis.

The safety information has additional constraints. Although arm-specific affected counts and serious-event terms are listed, the denominator for the term percentages and context for interpreting the Death NOS listing are not reported. Arm-level death counts are null. Information about event timing, follow-up, severity detail, and attribution is not reported in the registry record.

The record states the response assessment method and timeframe, but does not report whether assessment was blinded or how assessments were adjudicated. It also does not provide enough information to characterize follow-up for progression and survival outcomes. These limits are specific to the information recorded and do not establish what may have been collected outside the registry record.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: Allogeneic Natural Killer (NK) Cells for Ovarian, Fallopian Tube, Peritoneal and Metastatic Breast Cancer — NCT01105650.
• Registry record last updated 2017-12-28; status last verified 2017-12.
• Results first posted 2017-02-24.
• 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

NCT01105650 is a completed, single-center Phase 2 trial sponsored by Masonic Cancer Center, University of Minnesota. The registry lists a start of 2010-07, primary completion of 2014-04, and enrollment of 13. It describes a non-randomized, single-group study with no masking and four regimen arms.

The treatment platform included fludarabine and cyclophosphamide before allogeneic NK-cell infusion, with cyclosporine and IL-2. Arm descriptions differed in methylprednisolone use or dose and, in Arm 4, the IL-2 schedule. The primary endpoint was response at Month 3, defined to include CR, PR, or SD under RECIST v.1.1. Secondary outcomes were time to disease progression, progressive disease at one year, and overall survival.

The registry does not report numerical results for the primary or secondary outcomes in the registry record for this record. It lists adverse-event counts and serious-event terms, but the context needed for a full comparative interpretation is limited. The non-randomized arms are not like-for-like controls.