A Phase 2, Multicenter, Open-label Study of MEDI-551 in Adults With Relapsed or Refractory Chronic Lymphocytic Leukemia (CLL): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Phase 2, Multicenter, Open-label Study of MEDI-551 in Adults With Relapsed or Refractory Chronic Lymphocytic Leukemia (CLL) (NCT01466153), a randomized phase 2 study. The primary outcome is Objective Response Rate.

2. Clinical Question

The question is about response proportions in the randomized study groups, not about how long response lasted or how long participants lived without progression. Those are different outcomes. The registry lists time to response, time to disease progression, and progression-free survival as secondary outcomes, but their results are not included in the primary comparison.

The posted primary results report response percentages and 95% confidence intervals for the three groups. These provide a descriptive comparison and show the uncertainty around each group’s response estimate. Recalculated unadjusted comparisons can also describe the difference between a MEDI-551 group and the rituximab group, but they are not the same as the sponsor’s reported analysis. The registry record does not report enough detail to reconstruct every element of the sponsor’s analysis plan.

3. Trial Design

This was a phase 2, multicenter, randomized, open-label, parallel-group study of MEDI-551 and bendamustine versus rituximab and bendamustine in adults with relapsed or refractory CLL. The registry lists enrollment as 183. The study is marked completed. Its recorded start date is 2012-02-07, and its primary completion date is 2016-01-08.

Three treatment groups are described: rituximab plus bendamustine, MEDI-551 2 mg/kg plus bendamustine, and MEDI-551 4 mg/kg plus bendamustine. The treatment descriptions specify intravenous administration. Bendamustine was administered on Days 1 and 2 of each cycle, for a stated total of 6 cycles. Rituximab was described as given on Day 2 of Cycle 1 and then on Day 1 of up to 5 subsequent 28-day cycles. MEDI-551 was described as given on Days 2 and 8 of Cycle 1 and then on Day 1 of up to 5 subsequent 28-day cycles. The record says bendamustine was administered before the other study drug.

Randomization supports a comparison between assigned groups, while the open-label design means participants and study personnel were not masked to treatment assignment. The registry does not report the randomization ratio, the details of sequence generation or allocation concealment, or the reasons participants may not have received assigned treatment. These details should not be inferred from the treatment groups or enrollment figure.

The primary endpoint was objective response rate in the intent-to-treat population. The registry defines that population as all randomized participants. Its endpoint description counts participants with a complete or partial response, using National Cancer Institute–Working Group guidelines on CLL. The assessment period extended from treatment administration until specified disease or study events, with a maximum timeframe of up to 24 months. The record does not report the protocol’s original sample-size justification or other original design assumptions.

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: A total of 182 participants were screened and 159 participants were randomized.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent-to-treat (ITT) population includes all participants who were randomized into the 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.

Population: Intent-to-treat (ITT) Population includes all participants who were randomized into the study.

6. Primary Endpoint

The registry lists the primary outcome as Objective Response Rate, measured over From treatment administration (Day 1) until disease progression, death, initiation of alternative therapy, withdrawal of consent, or end of study (up to 24 months).

Objective response rate is the proportion of participants who had either a complete response or a partial response. The registry specifies that responses were assessed using National Cancer Institute–Working Group guidelines on CLL. This endpoint reduces response status to a binary outcome for each participant: a qualifying response or no qualifying response. The reported group percentage is the fraction of participants counted as responders, expressed as a percentage.

Response rate is useful in a clinical trial because it describes how often a defined response was observed during the specified assessment period. It is not a measure of response duration, progression-free survival, overall survival, or quality of life. A response proportion also does not indicate how quickly a response occurred; the registry lists time to response separately as a secondary outcome.

The primary endpoint timeframe began at treatment administration and continued until disease progression, death, initiation of alternative therapy, withdrawal of consent, or end of study, up to 24 months. The denominator was the intent-to-treat population, defined in the record as all participants randomized. This choice makes the endpoint population depend on randomization rather than only on treatment completion. The registry record does not describe how missing response assessments or participants without an assessable response were handled, so their treatment in the endpoint calculation should not be assumed.

Each group’s percentage is an estimate based on the participants counted in that group. A confidence interval communicates the uncertainty around that estimate. For comparisons, the difference between two percentages is also uncertain; a confidence interval for the difference indicates which directions and magnitudes remain compatible with the posted summary data. The endpoint is therefore best interpreted together with its uncertainty, rather than as a stand-alone percentage.

7. Results Posted in the Registry

• MEDI-551 2 mg/kg + Bendamustine: 52.8% (19/36) versus 59.7% (37/62) on control; risk difference -6.9 percentage points (95% CI -27.3 to 13.5), risk ratio 0.88 (0.61–1.28); unadjusted two-proportion z-test p = 0.506.
• MEDI-551 4 mg/kg + Bendamustine: 63.9% (39/61) versus 59.7% (37/62) on control; risk difference 4.3 percentage points (95% CI -12.9 to 21.4), risk ratio 1.07 (0.81–1.41); unadjusted two-proportion z-test p = 0.627.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Complete Response Rate

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • MEDI-551 2 mg/kg + Bendamustine: 5.6% (2/36) versus 6.5% (4/62) on control; risk difference -0.9 percentage points (95% CI -10.6 to 8.8), risk ratio 0.86 (0.17–4.47); unadjusted two-proportion z-test p = 0.858.
  • MEDI-551 4 mg/kg + Bendamustine: 11.5% (7/61) versus 6.5% (4/62) on control; risk difference 5.0 percentage points (95% CI -5.0 to 15.1), risk ratio 1.78 (0.55–5.77); unadjusted two-proportion z-test p = 0.329.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between MEDI-551 2 mg/kg + Bendamustine and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between MEDI-551 4 mg/kg + Bendamustine and the control. The interval gives the range of effects that remain compatible with the data.

Minimal Residual Disease Negative Complete Response (CR) Rate

• Endpoint type: Binary: the percentage of participants with the event (counts rebuilt from the posted percentages and group sizes).
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • MEDI-551 2 mg/kg + Bendamustine: 5.6% (2/36) versus 1.6% (1/62) on control; risk difference 3.9 percentage points (95% CI -4.2 to 12.1), risk ratio 3.44 (0.32–36.67); unadjusted two-proportion z-test p = 0.275.
  • MEDI-551 4 mg/kg + Bendamustine: 4.9% (3/61) versus 1.6% (1/62) on control; risk difference 3.3 percentage points (95% CI -3.0 to 9.6), risk ratio 3.05 (0.33–28.51); unadjusted two-proportion z-test p = 0.301.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between MEDI-551 2 mg/kg + Bendamustine and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between MEDI-551 4 mg/kg + Bendamustine and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants Who Developed Detectable Anti-drug Antibodies (ADA)

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • MEDI-551 4 mg/kg + Bendamustine: 14.0% (8/57) versus 12.9% (4/31) on control; risk difference 1.1 percentage points (95% CI -13.7 to 16.0), risk ratio 1.09 (0.36–3.33); unadjusted two-proportion z-test p = 0.883.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between MEDI-551 4 mg/kg + Bendamustine and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists a range of secondary outcomes, including complete response, minimal residual disease-negative complete response, time to response, time to disease progression, progression-free survival, and safety outcomes. The posted results therefore cover outcomes that describe response depth, timing, disease control, and adverse events, rather than only whether any complete or partial response occurred. These outcomes answer different questions and should not be treated as interchangeable measures of benefit or risk.

The posted binary secondary outcomes generally involve relatively few observed events, and their confidence intervals are wide. The estimates across the treatment groups do not provide a consistent, precise pattern that would independently settle the primary comparison. Some point estimates are higher in a MEDI-551 group and others are not, but point-estimate variation is expected when each outcome is uncertain. Detailed per-outcome interpretation should remain tied to the outcome definition, its denominator, and its own interval.

The primary endpoint remains the principal efficacy outcome named in the registry record. Secondary results may add context, such as whether observed responses were complete or accompanied by the specified residual-disease status, but they do not replace the primary analysis. Time-to-event endpoints address duration or progression-related questions and need methods appropriate to follow-up and censoring; their detailed estimates and analysis methods are not included in the registry record for this explanation.

Unadjusted secondary results carry less weight when many outcomes and comparisons are examined. Multiple testing increases the chance of a seemingly notable result arising by chance. The registry record does not report a multiplicity correction, a testing hierarchy, or which secondary outcomes were designated confirmatory. As a result, secondary findings should be considered supportive or exploratory rather than conclusive evidence of a treatment difference. The record also does not report whether any secondary analyses were prespecified in a detailed statistical analysis plan.

Safety outcomes have a different interpretive role from efficacy outcomes and are described separately. They reflect events recorded in the safety populations, which are not identical in size to the primary intent-to-treat groups. This distinction matters when comparing a percentage or count across endpoints: the population and observation rules may differ. The registry record does not provide enough information to infer why these denominators differ.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Number of Participants With Treatment Emergent Adverse Events (TEAEs), Treatment Emergent Serious Adverse Events (TESAEs) and Adverse Events of Special Interest (AESIs)
• Number of Participants With Abnormal Clinical Laboratory Parameters Reported as AEs
• Number of Participants With Abnormal Vital Signs and Electrocardiogram Reported as AEs
• Complete Response Rate
• Minimal Residual Disease Negative Complete Response (CR) Rate
• Time to Response
• Time to Disease Progression (TTP)
• Progression Free Survival (PFS)

9. Statistical Interpretation

10. Statistical Methods Explained

Objective response rate is a binary outcome summarized as a percentage. Each participant is classified according to whether a complete or partial response was recorded under the stated CLL guidelines. The group percentage is an estimate of the response proportion for that group. This format is straightforward to interpret, but it does not show the timing or duration of response and does not capture differences in outcomes among participants who did not meet the response definition.

What does a confidence interval express?

A confidence interval describes the precision of an estimate under the statistical method used to construct it. For a group response percentage, a wider interval indicates that the estimate is less precise. The intervals depend on both the observed proportion and the amount of information available. They are not a range that contains a known, fixed proportion with a particular probability after the data have been observed. Rather, the confidence level describes the long-run behavior of the interval procedure under its assumptions.

The registry reports 95% confidence intervals for each group’s primary response percentage. The interval endpoints should be read with the point estimate: a percentage near the middle of a broad interval is less precisely estimated than a percentage accompanied by a narrow interval. A confidence interval is also not a test of clinical importance. Clinical importance requires a meaningful threshold or context, neither of which is specified for the primary response comparison in the registry record.

How are group comparisons different from group estimates?

A difference between two response proportions is a contrast between groups. Its estimate depends on both group proportions, and its uncertainty depends on the information in both groups. A confidence interval for a difference that includes no difference indicates that the summary data do not rule out no difference under the calculation method. If the interval also permits values on either side, then a direction of effect is not established by that interval alone.

The recalculated comparisons are unadjusted re-calculations based on the posted summary statistics. They are useful as a transparent, simple comparison, but they do not reproduce any adjustment or stratification that may have been included in the sponsor’s analysis. The recalculated p-values are therefore not substitutes for the sponsor’s reported result. Since the registry record does not supply all analysis details, a complete independent reproduction is not possible from the record alone.

What does a p-value say?

A p-value describes how unusual data at least as inconsistent with a specified null hypothesis would be if that hypothesis and the test assumptions held. It is not the probability that the null hypothesis is true, and it is not a measure of the probability that a treatment is effective. It also does not indicate whether an observed difference is important to patients. A p-value should be considered with the estimated effect, its confidence interval, the outcome definition, the analysis method, and the design.

The recalculated primary comparison p-values are 0.506 for MEDI-551 2 mg/kg plus bendamustine versus rituximab plus bendamustine, and 0.627 for MEDI-551 4 mg/kg plus bendamustine versus rituximab plus bendamustine. These unadjusted values are calculated from the posted summaries, not reported as the sponsor’s primary analysis. The sponsor’s reported analysis of the 4 mg/kg group used the Cochran-Mantel-Haenszel method. This method can test association between treatment group and a binary outcome while accounting for stratification factors, but the registry record does not identify the strata or report enough detail to reproduce that analysis.

Why can adjusted and unadjusted results differ?

An unadjusted comparison treats the groups as aggregate sets. A stratified method evaluates the association within strata and combines information across them. If the study used stratification, or if the analysis otherwise accounts for variables not represented in the aggregate recalculation, estimates or p-values may differ. The record does not provide the sponsor’s full statistical analysis plan, so the exact source of the difference cannot be determined from the registry record.

Randomization is intended to support a fair comparison of assigned groups. It does not guarantee that every analysis method will produce identical results, especially when analyses use different assumptions or handle stratification differently. The distinction between an intent-to-treat population and a treatment-received population also matters. The primary endpoint is defined for all randomized participants, but the record does not detail whether every participant had an evaluable endpoint or how missing results were handled.

Why does multiplicity matter?

Multiplicity arises when several outcomes, comparisons, or analyses are examined. With more statistical tests, the chance of at least one apparently unusual result can increase even when there is no underlying difference for any test. A prespecified testing hierarchy or adjustment can help control that issue. The registry lists multiple secondary outcomes and comparisons, but does not report a multiplicity adjustment or a hierarchy for interpreting them. Accordingly, isolated secondary p-values should be considered exploratory unless the trial’s analysis plan establishes otherwise.

The methods available in the registry record describe the posted outcomes but do not fully document the original design and analysis choices. The original sample-size assumptions, interim analyses, missing-data methods, and complete statistical analysis plan are not reported in the registry record. General statistical explanations above describe how common methods are interpreted; they should not be mistaken for confirmation that every such procedure was used in this trial.

11. Confidence Intervals and Precision

For the primary endpoint, the 95% confidence interval was 46.4% to 71.9% for rituximab plus bendamustine, 35.5% to 69.6% for MEDI-551 2 mg/kg plus bendamustine, and 50.6% to 75.8% for MEDI-551 4 mg/kg plus bendamustine. Each interval spans a substantial range of response proportions, so the group estimates are not highly precise. The interval for the 2 mg/kg group is especially broad in relation to its point estimate, consistent with the smaller group size recorded for that arm.

Confidence intervals for group proportions and confidence intervals for group differences answer related but distinct questions. A group interval describes uncertainty in that group’s response proportion. A comparison interval describes uncertainty in the contrast between groups. The recalculated risk-difference intervals cross no difference for both MEDI-551 comparisons, and allow effects in either direction. This means the data do not pin down the direction or size of the comparison with precision.

Interval overlap among group estimates is a useful visual cue about uncertainty, but it is not a replacement for a direct comparison test. Conversely, a p-value alone cannot show the range of differences compatible with the data. The posted estimates, comparison results, and sponsor analysis should therefore be read together. The record does not report a clinically important response-rate difference threshold, so statistical precision cannot be translated into a conclusion about whether a difference would matter clinically.

Precision is influenced by the information available for each outcome. It can be limited when the group is relatively small or when the observed outcome is uncommon. General statistical principle: more information usually narrows uncertainty, while sparse outcomes can produce unstable estimates. These considerations are relevant to the secondary outcomes as well, but their uncertainty should be assessed from their own estimates and intervals rather than inferred directly from the primary endpoint.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size display is a planning illustration based on the observed effect, not the trial’s original sample-size design. For this binary primary endpoint, the entries are participants per arm. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect. Time-to-event displays, if present, use total events rather than participants per arm; that distinction does not apply to the primary binary endpoint.

The column for 80% power gives the number of participants per arm in a hypothetical study that would have an 80% chance of detecting the stated effect under those assumptions, if that effect and the assumptions were correct. The 90% power column describes a higher chance of detecting that effect under the same planning framework, and therefore generally requires more participants per arm. These are prospective planning quantities; they do not describe the chance that the completed study detected a real effect.

Rows that plan for a smaller fraction of the observed effect require larger participant numbers. This is because a smaller difference between groups is harder to distinguish from random variation than a larger difference, all else equal. The required numbers can become very large when the observed effect used as a starting point is small or uncertain. The observed effect in a modest study is itself imprecise, so it should not be treated as a reliable target for a future trial without further scientific justification.

The displayed figures are not observed power. Observed power calculated from the result is not a useful substitute for the confidence interval and does not add information beyond the effect estimate and its uncertainty. The planning figures also do not establish what enrollment was intended for this trial: the original sample-size justification and design assumptions are not reported in the registry record. The display should be read only as an illustration under its stated assumptions, not as the actual design rationale or a recommendation for a future study.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.597; treatment proportion p₁ = 0.528; average p̄ = 0.562.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7016 = 1.3751.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6999 = 0.5891.
4. n = (1.3751 + 0.5891)² / (-0.0690)² = 810.4, rounded up to 811 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.596774193548387                          # control proportion
p1 <- 0.527777777777778                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(1466153)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 811 (80% power), 1,085 (90% power)
 75% of observed effect: n per arm = 1,435 (80% power), 1,921 (90% power)
 50% of observed effect: n per arm = 3,213 (80% power), 4,301 (90% power)
power.prop.test: 810.4 
Simulated power at n = 811 per arm: 0.797 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry reports a sponsor analysis comparing rituximab plus bendamustine with MEDI-551 4 mg/kg plus bendamustine for the primary endpoint. It identifies the method as Cochran-Mantel-Haenszel and gives a reported p-value. That analysis is distinct from the unadjusted recalculation based on the aggregate response counts and group sizes. The sponsor’s method may account for stratification, while the recalculation does not.

The unadjusted recalculated p-value for the 4 mg/kg comparison is 0.627. The sponsor’s reported p-value is different. This difference does not, on its own, establish that either calculation is erroneous: methods can yield different results when they use different assumptions or account for different features of the study. The registry record does not report the strata, the full model specification, or enough detail to reproduce the sponsor’s analysis.

For the 2 mg/kg comparison, the registry record does not list a sponsor-reported analysis in the posted primary analyses. The unadjusted recalculation for that contrast therefore remains a separate descriptive calculation rather than a reproduction of an official sponsor result. Neither recalculated test should be represented as the complete prespecified trial analysis.

The most informative reading keeps the sponsor’s reported inferential analysis, the posted response estimates and confidence intervals, and the transparent unadjusted recalculations conceptually separate. The sponsor result addresses the analysis recorded by the study; the recalculation provides a simple check based on summary data. Without the full analysis plan and strata details, the registry record does not support a more detailed reconciliation.

15. Multiplicity, Missing Data and Analysis Populations

The registry record reports a primary endpoint and multiple secondary outcomes, with more than one treatment comparison for several binary outcomes. Examining many outcomes or contrasts creates a multiplicity problem: even if there were no true effects, repeated testing can produce apparently unusual results by chance. Prespecified procedures, such as a testing hierarchy or adjustment to significance thresholds, can address this issue, but they also need to be interpreted according to the plan set before results were known.

A multiplicity adjustment or formal hierarchy is not reported in the registry record. The sponsor’s analysis description for the reported primary contrast identifies the Cochran-Mantel-Haenszel method, but the record does not state how the family of outcomes and comparisons was handled. It would therefore be inappropriate to treat isolated secondary p-values as confirmatory or to assume that the nominal error rate applied to every result without adjustment.

The primary outcome population is specified as intent to treat: all participants randomized into the study. That definition preserves the randomized comparison in principle. However, the registry record does not explain how missing response assessments, withdrawals, initiation of alternative therapy, or other endpoint interruptions were handled in the analysis. The endpoint timeframe lists these events as stopping conditions, but that does not by itself reveal the statistical handling of participants without an observed response result.

The registry also does not report the complete analysis populations for every outcome, the original sample-size justification, interim-analysis procedures, or the full statistical analysis plan. The safety results use denominators that differ from those of the primary endpoint, but the reasons and inclusion criteria are not reported. These omissions limit assessment of whether the posted analyses match the planned analyses and whether missingness could have influenced the results.

These gaps do not prove that a particular method was or was not used. They identify what cannot be established from the registry record. Interpretation should distinguish the stated endpoint definition and posted results from unreported choices about multiplicity, missing data, and analysis populations.

16. 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 2% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: infusion related reaction (15.2%), pneumonia (12.1%), febrile neutropenia (10.0%), sepsis (5.0%), vomiting (3.3%).

The registry reports serious adverse events for safety groups with 60 participants at risk in the rituximab plus bendamustine group, 33 in the MEDI-551 2 mg/kg plus bendamustine group, and 57 in the MEDI-551 4 mg/kg plus bendamustine group. The numbers of participants affected by serious adverse events were 19, 16, and 19, respectively. These are participant counts, not counts of every serious event; one participant may experience more than one event, and the registry record does not provide event-level totals in this summary.

For other adverse events, the record reports 58 of 60, 31 of 33, and 54 of 57 participants affected in the respective groups. The safety denominators differ from the primary endpoint group sizes. The record does not explain the difference, so these safety figures should not be interpreted as if they used the intent-to-treat denominators.

The registry lists infusion related reaction at 15.2%, pneumonia at 12.1%, febrile neutropenia at 10.0%, sepsis at 5.0%, vomiting at 3.3%, and abdominal lymphadenopathy at 3.0% among the top serious terms. The record does not specify in this summary the denominator or treatment-group distribution for these terms. Death counts are not reported in the registry record.

These counts describe events recorded during the study; they do not establish comparative safety or causation. Differences in exposure, observation, and event definitions can affect adverse-event comparisons, and those details are not fully available in the registry record.

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

The posted results show objective response rate estimates of 59.7% for rituximab plus bendamustine, 52.8% for MEDI-551 2 mg/kg plus bendamustine, and 63.9% for MEDI-551 4 mg/kg plus bendamustine. The confidence intervals are broad and overlap. The simple recalculated comparisons have uncertainty spanning both directions around no difference, and their unadjusted p-values do not provide strong evidence against a null difference under those calculations.

These observations do not establish that the regimens have equal effects. Failure to demonstrate a difference is not the same as evidence of equivalence or noninferiority. Those conclusions require an appropriate design, prespecified margins, and analyses suited to the question. Such margins and a noninferiority or equivalence design are not reported in the registry record.

Nor do the results establish that one regimen is superior. Although the 4 mg/kg group has a higher point estimate than the comparator, the uncertainty around the comparison includes the possibility of no difference and effects in either direction. The estimate for the 2 mg/kg group is lower than the comparator estimate, but its uncertainty likewise prevents a precise conclusion about the direction and size of the difference.

Objective response rate addresses whether a complete or partial response was recorded during the specified period. It does not establish how long a response lasted, how long participants remained free of progression, or what happened to survival or quality of life. Those outcomes are distinct, and their results are not presented in the primary results. Treatment assignment was open label, which may be relevant to assessment processes, although the registry does not report whether response assessment was independently reviewed or blinded.

The study was phase 2 and included adults with relapsed or refractory CLL described as not eligible for autologous stem cell transplant. Its findings should be understood in that study context. The registry record alone does not support extending the observed estimates to populations or settings not described in the study, nor does it provide enough detail to resolve all issues concerning missing data, multiplicity, or the sponsor’s analysis.

Statistically, the main message is uncertainty rather than a definitive direction of effect. The group percentages are estimates, not exact underlying response rates. Their intervals and the comparison analyses are necessary to understand how much the data constrain possible differences. No medical treatment recommendation follows from these statistical results.

18. Why This Trial Matters Statistically

This trial illustrates why a randomized comparison should be read through both its point estimates and their uncertainty. The primary response percentages give a simple description of each group, but the confidence intervals show that each estimate remains uncertain. The comparison-specific intervals make clear that the observed ordering of point estimates does not settle the direction or size of a treatment difference.

It also illustrates the distinction between a sponsor-reported analysis and a recalculation from aggregate summaries. The sponsor reported a Cochran-Mantel-Haenszel analysis for the comparison involving MEDI-551 4 mg/kg, while the unadjusted recalculation uses only posted group summaries. Because the record omits the stratification details and full plan, the two results cannot be fully reconciled from the registry entry alone. Reporting both kinds of result without conflating them helps readers understand what each can and cannot establish.

The several secondary outcomes provide broader descriptive coverage of response depth, disease progression, timing, and safety. At the same time, multiple outcomes and comparisons make multiplicity and prespecification important. If the hierarchy and adjustment strategy are unknown, secondary p-values should not be treated as a collection of independent confirmatory tests.

Finally, the registry shows why analysis populations and denominators matter. The primary endpoint is defined for all randomized participants, whereas the safety denominators differ. Without details about the populations and missing-data rules, a reader cannot assume that every outcome has the same analytic basis. Understanding these distinctions is central to responsible interpretation of clinical trial statistics.

19. 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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

The registry record does not report the full statistical analysis plan, original sample-size justification, randomization ratio, sequence-generation details, allocation concealment, or the strata used in the sponsor’s Cochran-Mantel-Haenszel analysis. This limits the ability to verify prespecification and reproduce the sponsor’s inferential result. The registry also does not report whether interim analyses were planned or conducted.

The endpoint definition gives a timeframe and response criteria, but the record does not explain how missing assessments or participants with no evaluable response were handled. It does not state whether response assessment was independently reviewed or blinded, which may be relevant in an open-label study. The record also does not provide detailed follow-up distributions or clarify the handling of participants who stopped treatment or began alternative therapy.

Several secondary outcomes and multiple treatment comparisons are listed, but a multiplicity strategy is not reported. This limits the confirmatory interpretation of secondary findings. For some outcomes, the event counts are sparse and their confidence intervals broad, which constrains precision. Detailed results for all time-to-event outcomes and their censoring methods are not included in the registry record.

Safety results have different denominators from the primary endpoint, and the record does not explain the populations or exposure rules underlying those denominators. The top serious adverse-event terms are listed with percentages, but their denominator and treatment-group distributions are not given in this summary. Death counts are not reported in the registry record.

These limitations concern what the registry record documents. They do not show that the study failed to use sound procedures; they mean those procedures cannot be assessed from the available record. The results should be interpreted as posted, with caution about claims that require information the record does not report.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Phase 2, Multicenter, Open-label Study of MEDI-551 in Adults With Relapsed or Refractory Chronic Lymphocytic Leukemia (CLL) — NCT01466153.
• Registry record last updated 2017-05-31; status last verified 2017-04.
• Results first posted 2017-05-31.
• 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.

23. Record Summary

NCT01466153 is a completed, randomized, open-label, phase 2, multicenter, parallel-group study of MEDI-551 plus bendamustine versus rituximab plus bendamustine in adults with relapsed or refractory CLL. Enrollment is recorded as 183. The primary endpoint was objective response rate, defined as complete or partial response in the intent-to-treat population.

Posted response estimates were 59.7% for rituximab plus bendamustine, 52.8% for MEDI-551 2 mg/kg plus bendamustine, and 63.9% for MEDI-551 4 mg/kg plus bendamustine. Their confidence intervals are broad, and the unadjusted recalculated comparisons do not establish a clear difference. The sponsor reported a Cochran-Mantel-Haenszel analysis for the 4 mg/kg comparison; its full details are not reported in the registry record.

Secondary outcomes include response-depth, disease-progression, timing, and safety measures. Multiplicity procedures, missing-data handling, and several details of the original analysis plan are not reported in the registry record.