A Study of Rituximab and Bevacizumab in Patients With Follicular Non-Hodgkin's Lymphoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study of Rituximab and Bevacizumab in Patients With Follicular Non-Hodgkin's Lymphoma (NCT00193492), a randomized phase 2 study. The primary outcome is Overall Response Rate (ORR), the Percentage of Patients Who Experience an Objective Benefit From Treatment.

2. Clinical Question

3. Trial Design

This was a randomized, phase II trial with a parallel-group design and no masking. The registry lists an enrollment of 60. The posted outcome results include 31 participants in the rituximab arm and 29 in the rituximab/bevacizumab arm. These are the outcome denominators recorded for the primary comparison; the analysis population and any difference between enrollment and outcome populations are not reported in the registry record.

The trial compared rituximab alone with rituximab plus bevacizumab. Both arm descriptions specify an initial course of rituximab 375mg/m2 by slow IV infusion weekly for 4 consecutive weeks, on days 1, 8, 15, and 22. In the combination arm, bevacizumab 10mg/kg was scheduled on Days 3 and 15 during this course. The registry description for each arm ends before all later treatment details are complete. The record says that additional rituximab was planned for patients with objective response or stable disease at week 12 reevaluation, but the full schedule is not reported in the registry record.

The study is marked completed. The recorded start is 2005-09, and primary completion is 2013-03. The sponsor is SCRI Development Innovations, LLC. The brief summary frames the study around feasibility, efficacy, and safety in previously treated patients with disease progression. The posted primary result concerns overall response rate, and progression-free survival is listed as a secondary outcome. Further details on randomization procedures, allocation concealment, follow-up conduct, protocol deviations, and the original statistical design assumptions 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

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 Overall Response Rate (ORR), the Percentage of Patients Who Experience an Objective Benefit From Treatment, measured over 18 months.

The primary endpoint was overall response rate, defined as the percentage of participants experiencing an objective benefit from treatment, with an assessment timeframe of 18 months. The registry specifies response assessment using RECIST v1.0 for target lesions assessed by MRI or CT. Complete response required disappearance of all target lesions; partial response required at least a 30% decrease in the sum of the longest diameters of target lesions. Overall response was the combination of complete and partial response.

A response rate turns a clinical assessment into a binary outcome for each participant: a participant either meets the response definition or does not. The arm-level rate then describes the share of participants classified as responders under that definition. This endpoint is used to summarize tumor response over a stated assessment period, rather than the length of time until progression or another event. Progression-free survival is listed separately as a secondary outcome.

The posted response rates were 42.0% for rituximab and 48.0% for rituximab/bevacizumab. The corresponding confidence intervals are wide, so the estimates do not precisely identify the response rate in either arm. The registry record does not report additional details about assessment timing within the 18-month timeframe, independent review, handling of unevaluable participants, or the analysis population.

7. Results Posted in the Registry

• Rituximab/Bevacizumab: 48.3% (14/29) versus 41.9% (13/31) on control; risk difference 6.3 percentage points (95% CI -18.8 to 31.5), risk ratio 1.15 (0.66–2.02); unadjusted two-proportion z-test p = 0.622.

8. Statistical Interpretation

9. Statistical Methods Explained

What kind of outcome is overall response?

Overall response rate is a binary-outcome summary. Each participant is classified according to whether they met the registry’s response definition, and the arm-level percentage represents the fraction meeting that definition. The registry reports 13 responses among 31 participants in the rituximab arm and 14 among 29 in the rituximab/bevacizumab arm. The corresponding posted percentages are 42.0% and 48.0%.

General: a binary endpoint is straightforward to describe, but it compresses information. It records whether the specified response criterion was met, not the degree of change beyond that threshold, the duration of response, or the time to progression. For this endpoint, the response rules are based on RECIST v1.0 and include complete or partial response.

How should the arm percentages be understood?

The response percentage in each arm is an estimate of the response proportion for the participants represented by that arm’s outcome data. Such an estimate varies from sample to sample, especially when the groups are modest in size. A confidence interval expresses the uncertainty around an estimate under the statistical method used to construct it. It is not a range containing a fixed proportion of individual participant outcomes, nor does it guarantee that the underlying response rate lies within the interval.

The registry posts a 95% confidence interval for each arm: 25.0% to 61.0% for rituximab and 29.0% to 67.0% for rituximab/bevacizumab. General: a 95% confidence procedure is designed so that, over repeated studies under its assumptions, intervals constructed in that way would contain the true parameter at the stated rate. The interval’s width gives a practical sense of precision: broader intervals leave more uncertainty about the rate.

What does the between-arm comparison measure?

The comparison is reported as a risk difference, which contrasts the two response proportions by subtraction. The registry’s estimated risk difference is positive, and its confidence interval runs from a negative value to a positive value. Because the interval includes zero, the posted data are compatible with no difference as well as with differences in either direction. The unadjusted p-value is 0.622. This p-value assesses how unusual a result at least this discordant with a zero-difference hypothesis would be under the test’s assumptions; it is not the probability that the null hypothesis is true.

The comparison also includes relative-effect summaries, but their precise values are not necessary to interpret the main result. General: a risk difference describes an absolute contrast in proportions, whereas a risk ratio or odds ratio expresses a relative contrast. These measures answer different questions and can give different impressions when an outcome is common. For this registry result, the risk-difference interval conveys the key point: its uncertainty includes effects on both sides of no difference.

What does “unadjusted” mean?

The comparison is an unadjusted re-calculation from the posted summary statistics. It does not account for additional covariates or stratification factors. The registry does not report a sponsor analysis or a more detailed adjusted model for the primary endpoint. The discrepancy indicator is false, so no conflict between a posted sponsor result and this unadjusted calculation is identified in the record.

General: adjustment can account for measured baseline characteristics or design factors, but it does not automatically eliminate bias and may change precision. Without a stated analysis plan or adjusted result, the unadjusted comparison should not be described as adjusted or as reproducing a more complex trial analysis.

Why does uncertainty matter in a randomized trial?

Randomization helps balance factors between treatment groups on average, but it does not make estimates exact. The observed outcome counts remain subject to sampling variation. Confidence intervals make that uncertainty visible, while a p-value condenses evidence against a specified null model into a single statistic. Neither replaces examination of the effect estimate, its uncertainty, endpoint definition, and the conduct of the trial.

General: p-values do not measure the size or clinical importance of an effect. A large p-value does not demonstrate equivalence; demonstrating equivalence or noninferiority generally requires a prespecified margin and an analysis designed for that question. Such a margin and such an analysis are not reported in the registry record. The trial’s phase II designation and response endpoint do not, by themselves, establish what formal decision rule was used.

What statistical details are absent?

The registry record does not report the original statistical analysis plan, the response-analysis method, how missing or unevaluable outcomes were handled, or the analysis population. It lists no posted analyses. It also does not report the original sample-size justification, planned treatment-effect target, interim analyses, or any stopping rules. These details should not be inferred from the posted outcome summaries.

10. Confidence Intervals and Precision

The response-rate estimates have broad 95% confidence intervals: 25.0% to 61.0% for rituximab and 29.0% to 67.0% for rituximab/bevacizumab. The intervals show that the underlying response proportions are not estimated with high precision. In particular, the observed percentages should not be treated as exact descriptions of the response rates that would occur in a larger or different population.

For the comparison, the risk-difference interval extends below and above zero. This means the posted data do not pin down the direction of the underlying difference. The interval permits a range of possibilities, including a lower rate in the combination arm and a higher rate in that arm. Its width reflects considerable uncertainty in the estimated contrast. The p-value of 0.622 is consistent with that imprecision, but it should not be used as a substitute for considering the confidence interval.

General: confidence intervals are most useful when read as a range of effect sizes compatible with the data under the analysis assumptions. A wide interval may include effects that would be considered meaningfully different in either direction, even when the point estimate is close to zero or a p-value is not small. No clinical threshold for an important response-rate difference is reported in the registry record, so statistical precision cannot be translated into a judgment about clinical importance.

The intervals do not account for every possible limitation. Their interpretation depends on the statistical method and analysis population, neither of which is fully described in the registry record. Details about missing outcome data, response assessment, and any multiplicity strategy are also not reported. The available estimates should therefore be read as uncertain summaries of the posted trial results, rather than as precise predictions for future patients or studies.

11. 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 figures are planning illustrations based on the observed effect, not the trial’s original design justification. The registry record does not report the trial’s own sample-size assumptions. These illustrations use 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.

For this binary-outcome comparison, the figures refer to participants per arm, not a total across arms and not a count of events. The 80% power column represents the number of participants per arm that the planning calculation associates with an 80% chance of detecting the specified effect under its assumptions. The 90% power column uses a higher probability of detection and therefore requires more participants per arm for each effect scenario. These are planning targets, not observed trial results.

The rows consider planning for 100%, 75%, and 50% of the observed effect. For planning for 100% of the observed effect, the figures are 966 participants per arm for 80% power and 1,293 participants per arm for 90% power. For planning for 75% of the observed effect, they are 1,712 and 2,292 participants per arm, respectively. For planning for 50% of the observed effect, they are 3,839 and 5,138 participants per arm, respectively. Smaller effects need more participants because they are harder to distinguish from random variation under the same assumptions. These figures are planning illustrations, not observed power.

The calculations rely on a simplified planning framework and should not be mistaken for the trial’s protocol-based design. The registry does not report whether the original design used these assumptions or a different approach.

12. 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 = 0.419; treatment proportion p₁ = 0.483; average p̄ = 0.451.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7037 = 1.3792.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.7023 = 0.5911.
4. n = (1.3792 + 0.5911)² / (0.0634)² = 965.7, rounded up to 966 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 <- 0.419354838709677                          # control proportion
p1 <- 0.482758620689655                          # 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(193492)
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 = 966 (80% power), 1,293 (90% power)
 75% of observed effect: n per arm = 1,712 (80% power), 2,292 (90% power)
 50% of observed effect: n per arm = 3,839 (80% power), 5,138 (90% power)
power.prop.test: 965.7 
Simulated power at n = 966 per arm: 0.807 (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.

13. Multiplicity, Missing Data and Analysis Populations

The primary endpoint is overall response rate, and progression-free survival is listed as a secondary outcome. The registry reports 2 outcome measures but lists 0 analyses. It does not provide results for the secondary outcome in the posted record. The multiplicity strategy—how the trial handled testing more than one outcome or comparison—is not reported in the registry record.

General: when several outcomes, timepoints, subgroups, or comparisons are tested, the chance of obtaining at least one apparently unusual result can increase if each test is interpreted separately. A prespecified hierarchy, adjustment, or other testing plan can address that issue, but none is described in the registry record. This absence does not establish that no such plan existed; it means the posted record does not report it.

The registry record also does not report the primary endpoint’s analysis population, whether all randomized participants were included, or how missing and unevaluable response assessments were handled. The outcome denominators are 31 in the rituximab arm and 29 in the rituximab/bevacizumab arm, but the record does not explain whether these represent every randomized participant in each group or a subset with evaluable outcome data. It does not state how participants with incomplete follow-up were classified.

Details of randomization generation, allocation concealment, protocol deviations, and analysis by assigned treatment are not reported in the registry record. These details affect how confidently the randomized comparison can be interpreted. Because the study was unmasked, information about independent or masked response review would also help readers assess the potential influence of knowledge of treatment assignment; that information is not reported in the registry record.

These reporting gaps limit evaluation of the statistical analysis but do not justify assuming that a particular method was or was not used. The posted response estimates and their uncertainty can be described, while conclusions about missing-data handling, analysis populations, and multiplicity remain unresolved from the registry information.

14. 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: febrile neutropenia (3.4%), infections and infestations - other, pneumonia (3.4%), wound complication (3.4%), psychiatric disorders - other, dementia (3.4%), hematoma (3.4%).

The serious adverse event summary reports 1 affected participant among 31 at risk in the rituximab arm and 8 among 29 at risk in the rituximab/bevacizumab arm. For other adverse events, the record reports 29 affected participants among 31 at risk in the rituximab arm and 29 among 29 at risk in the combination arm. These categories should be read separately: serious adverse events are not interchangeable with the broader “other adverse events” count.

The listed serious-event terms include Febrile Neutropenia, Infections and infestations - Other, pneumonia, Wound complication, Psychiatric disorders - Other, dementia, Hematoma, and Epistaxis. Each term is listed with 3.4%. The record does not provide deaths for either group; death counts are not reported in the registry record.

These are descriptive registry counts. The record does not provide enough detail to determine event duration, recurrence, severity beyond the serious-event classification, relatedness to treatment, or whether the same participant experienced more than one event. The group counts should not be interpreted as a complete comparison of safety or as evidence of causation. Further adverse-event definitions and collection details are not reported in the registry record.

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

The result shows that the posted overall response rate was 42.0% with rituximab and 48.0% with rituximab/bevacizumab. The interval estimates are wide, and the unadjusted comparison has a p-value of 0.622. The risk-difference interval includes values on both sides of zero. Together, these features indicate that the trial’s posted primary result does not precisely establish the direction or size of a response-rate difference.

This does not show that the regimens have identical response rates. A result that does not provide strong evidence of a difference is not the same as a demonstration of equivalence. The registry record does not report an equivalence or noninferiority margin or a corresponding analysis. Nor does the numerical ordering of the two observed percentages, by itself, demonstrate a reliable benefit from adding bevacizumab.

The endpoint describes whether a participant met the RECIST v1.0 criteria for complete or partial response within the stated 18-month timeframe. It does not describe response duration, progression-free survival, overall survival, or quality of life. Progression-free survival is listed as a secondary outcome, but its results are not reported in the registry record. Consequently, the primary response-rate result cannot stand in for those other outcomes.

The safety record includes serious-event counts and a broader other-adverse-event count, but it does not provide the detail needed to characterize all aspects of comparative safety. The trial result should not be read as a complete benefit-harm assessment. It also should not be generalized beyond the registry’s described population without considering differences in eligibility, prior treatment, assessment, and follow-up; such contextual information is not fully reported.

In statistical terms, the result is an uncertain estimate from a randomized phase II comparison, rather than a definitive statement about the effect of adding bevacizumab. The registry record does not report the full analysis plan, missing-data approach, or the results of the listed secondary outcome. Those limits should remain part of any interpretation of the posted response rates.

16. Why This Trial Matters Statistically

This trial illustrates why a response percentage should be read together with its interval and the between-arm comparison. The observed response rates differ numerically, but the confidence intervals are broad and the risk-difference interval spans both directions. Looking only at the point estimates would obscure how much uncertainty remains about the comparison.

It also illustrates the distinction between a response endpoint and a time-to-event endpoint. Overall response rate records whether a response criterion was met; progression-free survival addresses a different question and is listed as secondary. Since progression-free survival results are not reported in the registry record, the posted primary endpoint cannot answer whether the groups differed on that outcome.

The design is randomized, but the record describes the study as unmasked and does not report allocation procedures or response-review methods. Statistical interpretation depends not only on the outcome count, but also on how participants were assigned, followed, assessed, and included in analysis. The registry leaves several of those details unspecified, making it important to distinguish what the posted estimates show from what cannot be evaluated from the record.

Finally, the planning illustrations demonstrate how strongly required sample size depends on the effect targeted and the desired probability of detecting it. Smaller targeted effects require more participants per arm under the stated assumptions. These are planning calculations based on the observed effect and are not evidence about the power actually achieved by the trial. The original design assumptions are not reported in the registry record.

17. 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 leaves several trial-specific details unresolved. The primary endpoint’s population field is blank, and the analysis population is not stated. Although arm-level denominators are posted, the record does not explain whether all randomized participants contributed to the outcome analysis or how any missing or unevaluable assessments were treated.

The study was unmasked, but the record does not say whether response assessments were subject to independent review or whether assessors were masked. It also does not report the randomization method, allocation concealment, the prespecified analysis plan, or adjustments for baseline characteristics. These omissions restrict assessment of potential sources of bias and of the exact method behind the primary comparison.

The registry lists progression-free survival as a secondary outcome but does not post its results. It reports 2 outcome measures and 0 analyses, so the record does not explain the multiplicity approach or provide broader inferential context. The original sample-size justification and trial-specific power assumptions are also not reported; the available planning figures are illustrations based on the observed effect, not the original design.

The arm descriptions are truncated before the complete treatment schedules are given. The record also does not report detailed timing and completeness of response assessments within the 18-month timeframe. For adverse events, the posted counts do not describe recurrence, relatedness, duration, or the number of participants with multiple events; deaths are not reported in the registry record. These limits do not negate the posted summaries, but they constrain what can be concluded from them.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: A Study of Rituximab and Bevacizumab in Patients With Follicular Non-Hodgkin's Lymphoma — NCT00193492.
• Registry record last updated 2015-01-05; status last verified 2014-12.
• Results first posted 2014-12-12.
• 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.

21. Record Summary

NCT00193492 is a completed, randomized, open-label, parallel-group phase II trial sponsored by SCRI Development Innovations, LLC. The registry lists an enrollment of 60, with a recorded start of 2005-09 and primary completion in 2013-03. It compared rituximab with rituximab plus bevacizumab in previously treated follicular non-Hodgkin's lymphoma.

The primary endpoint was overall response rate at 18 months, defined using RECIST v1.0. The posted rates were 42.0% for rituximab and 48.0% for rituximab/bevacizumab; the unadjusted comparison has a p-value of 0.622, with substantial uncertainty in the effect estimate. Progression-free survival is listed as a secondary outcome, but its result is not reported in the registry record. The record includes adverse-event counts but does not report deaths or several details needed to evaluate analysis methods, missing data, and safety fully.