Comparison Study of Rituximab Plus Sargramostim to Rituximab Alone for Relapsed Follicular B-cell Lymphoma, a Form of Non-Hodgkin's Lymphoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Comparison Study of Rituximab Plus Sargramostim to Rituximab Alone for Relapsed Follicular B-cell Lymphoma, a Form of Non-Hodgkin's Lymphoma (NCT00308087), a randomized phase 2 study. The primary outcome is Number of Participants With a Complete Response or Unconfirmed Complete Response at Week 8 With Confirmation at Week 12.

2. Clinical Question

3. Trial Design

This was a randomized, open-label, parallel-group phase II trial. Participants were assigned to rituximab alone or rituximab plus sargramostim. The registry lists enrollment as 75. In the posted primary outcome, the rituximab arm has 37 participants and the combination arm has 38. The registry does not report the allocation sequence method, allocation concealment, or treatment masking; masking is recorded as none.

The study is listed as terminated. Its start is recorded as 2006-05 and its primary completion as 2009-06. The reason for termination and whether it affected recruitment, follow-up, or analysis are not reported in the registry record. The sponsor is Genzyme, a Sanofi Company.

The primary endpoint is a participant count: complete response or unconfirmed complete response at week 8, with confirmation at week 12. The registry identifies the primary outcome population as intent to treat. The record lists 6 outcome measures and 3 analyses. Secondary outcomes include treatment-emergent adverse events, best response across visits, progression-free survival, duration of partial response or better, and cost effectiveness. Results for those secondary outcomes are not included in the registry record for this analysis, so their findings cannot be described. Details such as follow-up completeness, protocol changes, 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.

Before assignment: 82 patients screened

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Intent to treat population

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 Number of Participants With a Complete Response or Unconfirmed Complete Response at Week 8 With Confirmation at Week 12, measured over Week 8 (confirmed at Week 12).

The primary endpoint counts participants who achieved a complete response or an unconfirmed complete response at week 8 and had that response confirmed at week 12. The registry’s description defines complete response as the complete disappearance of all detectable clinical and radiological evidence of disease. Thus, this endpoint concerns a response assessment at a specified time, with confirmation required at a later assessment; it is not a measure of how long a response lasted.

A participant count is a binary-outcome summary: each participant is either counted as meeting the endpoint or is not counted as meeting it. Reporting the count alongside the number assessed in each arm helps readers understand how much information supports the comparison. Here, the registry reports 1 of 37 participants in the rituximab arm and 3 of 38 in the combination arm.

Response at a defined time is used to compare the observed frequency of this outcome between treatment groups. It does not, on its own, describe progression-free survival, duration of response, safety, or cost effectiveness. Those are distinct outcomes. The registry lists several such secondary measures, but their results are not reported in the registry record. The record does not describe the criteria for an unconfirmed complete response beyond the endpoint wording and confirmation requirement.

7. Results Posted in the Registry

• Rituximab + Sargramostim: 7.9% (3/38) versus 2.7% (1/37) on control; risk difference 5.2 percentage points (95% CI -4.8 to 15.2), risk ratio 2.92 (0.32–26.83); unadjusted two-proportion z-test p = 0.317.

8. Statistical Interpretation

9. Statistical Methods Explained

For the primary endpoint, each participant contributes a binary result: they either meet the response definition and confirmation requirement or they do not. The arm-level counts are 1 of 37 for rituximab and 3 of 38 for rituximab plus sargramostim. Comparing these proportions is one way to describe how often the endpoint occurred in each group. The registry also reports a count-based endpoint, rather than a continuous measurement such as a laboratory value.

What does the risk difference express?

The risk difference is an absolute contrast between the proportions meeting the endpoint in the two arms. The recalculated estimate is 0.052. Its confidence interval ranges from -0.049 to 0.152. The sign indicates the direction of the contrast under the calculation’s arm ordering; the interval spans both negative and positive values. In general, an absolute difference can be easier to interpret as a change in event frequency than a relative measure, but the uncertainty interval remains essential to interpretation.

How do relative measures differ?

The relative risk compares the event proportions as a ratio, while the odds ratio compares the event odds. The recalculated relative risk is 2.921, with an interval from 0.318 to 26.83; the odds ratio is 3.086, with an interval from 0.306 to 31.1. These are relative scales, so their values should not be treated as absolute increases in response. General statistical principle: relative estimates can appear large when event counts are small, and their intervals can be wide when the data contain little information about the effect.

What does a confidence interval communicate?

General explanation: a confidence interval expresses the uncertainty of an estimate under the method used to calculate it. A wider interval indicates less precision. It does not list all possible individual outcomes, and it does not mean that every value inside the interval is equally plausible. The interval should be considered together with the estimate, the endpoint definition, and the observed counts. Here, the broad intervals reflect limited precision around the comparative estimates, while the risk-difference interval includes no difference.

How should the p-values be read?

General explanation: a p-value describes how compatible the observed data are with a specified null hypothesis, under the assumptions of a particular statistical test. It is not the probability that the null hypothesis is true, and it is not a measure of treatment benefit. The registry reports a 2-sided Exact Conditional Test with p = 0.618. The simple unadjusted recalculation reports p = 0.317. Because the methods and reported values differ, they should be identified separately rather than blended into a single result. Details needed to reproduce the registry analysis are not reported in the registry record.

Why can over-dispersion matter?

General explanation: over-dispersion means that observed variability is greater than a model assumes. If a model assumes less variability than the data contain, uncertainty may be understated and intervals may be too narrow. The primary endpoint is a binary count, and no over-dispersion assessment or model-based adjustment is reported in the registry record. It would therefore be inappropriate to claim that over-dispersion was assessed or handled.

The outcome population is identified as intent to treat, but the registry record does not explain how missing or unevaluable response assessments were handled, or whether every randomized participant was included in the primary analysis. It also does not report adjustment for baseline characteristics, the full analysis specification, or multiplicity handling. These omissions limit the ability to reproduce or fully assess the analysis. General principle: such choices matter because they can affect which observations enter an estimate and how uncertainty or false-positive risk is characterized.

10. Confidence Intervals and Precision

The primary endpoint was uncommon in the posted results: 1 of 37 participants in the rituximab arm and 3 of 38 in the combination arm met the response definition. The small event counts mean the comparative estimates are imprecise. The recalculated risk difference is 0.052, with a 95% confidence interval from -0.049 to 0.152. That interval includes zero, so the unadjusted estimate is compatible with no absolute difference as well as with effects in either direction.

The relative estimates are also uncertain. The relative risk is 2.921, with a 95% confidence interval from 0.318 to 26.83, and the odds ratio is 3.086, with a 95% confidence interval from 0.306 to 31.1. These intervals extend across a wide range of values, including values below and above the no-association value for a ratio. Their breadth is a reminder that a point estimate can look substantial while remaining poorly determined by sparse data.

In general, a confidence interval describes uncertainty associated with the estimate under its calculation method; it is not a guarantee about the true effect. Precision is distinct from statistical significance and from clinical importance. The posted estimates do not pin down the size or direction of the underlying comparative effect with confidence. The registry’s own reported p-value is from a different stated analysis, the 2-sided Exact Conditional Test, and is p = 0.618. The record does not give a confidence interval for that analysis. Neither the interval nor the p-value resolves the limitations created by the small number of observed primary endpoint events.

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 planning figures are participants per arm, not total events. They are illustrations based on the observed effect, not the trial’s original design calculation. The assumptions listed for this illustration are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; each planning row uses the stated fraction of the observed effect as the effect to detect. The original design assumptions are not reported in the registry record.

The 80% power column gives the planned participants per arm for a hypothetical study designed to have an 80% chance of detecting the chosen effect under those assumptions, if that effect were present. The 90% power column gives the corresponding per-arm planning figure for a higher target chance of detection. These are prospective planning concepts, not descriptions of the observed trial result. The figures are 292 participants per arm for planning at 100% of the observed effect, 458 for 75% of the observed effect, and 894 for 50% of the observed effect at 80% power. At 90% power, the corresponding figures are 390, 613, and 1,196 participants per arm.

The pattern reflects a general principle: smaller effects are harder to distinguish from random variation, so detecting them generally requires more participants. These planning values are not observed power.

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 = 0.027; treatment proportion p₁ = 0.079; average p̄ = 0.053.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.3168 = 0.6209.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.3147 = 0.2648.
4. n = (0.6209 + 0.2648)² / (0.0519)² = 291.0, rounded up to 292 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 <- 0.027027027027027                          # control proportion
p1 <- 0.0789473684210526                          # 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(308087)
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 = 292 (80% power), 390 (90% power)
 75% of observed effect: n per arm = 458 (80% power), 613 (90% power)
 50% of observed effect: n per arm = 894 (80% power), 1,196 (90% power)
power.prop.test: 291 
Simulated power at n = 292 per arm: 0.824 (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. Analyses Reported by the Sponsor

The registry reports the sponsor’s own primary analysis as a 2-sided Exact Conditional Test, with p = 0.618. The unadjusted recalculation reports a risk difference of 0.052, a 95% confidence interval from -0.049 to 0.152, and p = 0.317. These figures come from different stated analyses and should not be presented as if they were one test result.

The recalculation is a simple unadjusted comparison from the posted arm-level summary statistics. The registry’s reported analysis identifies an exact conditional method, but the detailed calculation specification and accompanying estimate or confidence interval are not reported in the registry record. The difference between the two p-values should not be attributed to a particular analytical choice without those details.

The discrepancy flag is false. The registry record does not mark a discrepancy between the recalculation and the reported analysis. Readers should retain the distinction between the registry-reported test and the separate unadjusted summary, rather than treating one as a correction of the other. Neither p-value alone expresses the magnitude or precision of the effect; the recalculated interval is broad, and the primary endpoint counts are sparse.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists 6 outcome measures and 3 analyses. The listed secondary outcomes include treatment-emergent adverse events, best response across all visits, Kaplan-Meier estimates of progression-free survival, duration of partial response or better, and cost effectiveness. The registry does not report a multiplicity adjustment or explain how the primary and secondary outcomes were prioritized in the analysis. It is therefore not possible to determine from this record whether or how the possibility of chance findings across multiple outcomes was addressed.

The primary outcome population is identified as intent to treat. That label describes the stated analysis population, but the registry record does not provide a detailed definition of which participants were included, how assignment was used in analysis, or how missing response assessments were handled. It does not report the number of missing or unevaluable assessments, imputation methods, or sensitivity analyses for missing data. Those details matter because the denominator and assumptions about unavailable outcomes can affect an estimated response proportion.

The record also does not describe whether there were other analysis populations, such as a per-protocol population, or whether the primary analysis was adjusted for baseline factors. The stated registry analysis is a 2-sided Exact Conditional Test, but other elements of a full statistical analysis plan are not reported in the registry record. The timing and rationale for any interim analyses are not reported in the registry record. These are limits on what can be evaluated from the posted information, not evidence that a particular procedure was or was not used.

15. 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: anaemia (2.7%), myocardial infarction (2.7%), asthenia (2.7%), gastroenteritis (2.7%), pneumonia (2.7%).

The registry reports serious adverse events affecting 5 of 37 participants in the rituximab arm and 7 of 38 in the rituximab-plus-sargramostim arm. For other adverse events, the corresponding reported counts are 33 of 37 and 38 of 38. These figures are counts of participants affected, not event rates calculated for comparison in this explanation. They do not, on their own, establish whether a difference between groups is meaningful.

The record lists Anaemia, Myocardial infarction, Asthenia, Gastroenteritis, Pneumonia, and Hepatic enzyme increased among the top serious adverse event terms, each at 2.7%. The information available does not identify which arm each term occurred in or provide additional event-level context, so no arm-specific interpretation can be made from these entries. Death counts are not reported in the registry record.

Adverse-event counts are descriptive and require context, including definitions, severity, timing, attribution, and duration. Those details are not reported in the information. The registry lists treatment-emergent adverse events as a secondary outcome, but the summary figures alone do not provide a complete safety comparison.

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

The posted primary results show that 1 of 37 participants in the rituximab arm and 3 of 38 in the rituximab-plus-sargramostim arm met the week 8 response endpoint with confirmation at week 12. The unadjusted risk difference estimate is 0.052, but its 95% confidence interval ranges from -0.049 to 0.152. The interval is broad and includes no difference. The relative risk and odds ratio also have wide intervals, indicating substantial uncertainty about the size and direction of the comparative effect.

The registry’s own analysis reports p = 0.618 from a 2-sided Exact Conditional Test. The simple unadjusted recalculation reports p = 0.317. These values reflect distinct analyses and should be described with their methods. Neither value establishes that the treatments are equivalent, and neither gives the probability that a treatment effect exists. The available results do not provide a precise estimate of the comparative effect on this endpoint.

The endpoint is a response assessment at a specified time with confirmation, not a measure of response duration, progression-free survival, or overall safety. Although the registry lists secondary outcomes in those areas, their results are not reported in the registry record for this analysis. The adverse-event summaries are descriptive and likewise do not, by themselves, establish a comparative safety profile.

Accordingly, the posted data describe a small number of confirmed response outcomes and an uncertain unadjusted comparison. They do not support a definitive conclusion about the magnitude of a treatment difference, do not establish equivalence, and do not answer the listed secondary-outcome questions. The trial’s terminated status is recorded, but the reason and its implications for the results are not reported in the registry record.

17. Why This Trial Matters Statistically

This trial illustrates a basic challenge in interpreting randomized comparisons with a binary endpoint: the treatment groups can have different observed counts, while the number of events remains too small to estimate the size of the contrast precisely. Here, the primary outcome counts are 1 of 37 and 3 of 38. The point estimate alone would not convey the uncertainty; the broad confidence intervals make that uncertainty explicit.

It also shows why effect measures need to be read together. The risk difference presents the comparison on an absolute scale, while the relative risk and odds ratio express it on relative scales. Their point estimates and intervals answer related but distinct descriptive questions. With sparse outcomes, relative estimates can be especially unstable, and the interval is more informative than focusing on a single point estimate.

The record also distinguishes a registry-reported test from a simple unadjusted recalculation. The reported exact conditional test has p = 0.618, while the recalculated unadjusted comparison has p = 0.317. Identifying the method beside the result helps avoid treating p-values as interchangeable. The record does not provide enough analytical detail to account for the difference.

Finally, the trial’s registry status and design details should be kept separate from claims about its findings. It is listed as terminated, but the reason is not reported in the registry record. The original design assumptions and details of missing-data or multiplicity handling are also not reported. These omissions constrain statistical interpretation without changing the observed counts.

18. 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 endpoint is based on few observed responses, resulting in broad uncertainty intervals for the unadjusted comparisons. The interval for the risk difference includes zero, and the relative-measure intervals span wide ranges. The posted estimates therefore provide limited precision about the size and direction of the underlying effect.

The registry reports its own primary test as a 2-sided Exact Conditional Test with p = 0.618, while the simple recalculation gives p = 0.317. The record does not report enough detail to reproduce the registry analysis or explain the difference. It also does not provide a confidence interval or effect estimate alongside the reported test result in the information available for this analysis.

The study is listed as terminated, but the reason for termination and any effects on recruitment, follow-up, or analysis are not reported in the registry record. The original sample-size justification and design assumptions are not reported. The planning illustration is based on the observed effect and cannot establish what the trial was originally designed to detect.

Further limitations are that missing response assessments and their handling are not reported, nor are adjustment for multiple outcomes, detailed baseline characteristics, allocation concealment, or the full statistical analysis plan. The registry lists secondary outcomes, but their results are not included in the registry entries. These gaps prevent a fuller assessment of the trial’s analysis and of outcomes beyond the primary endpoint.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Comparison Study of Rituximab Plus Sargramostim to Rituximab Alone for Relapsed Follicular B-cell Lymphoma, a Form of Non-Hodgkin's Lymphoma — NCT00308087.
• Registry record last updated 2013-12-27; status last verified 2013-12.
• Results first posted 2010-08-30.
• 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.

22. Record Summary

The registry record describes a randomized, open-label, parallel-group phase II trial comparing rituximab plus sargramostim with rituximab alone for relapsed follicular B-cell lymphoma. Enrollment is listed as 75, and the study status is terminated. The primary outcome is the number of participants with complete or unconfirmed complete response at week 8, confirmed at week 12, in the intent-to-treat population.

The posted counts are 1 of 37 in the rituximab arm and 3 of 38 in the combination arm. The registry reports a 2-sided Exact Conditional Test with p = 0.618. The separate unadjusted recalculation gives a risk difference of 0.052, with a 95% confidence interval from -0.049 to 0.152. Those estimates are imprecise, and the registry’s other listed outcome results and several analysis details are not reported in the available record.