A Study to Investigate the Efficacy and Safety of Bendamustine Compared With Bendamustine+Obinutuzumab (GA101) in Participants With Rituximab-Refractory, Indolent Non-Hodgkin's Lymphoma (GADOLIN): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Investigate the Efficacy and Safety of Bendamustine Compared With Bendamustine+Obinutuzumab (GA101) in Participants With Rituximab-Refractory, Indolent Non-Hodgkin's Lymphoma (GADOLIN) (NCT01059630), a randomized phase 3 study. The primary outcome is Number of Participants With Progressive Disease (PD) as Assessed by Independent Review Committee (IRC) or Death.

2. Clinical Question

3. Trial Design

This was a randomized, open-label, multicenter, parallel-group phase III study. The registry record lists an enrollment of 413 and a completed study status. The treatment groups were bendamustine alone and obinutuzumab plus bendamustine. Masking was listed as none, so participants and study personnel were not masked to treatment assignment according to the recorded design.

The study summary describes the combination strategy as obinutuzumab with bendamustine followed by continued obinutuzumab treatment as maintenance monotherapy. For the bendamustine-alone group, the recorded regimen was 120 mg/m² by intravenous infusion on Days 1 and 2 of each 28-day cycle, for up to six cycles. The combination-arm description records bendamustine at 90 mg/m², with a schedule variation for the first 10 participants and a separate schedule for the remaining participants. The record’s description of that arm is cut off, so its full treatment schedule is not reported in the registry record.

The listed primary outcome uses the intention-to-treat population and independent review committee assessment. The study summary says the end of study was defined by completion of safety follow-up for all participants, described as 2 years’ follow-up from the last dose. The registry record lists study start as 2010-04-30 and primary completion as 2014-09-30. Randomization procedures, allocation concealment, stratification factors, and the original sample-size justification are not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: ITT 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 Progressive Disease (PD) as Assessed by Independent Review Committee (IRC) or Death, measured over Baseline until PD or death, whichever occurred first (assessed at baseline, 14 days prior to Cycle [Cy] 4 Day 1 [1 Cy=28days], 28-42 days after Cy 6 Day 1, then every 3 months up to 2 years and every 6 months for next 2 years [up to 4.5 years overall]).

The primary endpoint is the number of participants with progressive disease, as assessed by an independent review committee, or death. It covers baseline until the first of those events. The registry lists assessments at baseline, before Cycle 4, after Cycle 6, and then at scheduled intervals extending up to 4.5 years overall. The endpoint is described for the intention-to-treat population.

Progressive disease was assessed using modified response criteria for indolent non-Hodgkin lymphoma. The registry’s definition includes a new lesion greater than 1.5 cm in any axis, specified increases from the smallest recorded lesion measurements, and particular size and growth criteria for lymph nodes. These rules translate observed disease measurements into a progression event; the endpoint therefore depends on both the assessment schedule and the stated criteria.

Counting participants with progression or death provides a direct comparison of the proportion experiencing either event during follow-up. It does not by itself describe when events occurred, how long participants remained event-free, or the timing of censoring. A time-to-event endpoint such as progression-free survival addresses timing as well as event status. The registry record also lists progression-free survival among the secondary outcomes, but does not report a primary-endpoint analysis method.

7. Results Posted in the Registry

• Obinutuzumab + Bendamustine: 42.6% (87/204) versus 59.8% (125/209) on control; risk difference -17.2 percentage points (95% CI -26.7 to -7.7), risk ratio 0.71 (0.59–0.87); unadjusted two-proportion z-test p = <0.001.

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.

Number of Participants With PD or Death as Assessed by Investigator

• 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:
  • Obinutuzumab + Bendamustine: 64.7% (132/204) versus 72.7% (152/209) on control; risk difference -8.0 percentage points (95% CI -16.9 to 0.9), risk ratio 0.89 (0.78–1.01); unadjusted two-proportion z-test p = 0.079.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Obinutuzumab + Bendamustine and the control. The interval gives the range of effects that remain compatible with the data.

PFS as Assessed by Investigator

• Endpoint type: Time to event: compared as a hazard ratio.
• Test: Hazard ratio as reported by the sponsor (Log Rank); it was not recalculated.
• Result:
  • Bendamustine Alone vs Obinutuzumab + Bendamustine: hazard ratio 0.57 (95% CI 0.45–0.73), reported p < 0.001 (as reported in the registry).
• Interpretation:
  • The 95% confidence interval excludes no difference (a ratio of 1), so the hazard of the event is statistically lower for Bendamustine Alone vs Obinutuzumab + Bendamustine than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Participants With Objective Response as Assessed by IRC

• 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:
  • Obinutuzumab + Bendamustine: 75.5% (154/204) versus 77.5% (162/209) on control; risk difference -2.0 percentage points (95% CI -10.2 to 6.2), risk ratio 0.97 (0.88–1.08); unadjusted two-proportion z-test p = 0.628.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Obinutuzumab + Bendamustine and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes add perspectives beyond the primary count assessed by the independent review committee. They include investigator-assessed progression or death, investigator-assessed progression-free survival, and objective response assessed by either an independent review committee or an investigator. These endpoints overlap in subject matter but are not interchangeable: they differ in assessment source, event definition, and whether they capture a status or time-to-event outcome.

The broad pattern is not identical across all measures. Investigator-assessed progression-free survival has a reported hazard ratio whose confidence interval is below 1, while the unadjusted investigator-assessed progression-or-death risk difference has an interval that includes zero. Objective response assessed by the independent review committee has a small unadjusted risk-difference estimate with an interval that includes zero. This does not by itself mean the endpoints conflict. A response can occur without establishing how long it lasts, and an outcome based on time to progression or death carries information about event timing that a count over follow-up does not.

Secondary findings can help describe whether the primary result is accompanied by patterns in related outcomes, but their evidential weight depends on prespecification and multiplicity control. The registry record lists many secondary outcomes and does not report a multiplicity adjustment or a testing hierarchy. When many outcomes are tested, chance alone can produce apparently notable results. Unadjusted secondary p-values therefore deserve less weight than a clearly prespecified primary analysis, and should be interpreted as supportive or exploratory unless the analysis plan establishes otherwise.

The registry’s secondary outcome list includes 31 posted outcomes, but only selected endpoint summaries and analyses are detailed in the record. Analysis populations, missing-data procedures, and the status of each outcome as confirmatory or exploratory are not reported in the registry record. The results are best read as complementary descriptions of disease course, not as independent replications of the primary comparison.

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

• Number of Participants With PD or Death as Assessed by Investigator
• PFS as Assessed by Investigator
• Percentage of Participants With Objective Response as Assessed by IRC
• Percentage of Participants With Objective Response as Assessed by Investigator
• Percentage of Participants With Best Overall Response (BOR) as Assessed by IRC
• Percentage of Participants With Best Overall Response (BOR) as Assessed by Investigator
• Percentage of Participants With BOR at the End of Induction Treatment as Assessed by IRC
• Percentage of Participants With BOR at the End of Induction Treatment as Assessed by Investigator

9. Statistical Interpretation

10. Statistical Methods Explained

The registry describes a randomized parallel-group design, with no masking. Random assignment is generally used to make treatment groups comparable at the start of a trial, reducing the risk that known or unknown baseline characteristics systematically determine treatment received. Open-label assignment means that treatment knowledge could influence behavior, co-interventions, reporting, or assessment. Independent review committee assessment can reduce the influence of treatment knowledge on endpoint classification, but does not remove every possible source of bias. Details of the committee’s procedures are not reported in the registry record.

What does the primary count represent?

The primary outcome is a binary event outcome summarized as a count of participants with progressive disease or death. The registry gives the number with an event and the group size for each arm. A risk is the event count divided by the corresponding group size; the posted unadjusted risk estimates are 0.598 for bendamustine alone and 0.426 for obinutuzumab plus bendamustine. The risk difference is the combination-group risk minus the bendamustine-alone risk. Its reported unadjusted estimate is -0.172. A negative risk difference means the observed event risk was lower in the combination group in this simple comparison.

These are marginal, unadjusted comparisons: they use the posted group summaries without adjustment for other variables. The registry does not report a primary-outcome model or adjusted effect estimate. The listed p-value, <0.001, is also an unadjusted recalculation, not a reported primary analysis. It should be interpreted as a test result under the assumptions of that calculation, not as a measure of effect size or clinical importance.

How is the time-to-event result different?

Progression-free survival is a time-to-event outcome. It accounts for the time until progression or death and can also use information from participants whose follow-up ends before an event is observed, under the analysis’s censoring assumptions. The registry reports median progression-free survival and a hazard ratio using the Log Rank method. A median describes the time by which the estimated event-free proportion has reached its midpoint. A hazard ratio compares event rates over follow-up under a time-to-event framework; it is not the same quantity as a risk difference, and its interpretation depends on the analysis assumptions. The registry does not report the censoring details or a fuller statistical analysis plan.

The reported hazard ratio is 0.57, with a 95% confidence interval from 0.45 to 0.73 and p < 0.001. These values are a reported registry analysis, unlike the primary risk-difference recalculation. They should be read with the stated Log Rank method and with awareness that the record does not provide additional detail about model checks or the handling of missing follow-up.

Why distinguish reported analyses from recalculations?

A recalculation from aggregate event counts is useful for a transparent, simple comparison, but it may not reproduce a trial’s planned or adjusted analysis. The registry gives reported analyses for investigator-assessed progression-free survival and independent-review objective response, but it does not list a reported analysis for the primary outcome. No discrepancy is flagged for the posted results. Where a reported sponsor analysis is available, it is important to keep its method and estimate distinct from a separate unadjusted calculation.

The objective-response result illustrates this distinction. The unadjusted recalculation is a risk difference based on response proportions. The registry’s reported analysis uses the Cochran-Mantel-Haenszel method and gives a result of 0.98, a 95% confidence interval from 0.58 to 1.65, and p=0.930. The method can account for stratification when the analysis is structured around strata, but the strata used are not reported in the registry record. The two sets of statistics should therefore be described separately, not combined into a single estimate.

What can the p-values establish?

In general, a p-value measures how unusual data at least as extreme as those observed would be under a specified null model and its assumptions. It does not give the probability that the null hypothesis is true, the size of a treatment effect, or the chance that a result will replicate. The primary unadjusted p-value and the p-values for secondary outcomes arise from separate comparisons, and the record does not report a multiplicity adjustment. Consequently, unadjusted secondary p-values should be treated as exploratory rather than as a single family of confirmatory tests.

Likewise, a confidence interval communicates estimation uncertainty under the calculation’s assumptions. It does not cover uncertainty about all design, measurement, or analysis choices. The distinction matters because the record supplies endpoint summaries and selected analyses but does not report complete procedures for missing data, multiplicity control, or the primary analysis. The statistical interpretation can therefore be precise about posted values while remaining limited about the full trial analysis plan.

11. Confidence Intervals and Precision

The primary unadjusted risk-difference estimate is -0.172, with a 95% confidence interval from -0.267 to -0.077. In general, a confidence interval is a range of estimates compatible with the observed data under a specified statistical procedure. It conveys the estimate’s precision: a narrower interval indicates less sampling uncertainty for that procedure, while a wider interval indicates more. The interval does not account for every possible source of bias or uncertainty in design, measurement, follow-up, or analysis.

For the primary outcome, the interval lies below zero, while the unadjusted p-value is <0.001. Taken together, these summaries indicate that the observed unadjusted risk difference is estimated with uncertainty that does not include zero under the stated calculation. That interpretation is not equivalent to a conclusion about clinical importance, and the registry does not post a primary sponsor analysis to establish whether an adjusted or prespecified method gives the same result.

The secondary outcomes illustrate how precision varies by endpoint. The investigator-assessed progression-or-death risk-difference interval extends from -0.169 to 0.009, including zero. The objective-response interval extends from -0.102 to 0.062, also including zero. These intervals permit a range of possible differences around their point estimates. The reported progression-free survival hazard-ratio interval extends from 0.45 to 0.73. Its scale and interpretation differ from the risk-difference intervals, so the widths should not be directly compared as if they measured the same quantity.

Confidence intervals do not express the probability that the true effect lies within this particular calculated interval. They also do not show whether the result is robust to unreported analytic decisions. The registry record does not report the primary analysis method, handling of missing data, or multiplicity procedure, each of which can affect how results are interpreted. Precision should therefore be read alongside endpoint definition, assessment method, and analysis status—not as a standalone measure of certainty.

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 planning figures are illustrations derived from the observed effect, not the trial’s original design justification. For binary outcomes, the figures are participants per arm. For time-to-event outcomes, the figures are total events, not participants per arm. The listed assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation; each planning scenario targets the stated fraction of the observed effect. The original design assumptions are not reported in the registry record.

The 80% power column describes the planned number of participants per arm, or total events for a time-to-event outcome, that the illustration estimates would be needed to have an 80% chance of detecting the specified effect under those assumptions. The 90% power column uses a higher target chance of detecting that effect and therefore generally requires more information. These are planning concepts, not guarantees: actual power depends on the true effect, event rates, follow-up, and analysis assumptions.

The rows labeled 100%, 75%, and 50% of the observed effect represent planning for the full observed effect or for smaller fractions of it. Smaller effects need more participants or events to detect with the same target power, because a smaller difference is harder to distinguish from random variation. The figures are not observed power; they are hypothetical sample-size illustrations based on the observed effect and listed assumptions. They should not be read as evidence that the study was designed around those assumptions or that the planned target was achieved.

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 = 0.598; treatment proportion p₁ = 0.426; average p̄ = 0.512.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.7069 = 1.3855.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6964 = 0.5861.
4. n = (1.3855 + 0.5861)² / (-0.1716)² = 132.0, rounded up to 132 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 <- 0.598086124401914                          # control proportion
p1 <- 0.426470588235294                          # 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(1059630)
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 = 132 (80% power), 177 (90% power)
 75% of observed effect: n per arm = 235 (80% power), 314 (90% power)
 50% of observed effect: n per arm = 526 (80% power), 703 (90% power)
power.prop.test: 132 
Simulated power at n = 132 per arm: 0.810 (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. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates several endpoints, comparisons, time points, or analysis choices. In general, the more statistical tests considered, the greater the chance that at least one result will appear notable by chance if each is judged separately. A prespecified hierarchy or an adjustment procedure can control how confirmatory claims are made across a family of tests. The registry record lists 33 outcome measures and 12 analyses, but a multiplicity strategy, testing hierarchy, or adjusted secondary p-values are not reported in the registry record.

The primary endpoint is identified, but the registry does not post a reported primary analysis or its analysis plan. The unadjusted primary p-value is a recalculation from the posted summary counts. The secondary progression-free-survival result is explicitly reported with a Log Rank method, and the objective-response analysis is reported using Cochran-Mantel-Haenszel. The record does not state whether these secondary analyses were prespecified as confirmatory, how they relate to a testing hierarchy, or whether additional analyses were performed but not posted. Therefore, isolated unadjusted secondary p-values should not be treated as if they were multiplicity-controlled evidence.

Missing-data handling is also not reported in the registry record. The record does not describe how participants without a complete assessment were handled for the binary outcomes, whether follow-up ended at particular events or withdrawals, or how missing assessments affected the primary count. For progression-free survival, the record gives the endpoint and reported method but does not provide censoring details. These omissions matter because incomplete follow-up and assessment data can affect which events are observed and how estimates are calculated.

The primary outcome is specified for the intention-to-treat population, but details of the population definition and analysis implementation are not reported. The record does not give separate analysis populations for all secondary outcomes, describe protocol deviations, or explain any exclusions from particular analyses. Randomization supports a comparison based on assigned groups in principle, but the handling of participants after assignment is not fully documented in the posted record. These are limits on reconstructing the statistical plan, 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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pneumonia (5.9%), febrile neutropenia (5.4%), sepsis (3.4%), infusion related reaction (3.4%), neutropenia (2.9%).

The registry reports participants with serious adverse events and participants with other adverse events by treatment group. Serious adverse events were reported for 76 of 203 participants at risk in the bendamustine-alone group and 91 of 204 at risk in the obinutuzumab-plus-bendamustine group. Other adverse events were reported for 194 of 203 and 200 of 204, respectively. These are counts of participants, not counts of every event occurrence. No comparative safety effect estimate or confidence interval is reported in the record.

The registry lists serious adverse-event terms and percentages including pneumonia at 5.9%, febrile neutropenia at 5.4%, sepsis at 3.4%, infusion related reaction at 3.4%, neutropenia at 2.9%, and pyrexia at 2.9%. The record does not associate these listed term percentages with a particular treatment group in the posted summary, so they should not be assigned to an arm. Death counts are not reported in the registry record. Definitions, event attribution, and a detailed comparative safety analysis are also not reported in the registry record. These summaries describe recorded safety events; they do not establish a causal treatment effect or replace the disease-outcome interpretation.

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

The primary result describes the proportion of participants recorded as having progressive disease or death during the stated follow-up, using independent review committee assessment. The posted counts are lower in the obinutuzumab-plus-bendamustine group, and the simple unadjusted risk-difference interval lies below zero. This is a statistical description of the observed trial data. The registry record does not provide a reported primary analysis, so the recalculated estimate should not be mistaken for an adjusted or otherwise prespecified sponsor result.

The endpoint does not say when progression or death occurred. The secondary progression-free-survival outcome addresses timing, and its reported Log Rank analysis differs from the primary count comparison. Objective response is another distinct measure and does not by itself establish the duration of response or event-free time. Results across such endpoints can contribute to an overall description, but they should not be combined as though they measured the same quantity.

Statistical evidence is not the same as clinical importance. The record does not provide a threshold for a clinically important difference, nor does the primary estimate alone establish the balance of benefits and harms. The safety summaries identify participants with adverse events and list serious-event terms, but do not give a full comparative safety analysis. No medical recommendation follows from these summaries.

Generalizability is also limited to the population and circumstances represented by the trial. The study concerns rituximab-refractory indolent non-Hodgkin lymphoma, and further eligibility and setting details are not reported in the registry record. The open-label design and absence of detailed information on allocation procedures, missing data, multiplicity, and the primary analysis plan leave questions that cannot be resolved from the posted record alone.

The most defensible reading is therefore specific: the posted primary counts and unadjusted recalculation show a lower observed event proportion in the combination group, with the stated interval and p-value. Other outcomes supply different kinds of evidence, and the record’s omissions limit claims about prespecification, adjustment, missingness, safety comparison, and applicability beyond the recorded study population.

17. Why This Trial Matters Statistically

This trial is statistically informative because it puts a binary progression-or-death outcome alongside time-to-event and response measures. Those outcomes address related but distinct questions. Keeping the estimands separate helps prevent a count of participants with an event from being read as if it were a measure of time free from an event, or as if response itself established durable disease control.

The contrast between the primary unadjusted comparison and the reported secondary analyses also illustrates why the source and method of each estimate matter. A straightforward calculation from event counts offers an accessible risk-difference summary. A Log Rank analysis uses time-to-event information, while the reported Cochran-Mantel-Haenszel analysis uses a different method for response. Estimates and p-values from these methods answer different statistical questions and should not be compared as though they were outputs of one common model.

The record also illustrates the limits of registry summaries. It identifies the primary outcome and an intention-to-treat population, but does not report the primary analysis method, multiplicity handling, missing-data procedures, or original sample-size justification. These details affect how strongly results can support confirmatory interpretation. Clear separation between reported analyses and unadjusted recalculations helps preserve that distinction.

Finally, planning illustrations based on an observed effect are not a substitute for a trial’s original design rationale. They show how assumptions about effect size and target power shape hypothetical information needs, while the record does not state the original design assumptions. Read together, the endpoint definitions, analysis methods, uncertainty intervals, safety summaries, and reporting gaps demonstrate why statistical interpretation depends on more than a single p-value.

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 open-label design is a limitation for outcomes that may be influenced by knowledge of treatment assignment. Independent review committee assessment may reduce some assessment bias for the primary outcome, but the registry does not describe committee procedures or whether reviewers were masked. Investigator-assessed secondary outcomes are particularly distinct in assessment source, and the record does not provide enough detail to evaluate potential influence of treatment knowledge.

The primary analysis is not reported in the registry record. The available risk difference, confidence interval, and p-value are unadjusted recalculations from aggregate counts, so they do not establish what an adjusted or prespecified analysis would show. Randomization sequence generation, allocation concealment, stratification, and balance on baseline prognostic characteristics are not reported in the registry record.

Several analysis details are absent: how missing outcome assessments were handled, how censoring was defined for progression-free survival, which populations were used for secondary outcomes, and whether multiplicity was controlled. The record lists many outcomes, while confirmatory status and the testing hierarchy are not reported. These gaps limit reconstruction of the analysis plan and interpretation of secondary p-values.

The treatment description for the obinutuzumab-plus-bendamustine group is incomplete in the registry record. The primary outcome’s assessment schedule and criteria are described, but the record does not provide participant-level results, baseline characteristics, or detailed subgroup analyses. Safety information is limited to participant counts and listed serious-event terms; deaths and a full comparative safety analysis are not reported in the registry record. These reporting limitations do not show that the omitted procedures or analyses were absent from the trial, only that they cannot be evaluated from the registry record.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Study to Investigate the Efficacy and Safety of Bendamustine Compared With Bendamustine+Obinutuzumab (GA101) in Participants With Rituximab-Refractory, Indolent Non-Hodgkin's Lymphoma (GADOLIN) — NCT01059630.
• Registry record last updated 2020-01-13; status last verified 2019-12.
• Results first posted 2016-11-17.
• 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 NCT01059630, a completed, randomized, open-label, multicenter phase III parallel-group study in participants with rituximab-refractory indolent non-Hodgkin lymphoma. It compares bendamustine alone with obinutuzumab plus bendamustine followed by continued obinutuzumab maintenance monotherapy. The primary endpoint is the number of participants with progressive disease assessed by an independent review committee or death, in the intention-to-treat population.

The posted primary counts are 125 of 209 in the bendamustine-alone group and 87 of 204 in the combination group. The unadjusted recalculated risk difference is -0.172, with a 95% confidence interval from -0.267 to -0.077 and p < 0.001. The registry also reports secondary time-to-event and response analyses, alongside adverse-event counts. The primary analysis method, multiplicity strategy, missing-data procedures, and original sample-size assumptions are not reported in the registry record.