Allogeneic HCT With HLA-matched Donors : a Phase II Randomized Study Comparing 2 Nonmyeloablative Conditionings: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Allogeneic HCT With HLA-matched Donors : a Phase II Randomized Study Comparing 2 Nonmyeloablative Conditionings (NCT00603954), a randomized phase 2 study. The primary outcome is Percentage of Participants With Grade II-IV Acute GVHD Between the 2 Groups.

2. Clinical Question

The registry describes a comparison of two nonmyeloablative conditioning approaches used in centers in the United States. Both regimens precede infusion of cells from an HLA-matched donor, but they use different irradiation approaches and drugs. The main question is therefore a comparison of the recorded acute GVHD outcome between randomized groups, not a comparison of transplantation with no transplantation.

The primary endpoint is the percentage of participants with Grade II-IV acute GVHD within 180 days after HCT. The posted results report 12.2% for Flu-TBI and 8.9% for TLI-ATG. The unadjusted recalculation estimates a risk difference of -0.025, with a 95% confidence interval from -0.157 to 0.107 and an unadjusted p-value of 0.713. Those estimates describe the observed comparison and its uncertainty; they do not establish that either conditioning regimen causes a particular outcome.

Participants given a second allogeneic HCT were censored for GVHD analyses, according to the endpoint description. The registry record does not report further detail about how censoring was handled in the primary outcome calculation or how missing outcome data were treated.

3. Trial Design

NCT00603954 is recorded as a completed phase II randomized study with a parallel-group design and no masking. The registry lists an enrollment of 107. It identifies the sponsor as the University of Liege and the condition as hematological malignancies. The recorded start is 2008-01 and primary completion is 2016-07.

The comparison is between two conditioning regimens before allogeneic HCT with an HLA-matched related or unrelated donor. In the Flu-TBI group, the arm description specifies fludarabine 30 mg/m2 on days -4, -3 and -2, for a total dose of 90 mg/m2, followed by a single 2 Gy TBI dose on day 0 before cell infusion. It also describes a low dose-rate of approximately 7 cGy/min. The TLI-ATG arm is described as 8 Gy TLI with ATG. The registry says TLI was to be administered at 80 cGy daily, starting 11 days before transplantation, until 10 doses had been delivered. The arm description ends before fully specifying the irradiation fields, so further field details are not reported in the registry record.

The primary endpoint results use different denominators from the overall enrollment: the posted arm summaries list 48 participants for Flu-TBI and 40 for TLI-ATG. The reason for the difference between those denominators and the enrollment is not reported in the registry record. The primary endpoint description says that participants given a second allogeneic HCT were censored for GVHD analyses, but it does not provide a full participant-flow account or a detailed analysis-population definition.

Randomization supports a comparison between assigned groups, but the registry record does not report the allocation sequence method, allocation concealment, stratification factors, or the trial's original sample-size justification. It also does not report a detailed protocol-level statistical analysis plan. Those omissions limit the ability to assess how closely the posted analyses follow the prespecified design.

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: Patients given a second allogeneic HCT were censured for GVHD analyses.

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: 13/107 included patients (6 in the Flu-TBI and 7 in the TLI-ATG arm) were excluded from analysis because they did not meet the inclusion criteria at the time of the start of the conditioning (disease relapse (5), ineligible for further irradiation (3), donor refusal (2), HLA-mismatched donor (2), and poor PS (1)).

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With Grade II-IV Acute GVHD Between the 2 Groups, measured over 180 days after HCT.

The primary endpoint is the percentage of participants with Grade II-IV acute GVHD within 180 days after HCT. Acute GVHD is graded from I through IV in the registry description, with Grade IV identified as the worst grade. The endpoint combines Grades II, III and IV rather than counting Grade I in the primary event definition. Its purpose is to compare the proportion of participants who meet this specified acute GVHD threshold during the stated follow-up period.

The registry description gives clinical criteria involving skin rash, bilirubin and diarrhea for the grades. Some units and ranges in that description appear unusual or incomplete, and the registry record does not provide a separate clarification. The endpoint should therefore be interpreted according to the registry's stated Grade II-IV definition, rather than using the description to infer additional clinical definitions.

The posted results report 12.2% among 48 participants in Flu-TBI and 8.9% among 40 participants in TLI-ATG. The unadjusted recalculation is a risk difference of -0.025 for TLI-ATG relative to Flu-TBI. The corresponding 95% confidence interval ranges from -0.157 to 0.107. This interval expresses uncertainty in the between-group difference; it spans values in both directions, including no difference. The unadjusted p-value is 0.713.

Endpoint-specific interpretation depends on the observation period and analysis population as well as on the event definition. The record states that people who received a second allogeneic HCT were censored for GVHD analyses. Additional details on censoring, missing observations, adjudication, and the exact analysis procedure are not reported in the registry record.

7. Results Posted in the Registry

• TLI-ATG: 10.0% (4/40) versus 12.5% (6/48) on control; risk difference -2.5 percentage points (95% CI -15.7 to 10.7), risk ratio 0.80 (0.24–2.64); unadjusted two-proportion z-test p = 0.713.

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 Graft Rejection as Defined by Whole Blood and T Cell Chimerism

• 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:
  • TLI-ATG: 8.9% (4/45) versus 6.1% (3/49) on control; risk difference 2.8 percentage points (95% CI -7.9 to 13.5), risk ratio 1.45 (0.34–6.14); unadjusted two-proportion z-test p = 0.610.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between TLI-ATG and the control. The interval gives the range of effects that remain compatible with the data.

Percentage of Participants With Chronic GVHD in the 2 Groups

• 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:
  • TLI-ATG: 20.0% (8/40) versus 41.7% (20/48) on control; risk difference -21.7 percentage points (95% CI -40.3 to -3.0), risk ratio 0.48 (0.24–0.97); unadjusted two-proportion z-test p = 0.030.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically lower for TLI-ATG than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Percentage of Non Relapse Mortality in the 2 Groups

• 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:
  • TLI-ATG: 13.3% (6/45) versus 24.5% (12/49) on control; risk difference -11.2 percentage points (95% CI -26.8 to 4.5), risk ratio 0.54 (0.22–1.33); unadjusted two-proportion z-test p = 0.170.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between TLI-ATG and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover several aspects of transplantation and follow-up: graft rejection, chronic GVHD, infections, relapse, immunologic reconstitution, non-relapse mortality, and progression-free survival at stated follow-up times. The outcomes are not all the same type: some are percentages or participant counts, while others involve timing, incidence, or concentrations. Their follow-up periods also vary. The available secondary results therefore provide a broader but heterogeneous view of outcomes rather than a single confirmatory test of the primary question.

Across the posted secondary results, the direction and strength of comparisons are not uniform. For example, the listed results include a contrast for chronic GVHD with a sponsor analysis reporting p-values, while other posted comparisons have intervals spanning no difference and sponsor p-values that do not indicate a clear separation. Those results do not convert the primary endpoint's uncertain risk-difference estimate into evidence for a consistent overall advantage. The primary endpoint remains a specific Grade II-IV acute GVHD outcome at 180 days, whereas secondary outcomes address different events and time frames.

Unadjusted secondary results carry less weight as a collection because multiple outcomes and analyses create more opportunities for a nominally notable result to occur by chance. A p-value for one secondary outcome should therefore be viewed in the context of the full set of questions, not as independent confirmation. The registry record does not report a multiplicity adjustment or a hierarchy identifying which secondary analyses were confirmatory. It also does not provide full results for every listed outcome in the available record. The secondary findings are best treated as exploratory signals requiring careful interpretation, not as a replacement for a prespecified primary analysis.

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

• Number of Participants With Graft Rejection as Defined by Whole Blood and T Cell Chimerism
• Percentage of Participants With Chronic GVHD in the 2 Groups
• Incidences of Bacterial, Fungal and Viral Infections in the 2 Groups.
• Percentage of Relapse Rate in the 2 Groups
• Quality and Timing of Immunologic Reconstitution: Concentration of ATG Levels
• Percentage of Non Relapse Mortality in the 2 Groups
• Percentage of 4-year Progression Free Survival in the 2 Groups
• Percentage of 5-year Progression Free Survival in the 2 Groups

9. Statistical Interpretation

10. Statistical Methods Explained

The posted primary result is a binary outcome: each participant is classified according to whether Grade II-IV acute GVHD was recorded within 180 days after HCT. A percentage summarizes the proportion meeting that definition in each arm. The unadjusted comparison expresses the difference as a risk difference, with TLI-ATG compared with Flu-TBI. A negative value means the observed percentage is lower in TLI-ATG under this ordering; it does not by itself establish a treatment effect.

What does a confidence interval add?

A confidence interval gives a range of values compatible with the observed data under the method used. It conveys precision as well as direction: a narrow interval indicates that the estimate is relatively precise, while a wide interval indicates more uncertainty. The primary risk-difference interval, from -0.157 to 0.107, includes zero and values on both sides of zero. In practical statistical terms, the estimate does not tightly determine either the direction or the size of the difference. A confidence interval is not a guarantee that the true value lies within the particular calculated range.

How should the p-value be understood?

The unadjusted p-value of 0.713 describes how compatible the observed data are with a null model of no between-group difference, under the assumptions of that comparison. It is not the probability that the null hypothesis is true, and it does not measure the magnitude or clinical importance of a difference. A large p-value is not proof that two regimens have identical outcomes. It may occur when an estimate is near zero, when data are limited, or when variability is substantial. The interval is essential for seeing which sizes and directions of difference remain compatible with the data.

Why do sponsor analyses differ in form?

The registry lists a multivariate Cox model and Wilcoxon (Mann-Whitney) analyses among the sponsor's analyses. A Cox model is generally used for time-to-event outcomes and can estimate a relative event rate over follow-up while accounting for specified covariates. A multivariate model includes more than one predictor, although the covariates used in this trial are not reported in the registry record. A Wilcoxon rank-sum test compares ranks across groups and can be used for distributions of measurements or event-related quantities. The registry does not give enough information to determine exactly which data and estimands these tests address in this trial. Their p-values alone do not reveal a treatment effect estimate or its precision.

The simple recalculation is explicitly unadjusted and is based on the posted arm summaries. It is not a substitute for a model that accounts for time-to-event information, censoring, covariates, or any prespecified analysis features. Since the registry record is marked as discrepant, the sponsor's own analysis takes precedence. The discrepancy's nature and the rationale for choosing the sponsor analysis are not reported in the registry record.

What do over-dispersion and outcome type mean generally?

General statistical explanation: over-dispersion occurs when observed variability is greater than a model assumes. If it is ignored, standard errors can be too small and confidence intervals can be too narrow. It is especially relevant to count outcomes, where a basic model may assume a variance structure that does not fit the observed data. The primary outcome is recorded as a percentage of participants with an event, not as a count rate over person-time, so over-dispersion is not the central issue for interpreting its simple risk-difference calculation.

Different outcomes need different summaries. A fixed-time binary endpoint can be described with proportions and a risk difference. A time-to-event outcome requires attention to follow-up and censoring. Rank-based tests do not automatically provide the same effect measure as a risk difference. For a broad set of secondary outcomes, the method should match the endpoint's definition, timing, and data structure. The registry lists some analyses but does not report a complete statistical plan explaining the choice of method for every outcome.

What is not reported about the analysis?

The original sample-size justification, prespecified primary analysis details, missing-data procedures, covariate list for the multivariate Cox model, and a complete description of the analysis populations are not reported in the registry record. The registry record also does not explain the relationship between the primary endpoint's fixed 180-day percentage and the sponsor analyses listed for that outcome. These details matter because analysis choices can affect the estimand, uncertainty, and interpretation. Accordingly, the posted estimates should be read as registry results rather than as a full account of the trial's statistical protocol.

11. Confidence Intervals and Precision

The primary endpoint summary has denominators of 48 for Flu-TBI and 40 for TLI-ATG. The reported percentages are 12.2% and 8.9%, respectively. The unadjusted recalculated risk difference is -0.025, and its 95% confidence interval extends from -0.157 to 0.107. The interval crosses zero and covers a range of possible group differences in both directions. This indicates that the observed contrast is not estimated precisely enough to settle the direction of a difference.

Precision and statistical significance are related but not identical ideas. A confidence interval provides information about the range and uncertainty of an estimated effect, whereas a p-value summarizes compatibility with a specified null model. The unadjusted p-value is 0.713. It should not be read as a measure of the probability that the treatments are the same. Nor should the confidence interval be interpreted as proof that all values within it are equally plausible.

The sponsor analyses listed for the primary outcome report p-values but no effect estimates or confidence intervals. Without those estimates, their magnitude and precision cannot be assessed from the registry entries. The methods listed include a multivariate Cox model and Wilcoxon (Mann-Whitney) analyses; the registry does not specify enough detail to reconcile their targets with the fixed-time binary percentage.

The number enrolled in the trial is 107, while the primary outcome summaries use smaller arm denominators. The registry record does not explain the difference. In addition, it states that participants given a second allogeneic HCT were censored for GVHD analyses. The record does not give further details on missing endpoint observations or censoring. These reporting gaps make it harder to determine how the denominators and analysis methods affect the precision of the result.

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 figures are planning illustrations based on the observed effect, not the trial's original sample-size justification. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the arms, and a normal approximation. The assumed effect is a stated fraction of the observed effect. For binary outcomes such as the primary endpoint, the figures are participants per arm. For time-to-event outcomes, the figures are total events, as specified by the planning description.

The 80% power column gives the illustrative number of participants per arm, or total events for a time-to-event outcome, associated with an 80% chance of detecting the assumed effect if that effect is real and the planning assumptions hold. The 90% power column uses a higher target chance of detection and therefore generally requires a larger planned sample or event total. For the primary comparison, the planning illustration is 2,507 participants per arm for 80% power and 3,356 participants per arm for 90% power when using 100% of the observed effect. Planning for 75% of the observed effect gives 4,565 participants per arm for 80% power and 6,111 participants per arm for 90% power. Planning for 50% of the observed effect gives 10,513 participants per arm for 80% power and 14,073 participants per arm for 90% power.

Smaller effects are harder to distinguish from random variation, so planning for a smaller fraction of the observed effect requires more participants. These counts are illustrative and depend on the stated assumptions; they are not the trial's observed power. The original design assumptions and sample-size justification are not reported in the registry record.

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 = 0.125; treatment proportion p₁ = 0.100; average p̄ = 0.113.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4469 = 0.8758.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4465 = 0.3758.
4. n = (0.8758 + 0.3758)² / (-0.0250)² = 2,506.5, rounded up to 2,507 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 <- 0.125                          # control proportion
p1 <- 0.1                          # 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(603954)
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 = 2,507 (80% power), 3,356 (90% power)
 75% of observed effect: n per arm = 4,565 (80% power), 6,111 (90% power)
 50% of observed effect: n per arm = 10,513 (80% power), 14,073 (90% power)
power.prop.test: 2506.5 
Simulated power at n = 2,507 per arm: 0.802 (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 unadjusted recalculation uses the posted arm percentages and denominators to estimate a risk difference for the primary endpoint. It gives -0.025, with a 95% confidence interval from -0.157 to 0.107 and an unadjusted p-value of 0.713. The registry also lists sponsor analyses for the primary comparison: a multivariate Cox model with p = 0.508 and Wilcoxon (Mann-Whitney) results with p = 0.02 and p = 0.002.

The registry record is flagged as having a discrepancy. For that reason, the sponsor's own analysis takes precedence over the simple unadjusted recalculation when interpreting the sponsor's intended analysis. The record does not describe the discrepancy, provide effect estimates for the listed sponsor analyses, or explain exactly how the reported tests relate to the fixed-time binary endpoint. It is therefore not possible to use those p-values alone to recover the magnitude, direction, or precision of the sponsor's estimated contrast.

The two analyses should not be blended into a single conclusion. The unadjusted recalculation is a transparent summary from the posted proportions; it does not replace the sponsor's analysis or account for any features the sponsor's method may address. Conversely, the sponsor p-values without estimates and methodological detail do not make the uncertainty around the primary comparison clear. The registry record does not report a fuller explanation that resolves these differences.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 11 outcome measures, multiple secondary outcomes, and 21 analyses. Several outcomes are examined across different follow-up periods or with more than one statistical method. When many outcomes and tests are considered, the chance of at least one apparently notable result can increase even if no underlying difference exists. This is the general multiplicity problem. It does not mean every observed result is false, but it reduces the evidential weight of isolated unadjusted p-values when the analysis plan and testing hierarchy are unknown.

A prespecified primary outcome can help focus interpretation, while secondary outcomes can add context. The record identifies Grade II-IV acute GVHD at 180 days after HCT as the primary endpoint. However, the multiplicity strategy, any adjustment to significance thresholds, and any hierarchy among secondary endpoints are not reported in the registry record. It is also not reported whether the multiple sponsor analyses listed for the primary outcome were prespecified or how they were intended to relate to each other.

Missing-data and analysis-population details are also limited. The registry gives overall enrollment and arm-specific endpoint denominators, but does not explain the difference. It says participants given a second allogeneic HCT were censored for GVHD analyses. Further details about censoring, missing endpoint values, exclusions, and whether analyses followed randomized assignment are not reported in the registry record. These details affect the interpretation of denominators and the assumptions behind statistical comparisons.

The record does not report the original sample-size justification, interim analyses, or a complete statistical analysis plan. It is therefore not possible to establish from the registry entries whether the number of analyses or outcome definitions were prespecified, or whether additional analyses were conducted. This uncertainty is particularly relevant when interpreting secondary p-values and the differing methods listed for the primary outcome.

16. Safety Results

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

Deaths

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: relapse/progression (48.9%), chronic graft versus host disease (40.8%), lung (14.3%), viral infection (12.2%), acute graft versus host disease (11.1%).

The adverse-event summary lists 49 participants at risk for serious adverse events in Flu-TBI, with 41 affected, and 45 at risk in TLI-ATG, also with 41 affected. The summary separately lists 49 and 45 at risk for other adverse events, with 15 and 21 affected, respectively. Deaths are listed as 22 for Flu-TBI and 19 for TLI-ATG. These are counts in the registry summary; they should not be treated as comparative causal estimates.

The leading serious adverse-event terms are listed as relapse/progression, 48.9%; chronic Graft versus Host Disease, 40.8%; lung, 14.3%; viral infection, 12.2%; acute Graft versus Host Disease, 11.1%; and septic shock, 11.1%. The registry record does not report the denominator for these term percentages or clarify whether participants could contribute to more than one term. Thus the percentages should not be read as mutually exclusive categories or as arm-specific rates.

Serious adverse-event summaries describe events recorded during the trial; they do not on their own establish that a conditioning regimen caused an event. The registry record does not provide enough detail to assess event timing, severity grading beyond the serious-event classification, exposure-adjusted rates, or a formal between-group safety analysis.

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

The primary endpoint results describe the percentage of participants with Grade II-IV acute GVHD within 180 days after HCT. The registry reports 12.2% in Flu-TBI and 8.9% in TLI-ATG. The unadjusted risk-difference estimate is -0.025, with a 95% confidence interval from -0.157 to 0.107 and an unadjusted p-value of 0.713. These statistics show an observed difference in the direction of a lower percentage in TLI-ATG, accompanied by substantial uncertainty about the direction and size of the underlying difference.

The result does not demonstrate that the regimens are equivalent. Equivalence requires a prespecified margin and an analysis designed to assess whether the effect lies within that margin; neither is reported in the registry record. It also does not establish non-inferiority, which likewise requires a prespecified framework. A non-significant p-value is not evidence that two treatments have the same effect.

The result also does not show that either regimen works or fails. It is a comparison of two conditioning approaches among participants receiving HCT, and the data are limited to the outcomes and follow-up reported in the registry. The endpoint concerns acute GVHD, not every possible benefit or harm of transplantation. The secondary outcomes address other events but are varied in definition and timing, and their multiplicity strategy is not reported.

The discrepancy flag matters: the sponsor's own analysis takes precedence over the simple unadjusted recalculation, but the registry does not provide enough details to reconstruct the sponsor's effect estimate or to understand the discrepancy. The available p-values from sponsor analyses do not give the missing effect sizes and confidence intervals. As a result, interpretation should remain centered on what is documented: the posted primary percentages, the unadjusted comparison and its uncertainty, and the limits of the reported analyses.

No clinical recommendation follows from these registry summaries alone. The record does not report a complete protocol, full analysis plan, comprehensive participant flow, or full endpoint results. Such omissions prevent a definitive account of how design and analysis choices may have influenced the findings.

18. Why This Trial Matters Statistically

This trial is statistically informative because it compares two active conditioning strategies in a randomized parallel-group design, rather than comparing an intervention with no conditioning. Random allocation can support a fair between-group comparison, but the strength of that interpretation depends on clear reporting of allocation, follow-up, analysis populations, and the prespecified statistical plan. Some of those details are not reported in the registry record.

The primary endpoint illustrates why an observed percentage difference should be read alongside its uncertainty. The reported arm percentages are 12.2% and 8.9%, and the recalculated risk difference has a confidence interval spanning values on both sides of zero. A point estimate alone could suggest a directional pattern, but the interval shows that the data do not precisely resolve that pattern. The result is a useful example of the difference between an estimated effect and the certainty attached to it.

The registry also illustrates the importance of distinguishing an unadjusted summary from sponsor analyses using different methods. A Cox model, rank-based tests, and a fixed-time binary proportion do not automatically answer the same statistical question. Without effect estimates and a clear account of the estimands, p-values from different methods cannot be compared as if they were interchangeable.

Finally, the trial has a broad set of secondary outcomes and analyses. Those can enrich understanding, but they also raise questions about multiplicity, prespecification, and selective emphasis. The registry does not report a complete analysis plan or multiplicity procedure. This makes transparent reporting of estimates, confidence intervals, outcome definitions, and analysis populations especially important for interpreting randomized clinical evidence.

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 has several limitations specific to this trial. The enrollment is listed as 107, but the primary endpoint summaries use 48 participants in Flu-TBI and 40 in TLI-ATG. The reason for these smaller denominators is not reported in the registry record. Although the endpoint description says participants given a second allogeneic HCT were censored for GVHD analyses, it does not provide a full account of censoring or missing outcomes.

The sponsor's primary analyses are listed as a multivariate Cox model and two Wilcoxon (Mann-Whitney) analyses, with p-values but without effect estimates or confidence intervals. The registry record does not explain the target of each analysis or how these analyses relate to the primary endpoint's fixed 180-day percentage. Because the record is flagged as discrepant, the sponsor's analysis takes precedence over the simple unadjusted recalculation, but the discrepancy and its resolution are not described.

The record does not report the original sample-size justification, allocation concealment, randomization sequence method, covariates used in the multivariate model, interim analyses, or a complete statistical analysis plan. The multiplicity approach for the many outcomes and analyses is also not reported. These omissions limit assessment of prespecification, analysis choices, and the weight to assign secondary findings.

Some arm-description text is incomplete, including the description of the TLI irradiation fields. The record's primary endpoint description includes grading criteria with unusual or incomplete units, and no clarification is reported. In addition, serious adverse-event term percentages are listed without a stated denominator or an explanation of whether categories overlap. These reporting gaps constrain detailed interpretation of regimen delivery, endpoint measurement, and safety patterns.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Allogeneic HCT With HLA-matched Donors : a Phase II Randomized Study Comparing 2 Nonmyeloablative Conditionings — NCT00603954.
• Registry record last updated 2022-01-26; status last verified 2021-11.
• Results first posted 2022-01-26.
• 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

NCT00603954 is a completed phase II randomized parallel-group study comparing Flu-TBI and TLI-ATG conditioning before allogeneic HCT with HLA-matched related or unrelated donors. The registry lists enrollment of 107 and identifies Grade II-IV acute GVHD within 180 days after HCT as the primary endpoint.

The posted primary percentages are 12.2% for Flu-TBI and 8.9% for TLI-ATG. The unadjusted recalculated risk difference is -0.025, with a 95% confidence interval from -0.157 to 0.107 and an unadjusted p-value of 0.713. Sponsor analyses are also listed, but their effect estimates and full methodological details are not reported. The record is flagged as discrepant, so the sponsor's own analysis takes precedence over the simple unadjusted recalculation.

Secondary outcomes address graft rejection, chronic GVHD, infections, relapse, immunologic reconstitution, non-relapse mortality, and progression-free survival. The registry record does not report a complete statistical plan, multiplicity procedure, original sample-size justification, or full explanation of the analysis populations.