Tacrolimus/Sirolimus/Methotrexate vs Tacrolimus/Methotrexate or Cyclosporine/Mycophenolate Mofetil for GVHD Prophylaxis After Reduced Intensity Allogeneic Stem Cell Transplantation for Patients With Lymphoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Tacrolimus/Sirolimus/Methotrexate vs Tacrolimus/Methotrexate or Cyclosporine/Mycophenolate Mofetil for GVHD Prophylaxis After Reduced Intensity Allogeneic Stem Cell Transplantation for Patients With Lymphoma (NCT00928018), a randomized phase 3 study. The primary outcome is To Compare 2-year Overall Survival of Patients With Lymphoma Undergoing RIC SCT Between Those Receiving Tacrolimus/Sirolimus/Methotrexate and Those Receiving Tacrolimus/Methotrexate or Cyclosporine/Mycophenolate Mofetil.

2. Clinical Question

3. Trial Design

This was a phase III, randomized, parallel-group trial with no masking. Its official title identifies it as a multicenter comparison. The trial studied GVHD prophylaxis after reduced-intensity allogeneic stem cell transplantation for patients with lymphoma. The registry lists Dana-Farber Cancer Institute as sponsor, records the trial as completed, and gives a start of 2009-06 and primary completion of 2014-11. Registry enrollment is 139.

Participants were assigned to a sirolimus-containing regimen or a sirolimus-free regimen. The first included tacrolimus, sirolimus and low-dose methotrexate. The second allowed either tacrolimus with methotrexate or cyclosporine with mycophenolate mofetil. The available arm descriptions include medication details, but the full operational rules for selecting the sirolimus-free regimen are not reported in the registry record. The posted outcome summaries show 66 participants for the sirolimus-containing arm and 73 for the sirolimus-free arm.

Random allocation is intended to make treatment groups comparable, on average, with respect to both measured and unmeasured prognostic factors. Lack of masking means participants and study personnel were not blinded to assignment, as recorded by the registry. That feature is relevant to outcomes that may involve judgment or reporting, while the registry record does not give enough detail to assess how particular outcomes were ascertained.

The primary endpoint was 2-year overall survival. Secondary outcomes included progression-free survival, disease progression, non-relapse mortality and GVHD outcomes at specified follow-up times, as well as histology-specific comparisons. The posted record includes results for only selected secondary outcomes. Details such as the allocation concealment procedure, stratification factors, treatment adherence, crossover, and the full analysis population 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:

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 To Compare 2-year Overall Survival of Patients With Lymphoma Undergoing RIC SCT Between Those Receiving Tacrolimus/Sirolimus/Methotrexate and Those Receiving Tacrolimus/Methotrexate or Cyclosporine/Mycophenolate Mofetil, measured over 2 years.

The primary endpoint is 2-year overall survival among patients with lymphoma undergoing reduced-intensity allogeneic stem cell transplantation. Overall survival concerns whether a participant is alive at the stated follow-up point, regardless of the cause of death. The registry records the outcome as a percentage of participants and gives a 95% confidence interval for each treatment arm. It does not state the endpoint definition in greater detail, including the method used to estimate survival when follow-up times differ or when participants are lost to follow-up.

Overall survival is used because it provides a broad measure of the outcome of transplantation and its related care. It reflects fatal outcomes from lymphoma, transplantation complications, GVHD and other causes together. This breadth also means that it does not isolate a possible anti-lymphoma effect of sirolimus: a difference in overall survival, if present, could reflect several pathways, while a similar estimate cannot establish that relapse or non-relapse mortality was alike.

The posted arm estimates are 70.0% for the sirolimus-containing regimen, with a 95% confidence interval from 57.0% to 79.0%, and 68.0% for the sirolimus-free regimen, with a 95% confidence interval from 57.0% to 78.0%. These are estimates for the endpoint and follow-up specified in the registry. They should be read with their uncertainty intervals and in light of the comparison estimate, rather than as exact underlying survival probabilities.

7. Results Posted in the Registry

• Sirolimus-Free Regimen: 68.5% (50/73) versus 69.7% (46/66) on control; risk difference -1.2 percentage points (95% CI -16.6 to 14.2), risk ratio 0.98 (0.79–1.23); unadjusted two-proportion z-test p = 0.878.

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.

To Compare 2-year Progression-free Survival Between the Two Treatment Arms

• 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:
  • Sirolimus-Free Regimen: 57.5% (42/73) versus 60.6% (40/66) on control; risk difference -3.1 percentage points (95% CI -19.4 to 13.3), risk ratio 0.95 (0.72–1.25); unadjusted two-proportion z-test p = 0.713.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Sirolimus-Free Regimen and the control. The interval gives the range of effects that remain compatible with the data.

To Compare the 2-year Cumulative Incidence of Chronic GVHD Between the Two Treatment Arms.

• 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:
  • Sirolimus-Free Regimen: 63.0% (46/73) versus 59.1% (39/66) on control; risk difference 3.9 percentage points (95% CI -12.3 to 20.2), risk ratio 1.07 (0.82–1.39); unadjusted two-proportion z-test p = 0.636.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Sirolimus-Free Regimen and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary results follow a pattern of fairly similar arm-level estimates, with broad uncertainty around the comparisons. The 2-year progression-free survival estimates are 61.0% for the sirolimus-containing regimen and 58.0% for the sirolimus-free regimen. The 2-year chronic GVHD estimates are 59.0% and 63.0%, respectively. The direction of the arm-level difference is not the same across these outcomes. This overall pattern does not supply a consistent secondary signal that can be treated as confirmation of the primary result.

The secondary outcomes address aspects of the transplantation course that overall survival does not isolate. Progression-free survival incorporates disease progression as well as death according to its endpoint definition; chronic GVHD is a complication outcome. The registry also lists disease progression, non-relapse mortality, acute GVHD and histology-specific comparisons among the secondary objectives. Results for only selected secondary outcomes are posted; the unposted results are not reported in the registry record. The primary and secondary outcomes therefore should not be treated as if they were interchangeable measures of one effect.

Secondary analyses are generally less conclusive than a prespecified primary comparison, especially when several outcomes are examined. Each additional test creates more opportunity for a seemingly notable result to arise by chance, and a multiplicity adjustment or testing hierarchy can affect how evidence should be judged. No such procedure is reported in the registry record. The posted secondary comparisons are unadjusted recalculations, and the record lists no reported analyses for them. They are useful as descriptive context but do not carry the same confirmatory weight as a fully specified and reported primary analysis.

The broad intervals around the posted secondary contrasts also counsel against over-reading small differences in point estimates. A result that appears numerically higher in one arm does not establish a reliable difference, and a result that is not statistically persuasive does not establish equivalence. Taken together, the available secondary findings provide context about selected outcomes, but they neither resolve the uncertainty in overall survival nor establish a separate benefit or harm for any secondary endpoint.

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

• To Compare 2-year Progression-free Survival Between the Two Treatment Arms
• To Compare the 2-year Cumulative Incidences of Disease Progression and of Non-relapse Mortality Between the Two Treatment Arms
• To Compare the 180-day Cumulative Incidence of Grades II-IV and Grades III-IV Acute GVHD Between the Two Treatment Arms
• To Compare the 2-year Cumulative Incidence of Chronic GVHD Between the Two Treatment Arms.
• To Compare the 2-year of Overall Survival, Progression-free Survival, Cumulative Incidences of Progression and Non-relapse Mortality Between the Treatment Arms for Each Histology Studied.

9. Statistical Interpretation

10. Statistical Methods Explained

The posted outcomes are percentages of participants with an event or a stated endpoint value, and the registry gives 95% confidence intervals for some of them. The comparison statistics include a risk difference and an unadjusted p-value. The interpretation below explains these statistical ideas generally; the registry record does not provide a complete statistical analysis plan or the exact methods used for every posted estimate.

What does a risk difference describe?

A risk difference compares the event probabilities in two groups by subtracting one from the other. Its sign depends on which group is treated as the reference and the order of subtraction. For the primary comparison, the registry labels the comparison as the sirolimus-free regimen against the sirolimus-containing regimen and reports a risk difference of -0.012. This is an unadjusted comparison based on posted summary statistics. It does not account for covariates or other differences among participants beyond the group comparison itself.

A risk difference is an absolute measure: it expresses the distance between group-level probabilities on the probability scale. It is distinct from a risk ratio, which expresses a relative comparison, and an odds ratio, which compares odds. The registry record also lists relative measures in the comparison data, but a reader should not assume that different measures have identical interpretations. The risk difference is especially useful for seeing the estimated absolute separation between group outcomes, while its uncertainty interval is essential for judging how stable that estimate may be.

How should a confidence interval be read?

A confidence interval describes the uncertainty associated with an estimate under a statistical procedure. In repeated use of a method under its assumptions, intervals constructed in this way would cover the target parameter at the stated confidence level at the corresponding long-run rate. It is not a range that assigns a probability to the fixed parameter after the data have been observed. For a treatment comparison, an interval spanning zero means that no difference is among the values compatible with the data and method at that confidence level. It may also include differences in either direction.

The arm-level confidence intervals and the confidence interval for a between-arm contrast answer related but different questions. Overlap between two separate arm intervals is not itself a formal test of the treatment contrast. The direct comparison interval is the relevant interval for assessing uncertainty in the risk difference. For the primary result, that interval ranges from -0.166 to 0.142. It is broad relative to the point estimate, so the estimate alone should not be treated as a precise statement about the size or direction of the underlying contrast.

What does an unadjusted p-value say?

A p-value is calculated under a null hypothesis and specified statistical assumptions. It measures how unusual the observed data, or data more extreme, would be if that null model and its assumptions were correct. It is not the probability that the null hypothesis is true, and it is not a measure of the size or clinical importance of an effect. The primary unadjusted p-value is 0.878. That value does not establish equality; the point estimate and its confidence interval are needed to understand the estimated difference and its uncertainty.

The word unadjusted matters. An unadjusted comparison generally does not adjust for baseline covariates or other design features in a model. Randomization supports a fair comparison in expectation, but adjustment can still be part of a planned analysis and may affect precision or account for stratification. Whether such methods were planned or used is not reported in the registry record. The posted comparison should therefore be distinguished from any sponsor analysis that may be described elsewhere.

Why can outcome type affect the analysis?

Overall survival and progression-free survival are commonly understood as time-to-event outcomes: the timing of an event and the duration of observed follow-up can matter, and some participants may not experience the event before follow-up ends. Cumulative incidence outcomes can also involve competing events, where the occurrence of one event changes the possibility of observing another. These are general statistical considerations, not claims about the method used in this trial. Although the registry identifies the follow-up time and provides percentages, it does not report the estimator, censoring rules, competing-risk method, or model used for the posted results.

For binary outcomes, the observed proportion is a natural summary, but it does not by itself show how follow-up was handled or whether the estimate is a simple proportion or a time-to-event estimate expressed as a percentage. In a transplant study with differing event types and follow-up, those distinctions can matter. The available record does not resolve them. It also does not report an analysis population, missing-data strategy, or sensitivity analyses.

Why does multiplicity matter?

When a trial evaluates several endpoints, testing each separately can increase the chance of at least one apparently unusual result under a situation where none of the tested effects is real. This is the general multiple-comparisons problem. How much it matters depends on the number and relationship of tests, which outcomes were designated primary, and whether any adjustment or hierarchy was prespecified. The registry lists several secondary objectives but does not report a multiplicity procedure. Consequently, unadjusted secondary p-values, if reported, would need cautious interpretation and should not be treated as confirmatory evidence on their own.

The record identifies zero reported analyses for the primary outcome and no reported analyses for the posted secondary outcomes. It gives no basis for reconstructing a full model or a sponsor's inferential procedure. The posted arithmetic comparison remains useful as a transparent descriptive contrast, but it should not be mistaken for a complete account of the trial's original statistical analysis.

11. Confidence Intervals and Precision

Precision concerns how tightly an estimate is determined by the data and analysis. The primary arm estimates are accompanied by 95% confidence intervals: 57.0% to 79.0% for the sirolimus-containing regimen and 57.0% to 78.0% for the sirolimus-free regimen. Both intervals span a substantial range of possible endpoint values. The point estimates, 70.0% and 68.0%, should therefore not be read as exact values for the populations from which the participants were drawn.

For the between-arm primary comparison, the risk difference is -0.012 and its 95% confidence interval runs from -0.166 to 0.142. The interval includes zero and allows for effects in either direction. This is why the comparison is imprecise: the data do not locate the difference narrowly. The unadjusted p-value of 0.878 is consistent with the absence of a clear separation under the comparison procedure, but it does not show that the regimens have identical outcomes.

Arm-level intervals and the interval for the difference should not be conflated. Each arm interval describes uncertainty around that arm's estimate, whereas the contrast interval directly describes uncertainty around the comparison. The latter is the more relevant measure for a claim about how the regimens differ. Neither interval accounts for uncertainties that may arise from unreported analysis choices, missing follow-up, selection of outcomes, or limitations in the registry description.

In general, interval width depends on the amount of information and the variability of the outcome, as well as the analysis method. The registry reports arm sizes of 66 and 73 for this outcome summary, but the record does not provide the survival estimator, event counts, censoring details or completeness of follow-up needed for a fuller assessment of precision. The confidence intervals therefore provide an important visible indication of uncertainty, while the unavailable analysis details remain relevant to interpretation.

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 based on the observed effect, not the trial's original design calculation. For these binary outcomes, the listed figures are participants per arm, not total events. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is the stated fraction of the observed effect. The registry record does not report the trial's own original sample-size assumptions or justification.

The 80% power column gives the planned participants per arm for a hypothetical study designed to have an 80% chance of detecting the selected effect if that effect were the true one, under the stated assumptions and statistical model. The 90% power column describes the corresponding planning target with a higher chance of detection under the same conditions. These are prospective planning concepts: they do not say what chance this completed trial had of finding an effect after its results were known.

The rows represent planning for 100%, 75% or 50% of the observed effect. A smaller target effect is harder to distinguish from random variation, so it requires more participants for the same power target under the stated assumptions. These figures should not be read as recommended enrollment or as evidence that the observed trial was adequately or inadequately powered; the original design rationale and assumptions are not reported in the registry record.

This is not observed power. The planning illustrations use the observed effect as an input and do not provide an independent measure of evidential strength. The estimates and their confidence intervals are more informative for interpreting the actual posted results. The figures also assume equal allocation, while the posted arm sizes are not equal; that difference reflects the distinction between a hypothetical planning scenario and the observed trial 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 = 0.697; treatment proportion p₁ = 0.685; average p̄ = 0.691.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6535 = 1.2809.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6535 = 0.5500.
4. n = (1.2809 + 0.5500)² / (-0.0120)² = 23,129.6, rounded up to 23,130 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 <- 0.696969696969697                          # control proportion
p1 <- 0.684931506849315                          # 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(928018)
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 = 23,130 (80% power), 30,964 (90% power)
 75% of observed effect: n per arm = 41,010 (80% power), 54,900 (90% power)
 50% of observed effect: n per arm = 92,020 (80% power), 123,189 (90% power)
power.prop.test: 23129.6 
Simulated power at n = 23,130 per arm: 0.801 (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

The registry lists a primary overall-survival objective and several secondary objectives, including progression-free survival, disease progression, non-relapse mortality, acute GVHD, chronic GVHD and histology-specific comparisons. Evaluating multiple outcomes raises a multiplicity question: without a planned strategy, the chance of at least one nominally unusual finding can increase as more tests are considered. The record does not report whether outcomes were tested in a hierarchy, whether p-values were adjusted, or whether any secondary findings were intended to be confirmatory.

The primary comparison in the posted results is an unadjusted recalculation. The registry reports zero primary analyses and no reported analyses for the posted secondary outcomes. There is no sponsor-reported model, test specification, or description of adjustment for baseline characteristics in the registry record. As a result, the posted comparisons can be interpreted as unadjusted summaries, but cannot be assumed to reproduce the trial's planned inferential analysis.

Missing-data handling is also not reported in the registry record. For survival and cumulative-incidence outcomes, the treatment of incomplete follow-up, censoring and competing events can affect estimates and uncertainty. The record does not describe participant withdrawal, loss to follow-up, outcome ascertainment completeness, or sensitivity analyses for missing observations. It is not possible to determine from the posted information whether missingness could have influenced the comparisons.

The analysis populations are not defined. In particular, the record does not state whether the primary outcome was analyzed by intention to treat, per protocol, or another approach, nor does it explain how deviations from assigned treatment were handled. Randomization identifies the intended basis for a between-arm comparison, but interpretation depends on how randomized participants were included in analysis. These omissions do not erase the posted estimates; they limit how fully their analytic validity and robustness can be evaluated.

The record also does not report a prespecified clinically meaningful difference or an equivalence or non-inferiority margin. Accordingly, the non-significant primary comparison cannot be interpreted as a demonstration that the regimens are equivalent. A proper equivalence claim requires a specified margin and analysis designed to assess whether uncertainty excludes differences outside that margin. No such framework is reported.

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: infection (12.1%), gvhd (5.5%), acute kidney injury/renal failure (4.5%), hypoxia (4.5%), other: relapse/progressive disease (4.1%).

The registry reports serious adverse events in 15 of 66 participants in the sirolimus-containing arm and 15 of 73 in the sirolimus-free arm. It reports other adverse events in 41 of 66 and 54 of 73, respectively. These are counts of participants affected relative to the listed group sizes; the record does not provide a full description of event collection periods, severity grading beyond the serious-event grouping, or statistical comparisons between arms.

The most frequently listed serious-event terms include infection at 12.1%, GVHD at 5.5%, acute kidney injury/renal failure at 4.5%, hypoxia at 4.5%, relapse/progressive disease at 4.1%, and thrombocytopenia at 3.0%. The record does not clarify whether people could contribute more than one event term, so the terms should not be added or interpreted as mutually exclusive categories. Death counts in the adverse-event summary are not reported in the registry record.

These figures describe reported adverse events in this trial; they do not establish that a regimen caused a particular event or that the arms differ in safety. Randomized group comparisons and the way events were collected would be needed for stronger attribution. The posted safety counts are descriptive, and the registry record does not provide adjusted safety analyses or confidence intervals.

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

The posted result describes 2-year overall survival estimates of 70.0% for the sirolimus-containing regimen and 68.0% for the sirolimus-free regimen. The unadjusted risk difference is -0.012, with a 95% confidence interval from -0.166 to 0.142 and an unadjusted p-value of 0.878. The estimate is close to zero, but the interval is wide enough to include differences in either direction. The data therefore do not demonstrate a clear difference in the primary outcome, while also not establishing that the regimens are the same.

The result applies to the trial population and the way the endpoint was recorded in the registry. It should not be generalized automatically to other transplant populations, lymphoma subtypes, GVHD-prevention regimens, or care settings. The registry lists Hodgkin and non-Hodgkin lymphoma, but the posted primary result is not separated by histology. The record does not report subgroup estimates that would show whether the overall result varies by lymphoma type.

The clinical rationale involved the possibility that sirolimus could influence lymphoma relapse, but overall survival alone cannot identify whether relapse changed. It combines fatal outcomes from any cause. The posted secondary results include selected progression-free survival and chronic GVHD findings, but those do not isolate all pathways that could affect survival, and other secondary outcomes are not posted. The available summaries cannot determine why the arm estimates look as they do.

Nor does the result establish that sirolimus is beneficial, ineffective, safe, or harmful. The trial's randomized design is relevant to causal comparison, but the registry record lacks the detailed analysis plan, analysis population and missing-data methods needed for a complete appraisal. The defensible interpretation is limited to the posted estimates and their uncertainty: the primary comparison is imprecise and does not show a clear between-arm difference in 2-year overall survival.

17. Why This Trial Matters Statistically

This trial illustrates why a point estimate, an uncertainty interval and a p-value should be read together. The arm estimates are numerically close, but the confidence interval for the risk difference is broad and includes zero. Reporting only the point estimates could make the result appear more definitive than it is; reporting only the p-value would conceal the range of effects compatible with the data. Together, these statistics show that the observed comparison is not precise enough to rule out meaningful differences in either direction.

The study also highlights the distinction between a biological rationale and a clinical endpoint. Sirolimus was considered in the context of possible anti-lymphoma activity, but the primary endpoint was overall survival, which reflects all causes of death. Secondary endpoints can help explore progression and transplantation complications, yet multiple outcomes create interpretive challenges when a testing hierarchy or multiplicity plan is not reported. A trial may therefore have a clear clinical motivation while still leaving uncertainty about which pathway explains an observed result.

Finally, a randomized phase III design does not by itself make every registry summary sufficient for full statistical evaluation. The record identifies the broad comparison and posts selected estimates, but does not report the primary analysis, missing-data approach or analysis population. Those details affect whether the posted unadjusted comparison aligns with the study's intended analysis. The example underscores the value of transparent endpoint definitions, analysis methods, and uncertainty reporting alongside trial results.

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

Several limitations are specific to the information available for this trial. The primary result is presented as a percentage at 2 years, but the registry does not report the estimator or explain how incomplete follow-up, censoring or deaths before the assessment time were handled. That limits independent evaluation of the endpoint calculation. The record also does not describe the analysis population or whether participants were analyzed according to randomized assignment.

The sirolimus-free arm could use either tacrolimus with methotrexate or cyclosporine with mycophenolate mofetil, while the posted results combine these options. The registry record does not explain how the choice was made or provide separate results for the two control regimens. The comparison therefore concerns the reported combined sirolimus-free arm, not an independently estimated contrast against each alternative.

Results are not posted for every secondary objective. In particular, the record lists acute GVHD, disease progression, non-relapse mortality and histology-specific outcomes, but does not provide their results. The selected secondary results cannot show the full range of planned clinical outcomes. The multiplicity strategy is also not reported, so the evidential status of secondary comparisons is uncertain.

The trial was unmasked, as recorded in the registry. The implications of that feature depend on how outcomes were ascertained and how treatment knowledge could affect reporting or assessment; those details are not reported in the registry record. The posted adverse-event summary lacks event timing and arm-level comparative analyses, and does not report deaths in that summary. These limitations concern the completeness and interpretability of the registry information and should be kept distinct from the trial's randomized design.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Tacrolimus/Sirolimus/Methotrexate vs Tacrolimus/Methotrexate or Cyclosporine/Mycophenolate Mofetil for GVHD Prophylaxis After Reduced Intensity Allogeneic Stem Cell Transplantation for Patients With Lymphoma — NCT00928018.
• Registry record last updated 2019-02-01; status last verified 2019-01.
• Results first posted 2015-12-18.
• 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

NCT00928018 is a completed, randomized, unmasked, parallel-group phase III trial sponsored by Dana-Farber Cancer Institute. It compared a GVHD-prevention regimen containing tacrolimus, sirolimus and low-dose methotrexate with a sirolimus-free regimen using tacrolimus and methotrexate or cyclosporine and mycophenolate mofetil after reduced-intensity allogeneic stem cell transplantation for lymphoma. Registry enrollment is 139; the posted primary outcome lists 66 participants in the sirolimus-containing arm and 73 in the sirolimus-free arm.

The primary endpoint was 2-year overall survival. The posted estimates were 70.0% and 68.0%, with a risk difference of -0.012, a 95% confidence interval from -0.166 to 0.142, and an unadjusted p-value of 0.878. The interval is broad and includes zero, so the comparison does not demonstrate a clear difference or establish equivalence. Selected secondary outcomes and descriptive adverse-event counts are also available, while several planned results and important analysis details are not reported in the registry record.