Treatment of Peripheral T-cell Lymphoma: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Treatment of Peripheral T-cell Lymphoma (NCT01664975), a randomized phase 4 study. The primary outcome is Progression-free Survival.

2. Clinical Question

3. Trial Design

The registry describes a randomized, parallel-group, phase 4 study of treatment for peripheral T-cell lymphoma. Its official title identifies it as a multicenter clinical trial, and the listed interventions are GDPT and CHOP regimens. The study status is completed. The registry gives an enrollment of 100 and reports arm denominators of 52 for GDPT and 51 for CHOP; these are the denominators attached to the posted primary-outcome counts. The record does not explain how the enrollment figure relates to the arm denominators.

Masking is listed as none, so the study is recorded as open-label. Who knew each participant’s assigned regimen, whether outcome assessors were masked, and whether any other safeguards against assessment bias were used are not reported in the registry record. The randomization sequence method, allocation concealment, and any stratification factors are also not reported.

The recorded start date is 2011-08, and primary completion is listed as 2016-04. The primary outcome is progression-free survival, expressed as a number of participants and assessed up to the end of the follow-up phase, approximately 24 months. The posted arm results are counts: 35 of 52 for GDPT and 27 of 51 for CHOP. The registry also names response rate, overall survival, and median survival time as secondary outcomes. Details about their definitions, assessment schedules, and results are not reported in the registry record.

The record lists no analysis entries, although it does include arm-level outcome results and comparative estimates. The original statistical analysis plan, including any adjusted model or prespecified approach to handling missing observations, is not reported. Consequently, the posted unadjusted comparison is useful as a descriptive contrast, but it cannot establish whether it matches the trial’s planned primary analysis.

4. Participant Flow and Accounting

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

Overall Study

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Progression-free Survival, measured over up to end of follow-up-phase (approximately 24 months).

Progression-free survival is the primary outcome named in the registry. In general, this kind of endpoint is intended to describe the time participants remain alive without disease progression. Its use can capture events that matter to the course of illness before an overall-survival difference can be assessed. However, the registry reports the primary measure as a number of participants with an event, rather than posting a time-to-event summary such as a hazard ratio or a survival curve.

The stated timeframe is up to the end of the follow-up phase, approximately 24 months. That wording does not mean every participant was observed for the same length of time; individual follow-up durations and censoring information are not reported in the registry record. The exact event definition—such as the criteria for progression, whether death is included, and how events were adjudicated—is also not reported. These details matter because the interpretation of a progression-free-survival result depends on what counts as an event and how follow-up is handled.

The posted counts are 35 of 52 in the GDPT group and 27 of 51 in the CHOP group. The binary comparison therefore treats the registry’s recorded outcome as an event count. It does not preserve the timing of events, so it cannot by itself describe when events occurred or account for differences in follow-up in the way a time-to-event analysis can. The stated endpoint and the recorded counts should be interpreted within those limits.

7. Results Posted in the Registry

• CHOP Regimen: 52.9% (27/51) versus 67.3% (35/52) on control; risk difference -14.4 percentage points (95% CI -33.1 to 4.3), risk ratio 0.79 (0.57–1.08); unadjusted two-proportion z-test p = 0.136.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary in the posted results: participants are counted as having or not having the registry-recorded event. A binary analysis compares the event frequency between groups. The arm counts are 35 of 52 for GDPT and 27 of 51 for CHOP. These counts retain whether an event was recorded, but do not describe the timing of that event. In contrast, a time-to-event analysis uses both event occurrence and the time until an event or censoring. Although the outcome is called progression-free survival, the registry posts a participant count and does not report a time-to-event estimate.

The reported risk difference is an absolute contrast in event proportions. Its sign depends on the order of the groups in the comparison. Here, CHOP is compared with GDPT, and the estimate is -0.144. In general, an absolute difference is often easier to relate to a change in event frequency than a relative measure, but its clinical interpretation depends on the outcome definition, follow-up, and context. The registry does not report a clinically important threshold for interpreting this difference.

The risk ratio is a relative comparison of the event proportions, reported as 0.787, and the odds ratio compares event odds, reported as 0.546. These measures answer related but distinct questions. A risk ratio compares probabilities, whereas an odds ratio compares odds. Neither measure makes the result causal by itself: the design and conduct of the study, outcome ascertainment, missingness, and analysis choices all affect interpretation. The corresponding confidence intervals are important alongside the point estimates because they show the uncertainty compatible with the observed data.

The unadjusted p-value is 0.136. A p-value describes how unusual data at least as discrepant as those observed would be under a specified null model and its assumptions. It is not the probability that the null hypothesis is true, nor a measure of the size or importance of an effect. A p-value should be read with the estimate and interval, not as a stand-alone decision rule. The registry does not report the original planned test or a detailed statistical analysis plan. The comparative estimates are therefore best understood as unadjusted calculations from the posted summary counts.

Randomization is intended to make treatment groups comparable on average, including on characteristics that may not be measured. It does not guarantee that groups will be identical, particularly in a study of this size, and does not prevent bias from differential follow-up, outcome assessment, or missing data. The record identifies the study as randomized but does not report the sequence-generation method, allocation concealment, or baseline characteristics. Masking is listed as none, so awareness of treatment assignment could potentially influence care or outcome assessment; whether that occurred is not established by the record.

Confidence intervals communicate a range of values compatible with the observed data under the method used. Their width reflects both the amount of information and variability in the data, as well as the statistical procedure. The specific interval method is not reported in the registry record. General statistical interpretation should therefore avoid assuming a particular model or adjustment. The posted intervals for the risk difference, risk ratio, and odds ratio each include the respective no-difference value.

Over-dispersion is a general issue that can arise when outcomes vary more than a simple statistical model assumes. If present and ignored, uncertainty may be understated and intervals may be too narrow. For this posted binary comparison, the registry does not report an over-dispersion assessment or a model that would permit evaluation of it. This is not evidence that over-dispersion occurred; it is a reminder that model assumptions matter and are not fully documented in the record.

Several interpretive details remain unavailable. The registry does not report how progression was defined, how event dates were determined, how participants lost to follow-up were handled, or whether participants were analyzed according to randomized assignment. It also does not report an adjusted comparison or whether any planned adjustment was used. Because these details can affect estimates and their uncertainty, the posted unadjusted counts and comparisons should be treated as limited summaries rather than a full reconstruction of the trial’s analysis.

10. Confidence Intervals and Precision

The confidence interval for the risk difference, -0.331 to 0.043, describes uncertainty around the unadjusted estimate of -0.144 for CHOP compared with GDPT. The interval crosses zero, the value representing no absolute difference. Thus, the observed data are compatible with no difference and also with effects in either direction. The interval’s span makes clear that the point estimate alone does not capture the uncertainty in this comparison.

The relative measures also have intervals that include their no-difference value. The risk ratio is 0.787, with an interval from 0.571 to 1.084; a risk ratio of 1 represents equal event risks. The odds ratio is 0.546, with an interval from 0.246 to 1.215; an odds ratio of 1 represents equal event odds. Those estimates are not interchangeable with the risk difference, but each interval indicates that the data do not pin down a single direction of effect with precision.

A confidence interval is not a guarantee that the true value lies within its bounds, nor does it describe the probability that a particular hypothesis is true. In general, it reflects the range of estimates reasonably compatible with the observed data under the method’s assumptions. The registry record does not state the interval-construction method, so the intervals should be interpreted as posted rather than attributed to a particular procedure.

The arm counts and uncertainty indicate that precision is limited. The registry does not report the event-definition details, follow-up distribution, missing-data approach, or an adjusted analysis, each of which could affect how informative the estimate is. The p-value of 0.136 should not replace consideration of the interval: the interval makes visible that a wide range of effects remains compatible with the data.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm for this binary outcome, not total events. They are illustrations based on the observed effect, rather than the trial’s original design calculation. The 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 set to the stated fraction of the observed effect. The trial’s original sample-size justification and design assumptions are not reported in the registry record.

For planning at 100% of the observed effect, the illustration gives 182 participants per arm for 80% power and 243 participants per arm for 90% power. At 75% of the observed effect, it gives 318 participants per arm for 80% power and 425 participants per arm for 90% power. At 50% of the observed effect, it gives 703 participants per arm for 80% power and 940 participants per arm for 90% power.

In plain terms, the 80% power column describes the planned chance of detecting the chosen effect under the stated assumptions if that effect were real; the 90% column uses a higher target chance of detection. These are prospective planning quantities, not evidence about what the completed study was capable of detecting after its results were known. This is not observed power.

The pattern illustrates a general planning principle: smaller effects require more participants to detect, when the other assumptions are held fixed. The estimates are sensitive to the effect size chosen and to whether the assumptions are suitable for a future study. They should not be read as a recommendation or as a reconstruction of the trial’s actual design process.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.673; treatment proportion p₁ = 0.529; average p̄ = 0.601.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6925 = 1.3572.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6850 = 0.5765.
4. n = (1.3572 + 0.5765)² / (-0.1437)² = 181.2, rounded up to 182 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 <- 0.673076923076923                          # control proportion
p1 <- 0.529411764705882                          # 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(1664975)
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 = 182 (80% power), 243 (90% power)
 75% of observed effect: n per arm = 318 (80% power), 425 (90% power)
 50% of observed effect: n per arm = 703 (80% power), 940 (90% power)
power.prop.test: 181.2 
Simulated power at n = 182 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.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome, progression-free survival, and names three secondary outcomes: response rate, overall survival, and median survival time. It also records four outcome measures. The secondary outcome names alone do not supply their definitions, analysis results, or the order in which they were assessed. The record does not describe whether additional outcomes or subgroup analyses were planned.

When a study evaluates multiple outcomes or makes multiple comparisons, the chance of obtaining at least one apparently unusual result can increase. Statistical approaches can address that issue, but the choice depends on the prespecified questions and analysis plan. No multiplicity strategy, hierarchy among outcomes, or adjustment for multiple testing is reported in the registry record. It is therefore not possible to determine whether the posted primary comparison was treated as the sole confirmatory test or how secondary results were intended to be interpreted.

The analysis population is not reported. The registry does not say whether participants were analyzed according to randomized assignment, according to treatment received, or under another population definition. The randomization description alone does not settle that question. The numbers attached to the primary outcome are 52 for GDPT and 51 for CHOP, but the record does not explain exclusions, deviations, or the relationship between those denominators and the enrollment figure.

Missing-data methods are also not reported. There is no description of the number or pattern of missing outcome observations, loss to follow-up, censoring, imputation, or sensitivity analyses. These details are especially relevant when interpreting an outcome related to follow-up duration. Without them, the degree to which missingness could affect the comparison cannot be assessed.

The registry lists no analysis entries, and it does not report a detailed statistical analysis plan. This absence limits assessment of prespecification, multiplicity, and analysis populations; it does not establish that no analyses were conducted. The posted unadjusted comparison can be interpreted as a summary of the recorded counts, but should not be assumed to represent every planned or performed analysis.

14. Safety Results

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

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: grade3/4 myelosuppression (44.2%).

The registry reports serious adverse events for 23 of 52 participants in the GDPT group and 21 of 51 in the CHOP group. It separately reports other adverse events for 21 of 52 participants in GDPT and 19 of 51 in CHOP. These are counts of participants recorded as affected within each category. The record does not define the categories, give event-level detail, or explain how participants with multiple events were handled.

The registry lists Grade3/4 myelosuppression as a top serious adverse-event term, with 44.2%. The denominator and the method used to select or calculate this term are not reported in the registry record. The figures therefore should not be treated as directly comparable to the participant counts without further clarification. Death counts are not reported.

These data describe reported adverse events; they do not, by themselves, establish that a regimen caused a particular event or permit a complete comparison of safety. The record does not provide event timing, duration, resolution, severity distributions beyond the listed term, withdrawals due to adverse events, or exposure-adjusted rates. Interpretation is consequently limited to the counts and term recorded, rather than a comprehensive account of relative safety.

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

The posted primary-outcome results show recorded events in both treatment groups: 35 of 52 participants assigned to GDPT and 27 of 51 assigned to CHOP. The unadjusted risk difference is -0.144 for CHOP compared with GDPT, and the interval from -0.331 to 0.043 includes no difference. The unadjusted p-value is 0.136. Together, these statistics describe an observed contrast with substantial uncertainty; they do not establish that the groups differ, and they do not establish equivalence.

The result also does not answer every question implied by the endpoint name. The registry posts a participant count rather than time-to-event results, and the operational definition of an event is not reported. The stated follow-up timeframe is approximately 24 months, but participant-level follow-up and censoring are not described. The posted comparison therefore cannot show when events occurred or whether follow-up differed between groups.

The data do not establish a causal treatment effect on their own. Randomization is an important design feature, but interpretation also depends on allocation procedures, participant follow-up, outcome assessment, and analysis choices. The registry does not report several of these details, including the analysis population and missing-data handling. Masking is listed as none, while any protections against bias in outcome assessment are not described.

The registry names response rate, overall survival, and median survival time as secondary outcomes, but their results are not reported in the record. The primary comparison should not be taken as a substitute for those outcomes or as a complete summary of efficacy. Safety data are also limited to reported participant counts and a listed serious-event term, without the contextual details needed for a full assessment.

In short, the record gives a numerical comparison of the posted primary-outcome counts and indicates that the estimate is imprecise. It does not support a definitive conclusion about the comparative effects of GDPT and CHOP, nor does it resolve questions about the endpoint definition, missing information, or the trial’s full analysis plan.

16. Why This Trial Matters Statistically

This study illustrates why a randomized comparison should be read through both its point estimate and its uncertainty. The posted counts differ, and the risk difference is negative for the stated CHOP-versus-GDPT comparison. Yet the confidence interval includes zero and spans effects in both directions. A visible difference in the arm counts is not, by itself, enough to establish that a treatment-group difference is reliably present.

It also highlights a gap that can arise between an endpoint’s name and its posted statistical form. Progression-free survival is a time-to-event concept, but the registry reports a number of participants with an event. Without event dates, censoring information, and an explicit definition, readers cannot recover the timing component from the count. Knowing what is counted is as important as knowing the label applied to an outcome.

The record illustrates the value of transparent reporting. Randomized allocation and a stated follow-up timeframe provide useful context, but details about the sequence method, analysis population, missing data, multiplicity, and endpoint adjudication are not reported. Those omissions do not prove a problem occurred; they constrain what can be assessed from the registry.

Finally, the planning illustration demonstrates how the assumed effect size shapes sample-size requirements. Smaller target effects require more participants under the stated planning assumptions. This is distinct from interpreting the completed trial’s observed result. Keeping design planning, observed estimates, and uncertainty separate helps prevent a planning calculation from being mistaken for evidence that a study detected or failed to detect a particular effect.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several limitations specific to this record constrain interpretation. The primary outcome is named progression-free survival, but the posted measure is a binary participant count. The registry does not define the event criteria, indicate whether death is included, or report event timing, censoring, or adjudication. The stated follow-up extends to the end of the follow-up phase, approximately 24 months, but the distribution of individual follow-up is not reported.

The study is described as randomized and open-label, but the randomization sequence method, allocation concealment, and any stratification are not reported. The record does not describe masking of outcome assessors or other measures to limit assessment bias. Baseline characteristics and eligibility details are also not included in the registry record reported for this analysis, limiting assessment of group comparability and applicability.

The registry does not report the trial’s original statistical analysis plan, sample-size justification, analysis population, or missing-data procedures. It lists no analysis entries, while posting arm counts and comparative estimates. As a result, it is not possible to confirm whether the unadjusted comparison corresponds to the prespecified primary analysis or to assess exclusions, deviations, and the impact of missing follow-up.

Secondary outcomes are named, but their results are not reported. The record also does not describe any multiplicity approach, subgroup analyses, interim analyses, or outcome hierarchy. These are not assumed to have occurred; their status is simply not reported in the registry record. Safety reporting is limited: event definitions and timing are absent, the denominator for the 44.2% serious-event term is unclear, and death counts are not reported.

These limitations do not invalidate the recorded counts, but they narrow the conclusions that can be drawn from them. The posted estimates should be interpreted as unadjusted summaries of the registry results, with substantial uncertainty and incomplete methodological context.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Treatment of Peripheral T-cell Lymphoma — NCT01664975.
• Registry record last updated 2016-10-05; status last verified 2016-08.
• Results first posted 2016-10-05.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

NCT01664975 is a completed, phase 4, randomized, open-label, parallel-group study of GDPT and CHOP regimens for peripheral T-cell lymphoma. The official title describes a multicenter trial. The registry lists enrollment of 100, a start date of 2011-08, and primary completion in 2016-04.

The primary outcome is progression-free survival, reported as participant counts through the end of follow-up, approximately 24 months. The posted counts are 35 of 52 for GDPT and 27 of 51 for CHOP. The unadjusted risk difference for CHOP compared with GDPT is -0.144, with a confidence interval from -0.331 to 0.043 and p = 0.136. The interval includes no difference.

The registry names response rate, overall survival, and median survival time as secondary outcomes. It reports serious adverse events in 23 of 52 GDPT participants and 21 of 51 CHOP participants, and other adverse events in 21 of 52 and 19 of 51, respectively. Further outcome definitions, analysis details, and missing-data methods are not reported in the registry record.