Study of Pralatrexate & Gemcitabine With B12 & Folic Acid to Treat Relapsed/Refractory Lymphoproliferative Malignancies: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of Pralatrexate & Gemcitabine With B12 & Folic Acid to Treat Relapsed/Refractory Lymphoproliferative Malignancies (NCT00481871), a randomized phase 1, phase 2 study. The primary outcome is Objective Responses Assessed by International Workshop Criteria (IWC).

2. Clinical Question

3. Trial Design

The registry describes a Phase 1/2a, open-label study of pralatrexate and gemcitabine with vitamin B12 and folic acid supplementation in patients with relapsed or refractory lymphoproliferative malignancies. The study status is completed. The record lists the start as 2007-05 and primary completion as 2011-05, with enrollment of 119. The listed conditions include Hodgkin's lymphoma, peripheral T-cell lymphoma, B-cell lymphoma, and Waldenstrom's macroglobulinemia.

Three reported efficacy groups are Phase 1, Phase 2 - Group B, and Phase 2 - Group C. Their primary endpoint counts are 8.0 of 34.0, 5.0 of 38.0, and 7.0 of 34.0, respectively. These are response counts and analysis denominators, not the overall enrollment figure. The registry says the efficacy analysis included all patients who completed at least 1 cycle of treatment; the response definition also required measurable disease at baseline. It does not explain the relationship between that analysis population and overall enrollment.

The registry lists allocation as randomized, masking as none, and the model as single group. Those descriptors do not, on their own, establish the specific assignment process or whether the Phase 1 and Phase 2 groups were intended as directly randomized comparisons. The arm descriptions are empty, so the schedule distinction implied by the group names cannot be reconstructed from the record. Details of the original sample-size justification, treatment allocation procedures, and statistical plan are not reported in the registry record. Duration of response and progression-free survival time are listed as secondary outcomes, but results for those outcomes are not included in the posted results summarized for the 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.

Before assignment: 12 patients were enrolled but not treated. Of these, 9 patients had events after enrollment that rendered them ineligible; 2 patients had progressive disease (PD); 1 patient withdrew consent. Since they were never dosed, these 12 patients were not included in efficacy or safety assessments.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: All patients who completed at least 1 cycle of treatment were included in the efficacy analysis

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 Objective Responses Assessed by International Workshop Criteria (IWC), measured over Assessed every 8 weeks (+/- 1 week) for Phase II and no less than every 3 cycles for Phase I.

The primary endpoint counts participants who achieved an objective response, defined as a tumor response assessment of complete response or partial response under International Workshop Criteria. It is a binary outcome at the participant level: a participant either met the response definition or did not. The reported measure is the number of participants with an event, rather than a continuous change in tumor size or a time-to-event measure.

Response was assessed every 8 weeks (+/- 1 week) for Phase II and no less than every 3 cycles for Phase I. Only patients with measurable disease at baseline were eligible for the response determination. If an International Workshop Criteria assessment with PET was not done, the registry says an International Workshop Criteria assessment without PET was used for analyses.

Objective response is used in clinical studies to summarize whether a predefined level of tumor response was observed. It can provide a direct count for each analysis group, but it does not describe how long a response lasted or how long participants lived without progression. Duration of response and progression-free survival time are listed as secondary outcomes in this study; their results are not reported in the posted results used for this analysis. The endpoint also does not establish that a response was caused by treatment, especially when the group allocation and comparability details are not fully described in the registry record.

7. Results Posted in the Registry

• Phase 2 - Group B: 13.2% (5/38) versus 23.5% (8/34) on control; risk difference -10.4 percentage points (95% CI -28.2 to 7.5), risk ratio 0.56 (0.20–1.55); unadjusted two-proportion z-test p = 0.253.
• Phase 2 - Group C: 20.6% (7/34) versus 23.5% (8/34) on control; risk difference -2.9 percentage points (95% CI -22.6 to 16.8), risk ratio 0.88 (0.36–2.14); unadjusted two-proportion z-test p = 0.770.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary because it records whether a participant achieved an objective response. For this kind of outcome, the registry reports event counts and group denominators. The posted counts are 8.0 among 34.0 in Phase 1, 5.0 among 38.0 in Phase 2 - Group B, and 7.0 among 34.0 in Phase 2 - Group C. A count is easiest to interpret alongside its denominator, since the groups do not have identical sizes.

What does a risk difference describe?

A risk difference compares the response proportions in two groups by subtracting one proportion from the other. Its null value is zero, which means the estimated proportions are the same. The registry reports risk-difference estimates for Phase 2 - Group B and Phase 2 - Group C relative to Phase 1. Each estimate is negative, indicating a lower observed response proportion in the Phase 2 group at the point estimate. The confidence interval for each comparison includes zero, so the estimate is uncertain about whether the underlying difference is negative, positive, or absent.

A risk difference is expressed on the absolute scale. This makes it possible to ask how far apart the response proportions are, rather than how many times as large one is relative to the other. The reported interval communicates uncertainty around that estimated difference. It should be read as a range of values compatible with the data and the method used to construct the interval, not as a guarantee that the true value lies within it. The confidence level and the specific interval-construction method are not reported in the registry record.

Why are risk ratios and odds ratios also useful?

A risk ratio compares the response proportions as a ratio; its null value is one. An odds ratio compares the odds of response between groups and also has a null value of one. These measures describe relative rather than absolute differences. Their intervals in the posted comparisons include one, so they too leave the possibility of no group difference open. Risk ratios and odds ratios can be useful alternate summaries, but they do not remove the uncertainty arising from the small response counts or clarify whether the groups are causally comparable.

The risk difference, risk ratio, and odds ratio answer related but distinct questions. A relative measure can look different in scale from an absolute measure, even when both arise from the same observed data. For this reason, interpretation should consider the underlying event counts and denominators along with the selected effect measure. No measure should be read as a treatment effect without adequate information about group assignment and study design.

How should the p-values be read?

The registry posts unadjusted p-values of 0.253 for Phase 2 - Group B and 0.77 for Phase 2 - Group C. A p-value describes how unusual results at least this incompatible with a specified null model would be if that model were true and its assumptions held. It is not the probability that the null hypothesis is true, and it is not the probability that the observed finding occurred by chance. These p-values do not quantify effect size or precision; the estimates and their intervals address those questions more directly.

The p-values are labeled unadjusted. The registry record does not report a multiplicity plan or an adjustment for making more than one comparison. If several outcomes or comparisons are considered, repeated testing can affect the chance of obtaining at least one small p-value under a set of null hypotheses. The record does not provide enough information to determine whether or how that issue was addressed in the study's original analysis.

What design information matters for these comparisons?

The registry labels allocation randomized, masking none, and the model single group. It does not describe the assignment process or explain how those labels relate to the Phase 1 and Phase 2 groups. A randomized comparison can support causal interpretation when the groups are defined and analyzed according to an appropriate allocation process, but that process cannot be reconstructed from these registry entries. Because the phase-specific group descriptions are incomplete, the reported comparisons are best treated as comparisons of the posted results, with caution about causal interpretation.

The registry states that the efficacy population consisted of patients who completed at least 1 cycle. It also says that only patients with measurable disease at baseline were considered for the response determination. These eligibility and analysis details define what the primary endpoint counts mean. The record does not report how missing assessments, withdrawals, or participants who did not complete a cycle were handled. They should not be assumed to have been treated as nonresponders or excluded under a particular rule beyond the stated population description.

In general, over-dispersion matters when observed outcome variability is greater than the model assumes: standard errors may then be understated, and intervals may be too narrow if the model does not account for that extra variation. This is a general statistical consideration for models of event counts or proportions, not a reported feature or method of this study. The registry does not report a model-based analysis or a check for over-dispersion for the primary endpoint.

10. Confidence Intervals and Precision

Precision concerns how much uncertainty surrounds an estimate, not whether the estimate is important or whether a treatment caused the observed result. The posted risk-difference intervals for Phase 2 - Group B and Phase 2 - Group C both include zero. Thus, each interval is compatible with no difference and with differences in either direction. The risk-ratio and odds-ratio intervals also include their null value of one. Taken together, these intervals do not establish a clear direction or tightly determine the magnitude of a group difference.

The response counts are limited: 8.0 in Phase 1, 5.0 in Phase 2 - Group B, and 7.0 in Phase 2 - Group C. With few observed events, an estimate may change appreciably if the outcome status of individual participants differs. The different group denominators also matter when interpreting raw counts. The risk-difference estimates account for proportions, but the uncertainty remains substantial enough that the data do not sharply distinguish among possible differences.

The confidence level and interval method are not reported in the registry record. The intervals should therefore be interpreted as the posted uncertainty intervals without assigning an unreported confidence level or reconstruction method. Precision is also constrained by the incomplete design details: the registry does not explain how randomization relates to the phase-specific groups. Even a narrow interval would not resolve potential comparability concerns; conversely, uncertainty in these intervals is a reason to avoid strong claims about group differences. The observed estimates are compatible with more than one interpretation, and the record does not provide posted secondary-outcome analyses to add information about response duration or progression-free survival.

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 sample-size figures are planning illustrations based on the observed effect, not the trial's own design justification. For this binary outcome, the figures refer to participants per arm. The planning assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect: 100%, 75%, or 50%. The original trial's sample-size assumptions are not reported in the registry record.

The 80% power column gives the planned number of participants per arm for a hypothetical study with an 80% chance of detecting the chosen effect under the stated assumptions, if that effect is real and the assumptions hold. The 90% power column gives the corresponding planning number for a 90% chance. These are design targets, not observed outcomes. The figures are not observed power.

Planning for a smaller effect requires more participants per arm than planning for a larger effect, because a smaller difference is harder to distinguish from random variation. The figures for 75% and 50% of the observed effect therefore rise relative to planning for 100% of the observed effect. The displayed planning values are for two-arm comparisons; they do not establish the appropriate size for the original Phase 1/2a study or account for details not included in the listed assumptions. The registry does not report that these planning calculations were used in designing or conducting the trial. The posted primary endpoint is a participant event count, so the relevant unit is participants per arm, not total events. The separate total-events convention applies to time-to-event planning illustrations, not this binary endpoint.

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 = 0.235; treatment proportion p₁ = 0.132; average p̄ = 0.183.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5473 = 1.0728.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5424 = 0.4565.
4. n = (1.0728 + 0.4565)² / (-0.1037)² = 217.4, rounded up to 218 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 <- 0.235294117647059                          # control proportion
p1 <- 0.131578947368421                          # 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(481871)
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 = 218 (80% power), 291 (90% power)
 75% of observed effect: n per arm = 409 (80% power), 547 (90% power)
 50% of observed effect: n per arm = 966 (80% power), 1,292 (90% power)
power.prop.test: 217.4 
Simulated power at n = 218 per arm: 0.804 (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 the primary objective response outcome and names duration of response and progression-free survival time as secondary outcomes. It records three outcome measures and two posted secondary outcomes, but no analysis entries are posted. The primary response comparisons include unadjusted p-values. A multiplicity adjustment or prespecified hierarchy for outcomes and comparisons is not reported in the registry record.

Multiplicity is a general concern when several hypotheses are tested: the chance of at least one false-positive result can increase as the number of tests grows. Whether an adjustment is needed, and which method is suitable, depends on the study's prespecified objectives and testing strategy. Because that strategy is not reported, the unadjusted p-values should not be treated as if they were adjusted for all outcomes or comparisons. Nor is it possible to determine from the registry record whether additional analyses were conducted but not posted.

The stated efficacy population includes all patients who completed at least 1 cycle of treatment. For the response determination, patients also needed measurable disease at baseline. The registry does not report detailed rules for missing response assessments, discontinuations, or other missing data. It does not say how participants lacking an assessment were handled, whether missing outcomes were imputed, or whether sensitivity analyses were performed. No assumption about those procedures should be inferred from the event counts.

The analysis denominators are 34.0 for Phase 1, 38.0 for Phase 2 - Group B, and 34.0 for Phase 2 - Group C. The overall enrollment is 119, and the registry does not explain the difference between enrollment and the reported group denominators. The denominators therefore need to be interpreted within the stated efficacy population rather than as the full enrolled sample. Details of any intention-to-treat population, per-protocol population, covariate adjustment, subgroup analysis, or missing-data method are not reported in the registry record.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: pyrexia (20.6%), dyspnoea (11.8%), cellulitis (8.6%), pneumonia (8.6%), renal failure acute (5.9%).

The serious adverse event figures are counts of participants affected among those at risk in each group. Phase 1 reports 18 of 35; Phase 2 - Group B reports 13 of 38; Phase 2 - Group C reports 16 of 34. These are separate safety denominators and should not be substituted for the primary efficacy denominators. The registry also reports other adverse events affecting all 35, all 38, and all 34 participants at risk in the respective groups.

The leading serious-event terms listed are pyrexia at 20.6%, dyspnoea at 11.8%, cellulitis at 8.6%, pneumonia at 8.6%, renal failure acute at 5.9%, and hypoxia at 5.9%. These are registry-reported percentages for terms, not a complete account of every event or its clinical course. The record does not establish from these summaries whether an event was caused by study treatment, and the event counts do not describe severity beyond the serious-event classification.

Death counts are not reported in the registry record. Exposure duration, event timing, recurrence, and detailed event definitions are also not reported in the safety summary. Group counts should therefore be read as descriptive registry results, not as adjusted comparisons of safety risk or proof of a causal relationship.

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

The posted results establish that objective responses were recorded in each reported efficacy group: 8.0 in Phase 1, 5.0 in Phase 2 - Group B, and 7.0 in Phase 2 - Group C. The comparison estimates are lower for each Phase 2 group relative to Phase 1, but the risk-difference intervals include zero and the ratio intervals include one. The results therefore leave substantial uncertainty about the size and direction of any difference.

These data do not establish that the regimen works or does not work. A nonsmall p-value is not proof that groups are equivalent, and an observed response count does not by itself establish a causal treatment effect. The record's randomized-allocation label is not accompanied by sufficient detail to explain how assignment operated across the phase-specific groups. The listed single-group model and empty arm descriptions add to that uncertainty about the intended comparisons.

The primary endpoint captures complete or partial response under the stated assessment criteria among patients with measurable disease at baseline. It does not indicate how long any response lasted, nor does it directly answer whether progression was delayed. Although duration of response and progression-free survival time appear as secondary outcomes, their analyses are not posted in the registry record. The response endpoint also does not summarize the full benefit-risk profile.

Interpretation is limited to the stated efficacy population, including patients who completed at least 1 cycle of treatment. The registry does not describe the handling of missing assessments or explain how this population relates to all enrolled participants. The safety counts provide separate descriptive information, but the record does not establish causality for reported adverse events. Overall, the posted results are best understood as a small set of response and safety summaries with imprecise comparisons and incomplete design and analysis details.

16. Why This Trial Matters Statistically

This trial illustrates why the label attached to a design cannot replace a clear description of how participants were assigned and how groups relate to one another. The registry lists randomized allocation, no masking, and a single-group model, while presenting Phase 1 and Phase 2 groups for comparison. The assignment details and arm descriptions are not reported. That ambiguity affects how strongly the group estimates can be interpreted as causal contrasts.

It also illustrates the importance of reading event counts together with denominators and uncertainty. The response counts are modest, and the intervals include the null values for the posted effect measures. Reporting a point estimate alone would hide the range of values compatible with the data. Reporting a p-value alone would also omit the magnitude and precision of the estimated difference.

Finally, the record distinguishes a primary outcome from secondary outcomes but does not post analyses for the latter or describe a multiplicity plan. That gap matters because endpoint selection and analysis strategy shape how readers interpret evidence across a study. The sample-size illustrations further show the general planning consequence of targeting smaller effects: more participants per arm are needed under the stated assumptions. Those illustrations are not the original study's justification. Together, these features demonstrate why transparent reporting of design, analysis populations, missing-data handling, and prespecified statistical methods is essential for interpreting clinical trial results.

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.

The registry record does not give the treatment schedules or explain the difference between the Phase 2 groups, despite their distinct names. It also does not explain how the randomized-allocation designation applies to the phase-specific groups, or how the comparisons were planned. This makes it difficult to judge group comparability and limits causal interpretation.

The primary analysis is defined for patients completing at least 1 cycle, and response assessment requires measurable disease at baseline. The record does not explain how these criteria relate to the full enrolled population or how missing assessments and withdrawals were handled. It does not provide detailed baseline characteristics, subgroup results, or adjusted analyses in the posted information used for these comparisons.

Only the objective response counts and comparisons are available for the primary analysis. Although duration of response and progression-free survival time are listed as secondary outcomes, their results are not reported in the registry record. The confidence level and construction method for the posted intervals are also not specified. The record does not report a multiplicity strategy or the original sample-size justification.

Safety summaries provide affected-participant counts and selected serious-event terms, but do not report deaths, exposure duration, timing, or treatment attribution. These omissions limit assessment of the context and completeness of the safety findings. The registry status is completed, but completion status does not itself resolve the missing methodological and outcome details.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Study of Pralatrexate & Gemcitabine With B12 & Folic Acid to Treat Relapsed/Refractory Lymphoproliferative Malignancies — NCT00481871.
• Registry record last updated 2020-01-07; status last verified 2019-12.
• Results first posted 2012-09-18.
• 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.

21. Record Summary

The registry record describes a completed, open-label Phase 1/2a study of pralatrexate and gemcitabine with vitamin B12 and folic acid supplementation for relapsed or refractory lymphoproliferative malignancies. It lists randomized allocation, no masking, a single-group model, and enrollment of 119. The primary endpoint was the number of participants achieving an objective response by International Workshop Criteria.

Primary response counts were 8.0 of 34.0 in Phase 1, 5.0 of 38.0 in Phase 2 - Group B, and 7.0 of 34.0 in Phase 2 - Group C. The posted comparison intervals include the null values, and the unadjusted p-values are 0.253 and 0.77. The record also lists duration of response and progression-free survival time as secondary outcomes, without posted analyses. Design details, missing-data procedures, multiplicity methods, and the original sample-size rationale are not reported in the registry record.