A Prospective Study of Cyclophosphamide in Systemic Lupus Erythematosus Treatment: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Prospective Study of Cyclophosphamide in Systemic Lupus Erythematosus Treatment (NCT01689350), a randomized na study. The primary outcome is Adverse Reaction (Leucopenia).

2. Clinical Question

3. Trial Design

The registry describes a completed, randomized, parallel-group study with no masking. It lists an enrollment of 92 participants with systemic lupus erythematosus. The study start is recorded as 2012-10 and primary completion as 2013-12. Sun Yat-sen University is listed as the sponsor. The phase is recorded as NA.

Participants were assigned to genotype-based cyclophosphamide dosing or traditional dosing. The experimental group included 47 participants and the control group included 45. In the experimental group, genotype detection was used to classify participants as extensive, intermediate or poor metabolisers, with initial weekly doses of 0.2g, 0.4g and 0.6g, respectively. The control group received an initial weekly dose of 0.2-0.6g, selected according to clinical experience. Both descriptions specify injection.

Randomized allocation can support a comparison between assigned groups, but the registry record does not describe how the random sequence was generated, whether allocation was concealed, or how departures from assigned dosing were handled. Masking is listed as none; additional detail on who knew the assignments is not reported in the registry record. The record also does not report the original sample-size justification or planned allocation ratio. The observed group sizes are known, but they do not reveal the details of the trial’s original design assumptions.

The primary outcome was adverse reaction (leucopenia), measured as the number of participants meeting the white blood cell threshold within one month. Infection is listed as a secondary outcome. The registry lists two outcome measures and two analyses. Further information on the assessment schedule for the secondary outcome, analysis populations, and the statistical plan is 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

5. Baseline Characteristics

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

Population: SLE patients who met the ACR criteria and applied CPA medication were included.

6. Primary Endpoint

The registry lists the primary outcome as Adverse Reaction (Leucopenia), measured over one month.

The primary endpoint counts participants with leucopenia during one month after receiving cyclophosphamide. The registry defines an event as a white blood cell count < 4.0 × 10^9/L and describes it as cyclophosphamide-induced leucopenia. The posted results give 6 events among 47 participants in the experimental group and 19 among 45 participants in the control group.

A binary endpoint reduces each participant’s result to whether the specified event occurred during the assessment period. Reporting the count alongside the group size makes it possible to interpret the event frequency within each group, while a comparison of frequencies describes the difference between groups. The endpoint is relevant to this study because it directly records a blood-cell adverse reaction considered in the registry’s description of cyclophosphamide treatment.

That definition also limits what the endpoint can establish. It captures the specified threshold during one month, not the duration, severity, timing, reversibility or clinical consequences of an event. The registry record does not report repeated measurements, a grading scale, or how participants with missing blood-count data were handled. The endpoint should therefore be read as a count of participants meeting the recorded definition, rather than a complete summary of all safety outcomes.

7. Results Posted in the Registry

• Experimental Group: 12.8% (6/47) versus 42.2% (19/45) on control; risk difference -29.5 percentage points (95% CI -46.8 to -12.2), risk ratio 0.30 (0.13–0.69); unadjusted two-proportion z-test p = 0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary: each participant either met the recorded leucopenia definition during one month or did not. For this kind of outcome, the group counts and denominators describe how many participants experienced the event and how many were assessed. The posted counts are 6 of 47 in the experimental group and 19 of 45 in the control group. The reported proportions are 0.128 and 0.422.

What does the risk difference describe?

The risk difference subtracts the control-group event proportion from the experimental-group event proportion. The estimate is -0.295, with an interval from -0.468 to -0.122. A negative value indicates that the observed event proportion was lower in the experimental group under this ordering. The risk difference is an absolute-scale contrast: it describes the gap between the two observed proportions, rather than a proportional change. It is tied to the particular outcome definition and one-month timeframe used in the registry record.

How do relative measures differ?

The risk ratio compares the experimental event proportion with the control event proportion. Its estimate is 0.302, with an interval from 0.133 to 0.688. The odds ratio compares event odds rather than event proportions; its estimate is 0.200, with an interval from 0.071 to 0.567. Both relative measures use one as the value corresponding to no group difference. Their interpretation depends on the order of the groups: reversing the comparison changes which group is in the numerator and reverses the direction of the contrast.

Odds and risks are distinct quantities. For outcomes that are not rare, an odds ratio can differ appreciably from a risk ratio, so it should not be read as though it were a direct ratio of event proportions. For clinical interpretation, the group counts and the absolute contrast help retain the scale of the observed outcome, while a relative measure provides a separate way to express the comparison.

What does the p-value address?

The unadjusted p-value for the re-calculated comparison is 0.001. A p-value describes how compatible the observed data are with a null model of no group difference, under the assumptions of the test. It is not the probability that the null model is correct, and it does not measure the probability that the result will replicate. The sponsor’s chi-squared analysis reports p < 0.01. A p-value does not communicate the magnitude of the contrast; the effect estimates and their intervals provide that complementary information.

Why do confidence intervals matter?

A confidence interval presents a range of estimates compatible with the data under the method’s assumptions. The risk-difference interval ranges from -0.468 to -0.122. The intervals for the risk ratio and odds ratio remain on one side of their no-difference value of one. This is consistent with the unadjusted p-value, but the interval also conveys uncertainty about the magnitude. It does not show all sources of uncertainty, such as potential bias from missing data, imperfect outcome ascertainment, or limitations in the trial design.

What assumptions remain unknown?

The comparison is an unadjusted re-calculation from the posted group counts and summary statistics. The registry record identifies the trial as randomized, but does not report the randomization procedure, allocation concealment, or whether the analysis followed intention-to-treat principles. It does not report a detailed statistical analysis plan, adjustment variables, methods for missing observations, or whether any interim analyses were planned or conducted. These omissions make it difficult to evaluate whether the posted comparison fully matches the intended analysis.

General statistical considerations also apply to binary outcomes. Small numbers of events can make estimates sensitive to a few participants, and the precision depends on both event counts and group sizes. Randomization is intended to balance prognostic factors on average, but it does not guarantee balance in a particular study. The registry record does not report baseline characteristics needed to assess that balance. Thus the methods describe the posted comparison, while the limits of the available record constrain how fully its assumptions can be checked.

10. Confidence Intervals and Precision

The risk-difference estimate is -0.295, with an interval from -0.468 to -0.122. The relative estimates are 0.302 for the risk ratio, with an interval from 0.133 to 0.688, and 0.200 for the odds ratio, with an interval from 0.071 to 0.567. These intervals indicate a range of effect sizes compatible with the observed data under the calculation methods; they do not establish that the true effect lies within a particular range with certainty.

The risk-difference interval does not include zero, the value representing no absolute difference. The relative-measure intervals do not include one, the value representing no relative difference. Their placement is consistent with the unadjusted p-value of 0.001. The interval widths still matter: the compatible values span a range of possible magnitudes, so the point estimates should not be treated as exact descriptions of the underlying effect.

The sponsor’s chi-squared analysis reports an odds ratio of 4.99 with a 95% confidence interval from 1.76 to 14.14. That result uses the registry’s reported group ordering. The re-calculated odds ratio uses the experimental group relative to the control group, so its direction differs. The sponsor’s interval is also fairly broad, indicating uncertainty about the magnitude on that reported scale.

Confidence intervals quantify sampling uncertainty under their statistical assumptions. They do not account for every limitation, including unreported missing-data handling, outcome measurement, or possible analysis choices. The registry record does not report a detailed account of those issues. Precision should therefore be considered alongside the design and reporting limitations, rather than as a complete measure of certainty.

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, not total events. They are illustrations based on the observed effect, rather than the trial’s original design assumptions. The stated assumptions are a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect to detect set as a stated fraction of the observed effect. The original sample-size justification and design assumptions are not reported in the registry record.

For planning at 100% of the observed effect, the figures are 35 participants per arm for 80% power and 47 participants per arm for 90% power. At 75% of the observed effect, the corresponding figures are 68 and 91 participants per arm. At 50% of the observed effect, they are 164 and 218 participants per arm. These are planning illustrations, not counts of participants observed in this study.

In general, a power target describes the chance of detecting a specified effect if that effect is present and the planning assumptions hold. The 80% column gives the illustrated participants per arm for an 80% power target; the 90% column gives the number for a 90% target. A higher target generally requires more participants for the same effect. Smaller effects also require more participants because they are harder to distinguish from random variation. The figures are not observed power, and they should not be read as evidence that the completed trial had a particular probability of detecting an effect.

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 = 0.422; treatment proportion p₁ = 0.128; average p̄ = 0.275.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6314 = 1.2376.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.5961 = 0.5017.
4. n = (1.2376 + 0.5017)² / (-0.2946)² = 34.9, rounded up to 35 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 <- 0.422222222222222                          # control proportion
p1 <- 0.127659574468085                          # 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(1689350)
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 = 35 (80% power), 47 (90% power)
 75% of observed effect: n per arm = 68 (80% power), 91 (90% power)
 50% of observed effect: n per arm = 164 (80% power), 218 (90% power)
power.prop.test: 34.9 
Simulated power at n = 35 per arm: 0.812 (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. Analyses Reported by the Sponsor

The registry reports a sponsor analysis using a chi-squared method. It gives an odds ratio of 4.99, a 95% confidence interval from 1.76 to 14.14, and p < 0.01. The reported group order is experimental group followed by control group. The posted counts are 6 of 47 participants with leucopenia in the experimental group and 19 of 45 in the control group.

The re-calculated comparison expresses the experimental group relative to the control group. Its risk difference is -0.295, risk ratio is 0.302 and odds ratio is 0.200. The apparent difference in direction between the sponsor’s odds ratio and the re-calculated odds ratio is consistent with the contrast being expressed in the opposite group order. The group order must be considered before interpreting whether an odds ratio is above or below one.

The re-calculated p-value is 0.001, while the sponsor reports p < 0.01. Both indicate a result below the reported threshold, but the registry does not provide enough detail to compare every aspect of the underlying analysis. The discrepancy flag is false, so the posted sponsor analysis is not identified as discrepant from the re-calculation. The re-calculated estimates remain unadjusted summaries from the posted counts; they do not replace the sponsor’s reported analysis or supply missing details about the original statistical plan.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists two outcome measures and two analyses. The primary outcome is leucopenia; infection is listed as a secondary outcome. The record does not report whether the analyses were prespecified in a detailed statistical plan, how multiple outcomes or analyses were handled, or whether any multiplicity adjustment was planned. Without that information, the chance of at least one nominally unusual result across several tests cannot be assessed from the posted record alone. The reported p-values should therefore be read as the results of their respective analyses, not as evidence that multiplicity was controlled.

The primary endpoint is described as a count of participants meeting a white blood cell threshold within one month. The registry record does not state how missing blood-count results were handled, whether all randomized participants were included in the analysis, or whether analyses were conducted according to assigned group. It also does not describe exclusions after randomization, treatment discontinuation, or protocol deviations. These details matter because the analyzed participants may not represent everyone randomized if outcome data or follow-up are incomplete.

Randomization is listed, but the record does not report the sequence-generation method or allocation concealment. Baseline characteristics and any adjusted analyses are also not included in the information available for this comparison. It is therefore not possible to evaluate balance between groups or determine whether the sponsor’s analysis adjusted for other variables.

The registry does not report interim analyses, stopping rules, or the original statistical analysis plan. It also does not report imputation methods or sensitivity analyses for missing outcomes. These are unknowns, not evidence that such procedures were or were not used. The results can be interpreted as posted, while the absence of these details limits assessment of analysis choices and robustness.

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 3% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: leucopenia (42.2%).

The adverse-event record lists serious-event figures of 6 participants among 47 in the experimental group and 19 among 45 in the control group. It also lists other adverse-event figures of 8 among 47 and 16 among 45, respectively. The registry identifies leucopenia as the leading serious-event term, with 42.2% reported. These entries should be read as counts and proportions associated with the registry’s adverse-event categories, not as a complete account of every safety outcome.

The primary outcome counts also record 6 participants with leucopenia in the experimental group and 19 in the control group, but the endpoint and adverse-event listings are distinct registry entries. The record does not provide enough detail to determine how adverse events were classified, whether event counts represent unique participants in every category, or how severity and relatedness were assessed. Death counts are not reported in the registry record.

Comparing event counts without considering the group sizes can be misleading, so each count is presented with its at-risk denominator. These safety figures do not describe the timing, duration, seriousness criteria, or outcomes of individual events. The registry record does not report a full narrative safety assessment or details of how adverse events were monitored.

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

The posted results show a lower observed frequency of the defined one-month leucopenia endpoint in the genotype-based dosing group than in the traditional-dosing group: 6 of 47 versus 19 of 45. The unadjusted risk difference is -0.295, and its interval ranges from -0.468 to -0.122. The unadjusted risk ratio and odds ratio also indicate a contrast in the same direction when the experimental group is compared with the control group. The sponsor’s reported odds ratio uses the opposite group ordering.

These are findings from this randomized study as recorded in the registry. They do not, on their own, establish that genotype-based dosing caused the observed difference. The registry record does not report enough methodological detail to assess randomization procedures, missing outcome data, analysis populations, or all possible sources of bias. The record also does not report baseline characteristics that would help describe the groups.

The endpoint is limited to a white blood cell count < 4.0 × 10^9/L within one month. It does not summarize other possible adverse reactions, the duration or severity of leucopenia, or longer-term outcomes. Infection is listed as a secondary outcome, but its results are not included in the posted details used for this comparison. The safety listing gives additional event counts, though its definitions and assessment procedures are not fully described.

Statistical evidence of a group difference is not the same as a complete account of clinical importance. The estimates and intervals describe the recorded endpoint and the uncertainty associated with this comparison. They do not support conclusions about every participant, every outcome, or circumstances not represented by the study record.

17. Why This Trial Matters Statistically

This trial raises a statistical question about individualizing a dose using genotype information and comparing that approach with a traditional dosing approach. The study design places two dosing strategies in parallel randomized groups, and the primary outcome is a binary adverse reaction measured during a defined period. That structure makes the contrast interpretable through event counts, group proportions, and absolute and relative effect measures.

Several statistical perspectives are useful together. The counts and proportions show the observed event frequency in each group. The risk difference expresses the absolute contrast, while the risk ratio and odds ratio provide relative comparisons. Confidence intervals communicate uncertainty about the estimated magnitude, and the p-value addresses compatibility with a null comparison under assumptions. No single statistic answers every question, so the estimand, group ordering, outcome definition, and timeframe all matter.

The record also illustrates why reporting detail is essential. Randomized allocation is noted, but the registry does not describe sequence generation, concealment, the original sample-size justification, missing-data handling, or the analysis population. Those omissions limit independent evaluation of the methods and robustness of the result. The sponsor’s odds ratio also demonstrates why group ordering must accompany interpretation of relative measures: reversing the contrast reverses the direction.

Finally, planning calculations based on an observed effect are not a substitute for prospective design assumptions. Such illustrations help show how sample requirements change as the target effect changes, but they do not explain what assumptions guided the completed trial. Clear reporting of both design and analysis decisions is important for separating the observed comparison from the broader conclusions readers may wish to draw.

18. Limitations

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

The registry record gives limited detail about how randomization was implemented. Although allocation is listed as randomized and masking as none, sequence generation, allocation concealment, and the people aware of assignment are not described. The record also does not report baseline characteristics, making it difficult to assess the groups’ comparability in the completed study.

The primary outcome has a clear threshold and timeframe, but there is no reported information on the frequency of blood-count measurement, how missing measurements were handled, or whether assessment was consistent between groups. The analysis population and any post-randomization exclusions are not reported. These gaps limit interpretation of the counts as a complete account of all randomized participants.

The registry does not report a detailed statistical analysis plan, original sample-size rationale, interim analyses, or multiplicity procedures. It lists infection as a secondary outcome, but results for that outcome are not included in the posted details. The safety entries also lack detail on adverse-event definitions, ascertainment, severity grading, and event duration. Deaths are not reported in the registry record.

The intervention description specifies genotype categories and starting doses, but does not report the classification criteria, implementation details, or subsequent dose changes. The control dose range is described as selected according to clinical experience, without additional decision rules. Consequently, the comparison cannot be fully characterized in terms of adherence or exposure over time. These are limitations of the reporting available in the registry record and should not be mistaken for proof that procedures were absent.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: A Prospective Study of Cyclophosphamide in Systemic Lupus Erythematosus Treatment — NCT01689350.
• Registry record last updated 2014-09-08; status last verified 2014-08.
• Results first posted 2014-09-08.
• 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

NCT01689350 is a completed randomized parallel-group study of genotype-based versus traditional cyclophosphamide dosing in systemic lupus erythematosus. It enrolled 92 participants and lists no masking. The experimental group used genotype detection to assign initial weekly doses by metaboliser category; the control group received traditional dosing selected according to clinical experience.

The primary outcome was the number of participants with leucopenia, defined as a white blood cell count < 4.0 × 10^9/L within one month. The posted counts were 6 of 47 in the experimental group and 19 of 45 in the control group. The re-calculated unadjusted risk difference was -0.295, with an interval from -0.468 to -0.122; the p-value was 0.001. The sponsor reported a chi-squared odds ratio of 4.99, with a 95% confidence interval from 1.76 to 14.14 and p < 0.01, using the listed group order.

Infection is listed as a secondary outcome. Further details on the statistical plan, missing data, and original sample-size assumptions are not reported in the registry record.