Lamotrigine Bioequivalence: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Lamotrigine Bioequivalence (NCT01995825), a randomized phase 4 study. The primary outcome is AUC.

2. Clinical Question

The study description gives the reason for conducting the comparison: brand and generic lamotrigine had been compared through blood levels in healthy volunteers, but the study sought to compare them in patients with epilepsy. The description also notes concerns held by some physicians and patients about possible differences in efficacy and tolerability. Those concerns motivate the research question; they are not results of this trial.

The posted primary result is a mean AUC of 101300.0 ng*hr/ml for brand lamotrigine and 101400.0 ng*hr/ml for generic lamotrigine. The registry comparison reports a mean difference of 47.0, with an interval from -17600.0 to 17700.0. These values describe the posted pharmacokinetic comparison. They do not, by themselves, establish that the products are clinically interchangeable or resolve questions about seizure control, adverse effects, or individual patients.

3. Trial Design

The registry identifies this as a phase 4, randomized, quadruple-masked crossover trial sponsored by the University of Maryland, Baltimore. The condition listed is epilepsy. The study began in 2012-05, reached primary completion in 2013-10, and is recorded as completed. Enrollment was 35.

The crossover structure is central to interpreting the comparison. The two treatment sequences were brand lamotrigine followed by generic lamotrigine, and generic lamotrigine followed by brand lamotrigine. The arm descriptions state that each sequence received each product for two periods. Thus, the treatment sequences describe the order of exposure; they are not simply two groups in which one group received only brand and the other received only generic.

Randomization refers to allocation by chance, while masking means that treatment identities were concealed from specified parties. The registry lists quadruple masking, but does not identify the parties covered by that label. The allocation method beyond the randomized designation, the sequence-generation procedure, and the practical details of masking are not reported in the registry record.

The registry reports the AUC outcome by treatment, with 34 participants listed for each treatment result. It does not explain how those counts relate to the sequence groups, whether all enrolled participants completed every period, or how any incomplete treatment periods were handled. Those details matter in crossover studies because the treatment comparison can depend on paired observations and on the period and sequence structure. The analysis population and the handling of missing or incomplete measurements are not reported in the registry record.

The study description says blood levels were the primary focus and that adverse effects and seizure control would also be considered secondarily. The posted results include the AUC summary and adverse-event counts. No secondary outcome results are reported in the registry record. The original design assumptions, including any planned sample size, outcome-analysis model, or interim analysis plan, 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 AUC, measured over 0-12hr.

The primary endpoint is AUC, a measure of pharmacokinetic exposure. The registry describes it as exposure over 0-12hr and gives the unit ng*hr/ml. AUC reflects the amount of drug measured across a specified time interval, rather than a single blood-level measurement at one moment. The registry reports the mean for each treatment and gives standard error as the dispersion measure for those means.

Pharmacokinetic exposure is used in this study because the stated primary comparison asks whether brand and generic lamotrigine produce similar blood-level exposure in patients with epilepsy. Comparing exposure is a way to investigate whether the products behave similarly in the body under the study conditions. It is not itself a direct measure of seizure control, tolerability, or a patient’s overall clinical outcome.

The posted mean AUC is 101300.0 ng*hr/ml for brand lamotrigine and 101400.0 ng*hr/ml for generic lamotrigine. The registry’s reported mean difference is 47.0. The estimate concerns the average treatment comparison; it does not show how much individual participants’ measurements varied or whether every participant had similar exposure on both products. The record does not report a participant-level distribution or an analysis of individual differences.

Using AUC as the primary outcome aligns with the study description’s focus on blood levels. The registry record does not report a prespecified bioequivalence margin or a decision rule for declaring equivalence. Consequently, the posted mean comparison and its interval can be described, but should not be treated as a registry-reported formal conclusion of bioequivalence.

7. Results Posted in the Registry

• Generic Lamotrigine: mean 101,393 versus 101,346 on control; difference 47.00 (95% CI -17,603 to 17,697), standardized effect d = 0.00; z-test p = 0.996.

8. Statistical Interpretation

9. Statistical Methods Explained

The registry reports a continuous primary outcome summarized by treatment mean and standard error. The comparison is expressed as a mean difference for generic relative to brand lamotrigine. The reported estimate is 47.0, with bounds of -17600.0 and 17700.0, and the unadjusted p-value is 0.996. The record does not report the statistical model or the precise method used to obtain these comparison statistics.

What does a mean describe?

A mean is a measure of the central location of a set of measurements. It gives a compact description of the average AUC for a treatment, but it does not show the full distribution of measurements. Two treatments can have similar means while individual participants have different values, and a mean alone cannot describe whether the measurements are tightly clustered or widely dispersed.

The treatment means are 101300.0 ng*hr/ml for brand lamotrigine and 101400.0 ng*hr/ml for generic lamotrigine. The posted standard errors are 6314.0 and 6421.0, respectively. A standard error describes the statistical precision of an estimated mean, rather than the spread of individual participants’ measurements. It is affected by both the variability in measurements and the number of observations contributing to the mean. The registry does not provide the participant-level measurements needed to characterize the distribution directly.

What does a mean difference describe?

A mean difference expresses the contrast between two treatment means in the units of the outcome. The posted estimate of 47.0 concerns the generic-versus-brand comparison in AUC units. A difference estimate is a point estimate: it is the single value that summarizes the observed contrast under the analysis used. The interval around it conveys uncertainty about the value that could plausibly be compatible with the data and the method.

For this trial, the posted interval runs from -17600.0 to 17700.0. Because it crosses zero, it includes both a negative and a positive mean difference. A negative value would indicate a lower mean for generic relative to brand under the reported contrast direction; a positive value would indicate a higher mean. The bounds are wide relative to the point estimate, so the estimate is not precise. The confidence level and construction method for the interval are not reported in the registry record.

How should the p-value be read?

The unadjusted p-value is 0.996. In general, a p-value measures how unusual data at least as extreme as the observed data would be if a particular null hypothesis and statistical model were true. Its interpretation depends on the model and analysis choices. It does not give the probability that the null hypothesis is true, nor does it quantify the probability that treatments are equivalent. It also does not replace the estimate and its interval, which are needed to understand the range and precision of the effect estimate.

The record does not identify the null hypothesis in detail, state whether the p-value came from a paired or unpaired analysis, or describe adjustments for the crossover design. It is labeled unadjusted, but the registry does not explain what adjustment, if any, might otherwise have been considered. A p-value by itself cannot establish that the study had adequate precision for an equivalence claim.

Why does the crossover structure matter?

In a crossover design, participants receive more than one treatment over different periods. A within-participant comparison can account for the fact that measurements from the same participant tend to be related. In general, that pairing can make a treatment comparison more precise than treating all observations as though they came from unrelated people, provided the design and analysis assumptions are appropriate.

The registry describes two treatment sequences and two periods on each product within each sequence. The posted AUC summaries do not give paired differences, sequence-specific results, or period-specific values. The analysis method does not appear in the registry record. It is therefore not possible to determine from the posted summaries whether the treatment comparison used the pairing appropriately or how period, sequence, or carryover considerations were handled.

Why can over-dispersion matter?

As a general statistical concept, over-dispersion occurs when observed variability is greater than a model assumes. When that happens, standard errors can be too small if the model does not account for the extra variability, making intervals too narrow and p-values too optimistic. This concept is especially relevant when a model imposes a variance structure on measurements. The registry record does not report a model or evidence of over-dispersion for this trial, so no conclusion about it can be drawn from the posted summaries.

What does the registry leave unspecified?

The record does not report a formal analysis plan, the population used for the primary analysis, handling of missing observations, treatment of incomplete crossover sequences, or an equivalence margin. It also does not report whether the interval and p-value arise from a model that accounts for the repeated measurements. These are important details for evaluating the inferential meaning of a crossover comparison. The posted numerical results can be described as recorded, but the unreported methods should not be filled in by assumption.

For that reason, interpretation should keep three elements distinct: the observed treatment means, the estimated mean difference and its interval, and the design and analysis assumptions needed to support a formal conclusion. The means are close, the estimated contrast is near zero, and the interval is broad. The record does not provide enough analysis detail to turn those summaries into a formal declaration of equivalence.

10. Confidence Intervals and Precision

The mean difference for generic relative to brand lamotrigine is reported as 47.0, with an interval from -17600.0 to 17700.0. An interval is useful because it shows the uncertainty around a point estimate: rather than focusing only on the single estimated value, readers can see a range of values compatible with the data under the method used. The confidence level and the method for constructing this interval are not reported in the registry record.

The interval includes zero and extends in both directions. This means the posted result does not precisely establish whether the mean AUC is higher or lower for generic lamotrigine. Its width also indicates uncertainty about the magnitude of any difference. The point estimate being close to zero does not remove that uncertainty; precision is judged by the interval as well as by the point estimate.

The treatment means are 101300.0 ng*hr/ml for brand and 101400.0 ng*hr/ml for generic. The standard errors are 6314.0 and 6421.0. Standard errors describe uncertainty in estimated means, not the range of values among individuals. In a crossover trial, precision also depends on the within-participant relationship between measurements. The registry does not report the paired differences or their variability, so the displayed group summaries do not fully explain the precision of the treatment contrast.

The interval should not be interpreted as a formal equivalence assessment. The registry record does not report a prespecified acceptable difference or a confidence-interval decision rule for bioequivalence. Without those details, the interval describes uncertainty around the reported comparison but does not determine whether a difference is small enough to meet a formal equivalence criterion.

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 event counts. They are illustrations based on the observed effect rather than the trial’s original design. The 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 trial’s own original sample-size assumptions are not reported in the registry record.

For planning for 100% of the observed effect, the figures are 9,796,918 participants per arm for 80% power and 13,115,293 participants per arm for 90% power. For planning for 75% of the observed effect, the corresponding figures are 17,416,742 and 23,316,076 participants per arm. For planning for 50% of the observed effect, they are 39,187,669 and 52,461,170 participants per arm.

In plain language, 80% power means that, under the stated assumptions and if the specified effect were the true effect, the design would have an 80% chance of detecting it using the stated significance threshold. The 90% column describes the same idea with a 90% chance. These are planning concepts, not guarantees about a particular study. Smaller effects require more participants to detect, which is why the figures increase as the fraction of the observed effect becomes smaller.

These figures are not observed power. They are sensitive to the observed effect used as their basis and are not the trial’s original enrollment target. Because the study is crossover and the illustration specifies equal allocation between arms, the displayed planning assumptions should not be mistaken for a reconstructed crossover-specific design calculation.

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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 47.000; pooled standard deviation σ = 37129.897.
2. n = 2 × (1.96 + 0.842)² × 37129.897² / 47.000² = 9,796,917.2, rounded up to 9,796,918 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
delta <- 47                      # observed difference in means
sigma <- 37129.8967006373                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1995825)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 9,796,918 (80% power), 13,115,293 (90% power)
 75% of observed effect: n per arm = 17,416,742 (80% power), 23,316,076 (90% power)
 50% of observed effect: n per arm = 39,187,669 (80% power), 52,461,170 (90% power)
power.t.test: 9796918 
Simulated power at n = 9,796,918 per arm: 0.802 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-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

Multiplicity arises when several outcomes, treatment contrasts, time points, or subgroup analyses are tested. When many statistical tests are conducted, the chance of at least one apparently notable result can increase unless the analysis accounts for the testing strategy. The registry lists two outcome measures, but the record does not provide the complete outcome definitions or a multiplicity strategy. It reports zero analyses and no secondary results, so the extent of any testing beyond the posted primary comparison is not reported in the registry record.

The brief description identifies seizure control and adverse effects as secondary areas of interest. Their results are not reported in the registry record. It is therefore not possible to assess whether they were analyzed, whether they were prespecified in detail, or whether any adjustment for multiple comparisons was planned or used. The posted unadjusted p-value should not be interpreted as an account of multiplicity handling.

Missing data are also important in a crossover study. The registry lists enrollment of 35 and 34 participants for each treatment’s AUC result, but does not explain the relationship between those counts. It does not state how missing measurements, incomplete periods, or participants who did not complete both treatment sequences were handled. The analysis population—such as whether it included all randomized participants or only participants with usable measurements—is not reported in the registry record.

The record also does not describe how the analysis handled the repeated measurements expected in a crossover comparison. It does not specify whether treatment, period, sequence, or participant effects were included, or whether any carryover assessment was performed. These details are relevant to interpreting both the estimate and its uncertainty.

None of these omissions establishes that a particular analysis was or was not performed. They indicate that the registry record does not document the information needed to evaluate the analysis plan, missing-data decisions, or multiplicity control. The posted AUC comparison and adverse-event summaries remain descriptive evidence, but the inferential context is incomplete.

14. Safety Results

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

Deaths

Other (non-serious) adverse events

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

Most frequent non-serious adverse event terms

The registry reports serious adverse events for 35 participants at risk in each treatment group. It records zero serious adverse events for brand lamotrigine and zero for generic lamotrigine. It also reports zero deaths in each group. These are the counts recorded in the serious-event summary; they do not establish that no other kinds of adverse events occurred.

For the other adverse-event category, 35 participants were listed at risk in each group. The reported counts were 7 for brand lamotrigine and 6 for generic lamotrigine. The registry record does not provide event terms, severity details, exposure time, or a statistical comparison for these counts. The counts should therefore be read as reported events in the posted categories, not as a complete account of all possible harms or a formal test of a treatment difference.

In a crossover study, participants can receive both treatments, so treatment-specific adverse-event counts are not necessarily counts from independent groups of different people. The record does not explain whether an individual could contribute events to both treatment categories or how events were assigned to periods. The available safety summary is limited to these aggregate counts and does not report a detailed safety analysis.

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

The posted primary outcome is mean AUC over 0-12hr, a measure of drug exposure. The reported means—101300.0 ng*hr/ml for brand lamotrigine and 101400.0 ng*hr/ml for generic lamotrigine—are close. The reported mean difference is 47.0, and the interval from -17600.0 to 17700.0 spans zero and a wide range of possible differences. The result therefore describes a small point estimate alongside substantial uncertainty about the treatment contrast.

These findings do not establish that the treatments are equivalent. A formal bioequivalence conclusion depends on criteria such as a prespecified acceptable range and a suitable analysis, neither of which is reported in the registry record. The high unadjusted p-value does not prove equality, and the close means do not show that every participant had similar measurements on each treatment.

The endpoint is pharmacokinetic rather than a direct clinical outcome. The AUC comparison does not by itself answer whether seizure control differed, whether adverse effects differed in ways not captured by the posted counts, or whether either product is appropriate for an individual. The brief description says that seizure control and adverse effects were secondary interests, but results for those outcomes are not reported in the registry record.

The crossover design can support within-participant comparisons, but the aggregate record does not show paired results or explain the analysis method. The treatment-specific counts are 34 for AUC, while enrollment was 35; the relationship between those figures and any missing or incomplete observations is not reported. These limitations leave the precision and analysis details incompletely documented.

The most defensible reading is that the posted treatment means are similar in magnitude, while the reported interval remains broad. The findings are informative as a registry-reported AUC comparison in patients with epilepsy, but they do not settle questions that require a documented equivalence analysis, detailed crossover modeling, or secondary clinical outcome results.

16. Why This Trial Matters Statistically

This trial raises a useful statistical distinction between a small estimated difference and strong evidence that any difference is small. The reported AUC means are close and the mean-difference estimate is 47.0, but the interval from -17600.0 to 17700.0 is broad. A near-zero point estimate can be compatible with substantial uncertainty, so precision and the study’s decision criteria matter as much as the apparent similarity of the means.

The crossover design is also statistically important. Since each sequence received both brand and generic lamotrigine over multiple periods, an analysis can use within-participant information. That structure calls for methods that reflect repeated measurements and account for relevant design features. The registry record does not report the model, paired treatment differences, or handling of period and sequence, leaving readers unable to evaluate these details from the summary results.

The study description distinguishes pharmacokinetic exposure from secondary questions about seizure control and adverse effects. That distinction matters because an AUC result cannot stand in for every clinical outcome. The posted results include a primary exposure summary and limited adverse-event counts, while secondary efficacy results are not reported in the registry record.

Finally, the sample-size illustration emphasizes how planning depends on the effect selected for detection. Its very large figures are derived from the observed effect under stated assumptions, not from the trial’s original design plan. They should be read as an illustration of the consequences of using that estimated effect, not as a judgment about observed power or as evidence about the trial’s clinical conclusions.

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 posted record leaves several trial-specific details unreported. Although it identifies a randomized, quadruple-masked crossover design, it does not state which parties were masked, how the randomization sequence was generated, or how treatment allocation was concealed. The sequence descriptions establish the order and number of treatment periods, but do not provide period dates, washout details, or other operational information.

The primary AUC result is reported as treatment means, standard errors, and a mean-difference interval. The registry does not report the analysis model, confidence level for the interval, paired treatment differences, or the handling of period, sequence, and carryover effects. It also does not report a prespecified bioequivalence margin or decision rule. These omissions limit assessment of whether the statistical analysis matches the crossover design and supports a formal equivalence interpretation.

Enrollment was 35, while 34 participants are listed for each treatment’s AUC result. The record does not explain participant flow, whether the same participants contributed to both treatment summaries, or how missing and incomplete observations were handled. The analysis population and reasons for any absent measurements are not reported in the registry record.

The study description mentions seizure control and adverse effects as secondary interests, but secondary outcome results are not reported. The safety summary gives aggregate counts without event terms, exposure duration, or a treatment-comparison analysis. The registry also does not report the trial’s original sample-size justification, interim analysis plan, or detailed analysis plan. These limitations concern what is documented in the record; they do not establish what procedures may have been used outside it.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Lamotrigine Bioequivalence — NCT01995825.
• Registry record last updated 2020-12-01; status last verified 2020-11.
• Results first posted 2020-12-01.
• 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

NCT01995825 is a completed phase 4 randomized, quadruple-masked crossover study sponsored by the University of Maryland, Baltimore. It enrolled 35 patients with epilepsy and compared brand with generic lamotrigine through two treatment sequences, each involving two periods on each product. The study began in 2012-05 and reached primary completion in 2013-10.

The primary outcome was mean AUC over 0-12hr, reported in ng*hr/ml. The posted means were 101300.0 for brand and 101400.0 for generic; the mean difference was 47.0, with an interval from -17600.0 to 17700.0 and an unadjusted p-value of 0.996. The interval is broad, and the registry does not report a formal equivalence margin or detailed crossover analysis method.

The safety summary reports zero serious adverse events and zero deaths in each treatment group, with other adverse-event counts of 7 for brand and 6 for generic. Secondary results for seizure control are not reported in the registry record.