Therapy in Amyotrophic Lateral Sclerosis (TAME): Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Therapy in Amyotrophic Lateral Sclerosis (TAME) (NCT02118727), a randomized phase 2 study. The primary outcome is The Primary Comparison for Efficacy Will be Based on a Linear Mixed Effects (LME) Model Fit to the ALSFRS-R Data for the Patients Followed Over 36 Weeks..

2. Clinical Question

This is a question about the difference in average progression rates over the stated treatment period, not a comparison of a single score at one time. A difference near zero in the posted point estimates indicates little separation between the reported group averages. The interval around the unadjusted estimate is important too: because it includes values on either side of zero, the summary does not pin down the direction of a possible difference with precision. The registry identifies a linear mixed effects model as the intended basis for the primary comparison, but a model-based analysis result is not reported in the registry record.

3. Trial Design

This was a randomized, quadruple-masked, parallel-group phase 2 study of memantine for ALS. The registry lists an enrollment of 89 and identifies the University of Kansas Medical Center as the sponsor. The study is marked completed. Its recorded start date is 2018-11-07, and its primary completion date is 2021-07-22.

Participants were allocated to memantine or placebo. The memantine arm received up to 20 mg by mouth twice a day for 36 weeks; the placebo arm received up to 2 tablets by mouth twice a day for the same period. The registry describes the study purpose as determining whether memantine, when used in conjunction with riluzole, could slow disease progression and potentially improve neuropsychiatric changes. It also describes an aim of assessing serum biomarkers for diagnostic and prognostic use. Results for those biomarker aims are not reported in the registry record.

The primary outcome was the rate of change in ALSFRS-R scores during 36 weeks of therapy. The record says that the primary comparison for efficacy would be based on a linear mixed effects model fit to data from participants followed over that period. The displayed primary outcome summaries include 54 memantine participants and 29 placebo participants with more than one score. The registry separately lists 58 participants at risk in the memantine safety group and 31 in the placebo safety group; the reason the outcome and safety counts differ is not reported in the registry record.

Quadruple masking is recorded, but which groups or roles were masked is not reported in the registry record. The original sample-size rationale, allocation ratio planned by the investigators, and other design assumptions are not reported in the registry record. The planning illustrations associated with the posted effect are separate from the trial’s original design assumptions.

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

Analysis population for the primary outcome, as defined in the registry: Includes the 54/58 participants randomized to memantine and 29/31 participants randomized to placebo with more than one ALSFRS-R value.

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 The Primary Comparison for Efficacy Will be Based on a Linear Mixed Effects (LME) Model Fit to the ALSFRS-R Data for the Patients Followed Over 36 Weeks., measured over During 36 weeks of therapy.

The primary endpoint is the rate of change in the Amyotrophic Lateral Sclerosis Functional Rating Scale–Revised (ALSFRS-R) during 36 weeks of therapy. The scale consists of 12 questions about participants’ capability and independence in daily activities. Scores range from 0 to 48, with higher scores indicating better outcome. A rate expressed as score on a scale/week describes the average direction and pace of score change over time.

Using a rate rather than a single end-of-treatment score makes the question about how scores evolve during follow-up. The registry describes comparing the rate during active therapy with the rate in the placebo arm. It identifies a linear mixed effects model as the intended method for this repeated-measure outcome. The exact model specification and handling of missing measurements are not reported in the registry record.

The posted arm summaries are -0.29 score on a scale/week for memantine and -0.281 score on a scale/week for placebo. Both are negative, consistent with declining average scores over the reported interval. The unadjusted difference between the reported rates is -0.009 score on a scale/week. These summaries describe group averages; they do not show that every participant followed the average rate or that an individual participant’s course can be inferred from it.

ALSFRS-R is used because it captures functional activities relevant to ALS progression. A change-rate comparison can summarize whether average functional scores changed at different rates between randomized groups. The posted result should be interpreted alongside its uncertainty and the fact that the registry does not report a fitted primary model result.

7. Results Posted in the Registry

• Memantine: mean -0.29 versus -0.28 on control; difference -0.01 (95% CI -1.00 to 0.98), standardized effect d = -0.00; z-test p = 0.986.

8. Statistical Interpretation

9. Statistical Methods Explained

What does the primary analysis aim to estimate?

The registry describes the primary comparison as a linear mixed effects (LME) model fit to ALSFRS-R data collected during 36 weeks. The target is a contrast between the average rates of score change in the memantine and placebo arms. The posted arm estimates are -0.29 and -0.281 score on a scale/week, respectively, and the unadjusted estimated difference is -0.009 score on a scale/week.

General explanation: An LME model is commonly used for repeated measurements because it can represent change over time while accounting for the fact that scores from the same participant are related. A model can also use information from participants who have different numbers or timings of measurements, depending on its specification and assumptions. That general flexibility does not establish how this trial’s actual model was specified or how missing observations were handled. Those details are not reported in the registry record.

The registry’s outcome description gives the analysis population as participants with more than one ALSFRS-R value: 54/58 randomized to memantine and 29/31 randomized to placebo. The record does not state why other randomized participants did not meet that criterion. The posted summary comparison is unadjusted, rather than a reported fitted LME result. The registry record lists zero analysis entries, so model coefficients, model-based uncertainty estimates, and model diagnostics are not reported in the registry record.

What does an unadjusted comparison mean?

The recalculated comparison uses the posted arm summaries without adjustment for other variables or a repeated-measure model. The reported difference is calculated as the memantine mean rate minus the placebo mean rate. Its estimate is -0.009 score on a scale/week, and its unadjusted p-value is 0.986. The recalculation is useful for describing the posted summaries, but it should not be mistaken for the trial’s intended LME analysis.

General explanation: An unadjusted comparison focuses on the two group summaries being compared. An adjusted or model-based comparison can account for additional structure, such as repeated observations or prespecified covariates, when the analysis plan includes those features. Different approaches may produce different estimates or standard errors. The model specification and any covariate adjustment used in the trial are not reported in the registry record.

What do the p-value and confidence interval say?

The p-value of 0.986 is for the unadjusted comparison. In general, a p-value describes how compatible the observed data are with a specified null hypothesis under the assumptions of the test. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. A large p-value does not demonstrate equivalence or prove that there is no difference.

The unadjusted confidence interval crosses zero. General explanation: A confidence interval gives a range of estimates compatible with the data under the method’s assumptions and reflects statistical uncertainty. When an interval for a difference crosses zero, it includes no difference as well as differences in both directions. The interval in this comparison is broad relative to its near-zero point estimate, so the estimate is not precise enough to isolate a narrow range of plausible differences.

Why does variability matter?

The posted arm summaries include standard errors of 0.39 for memantine and 0.32 for placebo. A standard error describes uncertainty in an estimated mean, rather than the spread of individual participants’ measurements. The comparison also reports a pooled standard deviation of 2.53 for its recalculation. No participant-level outcome distributions are available in the registry record.

General explanation: More variability in outcomes, fewer usable observations, or both can make an estimate less precise. Repeated observations from the same person are correlated, so treating every observation as independent can misstate uncertainty. An LME model is one general approach to representing this correlation; the actual correlation structure used in this trial is not reported in the registry record.

What do the planning illustrations represent?

The planning figures are illustrations based on fractions of the observed effect. They assume 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. These are planning calculations, not the original design rationale. The trial’s own sample-size assumptions and justification are not reported in the registry record. The figures are not observed power and should not be interpreted as an assessment of how likely the completed trial was to find a result.

10. Confidence Intervals and Precision

The unadjusted estimated difference in mean ALSFRS-R rate is -0.009 score on a scale/week. Its confidence interval extends from negative to positive values and therefore crosses zero. The endpoints are not repeated because the key interpretive point is that the interval includes both no difference and estimates in either direction. The p-value for this unadjusted comparison is 0.986.

General explanation: A confidence interval conveys uncertainty about an estimated difference. A narrow interval would indicate that the estimate is relatively precise under the analysis assumptions; a broad interval leaves more possible values compatible with the data. An interval that crosses zero does not establish that there is no difference. It indicates that no difference remains among the values compatible with the interval, while nonzero differences also remain possible.

In this comparison, the point estimate is very close to zero, but the interval is broad enough to include negative and positive values. Thus, the posted summaries do not precisely determine whether the underlying average rate is lower or higher in either arm, or how large a difference might be. The result is not an equivalence finding: the registry record does not report an equivalence margin or an equivalence analysis.

The confidence interval discussed is for an unadjusted recalculation from posted summary statistics. The registry describes an LME model as the planned basis for the primary comparison, but a model-based estimate and interval are not reported in the registry record. Standard errors for the arm summaries are 0.39 and 0.32; those are uncertainties around the reported arm means, not the participant-level spread. Interpretation of precision should also account for the primary outcome population, which includes participants with more than one score, and the lack of reported detail about missing measurements and the fitted model.

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 sample-size plan used to design the completed trial. The registry’s original design assumptions and sample-size justification are not reported in the registry record.

The calculations assume a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set at a stated fraction of the observed effect. For planning for 100% of the observed effect, the figures are 1,240,458 participants per arm for 80% power and 1,660,622 participants per arm for 90% power. For planning for 75% of the observed effect, they are 2,205,259 participants per arm for 80% power and 2,952,216 participants per arm for 90% power. For planning for 50% of the observed effect, they are 4,961,832 participants per arm for 80% power and 6,642,486 participants per arm for 90% power.

In plain language, 80% power is the probability of detecting an effect of the specified size at the stated significance level if that effect is present and the assumptions hold. The 90% column uses a higher target probability of detection under those same assumptions. Smaller effects require more participants to detect with the same power, which is why the illustrated requirements increase as the target fraction of the observed effect decreases.

These figures are not observed power. They are planning illustrations based on the observed effect, not a description of the trial’s original design or a measure of what the completed trial achieved. Because they depend on the observed effect, they should not be treated as a reliable stand-alone guide to future study size.

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 = -0.009; pooled standard deviation σ = 2.530.
2. n = 2 × (1.96 + 0.842)² × 2.530² / -0.009² = 1,240,458.0, rounded up to 1,240,458 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 <- -0.00899999999999995                      # observed difference in means
sigma <- 2.52996608575327                     # 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(2118727)
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 = 1,240,458 (80% power), 1,660,622 (90% power)
 75% of observed effect: n per arm = 2,205,259 (80% power), 2,952,216 (90% power)
 50% of observed effect: n per arm = 4,961,832 (80% power), 6,642,486 (90% power)
power.t.test: 1240459 
Simulated power at n = 1,240,458 per arm: 0.801 (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

The primary outcome is the ALSFRS-R rate of change during 36 weeks, and the registry identifies an LME model as the intended basis for its primary comparison. No analysis entries are listed, and no secondary outcome results are posted in the registry record. The registry lists three outcome measures, but the registry entries do not identify results for secondary measures.

Multiplicity concerns arise when an analysis tests several outcomes, time points, treatment contrasts, or subgroups. General explanation: The more statistical tests considered, the greater the chance that at least one result appears notable by chance if each is interpreted separately. A prespecified testing sequence or adjustment can address this issue, but the registry’s multiplicity plan is not reported in the registry record. It is therefore not possible to determine from the posted summary whether or how multiplicity was addressed.

The primary outcome description includes participants who had more than one ALSFRS-R value: 54/58 randomized to memantine and 29/31 randomized to placebo. This is the stated population for the posted primary measure. The registry does not explain why the remaining randomized participants are not included in that description, or whether participants were excluded for particular clinical or data-related reasons. The effect of missing measurements on the estimate is not reported in the registry record.

General explanation: Missing observations can affect a longitudinal comparison if the likelihood of having a measurement is related to a participant’s outcome or other characteristics. Methods such as mixed models rely on assumptions about the missing-data process; other approaches may be used in different settings. Without the trial’s missing-data assumptions, imputation methods, sensitivity analyses, or participant-flow details, the robustness of the reported outcome summary to missingness cannot be assessed.

The registry does not report a full analysis population definition beyond the requirement for more than one value, the fitted model details, or a sensitivity analysis. These omissions limit assessment of how the posted summary relates to the intended primary analysis and how representative the analyzed participants are of all randomized participants.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: lung infection (5.2%), colitis (3.2%), hematoma (3.2%), aspiration (3.2%), thromboembolic event (3.2%).

The safety figures list 58 participants at risk in the memantine group and 31 in the placebo group. Serious adverse events affected 8 participants in the memantine group and 3 in the placebo group. The registry records 3 deaths in each group. These are counts, not a comparison adjusted for exposure time or other factors.

The registry also lists serious adverse event terms with percentages: lung infection, 5.2%; colitis, 3.2%; hematoma, 3.2%; aspiration, 3.2%; thromboembolic event, 3.2%; and flu like symptoms, 1.7%. The record does not specify an arm for these listed terms, so they should not be attributed to either treatment group from this information alone. The denominator used for each term is not reported in the registry record.

Counts of participants affected by serious events are not the same as counts of event episodes; the posted summary does not state whether participants experienced more than one event. It also does not provide timing, duration, severity details beyond the serious classification, or an exposure-adjusted comparison. The safety counts therefore describe what the registry reports, but do not by themselves establish a difference in safety between groups or a causal relationship between an intervention and an event.

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

The posted primary outcome summaries show similar average rates of ALSFRS-R score change over 36 weeks: -0.29 score on a scale/week in the memantine group and -0.281 in the placebo group. The unadjusted estimated difference is -0.009 score on a scale/week, and its p-value is 0.986. The confidence interval crosses zero and includes estimates in both directions. Taken together, the summaries do not precisely identify a difference in the average rates.

This result does not show that the groups are equivalent. Demonstrating equivalence requires an agreed margin defining what difference would be small enough to count as equivalent, together with a suitable analysis. Such a margin and analysis are not reported in the registry record. Nor does a large p-value prove that the true difference is zero. It indicates that the unadjusted comparison does not provide evidence against its null hypothesis under that test; it is not a probability that either treatment has or lacks an effect.

The registry names an LME model as the intended primary analysis, but the posted summary does not include a fitted model result. The reported comparison is an unadjusted recalculation from summary statistics. The result therefore should not be treated as a substitute for model details or interpreted as though it accounts for every feature of the longitudinal data. The analysis population includes those with more than one ALSFRS-R value, and the effect of missing measurements is not reported in the registry record.

The study purpose also refers to neuropsychiatric changes and serum biomarkers, but results for those aims are not reported in the registry record. The primary outcome and adverse-event summaries cannot answer those separate questions. The safety counts describe serious adverse events and deaths reported by group, but the available information does not establish causation or allow a complete comparison of safety.

Overall, these posted results describe the trial’s registered outcome summaries and their uncertainty. They do not establish that memantine works or does not work, and they do not support conclusions about outcomes for which results are absent. Interpretation is limited by the broad unadjusted interval, the lack of a reported fitted primary model, and incomplete information about missing data and analysis decisions.

16. Why This Trial Matters Statistically

This trial illustrates why a point estimate, a p-value, and a confidence interval answer different statistical questions. The estimated difference in ALSFRS-R rate is -0.009 score on a scale/week, which is close to zero. The p-value is 0.986, while the confidence interval crosses zero and also includes values in both directions. The point estimate is therefore not enough on its own: precision is central to understanding what the comparison can and cannot establish.

It also highlights the distinction between a planned analysis and a summary-based recalculation. The registry describes an LME model for the primary repeated-measure outcome, but does not report a fitted model result. The available comparison is unadjusted. LME models are generally suited to repeated observations, but the choice of model details and assumptions matters. Without those details, the posted recalculation cannot show whether the intended analysis would yield the same estimate or uncertainty.

The primary outcome population is not identical to the safety-group population: the outcome summaries include participants with more than one score, while the safety counts use the listed numbers at risk. The registry does not explain the difference. That makes clear why readers need to distinguish the population contributing to an efficacy summary from the population used for safety reporting.

Finally, sample-size planning depends strongly on the effect selected for detection. The illustrations based on fractions of the observed effect show that smaller target effects require larger participant counts under the stated assumptions. Because the original design rationale is not reported, those illustrations should not be treated as the trial’s own planning basis. Together, these features make the registry record a useful example of how analysis choices, missing-data information, and uncertainty shape interpretation beyond the headline estimate.

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 has important limits for interpreting the primary comparison. It identifies an LME model as the planned basis for efficacy but lists no analysis entries and does not report a model estimate, model standard error, covariates, time structure, or model checks. The posted unadjusted recalculation is not a substitute for those details.

The primary outcome population is described as 54/58 participants randomized to memantine and 29/31 randomized to placebo who had more than one ALSFRS-R value. The registry does not explain why other randomized participants are absent from that analysis population or report their outcomes. The missing-data approach, assumptions, and sensitivity analyses are not reported in the registry record, limiting assessment of possible bias.

The registry lists three outcome measures, but no secondary results are posted, and no analysis entries are listed. The planned statistical approach to multiplicity is not reported. The study purpose includes neuropsychiatric and serum biomarker aims, but the available result summaries do not address them. The record also does not report an equivalence margin or a result from an equivalence analysis.

For safety, the registry reports affected-participant counts and deaths, along with serious event terms and percentages. The group attribution and denominators for the listed serious event terms are not reported in the registry record. Details about event timing and participant exposure are also absent from the stated details. These limitations mean the posted counts should be read descriptively rather than as a complete comparative safety analysis.

The trial’s original sample-size justification and design assumptions are not reported. The planning illustrations are based on the observed effect and stated assumptions, so they do not clarify the investigators’ original planning rationale. These gaps constrain conclusions about precision, analysis robustness, safety comparisons, and the relationship between planned and reported analyses.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Therapy in Amyotrophic Lateral Sclerosis (TAME) — NCT02118727.
• Registry record last updated 2022-11-29; status last verified 2022-11.
• Results first posted 2022-11-29.
• 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 identifies NCT02118727, Therapy in Amyotrophic Lateral Sclerosis (TAME), as a completed phase 2 randomized, quadruple-masked, parallel-group study sponsored by the University of Kansas Medical Center. Enrollment is listed as 89. The recorded start date is 2018-11-07, and the primary completion date is 2021-07-22.

The study compared memantine, taken by mouth at up to 20 mg twice a day, with placebo for 36 weeks. The primary outcome was ALSFRS-R rate of change during therapy. The posted arm estimates are -0.29 score on a scale/week for memantine and -0.281 for placebo; the unadjusted estimated difference is -0.009, with p = 0.986 and a confidence interval spanning zero.

The registry describes an LME model as the intended primary analysis but does not report a fitted model result. It lists no analysis entries and no posted secondary outcome results. Safety summaries report serious adverse events affecting 8 participants in the memantine group and 3 in the placebo group, with 3 deaths in each group. The registry record does not report the original sample-size rationale or key details needed to assess missing data and the planned model.