A 3-Month Open-Label Safety and Efficacy Study of TNX-102 SL Tablets in Fibromyalgia Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A 3-Month Open-Label Safety and Efficacy Study of TNX-102 SL Tablets in Fibromyalgia Patients (NCT02589275), a na phase 3 study. The primary outcome is Newly Emergent Adverse Events.

2. Clinical Question

The study was designed to collect safety and efficacy information during an open-label extension. Its posted primary endpoint is specifically a count of participants experiencing newly emergent adverse events, rather than a measure of symptom change. The registry record lists no analyses, and does not give an efficacy endpoint result. The adverse-event summaries provide group counts, but the extension has no concurrent randomized comparator for attributing differences to treatment. Accordingly, its central statistical question is descriptive: what adverse-event experience was recorded among participants who entered this extension and received TNX-102 SL?

3. Trial Design

This was a Phase 3, multicenter, open-label extension study of TNX-102 SL in people with primary fibromyalgia. The official title describes daily bedtime use for 3 months. The record lists a single-group model, no allocation, and no masking. The planned study treatment was a 2.8 mg sublingual tablet taken daily at bedtime. The stated purpose was to evaluate longer-term safety and efficacy after participants completed the double-blind studies TNX-CY-F301 and TNX-CY-F302.

The brief summary says participants would not be made aware of which therapy they had received during the preceding double-blind study. That concealment of prior assignment is distinct from masking within the extension: treatment in the extension itself was open-label. The record does not describe a concurrent placebo group or randomized allocation for this extension. Although adverse-event summaries are separated into TNX-TNX and PBO-TNX groups, the record explicitly notes that these are not a like-for-like control comparison.

The study enrolled 375 participants and is listed as completed. Its recorded start date is 2015-08-04, and its primary completion date is 2016-09-28. The primary outcome timeframe is up to 3 months from the first dose. The registry states that the reported population includes patients with adverse-event data collected when the study was terminated. It does not explain in further detail how this population relates to the enrollment total, nor does it provide a participant-flow account or reasons for missing endpoint data.

The registry lists one outcome measure and no analyses. Its primary outcome is the number of participants with at least one adverse event beginning after the first dose of TNX-102 SL. The design therefore supports description of observed events in the extension, but not a randomized estimate of the treatment effect relative to a concurrent control.

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: Data is reported for all patients with AE data collected at the time when this study was terminated.

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 Newly Emergent Adverse Events, measured over Up to 3 months from first dose.

The primary endpoint is the number of participants with at least one adverse event that began after the first dose of TNX-102 SL. The stated observation window is up to 3 months from that first dose. The outcome is a participant count: a person with multiple qualifying events is still counted as a participant with an event, rather than contributing a separate count for every event.

Counting participants with any newly emergent adverse event is a way to describe how often participants experienced an event during exposure and follow-up. It is a safety-oriented summary, not a measure of symptom improvement. The registry also provides adverse-event summaries separated into serious and other events, but does not report a separate statistical analysis of the named primary endpoint. In particular, the record does not give an endpoint estimate with a confidence interval or a formal hypothesis test.

Interpretation depends on the endpoint definition and the population observed. Events are counted only when they began after the first dose, and the registry says results include people with adverse-event data available when the study was terminated. It does not report a more detailed definition of event attribution, the completeness of follow-up, or how missing adverse-event information was handled. The endpoint can therefore describe events recorded in this extension, but the registry record alone does not establish that an event was caused by treatment.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

The registry identifies the primary outcome as a count of participants with at least one newly emergent adverse event. This kind of endpoint reduces a participant’s event history to whether they experienced at least one qualifying event during the stated follow-up. It can be an accessible safety summary, but it does not retain the number of events per participant, their duration, or their clinical course. Those details are not reported in the registry record.

What does a participant count represent?

A count of participants with an event is different from a count of all events. A participant who experiences several adverse events still contributes once to a “participants with at least one event” outcome. This makes the endpoint a measure of how many people were affected, not how frequently events recurred. The registry’s endpoint definition also requires that the event begin after the first dose. It does not provide further detail about event adjudication or attribution.

How should group counts be read?

The adverse-event summaries give an at-risk population and an affected-participant count for each listed category and group. These are descriptive counts. A count depends on how many people contributed data and how much follow-up they had. The record says the endpoint population comprises people with AE data collected when the study was terminated, but does not report individual follow-up time or whether observation differed across participants. Without those details, the counts do not provide a time-adjusted event rate.

Even if a count is accompanied by a percentage, that percentage does not make an uncontrolled group comparison causal. A randomized comparison aims to make treatment groups comparable, on average, with respect to both measured and unmeasured factors. This extension is described as single-group, with allocation listed as not applicable and masking listed as none. The record explicitly warns that its TNX-TNX and PBO-TNX groups are not a like-for-like control comparison. Their observed counts therefore cannot be interpreted as a randomized treatment contrast.

What would a confidence interval add?

In general, a confidence interval expresses the uncertainty around an estimate under a specified statistical model and sampling framework. It helps distinguish a precise estimate from one compatible with a broad range of values. No confidence interval for the primary endpoint or a group comparison is reported in the registry record. A confidence interval would not, by itself, fix the lack of randomization or remove possible differences between the listed groups.

Why does follow-up matter?

In general, the opportunity to observe an event depends on how long participants are followed and whether their event information is complete. If observation differs, a simple count may not be directly comparable across groups. Time-to-event methods can incorporate follow-up time and censoring, but they require suitable event-time and censoring information. The registry record does not report such an analysis, event times, or a time-to-event endpoint. The stated outcome is a participant count over a window of up to 3 months.

What does the sample-size illustration mean?

In general, a sample-size calculation describes the number of participants that a hypothetical design might need under specified assumptions about the effect, significance level, allocation, and statistical method. Such a calculation is a planning illustration, not an observed result and not evidence that a study had a particular probability of detecting an effect after the fact. The registry record does not report the trial’s own sample-size justification. It also does not report a formal analysis of the primary endpoint, so no power conclusion for an observed primary result can be drawn from the posted record.

The sample-size assumptions described for planning use a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and an effect defined as a stated fraction of the observed effect. These are assumptions for an illustrative calculation, not the design features of this single-group extension. The record does not state that the extension was planned or analyzed as a randomized two-arm test. No participant-per-arm planning values are available in the registry record.

Finally, the registry lists no analyses. That absence means there is no reported inferential framework to assess for the primary endpoint: no stated model, covariate adjustment, hypothesis test, or estimate with uncertainty. General statistical methods can clarify what such analyses would aim to do, but they cannot supply missing results or establish that any particular method was used.

10. Exact Estimates for Each Arm

No two-arm comparison is shown for this trial, but each arm's result can still be summarized with exact interval estimates computed from the posted counts. The Clopper-Pearson interval is exact and conservative; the Wilson interval is a common closer-to-nominal alternative; the Bayesian credible interval uses a uniform Beta(1, 1) prior. For small groups the three can differ noticeably.

R code and output

arm    <- c("TNX-TNX", "PBO-TNX")
n      <- c(170, 186)
events <- c(77, 132)
for (i in seq_along(n)) {
  ci <- binom.test(events[i], n[i])$conf.int          # Clopper-Pearson exact interval
  cr <- qbeta(c(0.025, 0.975), 1 + events[i], 1 + n[i] - events[i])   # Bayesian, uniform Beta(1,1) prior
  wi <- (function(x, n) { p <- x / n; z <- qnorm(0.975); d <- 1 + z^2 / n
    c((p + z^2 / (2 * n) - z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d, (p + z^2 / (2 * n) + z * sqrt(p * (1 - p) / n + z^2 / (4 * n^2))) / d) })(events[i], n[i])
  cat(sprintf("ARM|%s|%d|%d|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f|%.1f\n", arm[i], n[i], events[i], 100 * events[i] / n[i],
              100 * ci[1], 100 * ci[2], 100 * wi[1], 100 * wi[2], 100 * cr[1], 100 * cr[2]))
}

ARM  TNX-TNX  170  77  45.3  37.7  53.1  38.0  52.8  38.0  52.8
ARM  PBO-TNX  186  132  71.0  63.9  77.4  64.1  77.0  64.1  77.0

11. Multiplicity, Missing Data and Analysis Populations

The registry lists one outcome measure and no analyses. It does not report a statistical analysis plan, a hierarchy of confirmatory outcomes, or any multiplicity adjustment. Multiplicity arises when several hypotheses or endpoints are tested: repeated testing can raise the chance of finding an apparently notable result by chance. With no reported hypothesis tests or analysis plan, it is not possible to determine whether multiplicity was relevant to any sponsor analysis or how it may have been handled. No inferential result should be assumed from the listed outcome alone.

The stated primary endpoint is a participant count for newly emergent adverse events, with a timeframe of up to 3 months from the first dose. The record says the reported population includes participants with adverse-event data collected when the study was terminated. It does not describe the extent or pattern of missing adverse-event information, how participants without complete data were treated, or whether the endpoint population differed from the enrollment population. The listed enrollment of 375 and the at-risk counts in the adverse-event summaries do not, by themselves, explain participant flow or missingness.

In general, missing data can affect interpretation if the probability of missingness is related to participants’ experiences or outcomes. Different assumptions about missing observations can lead to different conclusions. The registry record does not report a missing-data method, sensitivity analysis, or a detailed analysis population definition beyond the statement about AE data collected at termination.

The record also does not state whether any interim analyses were conducted, or describe a protocol-specific statistical plan. Those details are not reported in the registry record. Consequently, the posted information supports a descriptive reading of the adverse-event summaries but not an assessment of the full prespecified analysis strategy.

12. 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: lymphadenopathy (0.6%), chronic sinusitis (0.6%), ovarian cyst (0.6%).

The serious adverse-event summaries list 170 participants at risk in the TNX-TNX group, with 2 affected, and 186 at risk in the PBO-TNX group, with 0 affected. The record lists no deaths in either group. The most frequent serious-event terms reported are lymphadenopathy, chronic sinusitis, and ovarian cyst, each at 0.6%.

These are observed counts and reported percentages, not estimates from a randomized comparison. The group labels do not indicate concurrent randomized arms, and the registry cautions that the groups are not a like-for-like control comparison. The numbers should therefore be read as descriptions of events recorded in each listed group, not as evidence that one group had a higher or lower treatment-related risk.

A serious-event count is also distinct from the total number of adverse events: the endpoint is a count of participants with an event, and the summaries distinguish serious from other adverse events. The registry record does not provide event timing, duration, clinical outcomes, or more detailed severity information. A zero count in a group means no serious events were recorded for that group in this summary; it does not establish that the underlying possibility of such events is absent.

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

The results describe adverse events recorded during an open-label extension in which participants received TNX-102 SL daily at bedtime for 3 months. The registry provides counts for serious and other adverse events in the TNX-TNX and PBO-TNX groups, along with serious-event terms and deaths. These observations can inform a descriptive account of what was recorded among participants with AE data at the time the study was terminated.

They do not establish a causal treatment effect. The study is listed as single-group, with no allocation and no masking, and the registry explicitly says the listed groups are not a like-for-like control comparison. There is no concurrent randomized comparator described for the extension. Differences between group counts cannot therefore be attributed to the extension treatment on the basis of these summaries.

The record also does not report a separate result for the named primary endpoint, a confidence interval, a hypothesis test, or an efficacy analysis. The stated purpose includes safety and efficacy, but the posted primary outcome is safety-focused and no efficacy outcome result is listed. Thus, the record does not support a conclusion about efficacy or a formal comparison of treatment strategies.

In general, absence of a statistically reported difference is not evidence of equivalence, and absence of a recorded event is not proof that an event cannot occur. Those distinctions are especially important when the registry does not report an inferential analysis or a randomized control comparison. The data are most appropriately understood as descriptive observations from this extension and its registry-defined population, rather than as a definitive evaluation of benefit or comparative safety.

14. Why This Trial Matters Statistically

This trial illustrates why the design and the estimand—the precise quantity a study aims to describe or compare—matter when reading clinical results. An open-label extension can collect additional safety observations after participants finish double-blind studies. It can describe events recorded during ongoing treatment and follow-up. But an extension without a concurrent randomized control group answers a different question from a randomized trial designed to estimate comparative effects.

The registry’s primary endpoint is a participant count for newly emergent adverse events. That is a clearly interpretable descriptive quantity, but it does not capture the full pattern of events, such as recurrence or duration. The serious and other event summaries add context, yet the record does not report a formal endpoint analysis, uncertainty estimates, or an efficacy result. A statistical reader should distinguish the useful descriptive information from claims that would require additional analysis or a different design.

The record also highlights the importance of transparent reporting. It identifies the population as people with AE data collected when the study was terminated, but does not provide a detailed participant-flow account or a missing-data analysis. It lists no analyses and does not report the study’s own sample-size justification, interim analyses, or statistical plan. These omissions limit the ability to assess precision and the completeness of the analysis.

For clinical researchers and biostatisticians, the central lesson is not to treat group labels or event counts as though they created a controlled comparison. Study structure, endpoint definition, follow-up, and analysis reporting determine what conclusions the data can support. Here, the strongest statistical interpretation is descriptive and bounded by the registry record.

15. 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 absence of a concurrent randomized control group is a central limitation for comparative interpretation. The registry describes a single-group open-label extension and explicitly cautions that TNX-TNX and PBO-TNX are not like-for-like control groups. Observed differences in their adverse-event summaries cannot be cleanly attributed to treatment, because the record does not establish that the groups are comparable for a causal comparison.

The primary outcome is defined, but a separate primary endpoint analysis is not reported. The registry lists one outcome measure and no analyses. It gives no confidence interval, hypothesis test, adjusted estimate, or efficacy endpoint result. The study’s stated interest in safety and efficacy therefore exceeds what can be concluded from the posted primary outcome information.

The recorded endpoint population comprises participants with AE data collected when the study was terminated. The record does not explain the relationship between that population and the enrollment total of 375, describe reasons for incomplete data, or report a missing-data sensitivity analysis. Individual follow-up duration and event timing are also not reported, limiting assessment of whether groups had comparable opportunity for event observation.

Other design and analysis details are not reported in the registry record, including a study-specific sample-size justification, interim analyses, and a statistical analysis plan. The posted summary also does not provide event-level descriptions beyond the named serious-event terms and percentages, such as duration or outcome. These limits mean the safety counts are informative as recorded observations but do not support broad comparative or causal claims.

16. Related Tutorials

17. Related Statistical Calculators

18. Sources

• ClinicalTrials.gov: A 3-Month Open-Label Safety and Efficacy Study of TNX-102 SL Tablets in Fibromyalgia Patients — NCT02589275.
• Registry record last updated 2025-02-20; status last verified 2025-02.
• Results first posted 2024-12-11.
• 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.

19. Record Summary

NCT02589275 is a completed Phase 3, multicenter, open-label extension study of TNX-102 SL tablets in people with primary fibromyalgia. The intervention was a 2.8 mg sublingual tablet taken daily at bedtime for 3 months. The study enrolled 375 participants and was intended to evaluate safety and efficacy after completion of TNX-CY-F301 and TNX-CY-F302.

The primary endpoint was the number of participants with at least one adverse event beginning after the first dose, over up to 3 months. The record lists one outcome measure and no analyses. Its adverse-event summaries report serious and other events for the TNX-TNX and PBO-TNX groups, but the registry warns these are not like-for-like control groups. No deaths are listed. A separate primary endpoint analysis, efficacy result, and formal comparative estimate are not reported in the registry record.