Study of FX006 vs Normal Saline in Patients With Osteoarthritis of the Knee: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Study of FX006 vs Normal Saline in Patients With Osteoarthritis of the Knee (NCT02116972), a randomized phase 2 study. The primary outcome is Change From Baseline to Week 12 in the Weekly Mean of the Average Daily (24-hr) Pain Intensity Scores for 32 mg FX006 Versus Placebo.

2. Clinical Question

3. Trial Design

This was a phase 2, randomized, quadruple-masked, parallel-group study. The registry lists 310 enrolled participants and three intervention groups: FX006 16 mg, FX006 32 mg, and placebo. The treatment descriptions specify a single 5 mL intra-articular injection in each group; placebo was normal saline. The study is marked completed.

The primary endpoint concerns the FX006 32 mg and placebo groups, rather than all three arms. The posted outcome population is described as randomized patients who received study drug assigned to those two groups. The primary result therefore should not be read as a direct comparison of 16 mg with either 32 mg or placebo. Results for a 16 mg primary comparison are not reported in the registry record.

The trial began in 2014-04, and primary completion is listed as 2015-11. Its brief description states that the study assessed the magnitude and duration of pain relief after a single injection of two FX006 doses relative to normal saline. The primary endpoint is specifically a Week 12 comparison. Secondary outcomes include a function subscale and a patient global impression measure at Week 12, as well as pain scores at later time points. The registry record does not report further details about allocation procedures, masking procedures, follow-up completion, or 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.

Before assignment: Subjects were screened within 21 days of being randomized

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Randomized patients who received study drug assigned to the FX006 32 mg arm and the placebo arm

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: A total of 310 subjects were randomized to a treatment assignment but only 306 enrolled subjects received their randomized treatment. Four (4) subjects randomized to placebo arm were were not treated.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline to Week 12 in the Weekly Mean of the Average Daily (24-hr) Pain Intensity Scores for 32 mg FX006 Versus Placebo, measured over Baseline and Week 12.

The primary endpoint is change from baseline to Week 12 in the weekly mean of average daily 24-hour pain-intensity scores. Pain is recorded on an 11-point numeric rating scale: 0 means no pain, and 10 means pain as bad as you can imagine. The endpoint is continuous, so each group’s result is expressed as a mean change rather than as a count of participants meeting a response threshold.

Change from baseline focuses the comparison on movement in scores over the stated period, while the between-group mean difference assesses whether average change differs between FX006 32 mg and placebo. Negative changes indicate lower scores at Week 12 relative to baseline, but the relevant treatment comparison is the difference between the groups, not either group’s change considered alone. The registry reports least-squares means with standard errors for the arms and a mean difference for the comparison.

Pain intensity is used as the primary outcome because the study’s stated purpose was to assess the magnitude and duration of pain relief. The endpoint addresses the Week 12 time point; it does not by itself describe the entire course of pain scores at other visits or establish how individual patients changed.

7. Results Posted in the Registry

• FX006 32 mg: mean -3.08 versus -2.50 on control; difference -0.58 (95% CI -1.23 to 0.07), standardized effect d = -0.25; z-test p = 0.079.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Change From Baseline to Week 12 for WOMAC C (Function Subscale)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • FX006 32 mg: mean -1.01 versus -0.79 on control; difference -0.22 (95% CI -0.46 to 0.02), standardized effect d = -0.26; z-test p = 0.067.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between FX006 32 mg and the control. The interval gives the range of effects that remain compatible with the data.

Change From Baseline to Week 12 for Patient Global Impression of Change (PGIC)

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard errors.
• Result:
  • FX006 32 mg: mean 2.60 versus 2.70 on control; difference -0.10 (95% CI -0.49 to 0.29), standardized effect d = -0.07; z-test p = 0.614.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between FX006 32 mg and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes add context to the primary pain comparison by considering function, patient global impression of change, and pain at later time points. The posted Week 12 function contrast points in the same general direction as the primary pain estimate, whereas the patient global impression comparison shows less separation between groups. The later pain assessments broaden the planned outcome coverage, but their results are not reported in the posted secondary results described in the registry record.

The secondary comparisons do not provide a clearly distinct pattern that overturns the primary result’s uncertainty. Their intervals include zero, and the posted unadjusted p-values do not cross the planning illustration’s 0.05 significance level. These results should be read as additional, exploratory context rather than as independent confirmation of a treatment effect.

Unadjusted secondary results carry less weight when several outcomes or comparisons are considered because repeated testing can increase the chance of a seemingly notable result by chance. The registry record does not report a multiplicity adjustment or a testing hierarchy. The results therefore do not support treating any one secondary comparison as a definitive finding, nor can the secondary outcomes resolve the uncertainty in the primary comparison.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Change From Baseline to Week 12 for WOMAC C (Function Subscale)
• Change From Baseline to Week 12 for Patient Global Impression of Change (PGIC)
• Change From Baseline to Week 16 and Then Week 20 and Then Week 24 in the Weekly Mean of the Average Daily (24-hour) Pain Intensity Scores

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous, so the registry reports group means and a mean difference. A mean summarizes the central level of a measure across participants, while the mean difference expresses the contrast between the two group means on the original scale. Here, the posted change estimates are least-squares means. In general, a least-squares mean is a model-based estimate of a group mean; the registry record does not provide the model’s full specification or list the covariates used to obtain these estimates.

The sponsor describes its analysis as a longitudinal mixed-effects model and labels the contrast a mean difference in final values. In general, longitudinal models can use measurements collected over time and account for the fact that observations from the same participant are related. A mixed-effects model can represent both average patterns and participant-level variation. The precise repeated-measure structure, covariance assumptions, treatment of baseline values, and missing-data approach are not reported in the registry record, so those details cannot be assessed from the posted description.

The arm-level standard errors describe uncertainty in the estimated means. They are not standard deviations of individual participants’ scores. A standard error generally becomes smaller when an estimate is more precise, but its interpretation depends on the analysis and data used. The registry reports standard errors of 0.229 for FX006 32 mg and 0.238 for placebo.

The confidence interval describes a range of effect values reasonably compatible with the data under the analysis and its assumptions. A 95% confidence interval is generated by a procedure that, under repeated use in comparable settings, would cover the true effect in 95% of intervals. It is not a statement that there is a 95% probability that this particular interval contains the true effect. For the sponsor’s reported comparison, the interval extends from -1.22 to 0.07.

The p-value measures how unusual results at least as discrepant from a null hypothesis of no difference would be if that null hypothesis and the analysis assumptions were correct. It is not the probability that the null hypothesis is true, nor does it quantify the importance of the effect. The sponsor reports a p-value above 0.05. The recalculated unadjusted p-value is 0.079. The registry’s planning illustration uses a two-sided significance level of 0.05; that planning assumption should not be confused with evidence about the trial’s original design.

The recalculated comparison uses posted summary statistics and is explicitly an unadjusted recalculation. In general, a summary-statistic recalculation can be useful as a check on the direction and approximate size of a contrast, but it cannot reproduce a model-based analysis when the underlying participant-level data and model details are unavailable. The recalculation also uses an approximate pooled standard deviation derived from the posted summaries. Such an approximation does not capture all features of a longitudinal model. Since the registry does not mark a discrepancy, the sponsor result remains the reported model analysis, and the recalculation should be read as supplementary.

In general, over-dispersion refers to variability greater than a model’s assumed distribution allows. It matters because underestimated variability can make standard errors too small and uncertainty appear narrower than it is. The primary outcome is continuous, and the registry does not report a dispersion test or an over-dispersion adjustment; this general issue should not be mistaken for a finding about this trial.

The registry reports 13 outcome measures and one analysis for the primary result. The presence of multiple outcomes and comparisons raises the general issue of multiplicity: when many hypotheses are tested, the chance of at least one apparently positive result can increase. The registry record does not report a multiplicity adjustment or a prespecified hierarchy for interpreting the outcomes. The secondary comparisons are therefore described as unadjusted and should be interpreted with appropriate caution.

11. Confidence Intervals and Precision

Precision concerns how tightly the data estimate the between-group effect. The sponsor’s 95% confidence interval for the primary mean difference runs from -1.22 to 0.07 units. Because this interval includes zero, it is compatible with no average difference as well as with effects favoring either group. Its span also indicates that the exact size of the difference is uncertain.

The point estimate is -0.58 units, but that single value should not be mistaken for a known effect. The interval conveys uncertainty around it. The reported p-value is above 0.05, consistent with the interval’s inclusion of zero under the usual two-sided interpretation. Neither statistic determines whether a difference would be important to patients; the registry record does not report a threshold for clinical importance.

The sponsor’s interval and p-value come from the longitudinal mixed-effects model. The recalculated interval is based on posted summary statistics and an approximate standard deviation, so it is a separate unadjusted calculation rather than a more precise version of the sponsor analysis. Small differences in the numerical limits can arise because the methods differ. The sponsor’s reported model result is the relevant reported analysis, while the recalculation provides a check using a simpler approach.

Standard errors for the arm means give context for the uncertainty in each estimate, but the precision of the treatment contrast is best read from its confidence interval. The interval applies to the mean difference for the stated population and Week 12 endpoint; it does not describe individual variation or guarantee that all patients experienced changes close to the group mean.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The sample-size figures are planning illustrations based on the observed effect, not the trial’s own design. They are participants per arm, not total events. The stated 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 set as the stated fraction of the observed effect. The trial’s original sample-size justification and design assumptions are not reported in the registry record.

The 80% power column represents the number of participants per arm in a hypothetical future design that, under the listed assumptions and effect size, would have an 80% chance of detecting an effect if that effect were present. The 90% power column uses a higher target chance of detection and consequently requires more participants per arm. For the primary endpoint, planning for 100% of the observed effect corresponds to 260 participants per arm for 80% power and 348 for 90% power. Planning for 75% of the observed effect corresponds to 462 and 618 participants per arm, respectively; planning for 50% corresponds to 1,038 and 1,389.

Smaller effects are harder to distinguish from random variation, so the illustrations require more participants per arm when the target effect is a smaller fraction of the observed effect. These values are not observed power. Observed power is not a useful substitute for interpreting the estimate and its confidence interval; the planning figures instead show how required sample size changes under stated assumptions.

13. 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 = -0.580; pooled standard deviation σ = 2.357.
2. n = 2 × (1.96 + 0.842)² × 2.357² / -0.580² = 259.3, rounded up to 260 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 <- -0.58                      # observed difference in means
sigma <- 2.35733906876981                     # 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(2116972)
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 = 260 (80% power), 348 (90% power)
 75% of observed effect: n per arm = 462 (80% power), 618 (90% power)
 50% of observed effect: n per arm = 1,038 (80% power), 1,389 (90% power)
power.t.test: 260.3 
Simulated power at n = 260 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.

14. Analyses Reported by the Sponsor

The registry reports the sponsor’s primary analysis as a longitudinal mixed-effects model, with a mean difference in final values of -0.58. Its 95% confidence interval is -1.22 to 0.07, and its p-value is above 0.05. This is the trial’s reported inferential analysis and should be distinguished from the simple recalculation based on posted summary statistics.

The recalculation also gives a mean difference of -0.58 and an unadjusted p-value of 0.079. Its direction and broad interpretation align with the sponsor’s result: the estimated difference favors a greater reduction in the FX006 32 mg group, while the interval includes zero. The recalculation uses an approximate standard deviation and does not reproduce the sponsor’s longitudinal model. It is therefore best treated as a supplementary check, not as a replacement or correction.

The registry does not flag a discrepancy between the results. The sponsor’s model-based estimate remains the principal reported result; the simpler calculation helps illustrate how the posted summary values compare under an unadjusted approach.

15. Multiplicity, Missing Data and Analysis Populations

The registry lists 13 outcome measures and one analysis for the primary result. The study also has secondary outcomes, including function, patient global impression of change, and pain scores at later time points. When multiple outcomes are examined, multiplicity is a general concern: the more hypotheses tested, the greater the opportunity for chance findings if each is interpreted without adjustment.

The registry record does not report whether the protocol specified a hierarchy of endpoints, whether the analyses were prespecified in detail, or whether any multiplicity adjustment was used. The secondary comparisons are presented as unadjusted. Their p-values should therefore not be interpreted as if they arose from a single isolated test. This does not invalidate the estimates, but it limits the strength of claims based on any one secondary result.

The outcome population for the primary comparison is described as randomized patients who received study drug assigned to the FX006 32 mg and placebo arms. The registry record does not report a fuller definition of the analysis population, such as a detailed intention-to-treat rule, nor does it give participant flow or the number with endpoint data at each assessment. The method used to address missing outcome data is not reported in the registry record.

Those omissions matter because the treatment contrast can depend on who contributes data and how missing measurements are handled. A longitudinal model may accommodate incomplete repeated measurements under assumptions, but the registry record does not describe the assumptions or implementation. It is not possible from the posted record alone to assess the robustness of the result to alternative missing-data approaches or to determine how much missingness occurred.

Likewise, the registry does not report interim analyses, stopping rules, a detailed statistical analysis plan, or the trial’s original sample-size justification. These should not be inferred from the completed status or from the planning illustration based on the observed effect. The primary result can be interpreted as reported, while these unreported design and analysis details limit independent appraisal.

16. 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: femur fracture (1.0%), myocardial infarction (1.0%), osteoarthritis (1.0%), rheumatoid arthritis (1.0%).

The registry reports serious adverse events among 1 participant in the FX006 16 mg group, 3 in the FX006 32 mg group, and 0 in the placebo group. The corresponding numbers at risk are 102, 104, and 100. Deaths are reported as 0 in each group. The registry also lists other adverse events affecting 14 participants in the 16 mg group, 13 in the 32 mg group, and 21 in the placebo group.

These are descriptive counts, not a complete comparison of safety. The listed serious-event terms include Femur Fracture, Myocardial Infarction, Osteoarthritis, and Rheumatoid Arthritis, each reported at 1.0%; the registry data given do not identify these terms by treatment group. Counts alone do not establish whether an event was caused by treatment, and small numbers provide limited precision for comparing groups. The registry record does not provide further detail about event timing, duration, severity beyond the serious classification, or attribution.

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

The primary result estimates a Week 12 difference in average change in pain-intensity scores between FX006 32 mg and saline. The point estimate is -0.58 units, favoring a larger average reduction in the FX006 group. The sponsor’s confidence interval includes zero, and its reported p-value is above 0.05. The posted data therefore do not exclude no average difference, while also allowing effects in either direction within the interval.

This finding does not prove that the treatments are identical, and it does not establish that FX006 has a particular effect for every participant. Group averages can conceal variation in individual experiences. The registry does not report a threshold for clinically important change, so the numerical difference cannot be classified as clinically meaningful on the basis of this record alone.

The result is specific to the 32 mg versus placebo comparison, the stated outcome population, and the Week 12 pain endpoint. It does not supply a primary efficacy estimate for the 16 mg arm, establish effects at every other time point, or resolve the secondary outcomes’ uncertainty. The safety counts likewise describe reported events and do not establish causal relationships or comparative risks.

Overall, the posted estimate is in the direction of greater average reduction with FX006 32 mg, but its uncertainty is substantial enough that the interval includes no difference. Statistical evidence, clinical importance, and individual response are different questions; the registry record addresses the first through an estimated group contrast and its uncertainty, but does not provide all information needed to answer the others.

18. Why This Trial Matters Statistically

This trial illustrates why a treatment comparison should be read through both its point estimate and its uncertainty. The primary estimate favors a greater average reduction in pain scores for FX006 32 mg, but the sponsor’s confidence interval includes zero. Considering the direction alone would overstate what the data establish; considering a p-value alone would also omit the possible range of effects.

The example also distinguishes a model-based analysis from a simple recalculation using summary statistics. The recalculated mean difference is close to the sponsor’s estimate, but the recalculation does not reproduce the longitudinal mixed-effects model. Similarity between estimates is reassuring as a numerical check, not proof that the methods are interchangeable.

Finally, the trial highlights the importance of endpoint scope and multiplicity. The primary analysis concerns one dose and a specific Week 12 pain measure, while the registry lists additional doses and secondary outcomes. Those results address related but different questions. Without a reported testing hierarchy or multiplicity adjustment, secondary findings require cautious interpretation. The unreported details about missing data and the original sample-size design also limit how fully the analysis can be evaluated from the registry record.

19. 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 leaves several trial-specific details unreported. It does not give the full statistical model specification, including covariates, repeated-measure structure, covariance assumptions, or missing-data handling. It also does not provide participant flow or endpoint availability over time. These gaps make it difficult to assess how the analyzed population and model assumptions shaped the estimate.

The primary posted comparison is limited to FX006 32 mg versus placebo. Although the trial included FX006 16 mg, the registry record does not report a corresponding primary outcome estimate for that arm in the results. The primary endpoint is tied to Week 12 and does not by itself represent the full duration of follow-up or all pain measurements.

The analysis is one among a set of outcomes, yet the registry does not report a multiplicity strategy or a prespecified hierarchy. Secondary findings consequently require caution. The original sample-size rationale, interim analyses, stopping rules, and detailed analysis plan are also not reported in the registry record. The planning illustration based on the observed effect cannot fill those gaps.

Safety information is presented as event counts, with limited detail on timing, attribution, and group-specific serious-event terms. The counts should not be treated as definitive evidence about comparative safety. These limitations do not erase the posted estimates; they define what can and cannot be concluded from the registry record.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Study of FX006 vs Normal Saline in Patients With Osteoarthritis of the Knee — NCT02116972.
• Registry record last updated 2024-01-24; status last verified 2024-01.
• Results first posted 2018-03-20.
• 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.

23. Record Summary

NCT02116972 is a completed phase 2 randomized, quadruple-masked, parallel-group study of FX006 for knee osteoarthritis. The registry lists 310 enrolled participants and groups assigned to FX006 16 mg, FX006 32 mg, or normal saline placebo. The primary posted comparison is FX006 32 mg versus placebo for change in weekly mean average daily pain-intensity scores from baseline to Week 12.

The sponsor reports a longitudinal mixed-effects analysis with a mean difference of -0.58 and a 95% confidence interval from -1.22 to 0.07. The interval includes zero. The simple unadjusted recalculation has the same estimated direction and does not replace the sponsor analysis. Secondary outcomes and adverse-event counts provide additional context, while missing analysis details and multiplicity procedures are not reported in the registry record.