Safety, Tolerability and Preliminary Efficacy of Multiple Intra-articular Injections of LRX712 in Patients With Knee OA: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Safety, Tolerability and Preliminary Efficacy of Multiple Intra-articular Injections of LRX712 in Patients With Knee OA (NCT04097379), a randomized phase 2 study. The primary outcome is Change From Baseline in Cartilage Volume in the Index Region Measured by 7 Tesla MRI.

2. Clinical Question

3. Trial Design

This was a Phase 2, randomized, placebo-controlled, parallel-group study with quadruple masking. Its title describes a subject- and investigator-blinded study. The registry reports enrollment of 45 and a completed status. The recorded study start was 2020-07-20, and the primary completion date was 2024-08-08. These dates describe the registry record; they do not give the duration of treatment for an individual participant.

The arms were LRX712 15 mg, LRX712 25 mg, LRX712 75 mg and placebo. All were administered by intra-articular injection every four weeks for a total of three administrations. The primary outcome results are posted for the 15 mg, 25 mg and placebo arms. The registry record does not report primary outcome results for the 75 mg arm, so its absence from those comparisons should not be interpreted as a result for that dose.

The primary outcome population included participants in the safety analysis set from the 15 mg, 25 mg and placebo arms with an available value aligned with the protocol. The reported outcome counts are 11.0 for LRX712 15 mg, 13.0 for LRX712 25 mg and 16.0 for placebo. These are outcome-specific counts, not a restatement of overall enrollment. The safety set is defined as enrolled participants who received any investigational treatment.

Multiple pharmacokinetic outcomes were also listed, including plasma concentration timing and concentration measures for LRX712 and MAE344, along with synovial fluid concentrations. The registry lists 11 outcome measures and 10 posted secondary outcomes. A more detailed account of allocation procedures, the original sample-size rationale and the protocol's full analysis plan is not reported in the registry record.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: This study had a 7-week screening period.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Participants in the safety analysis set from the arms "LRX712 15 mg", "LRX712 25 mg" and "Placebo" who had an available value for the outcome measure as aligned with protocol. The Safety Set comprises all enrolled participants who received any investigational treatment.

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 Change From Baseline in Cartilage Volume in the Index Region Measured by 7 Tesla MRI, measured over Baseline, Week 28.

The primary endpoint was change from baseline in cartilage volume in the index region, measured by 7 Tesla MRI, with assessments at Baseline and Week 28. Volume is expressed in µL. The index region combines the femoral medial anterior (FMA), central (FMC) and posterior (FMP) cartilage subregions of the target knee.

MRI was used to visualize and quantify cartilage volume in this specified region. Change from baseline focuses the analysis on how the measured volume differs from each participant's starting value, rather than on the Week 28 measurement alone. In general, a structural imaging measure can describe an anatomical feature, while symptoms and function describe different aspects of a condition. The registry's primary endpoint is the structural measure; it does not itself establish how participants felt or functioned.

The endpoint is reported as a least-squares mean, a model-based estimate, with standard error as the dispersion measure. The sponsor analyzed change using a mixed effects model for repeated measures. The registry describes the outcome as preliminary efficacy assessed in the context of systemic safety and local tolerability. A change in cartilage volume should therefore be read as an imaging result, not as direct evidence of a change in pain, mobility or other clinical experience.

7. Results Posted in the Registry

• LRX712 15 mg: mean 63.30 versus 11.60 on control; difference 51.70 (95% CI -88.15 to 191.55), standardized effect d = 0.28; z-test p = 0.469.
• LRX712 25 mg: mean 49.80 versus 11.60 on control; difference 38.20 (95% CI -95.18 to 171.58), standardized effect d = 0.21; z-test p = 0.575.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is continuous: cartilage volume change is measured in µL, and the registry describes the arm results as means of a continuous measure. The sponsor reports least-squares means rather than only raw averages. A least-squares mean is a model-based estimate of an arm's outcome, constructed within the specified statistical model. It can differ from a simple arithmetic average because the model accounts for its included terms and the repeated observations.

What does the repeated-measures model do?

The registry says the analysis used a mixed effects model for repeated measures, often abbreviated MMRM. In general, repeated-measures models analyze outcome observations collected at multiple time points while representing the fact that measurements from the same participant are related. The model can use the observed outcome information across the scheduled assessments under its assumptions, rather than treating each observation as independent. The registry describes Baseline and Week 28 for the endpoint.

The specified model included baseline, treatment, timepoint and treatment-timepoint terms as fixed effects, and participant as a random effect. In general, fixed effects represent the modeled contributions of the named factors; the treatment-timepoint term allows the treatment contrast to vary across time. A participant-level random effect represents participant-specific variation and the association among measurements from the same participant. This is a general interpretation of the model components; additional implementation details are not reported in the registry record.

The registry states that missing data was assumed to be Missing at Random (MAR). In general, MAR means that, conditional on information included in the analysis, the probability that an outcome is missing does not depend on the unobserved value itself. This is an assumption, not a finding that missingness was harmless. The registry record does not report how much outcome data was missing, the reasons for missingness, or sensitivity analyses examining departures from MAR.

How should the arm comparisons be read?

The sponsor's reported contrasts are least-squares mean differences against placebo. A positive contrast means the model-estimated change in the LRX712 group is higher than the model-estimated change in placebo, using the stated direction of subtraction. The reported differences are 51.70 µL for LRX712 15 mg and 38.19 µL for LRX712 25 mg. The corresponding reported p-values are 0.237 and 0.289. A p-value describes how compatible the data are with a specified null model under the analysis assumptions; it is not the probability that the null hypothesis is true, and it does not convey the size or precision of an effect by itself.

The sponsor labels the comparisons as superiority analyses and reports 90% confidence intervals. In general, an interval estimates a range of values compatible with the observed data and the chosen procedure. Because both intervals extend across zero, the estimates do not sharply distinguish a positive contrast from no contrast or a negative contrast. The p-values should be interpreted alongside the estimates and intervals, not in isolation.

The recalculated comparisons are explicitly simple and unadjusted. They use posted arm summaries and approximate standard deviations from the reported standard errors. For LRX712 15 mg, the recalculated estimate is 51.7, with an interval from -88.15 to 191.5 and an unadjusted p-value of 0.469. For LRX712 25 mg, it is 38.2, with an interval from -95.18 to 171.6 and an unadjusted p-value of 0.575. These values provide a separate summary-statistic-based check; they are not a replacement for the sponsor's repeated-measures analysis. The registry marks no discrepancy between the reported analyses and recalculations.

Why can over-dispersion matter?

In general, over-dispersion occurs when the observed variability is greater than a model assumes. It can make standard errors too small if the model fails to represent that variability, leading to intervals that are too narrow and tests that appear more decisive than the data warrant. The primary outcome is continuous, and the posted arm standard errors are large relative to the estimated contrasts. The registry does not report an over-dispersion assessment or a separate correction for it. The relevant practical point is that variability and uncertainty are substantial in these results, which is visible in the reported intervals.

The calculations described as recalculated use a pooled standard deviation approximated from the reported arm standard errors, as indicated in the comparison details. This approximation does not reproduce the participant-level repeated-measures model. It also does not account for any analysis features beyond the posted summary statistics. For these reasons, small differences in estimates, intervals and p-values between the two approaches are not surprising and should not be treated as contradictory evidence on their own.

The registry record does not report further details such as the model's covariance structure, degrees-of-freedom method, exact handling of all visit-level observations, or any planned hierarchy for tests. These details can affect model-based uncertainty and interpretation. The clearest reading is therefore to distinguish the sponsor's stated MMRM results from the unadjusted recalculations and to retain the uncertainty expressed by both.

10. Confidence Intervals and Precision

The sponsor reports 90% confidence intervals for the two model-based contrasts. For LRX712 15 mg versus placebo, the estimate is 51.70 µL and the interval runs from -68.7 to 172.1. For LRX712 25 mg versus placebo, the estimate is 38.19 µL and the interval runs from -76.7 to 153.1. Both intervals include zero and span values on either side of it. The data therefore leave substantial uncertainty about the direction and size of the contrast.

In general, a confidence interval expresses the uncertainty of an estimate under a specified statistical method. A broad interval indicates that the data are compatible with a wide range of values, while a narrow one indicates greater precision. Here, the limits are far from their point estimates in both directions. The intervals should be considered with the outcome-specific group counts—11.0, 13.0 and 16.0—and the observed variability, rather than treating the point estimates as precise measurements of a treatment effect.

The simple unadjusted recalculations also produce intervals spanning negative and positive values: -88.15 to 191.5 for the 15 mg comparison and -95.18 to 171.6 for the 25 mg comparison. These recalculated intervals arise from a different method and should be kept distinct from the sponsor's 90% intervals. The confidence level for the recalculated intervals is not reported in the registry record. Neither set of intervals supports a narrow estimate of the contrast. Precision is limited, and the results do not rule out materially different values within the ranges represented by the reported analyses.

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 sample-size table and curve give planning illustrations based on fractions of the observed effect, not the enrollment or analysis size of this completed trial. The figures are participants per arm, not total events. 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 the stated fraction of the observed effect. The trial's own original sample-size assumptions are not reported in the registry record. These figures are not observed power.

For the LRX712 15 mg comparison, planning for 100% of the observed effect corresponds to 195 participants per arm for 80% power and 261 participants per arm for 90% power. Planning for 75% of the observed effect corresponds to 347 participants per arm for 80% power and 464 participants per arm for 90% power. Planning for 50% of the observed effect corresponds to 780 participants per arm for 80% power and 1,044 participants per arm for 90% power.

For LRX712 25 mg, planning for 100% of the observed effect corresponds to 358 participants per arm for 80% power and 479 participants per arm for 90% power. Planning for 75% of the observed effect corresponds to 636 participants per arm for 80% power and 851 participants per arm for 90% power. Planning for 50% of the observed effect corresponds to 1,430 participants per arm for 80% power and 1,914 participants per arm for 90% power.

In plain terms, the 80% and 90% columns describe the planned probability of detecting the stated effect under the listed assumptions if that effect were present. Higher target power requires more participants. Likewise, smaller effects require more participants to detect than larger effects under these assumptions. Because these illustrations use observed effects from this study, they are sensitive to uncertain estimates from small groups and should not be read as the trial's original design plan.

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 = 51.700; pooled standard deviation σ = 182.169.
2. n = 2 × (1.96 + 0.842)² × 182.169² / 51.700² = 194.9, rounded up to 195 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 <- 51.7                      # observed difference in means
sigma <- 182.168880218329                     # 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(4097379)
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 = 195 (80% power), 261 (90% power)
 75% of observed effect: n per arm = 347 (80% power), 464 (90% power)
 50% of observed effect: n per arm = 780 (80% power), 1,044 (90% power)
power.t.test: 195.9 
Simulated power at n = 195 per arm: 0.799 (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. Analyses Reported by the Sponsor

The sponsor's own analysis used a mixed effects model for repeated measures and reported least-squares mean contrasts. It estimated 51.70 µL for LRX712 15 mg versus placebo, with a 90% confidence interval from -68.7 to 172.1 and p = 0.237. For LRX712 25 mg versus placebo, it reported 38.19 µL, with a 90% confidence interval from -76.7 to 153.1 and p = 0.289.

The simple recalculations are close to the sponsor's point estimates: 51.7 and 38.2. However, the recalculated intervals and unadjusted p-values differ. The recalculation uses posted summary statistics and approximate standard deviations; it does not reproduce the sponsor's repeated-measures analysis. The estimates should therefore be identified by method rather than combined as though they were one analysis.

The registry indicates no discrepancy between the reported analyses and recalculations. The sponsor's MMRM results remain the trial's reported analysis, while the recalculations are a separate unadjusted comparison. Both approaches yield imprecise contrasts with intervals spanning negative and positive values. The registry record does not report additional protocol details that would allow a fuller reconstruction of the sponsor's model.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists 11 outcome measures and 10 posted secondary outcomes. The primary endpoint has two reported superiority comparisons, one for LRX712 15 mg versus placebo and one for LRX712 25 mg versus placebo. The secondary outcomes include plasma concentration timing and concentration measures for LRX712 and MAE344, and synovial fluid concentrations. Multiple outcomes and comparisons create a general multiplicity issue: when many statistical tests are considered, the chance of at least one apparently unusual result can increase. Whether and how that issue was addressed depends on the prespecified analysis plan.

Multiplicity adjustment, a testing hierarchy, or an explicit strategy for interpreting the collection of outcomes is not reported in the registry record. The posted p-values should therefore be read as the values reported for their individual comparisons, not assumed to be adjusted across outcomes or doses. The registry record does not state whether additional analyses were conducted but not posted.

For the primary outcome, the registry says missing data was assumed to be Missing at Random in the mixed effects model for repeated measures. This assumption supports a model-based analysis under specified conditions; it does not show that the missing-data process followed MAR. The amount and pattern of missing data, reasons for missingness, and sensitivity analyses are not reported in the registry record. Outcome-specific counts are 11.0 for LRX712 15 mg, 13.0 for LRX712 25 mg and 16.0 for placebo; the registry describes the analyzed population as participants with an available value aligned with the protocol.

The analysis population was drawn from the safety set, defined as enrolled participants who received any investigational treatment, and restricted to the listed arms with available aligned outcome values. The registry record does not report further information about exclusions, deviations from randomization, or analysis populations for each secondary outcome. These gaps limit assessment of how the observed-data analysis relates to all randomized participants and how robust the results are to alternative missing-data assumptions.

15. 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: breast cancer (8.3%), myocardial infarction (7.1%), multiple fractures (6.2%).

The registry reports serious adverse-event figures by treatment and post-treatment categories. During the treatment-period entries, the affected counts are 0 for LRX712 15 mg, 0 for LRX712 25 mg, 0 for LRX712 75 mg, 0 for the combined LRX712 15mg and 25mg entry, and 0 for placebo. The total entry reports 0 deaths and 0 serious adverse events among 45 at risk.

For the post-treatment entries, the reported serious-adverse-event affected counts are 1 for LRX712 15 mg, 1 for LRX712 25 mg, 0 for LRX712 75 mg, 2 for the combined LRX712 15mg and 25mg entry, and 1 for placebo. The post-treatment total reports 3 affected among 45 at risk, with 0 deaths. The combined entry and the individual dose entries are separate registry categories; they should not be added together.

The listed top serious terms are Breast cancer (8.3%), Myocardial infarction (7.1%) and Multiple fractures (6.2%). The registry record does not report further details about the context or clinical course of these terms in the figures listed. These summaries describe reported safety events; they do not establish causality or provide a complete assessment of risk.

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

The primary endpoint results describe model-estimated change in MRI-measured cartilage volume in a specified knee region from Baseline to Week 28. The reported contrasts favor a larger estimated change for each LRX712 dose than for placebo, but their confidence intervals include zero and extend across both negative and positive values. The estimates are therefore uncertain, and the posted data do not give a precise account of the direction or size of the underlying contrast.

These results do not establish that LRX712 restored cartilage, nor do they establish that it did not. The endpoint measures an imaging feature, and the registry record does not report symptom or functional outcome results alongside these primary comparisons. A structural measurement cannot, by itself, answer whether participants experienced a change in pain, mobility or quality of life. The record frames efficacy as preliminary and considers it in the context of safety and local tolerability.

The trial was randomized and placebo-controlled, but the outcome-specific groups were small and the estimates imprecise. The primary comparisons cover the 15 mg and 25 mg arms; although a 75 mg arm is described, its primary outcome result is not reported in the registry record. The posted results also do not provide a basis for conclusions about differences among all three LRX712 doses.

Safety results should be read separately from the imaging efficacy outcome. The registry reports no deaths and lists serious adverse-event counts by treatment and post-treatment categories. Those summaries do not establish whether an event was caused by treatment, and the record does not provide enough detail to infer comparative risk from the listed terms alone. Taken together, the results are a preliminary and uncertain account of a structural endpoint and reported safety events, not a definitive statement about clinical benefit or harm.

17. Why This Trial Matters Statistically

This trial is statistically informative because it illustrates the gap between a point estimate and the precision needed to interpret it. Both posted sponsor contrasts are positive, but both 90% confidence intervals extend from negative to positive values. The estimated direction alone therefore does not settle whether the underlying contrast is positive, absent or negative. Intervals make that uncertainty visible in a way that a single mean difference cannot.

It also shows why the analysis method matters. The sponsor used a repeated-measures model, while the separate recalculation uses summary statistics and an approximate standard deviation. Their point estimates are close, but the intervals and p-values differ. Presenting the two as distinct analyses helps readers understand that a model-based trial result cannot always be recreated from arm summaries alone.

The sample-size illustrations highlight how much planning needs can depend on the effect chosen for detection. Under the stated assumptions, planning for smaller fractions of the observed effect requires more participants per arm. At the same time, the observed effects come from small outcome groups, so they are uncertain inputs for future planning. These illustrations are useful for understanding the scale implied by the estimates, but they cannot replace a prospectively justified design calculation.

Finally, the endpoint is a structural MRI measure. Statistical evidence about that measure is not interchangeable with evidence about how participants feel or function. Keeping the outcome's meaning, the analysis assumptions and the uncertainty in view is central to an appropriately restrained reading of this Phase 2 study.

18. Limitations

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

The primary outcome groups are small, with reported counts of 11.0 for LRX712 15 mg, 13.0 for LRX712 25 mg and 16.0 for placebo. The sponsor's confidence intervals are broad and include zero, leaving substantial uncertainty around both estimated contrasts. These features limit the precision of conclusions from the posted primary results.

The primary outcome results are reported for the 15 mg and 25 mg arms and placebo, but not for the 75 mg arm described in the trial design. The registry record does not explain why the 75 mg primary result is absent. Consequently, the reported comparisons do not provide a complete outcome picture across all described doses.

The primary endpoint measures cartilage volume in an MRI-defined region through Week 28. The registry record does not report primary results for symptoms or function, or longer-term structural outcomes. It also does not report the clinical importance of a given volume change, so the practical meaning of the estimated differences cannot be established from the endpoint alone.

The analysis assumes missing data are Missing at Random, but the amount, pattern and reasons for missing data and sensitivity analyses are not reported in the registry record. Details of the model's covariance structure and other implementation choices are also not reported. Multiplicity handling across the primary comparisons and other outcomes is not reported. These limitations make it difficult to assess robustness beyond the stated analysis and posted estimates.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Safety, Tolerability and Preliminary Efficacy of Multiple Intra-articular Injections of LRX712 in Patients With Knee OA — NCT04097379.
• Registry record last updated 2026-01-30; status last verified 2026-01.
• Results first posted 2026-01-07.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

22. Record Summary

NCT04097379 is a completed Phase 2 randomized, placebo-controlled, quadruple-masked, parallel-group study of intra-articular LRX712 in participants with knee osteoarthritis. The registry reports enrollment of 45 and describes 15 mg, 25 mg, 75 mg and placebo arms, each scheduled for three administrations every four weeks.

The primary outcome was change from baseline in cartilage volume in the index region measured by 7 Tesla MRI at Week 28. Sponsor-reported comparisons versus placebo were 51.70 µL for LRX712 15 mg and 38.19 µL for LRX712 25 mg; both 90% confidence intervals include zero. The registry also reports serious adverse-event categories and 0 deaths. The primary outcome results do not include the 75 mg arm, and the posted estimates remain imprecise.