A Phase 3 Study of Tanezumab for Chronic Low Back Pain: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Phase 3 Study of Tanezumab for Chronic Low Back Pain (NCT02528253), a randomized phase 3 study. The primary outcome is Change From Baseline in Average Low Back Pain Intensity (LBPI) Score for Tanezumab Versus (Vs) Placebo at Week 16.

2. Clinical Question

The registry describes a Phase 3 randomized study with a parallel-group design and quadruple masking. Its stated primary objective was to evaluate tanezumab 5 mg and 10 mg against placebo for chronic low back pain. The design also included comparisons with tramadol prolonged release, a medication the study summary describes as commonly used for this condition. Secondary questions covered disability, pain at additional follow-up times, patient global assessment, and safety.

The primary outcome is a change score, so the contrast concerns how much the reported score changed from baseline rather than the Week 16 score alone. A more negative change indicates a larger reduction in the pain score, because higher scores represent more pain. The registry’s posted analyses report contrasts for each tanezumab dose versus placebo. The interpretation should distinguish those sponsor-reported adjusted analyses from the simple recalculations, which use posted summary statistics and do not reproduce the sponsor’s full model.

3. Trial Design

This was a randomized, quadruple-masked, parallel-group Phase 3 study of tanezumab in adults with chronic low back pain. The registry identifies Pfizer as the sponsor and describes the study as multicenter. Enrollment was 1832. The study status is completed; the recorded start date is 2015-08-18 and the primary completion date is 2017-10-17.

The intervention list includes tanezumab 5 mg and 10 mg by subcutaneous injection, placebo-to-Week-16 groups followed by tanezumab, and oral tramadol prolonged release. The summary describes tanezumab as administered seven times at eight-week intervals over 56 weeks. The posted primary efficacy outcome compares pooled tanezumab dose groups with placebo at Week 16. Tramadol serves as an active control in other comparisons, including a Week 16 pain-intensity question.

The registry describes the primary analysis population as randomized participants who received at least one dose of subcutaneous study medication. It also describes a prespecified approach for efficacy data up to Week 16 that grouped participants who received placebo from Day 1 and then received tanezumab 5 mg or 10 mg at Week 16. The text of that population description is incomplete in the registry record, so further details should not be inferred from it.

The registry record does not report the original sample-size justification or the full protocol-level statistical plan. Details such as the planned allocation ratio, covariate set for the ANCOVA, handling of missing outcome values, interim analyses, and multiplicity procedure are not reported in the registry record. The planning illustrations associated with the posted results are not a substitute for 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: Treatment period was up to Week 56. Safety follow up period started at Week 64, thus Weeks 64 and 80 time points were during safety follow up period. Percentage (%) reduction in low back pain intensity (LBPI) and participants global assessment (PGA) 2-point reduction are efficacy measures and not applicable during safety follow up.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: ITT population:randomized participants who received at least 1 dose of SC study medication(either tanezumab or matching placebo).Pre-specified intent of study for efficacy data up to W16 was to analyze participants who received placebo from Day1 and then received tanezumab 5/10 mg at W16,together,in placebo arm.Data has been reported per four arms.

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: The safety population was defined as all participants treated with tanezumab or placebo SC.

6. Primary Endpoint

The registry lists the primary outcome as Change From Baseline in Average Low Back Pain Intensity (LBPI) Score for Tanezumab Versus (Vs) Placebo at Week 16, measured over Baseline, Week 16.

The primary endpoint is change from baseline in average low back pain intensity at Week 16. Pain was assessed using an 11-point numeric rating scale captured through interactive response technology. Participants rated their average low back pain during the past 24 hours from 0, no pain, to 10, worst possible pain. Higher scores indicate higher pain. The posted arm summaries are least-squares means with standard errors.

A change-from-baseline endpoint focuses on the shift in reported pain over the study period. For this scale, a negative change represents a lower score at follow-up than at baseline; a more negative value therefore represents a larger reduction in the measured score. Comparing changes between randomized groups helps describe how the groups differed over the same follow-up interval. It does not, by itself, identify why an individual participant’s score changed or establish that every participant had the same experience.

Average pain intensity is a direct measure of the symptom that the study set out to assess. It is used alongside other outcomes, including disability and patient global assessment, because those measures address related but distinct aspects of chronic low back pain. The Week 16 result is the primary endpoint in the registry record; longer-term pain outcomes are listed separately as secondary outcomes. The registry record does not report a threshold defining a clinically important change on this scale, so statistical estimates alone do not establish the practical importance of a difference.

7. Results Posted in the Registry

• Pooled Tanezumab 5 mg: mean -2.98 versus -2.68 on control; difference -0.30 (95% CI -0.70 to 0.10), standardized effect d = -0.10; z-test p = 0.144.
• Pooled Tanezumab 10 mg: mean -3.08 versus -2.68 on control; difference -0.40 (95% CI -0.80 to 0.00), standardized effect d = -0.14; z-test p = 0.051.
• Tramadol: mean -2.81 versus -2.68 on control; difference -0.13 (95% CI -0.51 to 0.25), standardized effect d = -0.04; z-test p = 0.499.

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 in Roland Morris Disability Questionnaire (RMDQ) at Week 16 for Tanezumab Versus (Vs) Placebo

• 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:
  • Pooled Tanezumab 5 mg: mean -6.27 versus -4.95 on control; difference -1.32 (95% CI -2.30 to -0.34), standardized effect d = -0.18; z-test p = 0.009.
  • Pooled Tanezumab 10 mg: mean -6.69 versus -4.95 on control; difference -1.74 (95% CI -2.72 to -0.76), standardized effect d = -0.24; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Pooled Tanezumab 5 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Pooled Tanezumab 10 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Change From Baseline in Average Low Back Pain Intensity (LBPI) Score for Tanezumab Versus (Vs) Tramadol at Week 16

• 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:
  • Pooled Tanezumab 5 mg: mean -2.98 versus -2.68 on control; difference -0.30 (95% CI -0.70 to 0.10), standardized effect d = -0.10; z-test p = 0.144.
  • Pooled Tanezumab 10 mg: mean -3.08 versus -2.68 on control; difference -0.40 (95% CI -0.80 to 0.00), standardized effect d = -0.14; z-test p = 0.051.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Pooled Tanezumab 5 mg and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Pooled Tanezumab 10 mg and the control. The interval gives the range of effects that remain compatible with the data.

Number of Participants Who Withdrew Due to Lack of Efficacy

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Placebo Followed by Tanezumab 10 mg: 20.1% (41/204) versus 12.4% (25/202) on control; risk difference 7.7 percentage points (95% CI 0.6 to 14.9), risk ratio 1.62 (1.03–2.57); unadjusted two-proportion z-test p = 0.035.
  • Pooled Tanezumab 5 mg: 10.1% (41/407) versus 12.4% (25/202) on control; risk difference -2.3 percentage points (95% CI -7.7 to 3.1), risk ratio 0.81 (0.51–1.30); unadjusted two-proportion z-test p = 0.389.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the proportion with the event is statistically higher for Placebo Followed by Tanezumab 10 mg than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Pooled Tanezumab 5 mg and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes broaden the study beyond Week 16 pain intensity. They include disability measured by the Roland Morris Disability Questionnaire, pain comparisons with tramadol, pain intensity at later time points, and patient global assessment. The registry lists outcomes extending to Weeks 56, 64, and 80 for some measures, but a listed follow-up time does not itself establish that complete results or participant follow-up were available at every time point.

Among the secondary results with sponsor analyses in the record, the Week 16 disability comparisons with placebo have negative LS mean differences, and their reported confidence intervals exclude zero. By contrast, the reported Week 16 pain comparisons of tanezumab with tramadol have confidence intervals that include zero. The registry also includes later outcomes and withdrawals due to lack of efficacy, but the posted information in this record does not provide a complete statistical account of all 65 secondary outcomes. The entire pattern across every secondary endpoint therefore cannot be assessed from the available results.

These outcomes provide context for the primary pain result but answer distinct questions. Disability, patient global assessment, and pain intensity are related measures, not interchangeable versions of the same endpoint. Results at later time points also address a different duration of follow-up from the primary Week 16 comparison. They should not be used to substitute for the primary analysis or to imply that one measure confirms another.

Unadjusted secondary results carry less weight when many outcomes, treatment contrasts, and time points are examined. Repeated testing can produce apparently notable results by chance, and the registry record does not report a multiplicity adjustment or a complete testing hierarchy. The secondary findings are therefore best read as supportive descriptive evidence, with attention to their estimates and intervals and without treating isolated p-values as definitive.

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

• Change From Baseline in Roland Morris Disability Questionnaire (RMDQ) at Week 16 for Tanezumab Versus (Vs) Placebo
• Change From Baseline in Average Low Back Pain Intensity (LBPI) Score for Tanezumab Versus (Vs) Tramadol at Week 16
• Change From Baseline in Average Low Back Pain Intensity (LBPI) Score at Weeks 2, 4, 8, 12, 24, 32, 40, 48 and 56
• Change From Baseline in Daily Average Low Back Pain Intensity (LBPI) Score at Week 64
• Change From Baseline in Roland Morris Disability Questionnaire (RMDQ) Total Score at Weeks 2, 4, 8, 16 (for Tanezumab vs Tramadol) 24, 32, 40, 48 and 56
• Change From Baseline in Roland Morris Disability Questionnaire (RMDQ) Score at Weeks 64 and 80: Observed Data
• Change From Baseline in Patient's Global Assessment (PGA) of Low Back Pain at Weeks 2, 4, 8, 16, 24, 32, 40, 48 and 56
• Change From Baseline in Patient's Global Assessment (PGA) of Low Back Pain at Week 64: Observed Data

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: each participant’s pain rating is numerical, and the reported arm summaries are least-squares means. A mean summarizes the central value of a group’s observations, while a standard error describes the estimated sampling uncertainty around that mean. Standard errors are not the same as the spread of individual participants’ scores. The posted summaries report standard errors for each arm.

What the ANCOVA comparison represents

The sponsor reports ANCOVA comparisons as least-squares mean differences. ANCOVA is a general statistical method for comparing group outcomes while accounting for selected covariates. In a randomized trial, adjustment can improve precision and address baseline differences in variables included in the model. The registry record identifies ANCOVA as the method but does not report the model’s covariates, exact specification, or other implementation details. The adjusted estimate therefore cannot be reconstructed from the arm means and standard errors alone.

A mean difference is expressed in the original outcome scale. A negative difference for this change score indicates that the intervention group’s mean change was more negative than the comparator’s. The confidence interval gives a range of estimates compatible with the observed data under the model and its assumptions. If the interval includes zero, the data are compatible with no between-group difference as well as with values on either side of zero. A p-value describes how incompatible the data are with a specified null hypothesis under the analysis assumptions; it is not the probability that the null hypothesis is true, nor a measure of clinical importance.

Why the recalculation is not the sponsor model

The recalculated comparisons are simple, unadjusted estimates derived from the posted arm summaries. The standard deviations used in that process are approximate, inferred from the reported standard errors and sample sizes. This approach does not include the sponsor’s ANCOVA adjustment or any covariates not represented by the summaries. Accordingly, recalculated confidence intervals and p-values can differ from sponsor results. In this trial they do differ, and the sponsor’s own analysis takes precedence.

These recalculations can still help readers understand the scale of the arm contrast and its approximate uncertainty using the information reported. They should not be mistaken for a correction of the sponsor’s analysis, an independent reanalysis of participant-level data, or a way to recover undocumented model details. The registry record does not provide participant-level data or enough information to reproduce the ANCOVA.

Interpreting confidence intervals and p-values

A confidence interval conveys precision as well as direction. A narrower interval indicates greater precision under the analysis assumptions; a wider interval indicates greater uncertainty. Whether an interval excludes zero is one way to describe statistical evidence for a difference, but that alone does not establish whether the difference is important in practice. The registry record does not provide a threshold for clinical importance on the pain scale.

The sponsor reports p-values for the primary ANCOVA comparisons and for several other analyses. The record does not report whether the primary comparisons were adjusted for multiple testing or how any planned hierarchy was defined. Therefore, p-values should be read in the context of the number of endpoints, doses, comparators, and time points assessed, rather than as isolated measures of certainty.

Why over-dispersion matters in general

General statistical context: over-dispersion occurs when observed variability exceeds what a model assumes. If a model understates variability, its standard errors can be too small and its confidence intervals too narrow, potentially making evidence appear more precise than it is. The primary endpoint is continuous and the sponsor reports ANCOVA, but the registry record does not report a concern about over-dispersion or enough diagnostic information to assess model fit. This general issue should not be interpreted as a finding about this trial.

Analysis populations and missing observations

The registry describes the primary efficacy population as randomized participants who received at least one dose of study medication. The record also gives a prespecified description for efficacy through Week 16 involving participants assigned placebo from Day 1 who then received tanezumab at Week 16, but the text is incomplete. It does not report the detailed handling of missing Week 16 scores, whether imputation was used, or how departures from assigned treatment were addressed. These details can influence estimates when outcomes are unavailable or participants discontinue, so the reported results should be interpreted with that uncertainty in mind.

For a continuous endpoint, an analysis may use observed outcomes, model-based adjustments, or imputation strategies, each relying on assumptions about missingness. Without the registry’s full method, it is not possible to determine which assumptions underlie the primary estimate. The registry record also does not report interim analyses or the complete statistical analysis plan. These omissions limit assessment of whether the posted analyses match all prespecified procedures.

11. Confidence Intervals and Precision

The sponsor’s 95% confidence interval for the Week 16 comparison of tanezumab 5 mg with placebo runs from -0.66 to 0.07 around an LS mean difference of -0.30. Because it includes zero, the reported data do not sharply distinguish a small difference in one direction from no difference or a small difference in the other direction. The estimate itself is negative, but the interval shows meaningful uncertainty about its direction and size.

For tanezumab 10 mg versus placebo, the sponsor’s 95% confidence interval runs from -0.76 to -0.04 around an LS mean difference of -0.40. The interval is entirely negative, but its endpoints still span a range of plausible effect estimates under the sponsor’s model. A confidence interval is not a guarantee that the true value lies inside it, nor does it describe the range of individual participant responses. It quantifies uncertainty about a group-level estimate under statistical assumptions.

The recalculated intervals are wider in their upper reach: -0.702 to 0.102 for tanezumab 5 mg, and -0.802 to 0.002 for tanezumab 10 mg. These are approximate, unadjusted intervals based on posted summaries, not replacements for the sponsor’s ANCOVA intervals. The difference between the two sets illustrates that model choice and available detail affect precision estimates.

Precision is not the same as clinical relevance. The registry record does not report a minimally important difference for the pain scale, and no threshold should be inferred from whether an interval crosses zero. The intervals also do not resolve uncertainties about missing-data handling, multiplicity, or the exact ANCOVA specification. They describe statistical uncertainty conditional on the analyses, rather than every source of uncertainty in the evidence.

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 observed effects, not the original trial design calculation. For these continuous and binary outcomes, the listed quantities are participants per arm, not total participants across both arms and not total events. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each scenario plans for an effect equal to the stated fraction of the observed effect.

The column for 80% power gives the number of participants per arm that the illustration estimates would be needed to have an 80% chance of detecting the assumed effect under those planning assumptions, if the effect and model assumptions were correct. The 90% power column represents a higher target chance of detecting that assumed effect and generally calls for a larger group size. These are conditional planning targets, not guarantees about a future study’s result.

Rows that plan for 75% or 50% of the observed effect assume a smaller effect than the full observed estimate. Smaller effects are harder to distinguish from random variation, so they require more participants to reach the same power target. The reported figures are intended to show how sample-size requirements change as the planning effect changes. The original design assumptions and sample-size justification are not reported in the registry record. These planning figures are not observed power.

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 = -0.300; pooled standard deviation σ = 2.925.
2. n = 2 × (1.96 + 0.842)² × 2.925² / -0.300² = 1,492.2, rounded up to 1,493 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 <- -0.3                      # observed difference in means
sigma <- 2.92495917747394                     # 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(2528253)
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 = 1,493 (80% power), 1,998 (90% power)
 75% of observed effect: n per arm = 2,653 (80% power), 3,552 (90% power)
 50% of observed effect: n per arm = 5,969 (80% power), 7,991 (90% power)
power.t.test: 1493.2 
Simulated power at n = 1,493 per arm: 0.803 (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 sponsor’s own reported primary comparisons use ANCOVA and give LS mean differences with 95% confidence intervals and p-values. The simple recalculations use unadjusted contrasts from posted summary statistics, with approximate standard deviations inferred from standard errors. They do not reproduce the sponsor’s adjusted model, and their results differ from the sponsor’s estimates of uncertainty and p-values.

For the 5 mg comparison, the sponsor reports -0.30, a 95% confidence interval from -0.66 to 0.07, and p = 0.112; the recalculation gives -0.3, an interval from -0.702 to 0.102, and p = 0.144. For the 10 mg comparison, the sponsor reports -0.40, a 95% confidence interval from -0.76 to -0.04, and p = 0.028; the recalculation gives -0.4, an interval from -0.802 to 0.002, and p = 0.051. The displayed estimates are close in direction and magnitude, but their inferential details are not identical.

Because a discrepancy is present, the sponsor’s own analysis takes precedence when describing the trial’s reported primary results. The recalculation is useful as a transparent, approximate check based on summary information, but it cannot verify or supersede the sponsor’s analysis. The registry record does not include sufficient detail to reproduce the ANCOVA or establish the reason for the discrepancy.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates multiple hypotheses, such as several doses, outcomes, comparators, or follow-up times. If each test is considered on its own without adjustment, the chance of at least one false-positive result across the full set may increase. Common approaches include prespecified testing hierarchies or adjustments to the significance threshold, but the registry record does not report whether either approach was used for the primary or secondary analyses.

The study includes a primary pain outcome and many secondary measures. The record reports 66 outcome measures and 737 analyses, but it does not describe how these analyses were grouped into confirmatory families or which results, if any, were protected by a formal testing strategy. This limits the interpretation of individual p-values, particularly for secondary outcomes. The sponsor-reported primary p-values should be read as the analyses posted in the registry, not as proof that all multiplicity concerns have been addressed.

Missing-data methods are also not reported in sufficient detail. The registry names the primary analysis population and gives a partial description of a prespecified Week 16 efficacy population, but it does not explain how missing pain scores were handled, whether imputation was performed, or what assumptions were made about participants with incomplete follow-up. The record also does not report sensitivity analyses designed to test alternative missing-data assumptions.

The primary outcome is described for an intent-to-treat population of randomized participants who received at least one dose. That description is not identical to a full account of how every randomized participant contributed to each analysis. The registry does not provide enough information to assess exclusions, treatment discontinuations, adherence to assigned groups, or the exact analysis population for every endpoint. The full protocol and statistical analysis plan are not reported in the registry record, so these aspects remain uncertain.

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: rapidly progressive osteoarthritis (1.0%), osteoarthritis (0.6%), rotator cuff syndrome (0.6%), cardiac failure (0.5%), road traffic accident (0.5%).

The registry reports serious adverse events by safety group, with the number at risk and the number affected. For placebo at Week 16, 7 participants were affected among 215 at risk. In the tanezumab 5 mg safety evaluation group, 21 were affected among 506 at risk; in the 10 mg group, 37 were affected among 502 at risk. The tramadol group reports 25 affected among 602 at risk. These are group counts, not adjusted comparisons, and the groups have different denominators.

The registry also reports deaths: 1 in the placebo group, 2 in the tanezumab 5 mg group, 2 in the tanezumab 10 mg group, and 1 in the tramadol group. The listed serious adverse event terms include rapidly progressive osteoarthritis at 1.0%, osteoarthritis at 0.6%, rotator cuff syndrome at 0.6%, cardiac failure at 0.5%, road traffic accident at 0.5%, and deafness neurosensory at 0.5%.

These counts do not, by themselves, establish comparative risk or causation. Interpretation depends on definitions, exposure and follow-up time, event ascertainment, and statistical comparisons, details that are not fully reported in the registry record. Serious adverse event counts should be read as a safety summary rather than a complete account of all harms or a causal conclusion.

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

The primary endpoint measures change in average low back pain intensity from baseline to Week 16. The sponsor-reported analysis gives a negative estimated difference for each tanezumab dose compared with placebo. The 5 mg interval includes zero, while the 10 mg interval is entirely negative. These statements describe the direction and statistical uncertainty of the posted group-level results; they do not establish what an individual participant experienced.

The estimates do not, on their own, tell readers whether a difference is large enough to matter to patients. The registry record does not report a clinically important difference threshold for the pain scale. Nor can the results be extended automatically to other populations, treatment durations, outcomes, or settings. The primary comparison is at Week 16, while the study also lists longer-term outcomes that must be interpreted separately.

The registry’s secondary results add information about disability, active-control comparisons, later assessments, and safety. They are not interchangeable with the primary endpoint, and the full results for all listed secondary outcomes are not reported in the record. Without a reported multiplicity strategy, secondary p-values deserve cautious interpretation.

Safety counts describe reported events in the listed groups but do not establish that an event was caused by a study treatment or permit a definitive comparison of risk. The registry record also leaves important analysis details unreported, including missing-data handling, model specification, and the original sample-size justification. Taken together, the posted statistics are evidence about measured outcomes under the reported analyses, with limitations in precision and reporting. They do not support medical advice or a universal conclusion about outcomes for individuals.

18. Why This Trial Matters Statistically

This trial raises several important statistical issues for interpreting clinical results. First, the sponsor’s primary comparisons use ANCOVA, while a simple recalculation from posted summaries produces different confidence intervals and p-values. This illustrates why an adjusted trial analysis cannot generally be recovered from arm means and standard errors alone. The distinction matters: the sponsor’s analysis is the trial’s reported result, while a summary-statistic recalculation is an approximate check with a different basis.

Second, the study addresses more than one dose, comparator, outcome, and follow-up time. A primary result can be reported alongside a large collection of secondary analyses, but interpretation depends on whether the testing plan controlled multiplicity. That plan is not reported in the registry record. Without it, individual secondary p-values should not be treated as though each came from a single isolated test.

Third, precision and practical importance are separate questions. Confidence intervals show uncertainty in group-level estimates, but the registry does not report a threshold for a clinically important change in pain intensity. Statistical significance alone cannot supply that missing context.

Finally, interpretation depends on details that registry summaries often leave incomplete, including the covariates in the ANCOVA and procedures for missing outcomes. The study’s sample-size illustrations are based on observed effects and stated planning assumptions, not the original trial design calculation. These features make the trial a useful example of why reported estimates, model details, multiplicity, missing-data methods, and design assumptions should be considered together rather than relying on a single p-value.

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 gives an incomplete description of the primary analysis population: the text ends before the population definition is complete. It identifies ANCOVA as the sponsor’s method but does not report the covariates or enough model detail to reproduce the adjusted estimates. This makes the difference between sponsor analyses and summary-statistic recalculations impossible to resolve from the record alone.

Missing-data procedures are not reported, including the approach to unavailable Week 16 outcomes and the assumptions used for later follow-up. This is relevant because outcomes are listed at multiple visits and some are explicitly described as observed data. The record does not establish how incomplete follow-up affected each analysis.

The registry does not report a multiplicity procedure despite many outcome measures and analyses. As a result, the evidential status of secondary p-values cannot be fully assessed. It also does not report a clinically important threshold for the pain score, limiting interpretation of the practical size of the estimates.

The safety summaries provide counts and denominators but do not give a complete account of exposure-adjusted rates, causal assessment, or all safety-analysis methods. Group denominators also differ. The record does not report the original sample-size justification, interim-analysis plan, or complete statistical analysis plan. These limitations apply to interpretation of the registry results and do not establish that any particular method was or was not used outside what the record reports.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: A Phase 3 Study of Tanezumab for Chronic Low Back Pain — NCT02528253.
• Registry record last updated 2020-02-21; status last verified 2020-02.
• Results first posted 2020-02-21.
• 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

NCT02528253 is a completed randomized Phase 3 study of tanezumab for chronic low back pain, sponsored by Pfizer. It enrolled 1832 adults and compared tanezumab 5 mg and 10 mg with placebo, with tramadol prolonged release included as an active control. The primary outcome was change in average low back pain intensity from baseline to Week 16.

The sponsor-reported ANCOVA estimate was -0.30 for tanezumab 5 mg versus placebo and -0.40 for tanezumab 10 mg versus placebo. Their reported 95% confidence intervals were -0.66 to 0.07 and -0.76 to -0.04, respectively. Simple unadjusted recalculations differ from those sponsor analyses; because a discrepancy is present, the sponsor’s analyses take precedence.

The registry also lists disability, pain, global assessment, and safety outcomes. Important details about multiplicity, missing-data handling, model specification, and the original design assumptions are not reported in the registry record.