Minocycline Study in Pancreatic Cancer Patients: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Minocycline Study in Pancreatic Cancer Patients (NCT01693523), a randomized phase 2 study. The primary outcome is Average Area Under the Curve (AUC) of Apriori Selected MDASI Symptoms.

2. Clinical Question

3. Trial Design

This was a phase 2, randomized, double-blind, parallel study comparing minocycline with placebo for reducing symptom burden among pancreatic cancer patients. M.D. Anderson Cancer Center is listed as the sponsor. The record lists a study start of 2013-01-28, primary completion on 2018-09-19, and status as completed.

The registry lists enrollment of 80, while its summary says 44 patients were randomized. The posted primary-outcome population description says 30 (68% were evaluable for the primary efficacy analysis). The record does not explain the relationship among the enrollment figure, the randomized population, and the evaluable population. Outcome summaries also include an observational group, but its role in relation to the randomized comparison is not reported in the registry record.

The minocycline regimen was 100 mg by mouth two times a day (200 mg/day); the placebo group received matching capsules twice a day. The initial dose was to begin on the first day of the run-in phase or chemotherapy. Questionnaires were scheduled at baseline, 1 time each week during drug/placebo administration, and at the end-of-study visit. The registry lists questionnaires as an intervention as well as the two medication groups.

The study summary says the purpose was to learn whether minocycline could reduce chemotherapy side effects. The posted primary measure concerns a symptom-burden subscale rather than a single symptom. Detailed eligibility criteria, the duration of medication administration, allocation procedures, and the original statistical analysis plan are 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: Of the 80 participants enrolled 1 participant were excluded from the study before assignment to groups.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: Of the 44 randomized patients, 30 (68% were evaluable for the primary efficacy analysis).

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 Average Area Under the Curve (AUC) of Apriori Selected MDASI Symptoms, measured over AUC from baseline to 2 weeks.

The primary endpoint is the average AUC for a preselected subscale of five MD Anderson Symptom Inventory (MDASI) items: fatigue, pain, disturbed sleep, lack of appetite, and drowsiness. Each item is rated on a 0 to 10 scale. The registry describes the ratings as capturing symptom severity and interference with daily functioning, with higher ratings indicating more severe symptoms or interference. The subscale is the average of the five selected items, and the registry states that lower AUC values represent better outcomes.

An AUC combines symptom ratings across time, so it reflects both the level of reported symptoms and their duration across the measured period. It can therefore represent accumulated symptom burden rather than a rating at a single assessment. The outcome description gives an AUC range from 0 to 700 for the 70-day study period and explains that the placebo mean of 200.8 can be expressed as 2.87 on the 0 to 10 scale over that period.

There is an important timing ambiguity in the record: the primary outcome timeframe is listed as AUC from baseline to 2 weeks, while the description refers to a 70-day area during a 10 week study. The registry record does not clarify which timing description governs the posted values. This makes the endpoint’s intended measurement window less clear than its symptom content and direction of interpretation.

7. Results Posted in the Registry

• Minocycline: mean 51.50 versus 30.60 on control; difference 20.90 (95% CI 5.50 to 36.30), standardized effect d = 0.79; z-test p = 0.008.
• Observational: mean 36.20 versus 30.60 on control; difference 5.60 (95% CI -4.95 to 16.15), standardized effect d = 0.29; z-test p = 0.298.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary measure is continuous: it summarizes symptom burden as an AUC. For a continuous outcome, a group mean describes the average observed value in that group, while the standard deviation describes how spread out individual values are around the mean. A larger standard deviation indicates more variability among measured values. The posted means are 51.5 for minocycline and 30.6 for placebo; their posted standard deviations are 31.4 and 19.9, respectively. These summaries describe the participants included in the corresponding outcome summaries, not every possible patient with pancreatic cancer.

An AUC is a way to combine repeated measurements over a defined time interval. General explanation: if measurements are repeatedly collected, the area reflects the accumulated level over time, so it can capture both symptom intensity and how long symptoms are experienced. The registry describes a 0 to 10 scale for each of the selected symptoms and says the subscale averages five items. It describes the AUC as ranging from 0 to 700 for a 70-day period. However, the outcome timeframe field says AUC from baseline to 2 weeks. Because these descriptions do not match, the intended period for the posted endpoint is not clear from the registry record.

The mean difference compares the group means on the outcome scale. For the randomized comparison, the registry reports a mean difference of 20.9, with minocycline higher than placebo. Given the stated direction of the measure, a higher AUC corresponds to worse measured symptom burden. The numerical difference is an unadjusted summary and does not by itself account for other characteristics, missing observations, or the way participants entered the analysis. The record does not report an adjusted primary estimate.

A confidence interval expresses uncertainty around an estimated effect. General explanation: under the assumptions used to construct it, a confidence interval procedure would produce intervals that contain the underlying effect at its stated confidence level in the long run. It is not a range containing a known fraction of individual participant outcomes, and it is not a probability statement about the specific underlying effect. The posted interval for the randomized mean difference extends from 5.501 to 36.3. Its positive values are consistent with a higher mean AUC for minocycline, but its width also shows that the estimate is not exact.

The unadjusted p-value for that comparison is 0.008. General explanation: a p-value measures how unusual data at least as incompatible with the null hypothesis would be if the null hypothesis and the statistical model were correct. It is not a measure of effect size, clinical importance, or the probability that the result is false. It also does not address whether the model’s assumptions or the analysis population are appropriate. The record does not report a statistical plan that would show whether this was the prespecified primary test or whether other analyses were considered.

The posted standardized effect-size value is not needed to interpret the outcome in its original units. General explanation: standardizing a mean difference relates it to variability in the outcome, which can make results from different scales easier to compare, but it does not change the underlying measurement or remove uncertainty. No standardized effect-size interpretation can replace the scale-specific meaning of the AUC and the registry's statement that lower values are better.

The observational comparison is also a difference in means, but it is not the randomized minocycline-versus-placebo contrast. The registry reports a mean difference of 5.6, a confidence interval from -4.947 to 16.15, and an unadjusted p-value of 0.298. General explanation: an interval containing the null value is compatible with effects in either direction under the analysis assumptions. It does not prove the groups are equivalent. Since the record does not explain the observational group's selection or role, the comparison has additional interpretive limits beyond those of the randomized estimate.

The registry lists 0 analyses and 2 outcome measures, while also listing 1 secondary outcome. It does not describe the statistical model, covariate adjustment, handling of repeated measurements, missing-data assumptions, or multiplicity procedure. These omissions prevent a full assessment of how the posted estimates were calculated and whether the reported uncertainty incorporates all relevant sources of variation. General explanation: repeated measurements and incomplete outcome data can require analytic choices that affect estimates and their precision; without those details, interpretation should remain tied to the posted summaries.

10. Confidence Intervals and Precision

The randomized mean difference is 20.9, with a confidence interval from 5.501 to 36.3. The interval is entirely above the null value for a mean difference, and its direction agrees with the posted group means: the minocycline mean AUC is higher than the placebo mean AUC. Under the registry’s stated outcome interpretation, higher values indicate worse symptom burden. The interval does not identify a single precise value, however; it allows a range of positive differences.

Confidence intervals describe uncertainty in an estimated group contrast under the method and assumptions used to calculate them. They do not show the range of individual outcomes, and they do not guarantee that the true effect falls within this particular interval. The registry record does not state the confidence level or the method used to construct the interval, nor does it report whether the analysis adjusted for other factors. Those details matter when judging the interval’s coverage and scope.

For the observational comparison, the mean difference is 5.6 and the confidence interval extends from -4.947 to 16.15. This interval crosses the null value, allowing values on either side of no mean difference. It is therefore not a precise basis for distinguishing the observational group from placebo on this outcome. That result should not be interpreted as evidence of equivalence.

Precision also depends on the analysis population and outcome timing. The record reports 30 (68% were evaluable for the primary efficacy analysis), but the group summaries list 23 for minocycline and 21 for placebo; the relationship among those counts is not explained. Further, the endpoint timing is described as baseline to 2 weeks in one field and as a 70-day AUC elsewhere. The registry record does not report how these issues were handled in the interval calculations.

11. Planning a Similar Trial: Sample Size

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

The planning figures are participants per arm for the two-arm comparisons, not total events. They are planning illustrations based on the observed effect, not the original trial’s design requirements. The stated assumptions are a two-sided significance level 0.05, equal allocation between the two arms, a normal approximation, and an effect to detect set to the stated fraction of the observed effect. The trial’s own original design assumptions and sample-size justification are not reported in the registry record.

For the minocycline-versus-placebo comparison, planning for 100% of the observed effect gives 26 participants per arm for 80% power and 34 participants per arm for 90% power. Planning for 75% of the observed effect gives 46 participants per arm for 80% power and 61 participants per arm for 90% power. Planning for 50% of the observed effect gives 102 participants per arm for 80% power and 136 participants per arm for 90% power.

The 80% power column describes the planned chance of detecting the specified effect under the calculation assumptions if that effect were real; the 90% power column uses a higher target chance of detection. These are prospective planning quantities, not observed power. Smaller effects need more participants to detect under the same assumptions, which is why the figures rise as the assumed fraction of the observed effect gets smaller.

The record also gives planning figures for the observational comparison. Those figures are based on its own observed mean difference, not the randomized treatment contrast, and should not be used as if they represented the same design or question. Planning calculations do not establish that a study will detect a particular effect in practice; their usefulness depends on whether the assumed effect, variability, allocation, and analysis match the study being planned.

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 = 20.900; pooled standard deviation σ = 26.552.
2. n = 2 × (1.96 + 0.842)² × 26.552² / 20.900² = 25.3, rounded up to 26 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 <- 20.9                      # observed difference in means
sigma <- 26.5524279223469                     # 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(1693523)
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 = 26 (80% power), 34 (90% power)
 75% of observed effect: n per arm = 46 (80% power), 61 (90% power)
 50% of observed effect: n per arm = 102 (80% power), 136 (90% power)
power.t.test: 26.3 
Simulated power at n = 26 per arm: 0.792 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry record does not report whether the primary outcome was tested according to a prespecified statistical analysis plan or whether the p-value was adjusted for multiple comparisons. It lists 2 outcome measures, 1 secondary outcome, and 0 analyses in the posted analysis count. The listed secondary outcome concerns the relationship between dynamic changes in inflammation biomarkers and symptom outcomes, with adjustment for grouping variables, evidence of infection, ECOG PS, age, and gender. The record does not provide estimates or results for that secondary question.

Multiplicity matters when several outcomes, time points, or comparisons are examined. General explanation: the more statistical tests considered, the greater the chance that at least one result may appear notable even when there is no underlying effect, unless the analysis plan accounts for that testing context. The registry does not say whether there were additional analyses beyond those posted or how the primary comparison was prioritized.

Missing-data handling is not reported in the registry record. The record says 30 (68% were evaluable for the primary efficacy analysis) among 44 randomized patients, while the outcome summaries list 23 for minocycline and 21 for placebo. It does not explain the relationship between these figures, why some patients were not evaluable, or whether missing values were imputed, excluded, or addressed through another method. The posted group counts alone do not establish the reasons for missingness.

The analysis population is also not fully described. The record does not state whether the primary comparison followed an intention-to-treat principle, included only participants with observed outcome data, or used another definition. It does not describe how participants were handled if they did not follow their assigned regimen or lacked measurements at some visits. These omissions limit assessment of potential bias from exclusions and incomplete follow-up.

The registry’s primary endpoint timing is described both as AUC from baseline to 2 weeks and as a 70-day AUC. The record does not clarify which period was analyzed or how repeated questionnaire measurements contributed to the summary. Without that detail, the endpoint definition, missing-data approach, and analysis population cannot be fully connected to the posted mean differences and confidence intervals.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: septic shock (8.7%), sudden cardiopulmonare (4.3%).

The registry reports serious adverse events for the minocycline and placebo groups separately. For minocycline, 2 participants are listed as at risk and affected for serious adverse events, and 3 are listed as at risk and affected for other adverse events. For placebo, 0 participants are listed as affected for serious adverse events, and 3 are listed as affected for other adverse events. The record lists 0 deaths in each group.

The serious-event terms listed include “Septic shock” at 8.7% and “Sudden cardiopulmonare” at 4.3%. The registry record does not state the denominator or calculation basis for these term percentages, and the term wording is retained as posted. These percentages should not be interpreted as rates in all enrolled patients without a stated denominator.

Adverse-event counts are descriptive safety summaries. They do not, on their own, establish that an event was caused by the study medication, and they do not provide a complete comparison of risks. The record does not report event timing, severity grades beyond serious-event classification, exposure duration, or a statistical comparison of safety outcomes. The counts should therefore be read as recorded events in the listed groups, not as evidence of causation or as a comprehensive assessment of safety.

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

The posted randomized comparison describes a higher mean AUC for minocycline than for placebo: 51.5 versus 30.6. Since lower values represent better outcomes according to the registry, the observed group means are in the direction of greater symptom burden for minocycline. The reported mean difference is 20.9, and its confidence interval runs from 5.501 to 36.3. The unadjusted p-value is 0.008. These statistics describe the posted analysis; they do not, by themselves, establish that the medication caused the difference.

The result concerns an average across five selected symptoms and across a stated time period, not every symptom separately or every moment of treatment. The endpoint description says the selected items were fatigue, pain, disturbed sleep, lack of appetite, and drowsiness. The registry gives conflicting descriptions of the measurement window: baseline to 2 weeks in the timeframe field and a 70-day AUC in the outcome description. The record does not resolve that discrepancy.

The summary also does not establish whether the difference would persist outside the recorded period, how individual patients’ experiences varied, or whether changes in any specific symptom drove the group average. The posted standard deviations show variation within both groups, but the registry does not provide individual-level distributions or symptom-by-symptom estimates. The observational group’s comparison with placebo is not a substitute for the randomized contrast and is not described in enough detail to support causal interpretation.

The analysis population and missing-data methods are not fully described. Although the record reports that 30 (68% were evaluable for the primary efficacy analysis) among 44 randomized patients, the group summaries list 23 for minocycline and 21 for placebo. The relationship among these populations is not explained. The statistical analysis plan, any adjustment for multiple testing, and adjusted primary estimates are also not reported in the registry record.

Accordingly, the result is a reported difference in a symptom-burden summary, with the minocycline mean higher than the placebo mean and with the stated uncertainty. It should be interpreted in the context of the unclear time window, population definitions, and unreported analysis details. It does not support a broader conclusion about treatment effects beyond the specific posted outcome summaries.

16. Why This Trial Matters Statistically

This trial raises several statistical interpretation issues that commonly arise in clinical research. First, the primary endpoint compresses repeated symptom ratings into an AUC. That approach can represent accumulated symptom burden, but the meaning depends on a clearly defined measurement window and consistent follow-up. Here, the registry describes the timeframe as baseline to 2 weeks in one place and a 70-day area in another. Without clarification, readers cannot confidently connect the estimate to a particular duration.

Second, the group means and their difference need to be read alongside the direction of the scale. A higher AUC is described as worse, so the posted minocycline mean is not a favorable direction relative to placebo. A p-value alone cannot convey that direction, the size of the difference, or its precision. The mean difference and confidence interval are more informative for those aspects, while still depending on the analysis and population used.

Third, the record contains distinct population descriptions: enrollment of 80, 44 randomized, and 30 (68% were evaluable for the primary efficacy analysis), alongside group counts in the outcome summaries. The record does not explain the relationship among these figures. This matters because the population included in an analysis affects what the estimate describes and how missing outcomes may influence the result.

Finally, the observational group appears alongside the randomized groups, but its role is not reported. Keeping the randomized comparison separate from that additional summary is important: a group label alone does not establish comparable assignment or eliminate selection differences. The registry also does not report multiplicity handling, missing-data methods, or a detailed analysis plan. These gaps show why statistical interpretation depends not only on a numerical estimate, but also on transparent definitions of the endpoint, analysis population, and methods.

17. Limitations

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

The primary outcome timing is not described consistently. The timeframe field states AUC from baseline to 2 weeks, while the outcome description refers to the average of the 70-day area during a 10 week study. The registry record does not clarify which period corresponds to the posted group means. This limits interpretation of the time span represented by the primary result.

Population counts also require caution. Enrollment is listed as 80, the study summary reports 44 randomized patients, and the primary efficacy population is described as 30 (68% were evaluable for the primary efficacy analysis). The posted outcome summaries list 23 for minocycline and 21 for placebo. The registry record does not explain how these counts relate to one another or why they differ.

An observational group is included in the outcome summaries, but its source, selection, and analytic purpose are not reported. Its comparison with placebo should not be interpreted as a randomized treatment contrast. The registry record also does not provide adjusted estimates, a detailed statistical analysis plan, missing-data methods, or a multiplicity procedure.

The available endpoint is an average across five selected symptoms. Separate results for each symptom are not reported in the record, nor are individual-level outcome distributions. The listed adverse-event summaries do not give a full account of event timing or exposure, and the denominator for the serious-event term percentages is not stated. These limitations do not negate the posted summaries, but they restrict the conclusions that can be drawn from them.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Minocycline Study in Pancreatic Cancer Patients — NCT01693523.
• Registry record last updated 2020-03-09; status last verified 2020-02.
• Results first posted 2020-02-19.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

The record describes a completed phase 2, randomized, double-blind, parallel study of minocycline versus placebo for symptom burden in pancreatic cancer patients. M.D. Anderson Cancer Center is listed as sponsor. The primary measure is an AUC summary of five MDASI symptoms, with lower values defined as better outcomes.

Posted means are 51.5 for minocycline and 30.6 for placebo, with a mean difference of 20.9, a confidence interval from 5.501 to 36.3, and an unadjusted p-value of 0.008. The record also lists an observational group. Important details—including the primary-outcome timing, the relationship among the reported analysis populations, missing-data handling, and the statistical analysis plan—are not reported or are unclear in the registry record.