This page provides an independent statistical analysis and educational interpretation of publicly reported results. ClinicalTrials.gov provides the official trial registry record.
1. Trial at a Glance
LUCENT 2 was a phase 3 maintenance study evaluating mirikizumab in participants with moderately to severely active ulcerative colitis. The registry describes a double-masked, parallel study with non-randomized allocation, 12 arms, and enrollment of 1328 participants. The primary efficacy comparison reported in the registry evaluates clinical remission at week 40 among mirikizumab induction responders.
| Feature | LUCENT 2 |
|---|---|
| Trial name | LUCENT 2 |
| ClinicalTrials.gov identifier | NCT03524092 |
| Phase | Phase 3 |
| Status | Completed |
| Condition | Ulcerative Colitis |
| Therapeutic area | Gastroenterology & Hepatology |
| Allocation | Non-randomized |
| Design model | Parallel |
| Masking | Double |
| Primary purpose | Treatment |
| Enrollment | 1328 |
| Number of arms | 12 |
| Interventions | Mirikizumab SC; Mirikizumab IV; Placebo SC |
| Lead sponsor | Eli Lilly and Company |
| Sponsor type | Industry |
| Start | 2018-10-19 |
| Primary completion | 2021-11-03 |
2. Clinical Question
The central statistical question was whether maintenance treatment with mirikizumab produced a higher proportion of clinical remission at week 40 among mirikizumab induction responders when comparing the registry-defined maintenance groups Miri IR - PBO SC and Miri IR - 200 mg Miri SC.
Population
Participants with moderately to severely active ulcerative colitis enrolled in the LUCENT 2 maintenance study. The primary endpoint specifically concerns mirikizumab induction responders.
Intervention
Mirikizumab, with the registry identifying both subcutaneous and intravenous mirikizumab interventions across the overall study.
Comparator
The primary analysis compares Maintenance Period: Miri IR - PBO SC with Maintenance Period: Miri IR - 200 mg Miri SC.
Primary question
What is the difference in the percentage of participants achieving clinical remission at week 40 between the two primary-analysis maintenance groups?
3. Trial Design
Miri IR - PBO SC
- Mirikizumab induction responder population
- Placebo SC maintenance designation in the registry analysis
- Compared with the 200 mg mirikizumab maintenance group for the reported primary endpoint
Miri IR - 200 mg Miri SC
- Mirikizumab induction responder population
- 200 mg mirikizumab SC maintenance designation in the registry analysis
- Reference treatment group for the reported risk-difference comparison
The overall intervention list also includes mirikizumab IV and placebo SC, and the registry reports 12 study arms. The posted statistical analyses focus on specific maintenance-period comparisons rather than presenting all 12 arms as separate confirmatory comparisons.
4. Analysis Populations
The primary endpoint analysis uses a modified intention-to-treat population. The registry defines this population explicitly as all randomized participants who received at least one dose of mirikizumab and had a correctly measured Modified Mayo Score at baseline. Participants were analyzed according to the treatment arm to which they were randomized, regardless of treatment actually received.
| Population | Registry definition / role |
|---|---|
| Modified intention-to-treat | All randomized participants receiving at least one dose of mirikizumab with a baseline Modified Mayo Score; analyzed according to randomized arm. |
| Primary endpoint mITT | Participants with the baseline Modified Mayo Score required for the clinical-remission definition. |
| IBDQ analysis population | mITT participants with baseline and at least one post-baseline IBDQ measurement. |
| Fecal calprotectin analysis population | mITT participants with baseline and at least one post-baseline fecal calprotectin measurement. |
| Urgency NRS analysis population | mITT participants with baseline and at least one post-baseline urgency NRS measurement. |
The statistical-analysis records also state that, under Global SAP v4 and the China ME2 Addendum, ME2 participants were included in a combined China population. China analyses were prespecified as descriptive and exploratory, without multiplicity control or p-values, and ME2 arms were therefore not reported separately in the confirmatory primary and secondary analyses.
5. Primary Endpoint
| Endpoint | Time frame | Type | Primary analysis |
|---|---|---|---|
| Percentage of Participants in Clinical Remission at Week 40 (Mirikizumab Induction Responders) | Week 40 | Binary | Cochran-Mantel-Haenszel test; risk difference |
Clinical remission definition
Clinical remission at week 40 is defined as achieving a 9-point modified Mayo score with rectal bleeding = 0, stool frequency = 0 or 1 with at least a 1-point decrease from baseline, and endoscopy = 0 or 1 excluding friability.
The registry describes the stool-frequency subscore as ranging from 0 for a normal number of stools to 3 for 5 or more stools than normal. The rectal-bleeding subscore ranges from 0 for no blood to 3 for blood only passed. The endoscopy subscore is based on central reading of colonoscopy or sigmoidoscopy and ranges from 0 for normal or inactive disease to 3 for severe disease, spontaneous bleeding, or ulceration.
6. Statistical Methodology
Cochran-Mantel-Haenszel test
The primary and several secondary binary endpoints were analyzed with the Cochran-Mantel-Haenszel test. This approach is useful when a treatment comparison is made for a categorical outcome while accounting for prespecified strata. The registry identifies this method for the clinical-remission endpoint and for the reported binary secondary endpoints.
For a binary endpoint, the risk difference expresses the absolute difference in the probability of the outcome between two groups. A positive value in this comparison corresponds to a higher estimated proportion in the 200 mg mirikizumab SC group.
ANCOVA
ANCOVA was used for the reported changes in IBDQ total score and fecal calprotectin. The effect measure was a least-squares mean difference in final values. The registry notes that these analyses used ANCOVA with modified baseline observation carried forward (mBOCF).
Mixed-effects model
The change from baseline to week 40 in bowel urgency based on the Urgency Numeric Rating Scale was analyzed using a mixed-effects model. Mixed-effects approaches are designed for repeated or longitudinal observations and can account for correlation among measurements from the same participant.
Multiplicity
The statistical-analysis records identify multiplicity adjustment as an analysis concept for the reported primary and secondary analyses. The exact multiplicity procedure and adjusted significance thresholds are not specified in the posted statistical-analysis fields summarized here, so the individual p-values are presented exactly as reported rather than assigned an additional interpretation beyond the registry's stated superiority framework.
China ME2 analyses
The registry states that the China ME2 analyses were prespecified as descriptive and exploratory, with no multiplicity control and no p-values. These analyses were included in a combined China population rather than being reported as separate ME2 arms in the confirmatory analyses.
7. Primary Result: Clinical Remission at Week 40
The primary endpoint compared the percentage of mirikizumab induction responders achieving clinical remission at week 40 between the Maintenance Period: Miri IR - PBO SC group and the Maintenance Period: Miri IR - 200 mg Miri SC group.
Risk difference in clinical remission
95% CI: 15.2–31.2 · P < 0.001
Cochran-Mantel-Haenszel test · two-sided 95% CI · superiority
| Primary endpoint | Comparison | Effect measure | Estimate | 95% CI | P-value |
|---|---|---|---|---|---|
| Clinical remission at Week 40 | Miri IR - PBO SC vs Miri IR - 200 mg Miri SC | Risk difference | 23.2 | 15.2–31.2 | <0.001 |
The reported risk difference of 23.2 means that the estimated percentage achieving clinical remission was 23.2 percentage points higher in the 200 mg mirikizumab SC maintenance group than in the placebo SC maintenance group for this analysis. Because the registry presents the comparison in that group order, the positive estimate is interpreted in favor of the 200 mg mirikizumab SC group.
The estimate does not mean that every participant had a 23.2% improvement, nor does it mean that an individual participant's probability changed by exactly 23.2%. It is an absolute difference between group-level outcome proportions.
The two-sided 95% confidence interval of 15.2 to 31.2 describes statistical uncertainty around the estimated risk difference under the analysis framework. It does not describe the range of individual treatment responses.
The p-value of <0.001 addresses the strength of evidence against the null hypothesis under the specified test. It does not measure the size or clinical importance of the effect. The effect size is described by the risk difference and its confidence interval.
The interpretation also depends on the analysis population and design. The registry defines the endpoint analysis using an mITT population and randomized-arm assignment, while the overall registry allocation field is listed as non-randomized. That design terminology should be kept distinct when considering causal interpretation.
8. Secondary Efficacy Results
The registry contains eight posted secondary statistical analyses in addition to the primary endpoint analysis. Five are binary week-40 efficacy outcomes analyzed with the Cochran-Mantel-Haenszel test and reported as risk differences; three are continuous or longitudinal outcomes analyzed using ANCOVA or a mixed-effects model.
| Secondary endpoint | Method | Effect | Estimate | 95% CI | P-value |
|---|---|---|---|---|---|
| Endoscopic remission at Week 40 | Cochran-Mantel-Haenszel | Risk difference | 28.5 | 20.2–36.8 | <0.001 |
| Histologic remission at Week 40 | Cochran-Mantel-Haenszel | Risk difference | 22.5 | 14.5–30.5 | <0.001 |
| Symptomatic remission at Week 40 | Cochran-Mantel-Haenszel | Risk difference | 30.2 | 21.9–38.6 | <0.001 |
| Endoscopic response at Week 40 | Cochran-Mantel-Haenszel | Risk difference | 31.0 | 22.4–39.6 | <0.001 |
| Clinical response at Week 40 | Cochran-Mantel-Haenszel | Risk difference | 30.6 | 22.3–38.9 | <0.001 |
| IBDQ total score change to Week 40 | ANCOVA with mBOCF | Mean difference | 25.24 | 19.16–31.32 | <0.001 |
| Fecal calprotectin change to Week 40 | ANCOVA with mBOCF | Mean difference | -839.64 | -1323.08 to -356.21 | <0.001 |
| Bowel urgency NRS change to Week 40 | Mixed-effects model | Mean difference | -1.06 | -1.51 to -0.61 | <0.001 |
Endoscopic remission
The reported risk difference for endoscopic remission at week 40 was 28.5, with a two-sided 95% CI of 20.2 to 36.8 and P < 0.001. The analysis used the same primary maintenance-group comparison and the Cochran-Mantel-Haenszel method.
Histologic remission
The reported risk difference for histologic remission was 22.5, with a two-sided 95% CI of 14.5 to 30.5 and P < 0.001. This provides a separate binary measure of disease status from the clinical-remission endpoint.
Symptomatic remission
The reported risk difference for symptomatic remission was 30.2, with a two-sided 95% CI of 21.9 to 38.6 and P < 0.001. The magnitude is an absolute difference in outcome proportions, not a relative risk.
Endoscopic response
The reported risk difference for endoscopic response was 31.0, with a two-sided 95% CI of 22.4 to 39.6 and P < 0.001. The endpoint was analyzed with the Cochran-Mantel-Haenszel test.
Clinical response
The reported risk difference for clinical response was 30.6, with a two-sided 95% CI of 22.3 to 38.9 and P < 0.001.
9. Patient-Reported and Biomarker Outcomes
IBDQ total score
LS mean difference in final values
95% CI: 19.16–31.32 · P < 0.001
ANCOVA with modified baseline observation carried forward
The registry reports a mean difference of 25.24 for change from induction baseline to week 40 in IBDQ total score. The two-sided 95% CI is 19.16 to 31.32. The analysis used ANCOVA with mBOCF.
The effect is expressed as a difference in adjusted final-value means rather than a risk difference. The confidence interval therefore describes uncertainty around an adjusted mean difference, not uncertainty around the proportion of participants achieving a binary outcome.
Fecal calprotectin
LS mean difference in final values
95% CI: -1323.08 to -356.21 · P < 0.001
ANCOVA with modified baseline observation carried forward
The reported mean difference for change from induction baseline to week 40 in fecal calprotectin was -839.64, with a two-sided 95% CI of -1323.08 to -356.21. Because the estimate is negative and the confidence interval remains below zero, the two-group adjusted difference in final values was lower for the group represented first in the reported comparison according to the registry's effect-measure ordering.
Bowel urgency NRS
Mean difference in final values
95% CI: -1.51 to -0.61 · P < 0.001
Mixed-effects model
The reported change from baseline to week 40 in bowel urgency based on the Urgency Numeric Rating Scale had a mean difference of -1.06, with a two-sided 95% CI of -1.51 to -0.61. The longitudinal analysis used a mixed-effects model.
These outcomes illustrate why effect measures must be interpreted in the scale of the endpoint. A risk difference such as 23.2 describes an absolute difference in proportions. A mean difference such as 25.24 or -1.06 describes a difference in adjusted outcome values. The numerical magnitudes are therefore not directly comparable across endpoints.
10. Statistical Methods Explained
Why was the Cochran-Mantel-Haenszel test used?
The primary endpoint and several secondary outcomes are binary. The Cochran-Mantel-Haenszel framework provides a way to compare categorical outcomes while accounting for stratification. In this trial, the registry specifically identifies it as the method for the clinical-remission analysis and the other reported binary efficacy analyses.
What does a risk difference of 23.2 mean?
A risk difference of 23.2 means the estimated proportion achieving the binary outcome differs by 23.2 percentage points between the two comparison groups, with the direction determined by the registry's group ordering. It is an absolute measure, so it is distinct from a relative risk or odds ratio.
Why is the confidence interval important?
The 95% confidence interval of 15.2 to 31.2 shows the statistical uncertainty surrounding the estimated primary risk difference. A narrower interval would indicate greater precision, while a wider interval would indicate more uncertainty. The interval is not a prediction interval for individual participants.
What does P < 0.001 mean?
The p-value describes the evidence against the null hypothesis under the specified statistical model and testing framework. It does not quantify how large the treatment effect is. The risk difference and its confidence interval provide the effect-size information.
Why was ANCOVA used for IBDQ and fecal calprotectin?
ANCOVA is appropriate for comparing continuous outcomes while incorporating baseline information into the analysis. Here, the registry identifies ANCOVA as the method and specifies modified baseline observation carried forward for the IBDQ and fecal-calprotectin analyses. The resulting effect is reported as a least-squares mean difference in final values.
Why use a mixed-effects model for bowel urgency?
Bowel urgency is measured longitudinally, making observations from the same participant related rather than independent. A mixed-effects model can represent that repeated-measure structure and estimate treatment-group differences while accounting for within-participant correlation.
What does multiplicity mean in this trial?
Multiple endpoints and comparisons can increase the chance of observing a statistically significant result by chance if each is tested independently. The registry's analysis descriptions identify multiplicity adjustment as a concept for the reported analyses. The exact adjustment procedure is not specified in the posted analysis fields summarized here, so the individual p-values should be read together with the prespecified testing framework rather than treated as unrelated tests.
11. Interpreting the Primary Effect in Context
The primary effect is a risk difference, so the most direct interpretation is an absolute difference in the percentage of participants achieving clinical remission. The estimate of 23.2 is therefore expressed in percentage points rather than as a relative percentage reduction or increase.
The 95% CI extends from 15.2 to 31.2. The interval is entirely above zero, consistent with the reported superiority test and P < 0.001. Its width also matters: the point estimate alone does not communicate the uncertainty surrounding the comparison.
The p-value and the effect estimate answer different questions. The p-value describes evidence against a null hypothesis, whereas the risk difference describes the size and direction of the observed group difference. A statistically small p-value should therefore not be used as a substitute for reporting the treatment effect and its confidence interval.
The mITT definition is important because the endpoint is not simply a comparison of everyone enrolled in the overall study. The registry specifies eligibility for the mITT population, including receipt of at least one dose of mirikizumab and a correctly measured baseline Modified Mayo Score.
12. Serious Adverse Events
The registry reports serious adverse events as affected participants divided by participants at risk for several study periods and analysis groups. Because the denominators differ substantially across groups and periods, these values should be read as group-specific counts rather than as a single overall safety percentage.
| Study period / group | Serious adverse events | Affected / at risk |
|---|---|---|
| Maintenance Period: Miri IR - PBO SC | Serious adverse events | 16 / 192 |
| Maintenance Period: Miri IR - 200 mg Mir | Serious adverse events | 13 / 389 |
| Maintenance Period: PBO IR - PBO SC | Serious adverse events | 8 / 135 |
| Maintenance Period: Miri IR - PBO SC - M | Serious adverse events | 4 / 21 |
| Maintenance Period: Miri IR - 200 mg Mir | Serious adverse events | 7 / 52 |
| Maintenance Period: PBO IR - PBO SC - ME | Serious adverse events | 1 / 12 |
| LOR Rescue Period: LOR Cohort-300 mg Mir | Serious adverse events | 3 / 90 |
| LOR Rescue Period: LOR Cohort-300 mg Mir | Serious adverse events | 2 / 12 |
| Extended Induction: Induction Nonrespond | Serious adverse events | 24 / 461 |
| Extended Induction: Induction Nonrespond | Serious adverse events | 4 / 66 |
| Open Label Maintenance: Delayed Responde | Serious adverse events | 9 / 271 |
| Open Label Maintenance: Delayed Responde | Serious adverse events | 3 / 33 |
13. Design and Statistical Issues to Watch
Allocation terminology
The overall registry design is listed as non-randomized, while the primary analysis describes randomized participants and randomized-arm analysis. This discrepancy should be preserved rather than silently resolving it.
Multiple arms
The study contains 12 arms, but the posted primary and secondary analyses reported here focus on a specific pair of maintenance-period groups.
Induction responders
The primary endpoint is explicitly restricted to mirikizumab induction responders. The treatment comparison therefore concerns a defined responder population rather than all 1328 enrolled participants.
Endpoint scale
Binary outcomes are summarized using risk differences, while continuous outcomes use mean differences. These effect measures answer different statistical questions.
14. Limitations
- Overall allocation description: the registry lists the trial allocation as non-randomized, while the primary statistical analysis describes randomized participants and randomized-arm analysis. This terminology should be considered when interpreting causal conclusions.
- Responder population: the primary endpoint is restricted to mirikizumab induction responders, so the primary estimate does not describe the outcome for every participant enrolled in the study.
- Multiple study arms: the registry reports 12 arms, whereas the posted confirmatory analyses summarized here compare a specific pair of maintenance groups.
- Multiplicity: the analysis descriptions reference multiplicity adjustment, but the exact adjustment procedure is not specified in the posted statistical-analysis fields summarized here.
- Exploratory China analyses: the registry identifies China ME2 analyses as descriptive and exploratory, with no multiplicity control or p-values.
- Safety denominators: serious-adverse-event records span several periods and populations, limiting direct comparison across rows.
- Continuous endpoint assumptions: interpretation of ANCOVA and mixed-effects estimates depends on the underlying modeling framework, outcome scale, baseline adjustment, and repeated-measure structure.
15. Why This Trial Matters Statistically
LUCENT 2 provides a useful example of how a modern maintenance trial can combine categorical efficacy outcomes with continuous patient-reported and biomarker outcomes. The same treatment comparison can therefore be expressed through several different effect measures, each requiring its own interpretation.
| Concept | How it appears in LUCENT 2 |
|---|---|
| Cochran-Mantel-Haenszel test | Used for the primary clinical-remission endpoint and several secondary binary outcomes. |
| Risk difference | Primary and binary secondary outcomes are reported as absolute differences in percentages. |
| ANCOVA | Used for IBDQ total score and fecal calprotectin change from baseline to week 40. |
| Mixed-effects model | Used for bowel urgency measured by the Urgency Numeric Rating Scale. |
| Modified intention-to-treat | Primary efficacy analysis includes randomized participants meeting the specified treatment-exposure and baseline-measure requirements. |
| Multiplicity | The registry analysis descriptions identify multiplicity adjustment as part of the statistical framework. |
| Exploratory analysis | China ME2 analyses were prespecified as descriptive and non-inferential. |
16. Related Tutorials
Learn more about the methods used in this trial:
17. Related Statistical Calculators
18. Sources
- ClinicalTrials.gov: NCT03524092 — LUCENT 2.
- PubMed: PMID 42120652.
- PubMed: PMID 42104877.
- PubMed: PMID 40591171.
- PubMed: PMID 40589947.
- PubMed: PMID 40071779.
Continue through the Clinical Biostats knowledge graph
Connect this trial's endpoints and methods to deeper statistical tutorials and practical analysis tools.
19. Record Summary
LUCENT 2 illustrates a maintenance-trial analysis in which the primary endpoint is binary and assessed at a fixed week-40 time point, while additional outcomes span endoscopic, histologic, symptomatic, patient-reported, biomarker, and bowel-urgency measures. The primary clinical-remission analysis reported a 23.2 risk difference with a two-sided 95% CI of 15.2–31.2 and P < 0.001 using the Cochran-Mantel-Haenszel test. The secondary analyses extend the statistical picture through additional risk differences, ANCOVA-based mean differences, and a mixed-effects model for longitudinal urgency measurements.