← Clinical Trial Results
COPDLongitudinal CohortBiomarkersNCT01969344

SPIROMICS: Complete Statistical Analysis of COPD Subgroups and Biomarkers

An independent statistical analysis of SPIROMICS, a longitudinal clinical study designed to characterize COPD subgroups through clinical measurements, biomarkers, imaging, environmental assessments, and disease progression measures.

Study start: 2010-11 · Primary completion: 2031-06-17
Scope of this record

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

SPIROMICS is a study of COPD subgroups and biomarkers sponsored by the University of North Carolina, Chapel Hill. The registry describes a cohort study evaluating morbidity, lung function, mortality, radiology, biological measures, environmental exposures, and epigenetic features.

2981
Enrollment
4
Arms
18 years
Average study completion timeframe
NCT01969344
Registry ID
FeatureSPIROMICS
TitleStudy of COPD Subgroups and Biomarkers
StatusACTIVE_NOT_RECRUITING
PopulationCOPD; Chronic Obstructive Pulmonary Disease; Chronic Bronchitis; Emphysema
Lead sponsorUniversity of North Carolina, Chapel Hill
Sponsor typeOTHER

2. Clinical Question

Population

Participants in the SPIROMICS cohort with COPD-related conditions including COPD, chronic bronchitis, and emphysema.

Intervention

The registry describes assessments rather than a therapeutic intervention comparison.

Comparator

The registry does not describe a treatment comparator.

Primary question

How do clinical characteristics, biomarkers, imaging findings, environmental factors, and physiological measures characterize COPD subgroups and disease progression?

3. Trial Design

Design
Longitudinal cohort study
Enrollment
2981 participants
Arms
4 arms
Follow-up
Through study completion, an average of 18 years

The statistical structure of SPIROMICS differs from a conventional randomized treatment trial. Its primary purpose is characterization of disease patterns over time, requiring methods capable of handling repeated measurements, longitudinal follow-up, heterogeneous disease phenotypes, and time-to-event outcomes.

4. Endpoints

EndpointTime frameDescription
MorbidityThrough study completion, an average of 18 yearsAcute exacerbations in the SPIROMICS cohort.
Lung FunctionThrough study completion, an average of 18 yearsSerial measurements including FEV1, FVC, FRC, and IC using spirometry/plethysmography.
MortalityThrough study completion, an average of 18 yearsDeaths identified during follow-up and clinic scheduling attempts, with cause determined through chart review and adjudication.
RadiologyThrough study completion, an average of 18 yearsMeasures including percent emphysema, PRMfSAD, DPM classifications, and Pi10.

The registry also lists SPIROMICS III endpoints covering physiology, mucin biology, mucin-interacting proteins, environmental exposures, indoor measurements, geocoding, and epigenetic measurements.

5. Planned Analysis

No formal statistical analyses were posted to ClinicalTrials.gov. The registry describes measurements and endpoints but does not report estimates, confidence intervals, or p-values.

How endpoints of this type are typically analyzed

Longitudinal lung-function measures such as serial FEV1 are commonly evaluated using repeated-measures models that account for correlation within participants over time. Acute exacerbations and mortality are commonly evaluated using event-based approaches, including time-to-event models when appropriate.

Imaging, biomarker, and environmental measures may require multivariable modeling to evaluate relationships between measurements and clinical characteristics while accounting for multiple potential sources of variation.

6. Statistical Methodology

Longitudinal modeling

Repeated observations from the same participant are statistically related. Longitudinal methods account for within-participant correlation rather than treating each observation as independent.

Repeated measurements → participant-level correlation → longitudinal model

These approaches allow researchers to evaluate trajectories, such as changes in lung function over time, while using information from multiple visits.

Time-to-event analysis

Mortality and morbidity outcomes may be evaluated using methods designed for outcomes where the timing of an event matters and some participants may not experience the event during follow-up.

Multidimensional phenotype analysis

SPIROMICS combines physiological, radiological, biological, and environmental measurements. Statistical analyses of such data often require methods that can integrate multiple correlated variables while identifying clinically meaningful patterns.

7. Statistical Methods Explained

Why are repeated measurements important in COPD research?

COPD progression is dynamic. Repeated measurements allow investigators to study changes over time rather than relying only on a single baseline measurement.

Why is FEV1 measured serially?

The registry identifies serial measurements of FEV1 as an objective assessment of disease progression. Repeated FEV1 measurements provide information about lung-function trajectories.

Why are biomarkers combined with clinical measurements?

Biomarkers may provide additional biological information that complements symptoms, physiology, and imaging findings.

Why are imaging endpoints statistically challenging?

Imaging-derived measures such as emphysema volume and PRM metrics can involve large amounts of structured data. Analysis must account for measurement variability and relationships among multiple imaging features.

Why are environmental measurements included?

The registry includes outdoor and indoor exposure measurements, allowing researchers to evaluate environmental factors alongside clinical and biological characteristics.

8. Limitations

9. Why This Trial Matters Statistically

SPIROMICS illustrates the statistical challenges of modern respiratory research. COPD is not a single uniform disease state, and meaningful analysis requires integrating multiple domains of information: symptoms, physiology, imaging, molecular measurements, and environmental exposures.

From a biostatistical perspective, the study highlights the transition from simple endpoint comparisons toward longitudinal and multidimensional modeling approaches. Understanding COPD heterogeneity requires methods that can separate true biological patterns from measurement variability.

10. Sources