Asthma
Conditions
Brief summary
This study will utilize University of Michigan EMR data to form a retrospective cohort of severe asthma patients on biologic therapies for asthma with computed tomography (CT) imaging and known outcomes on therapy. These images will then be analyzed using morphomics, a combination of high-throughput image analysis and deep learning techniques, to derive imaging biomarkers that will be able to predict therapeutic response to biologics. These biomarkers will then be tested in a second cohort from the National Jewish Health to assess for validity.
Detailed description
The two arms for this study represent the two cohorts: the University of Michigan cohort, and the National Jewish Health cohort. The cohorts were split into arms, as the National Jewish Health cohort was intended to act as a validator group for results derived from participants in the University of Michigan cohort.
Interventions
None listed
Sponsors
University of Michigan
Study design
Observational model
COHORT
Time perspective
RETROSPECTIVE
Eligibility
Sex/Gender
ALL
Age
18 Years to 85 Years
Healthy volunteers
No
Inclusion criteria
* Asthma diagnosis defined as a positive methacholine test, supportive PFTs, or clinician diagnosis with appropriate supporting clinical context * Prescribed and taking an FDA-approved biologic medication for severe asthma (mepolizumab, benralizumab, reslizumab, omalizumab, or dupilumab) * Chest computerized tomography (CT) images taken for any clinical purpose
Exclusion criteria
* 10-year history of smoking * Evidence of emphysema on CT imaging * Severe bronchiectasis secondary to other causes (subjects with scattered bronchiectasis secondary to severe asthma will not excluded) * Lack of accessible adequate follow-up information (4 or more months from drug initiation with clear documentation of response) will be excluded from final analysis due to inability to assess response but will be included in the initial cluster identification
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Chest Wall Muscle Size | 28 months | Muscle size was measured as the skeletal muscle cross-sectional area in centimeters squared (cm2) measured at the level of the T10 vertebra to determine the cutoff value in chest wall muscle size that optimally classified patients that had and did not have any reduction in pre-treatment baseline exacerbation frequency (binary). |
| Exacerbation Reduction - Muscle Size | 28 months | Muscle size was measured as the skeletal muscle cross-sectional area in centimeters squared (cm2) measured at the level of the T10 vertebra to assess whether skeletal muscle size area could be used to predict exacerbation frequency. Results reflect the average muscle sizes when frequency could be predicted. |
| Exacerbations | pre-biologic at baseline (looking back 1 year), post-biologic (4-18 months after biologic) | The number of exacerbations post-biologic were compared to the the number of baseline exacerbations. Baseline measurements were counted in the year prior to biologic initiation. Any exacerbation recorded between 4 to 28 months after biologic initiation was recorded as a post-biologic exacerbation. Any reduction in between the pre- and post-biologic was counted as an exacerbation reduction. Results reflect the number of participants who experienced (True) or did not experience (False) a reduction in exacerbations. |
| Exacerbation Reduction - Muscle Quality | 28 months | Muscle quality was measured as the skeletal muscle density in Hounsfield Units (HU) to determine the cutoff value in chest wall muscle quality to assess whether skeletal muscle quality could be used to predict exacerbation frequency. Results reflect the average muscle sizes when frequency could be predicted. |
| Chest Wall Muscle Quality | 28 months | Muscle quality was measured as the skeletal muscle density in Hounsfield Units (HU) to determine the cutoff value in chest wall muscle quality that optimally classified patients that had and did not have any reduction in pre-treatment baseline exacerbation frequency (binary). |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Reduction in Oral Corticosteroid (OCS) | pre-biologic at baseline (looking back 1 year), post-biologic (4-18 months after biologic) | Muscle size was used to investigate body composition measures associated with any reduction in maintenance OCS use when compared to baseline OCS use. Subjects with any OCS reduction in use during the 4- to 18-month period following initiation of biologics were classified as responders. |
| FEV1 Improvement | pre-biologic at baseline (looking back 1 year), post-biologic (4-18 months after biologic) | Muscle size was used to investigate body composition measures associated with a 5% or greater improvement in forced expiratory volume in one second (FEV1). Muscle size was measured as the skeletal muscle cross-sectional area in centimeters squared (cm2) measured at the level of the T10 vertebra to determine the cutoff value in chest wall muscle size that optimally classified participants who had and did not have improvement in FEV1. Participants with ≥5% improvement in FEV1 during the 4 to 18-month period following initiation of biologics were classified as responders. Any participant who did not achieve a ≥5% improvement from baseline in that timeframe was classified as a non-responder. |
Countries
United States
Participant flow
Participants by arm
| Arm | Count |
|---|---|
| University of Michigan, Ann Arbor, MI Severe asthma patients treated at Michigan Medicine who are on biologic therapies for asthma and who have CT imaging and known outcomes on therapy.
No interventions were included in this retrospective study. | 152 |
| National Jewish Health, Denver, CO Severe asthma patients treated at National Jewish Health who are on biologic therapies for asthma and who have CT imaging and known outcomes on therapy.
No interventions were included in this retrospective study. | 81 |
| Total | 233 |
Baseline characteristics
| Characteristic | University of Michigan, Ann Arbor, MI | National Jewish Health, Denver, CO | Total |
|---|---|---|---|
| Age, Continuous | 52.39 years STANDARD_DEVIATION 14.31 | 51.75 years STANDARD_DEVIATION 15.58 | 52.17 years STANDARD_DEVIATION 14.73 |
| Age, Customized Female | 51.42 years STANDARD_DEVIATION 13.21 | 48.08 years STANDARD_DEVIATION 17.86 | 50.50 years STANDARD_DEVIATION 14.65 |
| Age, Customized Male | 54.55 years STANDARD_DEVIATION 16.46 | 55.34 years STANDARD_DEVIATION 12.15 | 54.92 years STANDARD_DEVIATION 14.53 |
| Body Mass Index (BMI) Female | 32.46 kg/m^2 STANDARD_DEVIATION 9.27 | 31.70 kg/m^2 STANDARD_DEVIATION 9.21 | 32.25 kg/m^2 STANDARD_DEVIATION 9.23 |
| Body Mass Index (BMI) Male | 29.97 kg/m^2 STANDARD_DEVIATION 4.92 | 30.62 kg/m^2 STANDARD_DEVIATION 5.61 | 30.27 kg/m^2 STANDARD_DEVIATION 5.23 |
| Ethnicity (NIH/OMB) Hispanic or Latino | 3 Participants | 12 Participants | 15 Participants |
| Ethnicity (NIH/OMB) Not Hispanic or Latino | 144 Participants | 68 Participants | 212 Participants |
| Ethnicity (NIH/OMB) Unknown or Not Reported | 5 Participants | 1 Participants | 6 Participants |
| Height Female | 162.31 centimeters STANDARD_DEVIATION 6.78 | 162.84 centimeters STANDARD_DEVIATION 6.05 | 162.45 centimeters STANDARD_DEVIATION 6.57 |
| Height Male | 176.95 centimeters STANDARD_DEVIATION 6.56 | 176.59 centimeters STANDARD_DEVIATION 7.78 | 176.78 centimeters STANDARD_DEVIATION 7.12 |
| Race (NIH/OMB) American Indian or Alaska Native | 1 Participants | 0 Participants | 1 Participants |
| Race (NIH/OMB) Asian | 3 Participants | 0 Participants | 3 Participants |
| Race (NIH/OMB) Black or African American | 21 Participants | 2 Participants | 23 Participants |
| Race (NIH/OMB) More than one race | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 2 Participants | 10 Participants | 12 Participants |
| Race (NIH/OMB) White | 125 Participants | 69 Participants | 194 Participants |
| Region of Enrollment United States | 152 Participants | 81 Participants | 233 Participants |
| Sex: Female, Male Female | 105 Participants | 40 Participants | 145 Participants |
| Sex: Female, Male Male | 47 Participants | 41 Participants | 88 Participants |
| Weight Female | 85.31 kilograms STANDARD_DEVIATION 23.83 | 83.85 kilograms STANDARD_DEVIATION 23.89 | 84.91 kilograms STANDARD_DEVIATION 23.77 |
| Weight Male | 93.99 kilograms STANDARD_DEVIATION 16.88 | 95.43 kilograms STANDARD_DEVIATION 18.33 | 94.66 kilograms STANDARD_DEVIATION 17.48 |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk |
|---|---|---|
| deaths Total, all-cause mortality | 0 / 0 | 0 / 0 |
| other Total, other adverse events | 0 / 0 | 0 / 0 |
| serious Total, serious adverse events | 0 / 0 | 0 / 0 |
Outcome results
Primary
Chest Wall Muscle Quality
Muscle quality was measured as the skeletal muscle density in Hounsfield Units (HU) to determine the cutoff value in chest wall muscle quality that optimally classified patients that had and did not have any reduction in pre-treatment baseline exacerbation frequency (binary).
Time frame: 28 months
Population: 8 participants in the University of Michigan arm and 1 participant in the National Jewish Health arm did not have either a baseline or post-treatment count of exacerbations so the outcome could not be computed for them. In addition, 20 participants in the National Jewish Health arm (including the 1 without complete outcome data) did not have complete skeletal muscle visible within the scan field, so muscle size and quality could not be adequately assessed.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| University of Michigan, Ann Arbor, MI | Chest Wall Muscle Quality | Did not have reduction in exacerbations | 39.18 HU | Standard Deviation 6.96 |
| University of Michigan, Ann Arbor, MI | Chest Wall Muscle Quality | Had a reduction in exacerbations | 36.57 HU | Standard Deviation 7.94 |
| National Jewish Health, Denver, CO | Chest Wall Muscle Quality | Had a reduction in exacerbations | 30.27 HU | Standard Deviation 7.28 |
| National Jewish Health, Denver, CO | Chest Wall Muscle Quality | Did not have reduction in exacerbations | 28.60 HU | Standard Deviation 6.43 |
Primary
Chest Wall Muscle Size
Muscle size was measured as the skeletal muscle cross-sectional area in centimeters squared (cm2) measured at the level of the T10 vertebra to determine the cutoff value in chest wall muscle size that optimally classified patients that had and did not have any reduction in pre-treatment baseline exacerbation frequency (binary).
Time frame: 28 months
Population: 8 participants in the University of Michigan arm and 1 participant in the National Jewish Health arm did not have either a baseline or post-treatment count of exacerbations so the outcome could not be computed for them. In addition, 20 participants in the National Jewish Health arm (including the 1 without complete outcome data) did not have complete skeletal muscle visible within the scan field, so muscle size and quality could not be adequately assessed.
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| University of Michigan, Ann Arbor, MI | Chest Wall Muscle Size | Had reduction in exacerbations | 95.49 cm2 | Standard Deviation 25.63 |
| University of Michigan, Ann Arbor, MI | Chest Wall Muscle Size | Did not have reduction in exacerbations | 87.98 cm2 | Standard Deviation 23.65 |
| National Jewish Health, Denver, CO | Chest Wall Muscle Size | Had reduction in exacerbations | 97.82 cm2 | Standard Deviation 29.04 |
| National Jewish Health, Denver, CO | Chest Wall Muscle Size | Did not have reduction in exacerbations | 102.86 cm2 | Standard Deviation 28.77 |
Primary
Exacerbation Reduction - Muscle Quality
Muscle quality was measured as the skeletal muscle density in Hounsfield Units (HU) to determine the cutoff value in chest wall muscle quality to assess whether skeletal muscle quality could be used to predict exacerbation frequency. Results reflect the average muscle sizes when frequency could be predicted.
Time frame: 28 months
Population: 8 participants in the University of Michigan arm and 1 participant in the National Jewish Health arm did not have either a baseline or post-treatment count of exacerbations so the outcome could not be computed for them. In addition, 20 participants in the National Jewish Health arm (including the 1 without complete outcome data) did not have complete skeletal muscle visible within the scan field, so muscle size and quality could not be adequately assessed.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| University of Michigan, Ann Arbor, MI | Exacerbation Reduction - Muscle Quality | 37 HU | Standard Deviation 7.8 |
| National Jewish Health, Denver, CO | Exacerbation Reduction - Muscle Quality | 29.8 HU | Standard Deviation 7 |
p-value: 0.007Regression, Linear
p-value: 0.5Regression, Linear
Primary
Exacerbation Reduction - Muscle Size
Muscle size was measured as the skeletal muscle cross-sectional area in centimeters squared (cm2) measured at the level of the T10 vertebra to assess whether skeletal muscle size area could be used to predict exacerbation frequency. Results reflect the average muscle sizes when frequency could be predicted.
Time frame: 28 months
Population: 8 participants in the University of Michigan arm and 1 participant in the National Jewish Health arm did not have either a baseline or post-treatment count of exacerbations so the outcome could not be computed for them. In addition, 20 participants in the National Jewish Health arm (including the 1 without complete outcome data) did not have complete skeletal muscle visible within the scan field, so muscle size and quality could not be adequately assessed.
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| University of Michigan, Ann Arbor, MI | Exacerbation Reduction - Muscle Size | 94.3 cm2 | Standard Deviation 25.4 |
| National Jewish Health, Denver, CO | Exacerbation Reduction - Muscle Size | 99.3 cm2 | Standard Deviation 28.8 |
p-value: 0.411Regression, Linear
p-value: 0.161Regression, Linear
Primary
Exacerbations
The number of exacerbations post-biologic were compared to the the number of baseline exacerbations. Baseline measurements were counted in the year prior to biologic initiation. Any exacerbation recorded between 4 to 28 months after biologic initiation was recorded as a post-biologic exacerbation. Any reduction in between the pre- and post-biologic was counted as an exacerbation reduction. Results reflect the number of participants who experienced (True) or did not experience (False) a reduction in exacerbations.
Time frame: pre-biologic at baseline (looking back 1 year), post-biologic (4-18 months after biologic)
Population: 8 participants in the University of Michigan arm and 1 participant in the National Jewish Health arm did not have either a baseline or post-treatment count of exacerbations so the outcome could not be computed for them.
| Arm | Measure | Group | Value (COUNT_OF_PARTICIPANTS) |
|---|---|---|---|
| University of Michigan, Ann Arbor, MI | Exacerbations | True | 122 Participants |
| University of Michigan, Ann Arbor, MI | Exacerbations | False | 22 Participants |
| National Jewish Health, Denver, CO | Exacerbations | True | 61 Participants |
| National Jewish Health, Denver, CO | Exacerbations | False | 19 Participants |
Secondary
FEV1 Improvement
Muscle size was used to investigate body composition measures associated with a 5% or greater improvement in forced expiratory volume in one second (FEV1). Muscle size was measured as the skeletal muscle cross-sectional area in centimeters squared (cm2) measured at the level of the T10 vertebra to determine the cutoff value in chest wall muscle size that optimally classified participants who had and did not have improvement in FEV1. Participants with ≥5% improvement in FEV1 during the 4 to 18-month period following initiation of biologics were classified as responders. Any participant who did not achieve a ≥5% improvement from baseline in that timeframe was classified as a non-responder.
Time frame: pre-biologic at baseline (looking back 1 year), post-biologic (4-18 months after biologic)
Population: 48 UM patients and 31 NJ patients that lacked a suitable CT and pre- and post-biologic FEV1 measure were not included in this analysis
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| University of Michigan, Ann Arbor, MI | FEV1 Improvement | Non-Responders | 88.31 cm2 | Standard Deviation 23.85 |
| University of Michigan, Ann Arbor, MI | FEV1 Improvement | Responders | 94.59 cm2 | Standard Deviation 25.06 |
| National Jewish Health, Denver, CO | FEV1 Improvement | Non-Responders | 102.81 cm2 | Standard Deviation 25.26 |
| National Jewish Health, Denver, CO | FEV1 Improvement | Responders | 96.93 cm2 | Standard Deviation 32.43 |
Secondary
Reduction in Oral Corticosteroid (OCS)
Muscle size was used to investigate body composition measures associated with any reduction in maintenance OCS use when compared to baseline OCS use. Subjects with any OCS reduction in use during the 4- to 18-month period following initiation of biologics were classified as responders.
Time frame: pre-biologic at baseline (looking back 1 year), post-biologic (4-18 months after biologic)
Population: 25 UM patients and 20 NJ patients that lacked a suitable CT and pre- and post-biologic OCS maintenance use were not included in this analysis
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| University of Michigan, Ann Arbor, MI | Reduction in Oral Corticosteroid (OCS) | Responders | 91.17 cm2 | Standard Deviation 23.84 |
| University of Michigan, Ann Arbor, MI | Reduction in Oral Corticosteroid (OCS) | Non-Responders | 96.09 cm2 | Standard Deviation 26.17 |
| National Jewish Health, Denver, CO | Reduction in Oral Corticosteroid (OCS) | Responders | 103.94 cm2 | Standard Deviation 32.96 |
| National Jewish Health, Denver, CO | Reduction in Oral Corticosteroid (OCS) | Non-Responders | 97.92 cm2 | Standard Deviation 27.71 |