Contributions
Abstract: P877
Type: Poster
Abstract Category: Clinical aspects of MS - 5 Epidemiology
Background: While there is no consensus on definition, high disease activity (HDA) relapsing-remitting multiple sclerosis (RRMS) can be characterized, in part, by frequent relapses or a high burden of brain lesions (by number of lesions). Valid identification of patients with HDA-RRMS allows deeper insight into characteristics and predictors of HDA-RRMS, and may lead to more effective treatment options.
Objectives: Determine feasibility of using unstructured MRI reports to determine number of lesions (and associated information) as part of an EHR-based algorithm for HDA-RRMS patient identification in a US Integrated Delivery Network (IDN).
Methods: Study criteria included MS diagnosis between 2010 to 2014, age ≥ 18 years, ≥ 1 year baseline history, RRMS subtype identified using NLP in the clinical notes, no progressive MS disease and no other demyelinating diseases. HDA groups were defined based on frequency of relapse and number / type of brain or spinal lesions (e.g., gadolinium-enhanced T1, T2). Unstructured MRI radiology reports were reviewed to assess the ability to identify brain lesion burden. MRI reports were assessed for:
1) T1 lesion assessment,
2) T2 lesion assessment,
3) gadolinium contrast lesion assessment,
4) number of lesions,
5) volume/size of lesions,
6) location of lesions,
7) comparison to previous exam, and
8) presence of gadolinium-enhancing lesions.
Results: Of the 102 MRI reports used representing 25 unique patients randomly sampled from the 837 RRMS patient cohort, 100% included T1 and T2 lesion assessment. Most included gadolinium contrast assessment, location of lesions and comparison to a previous exam (93.1%, 83.3% and 92.4% where applicable, respectively). However, 83.3% did not indicate the number of lesions, and 89.2% did not indicate the volume of lesions. 100% of applicable reports indicated if there were any new lesions or changes to existing lesions in the comparison to previous exam, or if there were any contrast-enhancing lesions, noting 10.8% did have a new lesion(s) or lesion change and only 2.9% showed a gadolinium-enhancing lesion(s).
Conclusions: Overall, the information contained in the MRI reports alone was not sufficient to longitudinally define HDA subgroup as originally conceived. Based on the information captured, further work is underway to identify how this data can be best used or supplemented to classify clinically relevant brain lesion burden in RRMS patients.
Disclosure: This study was funded by Merck KGaA.
- Aaron W.C. Kamauu received a research grant from Merck KGaA.
- Hoa Van Le is an employee of PAREXEL and a stockholder of GlaxoSmithKline, and was a Harry Guess-Merck merit scholarship recipient.
- Chi Thi Le Truong is an employee of MedCodeWorld,
- Hannah R. Crooke is an employee of PAREXEL.
- John R. Holmen has nothing to disclose.
- Christopher L. Fillmore has nothing to disclose.
- Meritxell Sabidó-Espin is an employee of Merck KGaA.
- Schiffon L. Wong is an employee EMD Serono.
Abstract: P877
Type: Poster
Abstract Category: Clinical aspects of MS - 5 Epidemiology
Background: While there is no consensus on definition, high disease activity (HDA) relapsing-remitting multiple sclerosis (RRMS) can be characterized, in part, by frequent relapses or a high burden of brain lesions (by number of lesions). Valid identification of patients with HDA-RRMS allows deeper insight into characteristics and predictors of HDA-RRMS, and may lead to more effective treatment options.
Objectives: Determine feasibility of using unstructured MRI reports to determine number of lesions (and associated information) as part of an EHR-based algorithm for HDA-RRMS patient identification in a US Integrated Delivery Network (IDN).
Methods: Study criteria included MS diagnosis between 2010 to 2014, age ≥ 18 years, ≥ 1 year baseline history, RRMS subtype identified using NLP in the clinical notes, no progressive MS disease and no other demyelinating diseases. HDA groups were defined based on frequency of relapse and number / type of brain or spinal lesions (e.g., gadolinium-enhanced T1, T2). Unstructured MRI radiology reports were reviewed to assess the ability to identify brain lesion burden. MRI reports were assessed for:
1) T1 lesion assessment,
2) T2 lesion assessment,
3) gadolinium contrast lesion assessment,
4) number of lesions,
5) volume/size of lesions,
6) location of lesions,
7) comparison to previous exam, and
8) presence of gadolinium-enhancing lesions.
Results: Of the 102 MRI reports used representing 25 unique patients randomly sampled from the 837 RRMS patient cohort, 100% included T1 and T2 lesion assessment. Most included gadolinium contrast assessment, location of lesions and comparison to a previous exam (93.1%, 83.3% and 92.4% where applicable, respectively). However, 83.3% did not indicate the number of lesions, and 89.2% did not indicate the volume of lesions. 100% of applicable reports indicated if there were any new lesions or changes to existing lesions in the comparison to previous exam, or if there were any contrast-enhancing lesions, noting 10.8% did have a new lesion(s) or lesion change and only 2.9% showed a gadolinium-enhancing lesion(s).
Conclusions: Overall, the information contained in the MRI reports alone was not sufficient to longitudinally define HDA subgroup as originally conceived. Based on the information captured, further work is underway to identify how this data can be best used or supplemented to classify clinically relevant brain lesion burden in RRMS patients.
Disclosure: This study was funded by Merck KGaA.
- Aaron W.C. Kamauu received a research grant from Merck KGaA.
- Hoa Van Le is an employee of PAREXEL and a stockholder of GlaxoSmithKline, and was a Harry Guess-Merck merit scholarship recipient.
- Chi Thi Le Truong is an employee of MedCodeWorld,
- Hannah R. Crooke is an employee of PAREXEL.
- John R. Holmen has nothing to disclose.
- Christopher L. Fillmore has nothing to disclose.
- Meritxell Sabidó-Espin is an employee of Merck KGaA.
- Schiffon L. Wong is an employee EMD Serono.