Systems modeling advances precision medicine in alopecia

 

PORTLAND – Alopecia areata can resist treatment stubbornly, but dermatologists might soon have better tools to predict response to therapy.

Personalized gene sequencing is key to this type of precision medicine, but conventional sequencing can be “extremely cumbersome and clinically impractical,” James C. Chen, PhD, said at the annual meeting of the Society for Investigative Dermatology.

At Columbia University in New York, Dr. Chen and his associates are working to solve that problem. Instead of evaluating “2,000 some-odd genes” to predict treatment response, they are using network analyses to pinpoint the minimum number of transcription regulators needed for a medicine to produce a therapeutic effect. Only about a dozen of these “master regulators” might need to be analyzed to begin matching therapies to patients with alopecia, he said. “We can build predictive models that track how patients progress,” he added. “If a patient was put on tofacitinib and did not respond, we can see whether we could have predicted that beforehand.”

During alopecia trials at Columbia, researchers routinely perform RNA sequencing of scalp biopsies to analyze therapeutic response on a molecular level. Using these RNAseq data from patients with untreated alopecia areata and gene regulatory network analysis data from the Algorithm for the Reconstruction of Accurate Cellular Networks, Dr. Chen and his associates modeled the molecular mechanisms of action of the pan–Janus kinase inhibitor tofacitinib, the JAK1/JAK2 inhibitor ruxolitinib, the CTLA4 inhibitor abatacept, and intralesional triamcinolone acetonide (IL-TAC). Heat maps of molecular responses to treatment showed distinct mechanisms of action between IL-TAC and abatacept, Dr. Chen said.

Furthermore, these therapies showed distinct and much less robust molecular effects than either ruxolitinib or tofacitinib. A Venn diagram of the biosignatures and molecular mechanisms of action of all four therapies showed little overlap. In fact, the probability of so little overlap between tofacitinib and IL-TAC occurring by chance was 0.023. The lack of overlap between the two JAK inhibitors was even more pronounced (P = 2.21 x 10^–11).

Only 5-10 transcription factors are needed to capture these molecular mechanisms of action, which could greatly streamline precision dermatology in the future, according to Dr. Chen. “Systems biology offers a foundation for developing precision medicine strategies and selecting treatments for patients based on their individual molecular pathology,” he concluded. “Even when patients with alopecia areata have the same clinical phenotype, the molecular pathways they take to get there are not necessarily the same. We need to define those paths to maximize our chances of matching drugs to patients.”

Dr. Chen acknowledged support from the National Institutes of Health, epiCURE, and the National Institute of Arthritis and Musculoskeletal and Skin Diseases. He had no relevant financial conflicts of interest.

 

PORTLAND – Alopecia areata can resist treatment stubbornly, but dermatologists might soon have better tools to predict response to therapy.

Personalized gene sequencing is key to this type of precision medicine, but conventional sequencing can be “extremely cumbersome and clinically impractical,” James C. Chen, PhD, said at the annual meeting of the Society for Investigative Dermatology.

At Columbia University in New York, Dr. Chen and his associates are working to solve that problem. Instead of evaluating “2,000 some-odd genes” to predict treatment response, they are using network analyses to pinpoint the minimum number of transcription regulators needed for a medicine to produce a therapeutic effect. Only about a dozen of these “master regulators” might need to be analyzed to begin matching therapies to patients with alopecia, he said. “We can build predictive models that track how patients progress,” he added. “If a patient was put on tofacitinib and did not respond, we can see whether we could have predicted that beforehand.”

During alopecia trials at Columbia, researchers routinely perform RNA sequencing of scalp biopsies to analyze therapeutic response on a molecular level. Using these RNAseq data from patients with untreated alopecia areata and gene regulatory network analysis data from the Algorithm for the Reconstruction of Accurate Cellular Networks, Dr. Chen and his associates modeled the molecular mechanisms of action of the pan–Janus kinase inhibitor tofacitinib, the JAK1/JAK2 inhibitor ruxolitinib, the CTLA4 inhibitor abatacept, and intralesional triamcinolone acetonide (IL-TAC). Heat maps of molecular responses to treatment showed distinct mechanisms of action between IL-TAC and abatacept, Dr. Chen said.

Furthermore, these therapies showed distinct and much less robust molecular effects than either ruxolitinib or tofacitinib. A Venn diagram of the biosignatures and molecular mechanisms of action of all four therapies showed little overlap. In fact, the probability of so little overlap between tofacitinib and IL-TAC occurring by chance was 0.023. The lack of overlap between the two JAK inhibitors was even more pronounced (P = 2.21 x 10^–11).

Only 5-10 transcription factors are needed to capture these molecular mechanisms of action, which could greatly streamline precision dermatology in the future, according to Dr. Chen. “Systems biology offers a foundation for developing precision medicine strategies and selecting treatments for patients based on their individual molecular pathology,” he concluded. “Even when patients with alopecia areata have the same clinical phenotype, the molecular pathways they take to get there are not necessarily the same. We need to define those paths to maximize our chances of matching drugs to patients.”

Dr. Chen acknowledged support from the National Institutes of Health, epiCURE, and the National Institute of Arthritis and Musculoskeletal and Skin Diseases. He had no relevant financial conflicts of interest.

 

PORTLAND – Alopecia areata can resist treatment stubbornly, but dermatologists might soon have better tools to predict response to therapy.

Personalized gene sequencing is key to this type of precision medicine, but conventional sequencing can be “extremely cumbersome and clinically impractical,” James C. Chen, PhD, said at the annual meeting of the Society for Investigative Dermatology.

At Columbia University in New York, Dr. Chen and his associates are working to solve that problem. Instead of evaluating “2,000 some-odd genes” to predict treatment response, they are using network analyses to pinpoint the minimum number of transcription regulators needed for a medicine to produce a therapeutic effect. Only about a dozen of these “master regulators” might need to be analyzed to begin matching therapies to patients with alopecia, he said. “We can build predictive models that track how patients progress,” he added. “If a patient was put on tofacitinib and did not respond, we can see whether we could have predicted that beforehand.”

During alopecia trials at Columbia, researchers routinely perform RNA sequencing of scalp biopsies to analyze therapeutic response on a molecular level. Using these RNAseq data from patients with untreated alopecia areata and gene regulatory network analysis data from the Algorithm for the Reconstruction of Accurate Cellular Networks, Dr. Chen and his associates modeled the molecular mechanisms of action of the pan–Janus kinase inhibitor tofacitinib, the JAK1/JAK2 inhibitor ruxolitinib, the CTLA4 inhibitor abatacept, and intralesional triamcinolone acetonide (IL-TAC). Heat maps of molecular responses to treatment showed distinct mechanisms of action between IL-TAC and abatacept, Dr. Chen said.

Furthermore, these therapies showed distinct and much less robust molecular effects than either ruxolitinib or tofacitinib. A Venn diagram of the biosignatures and molecular mechanisms of action of all four therapies showed little overlap. In fact, the probability of so little overlap between tofacitinib and IL-TAC occurring by chance was 0.023. The lack of overlap between the two JAK inhibitors was even more pronounced (P = 2.21 x 10^–11).

Only 5-10 transcription factors are needed to capture these molecular mechanisms of action, which could greatly streamline precision dermatology in the future, according to Dr. Chen. “Systems biology offers a foundation for developing precision medicine strategies and selecting treatments for patients based on their individual molecular pathology,” he concluded. “Even when patients with alopecia areata have the same clinical phenotype, the molecular pathways they take to get there are not necessarily the same. We need to define those paths to maximize our chances of matching drugs to patients.”

Dr. Chen acknowledged support from the National Institutes of Health, epiCURE, and the National Institute of Arthritis and Musculoskeletal and Skin Diseases. He had no relevant financial conflicts of interest.