Methods of treating a subject with an alternative to anti-TNF therapy
Inventors
Johnson, Keith J. • Ghiassian, Susan
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-11987620-B2
Publication Date
2024-05-21
Expiration Date
Abstract
Methods of treating a subject suffering from an autoimmune disease, disorder, or condition with an alternative to anti-TNF therapy.
Core Innovation
The disclosure relates to treating a subject suffering from an autoimmune disease, disorder, or condition with an alternative therapy to anti-TNF therapy by using a gene expression response signature and a trained machine learning classifier to classify a subject as being indicative of non-response to the anti-TNF therapy. The gene expression response signature comprises gene expression levels of a set of genes in a biological sample from the subject, where the set of genes are differentially expressed in a first population of subjects who respond to the anti-TNF therapy compared to a second population of subjects who do not respond to the anti-TNF therapy.
A trained machine learning classifier is applied to the gene expression response signature to classify the gene expression response signature as being indicative of non-response of the subject to the anti-TNF therapy. The trained machine learning classifier is obtained at least in part by determining, for each of the set of genes, a significance of correlation with response outcome to the anti-TNF therapy among the responders compared to the non-responders, selecting a reduced subset of genes with highest significance of correlation, training a classifier using gene expression levels of the reduced subset together with responder or non-responder outcomes, and validating the classifier on a second independent cohort of subjects who received the anti-TNF therapy and were determined as responding or not responding.
The validated classifier is further configured by selecting a cutoff score such that the validated classifier achieves a true negative rate (TNR) of at least 0.5 and a negative predictive value (NPV) of at least 0.9. Based at least in part on the indicated non-response, the subject is selected not to receive the anti-TNF therapy and instead to receive the alternative therapy to anti-TNF therapy, where the alternative therapy is selected from rituximab, sarilumab, tofacitinib citrate, leflunomide, vedolizumab, tocilizumab, anakinra, and abatacept.
Claims Coverage
The provided partial content includes one independent claim. It contains a full method pipeline that includes obtaining a gene expression response signature, applying a trained machine learning classifier, validating it with specified performance thresholds, and using the prediction to select and administer an alternative therapy to anti-TNF therapy.
Signature-based classification of anti-TNF non-response
Obtain a gene expression response signature comprising gene expression levels of a set of genes in a biological sample from the subject, wherein the set of genes are differentially expressed in responders to anti-TNF therapy as compared to non-responders, and apply a trained machine learning classifier to classify the gene expression response signature as being indicative of non-response to the anti-TNF therapy.
Correlation-driven reduced gene subset with trained classifier
Obtain the trained machine learning classifier at least in part by determining, for each gene, a significance of correlation with response outcome to the anti-TNF therapy among responders compared to non-responders, selecting a reduced subset of genes with highest significance of correlation, training a classifier to classify whether the gene expression response signature associated with the reduced subset is indicative of non-response using gene expression levels and responder/non-responder outcomes.
Independent-cohort validation with cutoff achieving TNR and NPV thresholds
Validate the classifier on a second independent cohort of subjects who have received the anti-TNF therapy and have been determined as either responding or not responding, and select a cutoff score for the validated classifier such that the validated classifier achieves a true negative rate (TNR) of at least 0.5 and a negative predictive value (NPV) of at least 0.9.
Therapy selection based on predicted non-response and administration of alternative anti-TNF therapies
Based at least in part on the determining of non-response to the anti-TNF therapy, select the subject to not receive the anti-TNF therapy and instead receive the alternative therapy to anti-TNF therapy, and administer the alternative therapy selected from rituximab, sarilumab, tofacitinib citrate, leflunomide, vedolizumab, tocilizumab, anakinra, and abatacept.
Across the independent claim, the core inventive aspects are using a differentially expressed gene expression response signature between responders and non-responders, training a classifier using correlation-based selection of a reduced gene subset, validating the classifier on a second independent cohort with specified TNR and NPV thresholds via cutoff selection, and selecting an alternative non-anti-TNF therapy for a subject predicted to be indicative of non-response.
Stated Advantages
Not explicitly described in patent.
Documented Applications
Not explicitly described in patent.
Interested in licensing this patent?