OBJECTIVE: Promptly establishing maintenance therapy could reduce morbidity and mortality in patients with bipolar disorder. Using a machine learning approach, we sought to evaluate whether lithium responsiveness (LR) is predictable using clinical markers. METHOD: Our data are the largest existing sample of direct interview-based clinical data from lithium-treated patients (n = 1266, 34.7% responders), collected across seven sites, internationally. We trained a random forest model to classify LR-as defined by the previously validated Alda scale-against 180 clinical predictors. RESULTS: Under appropriate cross-validation procedures, LR was predictable in the pooled sample with an area under the receiver operating characteristic curve of 0.80 (95% CI 0.78-0.82) and a Cohen kappa of 0.46 (0.4-0.51). The model demonstrated a particularly low false-positive rate (specificity 0.91 [0.88-0.92]). Features related to clinical course and the absence of rapid cycling appeared consistently informative. CONCLUSION: Clinical data can inform out-of-sample LR prediction to a potentially clinically relevant degree. Despite the relevance of clinical course and the absence of rapid cycling, there was substantial between-site heterogeneity with respect to feature importance. Future work must focus on improving classification of true positives, better characterizing between- and within-site heterogeneity, and further testing such models on new external datasets.

Centro Lucio Bini Cagliari e Roma Italy

Charité Universitätsmedizin Berlin Berlin Germany

Charité University Medical Center Institute for Social Medicine Epidemiology and Health Economics Berlin Germany

Department of Adult Psychiatry Poznan University of Medical Sciences Poznan Poland

Department of Mental Health Poznan University of Medical Sciences Poznan Poland

Department of Pharmacology Dalhousie University Halifax NS Canada

Department of Psychiatric Nursing Poznan University of Medical Sciences Poznan Poland

Department of Psychiatry Charles University Prague Czech Republic

Department of Psychiatry Dalhousie University Halifax NS Canada

Department of Psychiatry McGill University Health Centre Montreal QC Canada

Department of Psychiatry University of Toronto Toronto ON Canada

Faculty of Computer Science Dalhousie University Halifax NS Canada

Harvard Medical School and McLean Hospital Boston MA USA

Mood Disorders Center of Ottawa Ottawa ON Canada

Section of Neuroscience and Clinical Pharmacology Department of Biomedical Sciences University of Cagliari Cagliari Italy

Section of Psychiatry Department of Medical Sciences and Public Health University of Cagliari Cagliari Italy

Unit of Clinical Pharmacology San Giovanni di Dio Hospital University Hospital of Cagliari Cagliari Italy

Zobrazit více v PubMed

Grof P. Sixty years of lithium responders. Neuropsychobiology 2010;62:8-16.

Garnham J, Munro A, Slaney C et al. Prophylactic treatment response in bipolar disorder: results of a naturalistic observation study. J Affect Disord 2007;104:185-190.

Drancourt N, Etain B, Lajnef M et al. Duration of untreated bipolar disorder: missed opportunities on the long road to optimal treatment. Acta Psychiatr Scand 2013;127:136-144.

Hou L, Heilbronner U, Degenhardt F et al. Genetic variants associated with response to lithium treatment in bipolar disorder: a genome-wide association study. Lancet 2016;387:1085-1093.

Mertens J, Wang Q, Kim Y et al. Differential responses to lithium in hyperexcitable neurons from patients with bipolar disorder. Nature 2015;527:95-99.

Stern S, Santos R, Marchetto M et al. Neurons derived from patients with bipolar disorder divide into intrinsically different sub-populations of neurons, predicting the patients’ responsiveness to lithium. Mol Psychiatry 2018;23:1453-1465.

Hui T, Kandola A, Shen L et al. A systematic review and meta-analysis of clinical predictors of lithium response in bipolar disorder. Acta Psychiatr Scand 2019;140:94-115.

Sportiche S, Geoffroy P, Brichant-Petitjean C et al. Clinical factors associated with lithium response in bipolar disorders. Aust N Z J Psychiatry 2017;51:524-530.

Grof P. Responders to long-term lithium treatment. In: Bauer M, Grof P, Muller-Oerlinghausen B, eds. Lithium in neuropsychiatry: the comprehensive guide. UK: Informa Healthcare, 2006:157-178.

Grof P, Duffy A, Cavazzoni P et al. Is response to prophylactic lithium a familial trait? J Clin Psychiatry 2002;63:942-947.

Kessing L. Lithium as the drug of choice for maintenance treatment in bipolar disorder. Acta Psychiatr Scand 2019;140:91-93.

Kim T, Dufour S, Xu C et al. Predictive modeling for response to lithium and quetiapine in bipolar disorder. Bipolar Disord 2019;21:428-436.

Nierenberg A, McElroy S, Ketter T et al. Bipolar CHOICE (Clinical Health Outcomes Initiative in Comparative Effectiveness): a pragmatic six month trial of lithium vs. quetiapine for bipolar disorder. J Clin Psychiatry 2016;77:90-99.

Manchia M, Adli M, Akula N et al. Assessment of response to lithium maintenance treatment in bipolar disorder: a consortium on lithium genetics (ConLiGen) report. PLoS ONE 2013;8:e65636.

McGuffin P, Farmer A, Harvey I. A polydiagnostic application of operational criteria in studies of psychotic illness: development and reliability of the OPCRIT system. Arch Gen Psychiatry 1991;48:764-770.

Denicoff K, Ali S, Sollinger A, Smith-Jackson E, Leverich G, Post R. Utility of the daily prospective National Institute of Mental Health Life-Chart Method (NIMH-LCM-P) ratings in clinical trials of bipolar disorder. Depression Anxiety 2002;15:1-9.

Fay MP, Shaw PA. Exact and asymptotic weighted logrank tests for interval censored data: the interval R package. J Stat Softw 2010;36:1-34.

Breiman L. Random forests. Mach Learn 2001;45:5-32.

Pedregosa F, Varoquaux G, Gramfort A et al. Scikit-learn: machine learning in python. J Mach Learn Res 2012;12:2825-2830.

He H, Garcia E. Learning from imbalanced data sets. IEEE Trans Knowl Data Eng 2010;21:1263-1264.

Lemaitre G, Nogueira F, Aridas C. Imbalanced-learn: a python toolbox to tackle the curse of imbalanced datasets in machine learning. J Machine Learn Res 2017;18:1-5.

Schnack H, Kahn R. Detecting neuroimaging biomarkers for psychiatric disorders: sample size matters. Front Psychiatry 2016;7:1-12.

Nunes A, Schnack H, Ching C et al. Using structural MRI to identify bipolar disorders - 13 site machine learning study in 3020 individuals from the ENIGMA Bipolar Disorders Working Group. Mol Psychiatry 2018;1-14.

Grof P, Alda M, Grof E, Zvolsky P, Walsh M. Lithium response and genetics of affective disorders. J Affect Disord 1994;32:85-95.

Backlund L, Ehnvall A, Hetta J, Isacsson G, Ågren H. Identifying predictors for good lithium response-a retrospective analysis of 100 patients with bipolar disorder using a life-charting method. European Psychiatry. 2009;24:171-177.

Tondo L, Hennen J, Baldessarini R. Rapid-cycling bipolar disorder: effects of long-term treatments. Acta Psychiatr Scand 2003;108:4-14.

Kessing L, Hellmund G, Andersen P. Predictors of excellent response to lithium: results from a nationwide register-based study. Int Clin Psychopharmacol 2011;26:323-328.

Kessing L, Vradi E, Andersen P. Starting lithium prophylaxis early v. Late in bipolar disorder. Br J Psychiatry. 2014;205:214-220.

Turecki G, Grof P, Cavazzoni P et al. Evidence for a role of phospholipase C-γ1 in the pathogenesis of bipolar disorder. Mol Psychiatry 1998;3:534-538.

Turecki G, Grof P, Grof E et al. Mapping susceptibility genes for bipolar disorder: a pharmacogenetic approach based on excellent response to lithium. Mol Psychiatry 2001;6:570-578.

Exploring the genetics of lithium response in bipolar disorders

Association of Attention-Deficit/Hyperactivity Disorder and Depression Polygenic Scores with Lithium Response: A Consortium for Lithium Genetics Study

Combining schizophrenia and depression polygenic risk scores improves the genetic prediction of lithium response in bipolar disorder patients

Exemplar scoring identifies genetically separable phenotypes of lithium responsive bipolar disorder