Towards the First Biomarker Test for Bipolar Spectrum Disorder: An Evaluation of 199 Patients in an Outpatient Setting

Abstract

Bipolar spectrum disorder seems to be challenging to diagnose, particularly unspecified or subthreshold types. The delay in diagnosis in the UK for bipolar I and II types is a staggering 10-13 years, with only 15% correctly diagnosed without delay. In the USA, the delay is 6-8 years, and there is a 60% incorrect diagnosis rate. The HCL-32 questionnaire is adequate, but not sufficient by itself, and patients may find it difficult to complete, particularly if they are unwell. We have investigated a biomarker test which can be used in day-to-day clinical practice to assist diagnosis. We evaluated 199 patients diagnosed with ICD-10 bipolar I, II, and unspecified disorders, using the HCL-32 questionnaire with a cut-off point of 14 and above, supplemented by history taking and examination using the principles of the CIDI 3, interviews of relatives, and longitudinal mood charts where available. The results were compared to the general population and a sample of patients diagnosed with recurrent depression for assessment of sensitivity and specificity. We evaluated four mutations of SLCO1C1, DiO1, and two DiO2alleles as potential biomarkers for bipolar spectrum disorder, and identified three mutations that exhibited high sensitivity, with rates of up to 87% and specificity of up to 46% in distinguishing bipolar spectrum disorders from recurrent depressive disorder. Additionally, mutations in SLCO1C1 and DiO1 exhibited a sensitivity of up to 86% and a specificity of up to 60% in detecting bipolar spectrum disorder compared to the general population within a clinical setting. These biomarkers have the potential to be used as a diagnostic test that is not open to subjective interpretation and can be administered even if patients are very unwell, requiring, though, the patient's consent. Further studies confirming these results are needed to compare the validity of using individual or a best combination of single nucleotide polymorphisms to identify bipolar spectrum disorders, particularly subthreshold presentations, and to differentiate them from other mood disorders such as major depression and recurrent depressive disorder.

Keywords: bipolar disorder; diagnosis; genetic testing.

Figure 1

The prevalence of mutations (at least one allele) and their corresponding 95% confidence intervals for the bipolar, general population, and major depressive disorder (MDD) populations. The mean differences in prevalence between the bipolar and general population are as follows: DiO1: 29% (23.4% to 34.6%), p < 0.0001; DiO2 Gly3Asp: 1.5% (−5.2% to 8.1%), p = 0.67; DiO2 Thr92Ala: 21.1% (16.2% to 26.1%), p < 0.0001; and SLCO1C1: 43.0% (37.9% to 48.1%), p < 0.0001. The differences in DiO2 Thr92Ala prevalence between the bipolar population and the MDD population from Poland are 33.1% (24.3% to 41.9%), p < 0.0001, and between MDD and the general population are −12.0% (−19.3% to −4.7%), p = 0.0007. Further details are shown in Table 2.