RESUMO
Patients diagnosed with schizophrenia display substantial heterogeneity in terms of their clinical presentations, and treatment response. Accumulating research suggests that such high diversity may reflect distinct biological subtypes with differentially affected underlying neurobiology. Novel treatments, including sex hormone estradiol treatments, provide alternative efficacious treatment avenues but also should be studied within the context of potential heterogeneity. This repeated-measures study characterised the association between hormone levels (estrogen, progesterone, testosterone, prolactin, FSH, LH, DHEA) and symptom treatment outcomes (defined by The Positive and Negative Syndrome Scale (PANSS)) across a 56-day study of 200 ug adjunctive estradiol treatment in women with schizophrenia. Group-based trajectory models was used to account for potential heterogeneity (subgroups). Receiver operating characteristic (ROC) curves were evaluated to define the predictive value of endogenous estradiol levels as a treatment-response biomarker of estradiol treatment. The results generated two subgroups; a treatment-responder group who demonstrated decreasing PANSS scores across time, and a treatment non-responder group, demonstrating stable PANSS scores across time. The treatment-responder subgroup was significantly negatively predicted by estradiol blood level (b= -2.34, SE= 1.17, p = 0.047), while FSH blood level was positively associated with the treatment non-responders (b= 7.14, SE= 2.54, p = 0.008). ROC for day 28, 56 time points yielded area under the curve of 0.52 and 0.55, respectively. Harrell's C-statistic = 0.59. This is the first study to identify endocrine markers in blood serum predicting response to estradiol treatment in female schizophrenia patients, highlighting the existence of heterogeneity of response, indicative of molecular subtypes. Characterising the differential underlying biology of the subgroups may lead to better targeted, specific treatments in the future.(ClinicalTrials.gov Identifier: NCT00357006). https://www.clinicaltrials.gov/ct2/show/NCT00357006.
