Cross-trial prediction of treatment outcome in depression: a machine learning approach.

BACKGROUND: Antidepressant treatment efficacy is low, but might be improved by matching patients to interventions. At present, clinicians have no empirically validated mechanisms to assess whether a patient with depression will respond to a specific antidepressant. We aimed to develop an algorithm to assess whether patients will achieve symptomatic remission from a 12-week course of citalopram. METHODS: We used patient-reported data from patients with depression (n=4041, with 1949 completers) from level 1 of the Sequenced Treatment Alternatives to Relieve Depression (STAR*D; ClinicalTrials.gov, number NCT00021528) to identify variables that were most predictive of treatment outcome, and used these variables to train a machine-learning model to predict clinical remission. We externally validated the model in the escitalopram treatment group (n=151) of an independent clinical trial (Combining Medications to Enhance Depression Outcomes [COMED]; ClinicalTrials.gov, number NCT00590863). FINDINGS: We identified 25 variables that were most predictive of treatment outcome from 164 patient-reportable variables, and used these to train the model. The model was internally cross-validated, and predicted outcomes in the STAR*D cohort with accuracy significantly above chance (64·6% [SD 3·2]; p<0·0001). The model was externally validated in the escitalopram treatment group (N=151) of COMED (accuracy 59·6%, p=0.043). The model also performed significantly above chance in a combined escitalopram-buproprion treatment group in COMED (n=134; accuracy 59·7%, p=0·023), but not in a combined venlafaxine-mirtazapine group (n=140; accuracy 51·4%, p=0·53), suggesting specificity of the model to underlying mechanisms. INTERPRETATION: Building statistical models by mining existing clinical trial data can enable prospective identification of patients who are likely to respond to a specific antidepressant. FUNDING: Yale University.

Baseline trajectories of heavy drinking and their effects on postrandomization drinking in the COMBINE Study: empirically derived predictors of drinking outcomes during treatment.

The Combined Pharmacotherapies and Behavioral Interventions (COMBINE) Study sought to answer questions about the benefits of combining behavioral and pharmacological interventions (naltrexone and acamprosate) in alcohol-dependent patients. Our goals were to identify trajectories of heavy drinking before randomization in COMBINE, to characterize patients in these trajectories, and to assess whether prerandomization trajectories predict drinking outcomes. We analyzed daily indicators of heavy drinking 90 days before randomization using a trajectory-based approach. Each patient was assigned to the most likely prerandomization heavy-drinking trajectory, and the baseline characteristics of participants in the baseline trajectories were compared. The main and interactive effects of these trajectories and treatment factors (acamprosate, naltrexone, or combined behavioral intervention) on summary drinking measures during active treatment (16 weeks) were assessed. We identified five trajectories of heavy drinking prerandomization: "T1: frequent heavy drinkers"; "T2: very frequent heavy drinkers"; "T3: nearly daily heavy drinkers"; "T4: daily heavy drinkers"; and "T5: daily heavy drinkers stopping early" before randomization. Trajectory membership was significantly associated with all drinking outcomes. Patients in "T5: daily heavy drinkers stopping early" had comparable drinking outcomes to those in "T1: frequent heavy drinkers," whereas the remaining trajectories were associated with significantly worse outcomes. The baseline trajectory did not interact significantly with the treatment condition. These exploratory analyses confirmed the hypothesis that baseline trajectories predict postrandomization drinking outcomes. Interestingly, "T5: daily heavy drinkers stopping early" had outcomes that were comparable to the least severe baseline trajectory "T1: frequent heavy drinkers," and baseline trajectories of heavy drinking did not moderate the treatment effects.

Luteal-phase accentuation of acoustic startle response in women with premenstrual dysphoric disorder.

Alterations in central nervous system response to menstrual cycle-related fluctuations in neuroactive steroids are thought to underlie the emergence of negative affect in the luteal phase of the menstrual cycle in women with premenstrual dysphoric disorder (PMDD). Such changes in the neuroendocrine milieu may lead to heightened arousal and response to stress in women with PMDD. Using the acoustic startle paradigm, we sought to determine whether women with PMDD have an accentuated physiologic response to a mildly aversive stimulus during the luteal compared to follicular phase. Further, we also examined the impact of visual affective stimuli on acoustic startle response (ASR) magnitude. During the follicular and luteal phases of the menstrual cycle, acoustic stimuli (103 dB) were delivered to 15 women with PMDD and 14 healthy menstruating women of similar age. After obtaining baseline ASR, the procedure was repeated when subjects viewed pleasant, neutral and unpleasant pictures. There was a significant group by menstrual cycle phase interaction for baseline ASR magnitude, which can be attributed to the heightened startle magnitude in women with PMDD compared to healthy women during the luteal relative to the follicular phase. The direction and degree to which picture viewing modulated the startle magnitude did not vary by group or menstrual cycle phase. These data suggest that menstrual cycle phase has a powerful modulatory effect on physiologic reactivity in women with PMDD but not in healthy women. Physiologic response to affective stimuli appears to be intact in women with PMDD across the menstrual cycle.

Delta-9-tetrahydrocannabinol effects in schizophrenia: implications for cognition, psychosis, and addiction.

BACKGROUND: Recent advances in the neurobiology of cannabinoids have renewed interest in the association between cannabis and psychotic disorders. METHODS: In a 3-day, double-blind, randomized, placebo-controlled study, the behavioral, cognitive, motor, and endocrine effects of 0 mg, 2.5 mg, and 5 mg intravenous Delta-9-tetrahydrocannabinol (Delta-9-THC) were characterized in 13 stable, antipsychotic-treated schizophrenia patients. These data were compared with effects in healthy subjects reported elsewhere. RESULTS: Delta-9-tetrahydrocannabinol transiently increased 1) learning and recall deficits; 2) positive, negative, and general schizophrenia symptoms; 3) perceptual alterations; 4) akathisia, rigidity, and dyskinesia; 5) deficits in vigilance; and 6) plasma prolactin and cortisol. Schizophrenia patients were more vulnerable to Delta-9-THC effects on recall relative to control subjects. There were no serious short- or long-term adverse events associated with study participation. CONCLUSIONS: Delta-9-tetrahydrocannabinol is associated with transient exacerbation in core psychotic and cognitive deficits in schizophrenia. These data do not provide a reason to explain why schizophrenia patients use or misuse cannabis. Furthermore, Delta-9-THC might differentially affect schizophrenia patients relative to control subjects. Finally, the enhanced sensitivity to the cognitive effects of Delta-9-THC warrants further study into whether brain cannabinoid receptor dysfunction contributes to the pathophysiology of the cognitive deficits associated with schizophrenia.

The psychotomimetic effects of intravenous delta-9-tetrahydrocannabinol in healthy individuals: implications for psychosis.

Recent advances in the understanding of brain cannabinoid receptor function have renewed interest in the association between cannabinoid compounds and psychosis. In a 3-day, double-blind, randomized, and counterbalanced study, the behavioral, cognitive, and endocrine effects of 0, 2.5, and 5 mg intravenous delta-9-tetrahydrocannabinol (Delta-9-THC) were characterized in 22 healthy individuals, who had been exposed to cannabis but had never been diagnosed with a cannabis abuse disorder. Prospective safety data at 1, 3, and 6 months poststudy was also collected. Delta-9-THC (1) produced schizophrenia-like positive and negative symptoms; (2) altered perception; (3) increased anxiety; (4) produced euphoria; (5) disrupted immediate and delayed word recall, sparing recognition recall; (6) impaired performance on tests of distractibility, verbal fluency, and working memory (7) did not impair orientation; (8) increased plasma cortisol. These data indicate that Delta-9-THC produces a broad range of transient symptoms, behaviors, and cognitive deficits in healthy individuals that resemble some aspects of endogenous psychoses. These data warrant further study of whether brain cannabinoid receptor function contributes to the pathophysiology of psychotic disorders.