Predicting Onset of Bipolar Disorder in Children at High Risk: Part I
At the 2019 meeting of the American Academy of Child and Adolescent Psychiatry, one symposium was devoted to new research on predicting onset of bipolar disorder in children who have a family history of the disorder. Below are some of the findings that were reported.
Symptom Progression
In offspring of parents with bipolar disorder, researcher Anne Cecilia Duffy found that symptoms in the children tended to progress in a typical sequence. Childhood sleep and anxiety disorders were first to appear, then depressive symptoms, then bipolar disorder.
Different Types of Illness May Respond Best to Different Medications
Duffy’s research also suggested links between illness features and a good response to specific medications. Those offspring who developed a psychotic spectrum disorder responded best to atypical antipsychotic medication. Those with classical episodic bipolar I disorder responded well to lithium, especially if there was a family history of lithium responsiveness. Those offspring with bipolar II (and anxiety and substance abuse) responded well to anticonvulsant medications.
If parents with bipolar disorder had experienced early onset of their illness, their children were more likely to receive a diagnosis of bipolar disorder.
The offspring of lithium-responsive parents tended to be gifted students, while those from lithium non-responders tended to be poorer students.
Comparing Risk Factors for Bipolar Disorder and Unipolar Depression
Researcher Martin Preisig and colleagues also showed that parental early onset of bipolar disorder (before age 21) was a risk factor for the offspring receiving a diagnosis of bipolar disorder. Parental oppositional defiant disorder (ODD) was also a risk factor for bipolar disorder in the offspring. The emergence of depression, conduct disorder, drug use, and sub-syndromal hypomanic symptoms also predicted the onset of mania during childhood.
Conversely, sexual abuse and witnessing violence were strong risk factors associated with a diagnosis of major (unipolar) depressive disorder. Being female and experiencing separation anxiety were also factors that predicted unipolar depression.
Predicting Conversion to Mania
Researcher Danella M. Hafeman reported that mood swings (referred to in the literature as “affective lability”), depression/anxiety, and having a parent who had an early onset of bipolar disorder were linked to later diagnoses of mania. Immediate risk factors that predicted an imminent onset of mania included affective lability, substance abuse, and the presence of sub-threshold manic symptoms.
High Baseline Levels Of C-Reactive Protein Predict Better Response To Lurasidone in Bipolar Depression
In a study presented at the 2017 meeting of the International Society for Affective Disorders, Charlies L. Raison and colleagues examined whether baseline levels of the inflammatory marker C-reactive protein (CRP) affected antidepressant response to the antipsychotic drug lurasidone in bipolar depression. The participants were divided into three double-blind groups: one received 20–60mg/day of lurasidone, another received 80–120 mg/day of lurasidone, and the third received placebo over a period of six weeks. The effect was dramatic—in people with CRP levels above 5 mg/L at the beginning of the study, lurasidone (at either dosage level) had a very large effect size (d=0.85), while in people with baseline CRP levels below 5 mg/L the effect size was smaller (d=0.35).
Interestingly, 118 of the participants (24.5%) had CRP levels above 5mg/L at baseline, indicating a substantial amount of inflammation was present in a quarter of the bipolar depressed patients. Higher levels of CRP at baseline were correlated with better improvement on specific items on the Montgomery–Åsberg Depression Rating Scale (MADRS): “lassitude” (or lack of energy), “apparent sadness,” “reported sadness,” and “pessimistic thoughts.” Raison and colleagues concluded: “These findings suggest that the efficacy of lurasidone in patients with bipolar depression may in part be linked to the inflammatory status of patients prior to treatment. If confirmed in prospective investigations, [the results of a wide-range CRP assay] may prove useful as a predictive biomarker that could help optimize the use of lurasidone for the treatment of patients with bipolar depression.”
Editor’s Note: In many instances, high levels of CRP predict a poor response to treatment (such as to selective serotonin reuptake inhibitor antidepressants (SSRIs) in unipolar depression), so these findings are of considerable interest. They also suggest the untested possibility that lurasidone has anti-inflammatory effects, as those with high levels of inflammation at baseline often respond better to drugs with direct anti-inflammatory effects such as celecoxib (Celebrex) or the antioxidant N-acetylcysteine (NAC).
Inflammation Predicts Poor Response to Sleep Deprivation with Light Therapy
A 2017 article by Francesco Benedetti and colleagues in the Journal of Clinical Psychiatry reports that people with bipolar depression who have higher levels of certain inflammatory markers may have a poor antidepressant response to the combination of sleep deprivation and light therapy, compared to those with lower levels of inflammation.
The study included 37 participants with bipolar disorder who were in the midst of a major depressive episode. Of those, 31 participants (84%) had a history of poor response to antidepressant medication. The patients were treated with three cycles of total sleep deprivation and light therapy within one week, a combination that can often bring about a rapid improvement in depression.
Depression improved in a total of 23 patients (62%) following the therapy. Blood analysis showed that compared to those who had a good response, the non-responders had higher levels of five intercorrelated inflammatory markers: IL-8, MCP-1, IFN-gamma, IL-6, and TNF-alpha. Those with higher body mass index had more inflammation, indirectly decreasing response to the therapy.
Lithium Responders and Non-Responders Have Different Neuron Characteristics
After hyperpolarization (via Wikimedia Commons)
A 2017 study in the journal Molecular Psychiatry suggests that by observing the neurons of a person with bipolar disorder, you can predict whether they will respond to lithium treatment. The drug is effective in approximately 30% of those to whom it is prescribed.
Researchers led by Shani Stern and Renata Santos used stem cell research to analyze neurons from people with bipolar disorder and healthy controls.
People with bipolar disorder shared some neuron features, namely a large, fast after-hyperpolarization (a phase in which the cell’s membrane changes), which is followed by a resting period before the neuron can fire again. The large, fast hyperpolarization in people with bipolar disorder speeds up this cycle, leading to fast and sustained neuron firing. This replicated previous findings by the same researchers, which found that people with bipolar disorder are more sensitive to stimuli. In people with bipolar disorder, the threshold for a neuron to fire drops with each subsequent after-hyperpolarization.
Chronic lithium treatment reduced this hyperexcitability in some patients—and these were the patients who had a good response to lithium treatment.
Among the study participants with bipolar disorder, there were differences in the neuron profiles of those who responded well to lithium versus those who did not.
Stern and colleagues programmed a computer to recognize the electrophysiological features of neurons from lithium responders and non-responders. The computer could then analyze the neurons of a patient whose response to lithium was unknown and predict with a greater than 92% success rate whether that patient had responded well to lithium treatment.
Different Psychotherapies for Different Illness Characteristics
Psychotherapy can play an important role in treating mental illness. At the 2014 meeting of the International Society for Bipolar Disorders, researcher F. Colom gave a plenary talk indicating that just like pharmacotherapy, psychotherapy should differ depending on characteristics of the illness—both its severity and whether the patient has more manic or more depressive symptoms.
For less severe illness with more depression, Colom explained that cognitive behavioral therapy (CBT) is ideal.
Psychoeducation and family focused therapy (FFT) is recommended for intermediate severity, with a focus on maintaining remission. Family focused therapy also works for early (prodromal) symptoms, as reported by researcher David Miklowitz et al. in 2013.
Lars Kessing et al. recently reported that specialty treatment in a clinic (including psychoeducation and vigilance to breakthrough symptoms that may suggest a new episode is imminent) is highly effective following a first episode of mania.
For more severe illness, Colom recommends cognitive remediation and rehabilitation to decrease illness burden and increase functioning. Functional remediation focuses on communication, includes homework, and teaches skills such as how to deal with money, time, and organization. It also helps improve social cognition.
For the most severe illness, palliative care to relieve symptoms and decrease illness impact is recommended. Colom noted that cognitive behavioral therapy is less effective with patients who have experienced more than 12 episodes (reported by Jan Scott et al. in the British Journal of Psychiatry in 2006), as is psychoeducation (Renares et al. 2010, Colom et al. 2014). These data re-emphasize the importance of early intervention, when these psychotherapeutic approaches are more helpful. Colom stresses the importance of behavioral cognitive therapy (BCT) rather than cognitive behavioral therapy (CBT) for those late in the illness whose episodes often arrive spontaneously, unprecipitated by psychosocial stress, and one needs more behavioral approaches to the brain’s habit memory system located in the striatum, which may drive highly recurrent illness.
Biomarker Suggests Which Patients With Depression Respond to Cognitive Behavior Therapy Versus SSRI
Not every treatment for mood disorders works for every patient, and for the 60% of depressed patients whose first treatment is ineffective, this wrong guess can translate into months of suffering, wasted money, lost productivity, and risk of suicide. An important trend in treatment research is the search for biomarkers, that is, biological indicators that can predict which patients might be likely (or unlikely) to respond to a particular treatment. A 2013 study by McGrath et al. in the journal JAMA Psychiatry suggests that brain glucose metabolism is one such biomarker.
Patients with untreated major depressive disorder had their brain glucose metabolism measured and then were randomized to receive 12 weeks of treatment either with the SSRI antidepressant escitalopram oxalate (trade name Lexapro) or with cognitive behavior therapy. Low glucose metabolism in a part of the brain called the anterior insula (compared to the rest of the brain) predicted that patients would reach remission on cognitive behavior therapy and respond poorly to escitalopram, while high metabolism in the same area predicted the opposite, that patients would reach remission while taking escitalopram and respond poorly to cognitive behavior therapy.
Researchers will want to test this finding with patients over the long term, but the data from this study suggest that anterior insula glucose metabolism may be a successful biomarker that can guide initial treatment selection for patients with depression.
Lithium Increases the Volume of the Prefrontal Cortex in Responders
Studies have indicated that lithium increases gray matter and the volume of the cortex and hippocampus in patients with bipolar I disorder. A poster presented by S. Selek et al. at the 5th Biennial Conference of the International Society for Bipolar Disorders described a longitudinal study of fronto-limbic brain structures in patients with bipolar I disorder during lithium treatment.
This study reported that patients whose illness failed to respond to lithium had smaller right amygdalas than euthymic bipolar I patients or healthy controls. After treatment with lithium, those who responded well to the drug showed significant enlargement of the left prefrontal cortex and the left dorsolateral prefrontal cortex, while those who responded poorly to lithium showed decreases in the volume of their left hippocampus and right anterior cingulate cortex.
Editor’s Note: This is one of several studies that suggest a relationship between volume of brain regions and degree of response to lithium. These data add to the remarkably consistent literature suggesting that lithium may have neurotrophic and neuro-protective effects, potentially because of the drug’s ability to increase neuroprotective factors such as BDNF and Bcl-2 while decreasing cell death factors such as BAX and p53.
Family History Of Alcoholism Predicts Positive Response To Ketamine
The drug ketamine can bring about antidepressant effects rapidly when given intravenously, but these effects last only a few days. In a recent study, bipolar depressed patients with alcoholism or a family history of alcoholism in first-degree relatives had a more extended positive antidepressant response to IV ketamine than those without this history, and fewer adverse effects from the treatment. The study, published by David Luckenbaugh et al. from the National Institute of Mental Health in the journal Bipolar Disorders in December 2012, replicates similar findings in patients with unipolar depression, where positive family history of alcoholism also predicted better response and fewer adverse effects from IV ketamine.
Alcohol and ketamine have a common mechanism of action. They are both antagonists of the glutamate NMDA receptor, meaning they limit the effects of glutamate, the major excitatory neurotransmitter in the brain. This suggests a theoretical explanation for why a history of alcoholism might relate to ketamine response.
Editor’s Note: Family history appears to be linked to how patients respond to different mood stabilizers. Lithium works best in those patients with a positive family history of mood disorders (especially bipolar disorder). Carbamazepine works best in those without a family history of bipolar disorder among first-degree relatives. Lamotrigine works best in those with a positive family history of anxiety disorders or alcoholism.
Drugs that are effective in patients with a family history of alcoholism all target glutamate in the brain. Lamotrigine decreases glutamate release, while ketamine reduces glutamate’s effects at the receptor. Both decrease glutamate function or activity. Like lamotrigine, carbamazepine also decreases glutamate release and has good effects in those with a history of alcoholism.
Memantine is another mood-stabilizing drug that is an antagonist of the NMDA receptor, like ketamine and alcohol. It will be interesting to see whether memantine will also be successful in those with a personal or family history of alcoholism.
