All About Lithium (in bipolar I)
Why is lithium most likely to be the best mood stabilizer for bipolar I?
1) Anti-Suicidal Properties
Lithium has anti-suicidal properties not found in other pharmacological treatments for bipolar I (Guzetta et al., 2007; Goodwin et al, 2003; Angst et al., 2005).
Baldessarini, Lithium and Suicide Meta-Analysis, 2006.
Baldessarini, Lithium and Suicidal Behaviors in Mood Disorders, 2016.
2) Excellent Lithium Responders
Paul Grof introduced the term “Excellent Lithium Responders” (ELRs) and defined it as bipolar patients who have no future mood episodes in response to lithium (Grof, 1999).
Data from longitudinal studies in the 1970s and 1980s suggest that approximately one-third of patients with bipolar disorder may be "Excellent-Lithium-Responders”(Rybakowski et al., 2001).
Link here: Rybakowski, Lithium Longitudinal Data, 2001.
There are many possible critiques of this data set (e.g. sample bias), but, even if the findings are significantly diminished, no other drug for bipolar disorder can approximate this magnitude of prophylactic efficacy.
3) True Mood Stabilizer
For a drug to be a true “mood stabilizer,” it has to treat acute bipolar depression and bipolar mania; and prevent both bipolar depression and bipolar mania (Goodwin and Jamison, 1990; Calabrese and Rapport, 1999; Bauer and Mtichner, 2004).
Lithium demonstrates effectiveness across all four of these domains with clinically meaningful effect sizes (Goodwin and Jamison, Manic-Depressive Illness, 2007).
What’s lithium mechanism of action?
No one knows.
However, we do know that lithium - at therapeutic levels - has neurotrophic and neuroprotective effects.
Lithium has been found to increase hippocampal and cortical gray matter volume (Post, 2016); and, in animal models of neurodegenerative and neurovascualar disorders, lithium decreases lesion size (Post, 2016).
A patient with bipolar I disorder has difficulty tolerating Abilify and would like to switch to lithium; how should you make this switch?
For people with severe mental illness (SMI), unless the side effects are unbearable or medically dangerous, I recommend doing a plateau cross-taper rather than a direct cross-taper. “Plateau” means you uptitrate your patient to a therapeutic dose of the new drug prior to downtitrating the old drug.
The disadvantage of a plateau cross-taper is that your patient may have more side effects from taking two medications at therapeutic doses.
The advantages are that you minimize the risk of a your patient not being on a therapeutic medication; and you can more easily reverse or modify the process if the new medication fails.
For the patient above (assuming we have baseline lithium labs), a plateau cross-taper could look this:
1) Start lithium carbonate 300 mg at bedtime; if well-tolerated, increase by 300 mg every 3 to 5 nights to a target dose of 900 mg at bedtime.
2) After 3 nights on lithium carbonate 900 mg, decrease aripiprazole by 5 mg per night until discontinued.
Why should you NEVER (except for a medical emergency) discontinue lithium abruptly?
Baldessarini and colleagues found that discontinuing lithium (< 2 weeks) rather than tapering patients off of it (> 2 weeks) more than doubles the risk of a recurrent mood episode within 1 year; and increases this risk by twenty-fold within 3 years (Baldessarini et al., 1996). Baldessarini also found that abrupt lithium discontinuation compared to taper confers two times the risk of a suicidal act (Baldessarini et al., 1999).
So taper your patients off of lithium gradually, meaning a minimum of 15 to 30 days.
Click here for link to the Baldessarini study: Baldessarini, Lithium Discontinuation, 1996.
What's the right lithium level?
Lithium levels should be individualized for every patient by empirically exploring the effectiveness versus the side effects within a range of 0.6 to 1.0 mEq/L.
The most recent guidelines recommend a range of 0.6 to 0.8 mEq/L and note that levels of 0.8 to 1.0 mEq/L may confer greater efficacy but also produce a significantly higher side effect burden (Goodwin et al., 2017).
Link here: Goodwin et al., Evidence-based guidelines for treating bipolar disorder: Revised third edition recommendations from the British Association for Psychopharmacology, 2016.
Here are some data-based and expert-based ways to think about target dosing for lithium:
1) Maj and colleagues randomized patients to lithium doses ranging from 0.3 to 0.9 mEq/L and found that lithium reduced affective episodes at 0.46 mEq/L and above (Maj et al., 1986).
2) Mogen Schou, who established the efficacy of lithium for mania (Schou, 1954), recommended a range of 0.5 to 0.7 mEq/L (Goodwin and Jamison, Manic-Depressive Illness, p. 804, 2007).
3) A review of the lithium literature by Severus concluded that lower levels of lithium may be better for the prevention of depression and higher doses may be better for the prevention of mania, yielding a recommended range of 0.5 to 0.8 mEq/L (Severeus et al., 2005).
How long does lithium take to start working?
Usually between 1 and 5 weeks, and it depends on the phase of the illness, with antidepressant effects lagging behind anti-manic effects (Goodwin and Jamison, Manic-Depressive Illness, 2007).
This is why for a patient with acute mania we typically combine lithium with an antipsychotic, which has a much more rapid onset of action.
Or, you could consider combining lithium with Depakote using the following Depakote loading strategy for rapid action:
1. 30 mg/kg of Depakote on days 1 and 2 in divided doses
2. Then 20 mg/kg of Depakote on days 3 through 10 in divided doses
3. Then adjust at MD's discretion
1. Prior to beginning Depakote loading, obtain platelet counts and LFTs
2. Obtain Depakote levels on days 3, 5, and 10 with target level of 90 to 100 pg/mL
1. Severe liver disease
4. Urea cycle disorder
5. Proven allergy to valproic acid, valproate, or divalproex (Depakote)
Source: Hirschfeld et al: The safety and early efficacy of oral-loaded divalproex versus standard-titration divalproex, lithium, olanzapine, and placebo in the treatment of acute mania associated with bipolar disorder. J Clin Psychiatry. 2003 Jul;64(7):841-6. Other supporting studies: Bowden et al (1994); Hirschfeld et al (1999); Zajecka et al (2002).
Should you dose lithium once or twice a day?
I recommend once a day dosing at bedtime for the following reasons:
1) Once a day dosing is easier for the patient = decreases non-compliance.
2) Once a day dose may be less damaging to the kidneys (Gitlin, 1999).
3) Once a day bedtime dosing often reduces side effects, such as cognitive dulling, which correlate with peak levels (Goodwin and Jamison, Manic-Depressive Illness, 2007).
What labs should you order when prescribing lithium?
Baseline: TSH, BUN/Creatinine, antithyroid peroxidase antibody titers
2 weeks: lithium level, TSH, BUN/Creatinine
6 months: TSH, BUN/Creatinine
Then yearly: TSH, lithium level, chem 7
*The reason for ordering antithyroid peroxidase antibody titers is that lithium can cause autoimmune thyroiditis, which can manifest as hypothyroidism or hyperthyroidism (UpToDate, Lithium and the thyroid, 2017).
What side effects of lithium do you need to tell your patients about? How should you explain them? And how should you document this?
I recommend giving your patients a handout and using it to teach them about lithium's side effects and drug-drug interactions.
Link/Handout* = a handout I wrote for patients in list format: Soskin DP, Side Effects and Drug-Drug Interactions with Lithium, 2017.
*Less acute but clinically significant side effects of lithium include: upset stomach, diarrhea, weight gain, acne, edema, polydipsia, polyuria, fine hand tremor, tiredness, cognitive dulling.
Or smartphrase: .edlithium
After you’ve done this, you can autopopulate your note with the following smartphrase:
Informed consent disclosed for lithium; discussed toxicity, drug-drug interactions, rebound mood effects with discontinuation, gastrointestinal, renal, endocrine, and cognitive side effects; non-compliance. Warned. Handout given.
What should you do if your patient's creatinine levels increase?
If they are above 1.6 mg/dl or have increased by greater than 25% of the pretreatment baseline, refer your patient to his or her PCP for a more comprehensive renal evaluation (Goodwin and Jamison, Manic-Depressive Illness, 2007).
If the results of that evaluation indicate that lithium is the cause, share the prognostic data below with your patient:
1) Approximately 10 to 20% of patients taking lithium for > than 10 years will have a decline in renal function; however, progression to renal failure secondary to lithium is rare (Kallner et al., 2000).
2) Risk factors for progression to renal toxicity on lithium include having other medical conditions (e.g. hypertension); the duration of lithium exposure; and the cumulative dose (Presne et al., 2003; Bocchetta et al., 2013).
Then together you should decide whether to continue treatment at the same dose, lower the dose, or gradually taper and discontinue lithium.
What is the most common presentation of acute (e.g. a lithium naive patient overdoses on his father’s lithium) or acute-on-chronic (e.g. a patient who has been taking lithium for 10 years intentionally overdoses on lithium) lithium toxicity?
Gastrointestinal symptoms = Nausea, vomiting, diarrhea.
Neurological signs tend to develop later in the course of toxicity because of the additional time required for lithium to reach and penetrate the blood-brain barrier compared to its more immediate/direct effects on the GI system.
What is the most common presentation of chronic lithium toxicity?
Per UptoDate, "In contradistinction to patients with acute lithium poisoning, in whom neurologic signs develop late, patients on chronic lithium therapy develop toxicity gradually and often present with neurologic findings."
Mild lithium toxicity often corresponds to mild neurologic signs, such as a tremor, slurred speech, and mild lethargy.
As the toxicity increases, patients will exhibit pyramidal, extrapyramidal, cerebellar signs.
Severe lithium toxicity can produce seizures, stupor, coma, permanent neurologic damage, or death.
Source: UpToDate, Lithium Poisoning, 2017.
At what serum level can patients get severely lithium toxic?
Lithium toxicity is defined as serum levels > 2.5 mEq/L (Sands J, 1999).
However, there have been many cases of lithium toxicity that do not correlate with serum levels (Speirs J, Hirsch SR, 1978; Venkatarathnamma et al., 2011; Peng J, 2014), and, therefore, the adage - “treat the patient and not the lab value” - applies.
Given lithium's low therapeutic index (i.e. the ease with which a patient could kill himself or herself by overdosing), is it safe for you to prescribe 30-day supplies to a patient with acute on chronic suicide risk?
Either hospitalize your patient or prescribe limited supplies.
How can you motivate a patient to try lithium if he or she has heard bad things about it or previously stopped it because of side effects?
1) Point out that many of the side effects he or she may have heard about or experienced are dose-dependent. Back in the day, psychiatrists dosed lithium much higher than we do now (early studies recommended levels of 1.2 mEq/l), so you can offer a gentler and lower dose titration.
2) Recommend that your patient read the book an Unquiet Mind by Kay Redfield Jamison: