Lithium: Classic? Yes;First line? … Only sometimes

Zivanovic et al. (2017) pose the question of why lithium prescriptions have consistently decreased. The authors review the literature on lithium’s efficacy in the different phases of bipolar disorder, appropriately highlight the persuasive data on lithium’s anti-suicide effect and the emerging data on its neuroprotective effects. They acknowledge lithium’s potential side effects as well as organ toxicities, but contrast these burdens/risks with those of other mood stabilizers. At its core, the paper of Zivanovic et al. presents an impassioned plea for the restitution of lithium as the first line agent in bipolar disorders. The paper serves as a companion piece to other, recently published, papers (Nolen, 2015). Yet, as has been reviewed in more detail elsewhere (Gitlin, 2016a), there are legitimate reasons why lithium is prescribed less than it once was. Most importantly, at one time, lithium was the only mood stabilizer available to psychiatrists and patients. Now, we are blessed with a number of different anticonvulsants and antipsychotics, any one of which may be the optimal treatment for some patients.

Other than the availability of other mood stabilizers, a number of other factors explain the decreasing rate of lithium prescriptions. Citing an earlier study, Zivanovic et al. note that two of the unmet needs of bipolar disorder are better treatment of depression and better prevention of depression. This is an area in which lithium, for all its advantages, falls short of being an optimal mood stabilizer. As the authors acknowledge, the data in support of lithium as an acute antidepressant are somewhat weak. Additionally, for maintenance treatment, in a recent meta-analysis, lithium was more effective than placebo in preventing depression at only a trend level (p = 0.08) (Severus et al., 2014). It is curious that Zivanovic and colleagues did not discuss the efficacy of lamotrigine, which demonstrated somewhat greater efficacy in preventing depression compared to lithium in two studies (briefly reviewed in Gitlin, 2016a). Thus, in its effects on bipolar depression—possibly the most desired effects of a mood stabilizer from a patient’s viewpoint—lithium frequently falls short of optimal efficacy.

Furthermore, although Zivanovic and colleagues briefly review the side effects and organ toxicities of lithium, they minimize what, for many patients, are serious burdens and concerns about lithium treatment. Again, when lithium was the only mood stabilizer available, these burdens were appropriately considered as unavoidable risks for an extraordinary treatment. However, now, we have other agents—e.g., lamotrigine—that are truly weight neutral and do not require regular venipuncture and monitoring. Furthermore, there is no consistent evidence that lithium is better tolerated or associated with fewer overall side effects than other mood stabilizers (reviewed in Gitlin, 2016a).

Lithium’s potential for renal toxicity is also minimized in Zivanovic et al.’s paper. Although, as the authors correctly point out, the absolute risk for end stage renal disease (ESRD) is relatively low, lithium-treated patients have an up to eight-fold risk of ESRD (reviewed more extensively in Gitlin, 2016b). Since mood stabilizers will generally be taken for very long time periods and since lithium’s effect on renal function correlates with length of treatment, concerns about renal damage are legitimate in treating a lifetime disorder (bipolar disorder) with a lifetime treatment (lithium).

Finally, lithium’s other deleterious effects on thyroid and parathyroid gland function must also be considered. Although less concerning than the renal effects, partly because they are more easily treated, they are still substantial drawbacks to the long-term use of lithium. If we are to raise the caveat about tardive dyskinesia as a serious concern associated with the use of antipsychotics as long-term mood stabilizers more seriously (as we should), then we should acknowledge the potential long-term, irreversible effects of lithium too.

My comments should not be interpreted as being dismissive of lithium as a mood stabilizer. I have prescribed lithium for 40 years, and it has been a mainstay of treatment in the mood disorders clinic that I have directed for over 35 years. I am eternally grateful for the availability of lithium and for its (at times) extraordinary efficacy. Nonetheless, instead of simply encouraging our colleagues to prescribe more lithium, we should also acknowledge its treatment limitations: it works more slowly than antipsychotics in treating acute mania, it is less than optimally effective in treating acute bipolar depression and in preventing bipolar depression, and it is associated with both a significant side effect burden and organ toxicities that are legitimate sources of concern. Instead, let us continue to prescribe lithium for many of our bipolar patients and be grateful that other therapeutic options are available given the clinical and biological variability of our bipolar patients.

See Review by Zivanovic 51: 886–896