Dusunen Adam The Journal of Psychiatry and Neurological Sciences 2016;29:293-297 Guest Editorial / Misafir Editoryal DOI: 10.5350/DAJPN20162904001 Why is not Lithium Michael Gitlin1 1UCLA, Geffen School of Medicine, Department of Prescribed More Often? Psychiatry, Los Angeles CA-USA Here are the Reasons Address reprint requests to / Yazışma adresi: Michael Gitlin, Geffen School of Medicine, 300 UCLA Medical Plaza, Suite 2200, Los Angeles, CA 90095 E-mail address / Elektronik posta adresi: [email protected] INTRODUCTION with the conclusion that we should prescribe lithium more often (as in Professor Nolen’s (8) thoughtful ne of the current conundrums in the and wise review in this Journal recently). Yet, when a Opsychopharmacology of bipolar disorder is phenomenon-the decreased use of lithium-is usually phrased as: Why isn’t lithium prescribed more repeatedly observed, it may be wise to consider the often, especially as a maintenance treatment? After all: reasons for the observation instead of simply 1) It is our oldest and most well established agent; its exhorting our colleagues to act differently. In this efficacy has been established in many studies and article, despite my gratitude for the availability of, verified in a recent meta-analysis (1); 2) multiple experience with, and academic interest in lithium (9), Practice Guidelines from a variety of countries and I will present the counter argument, suggesting regions have consistently deemed lithium as the first answers to the question of why lithium is not line, “gold standard” of mood stabilizers; 3) befitting a prescribed more often. gold standard treatment, it is frequently utilized as an active comparator when testing new mood stabilizers How are Clinical Decisions in (2-4); 4) since it has been prescribed for over 50 years, Psychopharmacology Made? there is little worry that new long term toxicities or side effects will emerge; 5) equally, there is an Clinical decisions, made (hopefully) collaboratively astonishing amount of clinical experience with by physicians and patients, are constructed by using lithium’s use (including mine, having prescribed informal algorithms that combine four factors: lithium for 40 years, during over 30 years of which I perceived efficacy, perceived tolerability (i.e., side have directed an academic Mood Disorders Clinic). effect burden), potential toxicity and the Despite these compelling reasons to prescribe burdensomeness of treatment-e.g. need for ongoing lithium, evidence from multiple studies in both blood tests and so forth. (For the sake of the discussion, bipolar disorder and when it is used as an adjunctive I am ignoring nonclinical factors such as cost, which is antidepressant treatment demonstrate declining and/ relevant in some countries but not others). In or lower prescribing rates of lithium than would be psychopharmacology, the various options for anticipated (5-7). Frequently, this issue is reviewed treatment are then compared, either explicitly or Düşünen Adam The Journal of Psychiatry and Neurological Sciences, Volume 29, Number 4, December 2016 293 Why is not lithium prescribed more often? Here are the reasons implicitly with all four factors being considered by effective than lithium in preventing manias in another both physicians and patients. Of course, physicians study (13). Some of these studies (2,4) were enriched and patients may weigh these factors somewhat for the non-lithium comparator medication, which differently. One might assume that physicians would may have given the comparators an advantage. Other be more oriented towards efficacy whereas patients studies, however (4,13) did not employ an enriched might consider burdensomeness and side effects more design. Still, other studies demonstrated lithium’s highly than do physicians. However, ironically, some robust efficacy (14). The overall conclusion, however, data suggest psychiatrists consider side effects as is that lithium does not routinely stand out as the most more relevant in explaining nonadherence than do effective mood stabilizer and, in depression- patients (10). predominant patients, may not even be the most With these four factors in mind, let us consider effective mood stabilizer we have. how lithium may be perceived by patients and psychiatrists that explain its declining use. Tolerability of Lithium Efficacy of Lithium Most lithium treated patients, estimated between 67-90%, experience side effects (15). Additionally, the As confirmed in the recent meta-analysis (1), majority of lithium treated patients experience more lithium robustly prevents mood episodes and than one side effect. The most common side effects are specifically mania. Yet, depression is the dominant nausea and/or diarrhea, tremor, polyuria/polydipsia, pole in bipolar disorder with the average bipolar cognitive impairment and weight gain. Some side patient spending three times as much time depressed effects are more associated with treatment as manic/hypomanic (11). One large naturalistic nonadherence (e.g., weight gain, cognitive impairment) study found that bipolar II individuals had a 37:1 than others (16). Thus, side effect burden with lithium ratio of depressed vs. hypomanic weeks (12). is nontrivial. Lithium’s efficacy is more robust against mania than Comparing rates of side effects of lithium with depression. In the Severus et al. meta-analysis (1), those of other mood stabilizers is rather difficult given lithium’s efficacy prevented depression at only a the different methodologies of the studies. As trend level (p=0.08; RR=0.78, CI=0.59-1.03). Thus, examples, in enriched designs, the medication being for the typical bipolar patient for whom depression tested-e.g., quetiapine or lamotrigine- is given openly dominates the course of illness, lithium may be before random assignment to either continuing that perceived as only a weakly effective agent. medication or switching to lithium as an active Similarly, in comparing lithium to other mood comparator or to placebo (2,4). This design therefore stabilizers, lithium is not always necessarily more selects for subjects who can tolerate the experimental effective and, in some studies, is less effective than medication before the controlled phase of the study, active comparators. In the combined analysis of the thereby biasing the side effect data against lithium. In a two registrational lamotrigine studies, lithium and study using a non-enriched design, lithium was lamotrigine were equally effective mood stabilizers associated with a numerically higher discontinuation but lamotrigine statistically separated from placebo in rate compared to olanzapine (26% vs 19%) but preventing depression while lithium did not (2), a olanzapine was associated with significantly more finding that was replicated (albeit only numerically, weight gain (13). In two studies using non-enriched not statistically) by Licht et al. (3). Similarly, in the designs, lithium was associated with a higher quetiapine/lithium/placebo maintenance treatment premature termination rate for intolerance (35% vs. study, quetiapine was more effective than lithium in 22%) compared to valproate in one (17), whereas no preventing depression (4). Olanzapine was more significant differences in side effects leading to 294 Düşünen Adam The Journal of Psychiatry and Neurological Sciences, Volume 29, Number 4, December 2016 Gitlin M treatment termination were seen between lithium and tardive dyskinesia which, at times, is irreversible while valproate in the other study (14). Lithium was metabolic syndrome is a major risk with some agents associated with more side effects than lamotrigine in such as quetiapine, olanzapine and clozapine. the one non-enriched maintenance study (3). In the Of course, we have no mood stabilizers that are most recent, non-enriched study, quetiapine was devoid of any health risks. Yet, the long term worries associated with more side effect burden compared to with lithium treatment, especially the concern about lithium but no differences in discontinuation rates renal damage are legitimate concerns when treating were observed (18). bipolar patients for long time periods as is inherent for For now, then, it would be difficult to consistently a lifetime disorder. distinguish between lithium and many other mood stabilizers on side effect burden. Clinically, lamotrigine Lithium and Burdensomeness of Treatment is associated with the fewest side effects, consistent with the results of the study conducted by Licht et al. (3). More than any other mood stabilizer, lithium requires regular, albeit infrequent venipuncture to Toxicity of Lithium monitor lithium levels and parameters associated with the potential toxicities noted above-thyroid, It has long been known that lithium has toxic renal and calcium/PTH measures. Usual effects on the thyroid gland and the kidneys. The recommendations suggest monitoring between every thyroid toxicity, caused primarily by lithium’s three months to yearly (9). Although not particularly interference with thyroid hormones’ release from the frequent, these tests are mandatory elements of gland (19) affects up to 19% of treated patients (20). lithium treatment and are burdensome to many Although easily monitored and treated, lithium- patients. Monitoring of metabolic parameters is also induced hypothyroidism is a relevant clinical necessary when treating with second generation concern. No other mood stabilizer causes antipsychotics, but somewhat less frequently than hypothyroidism. lithium monitoring. Valproate