Umeå University's logo

Bipolar disorder (BD) and schizoaffective disorder (SZD) are two serious mental disorders mostly requiring life-long treatment with mood stabilisers. Lithium is a first-line treatment for BD, generally considered superior to other mood stabilisers in terms of affective relapse and suicide prevention. However, lithium is not without problems, and adherence rates vary. One concern with lithium is its narrow therapeutic index, which carries a risk of intoxications. Several other types of adverse effects are also recognised. Of these, renal impairment may be the most well-known and feared. It is mandatory to be able to understand and recognise early signs of potentially severe or even catastrophic adverse effects. Although lithium is well established, many safety aspects remain unclear and underexplored. The overall aim of this thesis was therefore to examine some of these safety aspects of lithium treatment that are clinically relevant but insufficiently explored.

In four different studies, this thesis has examined (a) the impact on toxic lithium concentrations on the cardiac conduction system, (b) the incidence, nature, and aetiology of serious adverse drug events (ADE) arising from lithium treatment compared to other psychotropic medications, (c) the reasons for multiple lithium discontinuations in patients with BD or SZD, and (d) the effect of lithium dosing regimen on glomerular and tubular kidney function, namely whether lithium given as single daily lithium dosing (SDD) causes less renal impairment than lithium given as multiple daily lithium dosing (MDD).

The first two studies were based on the LiSIE (Lithium- Study into Effects and Side Effects), a retrospective medical recordbased cohort study from Northern Sweden. LiSIE explores both effects and side effects of lithium treatment as compared to other mood stabilisers. In the first study we found that in summary statistics, on the electrocardiogram, lithium intoxication led to changes in heart rate (p=0.046), but not in QT interval (p=0.104, p=0.070). However, few patients had clinically relevant QT changes. In the second study, we found an incidence of 1.9/1000 person years (PY) for serious ADE in patients with BD and SZD. Patients on lithium treatment and older individuals were more affected (p=0.0094, p= 0.0007).

The third study was based on the South London and Maudsley Biomedical Research Centre Case Register (SLaM-BRC case register) in the United Kingdom. Anonymised medical records were reviewed with the Clinical Record Interactive Search (CRIS) application. We found that during the 11-year observation time frame, 123 patients with BD and SZD had discontinued lithium on at least three occasions. Psychiatric reasons, such as suspected lack of insight or feeling subjectively well, were the most common reasons for lithium discontinuations in this particular group. They accounted for over 70% of all lithium discontinuations.

Finally, in the fourth study, we conducted systematic review. Of 709 studies screened, a total of 20 studies were included, measuring various tubular and glomerular outcomes. No study was assessed to be of good quality for investigating the research question. Only one study met the criteria of fair quality regarding tubular outcome. Confounding factors such as the daily total lithium dose were rarely considered.

Our results indicate that cardiac adverse effects during lithium intoxication seem mostly discrete, but can occur. The same applies to serious ADE related to the treatment of BD or SZD; the overall incidence was classified as uncommon but not rare. However, older individuals and lithium treatment seem to carry a higher risk of developing serious ADE. In multiple lithium discontinuers, lithium seems mainly discontinued for psychiatric reasons rather than adverse effects, leading to a subsequent restart due to affective relapse. A better understanding is needed of how to best promote adherence in this group of patients, to prevent a perpetual cycle of remitting when on lithium and relapsing when off lithium. Lastly, based on the low quality of the available studies, the question of whether the lithium dosing regimen is relevant for the protection of kidney function cannot be decided. Hence, there is no current evidence supporting the assumption that lithium given as SDD would be to be superior to MDD for the prevention or containment of renal impairment.