Clinical evaluation of switching from immediate‐release to prolonged‐release lithium in bipolar patients, poorly tolerant to lithium immediate‐release treatment: A randomized clinical trial

Abstract Aim The effect of switching from lithium immediate release (Li‐IR) to lithium prolonged release (Li‐PR) on lithium‐induced tremor after 1 and 12 weeks of treatment was evaluated in a randomized, multicenter, open trial, in bipolar patients from the participating sites with a tremor severity ≥2 (Udvalg for Kliniske Undersøgelser [UKU] rating scale) despite optimal lithium titration. Methods The primary endpoint was the evaluation of tremor by means of the UKU scale after 1 week of treatment. Secondary endpoints included manic Young Mania Rating Scale (YMRS) and depressive symptoms (Montgomery–Asberg Depression Rating Scale), a global assessment of the patient's status (Clinical Global Impression), polyuria/polydipsia (UKU item 3.8) and patient‐reported outcomes. Results Owing to difficulties in including suitable patients the enrollment phase was closed when 73 patients were randomized. Notwithstanding the lower number of patients, in the modified intention‐to‐treat population (n = 70) the primary endpoint was statistically significant: tremor improved after 1 week in 62.9% in Li‐PR group against 20.0% of patients in Li‐IR group (p = .0006; two‐tailed Fisher's exact test). The difference remained statistically significant after 4 (p = .0031) and 12 weeks (p = .0128). The same analysis performed in the PP population confirmed these results. Among the secondary endpoints, only the factor convenience of the treatment satisfaction questionnaire showed a statistically significant difference between groups. There were no apparent differences in the safety profile of the two formulations. Conclusions This study is the first comparative documentation of a potential benefit of the prolonged‐release formulation in reducing the symptom tremor, a well‐known adverse effect of lithium therapy. Indeed, the study results should be interpreted taking into account the sample size lower than planned.

studies showing wide variability from 4% to 65%. Lithium tremor is classified as a postural tremor; it is generally symmetric, typically produced by voluntary maintenance of a particular posture held against gravity. It is usually limited to hands or upper limbs at resting state but worsens during activities that require fine motor control such as writ-ing or pouring water. Lithium tremor typically appears when started or titrated. Tremor can cause serious problems in patients' daily lives, can be troublesome and socially embarrassing and might eventually cause noncompliance for treatment (Baek et al., 2014;Burgess et al., 2001). Tremor is the most frequent reason given by patients for lithium discontinuation. Even though clinicians may use beta-blockers to treat lithium-induced tremor, their potential side effects, together with their limited efficacy in reducing tremor, prompt for investigations to confirm that prolonged-release lithium formulations might be devoid of tremor side effect and improve compliance (Morgan & Sethi, 2005;Shoutsuki & Terao, 2014).
Both prolonged-release (PR) and immediate-release (IR) preparations of lithium salts are available in the EU market. The PR formulation may allow preventing high fluctuations in lithium plasma levels and high plasma peak levels, thus providing more stable plasma concentrations compared to IR formulations. The sulfate salt of the PR formulation shows a pH-independent solubility (unlike carbonate and citrate salts in the IR formulation) that may contribute to reduce interindividual and intraindividual variability in gastro-intestinal absorption. Such variability is usually high and may favor lithium adverse effects or intoxication, despite the regular monitoring of lithium blood levels. Moreover, some lithium-related adverse events appear to be a function of the rate of increase of serum lithium concentration, which has implications regarding the type of lithium formulation prescribed for a patient (Grandjean & Aubry, 2009a). Furthermore, the expected better safety profile of the PR tablets due to the reduced fluctuations in lithium plasma concentrations, and the potential improvement of the patient's compliance related to the fewer daily administration, might represent advantages over the IR formulations.
In a previous observational study, patients who switched from multiple daily administrations of lithium IR to once-daily lithium PR declared their preference for the PR preparation over traditional one for its better tolerability and ease of administration. The occurrence of side effects in fact was significantly reduced among patients receiving the PR versus IR lithium preparation (Durbano et al., 2002).
Engagement and development of a therapeutic alliance are important in any lifelong disorder that needs long-term adherence (Grande et al., 2016). The effectiveness of interventions aimed at increasing adherence to therapy might be helpful. Consequences of nonadherence can be serious leading to poor outcomes, worsening of the quality of life, functional impairment, and increased risks of relapse, rehospitalization, and suicidality (McIntyre & Calabrese, 2019).
Despite the fact that a prolonged-release formulation has been on the market in several European countries since several years, there is a lack of well-controlled clinical studies documenting the effectiveness and safety of Li-PR in comparison to the plain formulation (Girardi et al., 2016;Martinotti et al., 2018;Nolen et al., 2019).
The present study was designed to compare the effect on tremor and other adverse effects (e.g., polyuria and polydipsia) of the switch from lithium carbonate IR formulation ((Carbolithium® immediaterelease 150 mg and 300 mg capsules, TEVA) to lithium sulfate PR formulation (Lithiofor ® prolonged-release 660 mg tablets, Vifor SA / Resilient™ 660 mg prolonged-release tablets, Angelini S.p.A), in patients affected by bipolar disorder who showed to be poorly tolerant to the lithium IR treatment. The primary objective of the study was the evaluation of the change in the lithium-induced tremor when switching from lithium IR to lithium PR formulation. The secondary objectives included the assessment of:

Assessment of efficacy
The evaluation of tremor was carried out at screening and baseline and after 1, 4, and 12 weeks of treatment by means of the UKU side effects rating scale for the registration of unwanted effects of psychotropics (Lingjaerde et al., 1987). The choice to adopt the single item 2.5 of the UKU scale for the assessment of tremor was led by the lack of specific instruments focused on this symptom. The primary endpoint was evaluated as the proportion of patients with an improvement in tremor assessed by a single item (2.5 tremor) of the UKU rating scale after 1-week treatment period compared to baseline, improvement being defined as a difference ≥1 between scoring at baseline and scoring after 1 week of treatment. The timing for the assessment of the primary endpoint was chosen considering that a treatment period of 1 week with a PR preparation of lithium is considered adequate to reach stable lithium plasma levels. To enhance the relevance of this clinical measurement, the same efficacy assessment was evaluated as secondary endpoint at 4 and 12 weeks.
The secondary efficacy assessments were the following: (i) the effect on the lithium-induced tremor up to 12 weeks; (ii) the evaluation of the changes in manic and depressive symptoms from baseline to 1, 4, and 12 weeks, as assessed by the YMRS (Young et al., 1978)  Patients were asked to rate their satisfaction with treatment by filling in the patient's treatment satisfaction questionnaire (TSQM, version II). The 14-item TSQM is a reliable and valid instrument to assess patients' satisfaction with medication, providing scores on four subscales-side effects, effectiveness, convenience, and global satisfaction (Atkinson et al., 2004(Atkinson et al., , 2005. Quality of life was assessed by means of the Q-LES-Q-SF a reliable and valid clinical instrument for the assessments of quality of life (Endicott et al., 1993;Mauri et al., 2008).

Assessment of tolerability
Safety assessments consisted of: (i) the assessment of polyuria/ polydipsia (single item 3.8 polyuria/polydipsia of the UKU side-effect rating scale) at baseline and after 1, 4, and 12 weeks of treatment, (ii) collection of lithium plasma/serum levels, (iii) all adverse events (AEs) and serious adverse events (SAEs), with their severity and relationship to study drug, and (iv) assessments of vital signs throughout the study.
Monitoring of hematology, blood chemistry, and urinalysis performed at study centers, physical conditions and ECG were also performed at screening and at week 12. Changes in concomitant medications were collected and analyzed. Serum lithium concentrations were monitored at each scheduled visit (after 12 h post dose) and assessed as plasma or serum lithium concentrations, according to the standard methods used at the local laboratory of the centers.

Sample size determination
Aim of the study was to demonstrate an improvement in the proportion of patients with lithium-induced tremor when switching from lithium IR to lithium PR. A sample of 110 evaluable patients (55 per treatment group) was considered to be sufficient to detect a clinically important difference of 20% between groups in the item 2.5 tremor of the UKU side-effect rating scale after 1 week compared to baseline (expected improvement 5% in the Li-IR group, and 25% in the Li-PR one) using Fisher's exact test with a power of 80% and a 5% two-sided significance level (nQuery Advisor). Considering a drop-out rate of patients equal to 20%, a total number of 138 patients (69 per treatment group) had to be enrolled.

Statistical analysis methods
Descriptive statistics was applied. Mean with standard deviation and 95% confidence intervals (CI) for variables subjected to statistical inference were calculated for quantitative variables. For qualitative variables, counts and percentages and 95% confidence interval (Wald method) were provided. All baseline characteristics were summarized by groups and overall, for the modified intention-to-treat (ITT) population.
The primary endpoint was analyzed using a Fisher's exact test to compare the proportion of patients with an improvement on item 2.5 tremor in the two treatment groups after 1 week, at a statistical significance of 0.05 (two sided).
For the secondary endpoints tremor (UKU item 2.5) and polyuria/polydipsia (UKU item 3.8), the same analysis was per- The Safety Population (SP) included all randomized subjects who took at least one dose of the study treatments. The PP population consisted of 50 patients. Main reasons for exclusion were nonevaluable compliance or compliance <80% and intake of not allowed concomitant medications.

Patients
One patient was randomized but did not take any dose, leaving 72 patients in the Safety population.
Thirty-two and 27 patients in the Li-IR and Li-PR arm, respectively, completed the 12-week treatment period. Patient disposition is displayed in Figure 1.
Overall compliance to study medication was ≥80% in 91.7% and 94.5% of patients in the LI-IR and in the Li-PR group, respectively.

Sample characteristics
The clinical characteristics of patients included in the m-ITT are pre-

Efficacy results
Primary and secondary efficacy results are described in Tables 2 and 3. In the m-ITT, after 1 week of treatment an improvement in tremor (assessed by item 2.5 tremor of the UKU side-effects rating scale) versus baseline was achieved in 7 (20.0%) patients out of 35 in Li-IR There was a progressive increase in the number of patients with an improvement in tremor over time in the Li-IR group while the number of improved patients in the Li-PR tended to remain stable.
In the PP population the difference was statistically significant after 4 weeks (p = .0145) but not after 12 weeks (p = .2317).
The symptom polyuria/polydipsia improved after 4 weeks in 6 (17.1%) patients out of 35 in the Li-IR group against 6 (21.4%) patients out of 28 in the Li-PR group (p = .7523; two-tailed Fisher's exact test).
The evaluation of the changes in depressive symptoms (MADRS) from baseline after 1, 4, and 12 weeks of treatment for the m-ITT showed a different qualitative pattern of the three changes between the two treatment groups (statistically significant interaction "Time by treatment"); indeed, in the Li-IR treatment there were three increasing changes whereas in the Li-PR treatment there was an increasing change followed by a decreasing one and by an increasing change ( Figure 2, Table 2). So, the interaction significance appears due mainly by chance rather than a different expression of the two treatments.
Furthermore, the changes are not correlated with the baseline values (not statistical significance for the covariate on the "Between" and "Within Subjects Effect").
Otherwise, the mean values of changes in manic symptoms (YMRS) over time (baseline, 1, 4, and 12 weeks) were similar (interaction "Time by treatment" statistically significant) ( Table 2) with not statistically significant difference between the two treatment groups. In addition, the changes were correlated with the baseline values (statistical significance for the covariate on the "Between" and "Within Subjects Effect").
The changes over the time of the TSQM domains effectiveness, side effects, and global satisfaction did not show significant differences over time and between the treatment groups. Otherwise, for the domain convenience the difference between mean changes of the values over time turned out to be statistically significant (p = .0012), being greatly increased in the Li-PR treatment.

Safety results
The treatment groups did not differ in terms of occurrence of adverse events, which was overall low and expected for this class of drugs and

DISCUSSION
To our knowledge, this study is the first randomized, controlled, comparative documentation of a potential benefit of the prolonged-release lithium formulation in reducing the severity of tremor, a well-known adverse effect of lithium therapy, in bipolar disorder patients, when shifting from an immediate-release formulation.
According to the scientific literature, several potential advantages of the PR lithium formulations over the IR formulations in the treatment of bipolar disorder were reported. Such reports pointed to more stable lithium serum concentrations while reaching the effective interval and, on the long term, a reduced risk of some peak-related adverse events and a more convenient dosing regimen that may also improve adherence to therapy (Girardi et al., 2016;Martinotti et al., 2018;Nolen et al., 2019).
Despite the fact that lower than planned patients were enrolled in the study, results seem to confirm the literature data. The analysis of the primary endpoint allowed to confirm the benefit of switching from an immediate to a prolonged-release formulation of lithium in terms of improvement of the symptom tremor. Tremor improved after 1 week and the benefit was maintained up to week 12, although the size of the difference between the two formulations decreased over time as a result of an increase in the number of patients with an improvement in tremor in the Li-IR group. Indeed, this could be expected as it is well known that lithium-induced tremor tends to decrease over time for immediate-release formulations of lithium (Schou et al., 1970).
Lithium tremor is generally related to dose and lithium blood levels (Baek et al., 2014). Nevertheless, in this study, the improvement in tremor did not seem to be related to lower plasma/serum lithium levels in the Li-PR since there were no noticeable differences in the mean plasma/serum lithium concentrations between the two groups (data not shown). Rather, the observed differences between groups may be related to different peak concentrations and not to the overall exposure. Indeed, registrative pharmacokinetic studies documented the slower increase in serum lithium concentrations and lower peak serum concentration values reached with the lithium sulfate PR formulations. However, since peak concentrations were not measured in this study, the correlation was not documented.
A raised observation to the study results was that assessing tremor at different times after each dose, as well as measuring lithium levels at different times, could help determine the relationship between tremor, dosage, and formulation.
Although we agree with this observation, it should be noticed that this was a Phase IV trial and the assessments were performed once during the day according to the routine clinical practice, at baseline and at all scheduled visits. In fact, according to the disease severity, patients included in this clinical trial were not hospitalized, thus it would have been not feasible to assess tremor at different times. To avoid bias, tremor assessment was recorded by a blinded assessor, at the same time of the day at each visit.
The switch from an IR to a PR lithium formulation did not affect the efficacy of the treatment: the two groups showed similar mean values of changes over time for the MADRS and the YMRS. This is expected because, even if prolonged-release formulations provide more stable, consistent serum drug concentrations than immediate-release formulations, overall exposure is essentially equivalent over the dosage interval and with long-term administration (Grandjean & Aubry, 2009a).
The analysis of the TSQM factors effectiveness, side effects, and global satisfaction did not disclose statistically significant differences between groups. This could be explained by the fact that misattribution of symptoms as side effects of the mood stabilizers with symptoms of disease is frequent. Indeed, patients' perception of apprehension of side effects, as opposed to the actual presence of side effects, in the most severe cases may contribute to nonadherence (Burgess et al., 2001). Instead, the TSQM factor convenience was greatly increased in the Li-PR treatment. This could be particularly important as the use of lithium PR formulations administered once daily has been suggested as one strategy (along with dosage reduction and combination therapy) to reduce nonadherence (Gitlin, 2016). Indeed, in an observational study in 47 patients who were switched from multiple daily administration of lithium IR to once daily administration of a PR carbolithium formulation, a preference for better tolerability and ease of administration of lithium PR over lithium IR was reported (Durbano et al., 2002).
From the extensive safety evaluations collected in the study there was no indication of potential safety issues when switching from an IR to a PR lithium formulation. In the management of bipolar disorder, the primary goals are to ensure the safety of the patient and to achieve clinical and functional stabilization with minimum adverse effects, as this is a lifelong disease.
The study results should be interpreted with caution, considering the lower than planned sample size and the relatively short observation period. Further long-term clinical data from larger patients' cohorts may be needed to corroborate the above results.