Discontinuation of two-injection aripiprazole long-acting regimen: a pharmacological and psychopathological analysis

Objectives

A double-injection regimen for initiating aripiprazole long-acting injection (ALAI) has been approved for use based only on a pharmacokinetic modelling study. This regimen involves administering two injections of ALAI plus a single oral dose of aripiprazole, as opposed to the original regimen which called for four weeks of oral treatment with one injection. To date, there have been no studies evaluating the relative effectiveness of the double injection regimen compared with a single ALAI dose start.

Methods

We carried out a non-interventional, year-long observational study of the single and double-injection regimen for ALAI initiation. We analysed time to discontinuation with both regimens and noted reasons for discontinuation.

Results

Data were collected on 173 patients initiated on ALAI (43 double-injection, 130 single-injection initiation). Median time to discontinuation of ALAI was 111 days (95%CI 56 – 254) in the double-injection group, and 272 days (95%CI 202 - > 365) in the single-injection group (p = 0.004). For inpatients (n = 118) initiated on ALAI, median time to discharge was 14 days in the double-injection group, and 21 days in the single-injection group. There was no clear difference in discontinuations due to adverse effects.

Conclusion

Patients initiated with a double-injection regimen discontinued treatment earlier and to a greater extent than those given single-injection initiation. It is not clear that this difference relates to the regimens themselves or the reasons for choosing the particular regimen (e.g. prior non-compliance; reluctance to accept treatment).

Introduction

The long-acting injectable form of aripiprazole (ALAI) has been shown to be effective at reducing relapse, hospital admissions and bed days in patients with schizophrenia ¹. Initially, the licensed initiation of ALAI was 14 days of oral aripiprazole, followed by a single ALAI dose of 400mg and a further 14 days of oral aripiprazole. A modified regimen was approved in the UK and EU in 2020 ². The regimen is as follows: two intramuscular injections of 400mg each into separate sites, in addition to a single 20mg oral dose of aripiprazole. The approval of this regimen was based on a pharmacokinetic modelling study ³. To date, one observational study has investigated adverse effect reporting of this regimen ⁴. However, there are currently no clinical studies that have explored the relative effectiveness of this regimen compared with the original regimen. Discontinuation of long-acting injectable antipsychotics is influenced not only by pharmacokinetic properties but also by the psychopathological profile of patients ⁵. Negative symptoms, impaired insight, paranoia, and cognitive dysfunction can contribute to early discontinuation, particularly in treatment-resistant patients ⁶. Thus, we conducted a non-interventional observational study in a large Inner London hospital investigating clinical outcomes and characteristics of patients a year after receiving the double regimen.

Materials and methods

Data were collected from the South London and Maudsley (SLaM) National Health Service (NHS) Foundation Trust, an NHS Foundation Trust specialised in mental health which comprises of four psychiatric hospitals in the London Boroughs of Lambeth, Southwark, Croydon and Lewisham. Ethical approval was received from the Drugs and Therapeutics Committee (DTC) for this study (approval number: DTC/2022/36). Our standard method had previously been designated by the DTC as not requiring informed consent as non-identifiable data was used. Moreover, those viewing patient-specific data as part of the data collection process were clinicians who had access to these data in the course of their normal working practice. To identify patients who were initiated with the double-injection regimen (DIR), we used pharmacy dispensing records to produce a report on instances where two ALAI were dispensed in one instance. We used the electronic patient records to confirm administration of the two-injection regimen for the DIR group, and for data collection in the follow up. The methodology of the study is displayed in Figure 1. As the initiation regimen was approved for use on 27^th October 2020, patients who received ALAI between 1^st November 2020 and 1^st March 2022 were included in this study. Since the new loading regimen can be used to reload patients onto ALAI when they have missed a maintenance dose, patients who were previously maintained on ALAI and required reloading were also included. Patients who had received ALAI in the 8 weeks prior to initiation were excluded from this study.

For comparative analysis, we collected data on patients who had received the single injection regimen (SIR) of aripiprazole in the same time period as the DIR group. Data were collected using the Clinical Record Interactive Search (CRIS) system developed for use within the National Institute for Health and Care Research (NIHR) Maudsley Biomedical Research Centre ⁷. The CRIS system allows regulated access to anonymised information extracted from the SLaM electronic clinical records system. A search of the CRIS system was carried out to identify those initiated on ALAI via the SIR from 1^st November 2020 and 1^st November 2021 for inclusion in this study. Follow up for the SIR group was carried out using CRIS. Ethical approval for the use of CRIS as a research data-set was given by Oxfordshire Research Ethics Committee (Project ID 22-028).

Baseline demographic details were collected from electronic patient records. The date of initiation of ALAI was recorded. We recorded the care setting (inpatient/outpatient) and Mental Health Act status at the time of initiation. “Duration of illness” was defined as time from first acceptance of referral to secondary care mental health services to time of ALAI initiation. “Treatment-resistance” was defined as having received or been considered for clozapine prior to initiation of ALAI. “Previous antipsychotic” was defined as the most recent non-aripiprazole antipsychotic given within 3 months prior to initiation of ALAI. Data were collected retrospectively one year after initiation of treatment. Patients who were lost to follow-up e.g. transferred out of the Trust, passed away, discharged from SLaM etc. were censored at the point they were lost to follow-up.

Primary Analysis

To assess the effectiveness of the DIR, the primary objective of this study was to evaluate discontinuation of ALAI. Discontinuation of ALAI was defined as eight weeks without receiving ALAI. The date of initiation and discontinuation was recorded. Patients who remained on ALAI at one year after initiation were censored. Patients who received maintenance doses of 400mg or 300mg ALAI monthly were included.

Secondary Analysis

To evaluate the tolerability of the new regimen, clinical notes were reviewed to gather reasons for discontinuation of ALAI. Reasons for discontinuation were grouped into the following categories: patient choice, inefficacy, side effects/tolerability, other/unclear. Time to admission to psychiatric hospital following initiation was also recorded, and for those who were already an inpatient in a psychiatric hospital (referred to as “inpatients” for the purpose of the study) at the time of initiation 1-year follow-up for admission began at the time of discharge. For those who were initiated on ALAI as an inpatient, time to discharge following initiation of ALAI was recorded.

Statistical Analysis

Descriptive statistics were used to summarise baseline demographics, and highlight any observed differences between the two groups. For continuous variables such as age and duration of illness, the mean and standard deviation (SD) were calculated. For categorical variables, the percentage was calculated. A Kaplan-Meier estimate was used to assess the time to discontinuation and time to admission. A log rank test was used to produce a p value for the Kaplan-Meier estimates. Statistical analyses were performed in R ⁸.

Patient Demographics

The number of patients identified through data collection is displayed in Figure 2. In total, 173 patients were initiated on ALAI within the time period stated. 43 patients were initiated with the DIR, and 130 with the SIR. Baseline characteristics for those included in the study are in Table I. There was a similar proportion of male patients in the SIR group (67.7% compared to 62.8% in the DIR group). The most common diagnosis in both groups was schizophrenia (48.8% in the DIR, 41.5% in the SIR group). Total length of follow-up for discontinuation was 100 person-years (80 person-years in the SIR group, 20 in the DIR group).

Most patients were initiated during an inpatient spell for both the SIR (62.3%) and DIR (86.0%). There was a smaller proportion of treatment-resistance in the single-injection regimen (10.0%) than the double-injection regimen (20.9%). 38 patients (88.1%) in the DIR group had received a previous trial of aripiprazole for two weeks or more. A greater proportion of patients in the DIR group had directly switched treatment from clozapine (7.0%) compared with the SIR group (2.3%).

Time to ALAI Discontinuation

For the SIR group, discontinuation at one year was lower (n = 65, 50.0%) than for the DIR group (n = 31, 72.1%). The Kaplan-Meier estimates for time to discontinuation for the respective initiation regimens are displayed in Figure 3. Median time to discontinuation was 272 (95% confidence interval (CI) 202 - > 365) and 111 (95%CI 56 - 254) days for the SIR and DIR groups respectively (p = 0.004). Over a third of patients receiving the double-injection regimen (37.2%) did not receive a single maintenance dose of ALAI, compared to the single-injection group (13.1%). For patients with treatment-resistance who received the single-injection regimen (n = 13), 46.2% discontinued, whilst all treatment-resistant patients who received the double-injection regimen (n = 9) that were successfully followed up for one year discontinued (n = 8), and one patient was censored before one year. Ten patients were lost to follow-up for ALAI discontinuation in the study (eight in the SIR group, two in the DIR group), as they had transferred out of the Trust.

Reasons for discontinuation

Reasons for discontinuation are summarised in Table II. 31 patients in the DIR group discontinued ALAI (72.1%), and 65 discontinued in the SIR group (50.0%). The most common reason for discontinuation in both groups was “patient choice” - 58.1% in the DIR group and 38.5% in the SIR group. Side effects described as reasons for discontinuation are outlined in Table III.

Time to admission to psychiatric hospital

45 patients (26.0%) were admitted within one year of initiation of ALAI - 35 in the SIR (26.9%) and 10 (23.3%) in the DIR group. The Kaplan-Meier estimate for time to admission following initiation is displayed in Figure 5. A median estimate could not be produced as more than half of patients were not admitted at one year following initiation of ALAI. For the patients who were censored before one year follow-up, six patients had transferred from the Trust prior to reaching 1-year post-initiation. One patient in the DIR group died in the follow-up period, and two patients were discharged from SLaM services.

Time from initiation of ALAI to discharge

For those who were inpatients at the time of initiation (n = 118), median time from initiation to discharge was 21 days (95% CI: 0-61.6 days) in the SIR group, and 14 days in the DIR group (95% CI: 0-37.3 days).

Discussion

Key Findings

Treatment discontinuation is a widely used and valid outcome to measure effectiveness of antipsychotic medication ⁹. Patients receiving DIR discontinued to a greater extent than those given SIR (72.1% discontinued at 1 year in the DIR group, compared with 50.0% in the SIR group). We found that over a third of patients in the DIR group did not receive a maintenance dose after initiation (37.2% discontinued after 56 days), and the respective figure was lower in the SIR group (13.1%). Patients in the DIR group also discontinued aripiprazole over 150 days sooner than the SIR group (median time to discontinuation for the DIR group was 111 days, compared with the SIR group median 272 days).

Differences in the demographics of the patients initiated on the DIR compared with the SIR are apparent in our sample. There was greater proportion of patients initiated during an inpatient stay in the DIR group (86.0%) compared to the SIR group (62.3%). Figure 4 shows a clear difference in time to discontinuation when stratifying by the setting of initiation, irrespective of the initiation regimen. This difference is more pronounced compared with the difference when stratifying by initiation regimen. The greater proportion of inpatients in the DIR group may therefore contribute to the higher rates of discontinuation observed in the DIR group. The DIR may be preferentially prescribed in inpatient settings, possibly because it allows earlier discharge as suggested by our findings. Additionally, there was a greater proportion of treatment-resistant patients in the DIR group (20.9%) than the SIR group (10.0%). All treatment-resistant patients in the DIR group that completed follow-up stopped aripiprazole before the end of the study period. Interestingly, discontinuation rates were lower in the SIR group (46.1%) among treatment-resistant patients compared with the DIR group. The high rate of discontinuation in the double-injection cohort reiterates findings that aripiprazole is not effective in treatment-resistant schizophrenia ¹. As aripiprazole is not indicated in treatment-resistant schizophrenia, a higher proportion of discontinuation may be expected as a result. The higher discontinuation rate in the DIR group could reflect underlying psychopathological differences, including greater treatment resistance, poor insight, and paranoia, which have been linked to increased rates of non-adherence in schizophrenia ⁶.

Given the reduced need for oral treatment, it is also possible that the DIR was preferentially prescribed in patients with poorer adherence to medication, a risk factor for relapse in patients with schizophrenia ¹⁰. Baseline data on adherence to medication prior to initiation were not collected, however discontinuations due to “patient choice” were more frequent in the DIR group (58.1%) compared with the SIR group (38.5%). Discontinuation due to “patient choice” may be a marker of poor adherence to medication. This is further emphasised by the fact that discontinuations due to perceived inefficacy made up a larger proportion for the reasons for discontinuation in the SIR group (30.8%) compared with the DIR group (9.7%). However, it is important to note that “patient choice” may represent discontinuation due to a mixture of reasons, including side effects and inefficacy, and exploration of this reasoning is warranted in further studies ⁹. In our results, we did not observe a discernible difference in discontinuation due to adverse effects. Patients with greater illness severity often exhibit lower adherence due to mistrust of treatment, lack of perceived need for medication, and negative symptoms reducing motivation ⁵. This suggests that clinicians should consider individual psychopathological profiles when selecting an initiation regimen.

There was no difference in time to admission between cohorts, suggesting that the two initiation regimens have similar efficacy in reducing admissions. Patients were only followed up only for one year during which time most participants were switched to other antipsychotics, which may limit the impact of this finding. Inpatients in the DIR group were discharged sooner (mean time to discharge was 21 days in SIR, 14 in DIR), a factor that may impact the higher discontinuation rates in the DIR group. It is unclear whether the relationship between initiation regimen and time to discharge is causal i.e. the DIR allowed for quicker discharge of patients, or whether the DIR was used in patients who were going to be discharged sooner. Regardless, it is possible that the initiation of ALAI with less inpatient follow-up may have impacted the time to discontinuation in the DIR group.

Comparison with other studies

To the best of our knowledge, this is the first study investigating the discontinuation rates of ALAI initiation using the DIR. A recent retrospective, observational study conducted in Italy investigated adverse effects reported following the DIR ⁴. This study concluded that there were no new safety concerns associated with the DIR compared with the SIR and that adverse effects occurred at a similar frequency. Our findings similarly did not detect a difference in adverse effects between the regimens as a factor for discontinuation.

Previous studies have shown the effectiveness of ALAI in reducing bed days and admissions in patients initiated using the SIR. In one mirror-image study, the discontinuation of ALAI following initiation with one injection was 49%, very similar to our result in the SIR group (50.0%) ¹. This may affirm our finding that there is a greater discontinuation rate in those initiated with the DIR (72.1%). In the comparison study, the number of patients who discontinued due to perceived inefficacy was similar (34%) to those in our SIR group (30.8%). A higher proportion of patients discontinued due to “patient choice” when initiated with the DIR, however the SIR group result we found was broadly similar to the mirror-image study’s result (38.5% in our SIR group, 43% total in the comparative study for “patient refusal” or “patient request”). This further emphasises the higher proportion of discontinuations due to “patient choice” found in the DIR group in our study.

Limitations

The study is limited by the relatively low number of patients in the double-injection group. However, this study aimed to capture all patients who received the double-injection regimen immediately following the approval of the regime, so smaller numbers may be explained by clinicians’ lack of awareness of the new regime. Additionally, the study was conducted within one NHS Trust within the UK, which may limit the generalisability of findings. An inherent limitation of observational studies is the bias that may result from this design. An adjusted analysis was planned but it was not feasible to collect some important prognostic and confounding variables (such as previous response to aripiprazole), so an unadjusted and descriptive analysis was conducted. Differences between initiation regimens were described qualitatively. As such, it is difficult to attribute any differences in outcome to the ALAI loading regimens conclusively. Finally, this study did not assess the role of baseline psychopathology in discontinuation rates.

The double-injection initiation of ALAI may reduce the need for oral aripiprazole, however patients initiated on ALAI with the double-injection regimen discontinued treatment earlier than those given a single injection initiation. This finding may be explained by preferential prescribing of the double-injection regimen in patients more likely to discontinue treatment and in patients who are discharged sooner. The higher discontinuation rate of the two-injection regimen raises concerns about its suitability for patients with severe psychopathology, including those with treatment resistance. Future studies should explore how individual psychopathological factors influence treatment adherence and response to different LAI initiation regimens.

Funding

This research received no specific grant from any funding agency, commercial or not-for-profit sectors.

Conflicts of Interest statement

David Taylor has received investigator-initiated research grants from Janssen; Has spoken at events for Lundbeck, Otsuka, Viatris, Janssen and Recordati; has provided consultancy services for Viatris, Teva and Idorsia, has shares in Myogenes Ltd, Myogenes Inc, Saladax Inc and 428-Pharma

Authors’ contributions

EW conceptualised the study. All authors contributed to the study design. NM and BP collected the data. NM analysed the data. NM and OD interpreted the data. All authors contributed to the drafting and revision of the manuscript.

Ethical consideration

Ethical approval was received from the Drugs and Therapeutics Committee for this study (approval number: DTC/2022/36).

Figures and tables

FIGURE 1. Visual of the different initiation regimens, methodology of the proposed study.

FIGURE 2. Visual of how patient cohorts were identified for the two initiation regimens.

FIGURE 3. Kaplan-Meier curves of time to discontinuation of ALAI stratified by initiation regimen. Median time (Days) to discontinuation for double-injection regimen = 111, single-injection regimen = 272.

FIGURE 4. Kaplan-Meier curves of time to discontinuation of ALAI stratified by setting at time of initiation.

FIGURE 5. Time to admission to psychiatric hospital following ALAI initiation.

TABLE I. Baseline demographics of patients who were initiated on ALAI.

	All patients initiated on ALAI n = 173	Single-Injection Regimen n = 130	Double-Injection Regimen n = 43
Age (years)	n	n
Mean (SD)	39.1	35.5 (15.3)	37.6 (12.2)
Duration of illness (years)	n	n
Mean (SD)	11.4 (18.4)	11.8 (20.7)	10.1 (8.3)
Gender	n (%)
Male	115 (66.5)	88 (67.7)	27 (62.8)
Female	58 (33.5)	42 (32.3)	16 (37.2)
Ethnicity	n (%)
White	43 (24.8)	32 (24.6)	11 (25.6)
Black	92 (53.2)	67 (51.5)	25 (58.1)
Asian	14 (8.1)	14 (10.8)	0 (0)
Other	18 (10.4)	11 (8.5)	7 (16.3)
Not stated	6 (3.5)	6 (4.6)	0 (0)
Diagnosis (ICD-11 category)	n (%)
Schizophrenia (F20)	75 (43.4)	54 (41.5)	21 (48.8)
Other psychosis (F29)	28 (16.2)	23 (17.7)	5 (11.6)
Schizoaffective disorder (F25)	17 (9.8)	8 (6.2)	9 (20.9)
Bipolar affective disorder (F31)	23 (13.3)	18 (13.9)	5 (11.6)
Other	30 (17.3)	27 (20.7)	3 (7.0)
Inpatient at initiation	n (%)
Yes	118 (68.2)	81 (62.3)	37 (86.0)
No	55 (31.8)	49 (37.7)	6 (14.0)
Under MHA Section	n (%)
Yes	118	80 (61.5)	38 (88.4)
No	55	50 (38.5)	5 (11.6)
Considered treatment-resistant	n (%)
Yes	22 (12.7)	13 (10.0)	9 (20.9)
No	151 (87.3)	117 (90.0)	34 (79.1)
Antipsychotic switched from	n (%)
Second-Generation Antipsychotic	100 (57.8)	83 (63.9)	17 (39.5)
Clozapine	6 (3.5)	3 (2.3)	3 (7.0)
First-Generation Antipsychotic	32 (18.5)	28 (21.5)	4 (9.3)
None documented	35 (20.2)	16 (12.3)	19 (44.2)

TABLE II. Reasons documented at the time of discontinuation why ALAI was discontinued.

All patients that discontinued ALAI following initiation
	Single-Injection Regimen n = 65	Double-Injection Regimen n= 31
Reason for discontinuation of ALAI	n (%)
Patient choice	25 (38.5)	18 (58.1)
Inefficacy	20 (30.8)	3 (9.7)
Side Effects	8 (12.3)	6 (19.4)
Other/unclear	12 (18.4)	4 (12.9)
Patients that discontinued ALAI within 60 days of initiation
	Single-Injection Regimen n = 17	Double-Injection Regimen n = 16
Patient choice	11 (64.7)	11 (68.8)
Inefficacy	5 (29.4)	3 (18.8)
Side Effects	0 (0)	2 (12.5)
Other/unclear	1 (5.9)	0 (0)
Patients that discontinued ALAI more than 60 days post-initiation
	Single-Injection Regimen n = 48	Double-Injection Regimen n= 15
Patient choice	14 (29.2)	7 (46.7)
Inefficacy	15 (31.3)	0 (0)
Side Effects	8 (16.7)	4 (26.7)
Other/unclear	11 (22.9)	4 (26.7)

TABLE III. Adverse effects reported for patients who discontinued ALAI due to side effects. Some patients reported multiple side effects.

	Single-Injection Regimen (n = 8)	Double-Injection Regimen (n = 6)
Adverse Effects Reported Leading to Discontinuation	(n)	(n)
Akathisia	1	1
Nausea	0	2
Weight Gain	1	1
Sedation	2	0
Tremor	2	0
Parkinsonism	1	0
Injection site pain	1	2
Sexual dysfunction	1	0
Blurred vision	1	0
Hypersalivation	1	0