Abstract
Background:
The efficacy of the long-acting injectable formulation of the antipsychotic paliperidone (paliperidone palmitate) has been investigated in randomized controlled trials. Due to the nature of study designs, these may not be representative of usual clinical practice. The aim of this study was to assess the clinical effectiveness of the long-acting injectable antipsychotic paliperidone palmitate using treatment continuation at 1 year as an outcome.
Methods:
Patients were initiated on paliperidone palmitate prior to December 2014 in a single health board in Wales (UK). Demographic factors that may have influenced outcome, including diagnosis, age at initiation, sex, inpatient or outpatient status on initiation, were analysed to assess whether they influenced patient outcome. For patients completing 1 year of treatment, inpatient stay in the 12 months prior to and following paliperidone palmitate initiation was compared.
Results:
Data were available for 64 patients; 41 had a diagnosis of schizophrenia and 7 had previously received clozapine. Continuation rates at 6 and 12 months were 69% and 63% respectively. Treatment continuation was not associated with demographic factors. For continuers, mean inpatient stay pre- and post-initiation was 83.2 ± 105.3 and 73.5 ± 103.3 days respectively (p = 0.61). The most common reason for discontinuation was lack of effect (n = 9).
Conclusions:
The proportion of patients remaining on treatment was comparable to that reported in other naturalistic studies. Prescribing for indications outside the product licence was relatively common, but did not appear to influence outcome, although the number of patients in each group was small. Treatment continuation at 6 months appeared to be a predictor of longer-term outcome.
Introduction
Schizophrenia is a chronic, relapsing, severe mental illness associated with significant morbidity and mortality. The long-term management of schizophrenia may be limited by non-compliance with antipsychotic medication, which has been shown to be a significant factor in relapse and rehospitalization. 1 Long-acting injectable formulations of antipsychotics (dopamine D2 receptor antagonists or partial agonists) have been developed as a possible mechanism to address covert non-compliance, and may have some advantages over oral therapy in the management of this chronic relapsing condition. 2 However, the influence of factors that might affect outcome, such as optimal dosing interval, remain to be fully investigated. 3 Risperidone was the first of the newer, atypical antipsychotics to be formulated as a long-acting injection. Due to its pharmacokinetic properties, risperidone long-acting injection (RLAI) must be administered on a fortnightly basis. 4 The active risperidone metabolite, 9-OH-risperidone, has been shown to have high affinity for dopamine D2 and serotonin 5-HT2A receptors, 5 and has been marketed as the atypical antipsychotic paliperidone. A long-acting injectable formulation of paliperidone as the palmitate salt received marketing authorization (formerly known as the product licence) in the UK in 2011 and had the possible advantage of a pharmacokinetic profile that allowed monthly dosing. 6
The efficacy of paliperidone palmitate monthly injection (PP) in the treatment of schizophrenia has been investigated in randomized controlled trials (RCTs). PP has shown evidence of benefit over placebo 7 and comparable efficacy and tolerability to RLAI. 8 While the methodology employed in RCTs allows assessment of the efficacy of interventions, such studies may not be generalizable to clinical practice. The strict inclusion and exclusion criteria result in participants being unrepresentative of real-world patients. Moreover, short study durations may not be adequate to assess the long-term effectiveness of medicines intended for the treatment of chronic conditions. 9 Naturalistic observational studies10,11 and studies designed with pragmatic outcome measures such as treatment continuation 9 have substantiated the effectiveness of antipsychotics in settings more relevant to clinical situations. These studies have included patients who would not typically be eligible for inclusion in RCTs. This may help to provide evidence of the possible benefit of medicines when used in practice, and guide clinical decision-making. Furthermore, findings from naturalistic observational studies have been incorporated into treatment guidelines such as those published by the British Association for Psychopharmacology. 12 The aim of the present study was to assess the effectiveness of PP in a clinical setting using treatment continuation at 1 year as an outcome measure. Secondary objectives were to determine whether demographic factors (such as patient sex, treatment history and diagnosis) influenced treatment continuation at 1 year.
Methods
Design
This was a retrospective, naturalistic, 1-year follow-up study of patients receiving PP within a single health board in Wales (UK), in both hospital inpatient and outpatient settings. For patients completing 1 year of treatment, a mirror-image comparison of hospital inpatient stay in the 12 months prior to and following PP initiation was conducted. Data were collected from pharmacy records and by retrospective case-note review during November and December 2015. The health board quality improvement department and mental health clinical board approved the study.
Participants and outcome measures
All patients initiated on PP prior to December 2014 in a single health board in Wales were identified from pharmacy records and included in the study. Patients were classified as ‘continuers’ (those completing 12 months on PP treatment), or ‘discontinuers’ (those discontinuing within 12 months of initiation). Demographic factors that may have influenced treatment outcome including age at initiation, diagnosis (schizophrenia versus other indications including schizoaffective disorder), patient sex, previous antipsychotic, inpatient or outpatient status on initiation, prior clozapine treatment, two previous antipsychotic treatment failures and recreational drug use were recorded. The duration of hospital stay in the 12 months prior to and following PP initiation was used as a measure of treatment effectiveness in continuers. In the case of discontinuers, reasons for treatment discontinuation (as described in patient records), dose at discontinuation and the antipsychotic to which they were switched were noted. Prior treatment with clozapine was considered an indication of treatment-refractory illness.
Statistical analysis
Demographic factors that may have influenced treatment outcome were analysed using unpaired Student’s t test for continuous variables and Fisher’s exact test for categorical data. The number of days spent as an inpatient prior to and following PP initiation was compared using paired Student’s t test. All statistical analyses were conducted using GraphPad Prism version 6.0 (GraphPad Software, La Jolla, CA, USA).
Results
In total, 66 patients received PP, of whom two moved abroad during the follow-up period and were lost to follow up. Accordingly, 1-year outcome data were available for a total of 64 patients. Patient demographics are shown in Table 1; 41 (36%) patients had a diagnosis of schizophrenia; the other common diagnoses were schizoaffective disorder (n = 5; 8%), bipolar disorder (n = 5; 8%) and persistent delusional disorder (n = 4; 6%). Forty-nine patients (76%) were switched to PP from risperidone (either oral or long-acting injectable formulations). Other antipsychotics prescribed prior to PP were aripiprazole (n = 5), flupentixol decanoate (n = 3), amisulpride (n = 2), haloperidol (n = 2), olanzapine (n = 2), quetiapine (n = 1), zuclopenthixol decanoate (n = 1) and promazine (n = 1). Two patients were co-prescribed aripiprazole with another antipsychotic prior to PP initiation. Seven (11%) patients had previously received clozapine. The number of patients remaining on treatment over time according to diagnosis is shown in Figure 1; 40 (63%) patients remained on PP at 1 year.
Patient demographics, reasons for initiation and dose at study endpoint.
Patients remaining on treatment over time according to diagnosis.
Continuers
Treatment continuation (versus treatment discontinuation) was not associated with age at initiation (p = 0.054, t = 1.965, df 62, unpaired Student’s t test), diagnosis, patient sex, previous risperidone treatment versus other antipsychotic, inpatient/outpatient status on initiation, prior clozapine treatment, two previous treatment failures or recreational drug use (p = 0.79; p = 1.0; p = 0.38; p = 1.0; p = 0.41; p = 0.44; p = 0.43 respectively, Fisher’s exact test). Five continuers were co-prescribed regular antipsychotics in combination with PP. Three continued to receive the antipsychotic prescribed prior to PP initiation (risperidone, promazine and aripiprazole) and two were co-prescribed olanzapine. In regard to patients completing 1 year on PP, the mean time spent as an inpatient in the 12 months prior to PP initiation was 83.2 ± 105.3 days, and in the 12 months post-PP initiation it was 73.5 ± 103.3 days (p = 0.61, t = 0.5175, df 39, paired Student’s t test).
Discontinuers
In total, 24 patients discontinued PP within 1 year, 12 of whom discontinued between 90 and 120 days and 4 of whom discontinued between 120 and 365 days (see Figure 1). Four discontinuers were co-prescribed a regular antipsychotic in combination with PP. Three continued to receive the antipsychotic prescribed prior to PP initiation (risperidone, aripiprazole and olanzapine) and one was co-prescribed olanzapine. Mean dose at discontinuation, reasons for discontinuation and antipsychotic treatment post-discontinuation are shown in Table 2. The most common reason for discontinuation was perceived lack of effect (n = 9). The doses at discontinuation for these patients were 100 mg (n = 4) and 150 mg (n = 5) (mean = 127.8 mg), and the antipsychotics prescribed post-PP were clozapine (n = 7), flupenthixol decanoate (n = 1) and amisulpride (n = 1). Seven patients discontinued due to adverse effects, which encompassed weight gain (n = 2), extra-pyramidal side effect (n = 2), sexual dysfunction (n = 2) and sedation (n = 1). Out of the five patients switching from risperidone to PP due to a lack of effect, three had discontinued PP at 1 year due to lack of effect (the other two patients continued on PP treatment).
Reasons for discontinuation and antipsychotic prescribed post-PP.
Discussion
This study assessed the effectiveness of PP in clinical practice using treatment continuation at 12 months as an outcome measure. Among the 64 patients who were included in the study, 40 (63%) remained on treatment (classified as ‘continuers’) at 1 year. There were no significant differences in demographic factors between continuers and patients discontinuing PP, therefore baseline characteristics did not appear to influence treatment outcome in this study. The majority (83%) of patients who discontinued treatment did so within the first 180 days, therefore completion of 6 months’ treatment appeared to be a good indicator of continuation at 1 year. In the case of treatment continuers, there was no significant difference in inpatient stay in the 12 months prior to and following PP initiation. The most common reason for treatment discontinuation was lack of effect.
The proportion of patients remaining on treatment was comparable to the 65% 10 and 60% 13 reported in similar 1-year naturalistic studies of PP. One study comparing the outcomes of patients treated with PP, RLAI and zuclopentixol decanoate reported a somewhat lower continuation rate of PP (48%), albeit over a longer 18-month follow-up period. 14 Continuation with PP was noticeably higher than that seen at the same time point in a similar study of RLAI (45%) in our health board 15 and that observed in another study of RLAI in the UK. 16 However, it must be noted that the observational nature of these studies does not allow direct comparisons to be made. In the case of continuers, there was no significant difference in the number of days spent as an inpatient in the 12 months prior to and following PP initiation. This contrasted with other similar naturalistic studies of PP which revealed a reduction in hospital admissions and inpatient stay post-initiation in hospital in the UK,17,18 albeit over longer periods (2 and 3 years post-PP initiation respectively) than seen in our study. Two 1-year mirror-image studies, one from Canada 19 and one from the UK, 20 showed a non-significant reduction in hospitalization days when using first PP injection as the mirror point, suggesting that study duration may be an important factor when measuring this outcome. However, it should also be noted that these were naturalistic studies, and any variation may simply reflect underlying differences in the study populations. It is nevertheless conceivable that with a longer duration of follow up, PP may have shown a benefit in relation to inpatient stay. A cost-effectiveness study utilizing this as one of the outcome measures would be a useful focus for future work.
None of the demographic factors analysed in this study appeared to influence treatment outcome. This was in contrast to the findings of other studies of PP which found that outpatient status on initiation, switching from risperidone 10 and a diagnosis of schizophrenia 13 were associated with treatment continuation at one year. Similarly, outpatient initiation and switching from risperidone have also been shown to predict PP continuation at 2 years. 21 The relatively small number of subjects in our study may be an explanation for these differences. Prescribing for indications outside of the marketing authorization accounted for 36% of patients in our study. In part, this may reflect the more limited indications of the long-acting injectable paliperidone preparation compared to the oral equivalent. However, it also highlights the relatively common phenomenon of off-label prescribing within the field of psychiatry. Nevertheless, patients treated for conditions other than schizophrenia were no more likely to discontinue at 1 year, providing some evidence for the effectiveness of PP in this group. Although demographic factors were not associated with outcome, treatment continuation at 6 months appeared to be conducive to or suggestive of continuation at 1 year. The majority (83%) of patients who discontinued PP did so within 6 months of treatment initiation, while 91% of those completing 6 months went on to complete 1 year. This was largely in accordance with a more long-term finding that completion of 1 year of treatment with RLAI or oral aripiprazole was an indicator of subsequent continuation at 5 years in a naturalistic cohort. 11 A perceived lack of effect was the most common reason for discontinuation, consistent with other naturalistic studies,10,13,14 and accounted for 14% of the total sample. Despite lack of effect being the reason for discontinuation, 4/9 (56%) of these patients did not receive the maximum licensed dose of PP (150 mg), while 3/9 (33%) had previously been treated with clozapine. Therefore, 6/9 individuals might have been expected to experience a poor response due to suboptimal dosing or a history of treatment-refractory illness. Regular co-prescription of other antipsychotics was relatively infrequent, accounting for nine patients. The proportion of discontinuers co-prescribed (approximately 30%) was somewhat greater than the continuers, which in part may reflect more treatment-refractory symptoms in this subgroup. Two of those co-prescribed (one continuer, one discontinuer) received aripiprazole as a strategy to address elevated prolactin levels. PP appeared to have similar tolerability to that seen in our previous study of RLAI with seven (11%) of the 64 patients discontinuing due to adverse effects, compared with 10% reported by Deslandes and colleagues. 15 In this context, other naturalistic studies of PP have reported both lower (5%) 10 and slightly higher (13%) 14 levels of discontinuation due to adverse effects.
Patient choice was the reason for PP initiation in 16 patients, 14 of whom chose PP on the basis of monthly rather than fortnightly dosing when switching from RLAI. However, this did not appear to influence outcome, with similar proportions of both continuers and discontinuers (25%) switching to PP for this reason. Furthermore, of the six of these patients who subsequently discontinued, three did so due to refusal of PP. Despite the chemical and pharmacological similarities between paliperidone and risperidone, five patients switched to PP from oral risperidone or RLAI due to lack of effect. The rationale for this was the higher licensed dose equivalence of PP (150 mg/month) compared with RLAI (100 mg/month). This did not appear to influence outcome, however, with three of the five patients subsequently discontinuing PP due to lack of effect. Across the whole study cohort, the mean PP dose at endpoint was only slightly greater than 100 mg, although approximately 30% of continuers received 150 mg and therefore a higher equivalent dose than could have been achieved within the RLAI licence.
This was a retrospective, naturalistic, observational study, and the methodological limitations were common to many such studies. There was no comparator group, no formal assessment of mental state nor randomization; therefore, any conclusions must be treated with a degree of caution. However, the study does report outcomes in real-world patients (many of whom would have been excluded from RCTs), being treated for both licensed and unlicensed indications and with an extended period of follow up. PP appeared to show equivalent effectiveness in both schizophrenia and other indications. This may offer some reassurance to prescribers given the relatively common use of PP for indications outside the product licence. However, it must be noted that patient numbers were small, and the study was not designed to compare the outcomes of these two groups. Possible underlying differences between the populations treated do not allow direct comparisons to be performed. Although the overall continuation rate seen in this study was only 63%, the subsequent completion rate for those who remained on treatment at 6 months (91%) was encouraging and suggested that outcome at 6 months may be a useful indicator of longer-term treatment continuation.
Footnotes
Acknowledgements
The authors thank Mrs W. Davies for the opportunity to conduct this work.
Funding
This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Conflict of interest statement
PND has received funding to attend scientific meetings and honoraria for speaking engagements from Janssen-Cilag Ltd. None of the other authors have any conflict of interest to declare.