Patient-reported outcomes in schizophrenia patients treated with once-monthly extended-release risperidone in a long-term clinical study

Introduction

Discontinuation or disruption of antipsychotic therapy in the treatment of schizophrenia is one of the primary causes of relapse among patients with schizophrenia or schizoaffective disorder.¹ Administration of long-acting injectable (LAI) medications to patients with schizophrenia has the potential to improve medication adherence and lower discontinuation rates compared to patients on daily oral therapy.^2,3 Previous research has found that switching patients with schizophrenia or schizoaffective disorder from shorter-acting oral to LAI formulations improved health-related quality of life (HRQoL)^4,5 and reduced hospitalization rates.^4,6

The safety and efficacy of risperidone in the management of positive symptoms of schizophrenia is well established.^7–9 Risperidone is marketed in a number of countries, including the US, in various formulations including tablets, orally disintegrating tablets, oral solutions and intramuscular injections.^10,11

After a 3-week initiation period during which patients have to be concurrently treated with oral risperidone to ensure adequate therapeutic plasma concentrations, the LAI intramuscular formulation of risperidone is administered every 2 weeks as deep deltoid or gluteal injections.¹² In contrast, RBP-7000 (PERSERIS™, Indivior, Inc., Richmond, VA, USA), provides a once-monthly subcutaneous (SC) extended-release risperidone formulation that does not require any oral supplementation for the treatment of schizophrenia in adults.¹³ An 8-week, randomized, double-blind, placebo-controlled Phase III study of RBP-7000 (NCT02109562) demonstrated the clinical safety and efficacy of RBP-7000¹⁴ and significant improvements in HRQoL and overall well-being.¹⁵ To build upon this evidence, the aim of the present analysis was to further describe the effects of RBP-7000 on patient-reported HRQoL, subjective well-being, treatment satisfaction and preference to medication over 52 weeks.

Methods

This analysis used data from a 52-week multicentre Phase III open-label outpatient study (NCT02203838) that was conducted at 50 sites in the US between June 2014 and September 2016 to assess the long-term safety and tolerability of RBP-7000 (120 mg) in patients with schizophrenia. The study methods and results for clinical efficacy and safety (eg, Positive and Negative Syndrome Scale [PANSS], Clinical Global Impression-Severity of Illness, Abnormal Involuntary Movement Scale) are reported elsewhere.¹⁶

The study was conducted in accordance with the International Conference on Harmonisation Good Clinical Practice guidelines, Food and Drug Administration regulations governing clinical study conduct and the Declaration of Helsinki (2013). The study protocol, all protocol amendments, written study patient information, the written informed consent form and all other appropriate study-related information were reviewed and approved by the Copernicus Group Independent Review Board ^® (CGIRB).

Participants

Two cohorts provided participants for the clinical study: a rollover cohort and a de novo cohort. The rollover cohort consisted of participants with schizophrenia, aged 18–55 years, who had completed an earlier 8-week, randomized double-blind, placebo-controlled RBP-7000 study (NCT02109562)^14,15 during which they had received 2 monthly SC injections of either placebo, RBP-7000 90 mg or RBP-7000 120 mg and provided that treatment continuation was clinically warranted and participants were judged stable enough to enroll by the investigator. After the end of the placebo-controlled study (EOS), participants entered the open-label outpatient study directly with their third injection on day 1. Participants received up to an additional 11 SC injections of RBP-7000 over the course of 40 weeks in this study.

The de novo cohort consisted of participants who were diagnosed with schizophrenia, aged 18–65 years, who were not previously exposed to RBP-7000, were considered clinically stable with a PANSS total score ≤70 at screening and were otherwise healthy based on physical examination; those with suicidal ideation (within 6 months) or behavior (within 1 year) based on Columbia-Suicide Severity Rating Scale assessment, or with a significant risk of suicide in the opinion of the investigator, were excluded. De novo participants not on oral risperidone or on doses other than 3or 4 mg entered a run-in or conversion phase where they were tapered off their current antipsychotic medications (Figure 1) and started on a 7- to 14-day regimen of oral risperidone, whereas de novo participants already receiving oral risperidone did not have to complete the run-in or conversion phase and entered the study directly. De novo participants received up to 13 SC injections of RBP-7000 over the course of 52 weeks.

Figure 1 Study design for the open-label phase of de novo and rollover participants. Rollover participants received 2 doses of RBP-7000 during the previous 8-week trial and up to an additional 11 SC injections over the course of 40 weeks in this study. De novo participants received up to 13 SC injections of RBP-7000 over the course of 52 weeks. If they were not on oral risperidone, or on doses other than 3 or 4 mg, they were tapered off their current antipsychotic medications and initiated on a 7- to 14-day regimen of oral risperidone. If they were already receiving oral risperidone, they did not have to complete the run-in or conversion phase and entered the study directly. Abbreviations: C, conversion; DB; double-blind; EOS, end of study; F/U, follow-up; OL, open-label; RI, run-in; SC, subcutaneous.

Exposure

All participants received RBP-7000 120 mg at study entry. A single dose modification for tolerability to RBP-7000 90 mg was allowed at the investigator’s discretion. Participants who were down-titrated to the RBP-7000 90-mg dose who exhibited a worsening in psychiatric symptoms (ie, total PANSS score >70 or a 20% increase in PANSS score from previous assessment [at the 120-mg dose prior to the dose reduction]) received a subsequent single dose modification to RBP-7000 120 mg at the investigator’s discretion.

Patient-reported outcomes

Key patient-reported outcome (PRO) measures included the EuroQoL 5D 5-Level (EQ-5D-5L); Short-Form 36-item Questionnaire, Version 2 (SF-36v2); Subjective Well-being Under Neuroleptic Treatment-Short Version (SWN-S); Medication Satisfaction Questionnaire (MSQ) and the Preference of Medication (POM) Questionnaire. EQ-5D-5L, SF-36v2 and SWN-S assessments were administered at screening and weeks 0 (injection 1), 12 (injection 3), 24 (injection 6), 36 (injection 9) and 52 (EOS); MSQ and POM were administered at weeks 4, 12, 24, 36 and 52 with MSQ also administered at screening.

Depending on the measure and its inclusion in the double-blind trial, observations were compared to active baseline, open-label baseline or the first measurement within this clinical study. Active baseline was defined as the last non-missing measurement taken prior to the first dose of RBP-7000 regardless of whether received during the double-blind or open-label studies (ie, the measurement taken on the same day as the first dose). Open-label baseline was defined for all participants as the last non-missing measurement taken prior to the date of first dose of RBP-7000 in this clinical study.

Health-related quality of life

HRQoL was measured using the EQ-5D-5L as well as the SF-36v2. The EQ-5D-5L is a standardized, patient-reported, generic instrument for measuring health outcomes^17,18 that has been validated in schizophrenic patients.¹⁹ It provides a simple descriptive profile and a single index value for health status. The instrument consists of the EQ-5D-5L descriptive system and the VAS. The descriptive system consists of 5 dimensions (mobility, self-care, usual activities, pain/discomfort, anxiety/depression) with 5 levels of severity (no problems, slight problems, moderate problems, severe problems, extreme problems) within each dimension, which are used to generate the index score anchored at 0 for death and 1 for no problems reported in any dimension. The VAS records the respondent’s self-rated health on a 20-cm, 100-point vertical VAS with endpoints labeled “The worst health you can imagine” at 0 and “The best health you can imagine” at 100.

The SF-36v2 is a self-reported, multipurpose, 36-item survey that measures 8 domains of health: bodily pain, general health, general mental health, physical functioning, role-emotional, role-physical, social functioning and vitality.^20,21 The reliability and validity of this instrument have been demonstrated in a schizophrenic population.²² Scores range from 0 to 100, with higher scores indicating better HRQoL. The SF-36v2 yields scale scores for each of the 8 health domains and 2 summary measures derived from the domain scores: physical component summary (PCS) and mental component summary (MCS). The PCS and MCS are constructed to have a mean of 50 and standard deviation of 10 in the general US population. The SF-36v2 was scored per convention using proprietary software.

Subjective well-being

The SWN-S scale is a 20-item (10 positive and 10 negative) patient-rated instrument designed to capture subjective well-being during antipsychotic treatment.²³ The reliability of this instrument has been demonstrated in patients with schizophrenia.²⁴ Each item is scored on a Likert scale with 6 response categories ranging from “Not at all” to “Very much.” Each item is scored from 1 to 6, with a minimum total score of 20 (low subjective well-being) and a maximum total score of 120 (good subjective well-being). The SWN-S has 5 subscales (mental functioning, self-control, emotional regulation, physical functioning and social integration) from which a total score is derived.

Treatment satisfaction

Medication satisfaction was measured using the MSQ. The MSQ is a single-item questionnaire that evaluates satisfaction with current antipsychotic medication and was found to be reliable in patients with schizophrenia.²⁵ The single item is scored from 1 (Extremely dissatisfied) to 7 (Extremely satisfied). The MSQ was assessed dichotomously as “Dissatisfied” (scores 1–4) and “Satisfied” (scores 5–7).

Preference of medication

The POM is a 2-item questionnaire assessing the preference for the current antipsychotic compared with the most recent pre-study antipsychotic.²⁶ Even though both the items ask the same question, one is addressed to the subject and the other to the subject’s caregiver. The item is scored from 1 (Much better; I prefer this medication) to 5 (Much worse; I prefer my previous medication). The POM was evaluated dichotomously as “Better” (scores 1–2) and “Same or worse” (scores 3–5).

Statistical analysis

All items and/or summary scales were descriptively summarized by cohort (de novo and rollover) and combined and presented as an “All Participants” cohort. Categorical variables were summarized using frequencies and percentages, and continuous measures were summarized using descriptive statistics (mean, SD).

Sensitivity analyses

To check for withdrawal bias, baseline characteristics were compared between completers and non-completers. Completers were defined as participants who reached EOS, while non-completers were those who stopped treatment or dropped out before reaching EOS, in which case an end of treatment measurement had been collected.

Patterns of missingness were explored to determine the type of missingness. Mixed-effect models with repeated measures (MMRM) and joint process models were used to investigate the impact of informative censoring of longitudinal PRO outcomes due to dropout/discontinuation of the treatment regimen. Joint process models were used to simultaneously model longitudinal PRO and time to dropout/discontinuation.

Results

Of the 482 participants with a baseline PRO assessment within the clinical study (395 de novo and 87 rollover) (Table 1), 234 participants (48.5%) completed the study. The most common reasons for study discontinuation among study non-completers were participants withdrawing consent (21%) and dropout due to clinical reasons (20%), including lack of efficacy, adverse events and withdrawn by investigator (Table 2). The majority of participants in the PRO cohort were male (67.6%) and African American (71.0%) with a mean age of 45.1 years.

Table 1 Baseline characteristics by cohort

Table 2 Reasons for dropout by cohort

Impact of RBP-7000 on HRQoL and subjective well-being

EQ-5D-5L index and VAS scores remained stable from active baseline to EOS (Table 3). At active baseline and EOS, participants reported the most problems in the EQ-5D-5L anxiety/depression dimension and the least problems in the self-care dimension.

Table 3 Mean (SD) HRQoL scores by cohort at baseline and EOS

Among all SF-36v2 dimensions and across all cohorts, participants reported the lowest scores in the vitality dimension and the highest scores in the bodily pain dimension at both open-label baseline and EOS assessments. Scores across all SF-36v2 dimensions remained stable from open-label baseline to EOS in the all participants cohort. Within the same cohort, mean PCS and MCS scores also remained stable from open-label baseline to EOS (Table 3).

Across all cohorts, stable SWN-S scores were reported from active baseline to EOS (Table 4).

Table 4 Mean (SD) SWN-S scores by cohort at active baseline and end of study

Impact of RBP-7000 on treatment satisfaction and preference

The proportion of satisfied participants based on the MSQ increased by 15 percentage points from first measurement after active RBP-7000 treatment to EOS (Figure 2). An increase in satisfaction with treatment was observed in the de novo treatment cohort with 80% reporting satisfaction at EOS compared to 64% at active baseline.

Figure 2 Proportion of de novo and rollover (RO) participants reporting satisfaction with RBP-7000, at active baseline (BL) and end of study (EOS). There was a 15% increase in satisfaction among all participants from baseline to end of study.

From first measurement (week 4) to EOS, the proportion of participants reporting preference on the POM for RBP-7000 as “better” compared to current antipsychotic medication increased among the all participants cohort (66% vs 72%) (Figure 3). By EOS, the proportion of participants reporting the preference for RBP-7000 on the POM in the rollover cohort remained stable, while the de novo cohort showed an increase in preference from first assessment (64–71%).

Figure 3 Proportion of de novo and rollover (RO) participants reporting preference for RBP-7000 as compared with the most recent pre-study antipsychotic medication, at first assessment (FA,week 4) and end of study (EOS). There was a 6% increase in the proportion of patients reporting preference for current antipsychotic medication among all participants.

Sensitivity analysis

A comparison of open-label baseline characteristics between completers and non-completers revealed no significant differences in age, sex, race or body mass index. There were some minor differences in the prevalence of certain comorbid conditions (hypertension and insomnia), but overall there were no major differences between completers and non-completers at baseline.

Upon exploration, it was determined that the pattern of missingness was non-ignorable, ie, missing not at random, and a joint process modeling approach was considered appropriate (longitudinal modeling of PRO and time to discontinuation/dropout). Withdrawal bias due to dropouts for key PRO measures (SF-36v2 PCS, SF-36v2 MCS, SWN-S, MSQ) was assessed by comparing the estimates obtained from 3 models: MMRM (completers only), MMRM (all participants including both completers and non-completers) and joint process survival model (all participants including both completers and non-completers). The results from these models suggested a minor effect of dropout on characterization of change over time in PRO measures (Table S1).

Discussion

Few LAI treatment studies conducted in patients with schizophrenia report on PRO measures even though the patient’s perception of medication benefit is a predictor of improved compliance.^5,27,28 The results of this study demonstrate the long-term sustainability of improvements seen in HRQoL and overall well-being associated with RBP-7000 treatment. This builds on the evidence generated within the previously completed double-blind, randomized, placebo-controlled study, which reported significant improvements in HRQoL as measured by the EQ-5D-5L questionnaire and overall well-being as measured by the SWN-S questionnaire.¹⁵

Participants receiving RBP-7000 for 12 months reported mean SF-36v2 PCS and MCS scores at their EOS visit that were within one standard deviation of the norm-referenced scores of the SF-36v2 (as these scores are norm-referenced against a US population with a mean of 50 and SD of 10).²¹ Compared to previous studies of risperidone for the treatment of schizophrenia or other psychiatric disorders, RBP-7000 participants reported higher PCS and MCS.^29,30 Similarly, EQ-5D-5L index scores for participants treated with RBP-7000 were higher than or similar to the mean EQ-5D-5L index scores reported by other studies conducted in non-US participants with schizophrenia.^31–33 The high HRQoL observed within this analysis is particularly important considering that HRQoL (as assessed by SF-36 PCS and MCS scores) has been identified as an independent predictor of relapse in schizophrenia.³⁴

Subjective well-being (as measured by SWN-S scores) remained stable in participants treated with RBP-7000 and overall reported higher total scores compared to a previous study with risperidone and paliperidone palmitate LAI users.³⁵ Previous work has indicated that SWN-S scores in a schizophrenic population are positively correlated to compliance to neuroleptic treatment.^23,36,37 Given that adherence is critical to the stability and positive outcomes in patients with schizophrenia,³⁸ the SWN-S scores reported by RBP-7000 users in the current study are important to consider.

Within this study, we found that over two-thirds of participants reported high satisfaction with and preference for RBP-7000 as early as week 4. Satisfaction with and preference for RBP-7000 are important considerations since treatment satisfaction is a crucial component of long-term outcomes. For example, a panel of 12 psychiatrists discussed treatment satisfaction as an important domain for treatment effectiveness and retention in patients with schizophrenia³⁹ and patient satisfaction has been associated with improvement in long-term outcomes in schizophrenia.⁶ Patient satisfaction has also been shown to be correlated with treatment retention in a non-schizophrenic population.⁴⁰

It is important to interpret the study results within the context of limitations. The overall dropout rate observed in this study (51.5%) was high, but remains consistent with clinical trials of similar duration with second-generation antipsychotics.^41–44 Our sensitivity analysis regarding potential withdrawal bias revealed no significant differences between study completers and non-completers (dropouts), ie, no bias was observed due to dropout. This study was not powered to detect differences in PRO endpoints as these endpoints were collected as tertiary endpoints in the study.

These results will be most generalizable to a population similar to our study participants and cannot be generalized to a larger population. As this analysis was conducted on patients treated within a clinical trial setting, further research will be needed to assess the real-world effectiveness of RBP-7000.

Conclusion

A large portion of participants were satisfied with and preferred RBP-7000 within this study, attaining HRQoL scores close to those of a general US population. Even though risperidone is a well-known atypical antipsychotic, additional research is needed to confirm whether HRQoL, satisfaction and preference for this new once-monthly risperidone SC formulation translate into improved outcomes such as adherence and ultimately fewer relapses in patients with schizophrenia.