Relapse risk factors
Use of LAIs in first-episode schizophrenia
LAI vs. oral antipsychotics
Use of aripiprazole once-monthly
Current recommendations for LAI use
Comment by Dr. Marc-André Roy
The clinical goals in the management of schizophrenia are to reduce the severity and duration of acute episodes of psychosis, promote recovery, maintain optimal functioning and prevent subsequent relapses (National Institute for Clinical Excellence, Clinical Practice Guidelines, No. 82;2009:210; Falkai et al. World J Biol Psychiatry 2006;7:5-4; Lehman et al. Am J Psychiatry 2004;161[suppl 2]:1-56).
Relapses can contribute to clinical destabilization, an increased risk of hospitalization, and a more intractable course of worsening symptoms and reduced responsiveness to antipsychotic therapies. There is emerging evidence that the duration of psychotic relapses is associated with brain volume changes (Andreasen et al. Am J Psychiatry 2013;170:609-615; free full text at www.ncbi.nlm.nih.gov/pmc/articles/PMC3835590/).
Inadequate treatment early in the course of illness may be associated with poorer outcomes. A 15-year follow-up study found that two-thirds of patients experienced at least one relapse, and that after each relapse, 17% could no longer achieve remission. Only 26.8% of first-episode patients experienced a complete and sustained remission after 1-2 episodes of psychosis (Wiersma et al. Schizophr Bull 1998; 24:75-85). The mean duration of subsequent psychotic episodes lengthened from nine months for the second episode, to 15 months for the third and 27 months for the fourth, indicating a worsening pattern of more chronic illness.
Similar data were reported in the EPPIC study (Henry et al. J Clin Psychiatry 2010;71:716-728). After a median follow-up of 7.4 years, 37-59% of patients had a symptomatic remission, depending on the criteria used, but only one-quarter achieved both a remission and social/vocational recovery. A separate study found that 13.7% of first-episode schizophrenia patients met criteria for a full recovery (remission of positive and negative symptoms, adequate social/vocational functioning for two years or more) at five years (Robinson et al. Am J Psychiatry 2004;161:473-479). In the recently reported AESOP-10 follow-up study of first-episode schizophrenia, a longer time to first remission (>2 vs. < 2 years) was associated with a 6.8-fold increased risk of natural-cause mortality (Reininghaus et al. Schizophr Bull 2014; epublished September 27, 2014). A longer duration of untreated psychosis (>2 vs. < 2 years) was also associated with a 2.7-fold increased risk of natural-cause mortality.
A further consideration is the association between episodes of psychosis and brain volume loss. A longitudinal voxel-based morphometry study showed substantial reductions in grey matter volume when patients were re-scanned 1.5 years after a first episode (Asami et al. Neuroimage 2012;59:986-996). Brain volume losses in the temporal and frontal regions were associated with positive and negative symptoms, and with a worse clinical course. An MRI analysis of patients from the Iowa Longitudinal Study of first-episode schizophrenia reported that relapse duration was related to decreases in total cerebral volume and frontal volume, suggesting that prolonged periods of psychosis may have a negative impact on brain integrity (Andreasen 2013). Indeed, a “relapse signature” for psychosis has recently been proposed, characterized by structural changes in the frontal and temporal regions, neurochemical alterations in dopamine and glutamate neurotransmission, and neuroinflammation resulting from microglial activation (Cropley et al. Int Clin Psychopharmacol 2013; epublished May 10, 2013; Cropley & Pantelis. Epidemiol Psychiatr Sci 2014;23:219-225).
These findings underscore the importance of early, effective intervention to prevent relapses, improve psychiatric and social functioning, and slow the pathophysiological changes that occur in schizophrenia.
Relapse risk factors
Relapses have been variously defined but generally the criteria include psychiatric hospitalization, worsening on the PANSS, CGI or BPRS scales, exacerbation or re-emergence of symptoms, and/or self-harm or violent behaviour, including suicidality (Olivares et al. Ann Gen Psychiatry 2013;12:32; free full text at www.ncbi.nlm.nih.gov/pmc/articles/PMC4015712/). In their review of the literature, Olivares and colleagues identified three key factors associated with relapse: non-adherence to antipsychotic medication; comorbid stress/depression; and substance abuse (Olivares 2013). Of these, treatment non-adherence was the most frequently cited relapse risk factor. For example, a study of first-episode schizophrenia found that patients who were non-adherent to treatment at one-year follow-up were significantly more likely to relapse compared to those who remained on therapy (70% vs. 25%) (Ucok et al. Eur Arch Psychiatry Clin Neurosci 2006;256:37-43). In newly-diagnosed patients, the relapse rate was 69% in non-adherent patients compared to 18% in adherent patients (Drake et al. J Clin Psychiatry 2007;68:81-86).
Patients with first-episode schizophrenia are especially vulnerable to treatment non-adherence, in part due to poor acceptance of their diagnosis or the feeling that treatment is not necessary or desirable (Kane & Garcia-Ribera. Br J Psychiatry Suppl 2009;52:S63-67). A University of Calgary study of the first 200 admissions to its Early Psychosis Program found that the one-year rate of treatment adherence was only 41% (Coldham et al. Acta Psychiatr Scand 2002;106:286-290). One-year adherence rates as low as 30% have been reported in first-episode patients (Emsley et al. J Clin Psychiatry 2012;73:e541–e547). Indeed, a Finnish cohort study found that only 54.3% of first-episode schizophrenia patients continued on medication for longer than two months after treatment initiation (Tiihonen et al. Am J Psychiatry 2011;168:603-609).
Use of LAIs in first-episode schizophrenia
Long-acting injectable (LAI) antipsychotics are often reserved for chronic schizophrenia patients and those with a history of poor adherence to oral agents. However, LAIs provide a number of advantages over oral agents in first-episode schizophrenia. The route of administration avoids first-pass metabolism by the liver, which lowers the risk of drug-drug interactions; reduces peak-trough effects due to more consistent bioavailability, which improves drug tolerability; and enables clinicians to use the lowest effective dose and minimize cumulative drug exposure (McEvoy J. J Clin Psychiatry 2006;67[Suppl. 5]:15-18; Gerlach J. Int Clin Psychopharmacol 1995;9[Suppl. 5]: 17-20).
One particular advantage is greater adherence with LAI compared to oral agents, which contributes to improved clinical outcomes. A post-hoc analysis of two studies found that the all-cause discontinuation rate at 24 months was significantly lower with LAIs compared to oral agents (26.0% vs. 70.2%) (Emsley et al. Clin Ther 2008;30:2378-2386). The LAI was superior to the oral antipsychotic with respect to reduction in PANSS total scores (-39.7 vs. -25.7), remission rate (64.0% vs. 40.4%), and relapse rate (9.3% vs. 42.1%). Similarly, relapse rates were 64% lower with an LAI versus oral antipsychotic in the Finnish cohort study (Tiihonen 2011). In clinical practice, LAI initiation in first-episode patients promotes more regular contact with the healthcare team (Pandarakalam JP. Hosp Med 2003;64:603-608), and enables clinicians to discreetly monitor adherence and respond more promptly if a patient stops therapy (Remington & Adams. Can J Psychiatry 1995;403 Suppl 1]:S5-11; NICE 2009).
LAI vs. oral antipsychotics
The comparative effectiveness of LAI and oral antipsychotics is a matter of some contention. A meta-analysis of 21 randomized controlled studies found that the effect of LAI and oral antipsychotics on relapse prevention was comparable (Kishimoto et al. Schizophr Bull 2014;40:192-213).
However, study populations do not necessarily reflect patients seen in practice, most notably with respect to adherence. The close attention paid to patients enrolled in RCTs contributes to higher adherence rates overall, so randomized controlled trials are less likely to detect a difference in adherence rates between LAI and oral antipsychotics (Kane et al. J Clin Epidemiol 2013;66[8 suppl]:S37-S1). Accordingly, Kishimoto and colleagues performed a second meta-analysis of 25 mirror-image studies (n=5,940), which look at treatment effects before and after LAI initiation and patients serve as their own controls (Kishimoto et al. J Clin Psychiatry 2013;74:957-965). The study found that LAIs were superior to oral antipsychotics in preventing hospitalization (risk ratio 0.43), and decreasing the number of hospitalizations (RR 0.38).
Use of aripiprazole once-monthly
A once-monthly (OM) LAI formulation of aripiprazole (Abilify Maintena; Otsuka Pharmaceutical Co., Ltd.), was recently approved by Health Canada for maintenance treatment of adult patients with schizophrenia (Long-acting injectable aripiprazole approved in Canada, NeuroSens, August 13, 2014). A newly-published phase III study has now examined the efficacy of aripiprazole OM for acute episodes of schizophrenia (Kane et al. J Clin Psychiatry 2014; epublished August 19, 2014; free full text at http://www.psychiatrist.com/jcp/article/Pages/2014/v75n11/v75n1107.aspx [requires registration]).
A total of 304 patients were randomized to aripiprazole OM 400 mg or placebo for 12 weeks. All subjects received concomitant oral aripiprazole 10-20 mg/day for 14 days. A significant treatment effect was seen at week 1, and this benefit was maintained at all time points throughout the 12-week study. At the primary endpoint (10 weeks), the least squares mean change from baseline was significantly greater with aripiprazole OM versus placebo for PANSS total score (treatment difference -15.1) and CGI-Severity of illness (treatment difference -0.8). The most common adverse effects were weight gain (16.8% vs. 7.0% with placebo), headache (14.4% vs. 16.3% with placebo) and akathisia (11.4% vs. 3.5% with placebo). The authors concluded that aripiprazole OM is a viable treatment option for acute schizophrenia.
The comparative effectiveness of oral and LAI aripiprazole over the longer term was examined in the ASPIRE EU trial (Fleischhacker et al. Br J Psychiatry 2014;205:135-144; free full text at http://bjp.rcpsych.org/content/205/2/135.full.pdf+html). The impending relapse rate at week 26 was 7.12% for aripiprazole OM 400 mg versus 7.76% for oral aripiprazole, and aripiprazole OM was shown to be non-inferior to oral aripiprazole. The six-month treatment continuation rate was higher with aripiprazole 400 mg OM compared to oral aripiprazole (81.1% vs. 75.6%), and the all-cause discontinuation rates were 25.3% and 32.7%, respectively.
An extension study that enrolled de novo subjects and patients from two registration trials reported that the overall discontinuation rate at 12 weeks was 12-13% (Fleischhacker et al. 4th Schizophrenia International Research Society Conference, Florence, April 5-9; abstract). At 52 weeks, the overall completion rate with aripiprazole OM was 79.4% (Peters-Strickland et al. APA annual meeting, New York, May 3-7, 2014; abstract). A total of 95.0% of subjects met stability criteria at their last assessment. During the one-year period, the proportion of patients with impending relapse (8.3%) or requiring oral antipsychotic therapy (7.1%) was low. The discontinuation rates due to lack of efficacy or adverse effects were 4.0% and 2.9%, respectively. Aripiprazole OM has also been shown to produce improvements in daily function, most notably in the domains of personal and social relationships, self-care, and disturbing and aggressive behaviour (Fleischhacker et al. Schizophr Res 2014;159:415-420).
In the community setting, a recent mirror-image study found that when patients with a history of psychiatric hospitalization during oral therapy were switched to aripiprazole OM, in-patient stays were significantly reduced compared to the pre-LAI phase when they were taking oral medication (Kane et al. J Med Econ 2014;November 10, 2014; free full text at http://informahealthcare.com/doi/abs/10.3111/13696998.2014.979936). The rates of psychiatric hospitalization at three months were 2.7% with aripiprazole OM versus 27.1% while on an oral antipsychotic; at six months, the hospitalization rates were 8.8% versus 38.1%, respectively.
A reduction in hospitalization rates would be expected to be associated with substantial reductions in the cost of care. Resource utilization in Quebec was determined in an analysis of the RAMQ (Régie de l’assurance maladie du Québec) database for a cohort of patients (n=1,992) who received a prescription for an LAI (Lachaine & Lapierre. Peripharm, January 2014). The one-year hospitalization rate declined from 74.5% to 48.1% after LAI initiation, and the mean duration of hospital stay was reduced 47.7%. Total healthcare costs (outpatient visits, inpatient and outpatient costs, medications) per patient were reduced from CDN$24,382 in the year preceding LAI use to CDN$13,090 in the year following LAI initiation.
Current recommendations for LAI use
While LAI antipsychotics offer a number of advantages over oral agents in reducing patient morbidity and the cost of care, a number of recent surveys have observed that clinicians are often reluctant to initiate an LAI antipsychotic early in the treatment course (Samalin et al. J Nerv Ment Dis 2013;201:553-559; Geerts et al. BMC Psychiatry 2013;13:58). This attitude appears to be especially common in Canada, where physicians often see LAIs as a treatment of last resort for patients with a history of poor adherence (Iyer et al. Can J Psychiatry 2013;58[5 Suppl 1]:23S-29S). As a result, LAI antipsychotic use in Canada is estimated to be 6.3% (Williams et al. Acta Psychiatr Scand Suppl 2006;113:12-21), lower than the rate reported for any other country (Manchanda et al. Can J Psychiatry 2013;58[5 Suppl 1]:5S-13S; free full text at http://publications.cpa-apc.org/media.php?mid=1506).
Canadian and Québec clinical guidelines now recommend that the use of LAI antipsychotics should be discussed with patients and families throughout the disease course, most importantly during the first 2-5 years after diagnosis (Malla et al. Can J Psychiatry 2013;58[5 Suppl 1]:30S-35S; free full text at http://publications.cpa-apc.org/media.php?mid=1509); Stip et al. Can J Psychiatry 2011;53:367-376). Patients and their families should be informed about the potential advantages of monthly injections in reducing the risk of relapses and hospitalizations, and the benefits of adherence throughout the treatment course.
Comment
Dr. Marc-André Roy: The underutilization of LAIs in Canada has been the subject of much attention in recent years. Besides patients’ reluctance in accepting them, prescribers’ misconceptions of the role of LAIs – often seen as inherently coercive or as an option only in very severely ill patients – contribute to this underutilization.
But when LAIs are presented properly in the course of shared decision-making between the clinician and patients, a fairly large proportion will choose this treatment option. Furthermore, most experts in the field advocate that LAIs be offered early on in the course of treatment, given their positive effect on treatment adherence, and given the dramatic negative impact of relapses on the outcome of people at the early stage of psychosis, as stressed in the present article.
Another issue that certainly contributes to underutilization is that there is still only a limited array of molecules available in an LAI formulation. Given the broad differences in side-effect profiles across molecules, and of efficacy and tolerability in given individuals, it may be difficult for some patients to find a long-acting injectable that will be both efficacious and well-tolerated. Hence, the addition of new long-acting forms of antipsychotics is always very good news for clinicians and their patients. The recent approval of a LAI formulation of aripiprazole is thus welcomed by clinicians, particularly given its relatively benign side-effect profile, most notably with regard to extrapyramidal and metabolic effects.
Dr. Roy is Associate Professor of Psychiatry and Neurosciences, Université Laval, Québec City, Canada.