Initially, we performed an analysis and a literature review concerning the indications and the use of LAI antipsychotics. A literature search using the keywords “antipsychotic”, “neuroleptic”, “first-generation antipsychotic”, “atypical antipsychotic”, “second-generation antipsychotic”, “long-acting injectable”, “depot”, “depot neuroleptic” was performed in PubMed and EMBASE to find all the relevant studies published. Additional references were identified from http://www.fda.gov and http://www.ema.europa.eu.

Data from all of these sources was discussed and an overview of the current evidence has been graded and summarized using the French National Authority for Health (HAS) “levels of evidence” criteria 16 .

Following this first step, the scientific committee (PML, LS, MA, PC, SG, SL) created a questionnaire consisting of 32 questions that covered 539 therapeutic options. The 32 questions were regrouped into 3 areas that were judged as essential:

Target-population: Description of the different indications of the LAI antipsychotics and of the most appropriate period of the illness to introduce the treatment.

Prescription and use: Choice of the molecule, methods of introduction, specific strategies depending on the psychiatric disorder or co-morbidities, and treatment monitoring.

Specific population: Use of LAI antipsychotics in pregnant women, elderly patients, subjects in a precarious situation, and subjects having to be treated in a prison establishment.

This questionnaire was designed to be completed by an experts’ panel. The time required for its administration was estimated at around 3 hours.

At the time of development, all the LAI antipsychotics available in France were proposed as therapeutic options (Table 1). They were regrouped into 2 categories:

Long-acting injectable first-generation antipsychotics (LAI FGA).

Long-acting injectable second-generation antipsychotics (LAI SGA).

Note: as paliperidone palmitate had a marketing authorization date after the development of these guidelines, it could not be taken into account.

This artificial separation FGA/SGA is not consensual due to their heterogeneous profiles of efficacy and tolerability (especially for SGA) 17 18 . However, we maintained both these categories to facilitate the elaboration, the reading and the understanding of this guideline.

The experts were able to express their level of agreement or disagreement for each question. The rules that describe, on the one hand the agreement (or the disagreement), and on the other hand the degree of convergence of the expert opinions, were predefined.

Each expert answered each question with the help of a graduated scale from 0 to 9 (0 meaning a “total disagreement” or “a formal contraindication” and 9 indicating a “total agreement” or “a formal indication”) (Figure 1).

This scale is derived from a variation of the “Nominal Group” method, developed by the Rand Corporation and the University of California in the USA (“RAND/UCLA appropriateness rating method”).

The scale has the advantage of being well standardized, used in medicine and of having been published 19 .

The Scientific Committee (Appendix 1) selected 53 French psychiatrists considered to be experts in the use of LAI antipsychotics.

The selection criteria were:

Clinical activity in the field of schizophrenia or bipolar disorder.

Publication (s) or communication (s) of research work in the field of LAI antipsychotics in national or international congresses.

Each expert was consulted individually for his or her expertise using the supplied questionnaire. A follow-up was established to ensure, essentially, the sending and returning of these documents. A reminder of the expert's mission was made. Written commitment from each expert was requested. The experts participating in this project were not paid.

The answers to each question or proposition were quantitatively analyzed (number of answers, median, mean, standard deviation, minimum, maximum) (Table 2).

First-line treatment/strategy was defined if at least 50% of the answers to the question were in the zone 7–9 and less than 20% were in the zone 0. The treatment/strategy of choice was kept if at least 50% of the experts had rated it 9.

Second-line treatment/strategy was defined if less than 50% of the answers to the question were in the zone 7–9, at least 50% were in the cumulated zones 7–9 and 4–6, and less than 20% were in the zone 0.

Third-line treatment/strategy was defined if less than 50% of the answers to the question were in the cumulated zones 7–9 and 4–6, and less than 20% were in the zone 0.

Contraindication was defined if at least 50% of the answers were in the zone 0.

For all other cases (-), the question was considered as non-consensual.

The following rules were used to conclude the analyzed therapeutic strategy.

For all other cases the question was considered as non-consensual. An example is given in Figure 2. The results were interpreted by the scientific committee and permitted the development of the recommendations. An independent committee (Appendix 1) validated the final version of recommendations (EH, CL, PT). Two members of the scientific committee elaborated the final document (LS, PML).

Forty-two experts completed the questionnaire (Appendix 2), representing 79% of those contacted. The reasons for the non-participation of the remaining 11 experts were that they had either too much consultancy work or insufficient availability to reply within the time limits. The socio-demographic data and professional activities of the experts’ panel are presented in Table 3.

In the 1^st/2^nd/3^rd trimester: The experts failed to reach a consensus for 1^st line strategies. As 2^nd line strategies continuation of the LAI antipsychotic or switching to an oral form (FGA or SGA at the minimum effective dose) is recommended.

Most guidelines for the treatment of psychiatric disorders are evidence-based guidelines (EBG) 11 20 . However, recommendations cannot be established if there is no evidence available, in which case, CBG methodology can be used. The French National Health agency 19 recommends the Formal Consensus method when two of the following conditions are met:

No or insufficient level of evidence addressing the question.

Possibility to decline the topic in easily identifiable clinical situations.

Need to identify and select the strategies deemed appropriate by an independent panel from amongst several alternative options.

This method is very close to the Expert Consensus Guidelines methodology and has been applied to a variety of psychiatric disorders 4 21 22 23 24 25 26 27 .

Combining EBG and CBG methodologies may help clinicians to have a real evidence-based clinical practice, including both clinical expertise and scientific evidence 20 .

In the field of LAI antipsychotic use and management, CBG methodology appears to be particularly appropriate. Evidence concerning LAI antipsychotic efficacy and tolerability exists but it is lacking in many areas (i.e. indications or preferential patient profiles, a ranking system between LAI antipsychotics, the introduction stage, process for switching, medication management, specific populations…). CBGs allow the clinician to be led by recommendations that bear a closer relation to the characteristics of the patients followed in clinical practice than to the restrictive inclusion criteria of randomized-controlled trials 20 .

According to our experts’ panel, LAI antipsychotics are recommended as first-line treatment in various psychiatric disorders:

Schizophrenia.

Schizoaffective disorder.

Delusional disorder.

But also as second-line treatment in:

Bipolar disorder.

Personality disorder.

If their use in schizophrenia is common and supported by evidence 5 6 7 28 , their use in bipolar disorder is less obvious. Nevertheless, several placebo-controlled relapse prevention studies have shown the efficacy of risperidone microsphere as a monotherapy or as an adjunctive therapy to lithium or valproate in bipolar I patients 29 . In September 2011, and based on this data, the Food and Drug Administration Agency approved risperidone microsphere as a long-term treatment for bipolar I disorder. Scientific literature is currently limited to risperidone microsphere but the development of new drugs should allow further studies with LAI SGA as maintenance treatment for bipolar disorder.

The use of LAI antipsychotics in other indications (schizoaffective disorder, delusional disorder, personality disorder) is not based on evidence for these populations but is instead based on the clinical experience of our experts’ panel. If scientific evidence is required then the sharing of this experience can be considered as a real support for the clinical use of these compounds.

In recent years the interest of using LAI SGA in the early phase of schizophrenia has increased because the duration of untreated psychosis is associated with the prognosis of the illness 30 . Recent studies have underlined the fact that their use, as early as the first psychotic episode, offers many advantages in terms of efficacy, tolerance and improved adherence 31 32 33 . The available literature presents a weak level of evidence (open label, post-hoc analysis, and small sample) and placebo-controlled studies are needed. The formalized consensus of our experts’ panel is consistent with these preliminary results and recommends LAI SGA after the first schizophrenic episode.

The extension of this data to the first manic episode in bipolar disorder could be assumed but, to date, no data has emerged that compares the effect of the early introduction of oral or LAI antipsychotics on the course of the illness. This is probably the reason why the experts’ panel did not recommend LAI SGA in the early course of bipolar disorder as a maintenance treatment.

Our experts’ panel considers that LAI antipsychotics should be used with any patients with schizophrenia for whom maintenance antipsychotic treatment is indicated. This is consistent with the results of a survey conducted among psychiatrists from Europe, Middle-East and Africa, in which clinicians considered switching to or adding an LAI antipsychotic as the preferential pharmacological approach for addressing adherence problems 34 .

LAI antipsychotics have long been viewed as a treatment that could only be used for a small subgroup of patients with non-compliance, frequent relapses or who pose a risk to others. A cluster analysis of French and German studies, surveying psychiatrists about patient attributes that potentially influence their qualification for depot treatment, identified two clusters of patients 35 36 . Cluster I corresponded to the classical patient profile in whom depot forms are used (past history of relapse and poor compliance with oral forms). Cluster II was more unexpected and included patients with high levels of insight and of therapeutic alliance. The usefulness of depot formulations compared with oral treatment in terms of relapse prevention is not demonstrated in this population 35 36 . However, even limited gaps of treatment with oral formulation (11–30 days a year) is enough to increase the risk of relapse by 2.81 in patients with schizophrenia 37 . The identification of the two clusters, replicated in numerous countries, is consistent with the recommendation of our experts’ panel.

Considering the risks associated with non-compliance in bipolar patients 38 39 , the experts’ panel recommends LAI antipsychotics as a second-line treatment in bipolar disorder.

If the superiority of LAI antipsychotics versus placebo, in terms of relapse prevention, has been demonstrated 5 28 for schizophrenia, no study compares the LAI SGA versus LAI FGA.

We can only extrapolate the results from studies on oral antipsychotics. Some individual oral SGA (amisulpride, clozapine, olanzapine, risperidone) were better in overall efficacy in patients with schizophrenia than oral FGA 17 . Other oral SGA were no more effective, even for negative symptoms. However, a meta-analysis which considered all oral SGA as a single group demonstrated that they were associated with fewer relapses, less treatment failures and fewer hospitalizations in the long-term treatment of schizophrenia 40 . Oral SGA induced fewer extrapyramidal side effects than oral FGA but some SGA induced more weight gain or metabolic side effects than oral FGA 41 42 . Tolerance profiles of oral SGA are more mixed and require the characteristics of each molecule to be taken into account on an individual basis.

In a one-year observational study including 1859 patients diagnosed with schizophrenia, an adjusted Poisson regression analysis showed that the use of risperidone microsphere was associated with a lower rate of hospitalization compared to the use of other LAI FGA 43 .

So, with no evidence available, the experts’ panel recommended that the clinician’s decision-making process takes into account the benefit/risk balance and prioritizes LAI SGA (except for olanzapine pamoate, due probably to the risk of post-injection syndrome 44 ) over LAI FGA, according to patient tolerance.

The management of LAI antipsychotics in clinical practice can sometimes be complex for clinicians and there are limited data or recommendations in the literature. Our guidelines try to propose practical recommendations to facilitate the introduction, switching and management of LAI antipsychotics in the different phases of schizophrenia or bipolar disorder.

Indeed, the current EBG for biological treatment of schizophrenia and bipolar disorder 8 9 10 45 46 47 48 49 50 51 52 53 propose few recommendations concerning LAI antipsychotics.

Most of them recommend the use of LAI antipsychotics only for patients with non-adherence, frequent recurrence or who prefer this formulation. The conditions of use and management are not, or are only briefly, described. LAI antipsychotics are presented separately from the oral medication strategies (except for the CANMAT guidelines in bipolar disorder).

The main reasons given in explanation for the limited number of recommendations regarding LAI antipsychotics are related to the lack of long-term studies and the lack of high-quality evidence comparing LAI SGA to oral SGA. Perhaps the follow-up period, lasting a year or less, may have been too short to reveal the longer-term benefits of depot treatment versus oral form 9 46 .

However, in our opinion, the current criteria for level of evidence are probably not adapted to the studies dealing with LAI antipsychotics. Indeed, randomized-controlled trials have a major selection bias and cannot assess the potential adherence benefits of LAI formulations (non-compliant patients do not participate in a trial and those who accept to be included are the most compliant). Therefore, it can be difficult to demonstrate the benefit of LAI antipsychotics compared with oral antipsychotics. Future studies with LAI antipsychotics should combine the strengths of the different study designs (randomized-controlled studies, mirror-image studies or cohort studies).

In addition to these EBG, there are some CBG focusing on the use and management of LAI formulations for the treatment of schizophrenia 4 27 54 55 56 57 .

The first guidelines, published in 1998, already recommended that LAI FGA should be considered for “any patients with schizophrenia for whom long-term treatment is indicated” 54 . However, with the emergence in the years that followed of oral SGA, which are better tolerated compared to FGA, most of the guidelines have been in favour of the use of the oral formulation. Since the market authorization (2002) of the first LAI SGA (risperidone microsphere), two other specific guidelines concerning LAI antipsychotics 27 57 have been proposed. These guidelines recommended LAI SGA as first-line treatment for patients who request the long-acting formulations. Their use after the first schizophrenic episode or for patients who are stable with oral antipsychotics has been discussed.

In 2009, Velligan et al. published expert consensus guidelines about adherence problems in patients with serious mental illness 4 . Use of LAI antipsychotics was a personal choice for patients with frequent relapses associated with non-adherence, relapses because they stopped taking the medication, or because they expressed a preference for the LAI formulation.

The Association des médecins psychiatres du Québec (AMPQ) has also recently developed guidelines concerning LAI antipsychotics with a decisional algorithm, which places the depot formulation in every step of treatment as soon as possible 56 .

•French Association of Biological Psychiatry and Neuropsychopharmacology (Association Française de Psychiatrie Biologique et Neuropsychopharmacologie - AFPBN -).

•Professor Pierre-Michel Llorca/Doctor Ludovic Samalin.

•Doctor Mocrane Abbar.

•Professor Philippe Courtet.

•Professor Pierre-Michel Llorca.

•Doctor Sebastien Guillaume.

•Doctor Ludovic Samalin.

•Sylvie Lancrenon.

•Professor Emmanuel Haffen.

•Professor Christophe Lançon.

•Professor Pierre Thomas.