Author's Accepted Manuscript
Hyperprolactinemia and medications for bipolar disorder: Systematic review of a neglected issue in clinical practice Isabella Pacchiarotti, Andrea Murru, Georgios D. Kotzalidis, C. Mar Bonnin, Lorenzo Mazzarini, Francesc Colom, Eduard Vieta
www.elsevier.com/locate/euroneuro
PII: DOI: Reference:
S0924-977X(15)00107-8 http://dx.doi.org/10.1016/j.euroneuro.2015.04.007 NEUPSY11011
To appear in:
European Neuropsychopharmacology
Received date: 15 September 2014 Revised date: 2 March 2015 Accepted date: 1 April 2015 Cite this article as: Isabella Pacchiarotti, Andrea Murru, Georgios D. Kotzalidis, C. Mar Bonnin, Lorenzo Mazzarini, Francesc Colom, Eduard Vieta, Hyperprolactinemia and medications for bipolar disorder: Systematic review of a neglected issue in clinical practice, European Neuropsychopharmacology, http: //dx.doi.org/10.1016/j.euroneuro.2015.04.007 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Hyperprolactinemia and medications for Bipolar Disorder: systematic review of a neglected issue in clinical practice Isabella Pacchiarotti 1, Andrea Murru 1, Georgios D. Kotzalidis 2, C. Mar Bonnin 1, Lorenzo Mazzarini2, Francesc Colom 1, Eduard Vieta 1
1
Bipolar Disorders Unit, Institute of Neuroscience, Hospital Clínic, University of Barcelona, IDIBAPS, CIBERSAM, Barcelona, Catalonia, Spain. 2
NESMOS Department (Neuroscience, Mental Health, and Sensory Organs), Sapienza University, School of Medicine and Psychology, Sant'Andrea Hospital, Rome, Italy.
Corresponding author: Prof. Eduard Vieta Bipolar Disorder Unit, Hospital Clínic, c/Villaroel 170, esc. 12 pl. 0 08036 Barcelona (ES) Tel: +349322754 int. 3130 Fax: +34932275795 Email: [email protected]
ABSTRACT Drug-induced changes in serum prolactin (sPrl) levels constitute a relevant issue due to the potentially severe consequences on physical health of psychiatric patients such as sexual dysfunctions, osteoporosis and Prl-sensitive tumors. Several drugs have been associated to sPrl changes. Only antipsychotics have been extensively studied as sPrlelevating agents in schizophrenia, but the extent to which bipolar disorder (BD) treatments affect sPrl levels is much less known. The objective of this systematic review is to summarize the evidence of the effects of drugs used in BD on Prl. This review followed the PRISMA statement. The MEDLINE/PubMed/Index Medicus, EMBASE, and Cochrane Library databases were systematically searched for articles in English appearing from any time to May 30, 2014. Twenty-six studies were included. These suggest that treatments for BD are less likely to be associated with Prl elevations, with valproate, quetiapine, lurasidone, mirtazapine, and bupropion reported not to change PRL levels significantly and lithium and aripiprazole to lower them in some studies. Taking into account the effects of the different classes of drugs on Prl may improve the care of BD patients requiring long-term pharmacotherapy. Based on the results of this review, lithium and valproate appear to be safer due to their low potential to elevate sPrL; among antipsychotics, quetiapine, lurasidone and aripiprazole appear to be similarly safe.
KEY WORDS Hyperprolactinemia, bipolar disorder, antipsychotics, mood stabilizers, antidepressants
INTRODUCTION There is growing interest on the burden of drug-related side effects in people with severe mental illnesses (Whiteford et al., 2013; Gladigau et al., 2014). The long-term, potentially lifelong need for treatment arose concerns about physical consequences, like weight gain and metabolic/endocrine disturbances (McElroy and Keck, 2014). Among these, treatment-related changes in serum prolactin (sPrl) and possible hyperprolactinemia (Byerly et al., 2007) raise important issues, due to potentially severe effects on patients’ physical health. Hyperprolactinemia is usually defined as sustained prolactin (Prl) elevation above normal laboratory levels. These are below 530 mIU/L (25 ng/mL) in women and 424 mIU/L (20 ng/mL) in men with the more commonly used assays (1 ng/mL is equivalent to 21.2 mIU/L, WHO Standard 84/500), even if ranges among laboratories may differ (Peveler et al., 2008). In some patients, sexual dysfunction or menstrual abnormality can be useful indicators of elevated Prl, although these symptoms are nonspecific, especially in psychiatric patients on medication (Bushe et al., 2010). In addition, it is important that clinicians realize the clinical significance of Prl elevations despite absence of symptoms (Bushe et al., 2010). Nonetheless, even an eight-year long treatment with Prl-elevating drugs may lead to significant bone mineral density loss in young women, and osteoporosis secondary to persistent elevation of sPrl levels doubles the risk of fractures (Byerly et al., 2007; Sørensen et al., 2013). Other worth-mentioning long-term effects are gynecomastia, hypogonadism, prolactinoma and Prl-sensitive tumors (Byerly et al., 2007; Rettenbacher et al., 2010). Several drugs have been associated to sPrl elevations and hyperprolactinemia. Among psychiatric drugs, antipsychotics (APs) have been extensively studied (de Hert et al., 2014; Peuskens et al., 2014). Studies mainly focused on samples with schizophrenia, despite APs are used widely as first- and second-line treatment of patients with bipolar disorder (BD) (Grunze et al., 2013). However, drug treatment strategies in BD include also other options, like “classical” mood stabilizers (e.g., lithium, valproate) and antidepressants (ADs), and frequently their combination. The extent to which BD treatments affect sPrl levels has not been systematically reviewed. The objective of this systematic review is to critically consider and summarize the evidence of how drugs approved for use in BD affect blood PrL levels.
EXPERIMENTAL PROCEDURES This review has been conducted according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement (Moher et al., 2009). LITERATURE SEARCH We systematically searched the MEDLINE/PubMed/Index Medicus, EMBASE, and Cochrane Library databases from any time to May 30, 2014, cross-checking the obtained references. We further searched the www.clinicaltrials.gov site to check literature results and to obtain information about starting, ongoing, or concluded studies. The systematic search was performed by two blind independent researchers (IP and AM), who searched for randomized controlled trials (RTCs), as well as open studies and systematic reviews on the topic, as follows:
• MEDLINE/Pubmed/Index Medicus: authors used the keywords (prolactin*[ti] OR hyperprolactin*[ti] OR galactorr*[ti] OR gynecomast*[ti]) AND (bipolar[ti] OR mania[ti] OR manic[ti]) to identify studies of potential interest. In order to catch a wider selection of search returns, a broader string was also used by eliminating the title specifier: (prolactin* OR hyperprolactin* OR galactorr* OR gynecomast*) AND (bipolar OR mania OR manic). • Cochrane Library: keywords were (hyperprolactin* OR galactorr* OR gynecomast*) AND (bipolar OR mania OR manic). • Clinicaltrials.gov: keywords were hyperprolactinemia bipolar disorder. • EMBASE: hyperprolactinemia, bipolar disorder. STUDY SELECTION The first step was to screen whether the study investigated the effects of drugs (i.e. APs, ADs or “classical” mood stabilizers) approved for treating BD or tested in randomized clinical trials (RCTs) of BD on Prl blood levels in BD samples, based on abstract or full text. For this reason, studies with no therapeutic objectives were excluded. Data were grouped as: 1. Of interest, 2. Duplicates, and 3. Of no interest, according to abovedescribed criteria. After selection, full papers were retrieved and studied. DATA COLLECTION, PROCESSING, AND ITEMS To be included in this review, a paper had to clearly report effects on blood (serum or plasma) Prl levels and details of design, sample description, and inclusion/exclusion criteria.
RESULTS SYSTEMATIC SEARCH RESULTS Figure 1 is a flow-chart of considered and finally selected studies, endorsing the aforementioned PRISMA statement. The global search first returned 410 titles. We excluded a small proportion of duplicates or irrelevant records (37), so that 373 underwent title and abstract examination. Further 265 opinions, case reports, and reviews were excluded. One-hundred-and-eight studies were deemed eligible for further assessment. Among them, 1 is still ongoing, 1 inaccessible, and 80 unrelated to the aim of this review. Further 4 studies (Yatham, 1996, Sher et al., 2003; Cordes et al., 2011; Lykouras et al., 2011) tested a serotonergic hypothesis with AD-challenge and were subsequently excluded. The final selection included 28 studies, 8 returned with strict Pubmed/Medline/Index Medicus criteria and 20 with broad criteria. Figure 1 about here CONTENT RESULTS
The 28 included studies are summarized in Table 1. Among them, 23 studies presented at least one treatment arm with APs, 6 with lithium, 2 valproate and only 1 with the selective serotonin reuptake inhibitor (SSRI) AD, citalopram. Table 1 about here ANTIPSYCHOTICS AP effects on sPrl have been widely studied in schizophrenic samples (Stip et al., 2011; Gentile et al., 2013; Leucht et al., 2013; De Hert et al., 2014; Peuskens et al., 2014). The only study included here that assessed the general and drug-specific incidence of sPrl changes in AP-treated BD populations is a naturalistic cohort with 194 patients with schizophrenia and 49 with BD who underwent routine Prl screening in absence of important symptoms (Bush and Shaw, 2007). Above-threshold Prl levels were found in 38% of the cohort, more commonly in female (52%) than in male patients (26%). In this study, most women on risperidone monotherapy, either long-acting injectable (RLAI) or oral (RSP), developed severe hyperprolactinemia (>1000mIU/L). A trend for greater prevalence of hyperprolactinemia was found with oral RSP compared to RLAI (p = 0.057). Patients on amisulpride (AMI) also showed abnormal levels, independently from gender. In contrast, hyperprolactinemia was seen only in one olanzapine (OLZ)treated woman. Prevalence of hyperprolactinemia in monotherapy patients was presumably associated to single APs. Hyperprolactinemia was found in 100% AMI(7/7), 69% RSP- (24/35), 33% typical AP- (TAP, 27/83), 6% OLZ- (1/16), 5% clozapine- (CLZ, 1/21), and 0% aripiprazole- (ARP) or quetiapine (QTP)-treated patients (0/3). Last, most polypharmacy involved the concomitant use of typical and atypical APs. TAPs combined with, or added to, a single atypical AP did not increase hyperprolactinemia rates significantly in patients on RSP (69 vs. 67%), or OLZ (6 vs. 32%). Substituted benzoamides In an open-label 8-week study (Chen et al., 2010), adjunctive ARP to AMI and sulpiride (SUL) was ineffective in restoring baseline hyperprolactinemia (p=0.099). In an openlabel study, AMI induced galactorrhea in one of 20 patients with mania (Vieta et al., 2005). Aripiprazole Chen et al. (2010) compared in a mixed sample of patients with schizophrenia, schizoaffective disorder, and BD the effects of 8-week adjunctive ARP treatment on RSP-, AMI-, and SUL-induced hyperprolactinemia. ARP dose was 5-20 mg/day during the study period. sPrl levels were measured at weeks 4 and 8; they decreased significantly from baseline in the ARP+RSP-treated (14 of 15 patients, 93.33%, p < 0.001), but not in the ARP-AMI/ARP-SUL-treated patients (1 of 10 patients, 10%). Brown’s et al. (2013) meta-analysis of 10 studies aimed at assessing the role of ARP in acute mania. Compared to placebo (n=2239), mean Prl levels in the ARP groups dropped, in some cases below the lower normal threshold. However, data from studies measuring prolactinemia were insufficient for conducting meta-analysis.
ARP was compared to Li by Keck et al. (2009). Patients on ARP showed significantly greater mean sPrl drops than lithium patients during 12 weeks (−6.6 ng/mL vs. −2.2 ng/mL; p < 0.001). In studies comparing Hal to ARP (Vieta et al., 2005; Young et al., 2009), fewer participants taking ARP showed Prl increases above normal levels compared to Hal (Vieta 2005; Young 2009). In the first study (Vieta et al., 2005), mean sPrl levels decreased from baseline in the ARP (713.4 ng/ml, 7284.1mU/l), and increased in the Hal group (7.7 ng/ml, 7163.2mU/l) at week 12; means differed significantly (p < 0.001). Similarly, in the Young et al. (2009) study, ARP-treated patients showed a decrease in mean sPrl levels from baseline over 12 weeks, whereas Hal-treated patients showed an increase (713.4 ng/ml vs. + 6.7 ng/ml; p < 0.001). At endpoint, fewer ARP- (12.8%) than Hal-treated patients (60.8%) had new-onset hyperprolactinemia. A post hoc analysis of 28 rapid-cycling BD-I patients from a 100-week, double-blind, placebo-controlled study assessing long-term efficacy, safety and tolerability of ARP in patients with recently manic or mixed BD type I showed no significant changes from baseline in Prl levels with either ARP or placebo at the 100th week (p > 0.05) (Muzina et al., 2008). Asenapine During a 3-week RCT with asenapine (ASE) for acute mania, Prl levels did not change significantly from baseline, compared to placebo and OLZ (McIntyre et al., 2009). In the 40-week extension (McIntyre et al., 2010), shifts to high Prl levels in patients with normal or low baseline Prl levels at any time during treatment occurred in 33.3% of placebo-ASE treated patients (3 weeks of placebo and then shifted to ASE with blind unbroken), 34.3% ASE-treated patients, and 61.9% of OLZ-treated patients. Lurasidone Low potential for hyperprolactinemia was reported for lurasidone (LUR) in a 6-week monotherapy, fixed-dose, RCT (Loebel et al., 2014a). Median endpoint change in Prl was not clinically meaningful, but women on LUR had higher Prl levels compared to placebo. In the 6-week RCT of LUR as add-on to Li or DVP in bipolar depression, reported Prl changes from baseline were not significant (Loebel et al 2014b). Olanzapine OLZ was tested in a 3-week, head-to-head RCT vs. RSP (Perlis et al., 2006), showing significantly lower Prl elevation than RSP (p>0.001). In a 3-week, placebo-controlled, double-blind RCT with DVP as active comparator, OLZ-treated patients showed significant Prl elevations compared to placebo-treated (p < 0.001) and DVP-treated patients (p < 0.001) (Tohen et al., 2008). OLZ was used as an active comparator to ASE both in acute (3-weeks) and long-term (40-weeks) clinical trials (McIntyre et al., 2009; 2010). At week 40, Prl increases from baseline in the OLZ-treated arm were double compared to the 2 ASE-treated arms. In a 3-week RCT of adolescents with acute bipolar mania (Tohen et al., 2007), mean baseline-to-endpoint Prl increases were significantly greater in the OLZ group relative to the placebo group (p < 0.001). Abnormally high Prl levels occurred significantly more often in patients receiving OLZ compared to placebo (girls, 25.7% vs. 0%, p = 0.007; boys, 62.5% vs. 5%, p < 0.001).
In a two-phase RCT of PAL in patients with manic or mixed symptoms, the active comparator OLZ increased Prl levels in men and decreased them in women from baseline to week 15 (Bewaerts et al., 2012b). Paliperidone Two randomized, PBO-controlled studies with PAL, using QTP (Vieta et al., 2010) or OLZ (Bewaerts et al., 2012b) as active comparators in manic and mixed patients, found that PAL acutely induced greater increases in sPrl levels (Vieta et al., 2010) which persisted during chronic administration. In another acute randomized, PBO-controlled trial in patients with manic or mixed symptoms, PAL was associated with higher sPrl levels in women across all three PAL doses, which were proportional to daily dose (Bewaerts et al., 2012a). Finally, increased sPrl levels from baseline to the 6-week endpoint were found in another randomized, PBO-controlled study with adjunctive PAL (Berwaerts et al., 2011). Quetiapine In QTP monotherapy studies (Bowden et al., 2005; McIntyre et al., 2005), mean Prl decreased in QTP- and placebo-treated groups from baseline to the 12-week end-point. More QTP-treated (61 of 209) than placebo-treated (46 of 198) patients shifted from elevated to normal Prl levels, but between-groups differences were not significant (Vieta et al., 2005). Effects on serum prolactin were not monitored in studies examining QTP combined with Li/DVP (Adler et al., 2007). In a PBO-controlled RCT of PAL in patients with manic or mixed symptoms (Vieta et al., 2010), QTP used as the active comparator induced less sPrl increases compared to PAL during both acute and maintenance phases. Risperidone In a 3-week, randomized, controlled, double-blind, parallel multicenter study (Perlis et al., 2006) RSP (1–6 mg/day; N = 164) was compared to OLZ (5–20 mg/day; N = 165) in inpatients with acute manic/mixed symptoms. RSP-treated patients were more likely to show Prl elevations (51.73 ng/mL vs. 8.23 ng/mL; p < 0.001). RSP-induced hyperprolactinemia was successfully reversed by adding ARP in another study (Chen et al., 2010). A 4-week study of RSP in AP-naïve patients with schizophrenia, psychotic depression or BD, showed significant Prl increases from 12.1 ± 1.9 ng/ml at baseline to 65.7 ± 12.2 ng/ml post-treatment (p
Hyperprolactinemia and medications for bipolar disorder: Systematic review of a neglected issue in clinical practice Isabella Pacchiarotti, Andrea Murru, Georgios D. Kotzalidis, C. Mar Bonnin, Lorenzo Mazzarini, Francesc Colom, Eduard Vieta
www.elsevier.com/locate/euroneuro
PII: DOI: Reference:
S0924-977X(15)00107-8 http://dx.doi.org/10.1016/j.euroneuro.2015.04.007 NEUPSY11011
To appear in:
European Neuropsychopharmacology
Received date: 15 September 2014 Revised date: 2 March 2015 Accepted date: 1 April 2015 Cite this article as: Isabella Pacchiarotti, Andrea Murru, Georgios D. Kotzalidis, C. Mar Bonnin, Lorenzo Mazzarini, Francesc Colom, Eduard Vieta, Hyperprolactinemia and medications for bipolar disorder: Systematic review of a neglected issue in clinical practice, European Neuropsychopharmacology, http: //dx.doi.org/10.1016/j.euroneuro.2015.04.007 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Hyperprolactinemia and medications for Bipolar Disorder: systematic review of a neglected issue in clinical practice Isabella Pacchiarotti 1, Andrea Murru 1, Georgios D. Kotzalidis 2, C. Mar Bonnin 1, Lorenzo Mazzarini2, Francesc Colom 1, Eduard Vieta 1
1
Bipolar Disorders Unit, Institute of Neuroscience, Hospital Clínic, University of Barcelona, IDIBAPS, CIBERSAM, Barcelona, Catalonia, Spain. 2
NESMOS Department (Neuroscience, Mental Health, and Sensory Organs), Sapienza University, School of Medicine and Psychology, Sant'Andrea Hospital, Rome, Italy.
Corresponding author: Prof. Eduard Vieta Bipolar Disorder Unit, Hospital Clínic, c/Villaroel 170, esc. 12 pl. 0 08036 Barcelona (ES) Tel: +349322754 int. 3130 Fax: +34932275795 Email: [email protected]
ABSTRACT Drug-induced changes in serum prolactin (sPrl) levels constitute a relevant issue due to the potentially severe consequences on physical health of psychiatric patients such as sexual dysfunctions, osteoporosis and Prl-sensitive tumors. Several drugs have been associated to sPrl changes. Only antipsychotics have been extensively studied as sPrlelevating agents in schizophrenia, but the extent to which bipolar disorder (BD) treatments affect sPrl levels is much less known. The objective of this systematic review is to summarize the evidence of the effects of drugs used in BD on Prl. This review followed the PRISMA statement. The MEDLINE/PubMed/Index Medicus, EMBASE, and Cochrane Library databases were systematically searched for articles in English appearing from any time to May 30, 2014. Twenty-six studies were included. These suggest that treatments for BD are less likely to be associated with Prl elevations, with valproate, quetiapine, lurasidone, mirtazapine, and bupropion reported not to change PRL levels significantly and lithium and aripiprazole to lower them in some studies. Taking into account the effects of the different classes of drugs on Prl may improve the care of BD patients requiring long-term pharmacotherapy. Based on the results of this review, lithium and valproate appear to be safer due to their low potential to elevate sPrL; among antipsychotics, quetiapine, lurasidone and aripiprazole appear to be similarly safe.
KEY WORDS Hyperprolactinemia, bipolar disorder, antipsychotics, mood stabilizers, antidepressants
INTRODUCTION There is growing interest on the burden of drug-related side effects in people with severe mental illnesses (Whiteford et al., 2013; Gladigau et al., 2014). The long-term, potentially lifelong need for treatment arose concerns about physical consequences, like weight gain and metabolic/endocrine disturbances (McElroy and Keck, 2014). Among these, treatment-related changes in serum prolactin (sPrl) and possible hyperprolactinemia (Byerly et al., 2007) raise important issues, due to potentially severe effects on patients’ physical health. Hyperprolactinemia is usually defined as sustained prolactin (Prl) elevation above normal laboratory levels. These are below 530 mIU/L (25 ng/mL) in women and 424 mIU/L (20 ng/mL) in men with the more commonly used assays (1 ng/mL is equivalent to 21.2 mIU/L, WHO Standard 84/500), even if ranges among laboratories may differ (Peveler et al., 2008). In some patients, sexual dysfunction or menstrual abnormality can be useful indicators of elevated Prl, although these symptoms are nonspecific, especially in psychiatric patients on medication (Bushe et al., 2010). In addition, it is important that clinicians realize the clinical significance of Prl elevations despite absence of symptoms (Bushe et al., 2010). Nonetheless, even an eight-year long treatment with Prl-elevating drugs may lead to significant bone mineral density loss in young women, and osteoporosis secondary to persistent elevation of sPrl levels doubles the risk of fractures (Byerly et al., 2007; Sørensen et al., 2013). Other worth-mentioning long-term effects are gynecomastia, hypogonadism, prolactinoma and Prl-sensitive tumors (Byerly et al., 2007; Rettenbacher et al., 2010). Several drugs have been associated to sPrl elevations and hyperprolactinemia. Among psychiatric drugs, antipsychotics (APs) have been extensively studied (de Hert et al., 2014; Peuskens et al., 2014). Studies mainly focused on samples with schizophrenia, despite APs are used widely as first- and second-line treatment of patients with bipolar disorder (BD) (Grunze et al., 2013). However, drug treatment strategies in BD include also other options, like “classical” mood stabilizers (e.g., lithium, valproate) and antidepressants (ADs), and frequently their combination. The extent to which BD treatments affect sPrl levels has not been systematically reviewed. The objective of this systematic review is to critically consider and summarize the evidence of how drugs approved for use in BD affect blood PrL levels.
EXPERIMENTAL PROCEDURES This review has been conducted according to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) statement (Moher et al., 2009). LITERATURE SEARCH We systematically searched the MEDLINE/PubMed/Index Medicus, EMBASE, and Cochrane Library databases from any time to May 30, 2014, cross-checking the obtained references. We further searched the www.clinicaltrials.gov site to check literature results and to obtain information about starting, ongoing, or concluded studies. The systematic search was performed by two blind independent researchers (IP and AM), who searched for randomized controlled trials (RTCs), as well as open studies and systematic reviews on the topic, as follows:
• MEDLINE/Pubmed/Index Medicus: authors used the keywords (prolactin*[ti] OR hyperprolactin*[ti] OR galactorr*[ti] OR gynecomast*[ti]) AND (bipolar[ti] OR mania[ti] OR manic[ti]) to identify studies of potential interest. In order to catch a wider selection of search returns, a broader string was also used by eliminating the title specifier: (prolactin* OR hyperprolactin* OR galactorr* OR gynecomast*) AND (bipolar OR mania OR manic). • Cochrane Library: keywords were (hyperprolactin* OR galactorr* OR gynecomast*) AND (bipolar OR mania OR manic). • Clinicaltrials.gov: keywords were hyperprolactinemia bipolar disorder. • EMBASE: hyperprolactinemia, bipolar disorder. STUDY SELECTION The first step was to screen whether the study investigated the effects of drugs (i.e. APs, ADs or “classical” mood stabilizers) approved for treating BD or tested in randomized clinical trials (RCTs) of BD on Prl blood levels in BD samples, based on abstract or full text. For this reason, studies with no therapeutic objectives were excluded. Data were grouped as: 1. Of interest, 2. Duplicates, and 3. Of no interest, according to abovedescribed criteria. After selection, full papers were retrieved and studied. DATA COLLECTION, PROCESSING, AND ITEMS To be included in this review, a paper had to clearly report effects on blood (serum or plasma) Prl levels and details of design, sample description, and inclusion/exclusion criteria.
RESULTS SYSTEMATIC SEARCH RESULTS Figure 1 is a flow-chart of considered and finally selected studies, endorsing the aforementioned PRISMA statement. The global search first returned 410 titles. We excluded a small proportion of duplicates or irrelevant records (37), so that 373 underwent title and abstract examination. Further 265 opinions, case reports, and reviews were excluded. One-hundred-and-eight studies were deemed eligible for further assessment. Among them, 1 is still ongoing, 1 inaccessible, and 80 unrelated to the aim of this review. Further 4 studies (Yatham, 1996, Sher et al., 2003; Cordes et al., 2011; Lykouras et al., 2011) tested a serotonergic hypothesis with AD-challenge and were subsequently excluded. The final selection included 28 studies, 8 returned with strict Pubmed/Medline/Index Medicus criteria and 20 with broad criteria. Figure 1 about here CONTENT RESULTS
The 28 included studies are summarized in Table 1. Among them, 23 studies presented at least one treatment arm with APs, 6 with lithium, 2 valproate and only 1 with the selective serotonin reuptake inhibitor (SSRI) AD, citalopram. Table 1 about here ANTIPSYCHOTICS AP effects on sPrl have been widely studied in schizophrenic samples (Stip et al., 2011; Gentile et al., 2013; Leucht et al., 2013; De Hert et al., 2014; Peuskens et al., 2014). The only study included here that assessed the general and drug-specific incidence of sPrl changes in AP-treated BD populations is a naturalistic cohort with 194 patients with schizophrenia and 49 with BD who underwent routine Prl screening in absence of important symptoms (Bush and Shaw, 2007). Above-threshold Prl levels were found in 38% of the cohort, more commonly in female (52%) than in male patients (26%). In this study, most women on risperidone monotherapy, either long-acting injectable (RLAI) or oral (RSP), developed severe hyperprolactinemia (>1000mIU/L). A trend for greater prevalence of hyperprolactinemia was found with oral RSP compared to RLAI (p = 0.057). Patients on amisulpride (AMI) also showed abnormal levels, independently from gender. In contrast, hyperprolactinemia was seen only in one olanzapine (OLZ)treated woman. Prevalence of hyperprolactinemia in monotherapy patients was presumably associated to single APs. Hyperprolactinemia was found in 100% AMI(7/7), 69% RSP- (24/35), 33% typical AP- (TAP, 27/83), 6% OLZ- (1/16), 5% clozapine- (CLZ, 1/21), and 0% aripiprazole- (ARP) or quetiapine (QTP)-treated patients (0/3). Last, most polypharmacy involved the concomitant use of typical and atypical APs. TAPs combined with, or added to, a single atypical AP did not increase hyperprolactinemia rates significantly in patients on RSP (69 vs. 67%), or OLZ (6 vs. 32%). Substituted benzoamides In an open-label 8-week study (Chen et al., 2010), adjunctive ARP to AMI and sulpiride (SUL) was ineffective in restoring baseline hyperprolactinemia (p=0.099). In an openlabel study, AMI induced galactorrhea in one of 20 patients with mania (Vieta et al., 2005). Aripiprazole Chen et al. (2010) compared in a mixed sample of patients with schizophrenia, schizoaffective disorder, and BD the effects of 8-week adjunctive ARP treatment on RSP-, AMI-, and SUL-induced hyperprolactinemia. ARP dose was 5-20 mg/day during the study period. sPrl levels were measured at weeks 4 and 8; they decreased significantly from baseline in the ARP+RSP-treated (14 of 15 patients, 93.33%, p < 0.001), but not in the ARP-AMI/ARP-SUL-treated patients (1 of 10 patients, 10%). Brown’s et al. (2013) meta-analysis of 10 studies aimed at assessing the role of ARP in acute mania. Compared to placebo (n=2239), mean Prl levels in the ARP groups dropped, in some cases below the lower normal threshold. However, data from studies measuring prolactinemia were insufficient for conducting meta-analysis.
ARP was compared to Li by Keck et al. (2009). Patients on ARP showed significantly greater mean sPrl drops than lithium patients during 12 weeks (−6.6 ng/mL vs. −2.2 ng/mL; p < 0.001). In studies comparing Hal to ARP (Vieta et al., 2005; Young et al., 2009), fewer participants taking ARP showed Prl increases above normal levels compared to Hal (Vieta 2005; Young 2009). In the first study (Vieta et al., 2005), mean sPrl levels decreased from baseline in the ARP (713.4 ng/ml, 7284.1mU/l), and increased in the Hal group (7.7 ng/ml, 7163.2mU/l) at week 12; means differed significantly (p < 0.001). Similarly, in the Young et al. (2009) study, ARP-treated patients showed a decrease in mean sPrl levels from baseline over 12 weeks, whereas Hal-treated patients showed an increase (713.4 ng/ml vs. + 6.7 ng/ml; p < 0.001). At endpoint, fewer ARP- (12.8%) than Hal-treated patients (60.8%) had new-onset hyperprolactinemia. A post hoc analysis of 28 rapid-cycling BD-I patients from a 100-week, double-blind, placebo-controlled study assessing long-term efficacy, safety and tolerability of ARP in patients with recently manic or mixed BD type I showed no significant changes from baseline in Prl levels with either ARP or placebo at the 100th week (p > 0.05) (Muzina et al., 2008). Asenapine During a 3-week RCT with asenapine (ASE) for acute mania, Prl levels did not change significantly from baseline, compared to placebo and OLZ (McIntyre et al., 2009). In the 40-week extension (McIntyre et al., 2010), shifts to high Prl levels in patients with normal or low baseline Prl levels at any time during treatment occurred in 33.3% of placebo-ASE treated patients (3 weeks of placebo and then shifted to ASE with blind unbroken), 34.3% ASE-treated patients, and 61.9% of OLZ-treated patients. Lurasidone Low potential for hyperprolactinemia was reported for lurasidone (LUR) in a 6-week monotherapy, fixed-dose, RCT (Loebel et al., 2014a). Median endpoint change in Prl was not clinically meaningful, but women on LUR had higher Prl levels compared to placebo. In the 6-week RCT of LUR as add-on to Li or DVP in bipolar depression, reported Prl changes from baseline were not significant (Loebel et al 2014b). Olanzapine OLZ was tested in a 3-week, head-to-head RCT vs. RSP (Perlis et al., 2006), showing significantly lower Prl elevation than RSP (p>0.001). In a 3-week, placebo-controlled, double-blind RCT with DVP as active comparator, OLZ-treated patients showed significant Prl elevations compared to placebo-treated (p < 0.001) and DVP-treated patients (p < 0.001) (Tohen et al., 2008). OLZ was used as an active comparator to ASE both in acute (3-weeks) and long-term (40-weeks) clinical trials (McIntyre et al., 2009; 2010). At week 40, Prl increases from baseline in the OLZ-treated arm were double compared to the 2 ASE-treated arms. In a 3-week RCT of adolescents with acute bipolar mania (Tohen et al., 2007), mean baseline-to-endpoint Prl increases were significantly greater in the OLZ group relative to the placebo group (p < 0.001). Abnormally high Prl levels occurred significantly more often in patients receiving OLZ compared to placebo (girls, 25.7% vs. 0%, p = 0.007; boys, 62.5% vs. 5%, p < 0.001).
In a two-phase RCT of PAL in patients with manic or mixed symptoms, the active comparator OLZ increased Prl levels in men and decreased them in women from baseline to week 15 (Bewaerts et al., 2012b). Paliperidone Two randomized, PBO-controlled studies with PAL, using QTP (Vieta et al., 2010) or OLZ (Bewaerts et al., 2012b) as active comparators in manic and mixed patients, found that PAL acutely induced greater increases in sPrl levels (Vieta et al., 2010) which persisted during chronic administration. In another acute randomized, PBO-controlled trial in patients with manic or mixed symptoms, PAL was associated with higher sPrl levels in women across all three PAL doses, which were proportional to daily dose (Bewaerts et al., 2012a). Finally, increased sPrl levels from baseline to the 6-week endpoint were found in another randomized, PBO-controlled study with adjunctive PAL (Berwaerts et al., 2011). Quetiapine In QTP monotherapy studies (Bowden et al., 2005; McIntyre et al., 2005), mean Prl decreased in QTP- and placebo-treated groups from baseline to the 12-week end-point. More QTP-treated (61 of 209) than placebo-treated (46 of 198) patients shifted from elevated to normal Prl levels, but between-groups differences were not significant (Vieta et al., 2005). Effects on serum prolactin were not monitored in studies examining QTP combined with Li/DVP (Adler et al., 2007). In a PBO-controlled RCT of PAL in patients with manic or mixed symptoms (Vieta et al., 2010), QTP used as the active comparator induced less sPrl increases compared to PAL during both acute and maintenance phases. Risperidone In a 3-week, randomized, controlled, double-blind, parallel multicenter study (Perlis et al., 2006) RSP (1–6 mg/day; N = 164) was compared to OLZ (5–20 mg/day; N = 165) in inpatients with acute manic/mixed symptoms. RSP-treated patients were more likely to show Prl elevations (51.73 ng/mL vs. 8.23 ng/mL; p < 0.001). RSP-induced hyperprolactinemia was successfully reversed by adding ARP in another study (Chen et al., 2010). A 4-week study of RSP in AP-naïve patients with schizophrenia, psychotic depression or BD, showed significant Prl increases from 12.1 ± 1.9 ng/ml at baseline to 65.7 ± 12.2 ng/ml post-treatment (p
