Monoamine oxidase inhibitors (MAOIs) for fibromyalgia syndrome

Description of the condition

Fibromyalgia (FM) syndrome is a chronic condition of unknown aetiology (Cathebras 1998), affecting 3.7 million people in the United States (Lawrence 1998), with an average cost of USD 2274 per patient/year (Wolfe 1997). The disease is characterised by widespread musculoskeletal pain which commonly co‐exists with cognitive dysfunction, sleep disturbance and significant fatigue (Wolfe 2010). Correspondingly, patients often report high disability levels and poor quality of life (Hawley 1988;Hawley 1991), along with extensive use of medical care (Wolfe 1997). Lacking a specific laboratory test, methods for diagnosis include both the 1990 and 2010 American College of Rheumatology (ACR) criteria (Wolfe 1990; Wolfe 2010). The more commonly used 1990 ACR criteria have been shown to be 88% accurate in identifying patients with the syndrome (Smith 1998). In the past other standardised and recognised criteria had been used to diagnose FM (Smythe 1981; Yunus 1981; Yunus 1982; Yunus 1984).

Much effort has been made to elucidate the pathophysiology of FM. Alterations in alpha‐non REM sleep (Moldofsky 1989), structural (Bengtsson 1986) and functional (Bartels 1986; Bengtsson 1986; Lund 1986) alterations in muscle fibres, disturbances of hypothalamic‐pituitary‐adrenal axis (Crofford 1994), abnormal metabolism of substances like serotonin (Moldofsky 1989), norepinephrine and substance P (Vaeroy 1988), and alterations in regional cerebral blood flow (Bradley 1996; Gracely 2002) have been observed and postulated as aetiologic mechanisms. Despite these findings, the aetiology of this syndrome remains unknown. Since specific treatment aimed at altering the pathogenesis is not possible, drug therapy focused on pain reduction is ubiquitously employed.

Description of the intervention

The current concept of FM suggests that changes in the functioning of neurons that lead to 'sensitisation' of the brain and spinal cord are physiologically responsible for FM symptoms. Theoretically, medications that attenuate aberrant function of the central nervous system can be of benefit in the treatment of FM symptoms. A popular class of medications with such central effects is antidepressants and they are the most frequently prescribed medications for FM (Miller 2002).

Several studies on antidepressants have shown effectiveness compared to placebo for the symptoms associated with fibromyalgia (Rossy 1999; Arnold 2000; O'Malley 2000; Goldenberg 2007; Häuser 2009), although those have mainly centred in amitriptyline and there is a need to assess the effectiveness and safety of other antidepressants such as monoamine oxidase inhibitors (MAOIs).

Moclobemide is the first of a new class of reversible inhibitors of monoamine oxidase A, with high selectivity for the type A isoenzyme (MAO‐A) which primarily deaminates serotonin, noradrenaline and dopamine relative to other monoamine substrates (Steinmeyer 1993; Da Prada 1994). It has been shown to modulate central nervous system neurotransmitter disposition and to have therapeutic applications (Holford 1994). It is an effective antidepressant with a mild adverse effect profile (Stabl 1989;Versiani 1989;Versiani 1990; Priest 1994;Newburn 1999;Papakostas 2006).

Pirlindole is a tetracyclic compound that has been characterised as a potential antidepressant drug. It has pharmacological characteristics in common with both tricyclic antidepressants and classical irreversible MAOIs. Its main mechanism of action consists of a selective and reversible inhibition of monoamine oxidase A (De Wilde 1996; Bruhwyler 1997;Tanghe 1997).

Other antidepressants agents (selective serotonin re‐uptake inhibitors (SSRIs), tricyclic agents, serotonin–norepinephrine reuptake inhibitors (SNRIs)) have been the objective of other Cochrane reviews (Häuser 2012; Nishishinya 2012; Walitt 2012), as the original protocol for the present review has been split into different reviews (Nishishinya 2006).

Criteria for considering studies for this review

Search methods for identification of studies

We ran an electronic search in the Cochrane Central Register of Controlled Trials (CENTRAL) (The Cochrane Library 2010, Issue 10), MEDLINE accessed through PubMed (1966 to November 2010) and EMBASE accessed through OVID (1980 to November 2010). See Appendix 1 for search strategies in all databases and retrieved results. We searched bibliographies from reviewed articles and we retrieved relevant articles. We contacted content experts for unpublished and further possible studies.

Our search included all languages. We contacted the corresponding authors of identified RCTs when possible for additional information about other relevant studies. We also searched for ongoing trials in relevant databases such as clinicaltrials.gov and controlled‐trials.com.

Data collection and analysis

Description of studies

We initially identified 2728 studies related to FM in the 2009 search and 786 in the 2010 search. As the search strategy was designed as part of a global search strategy to identify all the RCTs on pharmacological and non‐pharmacological treatments for FM (Nishishinya 2006), many of the obtained references were not related to MAOIs. We excluded 3509 references as they did not fulfil the inclusion criteria related to the interventions evaluated in this review. We identified five studies potentially related to these interventions and a full text could only be obtained for four of them. Of these five, we excluded three studies (see Characteristics of excluded studies for further details about reasons for exclusion and Figure 1 for study flow diagram). We ultimately included two studies (see Characteristics of included studies for full description of studies).

Risk of bias in included studies

Risk of bias was high in Ginsberg 1998 and low in Hannonen 1998. In the first study, allocation concealment, sequence generation and blinding were not reported. Additionally, the attrition rate was high (39%) and the intention‐to‐treat (ITT) analysis was only performed with safety data. Efficacy data were evaluated per protocol.

On the other hand, Hannonen 1998 properly described allocation concealment, sequence generation and blinding of patients and outcome assessors, and conducted an ITT analysis, although the attrition rate was also high (30%).

It was not possible to assess selective outcome reporting as we did not have access to the study protocols.

Sample sizes were small in both studies (fewer than 50 patients per group).

See Figure 2 and Figure 3 for a 'Risk of bias' summary and graph and Characteristics of included studies for detailed information regarding 'Risk of bias' assessments for every study.

Effects of interventions

See: Table 1

Summary of main results

The objective of this systematic review was to assess the effectiveness and safety of monoamine oxidase inhibitors (MAOIs) in the treatment of fibromyalgia (FM) compared with placebo or another active drug. We identified two studies with a total of 230 patients (mainly women) diagnosed with FM that evaluated two different MAOIs: pirlindole (Ginsberg 1998) and moclobemide (Hannonen 1998) in the short term (four and 12 weeks respectively) and showed inconsistent results. Pirlindole compared to placebo statistically significantly improved pain, tender points and global assessment by the patient and by the physician, whereas moclobemide did not show any statistically significant differences in comparison with placebo for the same outcomes. When moclobemide was compared to amitriptyline, the latter showed more favourable results in the percentage of responders assessed by physician and sleep quality.

Pooled results of these two studies showed a moderate effect of MAOIs on pain and a small effect in tender points. Of note, we observed moderate heterogeneity (59%) when pooling results for the outcome pain, which could be explained by the different drugs employed and different duration and design of studies. Meta‐analyses of other outcomes (global assessment by the patient and the physician) did not show any significant difference between groups.

Overall completeness and applicability of evidence

While the two studies included in this review demonstrate short‐term improvements in pain, it remains unclear if MAOIs provide any long‐term benefit. FM is a chronic condition, potentially requiring treatment over an entire lifetime. When MAOIs are prescribed for the treatment of FM, it is typically with the intent that it will be a long‐term therapy. The longest study reviewed only considered improvement at the end of 12 weeks. None of the studies reviewed provide any insight into the long‐term efficacy of MAOIs for FM symptoms. For these reasons, we are not confident that their results adequately estimate the utility of MAOIs when applied to the general population.

The most frequent adverse effects of MAOIs were headache and insomnia, although there were no statistically significant differences with placebo. Drop outs due to adverse events did not differ either compared with placebo. In the Hannonen 1998 study, a high percentage of patients with at least one adverse event were reported in all groups (77% in the moclobemide group, 74% in the amitriptyline group and 80% in the placebo group) but only a small number of patients withdrew for that reason. It is important to be mindful that there is potential for more serious side effects to be seen in clinical practice than in these trials. MAOIs are well known to cause potentially fatal hypertensive crisis, serotonin syndrome and psychosis when they interact with foods containing tyramine (fermented beverages, liver and aged cheese) and a variety of common medications. Many of the medications that lead to MAOI interactions are commonly used in fibromyalgia treatments, such as SSRIs, tricyclics, meperidine, tramadol, dextromethorphan and St John's Wort. The design of the two clinical trials considered had strict limitations on the use of concomitant medications; it may be more difficult to avoid unintended interactions with these medications in a clinical environment. These studies also do not include the potential for withdrawal symptoms on cessation of taking these medications.

The two medications that are the focus of this review are also more recent members of the MAOI class. These medications are reversible in their effects, unlike older MAOIs that demonstrate irreversible action. Thus, the safety data reported here should not be considered to reflect the MAOI class as a whole.

Potential biases in the review process

Limitations of this review include the low number of studies and issues associated with study design, such as short durations of intervention and follow‐up and low numbers of patients.The long‐term efficacy of treatments for FM is unknown. There are possible adverse events that might come up when using these drugs for a long time. Risk of bias is low in one study (Hannonen 1998) and high in the other one (Ginsberg 1998) as sequence generation, allocation concealment, blinding and incomplete outcome data were poorly reported. Both studies were published before the publication of the CONSORT statement (Moher 2010) which might partly explain the low quality in the reporting of the Ginsberg 1998 study. External validity is limited also because both studies were mainly conducted in women.

Agreements and disagreements with other studies or reviews

Our results agree with previous reviews and guidelines recommendations. The European League Against Rheumatism (EULAR) recommends the use of moclobemide and pirlindole for the treatment of pain (level of evidence Ib, recommendation grade A), although states that the evidence about these drugs is limited (Carville 2008). Häuser 2009 and Üçeyler 2008 also conclude that MAOIs showed a small effect size for reducing pain. On the other hand, other clinical practice guidelines do not mention MAOIs in their recommendations, i.e. the American Pain Society guideline (Buckhardt 2005) or the University of Texas School of Nursing guidelines (Guidroz 2009).

Characteristics of included studies [ordered by study ID]

Characteristics of excluded studies [ordered by study ID]

Internal sources

• Iberoamerican Cochrane Centre, Spain.

External sources

• Agència d'Avaluació de Tecnologia i Recerca Mèdiques (146/24/2004), Spain.