Introduction
Major depressive disorder (MDD) is one of the most disabling and common psychiatric disorders worldwide.Reference Ferrari, Charlson and Norman 1 Recent data estimate a lifetime prevalence of MDD at 16.6% and the 1-year prevalence at 6.7%. MDD is also a leading cause of premature death and ongoing disability.Reference Kessler, Berglund and Demler 2
The relationship between MDD and inflammation has been investigated in several studies.Reference Miller, Maletic and Raison 3 , Reference De Berardis, Conti and Campanella 4 Chronic inflammation is thought to play a role in the etiology of MDD, at least for some individuals.Reference Dowlati, Herrmann and Swardfager 5 In particular, C-reactive protein (CRP) levels have been extensively examined in MDD and other mood disorders.Reference Raison and Miller 6 , Reference Leonard 7 Overall, the results indicate a possible association between CRP, severity of depressive symptoms, and suicide risk.Reference Gambi, De Berardis and Campanella 8
Psychopharmacological treatment of MDD includes a number of antidepressant drugs.Reference Lam 9 However, treatment with some antidepressants, such as selective serotonin reuptake inhibitors (SSRIs), may lead to adverse metabolic changes and therefore to increased inflammatory markers, such as CRP.Reference Klemettilä, Kampman and Seppälä 10 Also, evidence shows that the level of CRP may be correlated with treatment-induced weight gain.Reference Devaraj, Valleggi, Siegel and Jialal 11
Agomelatine was first reported in the literature in 1992, and is among a series of synthetic naphthalene melatonin analogues.Reference Biggio 12 Agomelatine shows agonistic activity with high affinity for melatonin MT1 and MT2 receptors and an antagonist activity with moderate affinity for serotonin 5HT2C.Reference De Bodinat, Guardiola-Lemaitre, Mocaër, Renard, Muñoz and Millan 13 No significant affinity for any of the monoamine transporters or for adrenergic, noradrenergic (NA), dopaminergic (DA), muscarinic, histaminic, or benzodiazepine receptors has been reported.Reference Hickie and Rogers 14 The agonist activity at the MT1 and MT2 receptors at the suprachiasmatic nucleus (SCN) may modulate the functions of these receptors and normalize circadian rhythms, including sleep-wake cycle, that are often altered in MDD.Reference Fornaro, Bandini and Cestari 15 It has been demonstrated that agomelatine may increase dopamine and noradrenaline levels in the frontal cortex, also stimulating cell proliferation and neurogenesis.Reference De Berardis, Di Iorio and Acciavatti 16
The efficacy of agomelatine in the treatment of MDD has been demonstrated in several studies.Reference Kennedy, Avedisova, Belaïdi, Picarel-Blanchot and de Bodinat 17 , Reference Komaram, Nukala, Palla, Nambaru and Kasturi 18 Its adverse effect profile seems to be relatively favorable, as it is not associated with weight gain or metabolic abnormalities that are otherwise commonly seen with traditional antidepressants.Reference Fornaro, Prestia, Colicchio and Perugi 19 It may also be hypothesized that the better tolerability profile of agomelatine compared to standard antidepressants might be related to a lower impact on CRP levels.Reference Plesničar 20 It could therefore by hypothesized that eventual changes recorded in CRP levels before and after agomelatine treatment would better reflect patients’ clinical status rather than adverse effects. However, to date, no studies have been carried out investigating the effects of agomelatine on CRP levels in MDD patients before and after treatment.
The present study, to our knowledge the first of its kind, aims (i) to investigate the effects of agomelatine treatment on CRP levels in a sample of patients with MDD and (ii) to investigate if CRP variations were correlated with clinical improvement in such patients.
Methods
Thirty adult outpatients (12 males, 18 females) with a Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IV-TR) diagnosis of MDD were recruited from several mental health facilities in central Italy during everyday clinical practice. Diagnoses were all made by psychiatrists with at least 5 years clinical experience and confirmed with the Structured Clinical Interview for DSM-IV (SCID).Reference First, Spitzer, Gibbon and Williams 21 The decision to initiate agomelatine for each enrolled patient was only based on clinician judgment and patients’ agreement, as commonly happens in “real-world,” everyday clinical practice.
The exclusion criteria for the inclusion in the present study sample were (i) any other axis I diagnosis; (ii) pregnant, nursing, or lactating women or those at childbearing potential refusing to take any adequate contraception method; (iii) patients who met DSM-IV-TR criteria for abuse or dependence on any drug including alcohol within prior 8 months; (iv) patients with medical contraindications to therapy with agomelatine based on medical history and laboratory data (namely severe liver disease); (v) patients with a known allergy or hypersensitivity to agomelatine. Other exclusion criteria were severe medical diseases; current treatment with anti-inflammatory or immunosuppressant drugs (such as non-steroidal anti-inflammatory drugs, steroids, paracetamol), ciprofloxacine, or other drugs with any documented major potential pharmacokinetic interaction with agomelatine; or known familial hypercholesterolemia or obesity. Any current psychopharmacological treatment was an exclusion criterion. Concerning the latter exclusion criterion, cases of MDD with psychotic features also were excluded, as we evaluated 3 cases treated with agomelatine plus an antipsychotic drug.
Primary outcome measure was the 17-item Hamilton Rating Scale for Depression (HAM-D) total score.Reference Hamilton 22 Assessments were carried out at the baseline visit and after 12 weeks of active treatment. The Snaith-Hamilton Pleasure Scale (SHAPS)Reference Snaith, Hamilton, Morley, Humayan, Hargreaves and Trigwell 23 was employed to assess anhedonia.
Agomelatine was administered in the following fashion: the starting dose was 25 mg/day at bedtime. In the case of no clinical response and/or clinician’s judgment and/or patients’ agreement, the dosage was eventually increased to 50 mg/day (maximum licensed dosage) at any time during follow-up, always administered in a single dose at the bedtime. Liver function tests in all patients receiving agomelatine were evaluated as specified in accordance with the recommendations in the product information. All treatment-related emerging adverse effects were evaluated.
In the present study, patients with a reduction of 50% or more on the HAM-D total score at endpoint were considered responders to treatment; remission, which represents complete or near complete symptom resolution, including resolution of functional impairment, was defined as a HAM-D total score of ≤7.Reference Ballenger 24
Serum CRP was measured at baseline and after 12 weeks of treatment with agomelatine using a highly sensitive nephelometric assay (BN-II Nephelometer; Dade Behring, Deerfield, IL, USA), which is able to detect a minimal CRP concentration of 0.22 mg/dl. Blood samples were taken between 7:00 a.m. and 8:30 a.m., after the patient had fasted for at least 10 hours.
Weight was measured (in light, indoor clothing with shoes off) using a balance beam scale, and height was measured using a stadiometer. Weight and height were then used to calculate body mass index (BMI) in kg/m2.
The rating scales and CRP measurement were conducted as a part of everyday, real-world clinical practice evaluation and assessment of patients, and, therefore, no institutional review board approval was needed. However, each patient had to understand the nature of the study and signed an informed consent document. The study was conducted in accordance with the principles of good clinical practice and the Declaration of Helsinki (1964) and subsequent revisions.
Statistical Analysis
Descriptive statistics (means and standard deviations as appropriate) and percentages were computed for the study sample on demographic variables and all psychometric scales. Chi-squared analyses (with the Yates’ correction for 2×2 tables) were used for categorical variables. All analyses were conducted using nonparametric testing. The Friedman test and Wilcoxon test for paired samples were used to monitor changes in scores on psychometric scales and CRP levels. Correlations between scores on psychometric scales and CRP levels were measured by means of Spearman’s rank correlation coefficient. Cohen’s d values were also calculated to obtain effect sizes for a better evaluation of the magnitude of group differences on the continuous study measures. Cohen describes d values of 0.2 as small, 0.5 as medium, and 0.8 as large effect. Finally, a multivariate, stepwise, linear regression analysis was performed to determine which variables were associated with the with CRP level variation at endpoint (dependent variable). In the first step, age, gender, smoking, and BMI were entered. In the second step, baseline HAM-D scores were added to the model. The agomelatine dosage at endpoint was added in the last step. Statistical analysis was performed using SPSS for Windows, Version 20.0 (SPSS Inc., Chicago, IL, USA).
Results
Patients’ data are reported in Table 1. At baseline, MDD patients with melancholic features (n=10, 33.3%) as well as with atypical features (n=2, 6.7%) had no significant differences concerning CRP levels than MDD patients without melancholic or atypical features (respectively, 2.5±0.6 mg/l and 2.6±0.6 vs 2.4±0.6 mg/l, p=NS).
Table 1 Demographic and clinical data of patients before and after agomelatine treatment
Abbreviations: CRP, C-reactive protein; HAM-D, Hamilton Rating Scale for Depression; SHAPS, Snaith-Hamilton Pleasure Scale; BMI, body mass index.
Efficacy analyses
The mean HAM-D total score at baseline was 28.2±2.8 and reduced to 9.7±4.9 at week 12 (p<0.001). A significant reduction was seen also for SHAPS (6.6±2.2 vs 3.1±2.0, p<0.001). The mean dosage of agomelatine at week 12 was 39.2±12.6 mg/day (17 patients, 56.7%, were taking 50 mg/day, whereas 13 patients, 43.3%, were taking 25 mg/day).
A reduction of more than 50% in HAM-D scores (responders) was attained in 11 patients (36.7%; 5 males, 6 females). Twelve patients (40%; 6 males and 6 females) achieved remission (endpoint score ≤7 on HAM-D). Seven patients (23.3%; 1 male, 6 females) had a reduction of less than 50% in HAM-D scores and were considered nonresponders. Among responders, the mean agomelatine dosage was 40.9±12.6 mg/day (7 of 11 patients were taking 50 mg/day). Among remitters, the mean dosage was 35.4±12.9 mg/day (5 of 12 patients were taking 50 mg/day). Among nonresponders the mean dosage was 42.9±12.2 mg/day (5 of 6 patients were taking 50 mg/day). No differences between groups in agomelatine dosage were found. In particular, no differences were observed in terms of remission/response rate between patients who were taking 25 mg/day or 50 mg/day (χ2=0.40, df=1, p=0.52).
No patients had worsening of mood symptoms during the period of evaluation, nor were any hospitalized or had emergent suicidal ideation.
In general, side effects were in the mild range and agomelatine was well tolerated in the majority of the patients. The most commonly reported adverse events were dizziness in 3 patients (24%) and nausea with decreased appetite in 2 patients (12%), both transient and mild. No abnormal elevations of liver enzymes were detected in any patient.
CRP levels analyses
Concerning CRP levels, no gender differences were found, nor were any found for smokers vs nonsmokers. Mean serum CRP levels detected in the whole sample were 2.5±0.6 mg/l, which were reduced to 1.8±0.5 at week 12 (p<0.001). The median CRP level at baseline was 2.3 mg/l and was 1.8 mg/l at endpoint. At endpoint, no differences between responders/remitters and nonresponders were found in patients above and below median CRP level. At endpoint, MDD patients with melancholic features as well as with atypical features had no significant differences concerning CRP levels than MDD patients without melancholic or atypical features (respectively 1.6±0.4 mg/l and 1.6±0.1 vs 2.0±0.5 mg/l, p=NS). Correlation between CRP and HAM-D variation after 12 weeks of agomelatine treatment showed a significant correlation (Spearman’s ρ=0.65, p<0.001), but not between CRP and SHAPS variation (Spearman’s ρ=0.32, p=0.08).
It is worth of noting (Figure 1) that remitters showed a significant difference in CRP levels after 12 weeks of agomelatine treatment (p<0.001), while no difference was observed in nonresponders (p=0.94). Responders showed a trend toward a reduction of CRP levels, though not always statistically significant (p=0.08). Also, results from a multivariate stepwise linear regression analysis showed that more CRP level variation was associated with higher baseline HAM-D scores, when controlling for age, gender, smoking, BMI, and agomelatine dosage at endpoint (partial r=0.47, t=2.2, p=0.03).
Figure 1 CRP levels (mg/l) at baseline and week 12 between whole sample, responders, remitters, and nonresponders (*p<0.001).
Discussion
To our knowledge, this was the first study that investigated the effects of agomelatine on CRP levels in MDD patients before and after 12 weeks of treatment.
In this study, agomelatine was effective as an antidepressant during a 12-week treatment period, as response was observed in 11 patients (36.7%) and remission in 12 (40%), without dropouts due to adverse effects. With regard to anhedonia, agomelatine showed efficacy as manifested by the significant reduction in SHAPS scores. These observations confirm the clinical efficacy and the relatively good tolerability of agomelatine, even in a particular and complex context as that of everyday, “real-life” clinical practice. Moreover, the efficacy of agomelatine on the anhedonia dimension of MDD holds particular importance in the treatment of patients with anhedonic features, as commonly seen in such a context.
Concerning CRP levels, the results of our study provide preliminary evidence that agomelatine’s antidepressant properties were associated with a reduction in circulating CRP levels, as is the case for other antidepressant drugs such as SSRIs or tricyclics. This is in line with other studies. Lanquillon et alReference Lanquillon, Krieg, Bening-Abu-Shach and Vedder 25 assessed 24 inpatients with MDD, both before and during 6 weeks of antidepressant (amitriptyline) treatment, and found that the levels of CRP were significantly increased at admission and decreased in both responder and nonresponder patients during antidepressant treatment. Tuglu et al.Reference Tuglu, Kara, Caliyurt, Vardar and Abay 26 observed a decrease in CRP levels in 26 MD patients after 6 weeks of antidepressant treatment with SSRIs. Recently, Uher et alReference Uher, Tansey and Dew 27 measured CRP in MDD patients who were randomized to 12 weeks of treatment with escitalopram (a SSRI) or nortriptyline (a norepinephrine reuptake inhibitor, NRI). They observed that serum CRP levels at baseline differentially predicted treatment outcomes with both antidepressants. Patients with low levels of CRP (<1 mg/L) showed improvement on the Montgomery–Åsperg Depression Rating Scale (MADRS), with scores 3 points higher after escitalopram treatment, compared with nortriptyline. On the other hand, patients with higher CRP levels scored 3 points higher on MADRS after nortriptyline, compared with escitalopram.
It is worth noticing that in our study, a significant reduction in CRP levels was observed in patients who achieved full symptom remission (endpoint score≤7 on HAM-D), but not in those who only achieved response (reduction of more than 50% in HAM-D scores at endpoint with HAM-D score ≥ 7). This seems to indicate that only full symptom remission and not the simple treatment response may influence CRP levels. However, as responders showed a trend toward a reduction of CRP levels, even though not statistically significant, one can argue that the lowering of CRP levels may be a slow but constant process that remodulates CRP levels only after the achievement of remission.Reference Eyre, Stuart and Baune 28 In fact, it is arguable that the relatively short period of our study (12 weeks) may not have been long enough for us to observe the statistically significant reduction observed in remitters, as some patients may take more time than others to reach MDD remission.Reference Riedel, Möller and Obermeier 29 On the other hand, it is also possible to hypothesize that elevated levels of CRP in phases from acute depression to response may reflect a persistent activation of the inflammatory response system in MDD, regardless of symptomatic severity, at least in some patients.Reference Tsai, Chung, Huang, Chen, Lee and Kuo 30
Moreover, in our study, the multivariate stepwise linear regression analysis showed that a more prominent CRP level variation was associated with higher baseline HAM-D scores, while controlling for age, gender, smoking, BMI, and SHAPS variation. This finding is in line with previous data that demonstrated that the improvement in depressive symptoms following treatment with antidepressants was functionally correlated with the reduction in CRP levels.Reference O’Brien, Scott and Dinan 31
The effect of agomelatine on CRP levels may be explained in several ways. An intriguing explanation involves agomelatine’s specific action on the circadian rhythms through functional interaction with melatonergic MT1/MT2 receptors.Reference De Berardis, Marini and Fornaro 32 In fact, it has been suggested that chronic inflammation related to chronic diseases, such as rheumatoid arthritis, allergies, aging-related diseases, and, perhaps, MDD, may be modulated by some proteins encoded by clock genes that may affect the development of chronic inflammatory diseases or increase the severity of their symptoms.Reference Kizaki, Sato and Shirato 33 Therefore, agomelatine’s capability in restoring circadian rhythms may relieve the chronic subclinical inflammation in MDD patients through the action on such clock genes and, probably, on circadian misalignment often seen in MDD. In fact, in a recent study, higher CRP levels after chronic circadian misalignment decreased in the resynchronized control group across weeks of circadian entrainment.Reference Wright, Drake and Frey 34 On the other hand, agomelatine’s antagonist activity on 5HT2C in the frontal cortex enhances the DA and NA neurotransmission that is responsible for its antidepressant effect, together with the agonistic activity on MT1 and MT2 receptors.Reference MacIsaac, Carvalho, Cha, Mansur and McIntyre 35 The enhanced DA and NA neurotransmission may induce a T helper 1 (Th1) shift and a T helper 2 (Th2) shift,Reference Martino, Rocchi, Escelsior and Fornaro 36 thus inhibiting the production of Th2 cytokines such as interleukin 6 (IL-6) and the production of Th1 pro-inflammatory cytokines, including tumor necrosis factor-α (TNF-α).Reference Hashimoto 37 This re-modulation may also account for reduction in CRP levels seen in our patients treated with agomelatine. However, the effects of agomelatine on IL-6 and TNF-α currently are unknown, and further studies are needed.
This study was exploratory, and therefore several limitations must be considered. The 4 major limitations of our study were (1) the relatively small sample, (2) the open label design, (3) the lack of placebo or healthy control groups, and (4) the relatively short treatment period (12 weeks). Also, the chance of “selection by indication” bias (lack of random assessment) might be perceived as a limitation of the present study, though we adhered to a practical approach in order to better resemble real-world routine prescription trends.
Conclusions
In conclusion, agomelatine’s antidepressant properties were associated with a reduction in circulating CRP levels in MDD patients who achieved remission after 12 weeks of treatment. Moreover, more prominent CRP level variation was associated with more severe depressive symptoms at baseline. Moreover, this was the first study that assessed CRP levels during agomelatine treatment in “real-world,” everyday clinical practice. Further studies, on wider samples and for longer periods are needed to validate the role of CRP levels as markers of treatment response or remission in MDD patients.
Disclosures
The authors have nothing to disclose.
