Several studies have been performed with omega-3 fatty acids in schizophrenia. However, the results are inconsistent and the effect size is small in both first manifestation schizophrenia and chronic schizophrenia (Ross et al. 2007). Of great interest are the findings of a 1-year study by Amminger et al. (2010) in a sample of people at high risk for schizophrenia who were already showing prodromal symptoms. This study found a significantly lower rate of progression to psychoses in those people who received omega-3 fatty acids compared to the placebo group. The use of omega-3 fatty acids is discussed in detail in Chap. 18 of this book.

In addition to its other effects, erythropoietin has immunomodulatory effects. In a 12-week placebo-controlled study in chronic schizophrenia patients, cognition improved significantly more with rh-erythropoietin than with placebo. However, rh-erythropoietin was not superior to placebo in overall psychopathology, measured with the PANSS scale, or social functioning (Ehrenreich et al. 2007). Another interesting effect of rh-erythropoietin is that it is able to slow the loss of CNS volume in schizophrenia (Wüstenberg et al. 2011).

Recent studies have shown that the modern tetracycline antibiotic minocycline can also be effective as an add-on treatment in schizophrenia (Chaves et al. 2009). Although minocycline has numerous effects in the brain, the most relevant one is the anti-inflammatory effect via modulation of the oxidative system. Minocycline also inhibits microglia activation. PET studies show that microglia are more strongly activated in schizophrenia patients than in healthy controls (van Berckel et al. 2008). In addition to case reports of clinical effects of minocycline in schizophrenia (Ahuja and Carroll 2007) and reports from animal experiments (Mizoguchi et al. 2008), double-blind, controlled studies have also shown positive effects of minocycline on cognitive functions and negative symptoms (Levkovitz et al. 2010; Chaudhry et al. 2012).

A double-blind, prospective, randomized study of the virostatic agent valaciclovir found no advantage with regard to schizophrenic symptoms in patients seropositive for cytomegalovirus (Dickerson et al. 2003, 2009a). Another study with azithromycine also found no superiority over place in schizophrenia patients seropositive for toxoplasmosis (Dickerson et al. 2009b).

A blunted type 1 (acute) immune response and shift to a type 2 (chronic) response have been described previously in schizophrenia: serum levels of the pro-inflammatory type 1 cytokine IFN-γ and after in vitro stimulation, the IFN-γ production were lower in (unmedicated) schizophrenia patients than in healthy controls (Rothermundt et al. 2000; Schwarz et al. 2001), although the findings are in part controversial (Miller et al. 2011). Stimulation of the blunted part of the immune response and downregulation of the (upregulated) immune response by anti-inflammatory medication might be two sides of the same coin. Therefore the type 1 response stimulant IFN-γ was hypothesized to have a therapeutic effect in schizophrenia. The effects of adjunctive IFN-γ in two treatment-resistant schizophrenia inpatients were evaluated and both patients showed an impressive therapeutic benefit from the IFN-γ therapy (Grüber et al. 2014). However, these results have to be considered cautiously because of the very limited experiences in two patients, possible side effects of IFN-γ and the fact that the data are preliminary.

Possible unwanted immune effects of IFN-γ, a strong type-1 immune activator, have to be taken into account. On the other hand, the type 1 stimulation might be a further option to re-balance the type 1/type 2 imbalance, in particular in patients showing a blunted type 1 immune response.

Anti-inflammatory treatment would be expected to show antidepressant effects also in depressed patients, because of the increase of pro-inflammatory cytokines and PGE₂ in depression. COX-2 inhibitors in particular seem to have beneficial effects: in animal studies COX-2 inhibition can lower the increase of the pro-inflammatory cytokines IL-1α, TNF-α and of PGE₂ and it can also prevent clinical symptoms such as anxiety and cognitive decline that are associated with this increase of pro-inflammatory cytokines (Casolini et al. 2002).

Additionally, COX-2 inhibitors influence the CNS serotonergic system, either directly or via CNS immune mechanisms. In a rat model, treatment with rofecoxib was followed by an increase of serotonin in the frontal and temporo-parietal cortex (Sandrini et al. 2002). Since a lack of serotonin is one of the key features in the pathophysiology of depression, a clinical antidepressant effect of COX-2 inhibitors would therefore be expected. A possible mechanism of the antidepressant action of COX-2 inhibitors is the inhibition of IL-1 and IL-6 release. Moreover, COX-2 inhibitors also protect the CNS from the effects of quinolinic acid (Salzberg-Brenhouse et al. 2003). In the depression model of the bulbectomized rat, a decrease of cytokine levels in the hypothalamus and a change in behaviour have been observed after chronic celecoxib treatment (Myint et al. 2007). In another animal model of depression, however, the mixed COX-1/COX-2 inhibitor acetylsalicylic acid showed an additional antidepressant effect by accelerating the antidepressant effect of fluoxetine (Brunello et al. 2006).

Accordingly, a clinical antidepressant effect of rofecoxib was found in 2,228 patients with osteoarthritis, 15 % of whom had a co-morbid depressive syndrome, evaluated by a specific depression self-report. Co-morbid depression was a significant predictor for worse outcome (assessed by osteoarthritis-related pain) after rofecoxib therapy. Surprisingly, during therapy with 25 mg rofecoxib the rate of substantive depression decreased significantly from 15 to 3 % of the patients (Collantes-Esteves and Fernandez-Perrez 2003).

Moreover, we were able to demonstrate a significant therapeutic effect of the selective COX-2 inhibitor celecoxib on depressive symptoms in a randomized double-blind pilot add-on study in MD (Müller et al. 2006). In another clinical study, the mixed COX-1/COX-2 inhibitor acetylsalicylic acid accelerated the antidepressant effect of fluoxetine and increased the response rate in depressed non-responders to a monotherapy with fluoxetine in an open-label pilot study (Mendlewicz et al. 2006).

COX-2 inhibitors have also showed interesting effects in animal models of depression. Treatment with the COX-2 inhibitor celecoxib, but not with a COX-1 inhibitor, prevented the dysregulation of the HPA axis, in particular the increase of cortisol, one of the key biological features associated with depression (Casolini et al. 2002; Hu et al. 2005). This effect was expected because PGE₂, which stimulates the HPA axis in the CNS (Song and Leonard 2000), is inhibited by COX-2 inhibition. The functional effects of IL-1 in the CNS, which include sickness behaviour, were also shown to be antagonized by treatment with a selective COX-2 inhibitor (Cao et al. 1999).

Another randomized, double-blind study in 50 patients with MD also showed a significantly better outcome with the COX-2 inhibitor celecoxib plus fluoxetine than with fluoxetine alone (Akhondzadeh et al. 2009). A similar result was obtained with a celecoxib add-on approach to sertraline in MD (Abbasi et al. 2012). A meta-analysis on the use of COX-2 inhibitors in MD found an overall benefit of celecoxib add-on therapy (Na et al. 2013).

Although those preliminary data have to be interpreted cautiously and intense research is required in order to further evaluate the therapeutic effects of COX-2 inhibitors in MD, those results are encouraging for further studies on the inflammatory hypothesis of depression with regard to pathogenesis, course and therapy (Table 17.2).

Interestingly, in a study of patients with rheumatoid arthritis, eternacept, which blocks the interaction of TNF-α with the TNF-α cell surface receptors, showed a highly significant antidepressant effect on the beck depression inventory (BDI), a self-rating scale (Raison et al. 2012). Depression, however, was not the primary outcome criterion in this study (Tyring et al. 2006).

Another interesting study with the TNF-α receptor blocker infliximab in treatment-resistant patients with MD found no overall benefit, but did find a benefit in those with higher levels of inflammatory markers, such as CRP, TNF-α or soluble TNF-receptors (Raison et al. 2012), i.e. those patients who showed pronounced signs of inflammation responded better to infliximab than depressed patients without signs of inflammation.

Although those preliminary data have to be interpreted cautiously and further research is needed to evaluate the therapeutic effects of COX-2 inhibitors in MD, these results are encouraging for further studies on the inflammatory hypothesis of depression and the pathogenesis, course and therapy of the disease.

Several COX-2-inhibiting substances have been withdrawn from the market because of side effects, especially during long-term use. Compared to mixed COX-1/COX-2 inhibitors such as aspirin, the selective COX-2 inhibitors show fewer gastrointestinal side effects such as bleeding but a higher rate of cardiovascular side effects (Katz 2013). In a placebo-controlled study of rofecoxib—withdrawn from the market because of cardiovascular side effects—the increased relative risk for a cardiovascular event became apparent after 18 months of treatment; during the first 18 months, the event rates were similar in the rofecoxib and placebo groups. The results primarily reflect a greater number of myocardial infarctions and ischaemic cerebrovascular events in the rofecoxib group, while the overall and cardiovascular mortality was similar in the rofecoxib and placebo groups (Bresalier et al. 2005). These data for rofecoxib (at a dose of 25 mg/day) show that short-term treatment with COX-2 inhibitors is safe with regard to cardiovascular side effects. The cardiovascular effects, however, vary between the individual COX-2-inhibiting drugs. The strongest evidence for an increased risk of serious cardiovascular events is with rofecoxib therapy. Celecoxib therapy may be associated with an increased risk of cardiovascular events, but only when used at doses substantially higher than those recommended for the treatment of arthritis. There is a greater body of evidence supporting the relative cardiovascular safety of celecoxib when used at the doses recommended for the treatment of arthritis than for any of the other selective COX-2 inhibitors or NSAIDs (Howes 2007).

Increased risk for cardiovascular events was an exclusion criterion in our studies with celecoxib, during which we did not observe any cardiovascular events. In our recent studies, we estimated N-terminal probrain natriuretic peptide (NT-proBNP) as a marker for an increased risk for cardiovascular events; only patients showing normal NT-proBNP values were included into the studies. Plasma NT-proBNP is the best diagnostic marker for increased risks for cardiovascular events and has high sensitivity and specificity (Toufan et al. 2014).

In addition to the above-mentioned side effects of COX-2 inhibitors, some further concerns have been raised for the use of COX-2 inhibitors in depression (Maes 2012), for example the rebalancing effect of the type 1/type 2 immune response (Aid and Bosetti 2011). However, this effect seems to drive the therapeutic effect in schizophrenia. Many studies—including meta-analyses—have proven the beneficial effect of celecoxib and its tolerability in MD is well established. Therefore concerns about the use of COX-2 inhibitors, raised particularly on the basis of in vitro studies, have been disproved by clinical studies, although long-term studies are lacking.