Introduction
Increasing data indicate that the immune system in general and inflammation in particular may play a role in the pathophysiology of psychiatric disorders including depression and anxiety ( ). These data include evidence of low-grade inflammation in subgroups of patients with a variety of psychiatric diseases as well as effects of inflammatory mediators on neurotransmitter systems and neurocircuits that regulate behavior ( ). Interestingly, there appears to be a special relationship between increased inflammation and treatment response, best documented in depression ( ; ; ; ). Data indicate that increased inflammatory markers prior to initiation of therapy predict poor treatment response to conventional antidepressants ( ; ), and treatment-resistant depressed patients exhibit increased markers of inflammation ( ; ; ). Moreover, a rich literature describes the multiple pathways by which inflammatory factors can disrupt neurotransmitter metabolism and growth factors and ultimately neurocircuits that are the foundations of treatment response to conventional antidepressant medications ( ; ). Increased inflammation is also associated with a better response to ketamine and electroconvulsive therapy (ECT) ( ; ), suggesting that a primary effect of inflammation is to undermine the efficacy of conventional antidepressant treatments, while increasing the responsiveness to alternative treatment strategies.
Given the relationship between inflammation and pathophysiologic mechanisms related to depression and other psychiatric disorders, studies using drugs with antiinflammatory properties have been investigated ( ). In general, there is modest evidence that antiinflammatory therapies may have potential to treat depression including treatment-resistant depression (TRD). However, due to multiple confounds involving both the designs and the drugs used in these studies, the results are almost impossible to interpret, and therefore, the field is bereft of meaningful clinical trial data to serve as the foundation for future research ( ; ; ).
The goal of this chapter is to first provide the basis for the hypothesis that inflammation plays a role in depression, including a review of the neurobiological and immunological mechanisms involved. Next, we will address the special relationship between inflammation and treatment resistance. Finally, we will review relevant clinical trials that have targeted inflammation in depression including TRD and make recommendations regarding the drugs and clinical trials’ designs of future studies in this area. It is the position of the authors that there is great promise in further understanding the translational relevance of the rich database that has been developed regarding the impact of inflammation on the brain and behavior. Indeed, as will be described, patients with increased inflammation can be readily identified, and given our knowledge regarding inflammatory pathways to pathology, we believe that targeting inflammation and/or its downstream effects on the brain will provide an important platform for precision psychiatry. By the same token, we also recognize that failure to integrate what is known about inflammation and its role in depression and other psychiatric disorders will result in a missed opportunity based on outdated clinical trial methodologies, drugs with multiple off-target effects, and outcomes that are not well-aligned with our understanding of the impact of inflammation on the brain.
Foundation for the hypothesis of inflammation’s role in depression
Patients with depression exhibit all the cardinal features of a chronic inflammatory response
The earliest studies examining the relationship between the immune system and depression were derived from human and laboratory animal studies demonstrating that a variety of stressors could inhibit lymphocyte proliferation in vitro ( ). Indeed, numerous studies demonstrated that peripheral blood immune cells from depressed patients exhibited decreased in vitro B- and T-cell responses as well as decreased natural killer (NK) cell activity ( ). These data suggested that depression was associated with suppression of acquired immune responses. Interestingly, however, data began to surface that in contrast to acquired immune responses, there was evidence of activation of the innate immune response as revealed by increased concentrations of acute phase proteins ( ). This early work served as the foundation for subsequent studies demonstrating increased markers of an activated innate immune inflammatory response including increased inflammatory cytokines, increased chemokines, and increased cellular adhesion molecules as well as increased acute phase proteins. Hundreds of papers have been published in this area, and meta-analyses of the literature have established reliable elevations of the inflammatory cytokines interleukin (IL)-6 and tumor necrosis factor (TNF) and the acute phase protein C-reactive protein (CRP) in the peripheral blood of depressed patients ( ; ). Postmortem studies have also revealed evidence of an inflammatory response in the brain of depressed suicide victims including increased expression of toll-like receptors (TLRs) and other inflammatory signaling molecules in brain parenchyma ( ), activation of microglia and astrocytes ( ; ; ), and the presence of perivascular monocytes ( ; ). In addition, positron emission tomography (PET) using ligands that bind to the translocator protein (TSPO), which is expressed in microglia as well as other cell types in the brain, has provided evidence of microglial activation in depression ( ), although the specificity of these findings for neuroinflammation has yet to be established ( ). Finally, several large, longitudinal, epidemiological studies have shown that CRP and other inflammatory mediators can predict the development of depression, with less compelling data suggesting that depression leads to increased inflammation ( ).
Relationship with treatment nonresponse
Relevant to the response to antidepressants, a meta-analysis of the literature suggested that increased inflammatory biomarkers prior to treatment with conventional antidepressants (e.g., selective serotonin reuptake inhibitors (SSRIs) and serotonin norepinephrine reuptake inhibitors (SNRIs)) is predictive of a reduced therapeutic response in ambulatory depressed patients ( ). Moreover, studies have indicated that treatment-resistant depressed patients exhibit increased inflammatory markers including IL-6, TNF, and CRP ( ; ). Molecular markers of inflammation including mRNA for the purinergic receptor P2RX7, IL-1beta, TNF, the CXC chemokine CXCL12, and migration inhibitory factor have also been shown to predict antidepressant nonresponse to SSRIs and SNRIs ( ). In addition, two studies have suggested that CRP may predict a differential response to antidepressants, with both studies finding a poor response to escitalopram in individuals with a CRP > 1 mg/L ( ; ). One of these studies found bupropion and the other found nortriptyline to lead to better responsiveness in the subgroup with CRP > 1 mg/L. Interestingly, increased inflammatory markers also have been found to predict a positive response to ketamine and ECT ( ; ). It should be noted, however, that most studies identifying inflammatory markers as predictors of treatment response have used post hoc analyses. Few studies have a priori determined that inflammatory markers can predict response to antidepressant or antiinflammatory treatment. Taken together, these data suggest that inflammatory markers have the potential to help guide treatment decision-making, especially as it relates to treatment response. Moreover, as discussed below, increased inflammatory markers appear to have special relevance to well-known predictors of treatment nonresponse including obesity, early life stress, medical illnesses, and anxiety ( ; ).
Administration of inflammatory cytokines/stimuli causes depressive symptoms
A second major body of data that supports the relationship between inflammation and depression, while also addressing the cause-and-effect nature of this relationship, is work demonstrating that administration of inflammatory cytokines (e.g., interferon (IFN)-alpha) or inflammatory stimuli (e.g., endotoxin or typhoid or influenza vaccination) leads to behavioral changes that are consistent with those seen in patients with depression ( ). For example, chronic treatment with IFN-alpha leads to symptoms that meet criteria for major depression in 30%–50% of subjects (e.g., patients with cancer or infectious diseases) depending on the dose ( ; ). Moreover, comparison of symptoms induced by IFN-alpha with symptoms from otherwise healthy depressed individuals yield few meaningful differences between the two, although IFN-alpha-treated patients exhibit an overrepresentation of neurovegetative symptoms ( ). Endotoxin and typhoid and influenza vaccination, typically administered acutely to healthy volunteers in laboratory settings, also lead to symptoms of depression, albeit short-lived ( ; ; ). The capacity of these inflammatory stimuli to induce a behavioral syndrome like depression provides compelling data that inflammation can cause depression. Based on these data, much has been learned about the central nervous system (CNS) mechanisms by which inflammation affects neurotransmitter systems and neurocircuits in the brain to change behavior ( ).
Inhibition of inflammation reduces depressive symptoms
Further supporting the cause-and-effect relationship between inflammation and depression are studies indicating that inhibition of inflammation can reduce depressive symptoms. The most convincing data come from patients with autoimmune and inflammatory disorders who have been treated with anticytokine therapies that both reduce inflammation and reduce depressive symptoms ( ; ). In many of the studies, the decrease in depressive symptoms is independent of the improvement in disease activity, although the possibility that behavioral effects are tied to the response of the underlying disease process remains a consideration. Studies also suggest that drugs with antiinflammatory properties including celecoxib, aspirin, minocycline, statins, pioglitazone, and omega-3 fatty acids have antidepressant effects in otherwise healthy individuals ( ). Nevertheless, as indicated earlier, there are many complications in the clinical trial designs and the drugs used that make interpretation of these data challenging ( ; ; ).
Clinical predictors of treatment resistance and inflammation
Several clinical factors have been shown to be predictors of antidepressant treatment nonresponse and are related to inflammation. These include obesity, early life stress, medical illness, and anxiety ( ; ). There is a dose-response relationship between body mass index and inflammatory markers. In addition, early life stress has been reliably associated with increased inflammation and an increased inflammatory response to stress in exposed individuals ( ). A rich literature also describes the strong relationship between markers of inflammation and medical illnesses including diabetes, cardiovascular disease, and cancer ( ). Indeed, elevated CRP has been shown to be a reliable predictor of the development of these disorders, and the American Heart Association has defined levels of inflammatory risk based on CRP with a CRP < 1 mg/L representing low risk, a CRP between 1 and 3 mg/L representing moderate risk, and a CRP > 3 mg/L representing high risk ( ). Finally, data also strongly implicate increased inflammatory markers in patients with anxiety disorders ( ).
Mechanisms by which inflammation affects antidepressant treatment response
There are many pathways by which inflammation can undermine the fundamental mechanisms by which conventional antidepressants act including effects on monoamine metabolism, glutamate metabolism and neurogenesis and neural plasticity ( Fig. 15.1 ). These effects of inflammation in turn influence neural circuits in the brain, especially those that regulate motivation and motor activity as well as arousal, anxiety, and alarm ( ; ).
Monoamine metabolism
SSRIs and SNRIs act in large part by increasing the synaptic availability of the monoamine neurotransmitters serotonin and norepinephrine. Inflammation has multiple effects on monoamine metabolism including decreasing monoamine synthesis and release and increasing monoamine reuptake ( Fig. 15.1A ) . Through effects on tetrahydrobiopterin (BH4), inflammation decreases monoamine synthesis ( ). BH4 is an essential enzyme cofactor for the enzymes responsible for synthesis of the monoamines. These enzymes include tryptophan hydroxylase, which converts tryptophan to 5-hydroxytryptophan, a metabolic intermediate in the synthesis of serotonin; phenylalanine hydroxylase, which converts phenylalanine to tyrosine; and tyrosine hydroxylase which converts tyrosine to L-dihydroxyphenylalanine (L-DOPA) a metabolic intermediate in the synthesis of dopamine (DA) and norepinephrine. Relevant to inflammation, BH4 also serves as an enzyme cofactor for nitric oxide synthase (NOS), which leads to the production of nitric oxide, a key mediator of pathogen resolution ( ). Thus, BH4 can be usurped by excessive activation of NOS in the context of an immune challenge and in the process is converted to its inactive form BH2. BH4 is also highly sensitive to oxidative stress that is prevalent in the context of an inflammatory response, and as a result is irreversibly degraded into the inert dihydroxyanthopterin ( ). Of note, increased IL-6 in the CSF of patients receiving IFN-alpha was associated with decreased CSF BH4 ( ). Moreover, the phenylalanine to tyrosine ratio (a potential index of BH4 activity) was associated with increased scores on the Multidimensional Fatigue Inventory and decreased CSF DA ( ).
Inflammation is also associated with decreased monoamine release. Studies have shown that administration of inflammatory stimuli to humans and laboratory animals leads to reduced release of DA ( ). For example, nonhuman primates administered the inflammatory cytokine IFN alpha for 4 weeks exhibit decreased release of DA in response to amphetamine and K + in the striatum as determined by in vivo microdialysis ( ). In vivo microdialysis has also demonstrated decreased release of DA following intraperitoneal administration of IL-6 in the nucleus accumbens (part of the ventral striatum) in rodents ( ). In human studies, neuroimaging using F18-labeled DOPA demonstrated decreased release of radiolabeled ligand from DAergic neurons ( ). Interestingly, consistent with inflammation’s effects on DA synthesis, administration of the immediate DA precursor l -DOPA via reverse microdialysis was found to restore the DA response to amphetamine after chronic (4 weeks) IFN-alpha administration to non-human primates ( ). Of note, both IL-1 and TNF have been found to reduce expression of the vesicular monoamine transporter 2, which is responsible for packaging DA into vesicles prior to release ( ).
Finally, inflammation has been shown to increase the expression and activity of the serotonin and norepinephrine transporters ( ). These effects are mediated by p38 mitogen activated protein kinase (MAPK), and blockade of p38 MAPK has been shown to reverse the effects of endotoxin on depressive-like behavior in rodents, while also reversing endotoxin’s reduction of extracellular serotonin ( ). Similar effects of MAPK on the DA transporter have been described ( ).
Taken together, the effects of inflammation on monoamine metabolism serve to reduce monoamine availability and thus undermine a primary effect of conventional antidepressants, which is to increase synaptic monoamines.
Glutamate metabolism
Another pathway by which inflammation can undermine the effects of conventional antidepressants is to affect pathways that are not the targets of these drugs ( Fig. 15.1B ). One such pathway is related to the metabolism of glutamate, an excitatory amino acid, whose concentration is carefully controlled within the synapse ( ; ). Astrocytes are a major target of inflammation with well-established effects on excitatory amino acid transporters (EAAT) especially EAAT2, which is the primary glutamate transporter on astrocytes that removes glutamate from the tripartite synapse, consisting of pre- and postsynaptic glutamatergic neurons and astrocytes ( ). Several inflammatory cytokines including TNF and IL-1beta have been shown to decrease the expression of EAAT2 on astrocytes and can lead to reverse efflux of glutamate from these transporters ( ; ; ). In addition, inflammation can activate indoleamine 2,3 dioxygenase (IDO) which converts tryptophan to kynurenine (KYN) ( ). KYN is in turn transported to the brain through the large neutral amino acid transporter (LAT2) in the blood-brain-barrier (BBB) where it is converted by microglia into quinolinic acid (QUIN). QUIN can cause astrocytes to decrease the reuptake of glutamate and increase its release ( ). Inflammation also leads to increased oxidative stress that induces astrocytes to produce the antioxidant glutathione through a mechanism called the xC transporter that extrudes one molecule of glutamate in exchange for one molecule of cysteine ( ). Taken together, these and other effects of inflammation on glutamate metabolism can lead to increased synaptic glutamate and ultimately spillover of glutamate into the extrasynaptic space. Once in the extrasynaptic space, glutamate can bind to extrasynaptic N -methyl- d -aspartate (NMDA) receptors which in some reports can lead to decreased production of growth factors including brain derived neurotrophic factor (BDNF) and ultimately excitotoxicity ( ). Extrasynaptic spillover of glutamate can also influence local firing patterns of adjacent neurons leading to reduced synaptic precision and ultimately chaotic and incoherent local and interregional neuronal signaling and connectivity ( ). For example, patients with major depression and increased inflammation as indexed by CRP in conjunction with elevated basal ganglia glutamate as measured by magnetic resonance spectroscopy exhibit decreased regional homogeneity (ReHo), a measure of local coherence in firing of neighboring voxels in the brain ( ). Moreover, in depressed patients with both increased CRP and basal ganglia glutamate, multiple regions in the brain exhibit decreased ReHo including networks that are involved in motivation in association with symptoms of anhedonia ( ). Depressed patients with increased CRP plus decreased ReHo also exhibit treatment resistance to antidepressant therapy ( ). Thus, alterations in glutamate as a function of the impact of inflammation on astrocytes and glutamate metabolism is another mechanism by which inflammation can undermine the effects of conventional antidepressants, which have limited effects on glutamate neurotransmission.
Neural plasticity
Several studies have indicated that neurogenesis plays an important role in the response to conventional antidepressant therapy ( ). Indeed, blockade of neurogenesis in the brain using radiation for example can abrogate the antidepressant effects of conventional antidepressants seen in laboratory animals exposed to chronic stress ( ). Relevant in this regard is that inflammation also has been shown to inhibit neurogenesis both in vitro and in vivo and in the context of chronic stress ( Fig. 15.1C ) ( ). For example, laboratory animals exposed to chronic mild stress exhibit increased IL-1 and decreased neurogenesis in conjunction with the development of depressive-like behavior ( ; ). Blockade of IL-1 reverses the effects of chronic stress on depressive-like behavior in these animals, an effect ultimately mediated by inhibiting the inflammatory transcription factor nuclear factor kappa B (NF-kB), which decreases BDNF, a primary driver of neurogenesis ( ; ). The inhibitory effects of inflammation on neurogenesis represent another possible pathway by which inflammation can sabotage the effects of conventional antidepressants which are in part dependent on stimulating the growth and development of new neurons.
Impact of inflammation on neurocircuits
Given the impact of inflammation on neurotransmitter systems and neurogenesis/neural plasticity, much attention has been paid to the neural circuits that are affected by inflammation to lead to behavioral change. Based on a variety of neuroimaging strategies during inflammatory states, the circuits in the brain most reliably affected by inflammation involve those that regulate motivation and motor activity and those that regulate arousal, anxiety, and alarm ( ; ; ). For example, administration of the inflammatory cytokine IFN-alpha to patients with hepatitis C for 4 weeks was shown to significantly decrease activation of the ventral striatum (VS) using functional magnetic resonance imaging (fMRI) and a hedonic reward (gambling) task ( ). Decreases in VS activation in turn were correlated with increased anhedonia. Similar findings have been observed in healthy volunteers acutely exposed to endotoxin or typhoid vaccination, demonstrating the reproducibility of the impact of inflammation on these brain circuits ( ; ). In patients with major depression (MD), increased inflammation as indexed by CRP was also found to be associated with decreased functional connectivity between the VS and ventromedial prefrontal cortex (vmPFC) in conjunction with symptoms of anhedonia ( ). Increased CRP was also found in MD patients to be associated with decreased connectivity between dorsal striatum and motor regions of the PFC in relation to symptoms of psychomotor slowing ( ). These effects of inflammation on circuits regulating motivation and motor activity complement data demonstrating that inflammatory stimuli also activate circuits involving the dorsal anterior cingulate cortex (dACC), insula, hippocampus, and amygdala in association with symptoms of anxiety ( ; ). Indeed, increased inflammation as measured by CRP in patients with MD was associated with decreased connectivity between PFC and amygdala in conjunction with anxiety symptoms ( ). Increased inflammatory responses to stress have also been associated with increased activation of the dACC in association with task-based emotional distress during fMRI ( ). Interestingly, the effects of inflammation on these circuits relate in part to data demonstrating the impact of stress and inflammation on blood-brain barrier integrity in the region of the nucleus accumbens/VS and the trafficking of immune cells to brain vasculature and meninges including areas surrounding the amygdala ( ; ; ; ). Taken together, these data indicate that the effects of inflammation on specific neurocircuitry provide an excellent platform to establish relevant outcome variables for therapeutic interventions designed to reverse inflammation’s effects on the brain.
Therapeutic targets addressing TRD and inflammation
Given the many mechanisms by which inflammation can affect neurotransmitter systems, neural plasticity and neurocircuits in the brain to lead to treatment nonresponse, there is a target rich environment supporting opportunities for development of new treatments for TRD ( ). The most obvious target is inflammation itself followed by the metabolic pathways within immune cells that drive the production and release of inflammatory mediators. Downstream effects of inflammation on the brain are also viable treatment options including most notably the impact of inflammation on DA and glutamate. Moreover, recent data suggest that novel technological developments in neuromodulation may be relevant regarding inflammation and treatment resistance. Finally, although not explicitly discussed in this chapter, lifestyle interventions also have a role in reducing inflammation including diet and exercise as well as stress reduction techniques. Aggressive management of diseases associated with inflammation is additionally relevant in this regard.
Inflammation
A multitude of studies have examined the efficacy of a variety of drugs with antiinflammatory properties in the treatment of major depression ( ). Although some studies have specifically addressed the role of these agents in the context of TRD, most of the literature has focused on patients with MD who are presumably treatment responsive or are only partially treatment responsive. A vast array of agents has been tested including COX 1 and 2 inhibitors (e.g., aspirin, nonsteroidal antiinflammatory drugs (NSAIDs), celecoxib), omega-3 fatty acids, minocycline, statins, pioglitazone, and cytokine antagonists ( ; ; ). Metaanalyses of the data generated from these trials suggests that there may be modest efficacy of these agents in heterogeneous populations of patients with depression ( ). In general, however, these studies have been small, and given that only a subpopulation of depressed patients have increased inflammation, it is surprising that any effect has been observed at all. Indeed, in the largest trial to date, no effect was found for either celecoxib or minocycline compared to placebo ( ). In addition, longitudinal, epidemiological studies with large populations examining whether chronic exposure to antiinflammatory agents (commonly aspirin, NSAIDS, and statins) reduces the development of depression have been mixed and largely negative ( ; ; ; ). As discussed in detail below, there are many pitfalls in the studies that have been conducted to date, and therefore, the field is bereft of interpretable data to move the field forward. Nevertheless, several trials have yielded promising data using inflammatory biomarkers in post hoc fashion to identify patients more likely to respond, and a few trials have used inflammatory makers prospectively to identify likely responders. For example, two of the anticytokine trials (one of which was on TRD subjects) found that patients with increased inflammation at baseline were more likely to respond ( ; ), whereas a third study found early life stress (often associated with inflammation) was a predictor of response ( ). In addition, a recent trial using minocycline found that TRD patients with a CRP ≥ 3 mg/L were more likely to respond to minocycline augmentation than patients with a CRP < 3 mg/L ( ). Moreover, increased brain TSPO binding at baseline, possibly reflective of microglial activation, was associated with a greater response to celecoxib in an open label study ( ).
There are many antiinflammatory drugs available to be used in future trials including those targeting novel pathways such as Janus kinase (JAK)1 and JAK2, p38 MAPK, the inflammasome (via inhibition of P2X7 receptor), and anticytokine drugs targeting T-cell cytokines such as anti-IL-17 and anti-IL-12/23 as well as the anticytokine drugs that block TNF, IL-1, and IL-6 ( ). However, the most pressing issue at this juncture is to conduct trials that leverage the knowledge of the effects of inflammation on the brain and behavior to conduct trials that will yield interpretable data (see later).
Immunometabolism
Upon activation, immune cells undergo shifts in metabolism that involve a relative increase in glycolysis and a reduction in oxidative phosphorylation ( ). This metabolic reprogramming, which occurs in the presence of oxygen (aerobic glycolysis), is referred to as the Warburg effect and has also been described in cancer cells ( ). Of note, glycolysis is highly energy inefficient generating 2 ATP per molecule of glucose compared to 34 ATP per molecule of glucose during oxidative phosphorylation. The shift to glycolysis allows rapid cellular activation and proliferation without the need to generate more mitochondria where oxidative phosphorylation occurs. The role of metabolic reprogramming in the response to antiinflammatory medications was first observed in the context of TRD patients being treated with the TNF antagonist infliximab ( ; ). Gene expression analysis of peripheral blood immune cells revealed that infliximab responders versus nonresponders exhibited an overrepresentation of not only pathways related to TNF and NF-kB but also pathways related to glycolysis and gluconeogenesis as well as lipid metabolism, which is also associated with immunometabolic reprogramming ( ). More recent studies have revealed that molecular pathways in peripheral blood immune cells including hypoxia-inducible factor 1 alpha, protein kinase B (also known as AKT), and phosphoinositide 3-kinase/AKT/mammalian target of rapamycin (mTOR) as well as insulin signaling are associated with anhedonia in patients with high inflammation (CRP > 3 mg/L) and correlate with functional connectivity and reward-related pathways that are associated with motivational deficits ( ; ). These data are consistent with the impact of inflammation on symptoms of anhedonia and expand consideration of the immunologic pathways involved in inflammation to include cellular immunometabolism.
Of relevance to treatment targets for TRD, drug development in autoimmune and inflammatory disorders is targeting glycolysis and related molecular pathways as novel treatments, in contradistinction to targeting the molecular pathways and products of the inflammatory response itself ( ). Drugs of interest in this regard include dimethyl fumarate (used in multiple sclerosis), which inhibits the enzyme glyceraldehyde 3-phosphate dehydrogenase that occupies a central position in the glycolytic pathway ( ); imatinib (used for cancer) which inhibits tyrosine kinases that in turn inhibit hexokinase (HK) and glucose-6-phosphate dehydrogenase, enzymes that promote glycolysis ( ; ), and lonidamine, which inhibits mitochondrially bound HK ( ; ). In addition, recent data suggest that rapamycin (an mTOR complex 1 inhibitor) prolonged the antidepressant effects of ketamine in patients with MD, consistent with the prediction of response to ketamine by inflammatory mediators including IL-6 ( ). Finally, given the interaction between insulin resistance and inflammation in association with anhedonia, insulin sensitizing agents have been considered and, in some cases, tested in depression including thiazolidinediones (e.g., pioglitazone), which stimulates peroxisome proliferator-activated receptor gamma, and metformin, which promotes 5′ adenosine monophosphate-activated protein kinase ( ).
Neurotransmitter targets
Dopamine
As noted earlier, the effects of inflammation on DA synthesis, release and reuptake appear to converge on alterations in ventral striatal responses to rewarding stimuli as well as objective and clinical measures of motivational deficits ( ; ; ). Therefore, increasing the availability of DA is a reasonable strategy to reverse inflammation’s effects on behavior, while also providing a target in the brain as reflected by reward circuitry involving the VS and vmPFC ( ). Indeed, as noted earlier, two studies in depression have shown that increased CRP (> 1 mg/L) was associated with a poor response to escitalopram, but a much better response to nortriptyline or bupropion ( ; ). These latter drugs both impact DA metabolism, possibly providing the rationale for their relative success in the face of increased inflammation.
In terms of DA synthesis, supplementation of BH4 with various precursor molecules can facilitate the conversion of phenylalanine to tyrosine and tyrosine to l -DOPA ( ). Such molecules include folic acid, l -methylfolate and S -adenosyl methionine ( ). Relevant in this regard, treatment of inadequate responders to SSRIs with l -methylfolate led to improvement in depressive symptoms in those individuals with increased baseline levels of inflammatory markers including CRP, TNF, and IL-6 in conjunction with markers of metabolic dysfunction including BMI and leptin ( ). These data not only support the notion that inflammation and metabolism may interact to disrupt treatment responsiveness but also suggest that depressed patients with this immunometabolic profile may be most responsive to drugs targeted to DA in general and DA synthesis in particular. Of note, l -DOPA, the immediate precursor of DA, has also been shown to improve psychomotor speed and depressive symptoms in elderly depressed individuals who are often treatment resistant ( ). Additionally, l -DOPA is being tested for its capacity to reverse decreased functional connectivity between VS and vmPFC in patients with MD and increased inflammation ( NCT02513485 ).
Other medications that affect DA availability are those that increase DA release and inhibit synaptic reuptake. Drugs that increase DA release include amphetamine, methamphetamine, and lisdexamfetamine. Although there is limited data supporting the efficacy of these agents in TRD, and no data on patients with TRD and increased inflammation, studies have found efficacy of lisdexamfetamine as an augmentation strategy to treat persistent symptoms of depression in MD patients ( ; ). In addition, several drugs have the capacity to block DA reuptake including bupropion (discussed earlier) as well as methylphenidate and modafinil. Methylphenidate has been shown to reverse the effects of IL-6 on effort-based motivation in laboratory animals ( ), while improving fatigue in cancer patients who often exhibit increased inflammatory markers ( ; ). Modafinil has also been found to reduce fatigue in patients with depression or TRD ( ), although the efficacy of both methylphenidate and modafinil as antidepressants is not well established. Of note, bupropion is currently being tested (vs escitalopram) for superior efficacy for anhedonia in MD patients with CRP ≥ 2 mg/L ( NCT04352101 ).
Finally, there has been interest in the use of DA agonists to treat patients with TRD. Indeed, pramipexole has been shown to improve depressive symptoms in TRD patients at high doses. Pramipexole has also shown antidepressant efficacy in patients with Parkinson’s disease as well as patients with MD with special benefit for symptoms of anhedonia and psychomotor retardation ( ; ; ). Finally, aripiprazole and brexpiprazole both exhibit partial D2/3 agonism and both are FDA approved as augmentation strategies for patients with TRD ( ; ). Whether TRD patients with increased inflammation may be especially responsive to these medications remains untested.
Glutamate
Given the profound antidepressant effects of ketamine, a NMDA antagonist, and the prediction of ketamine response by inflammatory markers including IL-6 and CRP in TRD patients as well as a laboratory animal model of TRD ( ; ), there is reason to believe that drugs that target glutamate and its receptors may be especially relevant for TRD patients with increased inflammation ( ). Indeed, studies in laboratory animals have shown that ketamine can block the development of endotoxin-induced depressive symptoms ( ). However, despite the success of ketamine and its recently FDA-approved ( S )-enantiomer esketamine, drugs targeting glutamate and its NMDA, AMPA, and metabotropic glutamate receptors have yet to show efficacy in antidepressant trials to date. For example, memantine, an NMDA antagonist, has failed to show efficacy in multiple trials in MD ( ), and lanicemine and GLYX-13 (rapastinel) did not separate from placebo, despite initial promise in disorders including depression ( ; ). There are numerous other glutamate receptor drugs in development, however, review of these agents is outside the scope of this chapter ( ). Nevertheless, the possibility that these glutamatergic drugs may be uniquely effective in patients with increased inflammation has yet to be examined. Indeed, the compound AXS-05, a combination of bupropion and the NMDA receptor antagonist dextromethorphan, may be of special interest, given its potential capacity to address both DA and glutamate pathologies in patients with TRD and increased inflammation. AXS-05 has shown initial promise and has been given FDA Breakthrough Therapy designation ( ). Finally, riluzole is an interesting compound in that it has been shown to decrease presynaptic glutamate release and increase the activity of EAAT2 ( ). Riluzole has also been shown in laboratory animal models of chronic unpredictable stress to mitigate anhedonic behavior ( ). Riluzole has exhibited evidence of benefit in at least two small studies in TRD ( ; ), in addition to augmentation of the response to ketamine, delaying the time to relapse ( ). Current studies are underway to examine the efficacy of riluzole in treating depressive symptoms in the context of increased inflammation in cancer patients ( NCT02796755 ).
Another target relevant to glutamate is the IDO pathway and the generation of QUIN. As noted earlier, QUIN is largely derived from peripherally-produced KYN that is transported to the brain through LAT-2 in the BBB ( ). Leucine is an amino acid that competes with KYN for transport into the brain and was shown to block the development of depressive-like behavior following the administration of endotoxin to rodents ( ). Leucine is currently being tested in patients with MD ( NCT03079297 ). Inhibition of IDO is another relevant strategy to reduce QUIN. Several IDO inhibitors are currently in clinical testing ( ). However, the majority of current trials of IDO inhibitors are in cancer where the blockade of IDO activity spares tryptophan from being converted to KYN ( ; ). Tryptophan is an essential amino acid for T-cell activation, which is critical in the immunologic response to tumors.
Neuromodulation
The impact of inflammation on neural circuitry invokes the potential use of neuromodulation strategies to restore the integrity of these circuits by stimulating or inhibiting relevant brain regions using a variety of techniques including ECT, repetitive transcranial magnetic stimulation, deep brain stimulation, and MRI-guided ultrasound. As noted earlier, increased inflammatory markers including IL-6 has been shown to predict response to ECT ( ). Response to these other neuromodulation strategies has yet to be evaluated in humans.
Although these strategies focus on the downstream effects of inflammation on the brain, new neuromodulation technologies have been developed that can influence the immune system itself. Indeed, stimulation of the afferent vagus nerve can inhibit inflammatory responses through the release of acetylcholine, which binds to the alpha 7 subunit of the nicotinic receptor on immune cells that in turn inhibits NF-kB ( ). Such stimulation of vagal afferents has been shown to reduce mortality following endotoxin administration in laboratory animals, and more recently has received FDA approval for use in rheumatoid arthritis ( ). This use of vagal nerve stimulation is the first example of a neuromodulation technique being employed to target the immune system. Of note, implanted vagal nerve stimulation devices used for depression generate efferent signals to the brain and are therefore not designed to target the immune system. Nevertheless, recent work has shown that transcutaneous vagal nerve stimulation, which stimulates both efferent and afferent vagal nerve fibers, blocks stress-induced production of peripheral blood IL-6 and IFN-gamma in patients with PTSD and may have therapeutic potential ( ). Finally, recent data suggest that exposure to lights flickering at specific frequencies can drive relative frequencies in the brain, which in turn can influence the activation state of microglia ( ). This technology has only been applied to laboratory animals but holds intriguing promise for regulating CNS immune responses in neuropsychiatric disorders.
Future directions
Based on the information provided above, there are many opportunities to advance strategies targeting inflammation and its downstream effects on the brain to address depressed patients who have experienced nonresponse to conventional antidepressant therapies. Although the lion’s share of research on drugs in this area has not focused on patients with treatment resistance, there is sufficient data to suggest that drugs affecting inflammation’s impact on behavior may be especially relevant to patients with TRD, given that these patients are more likely to exhibit increased inflammatory markers and increased inflammatory markers predict treatment nonresponse to conventional antidepressants. Moreover, as noted earlier, increased inflammatory markers predict response to treatments including ECT and ketamine that are typically reserved for TRD patients.
Unfortunately, despite the many trials that have been conducted using antiinflammatory medications to treat patients with depression and TRD, there are significant problems in study design, and the drugs employed as well as the outcome variables make interpretation of the extant literature virtually impossible. The lack of attention to the unique aspects of a pathophysiologic pathway with defined effects on the brain has led to a literature poorly aligned with the subgroup of patients that are most likely to respond and if anything jeopardizes the field. Indeed, if what is known about inflammation in depression is not incorporated into clinical trial design, the field will not advance, and an extraordinary opportunity to develop new treatments for depression will be lost.
Clinical trial design
The opportunities afforded clinical trial designs by virtue of our understanding of the role of inflammation in depression are manifold ( Table 15.1 ).
Related posts:
Clinical and epidemiological predictors of treatment-resistant depression
The economic burden of treatment-resistant depression: Cost-of-illness perspective
The neurobiology of treatment-resistant depression
Monoamine oxidase inhibitors for treatment-resistant depression
Switching antidepressants in patients with treatment-resistant depression
Adjunctive strategies for treatment-resistant depression
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
