Injection of illicit drugs is associated with increased transmission of both human immunodeficiency virus (HIV) and hepatitis C virus (HCV). These infections, in turn, may lead to clinically significant neuropsychiatric syndromes. HIV is associated with neurocognitive disorders and psychiatric sequelae, and HCV leads to fatigue, disordered sleep, depression, and cognitive impairment. The medications used to treat these infections may themselves have neuropsychiatric side effects. Combinations of antiretroviral medications, referred to as highly active antiretroviral therapy (HAART), are used for the treatment of HIV, and direct-acting antiviral medications are increasingly used to treat HCV. Given the frequent occurrence of psychiatric illness in both HIV-positive and HCV-positive patients, it is important to consider the potential for drug-drug interactions resulting from the combination of these agents with psychotropic medications. These areas of clinical importance are the focus of this review.
Neuropsychiatry of HIV Infection
HIV infection is associated with both neurocognitive and neuropsychiatric complications. The term “HIV-associated neurocognitive disorders” (HAND) describes a range of potential neurocognitive impairments, including HIV-associated dementia (HAD). Potential neuropsychiatric presentations seen in HIV infection include depression, anxiety, fatigue, mania, and psychosis.
HIV-Associated Neurocognitive Disorders
The pathogenesis of neurocognitive impairment in HIV is poorly understood but may involve HIV infection of perivascular macrophages and brain micro-glia, neurotoxic effects of viral proteins released by infected monocytes, and secondary damage to the central nervous system (CNS) resulting from the inflammatory response against HIV infection.1
HAND is subdivided into three subcategories. The first is asymptomatic neurocognitive impairment (ANI), which refers to cognitive abnormalities in at least two areas but no impairment in function. The second is mild neurocognitive disorder (MND), which describes marked cognitive impairment with mild impairment in functioning. The third is HAD, which is more severe cognitive impairment combined with significant functional impairment.
HAART, which was introduced in 1996, combines multiple antiretroviral medications to maximize clinical efficacy. Although the overall prevalence of HAND has stayed relatively constant from the pre-HAART era to now, the prevalence of HAD has decreased from 14% to 2%, whereas the prevalence of ANI and MND have increased from 16% and 5%, respectively, to 32% and 10%, respectively.2 Similarly, the Multicenter AIDS Cohort Study3 did not find significant progression of neurocognitive disease in 197 HIV-positive patients, with 77% remaining in the same stage during the 4 years of the study.
Cognitive symptoms of HAND include complaints of poor concentration, slowed thinking, impaired short-term memory, and forgetfulness.4 These symptoms may occur even in the presence of a normal CD4+ count and in patients who are adherent to HAART. Apathy, agitation, and social withdrawal are among the behavioral signs that clinicians may encounter. Motor symptoms consist of tremor, unsteady gait, impaired coordination, or leg weakness.4 Of note, cognitive, behavioral, and motor symptoms may wax and wane over time and can occur with a normal CD4+ count. Depending on the specific clinical presentation, differential diagnosis may include CNS opportunistic infections, CNS lymphoma, cerebrovascular disease, major depression, HCV, neurosyphilis, substance abuse, or side effects of HAART.
HAD, the most severe form of HAND, may present acutely over weeks or insidiously over months. In this disorder, patients may report problems with concentration, new learning, memory, and attention. Apathy, social withdrawal, decreased spontaneity, and poverty of speech are also common. Physical examination may be notable for ataxia, imbalance, spasticity, and increased deep tendon reflexes. Processing speed and executive function may also be affected. Activities of daily living may take longer to complete and, over time, patients may need assistance with daily functioning. Deficits may be permanent. At present, HAART is the only treatment for people with HAND.
HIV-infected patients, therefore, should be screened for neurocognitive impairment. This is particularly important because poor memory and attention may impair adherence to HIV treatment regimens. Self-administered questionnaires,5–7 such as the Medical Outcomes Study-HIV Health Survey (MOS-HIV),5 can be employed for both baseline and follow-up evaluation.
Alternatively, the European AIDS Clinical Society6 suggests asking all HIV patients the following questions: (1) Do you experience frequent memory loss (eg, do you forget the occurrence of special events, appointments)? (2) Do you have difficulties paying attention (eg, reading a book, watching a movie)? (3) Do you feel that you are slower when reasoning, planning activities, or solving problems?
Positive answers to any question should be followed up with complete neuropsychological testing. This will help establish baseline functioning for longitudinal assessment and define the precise nature of the deficits.
Depression and Anxiety
The most common neuropsychiatric complication of HIV infection is depression, occurring at 2 to 4 times the rate in the general population;8 however, it remains underdiagnosed and undertreated in HIV-positive patients.9 Because it may occur early or late in the natural course of the illness, regular monitoring for depression is recommended. Factors that influence the development of depression in HIV-infected patients include social stigma, coping with a chronic illness, and the direct effects of HIV on the brain.10 Risk factors for depression with HIV infection include female gender, homelessness, unemployment, poor self-efficacy, and lack of social support.
Given the higher prevalence of depression in HIV infection, the biological link between HIV infection and depression has been studied extensively. One leading hypothesis proposes that activation of the immune system by HIV leads to the release of viral proteins and inflammatory cytokines that are toxic to brain tissue.11 Alternatively, depression may be a psychological consequence of coping with the psychosocial stigma of HIV infection. Supporting this theory is the high rate of social isolation experienced by patients with HIV.12
Differential diagnosis of depression and anxiety in the HIV-positive patient includes grief, a primary Axis I depressive disorder (eg, major depression, dysthymia), an Axis I anxiety disorder (eg, generalized anxiety disorder), HAND, hypoactive delirium, or one of many opportunistic infections. Opportunistic infections (caused by fungi, bacteria, or viruses) and neoplasms of the CNS are more likely to occur at later stages of the disease. One challenge is to differentiate the symptoms of depression, such as fatigue, insomnia, pain, and reduced appetite, as all of these may also result from HIV illness itself. In addition, cognitive symptoms of depression may resemble symptoms of HAND, or these disorders may coexist. HAND can present as apathy and psychomotor slowing, which is easily mistaken for depression. In this context, symptoms that seem to be more associated with depression include anhedonia, loss of appetite, insomnia, and poor concentration.13 In contrast, fatigue and psychomotor slowing are more likely attributed to HIV infection. More extensive neuropsychological assessment may help in differentiating the nature of the symptoms.
Depression is associated with decreased medication adherence and a worse quality of life, which leads to worse clinical outcomes.11 Beyond treatment nonadherence, there are also correlations between depression and lower CD4+ counts and higher viral loads; conversely, lower prevalence of depression has been noted in antiretroviral-treated patients compared to those not using HAART.14 It has been hypothesized that HAART tamps down the immune system activation seen with HIV infection, reducing immune cell activation and neuroinflammatory cytokines, which in turn may decrease depression.14 Regardless of mechanism, effective psychiatric treatment may be lifesaving when poor medication adherence is due to untreated depression and anxiety.
A secondary mania occurring late in the course of HIV illness has been described.15 This presents as a full manic episode (often with irritable rather than euphoric mood) and prominent psychosis (often paranoia). Delirium must be ruled out. Standard anti-mania treatment is indicated. Treatment with antipsychotics is preferred in more-ill HIV patients due to poor tolerance of lithium and carbamazepine in this population.
Psychosis in HIV-positive patients may occur at any time in the course of the illness. Differential diagnosis includes delirium, illicit drug use, HAD, medication side effects, and opportunistic infections. Second-generation antipsychotics are the preferred treatment due to their lower risk for extrapyramidal symptoms.
Delirium must be considered in any HIV-positive patient with altered mental status. Possible etiologies are broad and beyond the scope of this review, but it is important to note that opportunistic brain infections and cerebral neoplasms should be included in the differential diagnoses.
Fatigue is a common complaint among HIV-positive patients. There is a long list of possible causes, including anxiety and depression, insomnia, drug abuse, HAART, or psychotropic side effects, and HIV-related conditions such as anemia, diarrhea, wasting, or hypogonadism. Patients should be screened for depression. In cases where no physical or psychological cause is found, trial of a stimulant or modafanil may be considered.
Neuropsychiatric Complications of Hepatitis C
HCV targets others organs besides the liver. The term HCV syndrome was coined to describe both the hepatic and extrahepatic complications of HCV infection. Chronic infection is associated with neuropsychiatric and neurocognitive disorders in nearly 50% of patients.16 These occur independently of liver disease severity or HCV replication rates, and regardless of stage of fibrosis or infecting genotype. Furthermore, there is no associated structural abnormality on magnetic resonance imaging of the brain. Neuropsychiatric complications include depression, disordered sleep, fatigue, and cognitive impairment. The following text briefly reviews these clinical presentations.
Patients with HCV infection report a variety of psychiatric symptoms including anxiety, depression, somatization, compulsiveness, and hostility.17 In some patients, reaction to a chronic illness partially explains these symptoms, but HCV itself may play a role.18 Both brief and longer episodes meeting the Diagnostic and Statistical Manual of Mental Disorders, fifth edition19 criteria for major depression have been reported20 and can occur outside of treatment with interferon (IFN). Formal and structured depression screening is recommended in all patients with HCV infection. The specific role for HCV in the etiology of neuropsychiatric sequelae is still unclear.
Between 50% and 80% of HCV-infected patients report fatigue.18 This may present as low motivation, difficulty initiating action, impaired concentration and attention, or mental and physical exhaustion. Fatigue seems more prominent in HCV-infected patients relative to those with similar liver impairment due to other causes. Furthermore, it often improves with HCV treatment.21 Compared to noninfected controls, chronic fatigue syndrome occurs twice as frequently in HCV patients. It is challenging to distinguish HCV-related fatigue from that induced by IFN treatment.
Disordered circadian rhythm with reversal of day and night cycle has been associated with hepatic encephalopathy (HE). More commonly, abnormal sleep patterns have been identified in nearly 50% of patients with cirrhotic or noncirrhotic HCV disease.18 Affected areas in noncirrhotic HCV patients include dissatisfaction with sleep quality, altered sleep patterns, increased nocturnal stimulation, and reduced sleep efficiency. These occur regardless of level of circulating HCV. In contrast, fragmented sleep and reduced 24-hour activity has been observed in patients with cirrhosis. Pronounced daytime sleepiness and frequent awakening are seen in patients with minimal HE.23
Neurocognitive impairment in HCV, termed “HCV-associated neurocognitive disorder” (HCV-AND), is characterized by impairment in executive function, sustained attention, verbal learning, working memory, and verbal recall.18 HCV-AND is unrelated to degree of liver impairment and is thought to be distinct from minimal hepatic encephalopathy, which is characterized by reversible neuropsychological deficits, including impaired psychomotor speed and selective attention in the setting of known cirrhosis or portal-systemic encephalopathy. Impaired concentration and memory was observed in a group of patients with active viral replication.24 In contrast, people who had cleared HCV demonstrated cognitive performance equivalent to controls, implicating a putative role for HCV.
Neuropsychiatric Adverse Events of Haart
HAART comprises various categories of medications including non-nucleoside reverse transcriptase inhibitors (NNRTI), nucleoside/nucleotide reverse transcriptase inhibitors (NRTI), protease inhibitors (PI), a fusion inhibitor, an entry inhibitor, and an integrase inhibitor (Table 1). The NNRTI efavirenz has the highest risk of causing a wide range of neuropsychiatric side effects, including depression, insomnia, vivid dreams, hallucinations, agitation, confusion, and dizziness.25 The NRTI zidovudine at high doses has been associated with mania.26 PIs have been associated with sexual dysfunction.27 The NRTIs stavudine, etravirine, and zalcitabine may cause peripheral neuropathy.28
Classes of Antiretroviral Medications for Treatment of HIV
Neuropsychiatric Adverse Events of Hepatitis C Medications
It has long been known that pegylated IFN (Peg-IFN) can induce depression in susceptible people, particularly those who fail to clear HCV RNA.18 The range of reported symptoms associated with this agent include depressed mood, anxiety, hostility, slowness, fatigue, sleep disturbance, lethargy, irritability, emotional lability, social withdrawal, and poor concentration,29 and they typically present between 8 and 12 weeks after the start of treatment. A meta-analysis and systematic review of 26 observational studies reported a cumulative 25% risk of depression with IFN use in the general HCV-positive population.30 Risk factors for IFN-induced depression include female gender, history of major depression or other psychiatric disorder, low educational level, and presence of baseline subthreshold depressive symptoms.
Despite initial hesitation at providing IFN treatment to patients with psychiatric disorders, there is evidence that people with psychiatric illness can be safely treated with IFN-based regimens and achieve results similar to nonpsychiatric populations.31 For example, one study found similar rates of treatment completion and sustained virologic response (a marker of treatment response) in patients with major depression compared to nondepressed controls.32 No difference was found in frequency of neuropsychiatric side effects between the groups. Similar results have been found in patients with schizophrenia. Comorbid psychiatric illness, therefore, is no longer an absolute contraindication to IFN treatment for HCV.
Ideally, optimal control of psychiatric symptoms should be attained in all patients prior to starting IFN treatment, and close clinical monitoring during IFN therapy is warranted. There is inadequate evidence to support uniform pre-IFN antidepressant treatment in all patients, although antidepressants are useful for patients with elevated baseline depression or a history of IFN-induced depression.33 The largest amount of evidence supporting the safety and efficacy is for use of selective serotonin reuptake inhibitors (SSRIs), particularly citalopram, escitalopram, sertraline, and paroxetine, although none of these medications has a US Food and Drug Administration approval for this indication.
With the advent of the newer direct-acting antivirals (DAAs), the use of IFN has significantly decreased, although it is still included in some regimens. The DAAs are a relatively new group of medications and there are few published reports about their neuropsychiatric side effects. The information in this section comes from the prescribing information for each medication based on adverse events observed in phase III trials.34–39
Telaprevir is not associated with neuropsychiatric side effects.38 Boceprevir in combination with Peg-IFN is associated with insomnia, irritability, and fatigue at rates comparable to Peg-IFN interferon.35 A combination of elbasvir and grazoprevir (marketed as Zepatier; Merck & Co., Inc., Kenilworth, NJ) was associated with a 1% risk of depression compared with a 2% risk when combined with ribavirin, an older antiviral used for HCV.34 Simeprevir can cause fatigue in 16% to 32% of patients when administered in combination with Peg-IFN.37 Insomnia and fatigue occurred in 34% and 14%, respectively, with ombitsavir-paritaprevir-ritonavir plus dasabuvir (marketed as Viekera Pak; Abbvie Inc., North Chicago, IL) compared with 26% and 8%, respectively, in placebo.39 Finally, fatigue occurred in 13% to 16% of patients treated with ledipasvir and sofosbuvir.36 Suicidal ideation was reported in 1% of patients taking regimens containing sofosbuvir combined with ribavirin or Peg-IFN. Irritability also occurred in 7% to 8% of patients compared with 1% in placebo.36
Drug Interactions Between Direct-Acting Antivirals and Psychotropic Medications
Most DAAs are metabolized by liver enzymes and may influence cytochrome P450 enzymes. Most psychotropic medications are metabolized by cytochrome P450 (CYP450) enzymes. Therefore, co-administration of these agents creates the possibility of drug-drug interactions (DDIs). Antipsychotics, benzodiazepines, and tricyclic antidepressants have a more limited therapeutic range, making them more susceptible to clinically relevant DDIs. In contrast, SSRIs and DAAs have a wide therapeutic dose range and are therefore less susceptible to changes in plasma concentrations that may result from CYP450 induction or inhibition. There are few published reports concerning interactions between DAAs and psychotropic agents, but we will review the limited number of drug interaction studies that have been conducted.
The basic principles of DDI dictate that drugs that influence CYP450 enzymes may affect plasma concentrations of other medications that are substrates of these enzymes. Plasma concentrations will be increased by inhibitors of the enzymes and lowered by inducers. With the exception of lepidasvir, ombitsavir, and sofosbuvir, the DAAs are primarily metabolized by CYP3A4/5.40 Co-administration of a strong CYP3A4 inducer, such as St. John's wort (SJW), therefore, may cause lower plasma concentrations, introducing the possibility of virologic failure. Most psychotropic agents do not strongly inhibit or induce CYP3A4. Although formal interaction studies are scant, co-administration of boceprevir and SJW demonstrated a negligible decrease in boceprevir plasma concentration in one study.37
Several DAAs inhibit CYP3A4, including ritonivir (strong), boceprevir (strong), simeprevir (mild), and grazoprevir (mild). SSRIs undergo variable metabolism by CYP3A4 but have a wide therapeutic range. Escitalopram, when co-administered with boceprevir, simeprevir, or a combination of paritaprevir/ritonavir/ombitasvir plus dasabuvir, did not lead to a significant increase in escitalopram concentrations.40 In contrast, benzodiazepines have a narrow therapeutic range. Agents such as midazolam and diazepam are primarily metabolized by CYP3A4 and therefore are susceptible to DDIs with DAAs. One report found that co-administration with boceprevir led to a significant increase in midazolam concentrations.40 Similar increases in midazolam were also noted when co-administered with simeprevir and grazeprevir.
Boceprevir, a strong CYP3A4 inhibitor, should not be used with midazolam and may interact with diazepam, chlordiazepoxide, zolpidem, and zopiclone. Lorazepam, oxazepam, and temazepam are safer options. Simeprevir is a mild inhibitor of intestinal CYP3A4 and is therefore safer to use with intravenous midazolam. It has also been shown to be safe to administer with escitalopram.40 Due to its minimal impact on CYP3A4, daclatasvir should be relatively safe when co-administered with antipsychotics, antidepressants, and benzodiazepines including midazolam. SJW is theoretically contraindicated based on its CYP3A4 inhibition. Sofosbuvir is metabolized by p-glycoprotein and uridine 5′-diphospho-glucuronosyltransferase and is, therefore, not susceptible to any CYP3A4 drug interactions. Lepidasvir is metabolized by an unclear mechanism but is mainly excreted in the bile unchanged and, therefore, should not be susceptible to CYP450 DDIs. Ritonivir is part of multiple drug regimens, including paritaprevir, ombitsavir, and dasbuvir. It is a strong CYP3A4 inducer and is, therefore, susceptible to DDI with any agent metabolized by this enzyme. Nonetheless, studies have shown that this fixed-dose regimen can be safely administered with duloxetine, escitalopram, zolpidem, alprazolam, or diazepam.40 Ritonivir inhibits CYP2D6, the enzyme responsible for duloxetine metabolism. However, duloxetine plasma concentrations were not altered in drug interaction studies with the fixed-dose combination regimen,40 and this may be due to a balancing effect from ritonivir's induction of CYP1A2. The relatively low dose of ritonavir (100 mg) in this regimen suggests that administration of CYP2D6 substrates are not contraindicated with this regimen.39 Grazoprevir is a weak CYP3A4 inhibitor and thus should not be contraindicated with CYP3A4 substrates. Both grazoprevir and elbasvir should not be administered with the strong CYP3A4 inducer SJW.34 IFN is an indirect immune modulator that is not metabolized by P450 and, therefore, does not have DDIs with psychotropic drugs.
Drug Interactions Between Haart and Psychotropic Medications
Pharmacokinetic DDIs must be considered when prescribing psychotropic agents with antiretrovirals because both classes of medication are metabolized by the CYP450 enzyme system. It is often prudent to start psychotropics at a lower-than-normal starting dose due to the increased brain sensitivity of HIV-positive patients, especially those who are later in their disease course. NNRTIs and PIs are both metabolized by CYP3A4. They also may induce or inhibit CYP3A4. Ritonivir is the most potent inhibitor, and for this reason it is often added in small doses to HAART regimens to boost effect. PIs may affect 3A4-metabolized drugs such as carbamazepine that have a narrow safety index. Triazolobenzodiazepines—midazolam, triazolam and alprazolam—are metabolized by CYP3A4, and when co-administered with PIs can lead to excessive sedation secondary to supratherapeutic concentrations. As a result, these agents are relatively contraindicated in this setting and other benzodiazepines are more prudent choices.
Bupropion is metabolized by the CYP450 enzyme 2B6, which is inhibited by nelfinavir, induced by nevirapine and efavirenz, and both induced and inhibited by ritonivir. Fluoxetine is metabolized by 2D6 and 3A4, which are inhibited by PIs. 3A4 is induced by nevirapine and may therefore decrease fluoxetine levels. Venlafaxine and tricyclic antidepressants are metabolized by 2D6 and 3A4, which may be inhibited by ritonavir or other PIs. All of these potential interactions require close monitoring. Citalopram, in contrast, has been combined with PIs without any negative effect.
Antipsychotics are not strong inducers or inhibitors of HAART. NNRTIs and PIs may induce or inhibit antipsychotic metabolism, leading to either antipsychotic failure or increased risk of antipsychotic side effects. Pimozide and thioridizine are substrates of 3A4 and contraindicated with these antiretrovirals due to their risk for cardiac effects at supratherapeutic concentrations. Pimozide has a narrow safety index and can cause arrhythmias and seizures when administered with PIs. Aripiprazole, quetiapine, and risperidone are 3A4 substrates as well. Increased adverse effects have been reported when these agents are co-administered with PIs.41,42 An additional consideration is the risk for QT prolongation. Saquinivir and lopinavir/ritonavir are associated with prolonged QT intervals and should be avoided with antipsychotics. In contrast, olanzapine and clozapine levels decrease with ritonavir due to CYP1A2 induction.41 HIV-positive patients may have a predilection for extrapyramidal symptoms, so first-generation antipsychotics should be avoided. Lipodystrophy, which may present as fat accumulation in the abdomen, has been associated with NRTIs and PIs. PIs are also associated with hyperglycemia. Second-generation antipsychotics such as olanzapine and clozapine increase the risk for weight gain and diabetes. Concomitant use of these medications warrants a discussion of risks and consideration of switching agents.
The NRTIs and the fusion inhibitor enfuvirtide have no CYP450 interactions with psychotropic agents. The entry inhibitor mariviroc and the integrase inhibitor raltegrevir have a low risk of DDIs. As it is impractical to review every possible DDI involved with HAART, we recommend using resources available online.43