Defining "refractory" in schizophrenia can be difficult (3). Initial definitions of treatment response focused primarily on the alleviation of positive symptoms. More recently, greater emphasis has been placed on negative symptoms and cognitive symptoms. Although often difficult to quantify, negative and cognitive symptoms are believed to account for greater morbidity and be more influential on overall prognosis. Schizophrenia in at least 30% of patients can be considered treatment refractory by rigorous methods of determining refractoriness, with some estimates being even higher, toward 50%—60% (4, 5). Treatment refractoriness has been defined by the International Psychopharmacology Algorithm Project (IPAP) as 1) no period of good functioning in the previous 5 years, 2) prior nonresponse to at least two antipsychotic drugs of two different chemical classes for at least 4—6 weeks each at doses ≥400-mg equivalents of chlorpromazine or 5 mg/day risperidone, and 3) moderate to severe psychopathology, especially positive symptoms: conceptual disorganization, suspiciousness, delusions, or hallucinatory behavior (6). The above criteria focus on positive symptoms, but the IPAP also considers continued negative or cognitive symptoms, violence, suicidality, and recurrent mood symptoms as elements of treatment refractoriness.
The ultimate goal of treatment for any illness is not just the remission of symptoms but a full functional recovery from that illness, and schizophrenia need be no exception. Conceptualizing recovery in schizophrenia requires considering social and occupational domains in addition to clinical symptoms such that improvement is necessary in both areas to achieve a functional remission and increased autonomy. These functional gains should be thought to maintain stability in similar terms, i.e., time frame and level of improvement, just as one would consider lack of hallucinations for a clinical remission only significant if it was a consistent finding over the course of several months or years (7). Treating patients with the recovery model in mind generally requires a collaborative approach including occupational therapy, job training, and community integration. Individuals with schizophrenia should be treated with optimism, and attempts to decrease internalized stigma are also crucial to helping them set goals (8).
The Texas Medication Algorithm Project (TMAP) provides a structure for organizing treatment strategies for several mental illnesses, including schizophrenia (9). In the most recent update to the TMAP algorithm, the authors identified four key recommendations. The first is to dose medications differently in the first episode of treatment compared with chronic treatment, as first-episode patients are more sensitive to the side effects of antidopaminergic medications. Second, although second-generation antipsychotic (SGA) medications are considered first-line, first-generation antipsychotic medications are a reasonable second-line option instead of earlier recommendations suggesting that they only be used after two failed SGA trials. The third recommendation was to consider clozapine treatment after two antipsychotic medications fail to lead to sufficient symptom improvement, instead of after three treatment failures as in previous guidelines. The final recommendation states that in cases where the response from clozapine is insufficient, consideration should be made to augment treatment with another antipsychotic drug or with brain stimulation (ECT or repetitive transcranial magnetic stimulation (rTMS)) (10).
Persistent negative symptoms yield the greatest functional disruption for patients with schizophrenia. Although negative symptoms remain recalcitrant to most pharmacological treatment, research has shown that increased participation in the community, e.g., through various forms of rehabilitation such as social skills training, can improve these symptoms (11). In addition, psychotherapeutic interventions, such as with body-oriented psychotherapy, have demonstrated modest benefits in improving negative symptoms (12). Despite these findings, there remains a gulf between treatment success and an overall armamentarium for working on these problems.
Although there are no definitive treatments for negative symptoms, positive symptoms of schizophrenia generally respond well to various approaches including psychopharmacological, electrotherapeutic, and psychotherapeutic. Many patients benefit from these interventions; however, there are those who continue to experience hallucinations persistently. Even more difficult to target therapeutically are delusions, which often remain with patients indefinitely. There are patients who do not have a remission of positive symptoms, even with clozapine, which is generally considered the antipsychotic drug of choice for patients with refractory psychotic symptoms.
Regardless of how one conceptualizes recovery or remission from schizophrenia, many individuals who have this illness continue to experience significant morbidity despite treatment. This article will focus on treatment options for those who continue to have needs that are not met by initial antipsychotic treatments, including data on polypharmacy, brain stimulation, psychotherapy, and new medications that are in various stages of development.
Generally, 30%—40% of patients do not respond sufficiently to antipsychotic treatment and continue to have residual positive symptoms. In a community sample of individuals with schizophrenia or schizoaffective disorder, 42.5% of patients were receiving at least two antipsychotic medications, and 70% received an adjunctive psychotropic medication to go with the antipsychotic. Only 25% of patients in the sample were receiving monotherapy with an antipsychotic (13). In a study from 2002, a community population was shown to have a rate of concomitant pharmacotherapy of more than 50% (14). Patients who are able to adhere to a medication regimen at an adequate dosage are more likely to benefit from treatment, although strategies to enhance treatment adherence are often difficult (15).
The Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) and Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia Study (CUtLASS) were able to also provide prospective measurements of treatment effectiveness (16, 17). These studies assessed the outcomes of SGA medications in comparison with first-generation antipsychotics (FGA). In CATIE, the time to discontinuation of medication was the primary outcome measure. The study found that three-fourths of patients switched or discontinued their initial study medication before the conclusion of the study at 18 months' duration. This finding indicates that the initial choice of antipsychotic is difficult, and many patients are unsatisfied with their initial medication. Most discontinuations for lack of efficacy came within the first 6 months of treatment. Subjects who were randomly switched to a new medication, however, had an earlier time to discontinuation than those who were randomly assigned to their current medication (18, 19). CUtLASS focused on quality of life for a primary outcome measure. The results demonstrated that there was a similar effect on quality of life for patients switched to either FGAs or SGAs (17).
Switching antipsychotic medications is an essential part of treating patients with schizophrenia, although there are few data to guide those decisions. Switching becomes especially difficult in situations in which there is moderate effectiveness from the medication on core clinical symptoms, but side effects are intolerable or unacceptable. Further clinical effectiveness trials are needed to better appreciate the complexity of changing medication and guide decision making in this area.
Schizophrenia is an illness that presents with significant variability among individual patients and thus the search for prognostic factors is an ongoing research endeavor. Factors that can be seen early in the illness may yield long-term clues for overall success in treatment (20). Several factors that are important in determining the future course for an individual patient include demographic factors such as age of onset and gender, method of disease onset, and quality and variability of symptoms. The duration of untreated psychosis and initial response to treatment portend future illness trajectory. For some factors, such as substance abuse, it is unclear whether the association may correlate with a better or worse prognosis. The way that individuals understand their illness and the role it plays in their lives also influence outcome because internalized stigma can hinder individuals as they develop goals and difficulties with adherence to a treatment regimen hamper effectiveness. As the number of potential prognostic markers increases, so does the search for biomarkers to help allow earlier determination of proximal factors and ways to measure the effects of disease-modifying treatments.
The onset of illness presents with many variables that can influence the prognosis (21). More severe symptoms at the time of initial diagnosis are associated with worsened future course of illness. The mode of onset and the duration of untreated psychosis (DUP) correlate with outcome, with a longer DUP and insidious onset portending a more severe course (22, 23). Although the mechanism is not fully understood, there are plastic changes in the brain that occur the longer a person has unopposed psychosis (24). Patients with faster recovery times in the initial episode of psychosis had better treatment outcomes over the first 2 years of illness. Likewise, a slow and insidious initial progression to schizophrenia heralds a worsened prognosis for further illness than does a rapid initial episode (25).
There may be genetic factors to predict treatment refractoriness. For example, researchers are looking at the CNR1 gene, which encodes the CB1 cannabinoid receptor. Cannabis usage, particularly in high doses, is known to be a risk factor for schizophrenia (26). Use of cannabis has been shown to worsen positive and negative symptoms and decrease cognitive performance. CNR1 is located on chromosome 6.14—6.15, a putative region of association with schizophrenia, and several single nucleotide polymorphisms have been associated with increased risk of disorganized schizophrenia. In a study of 59 patients with treatment-refractory schizophrenia, Hamdani et al. (27) looked at the 1359 G/A polymorphism and its role in treatment refractoriness measured by changes in Positive and Negative Syndrome Scale (PANSS) scores. Those with the G allele were more likely to have illness refractory to antipsychotic treatments (up to 76% for those with GG genotype), whereas those with the AA genotype had a refractoriness rate of 5%.
There are several other correlates of treatment response, some of which change over the course of the illness. Adherence to treatment can be a difficult matter for both clinicians and patients. Poor adherence to medication is associated with worsened long-term functional outcome, increased hospitalizations, more suicide attempts, and decreased life satisfaction (28, 29). Adherence to treatment is also influenced by stigma, and thus the treatment course can be influenced by stigma as well. Internalized stigma can lead individuals to be reticent to seek treatment, to feel worse about their prognosis, and to feel that they have less self-worth. These delays can lead to a longer DUP and a more difficult illness to treat at initial presentation (30). Likewise, targeting these feelings with psychotherapy can further enhance treatment and functional outcomes (31).
Differing symptom patterns also lead to differences in treatment effectiveness. Positive symptoms are generally more responsive to medication and, thus, if these are the leading cause of disability, remission is more likely. Prominent negative symptoms, however, yield outcomes that are more difficult to treat. Patients with negative symptoms were less likely to be able to demonstrate the typical improvements in social functioning that came with sustained employment (32).
Deficits in cognition also lead to worse functional outcome. In a study of 78 patients with schizophrenia, those with poorer performance on neuropsychological testing were more likely to have decreased functional capacity as measured by the UCSD Performance-Based Skills Assessment (33). The presence of a deficit in processing speed has been shown to indicate a poorer prognosis for overall outcome on the Disability Assessment Schedule and may be a mediator that can explain overall worse performance on other neuropsychological tests (34). Deficits in processing speed and working memory also indicate a worse prognosis for social cognition (35).
In addition to deficits in cognition predicting worse outcome, higher levels of negative symptoms have long been linked with poorer prognosis (36). A recent study from Israel showed that when premorbid symptoms were taken into account, the greatest predictor of future outcome was from negative symptoms (37). A similar finding from New York prospectively showed that in first-episode patients, those with greater negative symptoms had a more severe course 2 years after the initial hospitalization (38).
Substance use has provided a conundrum for determining future risk. Although cannabis use has been associated with earlier onset of illness and greater likelihood to convert from a putatively prodromal phase of illness to a more severe diagnosis (39), drug use itself has only equivocally been associated with outcome. Those who are able to procure illicit drugs may have different cognitive and social abilities than those who are unable to do so and thus may provide a confound that limits the predictive value of drug use (40). Despite that confound, chronic substance abuse, particularly with stimulants, can cause persistent psychotic symptoms and interfere with treatment adherence. This behavior can worsen prognosis and interfere with clinical understanding of how the patient is progressing.
Other Axis I illnesses are known to complicate the treatment and prognosis of schizophrenia. Obsessive-compulsive symptoms are a common comorbidity with schizophrenia and are seen in almost 15% of first-episode patients (41). The numbers increase in individuals who have prolonged exposure to antipsychotics. Comorbid obsessive-compulsive symptoms are associated with poorer performance in tests of executive function. However, it is unclear whether obsessive-compulsive symptoms are associated with increased negative symptoms or a worse prognosis in the initial presentation of schizophrenia (42).
Depression is a common comorbidity in schizophrenia with nearly half of patients having affective symptoms. The coexistence of these symptoms can cause additional diagnostic difficulty. Differentiating between negative symptoms and depression requires keen clinical observation in the presence of multiple overlapping features for both. Depressive symptoms can occur spontaneously or can be due to multiple factors, including medication side effects. Early evidence pointed to a protective effect of comorbid affective symptoms; however, that feeling has been disproven with comorbid depression. Individuals with prominent depressive symptoms have a greater risk of chronic symptoms, increased risk of hospitalization, and greater overall symptom burden (43).
It is not currently known why certain medications are more effective for a particular patient than other medications. Predicting which patients will have a disorder that is refractory to conventional treatment is not yet possible; thus, the search for specific biomarkers to identify factors that may help to enhance prognostication is ongoing. Some initial studies have begun to attempt to describe differences in those whose disorders are refractory to treatment compared with those who have a better response. These findings from small studies include biomarkers such as short telomere length (44), corneal temperature (45), and density of dopaminergic receptors (46). Many more biomarkers are currently under evaluation to help with predicting various phases of the disorder.
Imaging has also been used to provide some further insights for the biological bases of treatment response (47). Results have been shown with many different imaging modalities, both functional and structural. Several studies have shown that there are ventricular differences, including larger and more asymmetrical shape, in patients who have worse clinical outcome. In addition, cerebellar and temporal cortical abnormalities have been correlated with worsened severity of social and positive symptoms, respectively. Decreased gray matter has also been seen in chronically hospitalized patients and those with other markers of poor outcome compared with healthy control subjects, but initial studies did not show statistically significant differences from other patients with schizophrenia (47).
More recent findings have implicated more posterior cortical regions for gray matter changes in patients with a more refractory course. In addition, positron emission tomography imaging has shown that there are functional changes that are worsened posteriorly compared with those who have a more treatable course of illness. For those with this finding, there is decreased metabolic performance in the cingulate, striatum, and the temporal lobe. Deficits in white matter have also shown a similar posteriorized pattern in patients with poor outcomes. In addition, decreased interhemispheric connectivity has been shown, despite patients with treatment-resistant disorders having a larger splenium in the corpus callosum. Given the overall decreased intracranial volume seen in schizophrenia, the increased size of the splenium was surprising and provides further support for neurodevelopmental theories of schizophrenia (48).
Many individuals with schizophrenia respond well to antipsychotic medication, although there are many patients for whom initial treatment is insufficient. For patients who do not respond to initial treatments, there are several strategies to guide further treatment. Treatment algorithms, such as the TMAP (9) and the IPAP (6), provide guidelines that can help to identify initial tactics that one can try after the first approaches are unsuccessful. Clozapine, which will be discussed in further detail, remains the primary medication for treatment-resistant schizophrenia. Data on other existing medications, sometimes at higher doses than typically used or in combination polypharmacy, continue to be described in the literature. There is also an increasing evidence base for new adjunctive medications to be used in conjunction with antipsychotics. Nonpharmacological treatments including brain stimulation and different forms of psychotherapy also will be discussed both on their own and in conjunction with pharmacotherapy.
Clozapine has been the treatment of choice in treatment-refractory schizophrenia for more than 20 years (49, 50). Clozapine was first discovered decades earlier; however, limited understanding of its risks and complications initially limited its use. In the United States, it was only available on a compassionate use basis for those with severe tardive dyskinesia, significant treatment resistance, or sensitivity to extrapyramidal side effects (51). Currently, clozapine is generally considered only for use in patients with treatment-refractory schizophrenia, although some would argue that it should be considered for use earlier in the course of illness, particularly for those who do not immediately respond to a second-generation antipsychotic (52).
The rationale for reserving clozapine for patients with refractory disease is generally based on its more serious side effect profile. Although management strategies are available for many of these side effects; nonetheless, other medications are more easily tolerated. All patients prescribed clozapine must be entered into a national registry for monitoring. Adherence to the prescribed regimen is also essential, because missing the medication for as few as 2 days requires retitration (53). The most serious risk from using clozapine is for agranulocytosis. Other black box warnings include seizures, myocarditis, hypotension, and the ubiquitous warning for all antipsychotics of the increased risk of mortality in elderly individuals. Pulmonary embolism and bowel perforation are also dangerous risks for patients on clozapine. Less serious, but nevertheless difficult, side effects include sialorrhea and fatigue.
The rates of agranulocytosis have dropped significantly since aggressive hematological monitoring and slow-dose titrations have been part of the practice of prescribing clozapine (51). Full dosing and monitoring instructions are available in the package insert and through the clozapine registry. The greatest risk of agranulocytosis is in the first 3 months of treatment. In an analysis of more than 11,000 patients from 1990—1991, the first years of the widespread use of clozapine, it was shown that just less than 1% of patients developed agranulocytosis. This finding was consistent with previous reports of agranulocytosis rates of approximately 1%—2%. In that cohort, there were two patient deaths attributed to hematological adverse events (54). Genetic markers of risk continue to be an ongoing area of research, after it was shown that in a small group of Ashkenazi Jews, there was an association between different HLA haplotypes and development of agranulocytosis (55).
There are strategies to try to manage decreased blood counts, although agranulocytosis remains a reason that patients usually have to stop taking clozapine. When agranulocytosis or another drop in blood counts occurs, a hematologist should be consulted to discuss possible options and management. In some cases, particularly for benign ethnic neutropenia, lithium treatment is able to increase white blood cell production. Although it is insufficient to use as treatment for agranulocytosis and can even potentially mask a developing case of agranulocytosis (56), the ability to increase white blood cell count is usually enhanced with lithium (57). In more potentially serious cases, granulocyte colony-stimulating factor may be considered (58). Although not a definitive treatment, it can be an effective tool for maintaining clozapine in a patient rather than necessitating discontinuation, and several cases of successful use are reported in the literature, as are some unsuccessful cases (59, 60).
Metabolic side effects continue to be difficult to manage. Weight gain, dyslipidemia, and risk for diabetes are all seen frequently with clozapine use (61). It is important to encourage active participation from the patient's primary care provider with respect to managing many of these metabolic effects, particularly if clozapine is uniquely suited to ameliorating their psychotic symptoms. Fortunately, treatments are available for many of the metabolic side effects, although it would still be preferable to have medications that were effective in treating psychosis without these risks. Long-term consequences of metabolic side effects can affect overall life expectancy if not properly managed.
There are some patients for whom clozapine is not an option or for whom it is ineffective. In that case, the art of psychopharmacology comes into play, although there are some data to guide treatments with other agents. Initial psychopharmacological strategies at this point include consideration of either polypharmacy with other antipsychotics or with non-antipsychotic medications. Other possibilities are to use a higher dose of a single antipsychotic. Looking at antipsychotic polypharmacy, there are reports of several different combinations, both with and without clozapine.
Given that clozapine is typically used in treatment-resistant schizophrenia, combinations of medications that include clozapine are often tried if clozapine monotherapy is unsuccessful. Sometimes the addition of another agent can help to lower the dose of clozapine and potentially help with lowering the side effect burden. Most reports describe combining clozapine with aripiprazole. In a double-blind study, the addition of aripiprazole to clozapine was not shown to have benefits on the primary outcome measure of the Brief Psychiatric Rating Scale but did show benefits of secondary measures including the Schedule for the Assessment of Negative Symptoms (SANS). No differences were seen in glucose levels, although prolactin was lower in the aripiprazole group compared with placebo (62). In a small (N=24), retrospective chart review, it was shown that addition of aripiprazole was associated with lower clozapine dose and that 18 of 24 patients lost weight, with an average loss of 5 kg (63). Although case reports have suggested benefits from aripiprazole, further studies need to be done to establish its efficacy as an adjunctive agent to clozapine. Risperidone has also shown only limited benefit when added to clozapine (64).
If combining antipsychotic medications is not possible or ineffective, there are several agents worthy of consideration as other augmenting agents. Although none of these are specifically approved for use in schizophrenia, there are various amounts of data to support their use. Several reports recently have indicated benefit for adding lamotrigine to antipsychotic treatment. Lamotrigine decreases glutamate release and increases γ-aminobutyric acid (GABA) release and, when given with clozapine, has shown enhanced diminution of locomotor activity in rats treated with phencyclidine, a common animal measure of potential antipsychotic efficacy. However, several small trials have failed to demonstrate a statistically significant result in the primary outcome measures. In a recent meta-analysis, the data from these small trials was aggregated for enhanced power to detect a difference between placebo augmentation and lamotrigine augmentation in patients who still had residual symptoms despite treatment with clozapine (65—67). With the use of rigorous meta-analysis methodology, it was shown that there was a potential benefit with the addition of lamotrigine in these patients. Benefit was only shown for lamotrigine as an augmenting agent to clozapine, leading to speculation that the glutamatergic properties of lamotrigine may be uniquely suited to working with clozapine. The number needed to treat for this study was four, meaning that a benefit on the PANSS score was seen in one of every four patients treated. Although longer trials are still needed to confirm the effect, lamotrigine may be worth considering in those with schizophrenia that is most refractory to antipsychotic medication (68).
In addition to lamotrigine, other agents have been studied in conjunction with antipsychotics for treatment resistant schizophrenia. Topiramate, similar to lamotrigine, also affects glutamatergic transmission. In a small crossover trial of 36 patients receiving various atypical antipsychotics, a positive effect was seen for topiramate on the PANSS score. Further work needs to be done with topiramate to fully establish efficacy (69). Valproate has also never been established as an effective adjunctive agent in long-term treatment of schizophrenia, although benefits in short-term potentiation of treatment have been seen in one study (70). Modafinil also showed a limited effect in treating refractory negative symptoms. Although it did not reach statistical significance on the primary outcome measure of the SANS, it was well tolerated without positive symptom exacerbation (71).
Antidepressants have long been considered possible augmenting agents, particularly for targeting negative symptoms. Some initial studies were promising as well, However, although a benefit may be seen in targeting depressive symptoms, which are a frequent comorbidity of schizophrenia, current evidence does not show a benefit of treating negative symptoms with an SSRI (72).
Benzodiazepines are also frequently used as adjunctive agents in schizophrenia. In acute situations, the sedative effects of benzodiazepines can be useful. However, there is limited evidence for overall symptom improvement with benzodiazepines as long-term adjunctive agents. Further clinical trials are needed to fully appreciate the impact of benzodiazepines, but, according to a Cochrane review, there is insufficient evidence to support their long-term use (73).
Using a higher dose of an antipsychotic is a reasonable strategy for management in some patients. The mechanism for such treatment is often unclear, because dopamine receptors are typically saturated at conventional doses, but individuality in treatment response can be difficult to predict. Many case reports are available for successful high-dose treatment with different antipsychotic medications. In a small (N=40) double-blind study comparing high-dose olanzapine with clozapine, it was shown both groups showed similar and robust improvement on the PANSS. Of the subjects taking olanzapine, 71% were taking more than 35 mg/day. Metabolic complications were prominent in both groups, although the subjects taking olanzapine gained considerably more weight (74). In a double-blind trial of high-dose olanzapine (maximum dose 30 mg) versus clozapine in young, treatment-resistant subjects, clozapine demonstrated an enhanced effect (75). In an open-label follow-up, clozapine was beneficial in 7 of 10 patients who did not have an adequate response to olanzapine. Both groups, as is often seen in young people, were exquisitely sensitive to metabolic complications, showing greater weight gain and changes in glucose than is typically seen in older patients (76).
Adherence to a medication regimen is essential for successful pharmacotherapy. Patients who are unable to maintain their medication regimen are at a greater risk of relapse, hospitalization, and poor functional outcomes (28). The more times that a patient discontinues his or her medication, the more difficult treatment becomes in the future and the greater the chance for becoming treatment refractory. Many patients have difficulty in organizing and managing a daily medication regimen. Although simplifying the regimen to using daily instead of multiple daily doses is a start, for many patients, a depot medication is the most parsimonious treatment. There are currently depot preparations for three second-generation antipsychotics: risperidone (77, 78), paliperidone (79, 80), and olanzapine (81, 82). These join FGAs haloperidol and fluphenazine as appropriate choices for maintenance therapy. In all cases, patients must be able to demonstrate that they can tolerate an oral dose of medication before starting the long-acting injectable preparation. The switch from oral to depot medication is generally well tolerated (83).
Increasingly, nonpharmacological treatments are gaining attention and compelling data for their use are being generated. Most notably, brain stimulation methods are showing promise for treatment for many patients. These methods, including ECT and especially rTMS, are becoming safer and easier to administer. Although ECT has been used to treat schizophrenia for decades, there is limited data to support its use (84). A Cochrane systematic review recently concluded that ECT could be effective in those patients who do not respond to treatment with antipsychotic medications, particularly those who need especially rapid treatment. ECT has been shown to be better than sham treatment, although benefits from ECT have not been shown to persist after treatment is concluded, necessitating maintenance treatment for continuation of benefits. The number needed to treat for ECT was six when all the data were compiled for the Cochrane review (85).
Transcranial magnetic stimulation was recently approved for treatment of major depressive disorder, but currently does not have an indication for treatment of schizophrenia. Unlike ECT, there is greater ability to specify treatment location with rTMS. Transcranial magnetic stimulation is also more easily tolerated and only rarely induces a seizure when properly administered. Treatment for depression typically involves rapid transcranial magnetic stimulation, which serves to enhance excitability in the brain. Schizophrenia treatments have been both rapid and slow wave. Treatments targeting auditory hallucinations have largely been focused on the temporoparietal cortex, whereas treatments targeting negative symptoms have largely used rapid-wave rTMS aimed to the prefrontal cortex (PFC). As in depression, the PFC has been an initial target of the magnet. Initial studies were small and often confounded by difficulties with blinding the sham treatment (86). In a meta-analysis of several small studies, rTMS was shown to be beneficial in alleviating treatment-resistant auditory hallucinations, although the benefit was not seen in overall measures of positive symptoms. In general, the safety profile of rTMS was reassuring, and treatment was not limited by complications (87). Decreased auditory hallucinations were also found in a recent randomized, sham-controlled trial of rTMS in the bilateral temporoparietal area. This study showed the same improvement on auditory hallucinations but not general improvements in psychopathology (88).
Transcranial magnetic stimulation has been studied for treatment of negative symptoms, although with limited success (89). In a meta-analysis, rTMS was seen to have a benefit on negative symptoms in initial, uncontrolled studies, but the effect diminished for controlled trials. Further work needs to be done to more thoroughly study the potential for this therapy to improve otherwise recalcitrant symptoms. Indeed, researchers will need to focus on stimulating other areas besides the PFC because negative symptoms may arise from other parts of the brain, including the anterior cingulate, cerebellum, occipital cortex, and posterior cortical parietal cortex (90).
Multimodal treatment with medication and psychotherapy is the standard of care in most other psychiatric disorders, and schizophrenia should be no exception (91). Several models of psychotherapy continue to demonstrate efficacy for the treatment of core schizophrenia symptoms. Cognitive behavioral therapy (CBT) is the most thoroughly studied treatment modality, and there are data to support its usage in several domains of illness including both positive and negative symptoms (92). Beck (93) first reported a case in 1952 of a patient who demonstrated enhanced insight and functional improvement through CBT. In addition, CBT can be helpful in enhancing patient adherence to the overall treatment regimen and patient insight. A recent post hoc analysis demonstrated a benefit of CBT on suicidal ideation as well (94). Although the benefits may wane after treatment stops, there are profound improvements in the short-term, and some benefits such as with insight, can be seen even 1 year after treatment (95). The benefits of CBT are more likely to persist than other more benign comparison interventions, such as "befriending" or general supportive care, in which subjects are given the same social benefits of having a therapist to see, but without a specific or targeted intervention (92).
In a study looking at insight, subjects were randomly assigned to CBT or treatment as usual. The CBT group was able to better relabel psychotic symptoms and to understand their need for treatment. There was no difference between the groups with respect to insight into having an illness (95). Although improvements can be seen with respect to insight, delusions remain difficult to address. However, general psychopathology has been shown to improve, as measured by PANSS score, even in patients who are refractory to clozapine treatment. An initial benefit was seen in quality of life as well, although that did not persist at a 6-month follow-up. General psychopathology improvements were maintained after 6 months, however, in comparison with a befriending comparator group (96).
The inclusion of psychosocial treatments in comprehensive care of the person with schizophrenia is important. According to recently updated recommendations by the Patient Outcomes Research Team, many psychosocial treatments have evidence for benefit. Among these treatments are assertive community treatment (discussed below), supported employment, skills training, CBT, token economy interventions, and family-based treatments. Supported employment programs have been shown to lead to increased wages and increased time spent in competitive employment for individuals with schizophrenia. Although the long-term benefits of supported employment are unclear, for many patients the benefits are enduring. Skills training refers to teaching people about ways to manage living independently. Studies have shown that these skills are often well incorporated and last long after the intervention is completed. It is unclear whether improvement in these skills has an overall impact on symptomatology; however, it does seem to have an impact on overall quality of life (97).
In addition to the above treatments, harnessing family members to help with treatment is valuable. This fact has been known for a long time and remains a crucial aspect for those individuals who maintain a relationship with family. Family interventions can also help caregivers reduce burnout and provide methods for coping with the stresses that can occur in taking care of a family member with schizophrenia. It has been shown that family interventions can reduce hospitalization rates, improve adherence to the medication regimen, and lead to higher levels of employment. These benefits can be seen in an intervention that lasts as little as 6—9 months (98).
The treatment for those with disease refractory to initial medication trials is complex and challenging. Many patients require close monitoring, and a team approach is often essential. In the most severe cases, an assertive community treatment (ACT) approach can be the most effective. In an ACT model of care, patients are closely followed in the community by social workers and are seen frequently with an emphasis on helping to manage their needs in their own community without the need for long-term hospital placement. Because providers are able to see patients in this manner, crises can be more effectively averted before escalating and causing a rupture in the treatment course. In the recently published results of the ACCESS trial, Lambert et al. (99) demonstrated that compared with regular treatment in the community, the assertive model significantly helped to keep patients engaged and in treatment over a 1-year period. ACT patients were five times more likely to continue taking their medication over the period of the study. More of the ACT patients were able to live independently and to have a job. Further research to determine the minimum effective length of treatment in the ACT model needs to be done to better know how to deliver this type of intervention to the most people.
The treatment of schizophrenia has remained relatively unchanged over the past 50 years. Although newer medications are available, all still focus primarily on treatment at the dopamine D2 receptor (100). The current armamentarium of medications does provide improvement in the positive symptoms of schizophrenia for many patients. Unfortunately, a great number of patients still have residual psychotic symptoms, and greater numbers have persistent negative and cognitive symptoms. New medications are being developed to focus on targets other than the dopaminergic system. Many of these targets have been chosen to help enhance cognition and negative symptoms. Work toward developing personalized medicine through pharmacogenetics has also been evolving, with a goal to help in determining a rational medication choice for patients based on their genetic makeup. These treatments are still years away, although some advances have been made recently.
Although there still is no unified mechanism to explain schizophrenia, progress is being made in understanding signaling pathways and how these pathways influence pathology. Many of these present potential targets for new clinical interventions (101). Among the most promising targets currently under investigation include the GABAergic, glutamatergic, nicotinic, and serotonergic systems (102, 103). Many other systems have candidate drugs as well, including other dopamine receptors besides D2 (104).
GABA activity in the dorsolateral prefrontal cortex (DLPFC) has been shown to correlate with abilities in working memory and cognition. In schizophrenia, a deficit in GABA has been seen in postmortem studies and in imaging. This has been demonstrated by deficits in GABA neurons that express parvalbumin. In addition, levels of GABA are controlled by the enzyme GAD67 (67-kDa isoform of the glutamic acid decarboxylase), and this enzyme has been shown to be deficient in the DLPFC of schizophrenic patients. MK-0777 is an allosteric modulator of the GABAA receptor that leads to increased GABA expression and transmission. Specifically, it increases GABA neurotransmission from chandelier neurons, leading to better neuronal synchronization across the DLPFC (105). In a double-blind, placebo-controlled trial of MK-0777, schizophrenic patients taking stable doses of antipsychotic medications were shown to have positive effects (although the study was underpowered for statistical significance) on three working memory tasks (106). There are currently ongoing studies of MK-0777 for cognition in schizophrenia (107).
Additional work has also been done at the GABAB receptor. In the search to regulate nitric oxide (NO) increases that are seen with GABAergic dysfunction, baclofen has been looked at in a rodent model for its effect of reversing the N-methyl-d-aspartate (NMDA) antagonist model of psychosis. It has been shown that NMDA antagonists disrupt the NO pathway in addition to disrupting the NMDA channels. Baclofen had been shown previously to diminish the effects on prepulse inhibition in rodents given MK-801 (an NMDA antagonist.) In a recent study, baclofen was shown to prevent the prepulse inhibition changes seen with phencyclidine (PCP), particularly when injected prefrontally in rats and when combined with a nitric oxide synthase inhibitor. These results provide further evidence that in the NMDA antagonist model of psychosis, NO may be a downstream mediator in the PFC and involved with information processes mediated by glutamate and GABAB (108).
Just as the GABAergic system has received attention for its role in cognition in schizophrenia, the glutamatergic system continues to be an area of active research. Agonism at the metabotropic glutamate receptor (mGluR) was first shown more than 10 years ago to reverse PCP effects in the rat, and, with that discovery, the rationale was there for further work on antipsychotic medications using this target (109). Eli Lilly and Company developed a compound that was tested and initially showed a benefit in both positive and negative symptoms of schizophrenia (110), although these findings are yet to be replicated. Further work has now shifted toward other ways of interacting with the mGluR, including more allosteric modulations (111). Although the type II mGluRs have garnered the most attention, other mGluRs are also increasingly the focus of research, including mGluR5 and mGluR8 (112).
The metabotropic glutamate receptor 5 is a current focus of innovation. The mGluR5 acts closely with the NMDA receptor and may play a role in its regulation, particularly in the forebrain. mGluR5 has been studied in rats, and those rats with a knockout of the mGluR5 show deficits in prepulse inhibition in a manner similar to that in rats that have had PCP administered. Likewise, rats pretreated with 2-methyl-6-(phenylethyl)-pyridine (a selective mGluR5 agonist) do not show the cognitive and psychotic-like changes typically seen with PCP administration (113).
It remains difficult to create medications that offer an adequate safety profile and selectivity for the specific metabotropic glutamate receptors. As such, a number of positive allosteric modulators (PAMs) are being developed to target the metabotropic glutamate system but without the selectivity otherwise needed for targeting specific receptors (114). Allosteric modulators also protect against the inherent problems of sensitization that are typically seen with agonist molecules. The PAMs bind at the seven-membrane loops of the G protein-coupled receptor, whereas glutamate binds to the N terminus of the receptor. When bound, the PAM potentiates the response for the mGluR5 to glutamate when it binds in the glutamate receptor. Whereas mGluR5 PAMs have been found and demonstrate excellent selectivity for the allosteric site, none has thus far had adequate solubility or pharmacokinetics to be suitable for a potential treatment. The most recent PAM discovered, 3-cyano-N-(1,3-diphenyl-1H-pyrazol-5-yl)benzamide, has been able to be tested in vivo in rats and has demonstrated benefits in blocking psychotic behaviors from rats administered amphetamine. Further work is needed on mGluR5 PAMs to improve their pharmacokinetics such that they could ultimately be medications that could work in humans. Most recently, ADX47273 has been developed and is showing initial promise in rodents and may be studied ultimately in humans (115). Additional promise has been shown in animal trials with an mGluR2 allosteric modulator (116).
Additional work on the NMDA receptor has focused on using glycine to provide a benefit in the treatment of negative symptoms. Although results for giving glycine, d-serine, or d-alanine to patients have ultimately been mixed and fraught with methodological challenges, there remains promise that inhibition of the glycine transporter may provide some of the same benefits that have been seen with glycine itself. The glycine transporter modulates the amount of glycine available to the NMDA receptor and, thus, when blocked may provide a better glycine reserve for the receptor than merely providing a direct glycinergic agonist (117). Sarcosine, a glycine transporter antagonist, has been seen in a small trial to have benefits as monotherapy on both positive and negative symptoms, although further work needs to be done with that compound (118). Previous trials of glycine transport inhibitors have primarily been conducted as add-on trials for patients already taking antipsychotic medication (119). Current work with sarcosine both as an adjunctive agent and as monotherapy is ongoing. There are also studies in the ultra high-risk population to evaluate its use in preventing switching into psychosis (120).
The α-7 nicotinic receptor continues to be an active area for research. It has been shown that changes in the region of chromosome 15 that codes for this receptor, CHRNA7, leads to decreased expression of α-7 perhaps because of single nucleotide changes at the promoter. Several α-7 agonist and partial agonist molecules are currently under investigation. Many similar molecules had been limited previously due to difficulties in developing molecules that are readily transformable into pharmaceuticals. In an early phase-2 trial of 3-(2,4-dimethoxybenzylidene) anabaseine, an α-7 agonist, it was shown that negative symptoms, as measured by the SANS, improved significantly. Although the initial goal was to enhance cognitive symptoms, which did not occur, enhancement of negative symptoms was an encouraging finding (121).
As the search continues for novel pharmacological treatments, continued emphasis is being placed on setting the target for treatment at full recovery. This high standard is essential for helping patients achieve their goals and have greater quality of life. Much work is also being done on psychosocial treatment, including cognitive remediation and other nonpharmacological approaches to target overall recovery from symptoms and integration into the community. Ultimately, treatment-resistant or refractory schizophrenia will need to be defined not just as refractory positive symptoms, but also by the overall reduction in quality of life that many with the illness currently endure. As the field progresses, these deficits will increasingly be able to be addressed and must continue to be a focus of the overall well-being and treatment of patients.