KarXT for schizophrenia–effectiveness and value: A summary from the Institute for Clinical and Economic Review’s New England Comparative Effectiveness Public Advisory Council

Schizophrenia is a serious mental illness that affects about 3.9 million people in the United States and 24 million people worldwide.¹ The symptoms are typically categorized into 3 general categories: positive (delusions, disorganized speech, or hallucinations), negative (lack of pleasure, enjoyment, or motivation), and cognitive (impaired executive function).² The annual economic burden of schizophrenia is estimated to be approximately $343 billion in the United States, with a substantial portion of the costs being societal not medical.¹

Treatment recommendations include both pharmacologic (eg, antipsychotic medications) and nonpharmacologic (eg, cognitive behavioral therapy or supported employment services) options.³ Current pharmacologic treatment options mainly target positive symptoms. Unfortunately, effective treatments can have significant side effects, such as movement disorders (eg, tardive dyskinesia), sedation, agitation, and weight gain, that can lead to metabolic syndrome. These side effects can often cause patients to discontinue treatment, leading to the recurrence of symptoms and hospitalizations.²

KarXT (Karuna Therapeutics) is an emerging treatment for schizophrenia, with a regulatory decision anticipated in September 2024. KarXT is a twice-daily oral medication with a novel mechanism of action. It combines xanomeline, which targets muscarinic receptors (M1 and M4 receptor agonist), and trospium, which works to reduce the side effects of peripheral muscarinic receptor activation.⁴

The Institute for Clinical and Economic Review (ICER) conducted a systematic literature review, network meta-analyses (NMAs), and cost-effectiveness analysis to evaluate the clinical and economic outcomes of KarXT for the treatment of schizophrenia. This report summarizes the findings and highlights the key policy recommendations discussed at the public meeting of the New England Comparative Effectiveness Public Advisory Council (CEPAC) on February 9, 2024. The detailed report is available at https://icer.org/wp-content/uploads/2024/03/Schizophrenia_Final-Evidence-Report_For-Publication_03112024.pdf.

Summary of Findings

CLINICAL EFFECTIVENESS

We identified 3 randomized controlled trials (RCTs) of KarXT.^5-7 The trials enrolled adults aged 18 to 65 years with schizophrenia who were hospitalized for an acute exacerbation or relapse of symptoms who were not considered to have treatment-resistant schizophrenia. The primary outcome of the 3 trials was the change from baseline in the total Positive and Negative Syndrome Scale (PANSS) at week 5.

Meta-analyses of the 3 placebo-controlled RCTs (N = 640) showed significant improvements in the total PANSS score (change from baseline = −9.67; 95% credible interval [CrI] = −12.25 to −7.1) and PANSS response (relative risk [RR] = 1.96; 95% CrI = 1.46-2.66) with KarXT compared with placebo at week 5. PANSS response was defined as at least 30% improvement in the total score.

In addition, in the absence of head-to-head trials, we conducted a Bayesian NMA to indirectly compare KarXT with 3 second-generation antipsychotic medications (aripiprazole, risperidone, and olanzapine). These were chosen because they are 3 of the most frequently prescribed antipsychotic medications and represent a range of side effect profiles and effectiveness. For the NMA, we included 30 trials in addition to the 3 trials of KarXT that met the established inclusion criteria (33 total trials). Because of the lack of long-term data for KarXT at the time of this review, we only included short-term RCTs (3-8 weeks), which enrolled adults with a diagnosis of schizophrenia who were experiencing an acute exacerbation of symptoms. Using the available evidence, we compared KarXT with these interventions on PANSS total score, PANSS response, weight gain, and all-cause discontinuation.

Table 1 presents the NMA results for KarXT and the 3 antipsychotics vs placebo. All interventions showed significant improvement in the PANSS total score and PANSS response compared with placebo. However, there were no significant differences between KarXT and the 3 antipsychotics in either outcome. KarXT resulted in significantly less weight gain compared with both olanzapine (mean difference = −2.86 kg; 95% CrI = −3.97 to −1.82) and risperidone (mean difference = −2.06 kg; 95% CrI = −3.29 to −0.87), but not with aripiprazole. In contrast, the all-cause discontinuation rates were statistically significantly higher for KarXT compared with olanzapine (RR = 1.67; 95% CrI = 1.21-2.29) and risperidone (RR = 1.58; 95% CrI = 1.14-2.2).

TABLE 1.

Network Meta-Analysis Results vs Placebo

Intervention	Outcome
	PANSS total score	Weight gain, kg	PANSS response	All-cause discontinuation
	Mean difference vs placebo (95% CrI)		Relative risk vs placebo (95% CrI)
KarXT	−9.78 (−14.83 to −4.74)a	−0.37 (−1.34 to 0.58)	2.03 (1.4-3.06)b	1.19 (0.89-1.59)
Aripiprazole	−8.38 (−12.04 to −4.68)a	0.26 (−0.52 to 1.04)	1.37 (1.01-1.88)b	0.86 (0.72-1.01)
Olanzapine	−10.67 (−13.11 to −8.24)a	2.49 (2.02 to 3)a	1.66 (1.28-2.17)b	0.71 (0.63-0.81)b
Risperidone	−8.05 (−10.99 to −5.03)a	1.69 (0.96 to 2.43)a	1.96 (1.36-2.83)b	0.75 (0.65-0.88)b

^a Estimates signify that the 95% CrI does not contain 0.

^b Estimates signify that the 95% CrI does not contain 1.0.

CrI = credible interval; PANSS = Positive and Negative Syndrome Scale.

Because of a lack of data or inconsistent reporting, other patient-important adverse events were explored qualitatively. Extrapyramidal symptoms, including akathisia, dyskinesia, dystonia, and extrapyramidal disorder, were reported by 3.2% of participants who received KarXT compared with 0.9% in the placebo group.⁸ Small increases in prolactin levels were observed in the KarXT group (0.75 ± 16.45 ng/L) compared with a small decrease in the placebo group (−1.38 ± 16.49 ng/L). Common treatment-emergent adverse events with KarXT were in line with expected side effects of muscarinic receptor activation and trospium, including nausea, vomiting, constipation, and headache. In addition, hypertension (5.9% vs 1.2%) and tachycardia (4.7% vs 2.0%) were more common in the KarXT group compared with the placebo group.⁹

LIMITATIONS OF THE CLINICAL EVIDENCE

Given the new mechanism of action, the lack of data on patients taking KarXT for longer than 5 weeks at the time of this review is the major source of uncertainty. Initial data from acute settings suggest that weight gain may not be an important side effect of KarXT, but longer-term data are needed to confirm this. Similarly, we have no data on other long-term harms, including the incidence of tardive dyskinesia and other long-term movement disorder side effects. In addition, the impact of gastrointestinal-related side effects, such as nausea and constipation, on long-term adherence is not yet clear. Furthermore, there are currently no data on long-term patient-important outcomes, such as relapse prevention, return to work and school, and improvements in relationships with friends and family. Lastly, the data from RCTs of KarXT may not be representative of the larger population of people living with schizophrenia because of the inclusion criteria and study design of the trials (eg, the EMERGENT trials excluded patients with treatment-resistant schizophrenia or other comorbid mental disorders).

LONG-TERM COST-EFFECTIVENESS

A de novo decision analytic model, informed by key clinical trials and prior economic models^10-12 was used to evaluate the lifetime cost-effectiveness of KarXT compared with the current standard of care, including second-generation antipsychotic medications. The model included both a decision tree for acute psychosis and a lifetime Markov model reflecting maintenance treatment. The decision tree assessed adequate clinical response (defined as a ≥30% improvement in the PANSS total score), and the Markov model used 3-month cycles to assess relapse, treatment-emergent adverse events, treatment switching/stopping, and death. Results from the NMA informed key treatment-specific model inputs (eg, adequate clinical response, probability of developing metabolic syndrome, and discontinuation).

The modeled population included adults with schizophrenia who did not have treatment-resistant schizophrenia at the start of the model. Individuals started on either KarXT or aripiprazole, and, if treatment was discontinued, they moved to a second treatment market basket, including risperidone and olanzapine. If the second treatment market basket was discontinued, individuals moved to a third market basket that included risperidone, olanzapine, and clozapine. The treatment sequence was informed by antipsychotic medication efficacy and safety profiles, market share data, and clinical input. The model assumed that KarXT was not associated with a risk of developing metabolic syndrome beyond that of the general population and patients had no metabolic syndrome, diabetes, or cardiovascular disease at baseline. In addition, patients were assumed to receive the treatment over their lifetime, except for a small proportion that stopped treatment at 20 years. The base case was from the health care sector perspective, and all future costs and outcomes were discounted at 3% per year.

The price of KarXT was unknown at the time of this evaluation. Therefore, we used a placeholder price of $20,000 per year in the economic model. Based on the highly favorable assumption that KarXT does not increase the risk of metabolic syndrome and its associated consequences beyond that seen in the general population, KarXT results in less time with diabetes and greater quality-adjusted life-years, greater life-years, and greater equal-value life-years (see Table 2 for the incremental cost-effectiveness ratios results). Using this same assumption, we calculated ICER’s Health Benefit Price Benchmark (ie, the price range that would achieve incremental cost-effectiveness ratios between $100,000 and $150,000 per quality-adjusted life-year or per equal-value life-year gained) to be between $16,000 and $20,000 per year. In a scenario analysis from the modified societal perspective (which included patient productivity, caregiver time, and criminal justice impacts), the incremental cost-effectiveness ratio was similar to the base case analysis.

TABLE 2.

Incremental Cost-Effectiveness Ratios for the Base Case

Treatment	Cost per QALY gained	Cost per evLY gained	Cost per life-year gained	Cost per year without diabetes
KarXTa	$163,000	$146,000	$347,000	$60,000

^a Assuming a KarXT placeholder price of $20,000 per year.

evLY = equal-value life-year; QALY = quality-adjusted life-year.

LIMITATIONS IN THE MODELING OF LONG-TERM COST-EFFECTIVENESS

If KarXT is approved, it will likely be used in the maintenance phase; however, we currently have no data on its long-term efficacy and harms. Therefore, we made numerous assumptions in the economic model about the effectiveness of KarXT in the maintenance phase based on the available acute phase data. In addition, we made the assumption that the long-term effectiveness of KarXT was similar to other approved second-generation antipsychotic medications. However, a key assumption that we varied based on short-term data was that KarXT was not associated with increased metabolic risk beyond that of the general population. This assumption is a key driver of the cost-effectiveness findings. Thus, we tested this assumption and also the impact of KarXT on tardive dyskinesia in numerous sensitivity and scenario analyses presented in our final report at https://icer.org/wp-content/uploads/2024/03/Schizophrenia_Final-Evidence-Report_For-Publication_03112024.pdf.

Policy Discussion

The New England CEPAC is one of the independent appraisal committees convened by ICER to engage in the public deliberation of the evidence on clinical and cost-effectiveness of health care interventions. The New England CEPAC is composed of medical evidence experts, including clinicians, methodologists, and patient advocates. The ICER report on KarXT for the treatment of schizophrenia was the subject of a CEPAC meeting on February 9, 2024. Following the discussion, the CEPAC panel members deliberated on key questions raised by ICER’s report.

A majority of the panel (10-2) voted that the evidence is not adequate to demonstrate that the net health benefit of KarXT is superior to that of aripiprazole. However, a slight majority of the panel (7-5) voted that the current evidence is adequate to demonstrate a superior net health benefit of KarXT from that provided by olanzapine and/or risperidone.

The New England CEPAC also voted on “potential other benefits” and “contextual considerations” as part of a process intended to signal to policymakers whether there are important considerations when making judgments about the long-term value for money that are not adequately captured in analyses of clinical and/or cost-effectiveness. They highlight several factors beyond the results of cost-effectiveness modeling that the CEPAC panel felt were particularly important for judgments of overall long-term value for money. As shown in Table 3, a majority of the panel voted that the treatments for schizophrenia should receive high or very high priority based on the acuity of need and the magnitude of the lifetime impact of the condition. There was no vote on long-term value for money because the price for KarXT was not available at the time of this review.

TABLE 3.

Votes on Contextual Considerations and Potential Other Benefits or Disadvantages

When making judgments of overall long-term value for money, what is the relative priority that should be given to any effective treatment for schizophrenia on the basis of the following contextual considerations?
Contextual consideration	Very low priority	Low priority	Average priority	High priority	Very high priority
Acuity of need for treatment of individual patients based on short-term risk of death or progression to permanent disability	0	0	2	3	7
Magnitude of the lifetime impact on individual patients of the condition being treated	0	0	0	1	11
What are the relative effects of KarXT vs clinically guided management using second-generation antipsychotics on the following outcomes that inform judgment of the overall long-term value for money of KarXT?
Potential other benefit or disadvantage	Major negative effect	Minor negative effect	No difference	Minor positive effect	Major positive effect
Patients’ ability to achieve major life goals related to education, work, or family life	0	0	3	7	2
Caregivers’ quality of life and/or ability to achieve major life goals related to education, work, or family life	0	0	2	7	3
Patients’ ability to manage and sustain treatment given the complexity of the regimen	0	4	7	1	0
Society’s goal of reducing health inequities	1	0	2	9	0

Following the discussion of the evidence, a policy roundtable was convened to deliberate on how best to apply the evidence on KarXT for the treatment of schizophrenia. The policy roundtable members included 1 patient organization representative, 1 caregiver, 2 clinical experts, 2 payer representatives, and 1 representative from Karuna Therapeutics, the manufacturer of KarXT. The full set of policy recommendations can be on the ICER website at https://icer.org/wp-content/uploads/2024/03/ICER_Schizophrenia_Policy-Recommendations_For-Publication_03112024.pdf.

The key policy recommendations are as follows:

• 1.If KarXT receives US Food and Drug Administration (FDA) approval, payers should use the FDA label as the guide to coverage policy and engage clinical experts and diverse patient representatives in considering how to address coverage issues. Specifically, given the significant uncertainty about the long-term effectiveness of KarXT and its presumed high cost in relation to available generic treatment options, it is reasonable for payers to use limited prior authorization as a component of coverage if KarXT receives FDA approval.
• 2.Private payers should ensure that benefit designs developed in conjunction with employers and other plan sponsors do not create requirements for out-of-pocket spending that lead to major barriers to appropriate access, especially for those with lower income.
• 3.Medicaid and Medicare should improve payment and resources for clinicians caring for people with schizophrenia to ensure that patients being discharged from inpatient care have adequate case management and support for housing and care in the community.
• 4.Manufacturers should set prices that will foster affordability and good access for all patients by aligning prices with the patient-centered therapeutic value of their treatments. There is both considerable hope associated with the promise of KarXT and substantial uncertainty regarding its longer-term safety and effectiveness. Launch pricing should reflect these considerations.
• 5.All stakeholders have a responsibility and an important role to play in ensuring that effective new treatment options for people living with schizophrenia are introduced in a way that will help improve comprehensive care for people with schizophrenia and reduce health inequities, particularly for Black Americans.

Funding Statement

Drs Whittington and Tice reports funding from the Institute for Clinical and Economic Review for the conduct of the study. Dr Whittington reports other research support from the National Pharmaceutical Council and from No Patient Left Behind outside the submitted work and other personal fees from Genentech outside the submitted work.