Real-World Treatment of Schizophrenia in Adults With a 22q11.2 Microdeletion: Traitement dans le monde réel de la schizophrénie chez des adultes atteints du syndrome de microdélétion 22q11.2

Definition of Treatment Resistance and Other Markers of Clinical Severity

The most common definition of treatment-resistant schizophrenia (TRS) is the failure of ≥2 nonclozapine antipsychotics (adequate dose/duration).^6,9,21 We chose a commonly used conservative proxy^22,23: any clozapine trial or offer/recommendation,^24–26 and/or a clinical diagnosis of TRS with ≥2 nonclozapine antipsychotics.

We also noted other markers of severe/difficult-to-treat illness: lifetime use of electroconvulsive therapy (ECT), ⁶ or ≥2 concurrent antipsychotics.^27,28

Comparison With Prevalence of Treatment Resistance in Community-Based Schizophrenia Sample

The prevalence of TRS ranges greatly, depending on the definition used and the population studied (e.g., higher for inpatients than outpatients).^6,29 We searched the literature for a large community-based (outpatient) sample to use as a comparison to the 22q11.2 microdeletion cohort, seeking one as comparable as possible on key parameters (age, duration of illness, and outpatient status), and definition of TRS used. We identified a study reporting on a random cluster sample of 1,739 outpatients with schizophrenia and analysed for TRS based on eligibility for treatment with clozapine. ³⁰ This sample was of similar mean age (40.0, SD 11.6 vs. 40.5, SD 12.3 years; t = −0.43, p = 0.67) and illness duration to the 22q11.2 microdeletion cohort, though had proportionately more males (70% vs. 54.2%, p ≤ 0.0001) and fewer of European ancestry (52% vs. 82%, p ≤ 0.0001).

Assessment of Treatment Response to Clozapine

We used the Clinical Global Impression–Improvement (CGI-I) scale to assess within-individual treatment response for 30 participants with sufficient pre- and post-treatment data for clozapine and a nonclozapine antipsychotic,^13,31,32 selecting the latter through random-generation (SAS 9.3, SAS Institute, Cary, NC). CGI-I ratings of 1, 2, 3 represent very much, much, minimally, improved, 4 no change, 5–7 minimally, much, and very much worse, respectively. ³¹ The CGI-I was completed independently by 2 psychiatrists (LV and ASB), using full data (LV), and 60 (2-per-participant) anonymized clinical vignettes prepared on pre- and post-treatment mental state and functioning from all records, blinded to antipsychotic and participant (ASB). CGI-I-rating interrater-reliability showed substantial agreement (weighted Cohen's kappa ³³ 0.69, 95% CI, 0.58 to 0.80).

Contemporary Subsample (n = 88) and Treatment at Last Follow-up

Given changes in prescribing over the prolonged time frame, we examined a contemporary subsample at the last follow-up. We derived a subset of 88 (82.2%) participants by restricting those with treatment data from January 1, 2012, to the study end, and recorded antipsychotics at the last follow-up (median-year: 2020; IQR: 2018, 2021). We converted dosages to chlorpromazine equivalent (CPZE) using chlorpromazineR (version 0.2.0).^34–37 We designated primary treatment as clozapine or nonclozapine antipsychotic with the highest CPZE when >1 was concurrently prescribed. We summed the total CPZE per individual. We also noted individuals receiving ECT.

Lifetime History of Treatment Resistance in Individuals With 22q11.2 Microdeletion and Schizophrenia, and Response to Clozapine

There were 63 (58.9%) individuals who met the criteria for TRS, 61 based on clozapine recommendation/trial, and 2 on a clinical diagnosis of TRS. The proportion with TRS in the 22q11.2 microdeletion cohort was significantly greater than reported for a large community schizophrenia sample (42.9%, χ² = 10.38, df = 1, p < 0.01). ³⁰

The 22q11.2 microdeletion-TRS group had a significantly longer median duration of psychotic illness (20.9 years vs. 15.4 years, p = 0.03) and a greater median number of lifetime antipsychotics prescribed (5 vs. 3, p < 0.0001) than the no-TRS group (Table 1). Interestingly, in the no-TRS group, a larger proportion had 22q11.2 microdeletion detected before onset of schizophrenia, than the TRS group (Table 1). There were no other significant between-group differences in demographic, genetic, or clinical variables, including proportion with moderate/severe intellectual disability (Table 1). Eleven (10.3%) individuals had received a course of ECT over their lifetime of treatment (Supplemental Table S1a).

Of the 61 individuals recommended to have clozapine, 49 (80.3%) subsequently had a clozapine trial. The response to clozapine was significantly better than to nonclozapine antipsychotics within individuals (median CGI-I: 2; IQR: 2, 2; median CGI-I: 3; IQR: 3, 4, respectively; S = 141.5, p < 0.0001), for the subset (30 out of 49) with before and after treatment data available.

Most Recent Documented Treatment for a Contemporary Subgroup (n = 88)

Treatments at the last follow-up in the 22q11.2 microdeletion contemporary subgroup (n = 88 out of 107, 82.2%) are shown in Figure 1. The median total antipsychotic dosage per individual was 450 (IQR: 300, 750) mg CPZE, in line with average dosage ranges commonly used in general population schizophrenia.^34–37 This included individuals treated concurrently with one or more antipsychotics or ECT (Figure 1). For most (n = 78 out of 88, 88.6%), the antipsychotic was a second-generation compound. Of the 12 primary antipsychotics recorded at the last follow-up, clozapine was the most commonly used (n = 29 out of 88, 33.0%; median clozapine dosage: 275 mg; IQR: 200, 350 mg), followed by olanzapine, risperidone, and paliperidone (Figure 1). Twelve (13.6%) individuals received a long-acting injectable (LAI) agent (Figure 1).

OPEN IN VIEWER

Figure 1.

Treatments documented at the last follow-up for 88 adults with schizophrenia and 22q11.2 microdeletion. Shown is treatment information available at the last follow-up for the contemporary subsample of 88 individuals with schizophrenia and a 22q11.2 microdeletion (see Methods for details). Dark solid bars indicate results for the primary antipsychotic treatment (for all n = 88), and light solid bars the subset of individuals concurrently prescribed a secondary/additional antipsychotic. Primary treatments were deemed to be clozapine, or the nonclozapine antipsychotic with the highest CPZE dosage when more than one nonclozapine antipsychotic was prescribed (see Methods for details). An asterisk (*) indicates a long-acting injectable antipsychotic for one or more individuals (total, 12 out of 88 individuals). Median total CPZE dose (in mg) per individual (i.e., primary and secondary antipsychotics) is shown for the 7 most commonly used primary antipsychotics (i.e., involving ≥5 individuals). The category “Other” represents one individual each with primary antipsychotic ziprasidone, loxapine, or perphenazine, and one with severe Parkinson's disease who was receiving no antipsychotic treatment (but who, after the study end, was prescribed low dose clozapine). With respect to the 11 out of 29 individuals on clozapine who were concurrently prescribed another antipsychotic, at the last follow-up the CPZE dosage was higher for clozapine than for the other antipsychotic in 6 cases. Adjunctive/secondary antipsychotic medications included those named in the figure as primary antipsychotics, and also brexipiprazole, methotrimeprazine, chlorpromazine, and flupentixol. On the right (grid patterned bar) are individuals receiving ECT at the last follow-up, all concurrently prescribed primary antipsychotic treatment (dark checkered bar), 2 receiving quetiapine, one olanzapine, and one clozapine, each represented to the left within the n = 88 (dark solid bars).

Demographic, clinical and genetic variables, and lifetime results for this subset, both overall and for TRS versus no-TRS comparisons (Supplemental Table S2), appeared similar to those for the full sample (Table 1). Of the 88 individuals, 54 (61.4%) were in the TRS group (Supplemental Table S2). Total CPZE dosages at the last follow-up were significantly higher for the TRS group than for the no-TRS group (median 525 vs. 360 mg, respectively, Z = −2.20, p = 0.03). Antipsychotic polypharmacy at the last follow-up was documented for 28 (31.8%) individuals, including 11 (37.9%) of the 29 receiving clozapine (Figure 1). These 29 individuals represent a majority of those ever tried on clozapine within this subset (29 out of 42, 69.0%) (see footnote in Supplemental Table S1a).

Treatment, Treatment Resistance, and Intolerance in 22q11.2 Microdeletion– Schizophrenia

Notably, dosages of nonclozapine antipsychotics used in this 22q11.2 microdeletion cohort were comparable to those commonly reported for schizophrenia in the general population.^34–37 Of the wide variety of antipsychotics used, second-generation antipsychotics were most common (Supplemental Table S1a; Figure 1), in line with international prescribing trends. These results are consistent with a recent review that recorded treatments reported in the literature for psychosis, including a subset diagnosed with schizophrenia, in individuals with a 22q11.2 microdeletion or related syndrome. ¹⁵

A novel finding was that an estimated 60% of those with 22q11.2 microdeletion–schizophrenia met conservative criteria for treatment resistance (eligibility for clozapine), a significantly greater prevalence than for a large outpatient sample of schizophrenia. ³⁰ This is also greater than figures for TRS reported in most studies of schizophrenia or as summarized in schizophrenia clinical practice guidelines.^6–8 Even individuals with 22q11.2 microdeletion in the no-TRS group had been treated with an average of 3 different antipsychotics. Although there has been speculation about antipsychotic treatment intolerance in 22q11.2 microdeletion, ¹⁵ the dosage results of the current study go against this as a main driver of TRS and clozapine use. Collectively, the findings suggest that a 60% TRS estimate for 22q11.2 microdeletion may be considered a minimum, and/or that some level of treatment intolerance may be a factor in eventually moving to clozapine. Tolerability of antipsychotics and treatment choices, however, appear likely to vary in a manner similar to that for any patient with schizophrenia. ⁴⁵

Clozapine in 22q11.2 Microdeletion–Schizophrenia

Results on within-individual response for this 22q11.2 microdeletion cohort suggest the superiority of clozapine over other antipsychotics, consistent with decades of literature on clozapine efficacy. ³⁹ The results were also broadly consistent with a previous report involving 20 of the total 107 individuals studied here, using a different study design, that found comparable clinical improvement but at a lower median dose of clozapine compared to a Canadian community schizophrenia sample without a 22q11.2 microdeletion. ¹³ In the current study, clozapine dosage was in the lower range of that generally recommended for schizophrenia (200–500 mg). ³⁵ A recent review of 25 articles involving 37 individuals with 22q11.2 microdeletion, most with psychotic illness, reported a mean clozapine dose of 202 mg, ⁴⁶ further supporting dosing at the lower end of typical ranges.

Notably, the majority of individuals with 22q11.2 microdeletion tried on clozapine remained on clozapine. However, the lifetime median of 5 antipsychotics in those with TRS suggests there are likely issues related to delays and underutilization of this special antipsychotic comparable to those in general clinical practice. ⁶ This is despite the many known advantages of clozapine treatment, including lower mortality for clozapine-treated individuals. ⁴⁷ The current study was underpowered to replicate this finding for individuals with a 22q11.2 microdeletion, as indicated by the nonsignificantly lower proportion of individuals who had died in the TRS group than in the no-TRS group (Table 1). Addressing clinician, patient, and caregiver fears and concerns about clozapine is likely to be as important in 22q11.2 microdeletion as for schizophrenia in the general population.^47–49 We have written previously about clozapine safety and related risk reduction measures for 22q11.2 microdeletion, including seizure prevention, using vitamin D and prophylactic anticonvulsants (Table 2),^2,13 in addition to the strict general monitoring system safeguarding against neutropenia risk. ⁴⁷

OPEN IN VIEWER

Table 2.

Clinical Considerations for the Treatment of Schizophrenia in Adults With a 22q11.2 Microdeletion.

Accurate diagnosis is the cornerstone of optimal treatment. For schizophrenia and related treatable psychotic conditions, standard psychiatric diagnostic criteria (e.g., DSM) are as applicable to individuals with 22q11.2 microdeletion as to others. 2 Likewise, clinical genetic testing for individuals with schizophrenia is available for detection of 22q11.2 microdeletion (and other pathogenic CNVs), with attendant management implications once the molecular diagnosis is made.2,40Follow clinical recommendations for 22q11.2 microdeletion, 2 including standard medical history, review of systems, and laboratory assessment (e.g., CBC, calcium, magnesium, PTH, creatinine, TSH, and liver function tests) for 22q11.2 microdeletion-related multisystem comorbidities 20 that may affect schizophrenia symptoms, treatment, treatment response, and development of treatment resistance. Knowledge of commonly associated features, for example, mild thrombocytopenia, will also be beneficial (vs. ascribing solely to medication effects). 2 Periodically assess for motor abnormalities given increased risk of movement disorders with 22q11.2 microdeletion (including early onset Parkinson's disease), and risk of extrapyramidal symptoms with antipsychotic medications that can delay diagnosis of movement disorders.2,59,60,64Other considerations in 22q11.2 microdeletion include catatonia that may be related to schizophrenia and/or Parkinson's disease. 2 Monitor cardiometabolic parameters (e.g. BMI, resting heart rate, blood pressure, HbA1c, glucose, and lipids), given the propensity to obesity, dyslipidemias, and type 2 diabetes in 22q11.2 microdeletion,56–58 in schizophrenia,8 and with antipsychotic medications. 2 Evaluate sleep and consider polysomnography as indicated, given the increased risk of obstructive sleep apnea in 22q11.2 microdeletion, which may be further increased in the presence of obesity and/or a history of palatal flap surgery. 65 A shared care model, for example, community care with a specialty clinic for 22q11.2 microdeletion, may be helpful to boost expertise. 2 Assess cognitive and adaptive functioning, and possible impact on the capacity to consent to treatment, and on medication adherence and monitoring, while providing opportunities for shared decision-making. 2 Involve family or other caregivers as much as possible to assist with support, providing collateral information, and with treatment adherence. 2 Treat using standard clinical practice guidelines for schizophrenia, including antipsychotic medications.6–8,47,49 A “start low, go slow” approach is likely to optimize outcomes, in terms of treatment response and managing side effects. 2 Determine if antipsychotic treatment resistance is present as soon as possible, and treat with clozapine as per standard guidelines for treatment-resistant schizophrenia.6–8,47,49 Starting on a low dose and titrating slowly, while assessing treatment response and side effects, is likely to be especially important for clozapine. 13 Ensure the standard recommended 2000 IU vitamin D daily to prevent hypocalcaemia, and attendant risk for seizures, movement disorders, and QT prolongation. 2 For individuals starting on clozapine, consider prophylactic anticonvulsant treatment to address increased seizure risk in 22q11.2 microdeletion, given that seizures may occur on clozapine even at low dosages (e.g., 100 mg).2,13Anticipatory monitoring for and management of other features of 22q11.2 microdeletion, for example, constipation, will facilitate care. 2

DSM = The Diagnostic and Statistical Manual of Mental Disorders; CNV = copy number variation; CBC = complete blood count; TSH = thyroid stimulating hormone; PTH = parathyroid hormone; BMI = body mass index; HbA1c = haemoglobin A1c.

Molecular Genetic Subtypes of Schizophrenia and Treatment Resistance

Our finding of elevated prevalence of TRS in 22q11.2 microdeletion is consistent with recent data available on CNVs that convey an increased risk for schizophrenia. ⁵ One study of patients with TRS reported an increased prevalence of clinically relevant CNVs, including 4 individuals with 22q11.2 microdeletion. ¹¹ Another reported increased prevalence of TRS in individuals with CNVs (n = 15 with typical 22q11.2 microdeletion). ¹² A review of case reports involving 8 individuals with a 3q29 microdeletion and schizophrenia suggested possible enrichment of TRS, efficacy of clozapine, and no relation to intellectual disability. ¹⁰ Here, we similarly found no relationship between moderate/severe intellectual disability and treatment resistance in 22q11.2 microdeletion. Collectively, these results provide support for the potential utility of clinical genetic diagnostics in schizophrenia, that is, genome-wide microarray for detecting pathogenic CNVs.^4,40,50 Clinically detectable schizophrenia-relevant CNVs, including 22q11.2 microdeletion, may be considered to be genetic markers of potential TRS.

Clinical Considerations for Treatment of Schizophrenia in 22q11.2 Microdeletion

The results of the current study appear broadly consistent with clinical practice guidelines for schizophrenia, in general, and for 22q11.2 microdeletion.^2,6–8 The findings support the likelihood that most issues that are important for the general population of genetically undiagnosed schizophrenia are also relevant for 22q11.2 microdeletion–schizophrenia. This would include treatment adherence, and underutilization despite known benefits of LAI antipsychotics to reduce the risk of relapse and hospitalization. ⁵¹ A recent Canadian study estimated the use of LAIs at 6.5%, ⁵² and our results showed 13.6% prevalence in 22q11.2 microdeletion (Figure 1).

For individuals with a 22q11.2 microdeletion, as for any patient with schizophrenia, there are no perfect choices in regard to antipsychotics, with all choices requiring consideration of the balance between benefits and risks. The benefits of antipsychotics for the treatment of schizophrenia are clear.^6–8,39 The risks – and feasible management of these risks – are also clear for antipsychotics, almost all now in long-term use, including 35 years of experience with clozapine. ³⁹

Table 2 outlines clinical considerations for managing schizophrenia in the context of a 22q11.2 microdeletion. Importantly, 22q11.2 microdeletion-associated conditions are responsive to standard treatment strategies for each, including schizophrenia, thus there are longstanding recommendations to follow routine clinical practice guidelines for individual conditions.^{1,2,6–8,53–55} The main caveat is that in all cases attention be paid to the multisystem nature of 22q11.2 microdeletion-related comorbidities.^2,17,20,53 There is also a general emphasis on the importance of family or other caregivers in obtaining a history for diagnosis, monitoring treatment response, facilitating medication management, and ensuring appointment attendance. ²

The 22q11.2 microdeletion conveys increased risks for multiple conditions, whether or not antipsychotics are used, and perhaps regardless of the specific antipsychotic. ² Notably, the onset of Parkinson's disease, obesity, hypertriglyceridemia, and type 2 diabetes is on average several decades earlier in 22q11.2 microdeletion than in the general population. For some individuals, and particularly for metabolic conditions, onset may precede, coincide with, or be detected soon after the onset of schizophrenia (Table 2).^2,56–60 Thus, it will be important to consider the propensity to, and/or presence of, pre- and post-treatment neurologic and cardiometabolic risks, and to monitor for and mitigate these risks (Table 2).^2,59,60 Looking ahead, new strategies such as concomitant metformin with antipsychotic treatment may be particularly relevant for individuals with a 22q11.2 microdeletion. ¹⁷

Diagnostic Considerations Relevant to Schizophrenia and the 22q11.2 Microdeletion

In the current study, clinical genetic testing for the 22q11.2 microdeletion occurred before the onset of schizophrenia in a significantly greater proportion of individuals in the no-TRS group than in those with TRS (Table 1, Supplemental Table S2). This suggests the possibility that earlier genetic diagnosis of 22q11.2 microdeletion can help facilitate anticipatory care and thus more timely diagnosis and effective management of schizophrenia. ² Potential clinical relevance includes reducing the duration of untreated psychotic illness, and reducing the time to clozapine, given the evidence of increased TRS.

In the general population, poorer outcomes of schizophrenia are associated with a longer duration of untreated psychotic illness, ⁶¹ and longer time to clozapine prescription,^48,49 however, there remain ongoing challenges with addressing these.^47–49,61 Similarly, in 22q11.2 microdeletion, and despite the >20-fold increased risk for schizophrenia, misattribution of symptoms to intellectual/learning disabilities, challenges in obtaining a longitudinal history, and/or reluctance to make a diagnosis of schizophrenia or TRS, may cause diagnostic delays and/or confusion about optimal management. ² These will be important issues to address in clinical care and to assess in a prospective study.

For 22q11.2 microdeletion, a genetic diagnosis can also facilitate the identification of treatable comorbidities, including those directly applicable to psychiatric care. ² A relevant example related to antipsychotics is the opportunity to forestall the misattribution of Parkinson's disease symptoms to extrapyramidal side effects, and thus prevent the serious consequences of diagnostic and treatment delays (Table 2).^59,60

Notwithstanding these advantages, there remain challenges with implementing genetic diagnostics in psychiatry. Many individuals who meet the criteria, especially adults, have not had standard clinical genetic testing to identify the 22q11.2 microdeletion. ⁵⁰ Barriers to implementation vary from location to location but lack of knowledge, training and practice in ordering genetic tests, and associated fears and concerns (in contrast to other more commonly used tests such as neuroimaging) are common in many areas of medicine. ⁵⁰

Strengths and Limitations

To our knowledge, this is the largest report of treatment of schizophrenia in individuals with 22q11.2 microdeletion. Strengths include a well-characterized cohort with molecularly confirmed typical 22q11.2 microdeletion and application of standard diagnostic criteria for schizophrenia. The duration of illness excluded early phase psychotic illness, serving to further reduce diagnostic uncertainty. Observational data allowed determination of real-world treatments, and assessment of treatment resistance used a conservative definition, as recommended for retrospective studies. ⁹ Limitations mainly involve the retrospective and naturalistic nature of the study, involving a period of 25+ years, multiple clinicians, and variable availability and quality of records documenting lifelong antipsychotic treatment. These limitations precluded obtaining consistent data on adherence, detailed response, relative timing/order and rationale for all trials of treatment or their discontinuation. Also, the community-based schizophrenia sample used for the TRS comparison had no genetic test data available. ³⁰ Some individuals could have had 22q11.2 microdeletion or another CNV with increased TRS propensity, thereby tending to diminish between-sample differences, but it is unlikely this would have substantively changed the results. There may have been unknown effects of ascertainment of the 22q11.2 microdeletion cohort. Nearly half of the individuals in this schizophrenia cohort were referred from psychiatric compared to other sources, for example, paediatrics/genetics. However, this variable showed no TRS-related between-group differences (Table 1). Although the cohort was ascertained through a specialty clinic for adults with 22q11.2 microdeletion, nearly two-thirds (63.7%) of the individuals studied had onset of schizophrenia, and thus antipsychotic treatment, before molecular diagnosis. Management could thus only be affected by adult specialty clinic services after that time, and after referral, as for all participants. Further research would be needed to assess the effects of specialty services on results.