Optimisation of clozapine use during early years of schizophrenia – can clozapine be considered a disease modifying drug?

Jones, Rowena M (2024). Optimisation of clozapine use during early years of schizophrenia – can clozapine be considered a disease modifying drug? University of Birmingham. Ph.D.

Preview

Jones2024PhD.pdf
Text - Accepted Version
Available under License All rights reserved.
Download (3MB) | Preview

Abstract

Introduction
It is increasingly recognised that schizophrenia may be an immunological disease. Although there has been some progress in the management of schizophrenia by way of innovative service development over the last two decades, drug treatment has remained essentially static, with no major breakthroughs since the arrival of clozapine in the 1980s. Contrast this with the progress that has been made in the treatment of established immunological conditions such as rheumatoid arthritis, where the use of disease modifying drugs has transformed the therapeutic landscape. New insights into the immunology of schizophrenia provide potential for better understanding of how current treatments work and also a road map for developing novel compounds. Clozapine remains the treatment of choice for treatment resistant schizophrenia (TRS). It is possible that clozapine’s superiority over traditional antipsychotics in TRS is down to its immunological effects. This raises the prospect that clozapine may be a disease modifying drug, capable of changing the course of disease, in which case earlier prescription of clozapine would be likely to produce better outcomes for patients. Clozapine is known to have effects on peripheral blood immune cells including neutrophils. Neutrophils have been shown to be elevated in schizophrenia, and there is much interest currently in the role of neutrophils in the pathogenesis of a number of immune conditions. A spike in neutrophils shortly after starting clozapine is commonly seen in clinical practice and has been reported in the literature. It is possible that these early changes in neutrophil counts following clozapine commencement may be linked with response to clozapine. In addition, clozapine has been found to be associated with secondary immunoglobulin deficiency which may also have value as a clinical biomarker in the treatment of TRS. In this thesis I have explored whether there is evidence in the literature that earlier clozapine is beneficial in TRS and have conducted my own study into the timing of clozapine and its effectiveness. I have also examined neutrophil and immunoglobulin trajectories with clozapine use and looked for evidence of an association with clinical outcome.

Hypothesis and methods
I have set out to test the following three hypotheses.
Hypothesis 1. Clozapine will be more effective if used earlier in the course of course of TRS. I have investigated this by conducting a meta-regression of the effect of age (as a proxy for duration of illness) on clozapine response from randomised controlled trial data of clozapine in comparison to other antipsychotic drugs. I have then conducted an observational study of anonymised patient records, using ordinal logistic regression, to look for an association between duration of prior psychotic illness and clozapine response.
Hypothesis 2. Patients can be classified by neutrophil trajectory following clozapine commencement, and neutrophil trajectory can predict response to clozapine. I have taken the database that I created for paper 2, and linked this with ZTAS full blood count data, in order to conduct a latent class growth analysis (LCGA) of baseline and early neutrophil counts with clozapine and a logistic regression of neutrophil trajectory class against clinical outcome.
Hypothesis 3. Immunoglobulin levels fall with clozapine treatment and reduction in globulin level can predict response to clozapine. Combining my database from paper 2 with calculated globulin (CG) results obtained by a fellow CRIS researcher, I have compared CG counts pre and post clozapine prescription and have carried out a logistic regression of change in globulin score against clinical outcome.

Results
Paper 1 – Chapter 3
From a meta-analysis of 34 papers, meta-regression results failed to demonstrate an effect of age on clozapine response (p = 0.79, [95% CI -0.03 – 0.03]), however a linear regression of age against response using individual patient data from one study did show a significant effect of age, with younger age associated with greater response to clozapine (p=0.00, [95% CI -110.71 - -27.20]). Individual patient data from a second smaller study did not show a significant effect.
Paper 2 – Chapter 4
From a sample of 425 patients obtained using routine electronic clinical data, ordinal logistic regression results showed a significant association between duration of illness prior to commencing clozapine, and clozapine response (adjusted OR = 1.04 [95% CI 1.01 – 1.06]), indicating a 4% increase in the odds of a higher (worse) outcome CGI-S score per additional year of illness.
Paper 3 – Chapter 5
Using the same sample from paper 2, LCGA suggested 3 distinct classes of neutrophil trajectories, differing from outset with high, high-normal and low normal counts. Logistic regression showed that high-normal neutrophils were associated with higher odds of clozapine response (adjusted OR = 2.10 [95% CI 1.31 – 3.36]).
Paper 4 – Chapter 6
From a sample of 343 patients, 149 had pre and post calculated globulin (CG) levels available whilst 341 had only post CG levels. Mean CG level fell significantly following clozapine treatment (t = 2.74 p = 0.007). Logistic regression showed no association between change in CG level and clozapine response (adjusted OR 1.02 95% CI [0.89-1.16]). There was a significant association between CG level 1 year post clozapine and clozapine response (adjusted OR 1.06 95% CI [1.00-1.12]), but the data was of poor quality. Sex and ethnicity differences were found in CG levels pre and post clozapine.

Conclusions
The results from papers 1 and 2 provide support for the hypothesis that earlier clozapine is associated with better response to clozapine. Whilst the meta-regression results were not significant, there was evidence from limited individual patient data of an association between younger age, and shorter duration of illness, and better response to clozapine. In the paper I provide a critique of the meta-regression, both in terms of the methodology and the data quality. Paper 2 in comparison provided clearer evidence of an association between shorter duration of prior illness and clozapine response, and whilst it was an observational rather than experimental study, it had the advantage of being a sample of real-world clinical data.
Paper 3 provided evidence in support of differing neutrophil trajectories following clozapine initiation. However, rather than variation between classes in terms of a spike in neutrophil count with clozapine, neutrophil counts varied from the outset, with high normal neutrophil counts associated with a greater response to clozapine. These results are in keeping with a hypothesis that patients with high normal neutrophil counts, reflecting an ongoing inflammatory process, are more likely to respond to clozapine, and indicate that neutrophils may have utility as a biomarker to predict clozapine response in TRS.
Results from paper 4 support the hypothesis that immunoglobulin levels fall with clozapine treatment but failed to show an association between fall in CG level and response to clozapine. There were sex and ethnicity differences in CG levels both pre and post clozapine, with most patients who developed low CG levels on clozapine being white males. However, the data quality for this paper was poor and results need to be interpreted with caution.
In summary, the key findings from this thesis i.e. firstly, that earlier clozapine may be associated with better outcome in TRS and secondly, that neutrophil count can help predict response to clozapine, support the concept of clozapine acting as a disease modifier by reducing the burden of inflammation in TRS.

Type of Work:

Thesis (Doctorates > Ph.D.)

Award Type:

Doctorates > Ph.D.

Supervisor(s):