Treatment in early psychosis with N-acetyl-cysteine for 6 months improves low-level auditory processing: Pilot study

Abstract

Sensory impairments constitute core dysfunctions in schizophrenia. In the auditory modality, impaired mismatch negativity (MMN) has been observed in chronic schizophrenia and may reflect N-methyl-d-aspartate (NMDA) hypo-function, consistent with models of schizophrenia based on oxidative stress. Moreover, a recent study demonstrated deficits in the N100 component of the auditory evoked potential (AEP) in early psychosis patients. Previous work has shown that add-on administration of the glutathione precursor N-acetyl-cysteine (NAC) improves the MMN and clinical symptoms in chronic schizophrenia. To date, it remains unknown whether NAC also improves general low-level auditory processing and if its efficacy would extend to early-phase psychosis. We addressed these issues with a randomized, double-blind study of a small sample (N = 15) of early psychosis (EP) patients and 18 healthy controls from whom AEPs were recorded during an active, auditory oddball task. Patients were recorded twice: once prior to NAC/placebo administration and once after six months of treatment. The N100 component was significantly smaller in patients before NAC administration versus controls. Critically, NAC administration improved this AEP deficit. Source estimations revealed increased activity in the left temporo-parietal lobe in patients after NAC administration. Overall, the data from this pilot study, which call for replication in a larger sample, indicate that NAC improves low-level auditory processing in early psychosis.

Introduction

Low-level sensory impairments, both in auditory and in visual processing, seem to constitute part of the core dysfunctions in schizophrenia (Ethridge et al., 2015, Javitt and Freedman, 2015, Javitt, 2009). Increasing evidence indicates that oxidative stress related to glutathione (GSH) synthesis deficits in conjunction with N-methyl-d-aspartate (NMDA) hypofunction are major contributors to the pathophysiology of schizophrenia (Hardingham and Do, 2016). Previous work has shown that add-on administration of the glutathione precursor NAC in chronic schizophrenia patients improves auditory MMN generation to tone deviants (Lavoie et al., 2008) and clinical symptoms (Berk et al., 2008). However, it remains unknown whether NAC can also improve general low-level auditory processing impairments and whether its efficacy extends to the early stages of the disease. The contribution of an impaired antioxidant system in schizophrenia is supported by a variety of findings. Polymorphisms in key genes for GSH synthesis have been associated with schizophrenia (Rodriguez-Santiago et al., 2010, Do et al., 2009, Gysin et al., 2007, Tosic et al., 2006) and are related with decreased GSH levels in the cerebrospinal fluid, prefrontal cortex and post-mortem caudate of patients (Do et al., 2000, Yao et al., 2006, Flatow et al., 2013, Gawryluk et al., 2011, Xin et al., 2016; for a review see Koga et al., 2016, Yao and Keshavan, 2011). GSH-deficient animal models reproduce schizophrenia phenotypes including hypofunction of NMDA receptors (NMDAR) (Steullet et al., 2006). Several studies have demonstrated that the administration of NMDAR agonists induces in healthy controls clinical symptoms as well as sensory processing impairments similar to those observed in schizophrenia and exacerbates these symptoms in patients. Specifically, impairments in the P50, N100, and P300 components of the AEP as well as in oscillatory activity have been observed after acute administration of NMDAR agonists (Chen et al., 2015, Javitt, 2009, Krystal et al., 1994).

MMN, an AEP component generated by deviant stimuli within an oddball paradigm, has been extensively used to measure NMDAR dysfunction in schizophrenia and it is often found decreased in amplitude in both chronic schizophrenia and earlier stages of the disease (Coyle, 2012, Moghaddam and Javitt, 2012, Turetsky et al., 2007). Previous studies have shown that administering NAC to chronic schizophrenia patients led to improved MMN responses (Lavoie et al., 2008), and increased EEG synchronization (Carmeli et al., 2012), in addition to improvement of negative symptoms (Berk et al., 2008). NAC has been used in various studies as a cysteine donor and given orally is quickly absorbed; peak plasma concentration of cysteine is reached within 120 mins (Borgstrom and Kagedal, 1990). NAC crosses the blood-brain barrier, and cysteine can be used in the brain as a GSH precursor (Farr et al., 2003, Conus et al., 2017). Animal studies have previously shown that administration of NAC protects the brain against GSH depletion (Atkuri et al., 2007, Fu et al., 2006, Kamboj et al., 2006) and its neurochemical and morphological consequences (das Neves Duarte et al., 2012, Cabungcal et al., 2013).

Deficits in the P50 and N100 components of the AEPs are well established in patients with schizophrenia (Bodatsch et al., 2015, Brockhaus-Dumke et al., 2008, Rosburg et al., 2008, Turetsky et al., 2008). Schizophrenia patients show a significant diminution in the amplitude of N100, especially for long ISIs (> 1 s) (Rosburg et al., 2008, Shelley et al., 1999). Two recent studies using an auditory oddball paradigm demonstrated that psychotic disorder patients show impairments in response to standard/frequent stimuli in both P50 and N100 (Geiser et al., 2017), as well as the alpha-frequency range, the latter of which is thought to reflect deficient thalamo-cortical connectivity (Lee et al., 2017). The low-level AEP deficit observed by Geiser et al. resulted from weaker responses within the left temporo-parietal lobe and it was correlated with peripheral measures of GSH levels, specifically the ratio of glutathione peroxidase and glutathione reductase activities (GPx/GR) which correlates negatively with brain GSH levels in early psychosis patients (Xin et al., 2016). This AEP deficit in response to standard stimuli within the oddball sequence points to impaired input to auditory cortex (Lee et al., 2017). Previous studies have shown that the responses to standard sounds within oddball ERP tasks are impaired not only in chronic schizophrenia, but also in first–episode schizophrenia patients and clinical high-risk individuals (Salisbury et al., 2010, del Re et al., 2015). In addition, Foxe et al. (2011) demonstrated a significant amplitude reduction of the N100 in clinically unaffected first-degree relatives and highlighted the importance of the use of high numbers of trials for the reliable quantification of the evoked responses of interest.

In the present study, we investigated if the effect of NAC on MMN responses also applies to low-level auditory deficits in psychosis, specifically to the reduced N100 response to standard sounds. Additionally, as the group of patients that participated in this study was in the early stages of the disease, we assessed whether the efficacy of NAC can extend to early psychosis (EP). We show for the first time, that NAC administration results in improved low-level auditory processing.