Elsevier

Clinical Neurophysiology

Volume 127, Issue 2, February 2016, Pages 1445-1451
Clinical Neurophysiology

The prognostic value of sleep patterns in disorders of consciousness in the sub-acute phase

https://doi.org/10.1016/j.clinph.2015.10.042Get rights and content

Highlights

  • Regular sleep structure is a good predictor of clinical outcome in sub-acute DOC patients.

  • Regular sleep structure is stronger than other known prognostic factors of DOC outcome.

  • Visual-quantitative sleep analysis is helpful in the prognostic evaluation of DOCs.

Abstract

Objective

This study aimed to evaluate, through polysomnographic analysis, the prognostic value of sleep patterns, compared to other prognostic factors, in patients with disorders of consciousness (DOCs) in the sub-acute phase.

Methods

Twenty-seven patients underwent 24-h polysomnography and clinical evaluation 3.5 ± 2 months after brain injury. Their clinical outcome was assessed 18.5 ± 9.9 months later. Polysomnographic recordings were evaluated using visual and quantitative indexes. A general linear model was applied to identify features able to predict clinical outcome. Clinical status at follow-up was analysed as a function of the baseline clinical status, the interval between brain injury and follow-up evaluation, patient age and gender, the aetiology of the injury, the lesion site, and visual and quantitative sleep indexes.

Results

A better clinical outcome was predicted by a visual index indicating the presence of sleep integrity (p = 0.0006), a better baseline clinical status (p = 0.014), and younger age (p = 0.031). Addition of the quantitative sleep index strengthened the prediction.

Conclusions

More structured sleep emerged as a valuable predictor of a positive clinical outcome in sub-acute DOC patients, even stronger than established predictors (e.g. age and baseline clinical condition).

Significance

Both visual and quantitative sleep evaluation could be helpful in predicting clinical outcome in sub-acute DOCs.

Introduction

The outcome of patients with disorders of consciousness (DOCs) remains difficult to predict. A key factor to consider is the duration of the DOC, as the prognosis worsens the longer the patient is in the condition (MSTF, 1994, Lammi et al., 2005, Katz et al., 2009). The identification, in the sub-acute phase, of further prognostic factors could allow better planning of treatment and rehabilitation for patients in the early stages of DOCs. The patient’s age, the type of brain injury (Bernat, 2006, Monti et al., 2010), and the severity of the DOC (Katz et al., 2009, Luauté et al., 2010) are well-known prognostic factors in this setting; patients in the minimally conscious state (MCS) have been found to have a better prognosis than those in a vegetative state/unresponsive wakefulness syndrome (VS/UWS). Even though MCS patients showed a better prognosis than VS/UWS patients, differential diagnosis between these two conditions continues to be challenging (Schnakers et al., 2009). Moreover, the boundaries between MCS and VS/UWS are elusive (Guldenmund et al., 2012) and there actually exist intermediate states between the two conditions (Monti et al., 2010). It is therefore believed that viewing DOCs as a continuum of different levels of consciousness may allow a more realistic assessment of patients’ clinical status and prognosis.

The presence of regular sleep patterns could reflect preserved brain functioning, and the diagnostic and prognostic value of sleep in DOCs is a topic of growing interest (Cologan et al., 2010). However, most of the available studies were published before the definition of MCS (Giacino et al., 2002), and thus refer to a clinical framework that does not reflect current opinion. The studies conducted prior to the definition of MCS showed persistence of non-REM (NREM) sleep patterns (D’Aleo et al., 1994a, D’Aleo et al., 1994b, Evans and Bartlett, 1995, Giubilei et al., 1995, Oksenberg et al., 2001, Isono et al., 2002, Valente et al., 2002) and the sporadic presence of REM sleep markers (Gordon and Oksenberg, 1993, D’Aleo et al., 1994b, Oksenberg et al., 2001, Valente et al., 2002) in VS patients. In addition, the presence of organised sleep patterns was proposed as a reliable prognostic marker in post-traumatic coma patients (Valente et al., 2002) and in severe head-injured patients (Evans and Bartlett, 1995). However, none of these studies compared the prognostic value of sleep analysis with that of other, known prognostic markers.

Five studies (Landsness et al., 2011, Cologan et al., 2013, de Biase et al., 2014, Kang et al., 2014, Rossi Sebastiano et al., 2015) have been conducted since the definition of MCS (Giacino et al., 2002). Three of these (Landsness et al., 2011, de Biase et al., 2014, Rossi Sebastiano et al., 2015) focused on the diagnostic rather than the prognostic value of sleep analysis in DOCs. However, in one of them (Rossi Sebastiano et al., 2015), an assessment protocol based on multiple neurophysiological tests, including sleep evaluation, was found to provide significant information about residual functioning in chronic DOC patients.

Most of the above studies investigated heterogeneous samples that included DOC patients in the acute, sub-acute or chronic stages. Cologan et al. (2013), studying a group of sub-acute DOC patients, found that a large amount of standard spindles was associated with a better clinical outcome. However, the prognostic value of sleep integrity compared to other prognostic factors was not assessed. Finally, Kang et al. (2014) recently proposed a bedside scoring system for predicting awareness recovery in sub-acute UWS patients. Their score (Kang et al., 2014) includes clinical and neurophysiological parameters, such as the presence/absence of sleep spindles. However, their study did not include MCS patients.

The present study set out to compare the potential prognostic value of sleep/wake patterns (their persistence and various features) to that of other, known prognostic factors in sub-acute DOCs. To this end, we combined clinical evaluation with the use of a sleep-structure integrity index, created on the basis of relevant literature data. We also explored whether and how quantitative EEG spectral analysis parameters might be of prognostic value. Both VS/UWS and MCS patients were included in the study. These conditions were considered to lie on a continuum of different levels of consciousness.

Section snippets

Patients

The study involved 31 non-sedated, spontaneously breathing DOC patients, admitted consecutively to the Coma Unit at the “S. Maugeri” Foundation (Pavia, Italy) over a period of 30 months. The patients underwent neurological examination, brain imaging (CT or MRI) and standard EEG studies. The Italian version of the Coma Recovery Scale-Revised (CRS-R) (Sacco et al., 2011) was used for clinical assessment. Once the patients were in a stable condition, they were admitted to the Coma Unit and

Results

The PSG findings are summarised in Table 2: NREM markers were found in 24/27 patients and were always present in recordings in which REMs were present, therefore, in these patients, the sleep-structure index in fact indicated increasing levels of sleep organisation. In nine patients, identifiable sleep stages could be scored (sleep-structure index: 4), clustered in sleep cycles of variable length and having a cumulative duration ranging from 1 to 8 h out of the total 24 h of the recording. In 15

Discussion

In the present study, the presence of structured sleep, as identified by means of 24-h PSG, was shown to be a statistically significant predictor of a positive clinical outcome in a group of sub-acute DOC patients, even more powerful than established predictors, such as younger age and a less severe baseline clinical condition (Luauté et al., 2010, Monti et al., 2010). As shown in Fig. 2, a clear correspondence emerged between increasing complexity of sleep architecture and CRS+ score at

Conclusions

In conclusion, our data confirm previous data about the importance of sleep integrity in DOC patients. Moreover, we suggest that persistent and more organised sleep patterns may reliably predict positive outcome in sub-acute DOC patients. These features may be even stronger predictors than existing, well-established prognostic factors, such as patient age and clinical status. The sleep-structure index presented herein (possibly supported by quantitative analysis) could be a useful measure of

Acknowledgements

The authors thank Federica Camasso and Laura Spelta for performing 24-hour bedside polysomnography.

Conflict of interest: None of the authors have potential conflicts of interest to be disclosed.

References (30)

  • V. Cologan et al.

    Sleep in the unresponsive wakefulness syndrome and minimally conscious state

    J Neurotrauma

    (2013)
  • G. D’Aleo et al.

    Sleep spindles in the initial stages of the vegetative state

    Ital J Neurol Sci

    (1994)
  • G. D’Aleo et al.

    Sleep in the last remission stages of vegetative state of traumatic nature

    Funct Neurol

    (1994)
  • B.M. Evans et al.

    Prediction of outcome in severe head injury based on recognition of sleep related activity in the polygraphic electroencephalogram

    J Neurol Neurosurg Psychiatry

    (1995)
  • J.T. Giacino et al.

    The minimally conscious state: definition and diagnostic criteria

    Neurology

    (2002)
  • Cited by (46)

    • EEG-based methods for recovery prognosis of patients with disorders of consciousness: A systematic review

      2022, Clinical Neurophysiology
      Citation Excerpt :

      On the other hand, gamma waves were evaluated only when applying transcranial alternating (Naro et al., 2016b) or direct (Naro et al., 2016c) current stimulation. Ten studies focused their analysis on the delta, theta, and alpha bands, which are typical frequencies generally investigated in patients with DoC; three added beta bands (Arnaldi et al., 2016; Straudi et al., 2019; Wu et al., 2011); and only one also included sigma band (Arnaldi et al., 2016) (Fig. 3b, inner plot). We found high heterogeneity in the type and number of features extracted (Supplementary Material Table 2).

    View all citing articles on Scopus
    View full text