Clinical Trial Details
— Status: Active, not recruiting
Administrative data
| NCT number |
NCT05463029 |
| Other study ID # |
851120 |
| Secondary ID |
|
| Status |
Active, not recruiting |
| Phase |
|
| First received |
|
| Last updated |
|
| Start date |
August 1, 2022 |
| Est. completion date |
August 2025 |
Study information
| Verified date |
January 2024 |
| Source |
University of Pennsylvania |
| Contact |
n/a |
| Is FDA regulated |
No |
| Health authority |
|
| Study type |
Observational
|
Clinical Trial Summary
Disorders of consciousness (DoC) remain a major clinical challenge in which high rates of
misdiagnosis and difficult prognostication stem from limitations in the ability to access the
disordered physiological processes mechanisms of coma in real world clinical settings. There
is a great need to develop, validate, and translate to clinical use reliable diagnostics to
detect brain recovery potential not evident on neurobehavioral assessment. While resting
state fMRI (rs-fMRI) has demonstrated potential to improve the diagnostic evaluation of DoC
by detecting features of consciousness that are occult at bedside evaluation, this technology
has yet to achieve widespread clinical utility. The investigators propose that recent
advancements in rs-fMRI capabilities can be combined with streamlined analysis and
interpretation approaches to overcome persistent intensive care unit to perform rs-fMRI in
patients with prolonged impaired consciousness due to several causes including TBI, cardiac
arrest, stroke, seizures, and severe CNS infection. The investigators will determine the
optimal methods of data acquisition, analysis and interpretation for predicting recovery of
consciousness in these patients. Our expectations are that this approach will produce highly
reliable functional connectomic characterization of individual DoC patients, thereby allowing
for more accurate outcome prediction. The investigators will additionally investigate the
utility of a novel, simplified radiological approach to rs-fMRI data interpretation in
comparison to computationally intensive connectomic approaches. This
exploratory/developmental project is expected to provide critical data needed to design and
appropriately power future R01 studies validating the efficacy of fMRI-based network
integrity in the clinical evaluation of DoC.
Description:
Disorders of consciousness (DoC) are states of altered consciousness stemming from an
acquired injury to brain networks regulating arousal and awareness, characterized by a loss
of excitatory neural activity across corticocortical, thalamocortical and thalamostriatal
connections. DoC may be due to several causes including traumatic brain injury, anoxic brain
injury from cardiac arrest, stroke, severe brain infections, prolonged seizures, or others. A
primary challenge in the clinical management of DoC is accurately determining the type of DoC
and level of consciousness, as well as the potential for recovery. This is crucial for
decision-making and clinical management, with misdiagnosis rates as high as 40% due to
limitations of clinical evaluation. Recent research has demonstrated that consciousness not
detectable at bedside evaluation may be present in as many as 15-20% patients with DoC. Thus,
to improve the clinical assessment of DoC and ultimately develop targeted interventions to
promote neurological recovery, better methods of assessing DoC in individual patients must be
developed, validated, and translated to clinical use. The use of rs-fMRI in the study of DoC
has led to important insights into the biological mechanisms of coma as well as recovery of
consciousness, yet the translation of rs-fMRI to the clinical management of individual
patients with impaired consciousness has been hampered by several factors. First, BOLD fMRI
reflects a complex combination of signals derived not only from neural activity but also from
non-neurological sources that ultimately comprise as much as 25% of the BOLD signal obscuring
true neural connectivity and introducing spurious connectivity. This limitation has usually
been circumvented by analyzing fMRI at the group level, leveraging the statistical benefits
of averaging data across individuals. Better strategies to distinguish between neural and
non-neural contributors to the BOLD signal are needed to translate rs-fMRI to clinical use in
individual patients. Additionally, conventional acquisition and analysis strategies have
generally been shown to produce reliable estimates of functional connectivity only at
clinically intractable scan times. Recent evidence suggests that highly reliable measures of
functional connectivity can be obtained at the individual level at clinically feasible scan
times using more advanced imaging protocols, outperforming even several-fold longer
conventional acquisitions. The measured BOLD signal can be represented as a function of the
signal intensity at excitation (S0), the echo time, and signal decay constant. Neural
activity affects the BOLD signal by inducing an increase in blood-flow, which reduces the
local concentration of deoxyhemoglobin and in turn decreases the local magnetic
susceptibility thereby reducing the rate of transverse magnetization decay. By examining the
signal decay across multiple echo times, changes in T2* can be distinguished from changes in
S0 thereby allowing neural and non-neural signals from BOLD data to be differentiated. With
sensitivity to BOLD signal being maximal near the T2*, multi-echo acquisitions can also
enhance BOLD sensitivity by placing greater weight on those echoes closest to the T2* of a
particular voxel ("optimally combining"), which is known to vary across the brain. Lastly,
the process of combining BOLD data across echoes carries the advantage of dampening thermal
noise which is randomly organized and tend to cancel out upon averaging. When combined with
additional strategies to remove global motion-related signals, the influence of motion on
BOLD data can be nearly eliminated. Only recently with the integration of other technical
advances such as simultaneous multi-slice has it become possible to acquire multi-echo
rs-fMRI without major compromises in spatial or temporal resolution acquisition. Practical
challenges to obtaining advanced MRI in critically ill patients include the need to transport
patients to research MRI instrumentation capable of the latest acquisition methods that have
traditionally been located on different floors or even different buildings than intensive
care units (ICUs). As a result, advanced MRI methods such as multi-echo rs-fMRI have mainly
been applied to healthy subjects. This project will take advantage of a newly installed
research-dedicated state-of-the-art 3 Tesla MRI system equipped with fast gradients a
32-channel receiver array embedded within the neurointensive care unit (NICU). This system is
capable of multi-echo rs-fMRI at high spatiotemporal resolution and a key goal of this
project is to determine whether functional network integrity in patients with DoC can be more
reliably detected using multi-echo rs-fMRI than with standard rs-fMRI. Optimal methods would
be unsupervised, capable of accommodating interindividual differences in functional topology,
and capable of reducing complex connectivity data to discrete, clinically actionable
insights. Another key goal of this project is to compare three approaches to connectome
construction and characterization in their reliability at the individual patient level and
degree of association with clinical measures of neurologic dysfunction and outcome. The first
and most traditional approach will rely upon transformation of each data to a template space
and node definition based upon group-level meta- analysis. The second approach will use a
previously-described method of subject-level parcellation based upon connectivity gradients
and ultimately compared across individuals based upon the degree of similarity to template
networks. The final approach will use a radiological interpretation scheme in which expert
neuroimagers trained to recognize typical spatial patterns of resting state networks will
render an interpretation of the degree network maps obtained from independent component
analysis.
