Abstract
The purpose of this study was to perform a systematic, scoping review of commonly described intermittent/semi-intermittent autoregulation measurement techniques in adult traumatic brain injury (TBI). Nine separate systematic reviews were conducted for each intermittent technique: computed tomographic perfusion (CTP)/Xenon-CT (Xe-CT), positron emission tomography (PET), magnetic resonance imaging (MRI), arteriovenous difference in oxygen (AVDO2) technique, thigh cuff deflation technique (TCDT), transient hyperemic response test (THRT), orthostatic hypotension test (OHT), mean flow index (Mx), and transfer function autoregulation index (TF-ARI). MEDLINE®, BIOSIS, EMBASE, Global Health, Scopus, Cochrane Library (inception to December 2016), and reference lists of relevant articles were searched. A two tier filter of references was conducted. The total number of articles utilizing each of the nine searched techniques for intermittent/semi-intermittent autoregulation techniques in adult TBI were: CTP/Xe-CT (10), PET (6), MRI (0), AVDO2 (10), ARI-based TCDT (9), THRT (6), OHT (3), Mx (17), and TF-ARI (6). The premise behind all of the intermittent techniques is manipulation of systemic blood pressure/blood volume via either chemical (such as vasopressors) or mechanical (such as thigh cuffs or carotid compression) means. Exceptionally, Mx and TF-ARI are based on spontaneous fluctuations of cerebral perfusion pressure (CPP) or mean arterial pressure (MAP). The method for assessing the cerebral circulation during these manipulations varies, with both imaging-based techniques and TCD utilized. Despite the limited literature for intermittent/semi-intermittent techniques in adult TBI (minus Mx), it is important to acknowledge the availability of such tests. They have provided fundamental insight into human autoregulatory capacity, leading to the development of continuous and more commonly applied techniques in the intensive care unit (ICU). Numerous methods of intermittent/semi-intermittent pressure autoregulation assessment in adult TBI exist, including: CTP/Xe-CT, PET, AVDO2 technique, TCDT-based ARI, THRT, OHT, Mx, and TF-ARI. MRI-based techniques in adult TBI are yet to be described, with the main focus of MRI techniques on metabolic-based cerebrovascular reactivity (CVR) and not pressure-based autoregulation.
Introduction
A
Of particular interest is the brain's ability to autoregulate CBF after traumatic brain injury (TBI). Failure of normal autoregulation post-TBI has may be attributed to both primary and secondary neurological insults, and may worsen mortality and morbidity. 2,3 Various techniques have been developed to measure the autoregulatory capacity, using a range of monitoring technologies, often in association with manipulations in systemic physiology. 3 –5 The information derived from these techniques varies widely, with some providing an intermittent snapshot of autoregulatory status, whereas others can produce a continuous index of autoregulatory reserve. 3,4
However, the literature base describing these techniques is quite vast and difficult to navigate. The nomenclature surrounding the classification of measurement techniques is not clear, but a common classification is into “dynamic” versus “static.” 4,5 Dynamic techniques refer to those tests in which the physiology is directly manipulated in a fast manner, with autoregulation being measured through various CBF responses of a transient nature (such as transient compression of the carotid artery and observation of post-compression transient hyperemia). Static tests refer to those in which autoregulation is derived in the absence of any manipulation of physiology. Perhaps an alternate method of classification, particularly in the context of TBI, could be to differentiate methods that provide either “intermittent” or “continuous” autoregulation measure, where “intermittent” techniques are those that produce a single measure of autoregulatory capacity at the time of the test, whereas “continuous” measures are those that provide a regularly updating index of autoregulatory status.
This is Part I of a two part article series. Our goal was to collect the available literature on pressure autoregulatory measurement techniques (both intermittent/semi-intermittent and continuous) in adult TBI, with two main intentions: 1) to outline the commonly described techniques, and 2) to provide a comprehensive reference library of the available literature describing the use of these measurement techniques in adult TBI. The goal was not to describe, in detail, the individual studies and their findings. Within this article, Part I, we focus on intermittent/semi-intermittent pressure autoregulatory measurement techniques in adult TBI. The techniques explored include: arteriovenous oxygen difference (AVDO2), computed tomographic perfusion (CTP)/Xenon-CT (Xe-CT), magnetic resonance perfusion (MRP)/functional MRI (fMRI), positron emission tomography (PET), thigh cuff deflation technique (TCDT) with calculation of autoregulatory index (ARI) via transcranial Doppler (TCD), transient hyperemic response test (THRT), orthostatic hypotension test (OHT), and the semi-intermittent TCD-based techniques of mean flow index (Mx) and transfer function ARI (TF-ARI).
Methods
Multiple small systematic reviews were conducted (nine in total), using the methodology outlined in the Cochrane Handbook for Systematic Reviews. 6 Each intermittent technique of pressure autoregulation measurement was searched separately. The following methodology outline is similar for this article (Part I) and the other article in the series (Part II).
Search question, population, and inclusion and exclusion criteria
The question posed for each individual systematic review was: What literature is available on “x” pressure autoregulation measurement technique in adult TBI? The following intermittent/semi-intermittent techniques were individually searched within their own individual systematic review: AVDO2, CTP/Xe-CT, MRP/fMRI, PET, TCDT with calculation of ARI via TCD, THRT, OHT, and the semi-intermittent TCD-based techniques of Mx and TF-ARI. We included Mx and TF-ARI within the intermittent/semi-intermittent category because, despite producing an updating index/measure of autoregulatory status, both techniques still rely on TCD and, therefore, are typically applied for only short periods of time (i.e., a few to 60 min, with longer recordings possible, but requiring very special care). Therefore, we refer to these techniques as “semi-intermittent.”
We specifically excluded studies that addressed cerebrovascular reactivity (CVR) in response to metabolic manipulation (such as with CO2 testing) from this review, because these methods test metabolic reactivity of the cerebral vasculature and not autoregulation as it is classically understood (which addresses the vascular response to variations in blood pressure).
Given that the main goal of this article was to produce a scoping review describing these intermittent/semi-intermittent measurement techniques and provide a comprehensive reference library of the available studies in adult TBI, we did not define primary/secondary outcomes of interest. Further, the individual studies were not discussed in detail within the body of the article, as the goal of the comprehensive searching was to provide a reference library for interested readers.
Inclusion criteria were: All studies including human subjects with TBI (any severity), studies with five or more patients, studies with adults only (≥ 18 years of age), the use of an intermittent/semi-intermittent autoregulation measurement technique (as listed), and focus on pressure autoregulation (i.e., not metabolic CVR). Exclusion criteria were: Animal studies, studies with pediatric patients, and studies of fewer than five patients. Non-English-language studies were excluded given the small number identified.
The only exception to the abovementioned criteria was for the Mx search. Given the large volume of literature available, we focused on studies with ≥50 patients. Further, in both of these searches we included only those documenting the association between Mx and patient functional outcome (i.e., morbidity and mortality). The reason for stricter inclusion criteria for the Mx search is that there is a large volume of literature on this technique (≥ 100 articles on TBI alone). In addition, we wanted to provide the core studies for the reference library to the reader (i.e., the literature with large patient cohorts [> 50 patient] and documentation between these indices and a clinically relevant parameter, patient outcome). There are many smaller studies on Mx available documenting various associations between this index and both patient outcome and other neurophysiology. These are not reviewed in this article or in the supplementary material.
Search strategy
Each intermittent/semi-intermittent technique was separately searched through individual systematic reviews of the literature. In total, there were nine separate systematic reviews conducted, one for each intermittent technique. The strategy described subsequently was conducted for each technique.
MEDLINE®, BIOSIS, EMBASE, Global Health, Scopus, and Cochrane Library from inception to December 2016 were searched using individualized search strategies. The search strategy for MEDLINE can be seen in the Supplementary Text for each intermittent/semi-intermittent technique, with a similar search strategy utilized for the other databases (see online supplementary material at
Study selection
A two-step review of all articles returned by our search strategies was performed. First, the reviewer independently screened titles and abstracts of the returned articles to decide if they met the inclusion criteria. Second, full texts of the chosen articles were then assessed to confirm if they met the inclusion criteria and that the primary outcome of patient functional outcome was reported in the study.
Any meeting abstracts identified within the database search results were crosschecked with MEDLINE to determine if any full-length articles were subsequently published based on these abstract results. If a full-length article was identified, the meeting abstract was discarded from the final review and the formal article was included in its place.
Data collection
Data were extracted from the selected articles and stored in an electronic database. Data fields included: patient demographics, type of study, article location, number of patients, autoregulation technique described, autoregulation based outcomes described, and complications associated with the technique. Basic study characteristics can be seen in Tables 1 and 2. All data on specific study outcomes can be seen in Tables S1–S3 (see online supplementary material at
Results
Search results
Nine separate systematic reviews were conducted based on the nine individual intermittent/semi-intermittent techniques for autoregulatory measurement that were selected. The flow diagrams of the search/filtering results for each individual technique can be seen in Figures S1–S8(see online supplementary material at
1. The CTP/Xe-CT search yielded 218 references, with 135 for the first filter once duplicates were removed. Eleven articles made it through to the second filter, with another four added from the reference sections of the selected articles. After application of the inclusion/exclusion criteria to the full manuscripts, 10 articles were deemed eligible for inclusion in the final review. 7 –16
2. MRI technique searches produced 320 studies, with 194 remaining after removal of duplicates. After application of the first and second filtering processes, no articles were deemed eligible for inclusion. Even after reviewing the reference sections of all applicable articles, we were unable to identify any studies on MRI-based techniques for pressure autoregulatory measurement in adult TBI.
3. Searching for PET-based studies produced 365 references, with 323 for first filter after removal of duplicates. Twelve references made it to the second filter stage, with an additional one article added from the reference sections. After the second filter stage, six articles were deemed eligible for inclusion in the final review. 17 –22
4. The AVDO2 search produced 61 references, with 25 remaining after removal of duplicates. After passing through the first and second filtering stages, with the addition of 2 articles from the reference sections, a total of 10 articles were deemed eligible for inclusion in the final review. 20,23 –31
5. The TCDT search yielded 233 references, of which 120 remained after removal of the duplicate references. Through application of the first and second filters, and the addition of five articles from reference sections of other articles, a total of nine articles were deemed eligible for final inclusion. 12,32 –39
6. The THRT search produced 29 references, with 14 remaining after duplicate removal. Two were added from reference sections of relevant articles. Six articles were included in the final review. 26,40 –44
7. OHT searching produced 36 references with 28 remaining after duplicate removal. Two articles were added from relevant reference sections. In total, three articles met the inclusion criteria and were deemed eligible for inclusion in the review. 43,45,46
8. Mx-based searching produced 756 studies, with 340 remaining after removal of duplicates. After application of the first and second filtering processes, 17 articles were deemed eligible for inclusion. 18,47 –62
9. The search for continuous TF-ARI produced 756 references with 340 remaining after duplicate removal. One article was added from relevant reference sections. In total, six articles met the inclusion criteria and were deemed eligible for inclusion in the review. 59,63 –67
Technique overview and selected literature commentary
CTP/Xe-CT techniques
The use of CTP for intermittent autoregulatory assessment hinges on the acquisition of two separate scans, one before and one after a deliberate manipulation of systemic blood pressure. CTP is conducted utilizing a bolus dose of iodinated contrast agent, followed by assessment of contrast attenuation on CT as it relates to the time post-bolus, based on the known relationship between attenuation and serum concentration. 11,12 Contrast time-concentration curves are produced for arterial, venous, and parenchymal compartments for each individual patient. The manipulation of these curves yield the standard measures from CTP: cerebral blood volume (CBV), CBF, time to peak (TTP), and mean transit time (MTT). These values can be defined for various regions of interest (ROI), which may be observer defined or reflecting major arterial territories.
For autoregulatory testing, a baseline CTP is obtained, and is repeated after a manipulation of systemic blood pressure, targeting a specific mean arterial pressure (MAP) or cerebral perfusion pressure (CPP), typically 20 mm Hg above baseline. The two scans can be compared with a range of approaches, using both quantitative analysis of the source data and qualitative description of changes in the colorimetric maps. 11,12,15,16 In particular, attention is paid to what happens to CTP-based regional CBV and CBF during the physiologic manipulation. In someone with “intact” autoregulation, there should be little or no change in CBF during the change in MAP/CPP. Those with “impaired” autoregulation display a positive correlation between increased MAP/CPP and CTP-based CBF.
In adult TBI, only four studies were identified utilizing CTP as an intermittent assessment of autoregulatory status. 11,12,15,16 A total of 255 patients were described across these four studies, with mean age ranging from 32 to 37.5 years. All were conducted in severe TBI patients (i.e., Glasgow Coma Scale [GCS] ≤8). The primary goal of these studies was to evaluate CTP as a method of intermittent autoregulatory assessment. All studies found varying degrees of autoregulatory function post-TBI, 11,12,15,16 with up to 36% of patients demonstrating “impaired” autoregulatory function. 16 Study characteristics can be seen in Table 1, with outcomes reported in Table S1.
ARI = autoregulatory index, AVDO2 = arteriovenous difference in oxygen, CBF = cerebral blood flow, CBFV = cerebral blood flow velocity, CBV = cerebral blood volume, cm = centimeter, CMRO2 = cerebral metabolic rate of oxygen consumption, CPP = cerebral perfusion pressure, CO2 = carbon dioxide, CT = computed tomography, CTP = computed tomographic perfusion, CVR = cerebrovascular reactivity, HTS = hypertonic saline, ICP = intracranial pressure, MAP = mean arterial pressure, Mx = moving correlation coefficient between mean CBFV and arterial blood pressure, OHT = orthostatic hypotension test, PET = positron emission tomography, PRx = pressure reactivity index, rCBF = regional CBF, SAH = subarachnoid hemorrhage, TBI = traumatic brain injury, mTBI = mild TBI, TCD = transcranial Doppler, TCDT = thigh cuff deflation test, THRT = transient hyperemic response test, Xe = xenon,
In contrast to CTP, where CBF are estimated based on changes in image intensity following intravenous iodinated contrast, Xe-CT is based on the inhalation of a diffusible radio-opaque tracer gas (133Xenon [133Xe]) which is delivered at a concentration of ∼30%, for a duration of inhalation of 3–5 min. 14 The change in image attenuation in response to the wash-in and/or wash-out of 133Xe CT combined with estimated arterial 133Xe levels (approximated from end-tidal levels obtained from a xenon detector) allows calculation of CBF using the modified Kety–Schmidt technique. The CT scanner is equipped with a xenon detector.
To assess autoregulation, Xe-CT is conducted in a similar manner to CTP based autoregulatory assessments. A baseline Xe-CT is obtained, followed by a physiologic manipulation in MAP or CPP (typically to raise MAP/CPP ≥20 mm Hg above baseline). 7 –10,13,14 Once the target MAP/CPP is reached, a repeat Xe-CT is conducted. The two imaging sessions are compared, with comparison between the CBF calculated in predefined ROI. Autoregulatory capacity is defined as “intact” if there is little (i.e. even up to <1% change in CBF/CBV per mm Hg change in MAP/CPP) 14 change in CBF/CBV during the manipulation of MAP/CPP. Conversely, “impaired” autoregulation is defined as an increase in CBF/CBV with an increase in MAP/CPP. Various thresholds for “abnormal” autoregulation have been described, even down to 1% change per mm Hg increase in MAP/CPP, although it is currently unclear as to which thresholds definitively highlight those with autoregulatory dysfunction and those without.
Our search of the adult TBI literature produced only six studies utilizing Xe-CT for the assessment of autoregulation. 7 –10,13,14 A total of 187 patients were described within these studies, all with severe TBI. The primary goal of these studies was to assess autoregulatory capacity with Xe-CT as an intermittent technique. As with the CTP literature, the outcomes described were varied with “impaired” autoregulation and could be seen in up to 83% of the studies conducted. 10 Study characteristics can be seen in Table 1, with outcomes described in Table S1.
PET techniques
PET-based techniques for the assessment of autoregulatory capacity are routed in the measure of CBF and CBV with 15 O based tracers. 17 –22 Typically, 15 O- labelled water (H2 15 O) is utilized as the intravenous tracer agent, using either steady-state or bolus techniques. Emission scans are obtained and CBF maps are calculated using kinetic models that include measurement of arterial radiotracer concentrations. Imaging of CBV is undertaken in a similar manner, using 15 O-labelled carbon monoxide (C 15 O) as a tracer.
To assess autoregulatory capacity, CBF maps are obtained at baseline physiology and following a change in CPP, with a goal of an ∼20 mm Hg increase in CPP from baseline. 19 –21 Similar to CT studies, autoregulation is defined as “intact” if there is little or no change in CBF/CBV during CPP manipulations. Conversely, “impaired” autoregulation is defined as an increase in CBF/CBV during the CPP elevation.
Our search of the adult literature produced only six articles utilizing 15 O PET for the intermittent assessment of autoregulatory capacity. 17 –22 All of these articles came from a single center of excellence in neuroimaging and TBI research. A total of 119 patients were described in these studies; however, there may be an overlap in patient reporting among studies, so the total number of unique patients described is likely smaller. The primary goal of these studies varied: some focused on pericontusional CBF during manipulations in CPP, 21 whereas others were addressing autoregulation with 15 O PET based and other continuous techniques. 17 –20,22 Given that the focus of the PET-based studies was on predefined ROIs, the overall incidence of “impaired” autoregulatory status was not mentioned; only quantitative CBF numbers were described during testing. Further details on study characteristics can be seen in Table 1, whereas the specific outcome data can be seen in Table S1.
MRI-based techniques
The use of MRI-based techniques for intermittent autoregulatory assessment is lacking in adult TBI, as was highlighted by our systematic review failing to produce a single article on the use of MRI techniques for this purpose in adult TBI. This is in contrast to a large and growing literature base on CVR measurement using CO2 challenge, accompanied by imaging using blood oxygen level dependent (BOLD) fMRI or more quantitative estimation of CBF (e.g., using arterial spinal labelling) across the severity of TBI. 68,69 It is important to emphasize that such techniques do not measure autoregulation, but rather “probe” CVR in response to CO2 changes. It is important to emphasize this, as CVR and pressure autoregulatory capacity are distinct physiologic properties of the cerebral vasculature and may not be entirely interlinked. 41,70
The lack of studies utilizing MRI for intermittent autoregulatory assessment likely stems from issues in feasibility, cost, complexity of patient care while in the magnet, and the presence of alternatives that are faster and cheaper and yield quantitative assessments of CBF. Further, many MR perfusion (MRP)-based studies utilize gadolinium-based contrast boluses, which carry potential risk. 71 The development of arterial-spin-labelling (ASL) techniques for a gadolinium-free-based assessment of MRP is an exciting direction in which to take MRP-based autoregulation assessments. 72 However, to date, the literature on ASL MRP is focused on animal models and small clinical series, without mention of autoregulatory testing. 72
There does exist room for further exploration of MRI as a modality for intermittent autoregulatory assessment. With the use of MRI during manipulations in MAP/CPP, via vasopressors or TCDT (see subsequent technique description), one could conduct baseline and post-manipulation MRP studies that would allow for a comparison of CBF. A TCDT base MRI assessment of autoregulation has been described in healthy humans. 73
AVDO2 technique
The difference between arterial and venous blood oxygen content can be utilized in the intermittent assessment of autoregulatory capacity. The technique involves sampling arterial blood, typically from a radial arterial line, and venous blood from a catheter with the tip situated in the jugular bulb. The position of this jugular venous catheter can be confirmed via anterior-posterior plain radiography, with the optimal position near or just below the level of the mastoid process but above the level of the arch of C1. 31 The oxygen saturation is then determined within both the arterial and venous samples, with the AVDO2 value calculated as: AVDO2 = 1.34 × Hb[(SaO2 − SjO2)/100]; where Hb = hemoglobin level (grams per deciliter), SaO2 = % saturation of oxyhemoglobin in arterial blood (micromole/milliliter), and SjO2 = % saturation of oxyhemoglobin in venous blood (micromole/milliliter). 29 –31 With the AVDO2, CBF can be estimated as: 1/AVDO2; assuming a steady cerebral metabolic rate of oxygen consumption (CMRO2). 29 –31
Autoregulation is then determined based on estimations of CBF using the AVDO2 values at baseline and during a manipulation of MAP/CPP, with a goal increase in MAP/CPP of ∼20–30% from baseline. Patients with “intact” autoregulation should have a negligible (i.e., <20% change from baseline) change in 1/AVDO2 with MAP/CPP manipulation, whereas those with “impaired” autoregulation should display a significant change in 1/AVDO2 from baseline (i.e., ≤20% or more increase). 31
Our search of the adult TBI literature identified 10 studies documenting the application of the AVDO2 techniques for intermittent autoregulatory assessment. 20,23 –31 A total of 258 patients were described across these 10 studies with moderate–severe TBI and a mean age of ∼33 years in each study. The main outcome for each of these studies was to characterize autoregulatory status in TBI patients at baseline and during various interventions, including indomethacin therapy 26,27 and CO2 challenges. 27 Various rates of autoregulatory impairment and poor flow-metabolism coupling were documented across the studies. 20,23 –31 The incidence of “impaired” autoregulatory function within the AVDO2 studies was sparsely documented. Those studies that documented the number of patients with impaired autoregulatory function recorded rates of ∼60–70% in the small populations described. Further details on study characteristics can be seen in Table 1, whereas autoregulatory outcomes can be seen in Table S1.
ARI–TCDT
The TCDT for intermittent autoregulatory assessment in humans was originally described in 1989. 74 The classic form of the technique involves the use of bilateral thigh cuffs inflated for 2–3 min above systolic blood pressure, producing leg hyperemia and, systemically, a vasodilatory state given the reduction in venous return and cardiac preload. Both cuffs are then rapidly deflated. A dramatic drop in the MAP, sustained for a period of up to 30 sec, is subsequently seen, with an “adequate” test defined as at least a 15–20 mm Hg drop in pressure from baseline. This step decrease in blood pressure is subsequently followed by a reflex cerebral vascular response to correct for the change in blood pressure. Cerebral blood flow velocity (CBFV) within the middle cerebral artery (MCA) is measured using TCD at baseline and throughout the duration of TCDT, with cerebrovascular resistance (CVRes) estimated by dividing MAP by CBFV.
Autoregulation can then be measured in two main ways with these data from TCDT. First, one can calculate the rate of regulation (RoR) by: RoR = (change in CVRes/time interval)/change in MAP. Various thresholds are described including an RoR of 0.2/sec being characterized as “normal.” 74 The second commonly quoted method of autoregulation measurement with TCDT data is the determination of ARI, introduced by Tiecks and coworkers in 1995. 75 This method uses a second order differential model of cerebral autoregulation governed by three physical properties, the time constant, the damping factor, and the autoregulatory gain. For practical reasons, 10 sets of parameter values were defined grading the overall “strength” of autoregulation on a scale of 0 to 9, with 0 corresponding to complete loss of autoregulation, and 9 corresponding to hyperactive regulatory response. On this scale, normal autoregulatory capacity is defined as an ARI of 4–7, and an abnormal one is defined as ≤3. 32,75 The TCD-based CBFV changes recorded during the step change in MAP are normalized and compared with the model-generated responses corresponding to the 10 grades in order to determine which model response constitutes the “best fit,” and the grade of that response is returned as ARI. 75
Given that ARI effectively superseded the RoR metric with respect to analysis of TCDT, we decided not to pursue the latter any further in our literature search, and to concentrate only on ARI. Within the adult literature, there are nine articles describing the application of an intermittent TCDT ARI measurement as a means of autoregulation assessment in adult TBI. 12,32 –39 A total of 522 patients were described within these nine studies, with a mean age ranging from 32 to 56 years. All patients had severe TBI. The main focus of these studies was to describe the assessment of autoregulatory capacity utilizing the TCDT-based ARI calculation. Varying rates of “impaired” ARI (i.e., ARI <3) were described across the studies, with some reporting as much as ≥80% of impaired values 35,39 Further details on study characteristics and ARI-based outcomes can be found in Table 1 and Table S1, respectively. A continuous method of ARI measurement, not requiring any intermittent interventions, has also been developed, which will be discussed in more detail in Part II of this article series.
THRT
THRT was first described in humans TBI by Giller in 1991. 40 In this method, TCD-based MCA velocities are measured continuously on the side of interest for autoregulatory testing. The test involves a brief compression (all the way to total occlusion, if possible) of the ipsilateral internal carotid artery in the neck. Adequate compression occurs when the MCA velocities fall to 30–50% of baseline values. The carotid compression is released after 3 sec of stable FV reduction, while any changes in MCA velocity are measured with TCD. An increase in systolic CBFV of ∼10% or more from baseline values is considered significant and a sign of “intact” autoregulatory capacity. 40 The entire premise of this test is to provide transient hyperemia once the carotid compression is released, with the attempt to evaluate the cerebrovascular autoregulatory capabilities. Undershoot of CBFV (suggest to be <10%) on TCD indicates impaired autoregulatory function. It is unclear if excessive overshoot of CBFV (i.e., values well in excess of 10%) signify an important aspect of autoregulatory capacity. This concept of “at least a 10% increase in CBFV” during a THRT indicating intact autoregulation has been validated in small cohorts of neurosurgical patients with varying pathology. 40
Our search of the adult TBI literature yielded only six studies documenting the use of a THRT for the intermittent assessment of autoregulatory function. 26,40 –44 A total of 116 patients with moderate to severe TBI were described across the six studies. The primary outcome of interest for most of these articles was the documentation of autoregulatory capacity post-TBI. 40,42 –44 One study compared CVR to THRT-based autoregulation and Xe-CT. 41 One study looked at the impact of intravenous indomethacin on THRT-defined autoregulation. 26 Varying degrees of THRT-based autoregulatory “impairment” were documented, with up to ∼70% of patients displaying an “impaired” THRT. 26 CVR was found to have varying associations with THRT autoregulation, during which autoregulation could be “impaired” with intact CVR, or the converse. 41 One study found no correlation between THRT results and those of TCDT and OHT. 43 Further details on study design can be found in Table 1, whereas the details on autoregulation-based outcomes in these studies can be found in Table S1.
OHT
The OHT has been utilized in a limited fashion within TBI. 43,45,46 The lack of application likely stems from the complex care that moderate to severe TBI patients undergo within the ICU and the need to change patient position for autoregulatory testing. The test is completed in the following steps. First, TCD-based CBFV of the MCA is obtained at baseline with the bed set at a 0 degree tilt. Next, a sudden change (typically ≤3 sec) in head position is conducted, with varying levels of tilt from 30 up to 80 degrees. 43,45,46 A drop in CBFV of >20% with a raise in head position may be deemed abnormal, as a 10–15% change has been quoted for normal patient populations in the past. 76,77 The test is typically preformed multiple times, with varying durations of break in between testing periods.
Searching the literature produced only three articles documenting the use of OHT as an intermittent autoregulatory assessment in adult TBI patients. 43,45,46 Fifty-one patients with varying degrees of TBI severity were assessed with OHT across the three studies. The primary outcome of these studies was to describe OHT-based autoregulatory status post-TBI. Outcomes were mixed and unclear as to the accuracy of autoregulation assessment. One study found a weak correlation (r = 0.23) between TCDT and OHT. 43 The exact incidence of “impaired” autoregulation as measured from OHT is difficult to discern from the included literature, given that most of the studies were designed to compare OHT with other forms of intermittent autoregulatory testing. Further details on these studies can be found in Table 1 and Table S1.
TCD-based Mx
This index is derived from a moving Pearson correlation coefficient between two continuously measured signals, mean CBFV (mFV) and CPP. Given the need for continuous TCD monitoring to calculate this updating index, the duration of recording is typically limited to ≤60 min. Therefore, we refer to this technique as a “semi-intermittent” autoregulatory measurement technique.
The concept of this autoregulatory measure is that the response in mFV to slow wave changes in CPP are believed to be governed by the autoregulatory capacity. Transcranial Doppler ultrasound is used to capture the MCA velocity. As with all the other TCD-based measures, the signal can be acquired either unilaterally or bilaterally, thus providing insight into the symmetry of cerebral autoregulation. In addition, continuous recording of MAP (via either invasive arterial line or continuous noninvasive techniques) and ICP are required for determination of CPP (based on: CPP = MAP – ICP). The signals are recorded and stored, with the option of both off-line and real-time calculation of Mx. Both mFV and CPP are processed via a 10 sec moving average filter and sampling frequency decimated to 0.1 Hz (in other words replacing original mFV and CPP by time series composed of their non-overlapping 10 sec averages). The correlation coefficient is then calculated based on 30 consecutive 10 sec mean values (i.e., 5 min), with the coefficient updated every 10 sec. This produces an Mx time series with the same sampling rate as mean mFV/CPP. Variations in the correlation window length have been described, although the abovementioned parameters are the most commonly mentioned, particularly in more recent years. Similar to pressure reactivity index (PRx – derived from invasive ICP and MAP), Mx in routine autoregulation monitoring should be averaged at minimum for a period of 30 min, to reduce its inherent variance.
Some variations in the technique have been described. The use of MAP instead of CPP has been explored (termed Mxa), as this renders the technique potentially noninvasive. 51,57,62 However, by using MAP and not CPP, the impact of ICP on mFV is not considered during the calculation of the correlation coefficient. Literature comparing standard Mx to Mxa has demonstrated that Mx is superior in terms of outcome prediction, 51 hence Mxa is limited to those cases in which ICP is not directly monitored. Similarly, instead of using mFV in the correlation, there has been investigation into utilizing either only systolic or diastolic flow velocities, creating two new indices (Sx and Dx, respectively). Again, as with Mxa, Sx and Dx are not grossly superior to Mx. 50,51 Therefore, Mx is most commonly quoted.
The value generated by calculating Mx ranges from −1 to +1, as it is a correlation coefficient. A value that is highly positive indicates potentially “impaired” autoregulation, as it signifies a passive, linear realationship between CPP and mFV. Similarly, a non-positive Mx value indicates that fluctuations in CPP are heavily attenuated and phase shifted in FV, therefore suggesting “intact” autoregulation. The exact Mx value at which autoregulation is “impaired” and pathological is not clear. Sorrentino and coworkers evaluated thresholds of Mx as they related to mortality and Glasgow Outcome Scale (GOS) score at 6 months post-TBI. 62 A double peaked threshold was found for Mx with threshold of 0.05 and 0.3 discriminating between both life and death, and between good and poor outcome. 41 Both thresholds are utilized within the literature to denote “impaired” autoregulation.
Our search of the adult TBI literature yielded 17 studies, with ≥50 or more subjects, describing the association between Mx and patient outcome. 18,47 –62 The patients had moderate to severe TBI, and there was an average of 212 patients per study. It is of note that all of these articles originated from Addenbrooke's Hospital in Cambridge, as part of either retrospective reviews of a prospectively maintained database on TCD in TBI, or as prospective cohort studies. Therefore, the actual number of unique patients overall across all of the studies is much lower, because there is overlap in patient inclusion across the multiple studies. Despite this, across these 17 articles, Mx is noted to be strongly associated with patient outcome. Further, given the ability to obtain MCA velocities bilaterally, asymmetry in autoregulatory capacity has been documented, with the greater right–left difference in Mx being associated with worse outcome. 61 The incidence of “impaired” autoregulation as defined by Mx was not clear in most of the included studies, as the search for TCD-based Mx studies was for those documenting an association with patient functional outcome. In addition, “impaired” autoregulation detection using Mx is difficult, given the lack of an exact threshold for this physiologic event, and only thresholds for morbidity and mortality that have been well defined. For further detail on the main studies on Mx in adult TBI please refer to Table 2 and Table S2.
ABP = arterial blood pressure, ARI = autoregulatory index, CPP = cerebral perfusion pressure, CPPopt = optimal CPP, CT = computed tomography, Dx = reactivity index between diastolic TCD flow velocity and ABP, GOS = Glasgow Outcome Scale, GOSE = GOS extended, HRV = heart rate variability, ICP = intracranial pressure, ICU = intensive care unit, LF = low frequency, LPR = lactate:pyruvate ratio, L-PRx = long PRx, mm Hg = millimeters of mercury, MAP = mean arterial pressure, Mx = reactivity index between mean flow velocity via TCD and CPP, Mxa = MAP based Mx, PAx = reactivity index between pulse amplitude of ICP and ABP, PRx = pressure reactivity index, Sx = reactivity index between systolic TCD flow velocity and CPP, TBI = traumatic brain injury, TF-ARI = transfer function autoregulation index
TF-ARI
The determination of an ARI value has been discussed in the ARI–TCDT section of this article, and is based on the work from Aaslid and coworkers 74 and Tiecks and coworkers. 75 Classically, this index of autoregulation is an intermittent measure that is obtained through manipulation of the patients' physiology via a step response in MAP. Recently, TCD waveform analysis techniques have enabled production of a continuously updating ARI for a patient based on MCA CBFV and continuous MAP measures. Given dependency on TCD monitoring in order to calculate this updating index (similar to Mx), the duration of recording is typically limited to ≤60 min. Therefore, we refer to this technique as a “semi-intermittent” autoregulatory measurement technique.
The technique is complex and requires application of the Fourier transform (FT) algorithm to the beat-to-beat changes in mean arterial blood pressure (ABP)/CPP and mean CBFV. 64 The following technique is based on the study by Liu and coworkers. 64 CBFV and ABP/CPP are normalized into z-scores (i.e., each value subtracted from the mean, and subsequently divided by the standard deviation). Transfer function is calculated as frequency spectrum of FV signal divided by frequency spectrum of ABP. Inverse FT of transfer function can be used to obtain the CBFV impulse response in time, which can be integrated in time to yield an estimated CBFV response to “step change” in ABP/CPP. This response can then be fitted to the 10 responses of the ARI model proposed by Tiecks and coworkers. 75 The calculation may be conducted with 300 sec of data, updated every 10 sec. This yields a continuously updating ARI, the TF-ARI. The benefit of this method is that it may be conducted entirely noninvasively, assuming the use of MAP for the input.
This is a relatively little used method of semi-intermittent autoregulatory assessment in the context of TBI. Our search of the adult TBI literature produced six studies describing TF-ARI, 59,63 –67 and most of them were very recent. The largest studies were conducted retrospectively on a prospectively maintained database of TCD in adult TBI patients. 59,63,66 Only one study originated outside of the University of Cambridge, describing TF-ARI in 32 adult TBI patients. 65 The primary goal of each of these studies varied, with some measuring the association with Mx, 59,63,66,67 and others looking at the association with patient outcome. 59,65,66 Mx and TF-ARI display a strong negative correlation, 59,63 with TF-ARI associated with GOS score at 6–12 months post-TBI. 58,65 Given that this technique has only been described in a few studies, the incidence of “impaired” autoregulation based on TF-ARI is difficult to estimate based on the current literature body. For further details on TF-ARI in adult TBI, please see Table 2 and Table S2.
Discussion
Through this scoping review, we have been able to outline both the methodology and the available literature for the main intermittent/semi-intermittent autoregulatory techniques in adult TBI. We hope that this article will serve as both a reference for technique methodology and a comprehensive reference library for literature on these techniques in adult TBI. As such, the individual studies were not commented on in detail within the body of this article, given that this was not the goal. For those interested, tabulated study outcome data regarding autoregulation-based outcomes can be found in the supplementary material.
Other than Mx, a small literature base exists for the abovementioned techniques, with CTP/Xe-CT, PET, AVDO2, and TCDT ARI dominating the literature on intermittent/semi-intermittent autoregulation assessment in adult TBI. Most studies focused on small patient populations in the moderate to severe TBI cohort. The exact incidence of “impaired” autoregulatory capacity, as determined by each intermittent technique, varied and was not clear in some cases. Overall, AVDO2/Xe-CT/ARI/THRT recorded impaired autoregulation in 60– 80% of patients tested/studies conducted. CTP-based assessment only recorded impaired autoregulation in up to 36% of studies. PET, OHT, and TF-ARI literature lacked clear descriptions of the exact incidence of impaired autoregulation.
The overarching premise behind all of the intermittent (i.e., excluding Mx and TF-ARI) techniques is manipulation of systemic blood pressure/blood volume via either chemical (such as vasopressors) or mechanical (such as thigh cuffs or carotid compression) means. The method for assessing the cerebral circulation during these manipulations varies, with both imaging based techniques and TCD being used. One major concern with the intermittent techniques is the lack of standardization in testing protocols. Further, the exact definition of “intact” versus “impaired” autoregulatory status varies depending on the study (with the exception of TCDT ARI-based studies).
It is of note that the literature highlighting Mx mainly stems from a single center, Addenbrooke's Hospital at the University of Cambridge, with a small number of articles on Mx (with >50 patients) arising outside of this center. Therefore, despite the seemingly large number of large studies on Mx, one must understand that many of these studies are reporting overlapping patient cohorts. This highlights the need for multi-center collaboration in brain physics. With the development of a “brain physics consortium,” this could allow for multi-center sharing of high resolution monitoring data and could potentially address some of the issues with multiple single center retrospective studies conducted on identical patient cohorts.
Despite the somewhat small literature base for intermittent/semi-intermittent techniques (minus Mx) in adult TBI, we feel that it is important to acknowledge the availability of such tests given that they have provided fundamental insight into human autoregulatory capacity in various neuropathological states, leading the development of continuous and more commonly applied techniques in the modern ICU. Further, these intermittent techniques still provide valuable information on the autoregulatory status of a patient, without the need for invasive monitoring devices (with the exception of the AVDO2 technique and Mx derived from CPP) as is required in many continuous techniques. Therefore, these intermittent methods can play an important role in the assessment of autoregulation during follow-up for TBI patients, and those with other neuropathology. Finally, another advantage of the intermittent techniques specifically is the clarity of what constitutes input and what is the output (response); that is, that the causality is ensured. As will be seen in Part II of this series, the continuous index-based assessments of autoregulation are based on the correlation between continuously monitoring variables, such as ICP and MAP, typically without an imposed stimulus (i.e., a vasopressor-induced elevation in MAP). Therefore, with a large MAP/CPP based stimulus seen during the intermittent techniques, one can potentially achieve more reliable results because of a higher “signal-to-noise” ratio.
If one were to decide to undertake intermittent/semi-intermittent autoregulation testing, the current literature favors the use of ARI/TF-ARI-, TCDT-, or TCD-based Mx at the current time. There are a few reasons for this. First, these techniques have the largest literature base supporting their accuracy and reproducibility, with validation from multiple centers and application in other types of neuropathology. Second, the imaging-based approaches are cumbersome, require post-acquisition processing and appropriate timing of radiotracer/contrast administration in order to obtain a viable image. Further, hemodynamic or metabolic manipulation is typically required, leading to a more labor-intensive process, with a small literature base to guide one's definition of “intact” versus “impaired” autoregulation. Finally, the relative ease of implementation of ARI/TF-ARI-, TCDT-, and TCD-based Mx, allows for easily repeated measures in the same patient, in order to confirm one's results. The same “repetition” technique is not available for most imaging-based modalities.
Given the advent of continuous measures of autoregulatory capacity, such as the PRx, the literature on autoregulation in TBI has shifted to the application of these continuous techniques. This is despite the potential advantages of the intermittent/semi-intermittent techniques, as described. The commonly applied continuous techniques will be addressed in detail in Part II of this article series.
Conclusions
Numerous methods of intermittent/semi-intermittent pressure autoregulation assessment in adult TBI exist, including: CTP/Xe-CT, PET, AVDO2 technique, TCDT-based ARI, THRT, OHT, Mx, and TF-ARI. MRI-based techniques in adult TBI are yet to be described, with the main focus of MRI techniques being on metabolic-based CVR and not pressure-based autoregulation. The techniques of intermittent and semi-intermittent cerebrovascular autoregulation testing reviewed in this article have provided the basis for the development of continuous techniques now employed in many ICU settings. These methods may still play a role in follow-up and noninvasive assessment of autoregulatory capacity in adult TBI.
Footnotes
Acknowledgments
This work was made possible through salary support through the Cambridge Commonwealth Trust Scholarship, the Royal College of Surgeons of Canada – Harry S. Morton Travelling Fellowship in Surgery, the University of Manitoba Clinician Investigator Program, R. Samuel McLaughlin Research and Education Award, the Manitoba Medical Service Foundation, and the University of Manitoba Faculty of Medicine Dean's Fellowship Fund. J.D. is supported by a Woolf Fisher Scholarship (NZ). These studies were also supported by the National Institute for Healthcare Research (NIHR), UK through the Acute Brain Injury and Repair theme of the Cambridge NIHR Biomedical Research Centre, an NIHR Senior Investigator Award to D.K.M. Authors were also supported by a European Union Framework Program 7 grant (CENTER-TBI; Grant Agreement No. 602150).
Author Disclosure Statement
F.A.Z. has received salary support for dedicated research time, during which this project was partially completed. Such salary support came from: the Cambridge Commonwealth Trust Scholarship, the Royal College of Surgeons of Canada – Harry S. Morton Travelling Fellowship in Surgery, the University of Manitoba Clinician Investigator Program, R. Samuel McLaughlin Research and Education Award, the Manitoba Medical Service Foundation, and the University of Manitoba – Faculty of Medicine Dean's Fellowship Fund. M.C. and P.S. have financial interest in a part of the licensing fee for ICM+ software (Cambridge Enterprise Ltd, UK). M.C. is an honorary co-director of Technicam Ltd., producer of the Cranial Access Device used for cerebral microdialysis (CMD) insertion. The other authors have nothing to disclose.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
