Abstract
Preservation of optimal cerebral perfusion is a crucial part of the acute management after aneurysmal subarachnoid hemorrhage (aSAH). A few studies indicated possible benefits of maintaining a cerebral perfusion pressure (CPP) near the calculated optimal CPP (CPPopt), representing an individually optimal condition at which cerebral autoregulation functions at its best. This retrospective observational monocenter study was conducted to investigate, whether “suboptimal” perfusion with actual CPP deviating from CPPopt correlates with perfusion deficits detected by CT-perfusion (CTP). A consecutive cohort of aSAH-patients was reviewed and patients with available parameters for CPPopt-calculation, who simultaneously received CTP, were analyzed. By plotting the pressure reactivity index (PRx) versus CPP, CPP correlating the lowest PRx value was identified as CPPopt. Perfusion deficits on CTP were documented. In 86 out of 324 patients, the inclusion criteria were met. Perfusion deficits were detected in 47% (40/86) of patients. In 43% of patients, CPP was lower than CPPopt, which correlated with detected perfusion deficits (r = 0.23, p = 0.03). Perfusion deficits were found in 62% of patients with CPP<CPPopt compared to 34% in patients without deviation or CPP>CPPopt (OR 3, p = 0.01). These findings support the hypothesis, that a deviation of CPP from CPPopt is an indicator of suboptimal cerebral perfusion.
Keywords
Introduction
Aneurysmal subarachnoid hemorrhage (aSAH) leads to acute disturbance of cerebral perfusion by a sudden increase in intracranial pressure (ICP) and a decline of cerebral perfusion pressure (CPP). 1 This pathophysiological reaction usually restores within hours after ictus depending on the bleeding’s severity and the function of self-protecting mechanisms, like the cerebrovascular autoregulatory capacity. 2 Intact cerebral autoregulation at the microcirculatory level can prevent brain injury by maintaining a relatively constant cerebral blood flow despite CPP variations. 3 Since CPP is regarded as a general indicator of brain perfusion, continuous CPP monitoring is deemed imperative during the management of high-grade aSAH patients. The clinical implication of the CPP concept is limited by the lack of direct consideration of cerebral autoregulation, which is often impaired after aSAH.4,5 Monitoring cerebral autoregulation may help in establishing individualized treatment targets in patients with cerebrovascular diseases.6,7 Setting individualized targets based on cerebral autoregulation like an individually calculated optimal CPP (CPPopt) introduces a transition from the “one treatment fits all” paradigm derived from current guidelines toward a personalized medicine approach tailored to the individual patient. Aside from the fact that CPP monitoring does not directly consider the functional status of cerebral autoregulation, CPP also represents a volume surrogate and not a direct measure of cerebral blood flow. Thus, a correlation of CPP and CPPopt with techniques allowing direct assessment of cerebral perfusion like computed tomography perfusion (CTP) is critical for validation of their clinical value. CTP provides valuable real-time information about cerebral microcirculation, which is the reason why CTP is used for the prediction and detection of delayed cerebral ischemia after aSAH.8,9 A few studies indicated possible benefits of maintaining a CPP near the calculated CPPopt, representing an individually optimal condition at which cerebral autoregulation functions at its best.10–12 However, the CPPopt concept has not been adopted yet in clinical practice. This study aimed to investigate, whether “suboptimal” perfusion with actual CPP deviating from the calculated CPPopt correlates with perfusion deficits detected by CTP.
Material and methods
Study population and study design
This is a retrospective observational single-center study. A consecutive patient cohort with aSAH treated at our center in the time between January 2012 and December 2020 was reviewed and patients with available parameters for CPPopt-calculation, who simultaneously received CTP during the acute phase after the bleeding, were eligible for inclusion in the study. Since continuous CPP monitoring requires an invasive ICP monitoring only patients with ICP monitoring could be enrolled in the study. For data consistency reasons, only patients with ICP monitoring via intraparenchymal ICP probe were considered in this study. The inclusion criteria were age ≥18 years, confirmed aSAH, continuous invasive ICP monitoring and invasive blood pressure monitoring, and a CTP scan during the acute phase after ictus (first 14 days after aneurysm rupture). Patients with non-aneurysmal SAH, age <18 years, without continuous ICP monitoring, and without a CTP scan during the acute phase after the bleeding event were excluded. All patients were treated according to a predefined management protocol. 13 For ICP monitoring an intraparenchymal ICP probe (Codman Microsensors ICP Transducer; Codman & Shurtleff Inc) was inserted into the right frontal lobe. Mean arterial pressure was monitored with a standard pressure monitoring kit (DTXPLUS® Becton Dickinson Infusion Therapy Systems Inc) with the transducer referenced to the foramen of Monro. For multimodal data acquisition, IntelliSpace Critical Care and Anesthesia (ICCA) ICU software was used. The pressure reactivity index (PRx) was calculated as the correlation coefficient between ICP and mean arterial blood pressure. By plotting PRx versus CPP, CPP correlating the lowest PRx value, was identified as CPPopt. The CTP parameters were qualitatively and quantitatively assessed according to the previously described protocol, and perfusion deficits were documented. 13 Following cutoff values were applied: cerebral blood flow <53.93 ml/100 ml/min, cerebral blood volume (CBV) <3.14 ml/100 ml, mean transit time (MTT) >4.25 s, time to start (TTS) >0.94 s, time to peak (TTP) >9.28 s, time to drain (TTD) >4.93 s. 9 This is a retrospective observational study. No study-specific interventions were performed. All procedures were in accordance with the ethical standards of the responsible committee on human experimentation and with the Helsinki Declaration of 1975, as revised in 2013. 14 The study was approved by the local institutional review board of the University Medical Center Göttingen (16/9/20, approval date 14.09.2020). Due to the retrospective study design, informed consent was not required.
Statistical analysis
The statistical analyses were performed by means of the GraphPad Prism software (Version 9, GraphPad Software, San Diego, CA, USA). For the presentation of baseline data, descriptive statistics and frequency distribution analysis were done. Continuous variables are depicted as mean ± standard deviation (SD), and categorical variables as frequency or percentages. Descriptive statistics was used for the calculation of baseline characteristics in the study population. Fisher’s exact test was performed to calculate odds ratios (OR), sensitivity, and specificity. A p-value of <0.05 was considered statistically significant.
Results
Patient characteristics
In 86 out of 324 patients, the inclusion criteria were met. The mean age was 54 ± 11 years, 65% of whom were female. The median WFNS grade was 4, and all patients had a high Fisher grade of ≥3. A summary of patients’ baseline characteristics is given in Table 1.
Baseline characteristics of study population.
Cerebral perfusion on CTP vs CPP-CPPopt deviations
In 79% of patients, CTP scans were routinely performed in the early phase at the beginning of the vasospasm phase (day 3–5), and in 21% of cases during the peak phase of vasospasm (day 7–9). The average time of CTP assessment in relation to the admission day was 4 days (SD 1.9 days). Perfusion deficits were detected in 47% (40/86) of all patients. Increased blood flow velocity as measured by transcranial Doppler sonography (TCD) at the time of CTP had 26% (12/46) of patients without perfusion deficits and 38% (15/40) of patients with perfusion deficits on CTP. In 43% of analyzed patients, CPP was lower than CPPopt, which correlated with a higher rate of detected perfusion deficits on CTP (r = 0.23, p = 0.03). Perfusion deficits were found in 62% of patients with CPP < CPPopt compared to 34% in patients without deviation or CPP > CPPopt (OR 3, p = 0.01). A scatter plot depicting the deviations of CPP from CPPopt in patients with perfusion deficits compared to those without perfusion deficits is shown in Figure 1. The detected perfusion deficits on CTP were reversible in 55% (22/40) of patients and irreversible in 45% (18/40), resulting in delayed infarction (Supplemental material with examples of reversible and irreversible perfusion deficits was uploaded as Figures 2 and 3). The rate of delayed infarctions at discharge in the patient group with perfusion deficits was significantly higher compared to that in the group without perfusion deficits (45% vs. 11%, p = 0.0003). Patients with reversible perfusion deficits had less frequent deviations of CPP from CPPopt (45% vs. 72%, p = 0.09) and the deviations were smaller compared to patients with irreversible perfusion deficits, but the difference did not reach significance (mean −0.33 vs 1.73, p = 0.11).
The scatter plot is showing the deviation of CPP from CPPopt in the patients with perfusion deficits compared to those without perfusion deficits with a significantly higher incidence of perfusion deficits in the case of CPP < CPPopt (p = 0.001).
Examples (1. Example, on the left side, 2. Example, on the right side) of two patients with reversible perfusion deficits with depicted CT scan on admission (top) with red arrows pointing at the initial blood distribution, CT perfusion findings (middle) with red arrows pointing at the detected perfusion deficits, and CT scans at discharge (bottom) showing no infarctions (bottom).
Examples (1. Example, on the left side, 2. Example, on the right side) of two patients with irreversible perfusion deficits with depicted CT scan on admission (top) with red arrows pointing to the initial blood distribution, CT perfusion findings (middle) with red arrows pointing at the detected perfusion deficits, and CT scans at discharge (bottom) with red arrows pointing at the detected delayed infarctions (bottom).
Discussion
This study demonstrated significant differences in rates of perfusion deficits detected by CTP between patients exhibiting CPP values near the individually calculated CPPopt and those with CPP lower than CPPopt. While several techniques have been already introduced for studying cerebral perfusion, like magnetic resonance perfusion, Xenon computed tomography, positron emission tomography, and single photon emission computed tomography, their routine use in clinical practice remained limited by their availability and cost. 15 Computed tomography perfusion (CTP) is a widely available technique, that allows qualitative and quantitative real-time assessment of cerebral perfusion. Its main limitation is the radiation exposure, emphasizing the need for clearly defined indications and good patient selection. 16 During the last two decades, CTP has become an integral part of imaging protocols for acute management of aSAH patients allowing prediction and detection of delayed cerebral ischemia.10,13 While CTP represents a flow-related surrogate for cerebral perfusion providing direct cerebral blood flow measurements, CPPopt is a volume-related surrogate for cerebral perfusion derived from arterial blood pressure and ICP under consideration of cerebral autoregulation status. The relationship between these two surrogate parameters for cerebral perfusion has not been evaluated yet in aSAH patients. With this study, we aimed to investigate, whether a deviation of CPP from individually calculated CPPopt indeed reflects perfusion deficits as measured by CTP. The results of our study confirmed a significant correlation between the deviation of actual CPP from calculated CPPopt and the detection of perfusion deficits on CTP. These findings are supported by a previously published work, where a significant correlation of invasive CPP measurements with CTP parameters as well as with the extent of CTP-abnormalities in patients with brain trauma has been demonstrated. 17 Differences between successive CTP measurements showed a strong correlation with the corresponding differences in CPP values. Patients in whom brain perfusion was strongly dependent on CPP values, were deemed to have impaired cerebral autoregulation. Conversely, in patients with relatively independent brain perfusion from CPP values, cerebral autoregulation seemed to be preserved.17,18 Lagares et al. found a correlation of increased circulation times with worse neurological function of the patients. 19 The mechanism related to perfusion abnormalities is thought to be induced by a dysfunction of cerebral microcirculation, which can be captured by CTP. In our study, patients with CPP < CPPopt had a threefold higher probability of a perfusion deficit on CTP compared to patients with CPP near the calculated CPPopt. Since the study was not powered to answer the causality question, of whether keeping a CPP near the CPPopt would possibly prevent the occurrence of perfusion deficits, we cannot draw any conclusions concerning this subject. Although the deviations of CPP from CPPopt found in patients with irreversible perfusion deficits were higher compared to those with reversible perfusion deficits, the difference was not significant. Hence, our data does not allow a definition of cutoff values with a reliable discrimination between deviations of CPP from CPPopt leading only to reversible perfusion deficits and deviations, that will result in irreversible perfusion deficits. The reason for this could be the still small number of included patients. Overall, the findings of our study support the hypothesis, that a CPPopt-adapted management of aSAH patients may be beneficial for preventing perfusion deficits, which can be detected by CTP. Johnson et al. have performed a retrospective analysis of 82 aSAH patients, in whom data of Xenon CT-scans were available, and correlated these cerebral blood flow measurements by the Xenon CT with deviations of actual CPP from calculated CPPopt. The authors reported an association of CPP lower than CPPopt with a higher number of regions with low CBF on Xenon CT. 20 These results are in line with the findings in our study. Furthermore, the study results encourage further analysis to better define diagnostic and therapeutic consequences for preventing irreversible perfusion deficits ending in delayed infarctions.
Implications of measuring cerebral autoregulation in clinical practice
A deviation of CPP from calculated CPPopt was already reported to occur before the manifestation of delayed cerebral ischemia.12,21 Furthermore, previous studies have demonstrated a correlation of CPP-deviation from CPPopt and the PRx with functional outcome, whereby a PRx > 0.2 and a CPP < CPPopt were associated with worse outcomes at 6 months follow up. 11 Another study found better long-term outcomes at 12 months follow-up in patients with lower ICP values in the first days after aSAH and with higher CPP values during the vasospasm phase. 22 Considering this, the establishment of continuous cerebral autoregulation monitoring tools for clinical management of aSAH patients seems to have the potential of preventing delayed cerebral ischemia and improving functional outcome. 12 Several techniques are currently available for continuous monitoring of cerebral autoregulation, like TCD, near-infrared spectroscopy (NIRS), and the calculation of PRx, which was applied in our study. While TCD is a wide-studied method for assessment of cerebral autoregulation (by calculation of mean velocity index = Mx, as a correlation of blood flow velocity of the middle cerebral artery with mean arterial blood pressure or with CPP), the role of NIRS needs to be defined yet. On the contrary, NIRS measures regional cerebral oxygenation saturation (rSO2), whose correlation with mean arterial blood pressure allows the calculation of cerebral oximetry index (Cox) as a surrogate for cerebral autoregulation. 19 A good correlation of Cox with Mx was shown in comatose patients with acute brain injury in a retrospective comparison study. 23 Detected fluctuations in tissue oxygenation as measured by NIRS can be attributed to changes in cerebral blood flow, assuming all other determinants like arterial oxygenation, and hemoglobin are constant. 24 Since cerebral autoregulation indices are frequently influenced by several parameters, their correct interpretation is critical for taking the right actions to adjust these values and to provide the best possible condition for the patients, which ensures the preservation of cerebral perfusion. The findings of our study support the hypothesis, that a deviation of CPP from CPPopt may indicate possibly suboptimal cerebral perfusion and should trigger further diagnostic and therapeutic measures. Further prospective evaluation is necessary to develop disease-specific management protocols including cerebral autoregulation metrics in the intensive care unit for routine use.
Limitations of the study
The main limitation of the study is the retrospective data analysis. Hence, the cerebral autoregulation indices were calculated retrospectively. Another limitation is the relatively small number of patients, that resulted from the pre-defined inclusion criteria of the study. For data consistency reasons, only patients with invasive intraparenchymal ICP monitoring for at least 7 days, which was required for the calculation of PRx and individual determination of CPPopt, were included in the study. The PRx, that was used in this study, is a global parameter reflecting the functional status of cerebral autoregulation. Considering the fact, that delayed cerebral ischemia mostly occurs as a local disturbance of cerebral perfusion, other locally measured parameters like the partial tissue oxygen pressure (ptiO2) might have led to different findings. However, local monitoring techniques via probes usually placed within the right frontal lobe have the disadvantage of missing local perfusion disturbances located in brain areas other than the right frontal lobe. A whole-brain CTP, which was also performed in this study, allows the evaluation of perfusion deficits in different brain areas.
Reporting checklist
A STROBE checklist for reporting the findings in the manuscript was uploaded with the manuscript.
Footnotes
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Authors’ contributions
Vesna Malinova contributed to conceptualization, methodology, statistical analysis, data interpretation, writing of the original manuscript draft; Beate Kranawetter contributed to data curation, data analysis; Sheri Tuzi contributed to data curation; Onnen Moerer contributed to data interpretation, validation and approval of the final version of the manuscript; Veit Rohde contributed to data interpretation, supervision, validation; Dorothee Mielke contributed to data interpretation, supervision, reviewing and validation.