Keywords
COVID-19 - Intracranial Pressure - Hemodynamic Brain Response - Neurophysiological
Monitoring
Palavras-chave
COVID-19 - Pressão Intracraniana - Acoplamento Neurovascular - Monitorização Neurofisiológica
INTRODUCTION
The incidence of neurological symptoms, including headache, dizziness, myalgia, hypogeusia/dysgeusia,
and hyposmia/anosmia, in individuals with the coronavirus disease-19 (COVID-19) is
substantial, accounting for approximately 36% of reported symptoms.[1]
[2] However, the pathophysiology underlying the neurological manifestations of severe
acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection remains incompletely
understood.[3]
[4]
[5]
[6]
[7]
[8]
[9]
Neuropathological changes resulting from the coronavirus infection are believed to
arise from direct virus invasion or molecular alterations, secondary to a systemic
inflammatory response. The virus may access the central nervous system through hematogenous
or retrograde routes, such as the olfactory nerve.[8]
[9]
[10]
[11]
[12]
[13]
[14] Viral ribonucleic acid (RNAs) have been detected in the cerebrospinal fluid and
brain tissue during postmortem examinations of selected patients affected by the disease.[15]
These neuropathological changes can lead to alterations in vascular permeability,
a crucial factor in maintaining the integrity of the blood–brain barrier, regulating
gas exchange, and governing cerebral blood flow (CBF). Several factors influence CBF,
including arterial pressure, intracranial pressure (ICP), and cerebrovascular resistance.
Any factor that affects these determinants can lead to changes in cerebrovascular
hemodynamics. Additionally, mechanical ventilation (MV) may induce cardiac overload
in patients, as evidenced by increased jugular and central venous pressures, diminished
cerebral venous return, and consequent elevation of ICP levels.[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
Even small positive end-expiratory pressure (PEEP) values were associated with increased
ICP in patients with brain injury. While the PEEP's impact on ICP varies among patients
with different neurological injuries, its overall effect is minor.[23]
[24]
[25]
[26]
[27]
[28] Patients with severe SARS-CoV-2 infection usually require prolonged MV with extreme
parameters. Nonetheless, the influence of ventilator settings on ICP and compliance
in patients lacking brain injury still needs to be more adequately explored. Multimodal
brain monitoring offers a means to assess cerebrovascular hemodynamics (CVH) and evaluate
the effect of protective lung ventilation, particularly the arterial partial pressure
of carbon dioxide (PaCO2) and PEEP, on cerebral blood flow (CBF) and intracranial
compliance (ICC).[29]
The objective of this study was to evaluate the ICC in patients with COVID-19 infection
on MV, compared to patients with other diagnoses, to better characterize its effects.
We also compared the ICC in patients with COVID-19 infection not requiring mechanical
ventilation and healthy volunteers, to assess the isolated effects of this disease.
METHODS
Study design and setting
This prospective, observational, exploratory, multicenter study was conducted in four
tertiary care centers and one university (Federal University of São Carlos) registered
on Clinical Trials.gov (registration number 31589920.7.1001.5505). Our study followed
the consolidated standards of reporting trials (CONSORT) recommendation for observational
studies.[30]
[31]
Our convenience sample was recruited and followed for 15 days after study inclusion.
All patients or legal representatives signed an informed consent form. The present
study followed the declaration of Helsinki, and it was approved by the Ethics Committee
of the Federal University of São Paulo (UNIFESP) and São Carlos (UFSCar) under the
protocols 31589920.7.1001.5505 and 32338920.5.0000.5504, respectively.
Selection of participants
All COVID-19 participants tested positive on the reverse transcription polymerase
chain reaction (RT-PCR), and symptoms onset was < 15 days from study inclusion. For
MV patients, the time between hospital admission and study inclusion was ≤ 72 hours.
The MV non-COVID-19 group was composed of patients in MV due to alternative diagnoses.
The healthy volunteer group comprised healthy subjects with no acute respiratory symptoms
during evaluation. We excluded patients presenting with acute central nervous system
disorders.
Data collection and outcomes
Data obtained from electronic medical records included demographics, anthropometric
measurements (weight and height), clinical characteristics, the timing of symptoms,
and results of diagnostic tests, including chest imaging and arterial blood gas analysis.
Physiological data (heart and respiratory rates, oxygen saturation, and blood pressure)
and the utilization of ventilatory support were systematically collected during ICC
monitoring. Patients were monitored for 20 to 60 minutes, while healthy controls were
evaluated for 90 minutes in a room with appropriate climatization and temperature
after 15 minutes of rest.
A certified evaluator applied the modified Rankin scale (mRS) on day 15 of the study
participation, either in person or by telephone, to discharge patients. A poor outcome
was defined as mRS > 2. A missed outcome was the impossibility of contacting the patient
after discharge.
Intracranial compliance measurements
We evaluated ICC with a noninvasive ICP waveform monitoring device developed by Brain4Care
Inc. (Johns Creek, GA, USA). The Braincare sensor was placed on the patient's scalp
without shaving, surgical incision, or drilling, as previously described by Moraes
et al.[32] ([Figure 1A]). Minimal changes in the skull caused by changes in ICC were captured by the sensor
and provided the ICP waveform, as a proxy.[32]
[33]
[34]
Figure 1 The Brain4Care device in use.
Each cardiac beat generated an ICP waveform with three peaks: P1, associated with
systolic arterial pressure transferred from the choroid plexus to the cerebrospinal
fluid; P2, associated with the reflection (rebound) of the blood pressure wave in
the brain tissue; and P3, related to the closure of the aortic valve. These waveforms
closely resembled those obtained through invasive ICP measurements, and the relationship
between their components provided insights into the ICC ([Figure 1B]).[32]
[33]
[34]
The B4C (Brain4Care Inc.) analytics system validated all sensor-collected data, including
the P2/P1 ratio, a parameter indicating the morphology of the ICP pulse wave. The
software automatically determined P1 and P2, which were visually confirmed by inspecting
the waveforms. The amplitudes of the peaks were measured by subtracting the baseline
value of the ICP waveform. The P2/P1 ratio was calculated by dividing the amplitude
at these two-time points. The mean pulse and its corresponding 95% confidence interval
(CI) were computed using all valid alignment pulses through a nonparametric bootstrap
procedure with 1,000 replications. When P2 > P1, the ICC was categorized as “abnormal”
([Figure 1C]). The minute-by-minute analysis compared the defined indices with previously reported
values.[34]
Statistical analysis
Qualitative variables were summarized in absolute (n) and relative (%) frequencies.
Continuous variable distributions were assessed for normality by skewness, kurtosis,
and graphical methods. Those with normal distribution were presented as mean and standard
deviations and compared with the independent samples Student t-test. Otherwise, they
were presented as medians and interquartile ranges and compared with the Mann-Whitney
nonparametric test. Categorical variables were analyzed using the Chi-Square test.[35]
[36]
[37]
[38]
The P2/P1 ratios were analyzed using a mixed linear model with random effects in four
groups: MV patients (COVID-19 and non-COVID-19), nonmechanically ventilated COVID-19
patients, and healthy volunteers.[39]
[40]
[41] The P2/P1 ratio was obtained from the average of all valid pulses each minute; all
results outside 0.5 to 1.8 were considered artifactual and excluded.
For all analyses, statistical significance was set at p-value < 0.05. The R (R Foundation for Statistical Computing, Vienna, Austria) software,
version 4.0.5, was used for all analyses.
RESULTS
Between June 2020 and September 2021, 192 participants were recruited for this research.
However, only 78 participants were included to the final sample, among whom 15 were
mechanically ventilated COVID-19 patients (MV-COVID-19), 15 mechanically ventilated
participants without COVID-19 (MV non-COVID-19), 24 were nonmechanically ventilated
COVID-19 patients (non-MV-COVID-19), and a control group with 24 healthy individuals
([Figure 2]). In all four groups, the majority were men (60% MV COVID-19, 60% MV non-COVID-19,
67% non-MV-COVID-19, and 67% healthy volunteers) ([Figure 2]).
Figure 2 Flow diagram of the study.
Mechanically ventilated patients (COVID and non-COVID) were similar in age, sex, and
body mass index (BMI). There was no difference in P2/P1 ratios in mechanically ventilated
patients (COVID-19 vs. non-COVID-19), p = 0.65 ([Figure 3]). The MV COVID-19 patients had a higher frequency of systemic arterial hypertension
and type II diabetes (p = 0.03) ([Table 1]).
Figure 3 Summary of all research.
Table 1
Mechanical ventilation sample characterization
|
MV COVID-19 (n = 15)
|
MV non-COVID-19 (n = 15)
|
p-value
|
|
Male sex (%)
|
60
|
60
|
1.00
|
|
Age, years – median (IQR)
|
66 (53–72)
|
55 (42-70)
|
0.22
|
|
BMI – median (IQR)
|
26 (24–28)
|
25 (23-27)
|
0.17
|
|
Comorbidities (%)
|
SAH
|
60
|
20
|
0.03*
|
|
DM2
|
60
|
20
|
0.03*
|
|
Current smoker
|
27
|
40
|
0.44
|
|
Obesity
|
21
|
0
|
0.07
|
|
CKD
|
20
|
13
|
0.62
|
|
Other
|
60
|
41
|
0.46
|
Abbreviations: BMI, body mass index; COVID-19, coronavirus disease-19; CKD, chronic
kidney disease; DM2, diabetes mellitus type 2; IQR, interquartile range; MV, mechanical
ventilation; SAH, systemic arterial hypertension. Note: *Statistically significant
p-value.
The non-MV patients (both COVID-19 and healthy volunteers) were also similar in age
and sex. Non-MV-COVID-19 patients had a higher BMI (p < 0.01) and a higher frequency of comorbidities than healthy volunteers ([Table 2]). There was no difference in P2/P1 ratios in non-MV patients (COVID-19 and healthy
volunteers, p = 0.70) ([Figure 3]).
Table 2
Non-mechanically ventilated participants
|
Non-MV COVID-19 (n = 24)
|
Healthy volunteers (n = 24)
|
p-value
|
|
Male sex (%)
|
67
|
67
|
1.00
|
|
Age, years – median (IQR)
|
52 [45-65]
|
45 [43-55]
|
0.09
|
|
BMI – median (IQR)
|
31 [27-32]*
|
24 [22-26]
|
<0.01*
|
|
Comorbidities (%)
|
SAH
|
42
|
0
|
<0.01*
|
|
DM2
|
33
|
0
|
<0.01*
|
|
Current smoker
|
17
|
0
|
0.04*
|
|
Obesity
|
54
|
29
|
0.07
|
|
CKD
|
21
|
0
|
0.02*
|
|
Other
|
28
|
0
|
<0.01*
|
Abbreviations: BMI, body mass index; COVID-19, coronavirus disease-19; CKD, chronic
kidney disease; DM2, diabetes mellitus type 2; IQR, interquartile range; MV, mechanical
ventilation; SAH, systemic arterial hypertension. Note: *Statistically significant
p-value.
The MV COVID-19 patients were older than non-MV-COVID-19 patients (median age 66 [53–72]
vs. 52 [45–65], p = 0.04). Other demographic and clinical characteristics were similar between the
two groups. The P2/P1 ratio was higher in the MV COVID-19 patients than in the non-MVCOVID-19
(1.13 ± 0.27 vs. 1.07 ± 0.58, p < 0.01), as shown in [Figure 3].
At the follow-up, 15 days after study inclusion, 40% of the MV-COVID-19 patients were
still on MV, while 75% of the non-MV-COVID-19 patients had been discharged. A poor
functional outcome (mRS 3–6) at 15 days was observed in 87% of the MV-COVID-19 and
80% of MV-non-COVID-19 patients. A good functional outcome (mRS 0–2) was observed
in 50% of the non-MV COVID-19.
DISCUSSION
Our exploratory study showed no difference in P2/P1 ratios in mechanically ventilated
patients (COVID vs. non-COVID). The P2/P1 ratio was higher in MV COVID-19 patients
than in non-MVCOVID-19 patients. This finding is suggestive that changes in ICC previously
described in COVID-19 patients might have been an effect of MV itself.
There were two studies that evaluated COVID-19 patients under MV within 72 hours of
intubation using the B4C and other hemodynamic cerebral parameters.[29]
[42] Patients who were obese and nonobese were compared, and an ICC/CVH score was altered
in obese patients.[42] The authors suggested an association between ICC impairment and obesity, which may
have led to unfavorable prognosis in patients with severe COVID-19. In another series,
the P2/P1 ratio was abnormal in 66% of subjects, with the P2/P1 ratio between 1.01
and 1.2 in 48%.[29] However, as showed by the authors, neither of these studies used a control group
or aimed to evaluate the effect of COVID-19 on ICC, making it impossible to disentangle
the impact of COVID-19 from that of MV alone.
A systematic review[22] regarding brain-injured patients and MV concluded that PEEP could reduce CBF. However,
there are still many questions regarding the impact of airway pressure on ICP, especially
in nonneurological patients.[29]
[42]
[43]
[44] The influence of MV parameters on cerebral blood flow and ICC must be further evaluated.
Permissive hypercapnia leading to vasodilation, which is frequently seen in MV-COVID-19
patients, might play a role in derangements of CBF associated with MV.[23]
[45] Therefore, as used in our series, noninvasive neurological monitoring might be important
in preventing cerebral complications in MV patients.
Our exploratory study has several limitations. First, we utilized a convenience sample.
Second, we obtained data from the initial monitoring day, thus providing a single
instance of P2/P1 behavior during the intensive care unit stay. Third, we conducted
our study in the opening year of the COVID-19 pandemic, when higher mortality rates
were witnessed internationally due to lack of familiarity with the disease and understaffed
hospitals. Finally, due to the short follow-up period, we did not have enough time
to assess our population's functional outcome in the long term.
In conclusion, our data suggest that COVID-19 does not impair ICC, as measured by
a noninvasive ICP waveform monitor. However, these results must be interpreted carefully
since this study is exploratory. Further studies, with a more elaborate design correlating
ventilatory parameters, sedation, and long-term cognitive parameters at follow-up,
are of utmost importance to understanding the real impact of MV and COVID-19 upon
ICC.
Bibliographical Record
Ana Flávia Silveira, Marcella Barreto Santos, Nelci Zanon Collange, Cintya Yukie Hayashi,
Gustavo Henrique Frigieri Vilela, Samantha Longhi Simões de Almeida, João Brainer
Clares de Andrade, Salómon Rojas, Fabiano Moulin de Moraes, Viviane Cordeiro Veiga,
Uri Adrian Prync Flato, Thiago Luiz Russo, Gisele Sampaio Silva. Intracranial compliance
in patients with COVID-19: a multicenter observational study. Arq Neuropsiquiatr 2024;
82: s00441788669.
DOI: 10.1055/s-0044-1788669
