Traumatic cervical spinal cord injury: Comparison of two different blood pressure targets on neurological recovery

Controversy exists whether blood pressure augmentation therapy benefits patients suffering from spinal cord injury (SCI). This retrospective comparative study was designed to assess the impact of two different mean arterial pressure (MAP) targets (85–90 mmHg vs. 65–85 mmHg) on neurological recovery after traumatic cervical SCI.

injured spinal cord.Because intraspinal pressure is rarely measured, this is commonly done by elevating the mean arterial pressure (MAP).In this retrospective observational study, Långsjö et al. compared a group of patients treated at a lower MAP target compared to those with a higher MAP target.The authors found no difference between groups with regard to neurological outcomes.The optimal blood pressure for treating acute cervical SCI remains a subject of debate.

| INTRODUCTION
In light of current understanding, hypotensive episodes in the acute phase of spinal cord injury (SCI) decrease spinal cord blood perfusion and result in poorer neurological recovery. 1Since hypotension is common in patients suffering from cervical SCI, it has been suggested that blood pressure augmentation therapy by holding mean arterial pressure (MAP) between 85 and 90 mmHg for the first 7days following the injury would improve perfusion in the injury site and lead to improved neurological outcome via limiting secondary injury. 1 Blood pressure augmentation therapy has been implemented into current SCI treatment guidelines, although the benefit of this treatment has not been unequivocally demonstrated. 1The low level of evidence (classified as class III: case series and expert opinions) regarding the blood pressure augmentation therapy in SCI has been demonstrated, for example, in two recent systematic review articles. 2,3][6][7][8][9][10][11] Randomized controlled trials on the subject are lacking.In addition, there are no clear-cut comparative studies assessing the effects of distinctly different blood pressure targets on neurological recovery after SCI.
The aim of this study was to explore the effectiveness of blood pressure augmentation therapy on the neurological recovery after cervical SCI by combining high-resolution blood pressure monitoring methodology and two groups of patients treated with distinctly different MAP goals.

| METHODS
This retrospective comparative study was conducted in Tampere University Hospital, Finland.Data retrieval was approved by the Finnish Institute for Health and Welfare.Tampere University Hospital is one of the five university hospitals in Finland treating the most severely injured patients and offering specialized medical care for over 900,000 inhabitants.Tampere University Hospital is also one of the three national SCI centers in Finland (status received in 2011).
All consecutive adult patients admitted to Tampere University Hospital intensive care unit (ICU) for acute traumatic cervical SCI (defined by symptoms of tetraplegia associated with cervical cord lesion seen in the magnetic resonance imaging) between April 24, 2013, and August 31, 2018, were reviewed.Among the 120 patients treated, we excluded the patients aged <16 years The neurological status of the patients was evaluated two times during the study according to the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI). 12,13The first neurological evaluation was performed during the patient's stay in the ICU and the second during rehabilitation.Based on ISNCSCI, the American Spinal Injury Association (ASIA) impairment scale (AIS) grades and the neurological level of the injury were determined.The neurological recovery was evaluated based on the changes observed in AIS grades and the neurological level (ΔAIS and ΔNeurological level).
Based on the blood pressure measurements performed in the ICU, the individual mean MAP values were calculated for each patient.
In addition, to take all gathered data into account better, we calculated the proportions (%) of MAP values ≥85 mmHg for the first 3 days (3d-MAP% ≥85 ) and for 7 days (7d-MAP% ≥85 ).Mean MAP values and 3d-MAP% ≥85 were correlated with neurological recovery (ΔAIS and ΔNeurological level).Individual mean MAP values, 3d-MAP % ≥85 , 7d-MAP% ≥85 , ΔAIS, and ΔNeurological level were all statistically tested for differences between the lower and the higher MAP groups.
Additionally, 3d-MAP% ≥85 , 7d-MAP% ≥85 , and the rehabilitation time were tested for differences in the patient group experiencing neurological improvement against the patients who did not.
Statistical analyses were performed with IBM SPSS statistics software (version 28.0.1.1).Except for four variables (the timings of the neurological evaluations, rehabilitation time, and the APACHE score), all continuous variables were non-normally distributed.Medians and quartiles of all continuous variables are presented and compared using the Mann-Whitney U test.For categorical variables, the frequencies (n) and percentages were calculated.The Fisher's exact or chi-square test was used to analyze the group differences of the categorical variables.Pearson correlation was applied to analyze the relationships between neurological recovery (ΔAIS and ΔNeurological level) and blood pressure (mean MAP and 3d-MAP% ≥85 ).A p value <.05 was considered statistically significant.
Thirty-two patients (62.7%) received blood pressure augmentation therapy according to the current AANS guidelines (MAP 85-90 mmHg) and were included in the higher MAP group.Nineteen patients (37.3%) were treated with lower MAP target and formed the  1.The only difference between the groups was the rehabilitation time (91.0 vs. 56.5 days, p = .006),which was longer in the lower MAP group.However, the rehabilitation time was not statistically different between the patients experiencing neurological improvement vs. those who did not.

Neurological recovery measured with ΔAIS and ΔNeurological
level was not different between the MAP groups (Table 3).No association was found between the neurological recovery and individual mean MAP values or 3d-MAP% ≥85 (Figure 2).The 3d-MAP% ≥85 or 7d-MAP% ≥85 was not different between the patients showing neurological improvement and those who did not.
The used cumulative median dose of norepinephrine in the ICU was significantly higher in the higher MAP group compared to the lower MAP group (54.8 and 30.0 mg, respectively, p = .025).However, the daily (cumulative dose/days in the ICU) norepinephrine dose did not differ significantly between the groups (Table 3).

| DISCUSSION
In this study, comparing two patient groups of cervical SCI receiving distinctly different blood pressure therapies for 3-7 days, no significant difference was detected in neurological recovery between the groups.
Clinical evidence supporting the use of blood pressure augmentation therapy is scarce. 2,3Experimental studies on rats have demonstrated that the severity of SCI correlates with the reduction of spinal cord blood flow and the impairment of spinal cord axonal signaling (measured with motor and somatosensory evoked potentials). 14Also, in humans, increases in intraspinal pressure and decreases in spinal

Lower
Higher cord perfusion pressure (SCPP) have been shown to result in lowered tissue oxygen and glucose levels as well as increased lactate/pyruvate ratio in the site of the injury and to cause fluctuations in the motor performance of the limbs. 15In rats with mild SCI, induced hypertension improved spinal cord blood flow while hypotension decreased it linearly. 16Restoration of spinal cord blood flow was then observed to improve the axonal function in the injured rat spinal cord. 17sed on this experimental data, it would seem comprehensible that elevated MAP would automatically result in better neurological outcomes in SCI.In a study by Visagan et al., spinal cord tissue oxygen tension reached its peak values with SCPP levels of 90-100 mmHg. 15 that same study, however, the correlation of tissue oxygen and limb motor score formed an inverted U-shaped curve suggesting that not only hypoxia but also hyperoxia may impair motor function.Using microdialysis technique in 14 SCI patients, Phang et al. concluded that metabolically optimal conditions (highest glucose, lowest levels of glutamate, and lactate/pyruvate ratio) were reached when SCPP was elevated up to 90-100 mmHg; thus, corroborating the findings by Visagan et al. 18 However, SCPP levels >100 mmHg were again associated with deranged metabolism.These studies suggest that there could exist a "sweet spot" range regarding the optimal blood pressure level.Also, severity of the SCI may play a role.In severe SCI rat model, a decrease in spinal cord blood flow could not be restored at the injury site even with extreme adrenaline-induced hypertension.
The therapy only caused hyperemia in the adjacent parts of the spinal cord. 16Thus, at least in severe SCI, there may be a higher limit in induced hypertension after which a risk of edema in other parts of the spinal cord may become significant.
Although the experimental studies offer interesting results, it is nevertheless the clinical outcome that ultimately matters.The current AANS recommendation has been perhaps mostly influenced by the study of Vale et al. 4 In that study, 77 patients suffering from C1-Th12 SCI were recruited prospectively to receive blood pressure augmentation therapy of MAP>85 mmHg for 7 days.Fluids and vasopressors were used to reach the target and 64 of the patients were followed over 12 months after the injury.These authors reported that the applied MAP-target was chosen arbitrarily as it is considered "high normal" for non-injured 15-to 30-year-old males.The authors additionally reasoned that the chosen MAP-level was justified based on their previous experience with patients suffering from cerebral ischemia. 4The length of the therapy (7 days) was chosen based on the experimental animal study on SCI, where maximal spinal cord edema T A B L E 3 Blood pressure variables, neurological outcome, and norepinephrine usage.

MAP group
Lower Higher  was measured on Day 5 after the injury. 19 (MAP >90 mmHg) with fluids and dopamine for 7 days after injury. 5urological status improved in 20 patients, while 21 remained unchanged and 9 died.A control group was not included, so this study was descriptive and provided no information on the benefit of the therapy given.
There are no previously published studies directly comparing the effects of different blood pressure targets on neurological recovery after SCI.One previous study has presented worsened outcomes in 41 SCI patients experiencing hypotensive episodes (defined as MAP <85 mmHg lasting at least 2 h based on hourly measurements) compared to 9 patients whose MAP maintained constantly over 85 mmHg during their 5-day post-SCI care. 7However, the study was burdened by obvious unbalance between the patient groups (9 vs. 41) and quite sparse blood pressure monitoring (once per hour).The study also included thoraco/lumbar SCI patients as well as patients with cervical SCI.In the study, 83% of the patients with hypotensive episodes (vs.40% of patients maintaining MAP >85 group) failed to show improvement in their neurological status 26.7 days after the injury.
However, the difference in long-term recovery (at 115 and at 252 days after injury) was no longer statistically significant.
Using high-resolution blood pressure measurements (measurement every minute, MAP-values for the first 3 days after the injury used in the analyses) of 62 SCI patients (40 cervical), Catapano et al.
observed that AIS A-C patients showing neurological recovery had significantly higher blood pressure values than those who did not recover. 8 6 thus, supporting the findings of Catapano et al. 8 Despite the meticulous blood pressure monitoring in these studies, on average 25% of the MAP values still failed to reach the level of ≥85 mmHg. 6,11In fact, it has been shown that even in SCI dedicated centers, meeting the actual MAP targets may be difficult. 20 the higher MAP group of our study, only 14.4% of MAP values failed to reach the threshold of ≥85 mmHg which can, thus, be considered a fairly successful compliance.
Our current results do not concur with these previous findings.
Also, Martin et al. observed that episodes of hypotension and the need of vasopressors did not affect neurological recovery during acute (72 h) care regardless of the MAP goal and argued that arbitrary MAP augmentation may not be effective. 10There are many strengths supporting the validity of our findings.First, we used continuous invasive blood pressure monitoring with meticulous high-frequency storage for the MAP data and achieved a successful blood pressure titration with clinically relevant differences in the MAP endpoints of the groups.Second, to standardize our patient population, we only recruited cervical SCI patients and applied careful exclusion criteria.
Third, as a national SCI center, our hospital provides state-of-the-art care for SCI patients starting from acute intensive and neurosurgical care followed by rehabilitation within our own institution.Prior evidence states that SCI patients benefit from being treated in specialized centers. 21Fourth, in addition to AIS, we also measured another important parameter in recovery, the level of neurological injury, which has not been systematically applied in previous studies.
Although important measure, AIS may be somewhat rough in patients' follow-up.For this reason, we feel that adding the information on the neurological level of injury is beneficial.
Some limitations are present as well.The most profound is obviously the retrospective nature of the study, although the blood pressure data was, after all, collected prospectively.The careful exclusion criteria resulted in a homogenous, yet small sample of patients.Nevertheless, a rare comparison of distinctly different MAP therapies on SCI recovery is presented with uniquely balanced groups of patients.The only difference in the demographic data of the patient groups was the rehabilitation time which was approximately 1 month longer in the lower MAP group.However, longer rehabilitation time does not automatically result in improvement in neurological recovery (i.e., more motor skills, more sensation).
For example, AIS A patients with tetraplegia have only limited training possibilities due to complete SCI and total absence of motor function.
Thus, rehabilitation time can be quite short as it simply aims to improve patient's ability to manage in their home environment with ancillary equipment, personal assistant, and so on.Nonetheless, we found no difference in the rehabilitation time between the patients who showed neurological improvement in comparison to those who did not.
Patients included in our study were treated in our ICU during a 6-year period.Most of the patients in the lower MAP group were treated during the first 3-4 years, while the patients in the higher MAP group were treated later, during the last 2-3 years.During the whole 6-year recruitment period, however, treatment of SCI in our institution has remained the same.For this reason, the bias from the temporal difference regarding the treatment in the ICU is highly unlikely.
Altogether, 51 cervical SCI patients were analyzed in this study.
One consequence of the retrospective nature of the study was the inability to control the timing of neurological examinations.First, only the patients whose initial neurological examination was performed within 1 week after the injury were chosen for this study.Secondly, although the median times of the control neurological examinations were spread apart (110 days vs. 69 days), we found this difference to be not statistically significant between the groups.
Multimodal monitoring may be the way of the future also in SCI treatment.Werndle et al. 22 demonstrated that intraspinal pressures were significantly elevated in patients experiencing SCI.Furthermore, they were able to show that intraspinal pressure was higher at the injury site compared to pressures measured below the injury.Thus, for accurate measurement of spinal pressure and calculation of SCPP, probe must be placed at the injury site.Interestingly, surgical realignment of fractured spinal cord or laminectomy did not effectively lower the elevated spinal pressure.However, elevation of blood pressure using inotropic medication successfully increased SCPP and was also associated with increased amplitudes of motor-evoked potentials recorded below the injury level giving feedback on patient's motor performance in real time during the treatment. 22With this kind of monitoring methodology, individual treatment of SCI patients could become feasible.The downside in this sort of monitoring is of course the invasiveness.Nevertheless, as the main message in the early prospective studies seems to be to take good overall care of your SCI patients, 4,5  This study lays a foundation for future comparative randomized controlled trials by justifying the question whether the lower blood pressure target could yield a similar state of neurological recovery than a higher target.

AUTHOR CONTRIBUTIONS
JL designed the study, supervised and helped in data collection, analyzed the data, and wrote the final manuscript.SJ collected the data, analyzed the data, and helped in writing the original draft of the manuscript.SL wrote the original draft of the manuscript.MP helped in data analysis and commented on the manuscript.TT helped in writing the original draft of the manuscript.HH helped in designing the statistical analyses, helped in data analysis, and commented on the manuscript.AR helped in writing the manuscript.
SK offered the resources for the study and commented on the manuscript.EK helped in designing and supervising the study, and helped in writing the manuscript.TL offered the funding for the study, and helped in designing the analyses and in writing the final manuscript.

( 3
patients), with nontraumatic SCI etiology (3 patients), with ICU length of stay (LOS) <3 days (29 patients), the patients expressing no visible acute traumatic cord lesions on the magnetic resonance imaging (2 patients), patients who were admitted into Tampere University Hospital ICU >48 h postinjury (6 patients), the patients with insufficient or missing data on neurological recovery (11 patients), the patients who died before rehabilitation (7 patients) and the patients whose primary neurological evaluation was performed >7 days postinjury (8 patients).After exclusion, 51 patients comprised to final study cohort (Figure1).Severity of the disease was measured with SAPS (Simplified Acute Physiology Score) classification, APACHE IV (Physiologic Assessment and Chronic Health Evaluation IV) classification, and SOFA (Sequential Organ Failure Assessment) scores.SAPS and APACHE scores were calculated based on the data collected within the first 24 h after the patient was admitted to ICU.SOFA scores were calculated daily.The patient demographics, ICU-LOS, timing of surgery, and the cumulative dosage of norepinephrine were recorded.During the early phase of our period-of-interest (approximately years 2013-2016), the currently recommended blood pressure augmentation therapy guideline of MAP 85-90 mmHg by AANS was not systematically followed due to lack of solid experimental and clinical evidence of its effectiveness.During this time period, MAP-target was decided based on the discretion of treating intensivist and neurosurgeon upon the patient's admittance to the ICU.This resulted in 19 patients whose MAP target was set to slightly lower level (to 80-85 mmHg for 16 patients and to 65-70, 70-75, and 75-80 mmHg for the remaining three patients).This group of patients was then named the lower MAP group (MAP target 65-85 mmHg) in this study.However, in 2016, a collective decision was made in our ICU to simply follow the AANS recommendation for the MAP target.For this reason, most of the patients treated after the mid-year 2016 up until 2018 (and today) were treated with a MAP target of 85-90 mmHg.These patients formed the higher MAP group (MAP target 85-90 mmHg) consisting of 32 patients.Blood pressure was measured invasively from a catheter most often placed in the radial artery.MAP data was recorded continuously and stored as approximately 2-min median values with an electronic information and data management system (Centricity Critical Care, GE Healthcare, General Electric Company, Boston, MA, USA).Only the patients staying in the ICU and receiving blood pressure augmentation therapy for minimum of 3 days were included in the study.This was based on a previous study by Hawryluk et al. demonstrating the strongest association between MAP and neurological recovery during the first 3 days after SCI. 11If the ICU-LOS exceeded 7 days, only the MAP values recorded during the first 7 days were included in the analyses.To exclude the artifact blood pressure values (e.g., due to arterial line sampling and flushing), the measured MAP values were filtered to include values between 10 and 150 mmHg only.

cMAP=
mean arterial pressure.d 7d-MAP% ≥85 = proportion of MAP values ≥85 mmHg upon up to 7 days after the injury.e 3d-MAP% ≥85 = proportion of MAP values ≥85 mmHg upon the first 3 days after the injury.f AIS = American Spinal Injury Association (ASIA) impairment scale.F I G U R E 2 Correlation of neurological recovery and blood pressure.Correlation (Pearson) of neurological recovery measured as ΔAIS (A and C) and ΔNeurological level (B and D) and blood pressure measured as mean MAP (mmHg, A and B) and proportion of MAP values ≥85 mmHg upon the first 3 days (3d-MAP% ≥85 ) after the injury (%, C and D).Orange circles denote the higher MAP group, blue circles denote the lower MAP group.Regression lines (solid lines) with 95% confidence intervals (dashed lines) are presented.AIS, American Spinal Injury Association (ASIA) impairment scale; MAP, mean arterial pressure; r, Pearson correlation coefficient.
Patient selection and exclusion criteria.ICU, intensive care unit; LOS, length of stay; MAP, mean arterial pressure; MRI, magnetic resonance imaging; SCI, spinal cord injury.lower MAP group.In the higher MAP group (median [IQR], hereinafter presented similarly), 82.0% (74.2-85.7) of the MAP values recorded upon the ICU stay reached the suggested AANS guidelines lower limit of 85 mmHg.In the lower MAP group, 36.8% (22.4-50.9) of the MAP values reached 85 mmHg (p < .001 between the groups, Table 3).The median of the individual mean MAP values was 90.2 (88.7-91.7)mmHg in the higher MAP group and 81.4 (79.2-84.4)mmHg in the lower MAP group (p < .001).Comparison of the demographic and clinical characteristics of the patient groups is given in Table With this protocol, Vale et al. found that 60% of the patients initially suffering from complete (AIS A) SCI, recovered at least one AIS-grade, 30% recovered their ability to walk, and 20% their ability for voluntary bladder control.A control group was not included in this study by Vale et al.However, compared to previous reports, the recovery was enhanced.
It is of note, however, that the mean MAP values in the moving towards multimodal SCI monitoring could help us in this respect.This study offers a rare comparison of two groups of cervical SCI patients receiving care with distinctly different MAP targets.Results suggest that recommended MAP 85-90 mmHg may not result in better neurological outcome than lower MAP target (approx.80 mmHg).