Mean arterial pressure (MAP) augmentation in traumatic spinal cord injuries: Early hyperperfusion treatment influences neurologic outcomes

Abstract

Context: Hyperperfusion therapy, mean arterial blood pressure (MAP) > 85 mmHg, is a recommended treatment of blunt traumatic spinal cord injury (SCI). We hypothesized the first 24 h of MAP augmentation would be most influential on neurological outcomes.

Design: This retrospective study from a level 1 urban trauma center dating 1/2017 to 12/2019 included all blunt traumatic spinal cord injured patients receiving hyperperfusion therapy. Patients were grouped as “No improvement” vs “Improvement” measured by change in American Spinal Injury Association (ASIA) score during their hospitalization. MAP values for the first 12, first 24 and last 72 h were compared between the two groups; P < 0.05 was significant.

Results: After exclusions, 96 patients underwent hyperperfusion therapy for blunt traumatic SCI, 82 in the No Improvement and 14 in the Improvement group. Groups had similar treatment durations (95.6 and 96.7 h, P = 0.66) and ISS (20.5 and 23, P = 0.45). The area under the curve, calculation, to account for time less than goal and MAP difference from goal, in the No Improvement group was significantly higher (lower and more time below MAP goal) compared to the Improvement group for the first 12 h (40.3 v. 26.1 P = 0.03) with similar findings in the subsequent 12 h of treatment (13–24 h; 62.2 vs 43, P = 0.09). There was no difference between the groups in the subsequent 72 h (25–96 h; 156.4 vs 136.6, P = 0.57).

Conclusions: Hyperperfusion to the spinal cord in the first 12 h correlated significantly with improved neurological outcome in SCI patients.

Introduction

The pathophysiology of spinal cord injuries involves both mechanical forces (compression, distraction and transection) and nutrient delivery systems (blood flow, oxygenation) (1, 2). A critical aspect of spinal cord recovery is maintaining adequate blood flow to the injured neural tissue (1, 2). Augmentation of blood pressure with elevation of mean arterial pressure (MAP) is part of the modern treatment for traumatic spinal cord injuries (SCI) (1, 3–7). Loss of blood flow autoregulation within the injured spinal cord creates an environment for continued ischemia that can be seen with laser speckle contrast imagery (8). Elevated MAP goals (>85 mmHg) for four to seven days post injury correlate with improved neurological outcomes, measured by American Spinal Injury Association (ASIA) scores (1, 3–7). Spinal cord hyperperfusion protocols (Appendix A) appear to be an essential part of early spinal cord injury treatment showing improved neurologic outcomes (1, 3–7).

Spinal cord hyperperfusion protocols start with fluid resuscitation and vasopressor support. The Consortium for Spinal Cord Medicine recommends the use of vasopressors dopamine, norepinephrine or phenylephrine based on injury location (5–7). Initiation of therapies in injured patients does not appear to be without risk. Complications of vasopressor use are more common in elderly trauma patients and are typified by arrhythmias and elevated cardiac enzymes (9–12). Along with treatment risks, other details regarding treatment timing, duration and need to achieve set MAP goals during the entire 72–96 h are lacking in the current literature.

Optimal hyperperfusion care during early spinal cord injury therapy is MAP > 85 mmHg 100% of the treatment protocol (72–96 h). Even when patients are managed in intensive care units at level one trauma centers, these goals are achieved only about 75% of the time (1,13). Hyperperfusion goals are even less likely to occur in patients requiring operative fixation, with one recent study showing 100% of SCI patients had a MAP < 85 during operative interventions (14).

Approximately 80% of all SCIs occur concurrently with polytrauma, creating tenuous resuscitative circumstances in the setting of multiple potential sources of shock (15–17). Traumatic SCI treatment typically includes early operative intervention (within first 24 h post injury) to increase the odds of neurological improvement (2). Previous studies have demonstrated that the desire for both spinal cord hyperperfusion and early spine fixation creates competing treatment priorities (14). The interplay between hyperperfusion timing, spine fixation and complications associated with hyperperfusion therapies during early SCI treatment are yet to be delineated. In this paper, we aim to clarify the relationship between hyperperfusion therapy and neurologic recovery following spinal cord injury. We hypothesized that achievement of a systemic MAP > 85mmHg during the first 12–24 h of SCI treatment would have the greatest positive impact on neurological function.

Materials and methods

This was a retrospective IRB-approved study of spinal cord injured patients at an urban level one trauma center. Inclusion criteria were blunt traumatic spinal cord injury patient at least 18 years of age date range of January 1, 2017 to December 31, 2019. Exclusions for the study were penetrating mechanism, pregnancy, prisoners, incomplete neurological exam on arrival, missing neurological exam documentation, non-trauma admission and death prior to completion of hyperperfusion protocol.

Data were extracted using a combination of records from TraumaBase and chart review. Patient specifics included age, spinal cord level of injury, injury severity score (ISS) and abbreviated injury score (AIS) by spinal cord region. Hospital arrival MAP was recorded along with hourly MAP values during the 96 h of hyperperfusion therapy. ASIA scores were calculated on hospital day one, at the completion of MAP treatment and day of discharge; these scores were calculated using neurosurgical and physical medicine and rehabilitation documentation. Operative specifics including MAP values (every 3 min) and operative time were recorded. Vasopressor data including type of vasopressor and cumulative dose and daily doses were recorded (calculated in norepinephrine equivalents).

Patients were divided into two groups based on ASIA score changes from hospital day one to discharge. ASIA scores were calculated based on documentation from the Neurosurgical team and trauma team on hospital day one. Discharge ASIA scores were calculated based on the Physical Medicine and Rehabilitation documentation as well at the trauma team documentation. Subjects were placed in the “No Improvement” group if their ASIA score worsened or did not change while the “Improvement” group if their ASIA score improved throughout the hospitalization. Demographics and outcomes were compared between the groups using two-sample t-tests and nonparametric Wilcoxon two-sample tests for continuous or semi-continuous variables and chi-square tests for categorical variables. When statistically significant, chi-square tests were accompanied by estimated odds ratios. MAP differences were calculated as the difference from the goal MAP of 85 to the actual MAP (when less than 85mmHg). Area under curve (AUC) calculation was performed for each patient as a function of time spent below a MAP of 85 mmHg.

Where a₁:b₁ are the first time points where the patients MAP value crosses below MAP of 85 mmHg. f(x) represents a single patients MAP pressure curve for 96 h of the hyperperfusion therapy (Fig. 3a).

$$\mathrm{AUC}=\underset{a_1}{\overset{b_1}{\int }}f\left(x\right)\mathrm{d}x+\underset{a_2}{\overset{b_2}{\int }}f\left(x\right)\mathrm{d}x+\cdots \underset{a_x}{\overset{b_x}{\int }}f\left(x\right)\mathrm{d}x$$

Figure 3.

(A) Graphical representation of area under the curve for MAP values less than 85mmHg per MAP treatment time period. (B) Area under the curve for MAP values less than 85mmHg per MAP treatment time period. First 0–12 h the No Improvement group had statistically significantly longer amounts of time and greater severity of MAP < 85mmHg (P = 0.03). From 13–24 h the No Improvement continue to have a greater area under the curve (59.6, 33.2; P = 0.07). The final 72 h of treatment the areas were basically the same (156.4, 136.6; P = 0.57)

Results

A total of 124 patients met the original inclusion criteria of blunt spinal cord injury; after exclusions 96 patients were included for final analysis. A total of 82 patients had no improvement of their ASIA score and 14 had ASIA score improvement. Groups (No Improvement and Improvement) were similar in demographics including age, sex, race and comorbidities (Table 1). The two groups had similar levels of injury severity (mean ISS 23 and 20.5, P = 0.45) and there was no statistical difference between groups in their Abbreviated Injury Severity (AIS) scores by head, cervical, thoracic, and lumbar regions. In the Improvement group, 28.6% (n = 4) had an increase of 2 ASIA scores while the remaining 71.4% (n = 10) improved by 1 ASIA score. In the No Improvement group hospital transfer was statistically significantly different between groups with 7% of the Improvement group consisting of transfers compared to 39% of the No Improvement group (P = 0.02).

Table 1.

Patient demographics.

	No Improvement	Improvement	P value
	N = 82	N = 14
Patient Demographics
Male sex, n (%)	63 (76.83%)	11 (78.57%)	0.89
ISS (median) [25th, 75th]	20.5 [16,29]	23 [17,41]	0.45
Age (mean) (SD)	52.9 (18.57)	49.21 (15.77)	0.49
BMI (mean) (SD)*	28.58(6.13)	25.04 (6.15)	0.049
Race, white, n (%)	61 (74.4%)	8 (57.14%)	0.18
Tobacco use, n (%)	29 (35.7%)	7 (50%)	0.30
Comorbidities
Hypertension, n (%)	35 (42.68%)	5 (35.71%)	0.63
Coronary artery disease, n (%)	5 (6.10%)	2 (14.29%)	0.28
Atrial fibrillation, n (%)	5 (6.10%)	1 (7.14%)	0.88
Diabetes Mellitus, n (%)	18 (22.05%)	1 (7.14%)	0.20
COPD (%)	3 (3.66%)	1 (7.14%)	0.55
Chronic kidney disease, n (%)	5 (6.10%)	0 (0%)	0.34
Injury Specifics
Fall, standing, n (%)	29 (35.37%)	5 (35.71%)	0.87
Fall, height, n (%)	10 (12.2%)	2 (14.29%)
MVC, n (%)	22 (26.83%)	5 (35.71%)
Other mechanisms, n (%)	21 (25.61%)	2 (14.29%)
Hospital transfer, n (%)*	32 (39.02%)	1 (7.14%)	0.02
Head AIS (mean) (SD)	2.23 (1.02)	2 (0)	0.89
C-spine AIS (mean) (SD)	3.76 (1.63)	4 (0.43)	0.21
T-spine AIS (mean) (SD)	3.5 (1.45)	3.67 (1.53)	0.92
L-spine AIS (mean) (SD)	2.85 (1.28)	2.75 (1.5)	0.86
HOD1 ASIA A	25 (30.49%)	3 (21.43)	0.49
HOD1 ASIA B	11 (13.41)	3 (21.43)	0.43
HOD1 ASIA C*	12 (14.63)	6 (42.86)	0.012
HOD1 ASIA D	34 (41.46)	2 (14.29)	0.052

ISS = injury severity score, MVC = motor vehicle collision, HOD = hospital day.

Blood pressure augmentation treatment durations were similar between groups with means of 96.71 h for the Improvement group and 95.55 for the No Improvement group (P = 0.66). The in-hospital mortality rate was higher in the No Improvement group (9/82, 11%) compared to the Improvement group (0/14, 0%). The causes of death are detailed in Table 2 with all nine mortalities occurring after completion of the MAP treatment protocol duration of 96 h.

Table 2.

Hospital course.

	No Improvement	Improvement	P value
	N = 82	N = 14
Admission lactate (mean) (SD)*	2.11 (1.41)	2.53 (0.83)	0.037
Length of MAP protocol (hours)	95.55	96.71	0.66
Total hospital LOS (hours) (median) [25th, 75th]	276 [182,456]	312 [264,720]	0.26
Total ICU LOS (hours) (median) [25th, 75th]	5 [4,8]	5 [3,7]	>0.99
Total vent days (days) (median) [25, 75th]	0 [0,6]	1 [0,1]	0.93
Discharge to IPR, n (%)	47 (57.32%)	10 (71.43%)	0.32
Discharge to SNF, n (%)	7 (8.54%)	1 (7.14%)	0.86
Discharge to home, n (%)	10 (12.20%)	1 (7.14%)	0.58
Discharge to LTACH, n (%)	9 (11.0%)	2 (14.29%)	0.72
Discharge to COR, n (%)	9 (11.0%)	0	0.19
Death After Completion of MAP	Cause	Number
	Brain injury	2
	Respiratory failure, DNR-CC	4
	Cardiac arrest	2
	Septic shock	1

Average MAP values at 12 h intervals differed between groups to varying degrees throughout the entire treatment. In the first 12 h, the Improvement group had significantly higher average MAPs (97 mmHg) compared to the No Improvement group (90.5 mmHg) (P = 0.036) (Fig. 1). There was no statistical difference in average MAP values between groups after the initial 12 h of augmentation treatment. Although the average MAP for each group was greater than 85mmHg throughout the treatment duration, the variability between MAP high and low values were not accounted for with the mean calculation.

Figure 1.

Average recorded MAP value during each time period for the total treatment duration of 96 h. During the first 12 h of treatment the Improvement group had higher average MAP values. The first 12 h the Improvement group had statistically significantly higher values (P = 0.037). The trend continued at 24 h with higher average MAP values in the Improvement group, although not statistically significantly higher (P = 0.59)

The No Improvement group failed to reach the MAP goal of 85mmHg in the first 12 h of treatment 34.6% of the time compared to the Improvement group of 18.5% (P = 0.012) (Fig. 2). From 13–24 h of treatment both groups continued to fail at MAP goal achievement with the No Improvement group not reaching a MAP of 85mmHg 26.7% of the time compared to the Improvement group 21.7% (P = 0.44). There was no difference in the likelihood of achieving the MAP goal of 85 mmHg between groups in the final 72 h of treatment (24.4%, 23%, P = 0.73). Nearly all patients went to the operating room for spinal fixation and/or decompression during their blood pressure augmentation treatment within the first 24 h of hospital admission. In the Improvement group 100% (n = 14) of patients went to the OR at a median of 20 h after hospital arrival compared to 87% (n = 71) of patients in the No Improvement group at a median of 17 h (P = 0.64).

Figure 2.

Percentage of time MAP value was less than the goal of 85mmHg during each time period of treatment. The percentage of time the MAP value was less than goal was higher in the No Improvement group throughout MAP treatment duration. For the first 12 h of treatment the No Improvement group had statistically signifantly greater treatment time not at the MAP goal of 85mmHg; 15% compared to 8% in the Improvement group (P = 0.03). At 13–24 h of treatment time the No Improvement continued to not meet the MAP goal of 85mmHg almost 30% of the treatment time

To account for both the amount of time MAP was below 85mmHg and severity of MAP value less than 85mmHg the area under the curve (AUC) was calculated for the components of treatment that occurred below the MAP goal. The AUC in the No Improvement group was significantly higher (lower MAP and more time below MAP goal) compared to the Improvement group for the first 12 h (40.3, 26.1, P = 0.03) and nearly significant for the first 24 h (62.2, 43, P = 0.09) hours of treatment (Fig. 3b). The AUC was not significantly different between groups during the final 72 h of treatment.

The majority of patients required vasopressors to maintain MAPs of > 85 mmHg (79% Improvement, 83% No Improvement, P = 0.69) for similar average hours (60 h Improvement, 68 h No Improvement, P = 0.3). Both groups required nearly the same amount of vasopressors, based on norepinephrine equivalents, to maintain the desired MAP. Average total norepinephrine equivalents in the No Improvement group totaled 51,499.14 mcg and 42,349.55 mcg for the Improvement group (P = 0.35). In both groups tachycardia (heart rate > 130) was the most common complication followed by atrial fibrillation; no documented ventricular arrhythmias were present in either group. There was no difference in endpoints of resuscitation between groups (daily lactate, pH, urine output (UOP) and intravascular fluid volumes). There were no documented invasive line complications (Appendix B).

Discussion

Spinal Cord Injury patients have unique treatment goals which focus on increasing mean arterial pressure to preferentially perfuse the spinal cord thus limiting damage to ischemic/injured neural tissue (1, 3–7). Over a three-year period, our level one trauma center treated a considerable amount of SCI patients with MAP augmentation protocols aimed at a MAP of 85mmHg for the first 96 h post injury. Our results show those patients that had neurologic improvement achieved higher mean MAP values and more time at the MAP goal of 85mmHg in the first 12 h. The patients with neurologic improvement had both significantly less time and severity of hypotension below MAP goals (AUC calculations) compared to those patients without neurologic improvement. Physiologically and demographically patients were similar in both the Improvement and No Improvement groups. From our data, we found it essential that all blunt SCI patients be started on MAP augmentation and minimize time below MAP goals within the first 12 h of treatment. Although the overall ISS in both groups was significant, it was similar which does help eliminate bias between the two groups. This relationship was also consistent across AIS head and spinal cord levels which additionally indicate these two groups were similarly injured overall.

Universally reported, all SCI patients have MAP recordings less than goal at some point during their treatment protocol (1, 12). Compared to other studies, our patients had similar amounts of relative hypotension, with less hypotension in the first 0–12 h. A similar large population study report 24.9% of all MAP recordings were below 85mmHg threshold in the first five days, this is similar to our findings in both groups (1). The average MAP for the Improvement and No Improvement group in the first 12 h (91, 97mmHg respectively) was greater than the goal of 85mmHg. This is consistent with literature reporting that a MAP of >90mmHg is not only achievable in SCI patients but also a common occurrence in these patients (1, 3). However, this is the first study showing neurologic improvement differences with higher MAP in the first hours of treatment. Since a MAP of 90 was demonstrated in this study to be achievable and beneficial future studies aimed at identifying the most beneficial MAP threshold are necessary. In our study, the average MAP accounted for several differences between groups but failed to capture MAP variability.

In a previous study investigators compared the percentage of time patients achieved the MAP goal (85 mmHg) (1). A lower proportion of MAP less than 85mmHg was associated with improvement. Our study echoes these results with 8.6% of MAP values less than 85mmHg in the Improvement compared to 15.5% in the No Improvement group during the first 12 h of treatment. Although not statistically significant, this relationship continued in the 24 h and 25–96 h time periods. In an effort to account for time below MAP goals and amount of relative hypoperfusion we did an AUC sum calculation. The Improvement group had significantly less hypotensive time and severity, especially in the first 12 h. The early hours of SCI MAP treatment (0–12 h) continued to be associated with neurological improvement (average, percentage of time AUC). These results support the importance of early MAP augmentation, minimizing relative hypoperfusion and suggest further investigation into higher early MAP goals.

A significant number of our SCI patients were patients transferred from other care facilities. This patient population was significantly more likely to be found in the No Improvement group. MAP augmentation protocols begin at time of injury regardless of how long transport to another facility may take rendering these patients specifically vulnerable to MAP values less than goal. Suboptimal hemodynamic monitoring is often documented in both community/transferring hospitals and during patient transport (18). Specifically cited from a 40 patient case series, community/transferring hospitals had 50% of MAP recordings less than 80mmHg and 20% of MAP recordings less than 80mmHg during transport (18). Although this study did not include correlations to neurologic outcomes, our study examined both presence of hypotension and neurologic outcome (18). Our study found the transferred patients did have significantly lower MAP values in the first 12 h after injury compared to the direct from scene patients. This findings indicates further study into the specifics of transferred patients is necessary to determine what factors are contributing to the lower blood pressures found.

Although this study has some intriguing findings, it is not without several limitations. The retrospective nature of the study design makes intervention and acquisition of new data difficult. We were unable to account for a reliable time of injury to initiation of hyperperfusion therapy for all patients. Although the population size of 96 is one of the larger studies performed on hyperperfusion therapies in spinal cord injured patients, this still represents a relatively small population and minimizes important subgroup analysis, understanding of rare complications of therapy and the benefit of hyperperfusion in different levels of spinal cord injury. Finally, the study results are based on hospital outcomes with no long-term data. In future prospective studies, we intend to rectify several of these limitations and provide long-term outcomes in spinal cord injury with early hyperperfusion therapy.

Conclusion

Blunt SCI patients suffer complex physiologic changes due to the inherent nature of their injury (2). Mean arterial pressure augmentation to achieve higher than average MAP values is a cornerstone of modern treatment. Our findings echo the importance of maintaining an elevated MAP >85mmHg for neurologic improvement. In addition, focus on the first 12 h of treatment after injury with maintaining a MAP > 85mmHg in all patients regardless of hospital transfer status or direct from scene transport. Future prospective studies including multicenter data from large populations will lend to our evolving understanding of the importance of early therapies in spinal cord injury treatment.

Appendices.

Appendix A

Appendix B

	No Improvement	Improvement	P value
	N = 82	N = 14
Vasopressor use, n (%)	68 (82.93%)	11 (78.57%)	0.69
Mean total vasopressor (h)	67.88	60.3	0.30
Mean Norepinephrine Equivalents 24 h (mcgs)	10657.45	6788.64	0.24
Mean Norepinephrine Equivalents 48 h (mcgs)	15203.85	10865.91	0.21
Mean Norepinephrine Equivalents 72 h (mcgs)	14582.78	12667.27	0.54
Mean Norepinephrine Equivalents 96 h (mcgs)	11055.06	12027.73	0.58
Mean Norepinephrine Equivalents total (mcgs)	51499.14	42349.55	0.35
Mean number of arterial lines	1.21	1.14	0.59
Mean % of arterial line complications	0%	0%	NA
Mean number of central venous catheters	0.98	0.93	0.79
Mean % of central venous catheter complications	0.00%	0.00%	NA
Vasopressor Complications
Atrial fibrillation, n (%)	7 (8.54%)	1 (7.14%)	0.86
Ventricular tachycardia	1 (1.22%)	0 (0%)	0.68
Tachycardia (>130)	13 (15.85%)	3 (21.43%)	0.60
ST elevation	0 (0%)	0 (0%)	NA
Ventricular fibrillation	0 (0%)	0 (0%)	NA
Troponin Elevation	3 (3.66%)	0 (0%)	0.47
Resuscitative Points
Mean IVF (L) 24 h	3.14	2.45	0.46
Mean IVF (L) 48 h	2.35	2.02	0.69
Mean IVF (L) 72 h	1.48	1.86	0.21
Mean IVF (L) 96 h	1.04	1.14	0.36
Mean Hemoglobin 24 h	12.34	11.67	0.10
Mean Hemoglobin 48 h	11.32	10.51	0.09
Mean Hemoglobin 72 h	10.68	9.77	0.13
Mean Hemoglobin 96 h	10.36	9.86	0.36
Mean pH 24 h	7.32	7.35	0.74
Mean pH 48 h	7.38	7.34	0.09
Mean pH 72 h	7.39	7.38	> 0.99
Mean pH 96 h	7.39	7.39	0.76
Mean Lactate 24 h	1.82	1.41	0.63
Mean Lactate 48 h	1.25	0.82	0.065
Mean Lactate 72 h	0.98	1.06	0.39
Mean Lactate 96 h	0.87	0.9	0.74
Mean UOP 24 h	2.32	2.09	0.99
Mean UOP 48 h	2.26	2.08	0.74
Mean UOP 72 h	2.34	2.78	0.45
Mean UOP 96 h	2.29	2.44	0.60

There were no significant differences between use, duration, and quantity of vasopressors between the two groups. There were minimal complications in both groups, with tachycardia being the most common. There were also no significant differences between resuscitation points.

Disclaimer statements

Contributors None.

Funding None.

Conflicts of interest Authors have no conflict of interests