Introduction

The incidence of cardiovascular abnormalities following acute spinal cord injuries (ASCIs) is very common, particularly after cervical injuries.1, 2, 3 Many patients exhibit hypotension and bradycardia due a sudden loss of sympathetic outflow and relative hypovolemia. This combination, referred to as ‘neurogenic shock’, is common in patients with acute cervical tetraplegia or high thoracic paraplegia. With impaired spinal cord autoregulation, this neurogenic shock can contribute to hypoperfusion of the spinal cord and further ischemic insult. In a study conducted by Lehmann et al.1 100% of the patients that had a severe cervical injury suffered from persistent bradycardia and 68% of the patients suffered from hypotension.

Given the paucity of effective interventions for acute human SCI, it is important that all possible clinical measures be taken to minimize secondary damage and potentially improve neurologic outcome. In traumatic brain injury, restoring normal blood pressure has proven to be beneficial for neurologic recovery, and aggressive vasopressor support with fastidious monitoring of systemic and intracranial pressure is standard of care for these patients in most neurotrauma institutions.4 Comparatively less is known about traumatic spinal cord injury (SCI), however. In two studies, aggressively providing hemodynamic support to maintain adequate perfusion and blood pressure was shown to improve the mortality rates and neurological outcome of patients with ASCI.5, 6

Guidelines have recently been published by the Spinal Cord Medicine Consortium that provide direction for clinicians in their hemodynamic management of patients with ASCI.7 Here, we performed a systematic review of the literature on the topic of hemodynamic support for ASCI, with the goals of characterizing the clinical phenomenon of neurogenic shock, evaluating appropriate vasopressor agents, determining the optimal duration of treatment and identifying the hemodynamic parameters that warrant commencement of vasopressors.

Materials and methods

An electronic English literature search was conducted using the MEDLINE (1950 to April 2008) and EMBASE databases (1974 to July 2009). The search strategy included the following terms: acute spinal cord injury, hypotension, hypertension therapy, vasopressors, cardiovascular abnormalities, spinal shock and neurogenic shock. In addition, the Cochrane Central Register of Clinical Trials and the Cochrane Database of Systematic Reviews were searched in the updated registry of the first quarter of the year 2009. Additional sources included conference proceedings and systematic reviews published from January 2005 to April 2008. Both animal and human studies were included. Full text articles were found for abstracts that referred to treatment options and clinical outcomes for arterial pressure changes following ASCI. The references from these articles were also searched for relevant articles. The first two authors (AP and NY) reviewed the articles and reached a consensus opinion.

The articles were narrowed down to those that answered four questions:

  1. 1

    What patient groups (complete or incomplete SCI, cervical or thoracic level of injury) incur a neurologic benefit from having aggressive hemodynamic support?

  2. 2

    What is the most effective vasopressor regimen?

  3. 3

    What is the optimal duration of treatment?

  4. 4

    What is the optimal mean arterial pressure (MAP) to sustain adequate spinal cord perfusion?

These articles were rated according to their level of evidence.8

Outcome measures in patients with SCIs included the incidence of instituting vasopressor support, the percentage of neurologic improvement and increase in blood pressure.

For comparisons, vasopressor support outcomes were grouped under separate categories (that is, complete vs incomplete injuries, cervical vs thoracic/lumbar injuries, criterion MAP<85 vs <90 mm Hg). Mantel–Haenszel method (χ2-test) was used to combine results and analyze with a fixed effects model. Meta-analyses on these limited data set were performed using Review Manager5 (The Cochrane Collaboration, Copenhagen, 2008).9

Results

Our search showed 374 articles relevant to vasopressor support in patients with ASCI. Most of the studies were review papers but there were 25 papers of experimental clinical studies and 7 human clinical studies. All of the clinical studies were either level III or level IV studies without a control group for comparison.

Experimental studies of hypertensive therapy in animals with SCI

Following experimental spinal cord trauma immediate hypotension occurs (acute phase) followed by hypertension at a later stage (chronic phase). This reaction simulates neurogenic shock and autonomic hyperreflexia in humans.10, 11 Nine studies specifically tested the use of various pharmacologic agents for hemodynamic support in animals with SCIs (Table 1). Mean blood pressure and more specifically, spinal cord blood flow increased significantly using a combination of nimodipine (calcium channel blocker) with epinephrine (vasopressor), dextran (colloid), phenylephrine (vasopressor) and mildly with epinephrine and whole blood.

Table 1 Experimental studies accessing hypertensive therapy in animal models of spinal cord injuries
Full size table

Guha et al.12 found that extreme hypertension induced by epinephrine after SCI did not significantly increase spinal cord blood flow.

What patient groups need vasopressor support (complete or incomplete SCI, cervical, thoracic or lumbar level of injury) according to their neurologic improvement?

Levi et al.5 reported that patients with complete motor deficits are 5.5 times more likely to have a systolic blood pressure (SBP) less than 90 mm Hg at admission than patients with motor function. Tuli et al.13 reported that at admission the SBP is significantly higher for AIS C and D compared with AIS B. Regarding the severity of paralysis (complete vs incomplete), four studies were found comparing the institution and effect of vasopressors in the setting of complete and incomplete spinal cord lesions (Table 2). Patients with complete SCI more frequently needed vasopressor support (P<0.01).

Table 2 Effect of severity of injury on need for vasopressor support
Full size table

A complete cervical cord injury leads to loss of central supraspinal sympathetic control and, therefore, more pronounced hypotension.14 Patients with severe cervical cord injuries are more prone to hemodynamic and cardiac abnormalities within the first 1–2 weeks after injury.1, 2 Lehmann et al.1 found hypotension in 68% of the studied patients with SCI with complete cervical cord injury and no hypotension in either incomplete cervical or thoracolumbar cord injuries. In a retrospective study of patients with SCIs receiving appropriate medical and surgical treatment, motor complete patients initially showed a neurologic motor recovery rate of up to 15% whereas motor incomplete patients showed a motor recovery rate of 80% on average.15

As to the level of paralysis (cervical vs thoracic), only four studies (including Lehmann et al.) evaluated the institution of vasopressor support based on level of injury. Patients with cervical cord injuries required vasopressor support more frequently than patients with thoracolumbar injuries (P<0.001; Table 3).

Table 3 Effect of level of injury on need for vasopressor support
Full size table

This review confirms that the need for vasopressors is more pronounced in cervical cord injuries with complete neurologic deficit. This does not mean that vasopressors are unnecessary in cases of incomplete injuries or thoracic/lumbar cord injuries, but rather, that hypotension is less commonly experienced in these settings.

What is the best hypertensive drug regimen depending on its actions?

In a study by Levi et al.5 on 50 patients with cervical ASCI, dobutamine (5–15 μg kg−1 per min) and/or dopamine (2–10 μg kg−1 per min) were used along with fluids to control patient's hemodynamic profile. Dobutamine was used primarily as it improved cardiac contractility more effectively than dopamine. Vale et al.,6 in a study of 77 patients with ASCI, treated patients with an MAP less than 85 mm Hg with dopamine (2.5–5 μg kg−1 per min) followed by norepinephrine (0.01–0.2 μg kg−1 per min) if necessary.

Five studies using different vasopressor agents were reviewed (Table 4). All of them were successful in improving blood pressure above the targeted critical value. Owing to the heterogeneity of the studies, no statistical analysis could be performed to determine which, if any, of the vasopressor agents are better.

Table 4 Effectiveness of vasopressor support in increasing blood pressure in patients with SCI
Full size table

What is the optimal duration of treatment?

There were four studies reporting the optimal duration of vasopressor support (Table 5). Treatment ranged from 5 to 7 days on average. Most of the reported findings were not specific for their effect on the cardiovascular system or neurologic recovery. Therefore, optimal treatment duration could not be determined.

Table 5 Duration of vasopressor support in patients with SCI
Full size table

What is the MAP below which one should commence vasopressor support?

Four studies (Table 2) evaluated neurologic improvement following pressure support initiation for MAP<85 or <90 as criterion for initiating therapy. There was no significant difference (P>0.05) in terms of neurologic improvement between the initiation of pressures for an MAP<85 or <90 as a criterion for initiating therapy.

Discussion

Hypotension in acute spinal cord injured patients leads to decreased cord perfusion and the potential for further secondary ischemic cord injury.16 Before initiating any vasopressor agent, proper fluid replacement is required to enhance the action of the vasopressors.17 The target pulmonary capillary wedge pressure is approximately 18 mm Hg with a systolic arterial pressure between 80 and 100 mm Hg.18, 19 Volume resuscitation is undertaken first by crystalloids and then by colloids (that is, dextran, fresh frozen plasma or red blood cell units when hemoglobin is low). Other causes of hemodynamic instability (that is, bleeding, tension pneumothorax, myocardial infraction, cardiac tamponade, sepsis) should be excluded before attributing hemodynamic instability to neurogenic shock. Spinal cord blood flow has been shown to be adversely affected following traumatic SCI and an increase in blood pressure leads to significant improvement in axonal function both in the motor and somatosensory tracts of the cord.20, 21

The sympathetic supply to the heart originates in T1–T4 cord level. Injuries at or above this level lead to decreased myocardial contractility and heart rate. In addition, irrespective of the level of cord injury, supraspinal sympathetic drive to the peripheral vessels is interrupted, and the pooling of blood within the peripheral circulation leads to hypotension.22 In Table 6, the most widely used vasopressors are listed. Vasopressors that are α-receptor agonists result in peripheral vasoconstriction and elevation of blood pressure. β-Receptors agonists are responsible for increased cardiac contractility and heart rate. According to the recommendations by the Consortium for Spinal Cord Medicine,7 cervical and upper thoracic cord injuries down to T6 warrant a vasopressor with both inotropic and chronotropic as well as vasoconstrictive properties. Dopamine and norepinephrine have an effect on both α1- and β1-receptors and are reasonable choices.16, 18, 23 For lower thoracic lesions, a peripheral vasoconstrictor is needed. Phenylephrine specifically regulates peripheral vasodilation by acting only on α1 receptors and can be used for lower thoracic and lumbar cord injuries but not for upper thoracic or cervical injuries.16, 18, 23 Dobutamine exerts its effect prominently as an inotropic agent and its use in SCI is limited because of its effect on vasodilation and possible reflex bradycardia.24 Epinephrine, even though it is both an α-receptor and β-receptor agonist, may cause arrhythmias and require close monitoring. Vasopressin is not used in the setting of SCI because of its antidiuretic effects leading to water retention and hyponatremia.17 It is only used in cases of hypotension refractory to catecholamines.23 Isoproterenol, amiodarone and milrinone cannot reverse hypotension in SCI.23

Table 6 Most commonly used vasoactive agents16, 18, 23
Full size table

In SCIs, the choice of the appropriate vasopressor depends on the patient's hemodynamic profile and the level of the cord injury and usually is an agent containing mainly α1-adrenergic activity but also sometimes β1-adrenergic activity.

Neurogenic shock starts immediately after injury when the spinal sympathetic pathways are disconnected at the site of injury while parasympathetic discharge continues unopposed through the vagal nerve. This may last for several days to 6 weeks until segmental neuronal connections and reflex cycles gradually reappear distal to the level of injury, even without sympathetic discharge return.25 At that time, reflex hyperactivity begins manifesting as muscle spasms and elevation of blood pressure.14, 26 Furthermore, Ko and co-workers27, 28 specified that the time of return of deep tendon reflexes after SCI is within the first couple of weeks after injury with the delayed plantar reflex being the first followed by the bulbocavernous and cremasteric reflex. However, no established connection between spinal shock and neurogenic shock does exist.

Vale et al. used a minimum duration of 7 days after injury for hemodynamic treatment of symptomatic patients.6, 16 This duration was determined based on an experimental animal SCI study that showed that between days 3 and 5 after injury, the spinal cord experienced the greatest degree of cord edema and vascular congestion.29 Levi et al.5 used dobutamine at a mean dose of 5.4 μg kg−1 per min for a mean duration of 5.7 days.

Our study found that most clinicians performed vasopressor support for approximately 5–7 days. Longer duration of treatment may interfere with renal function due to the adrenergic effect on the renal arteries.

Mean arterial pressure is determined by arterial catheter placement. It can be determined by the formula MAP=DAP+1/3(SBP−DBP), where DAP stands for diastolic blood pressure and SBP for systolic blood pressure.23 Maintaining MAP above 85–90 mm Hg has been shown to improve the outcome of patients with ASCI.5, 13, 30 MAP should not be elevated above the normal range even in severe cord injuries, as experimental studies have shown.12 The goal of treating hypotension in patients with SCI should be to maintain an SBP at or above 85–90 mm Hg, according to consensus panel of Consortium for Spinal Cord Medicine.7

Conclusions

Given that rectifying systemic hypotension is one of the only clinical interventions that currently appear to influence neurologic outcome after ASCI, it is surprising that so little clinical evidence exists to address the fundamental questions that we posed in this systematic review. This relates to the problems that have plagued the SCI field in evaluating interventions that potentially alter neurologic outcome: the relatively low incidence of ASCI (making patient recruitment into clinical trials difficult), and the strikingly variable extent of spontaneous neurologic recovery among individuals of different ASIA impairment grades (making it necessary to recruit many patients to have sufficient statistical power).31 Intensive management of blood pressure appears to be of neurologic benefit, as it is in traumatic brain injury, but addressing the fundamentally important questions such as ‘what should the target MAP be set at, for how long and with what drugs?’ is obviously challenging. Complicating the matter is the recent report by Kwon et al.32 that indicates that intrathecal pressure may increase postoperatively in patients with ASCI (with a concomitant decrease in spinal cord perfusion pressure). This confounds the interpretation of how well the systemic MAP truly reflects cord perfusion, and raises the question of whether intrathecal pressure monitoring analogous to that performed in traumatic brain injury is warranted. For now, the recommendations are to avoid hypotension and maintain MAP above 85–90 mm Hg for the first 5–7 days after injury with volume expansion and vasopressors as needed. Much more clinical research is necessary in the future to make more refined recommendations for this patient population.