Advances in the Rehabilitation of the Spinal Cord–Injured Patient

The Orthopaedic Surgeons' Perspective

Brian K. Kwon, MD, PhD, FRCSC; Dan Banaszek, MD, FRCSC; Steven Kirshblum, MD


J Am Acad Orthop Surg. 2019;27(21):e945-e953. 

In This Article

The Prognostication of Outcome

Mortality After Spinal Cord Injury

Although SCI was historically a universally fatal condition, notable improvements in life expectancy have been achieved over the past 50 years.[1] However, life expectancy is markedly reduced in those who suffer an SCI, particularly in those with cervical injuries and in the elderly.[2,3]

The world's aging population has resulted in a great shift in the demographics of SCI whereby a notable rise in elderly patients who suffer cervical SCI from low-energy falls has generated a bimodal age distribution for SCI incidence and an increase in the mean age of acute SCI to 43 years.[2] Age is the strongest independent risk factor for survival in traumatic SCI, with the incidence of mortality increasing with each additional decade of life.[4] Patients aged 70 to 74 years have a 31.53 times odds ratio for death compared with a patient aged 35 years with similar injuries.[1] Spine surgeons are increasingly faced with elderly tetraplegic patients in whom the risks and benefits of aggressive surgical and rehabilitative treatment and its associated morbidity must be weighed against the expectations of survival and long-term recovery.

Neurologic Recovery After Spinal Cord Injury

Spine surgeons are inevitably asked about the chances for neurologic recovery after injury, highlighting the importance of knowledge about prognostication. At this time, the most commonly used predictor of functional outcome is the severity of injury at baseline, as measured with the International Standards for Neurological Classification of Spinal Cord Injury (ISNCSCI) examination (Supplemental Digital Content 1, with an American Spinal Injury Association Impairment Scale (AIS) A, representing a neurologically complete injury, having the poorest functional prognosis and AIS D having the best. The concept of neurologic recovery has been traditionally characterized by the improvement of AIS grade (conversion), although it is recognized that such AIS grade improvement (although obviously desirable) does not necessarily confer notable functional gains. A detailed review of AIS conversion from the US Model Systems, the prospective randomized Sygen trial,[5] and the European SCI network indicates that approximately 20% of individuals deemed AIS A at baseline convert to incomplete, with half regaining only sensation (AIS B) and half achieving some degree of motor function (AIS C or D).[6] It should be noted, however, that much of these data reflect conversion when the baseline neurologic examination is performed at the time of admission to rehabilitation. If this baseline AIS grade is assigned in the acute setting post-injury, the rate of AIS conversion may be much higher. Since this hyperacute period is typically when the surgeon is involved, he or she should be aware that the initial examination may be less reliable at predicting final outcome. For example, the surgical timing in acute spinal cord injury study, which prospectively evaluated the timing of surgery in patients with acute cervical SCI reported a conversion rate for complete AIS A cervical SCI of approximately 40% in both the early and late surgery groups.[7] This may better reflect the real-world scenario of ISNCSCI examinations performed in the emergency department setting on acutely injured individuals by members of the surgical team with limited time and varying degrees of familiarity with the ISNCSCI examination.

Aside from the baseline severity of neurologic impairment as being a predictor of recovery, surgeons may also consider imaging. Specifically, MRI has been studied extensively as a biomarker of injury severity, and measurable features on MRI may be helpful in establishing prognosis. These include intramedullary lesion length (IMLL) (Supplemental Digital Content 2,, presence of hematoma, and the degree of cord compression. Aarabi recently reported that the IMLL expansion could be used to predict whether AIS grade conversion would occur. The authors found IMLL to be the strongest indicator of AIS conversion, with 1- and 10-mm increases in IMLL corresponding to 4% and 40% decreases in the odds of AIS grade conversion, respectively.[8] The presence of hemorrhage within the cord, the extent of rostro-caudal edema, and the severity of compression cord are all poor prognostic signs and, if present, can give surgeons more confidence in their prognostication discussion with families.

There is also interest in the use of neurochemical biomarkers (eg, proteins, metabolites) within cerebrospinal fluid for biologically stratifying injury severity and predicting neurologic recovery.[9–11] Although there is some evidence that this biologic approach may be even stronger than MRI at predicting outcome,[12] the evaluation of cerebrospinal fluid biomarkers remains investigational at this time.

Ambulatory Function After Spinal Cord Injury

"Will I ever walk again?" is an inevitable question posed to the surgeon, and hence, the prognosis for regaining walking ability is worth considering separately. Several studies have attempted to determine the predictors of recovery of walking ability based on baseline neurologic characteristics. In general, patients with cervical SCI classified as AIS D at admission will regain ambulation, whereas in those classified as AIS C individuals, age plays a strong factor, with those younger than 50 years of age more likely to achieve ambulation than those older than 50.[13] In those with motor complete SCI but who have spared pinprick sensation in the lower extremities (LEs) (L2-S1) at the 72-hour examination, there is an improved likelihood of ambulation at 1 year after injury.[14]

A clinical prediction rule for ambulation outcomes after a traumatic SCI, within 15 days, was described by van Middendorp et al[15] where a combination of age (<65 versus ≥ 65 years), motor scores of the Quadriceps Femoris (L3) and Gastrocnemius-soleus complex (S1), and light touch sensation in dermatomes L3 and S1 provides an accurate prediction rule for a patient's ability to walk at 1-year after injury. This prediction algorithm has been validated, and recently, a simplified model was reported by Hicks et al[16] that uses just three differentially weighted variables to predict the ability to independently ambulate: age older than 65 years, motor score at L3, and light touch at S1. As might be expected, age younger than 65 years with normal motor function in L3 and normal light touch in S1 portends a good prognosis for independent ambulation.