Research news and notes

1. Traumatic brain injury in the elderly – attitude and reality 

Older patients are expected to do worse after traumatic brain injury. In the Dec 2008 issue of the journal Neurotrauma, Tokutomi et al [7] tried to improve our understanding of the causes contributing to this general correlation. They collected data on 797 patients from Japan's Neurotrauma Data Bank who were aged 6 years or older and had Glasgow Coma scale of 8 and below on admission or within 48 hours of admission. A variety of ages were represented; 25% of the patients were 69 years old or older. They confirmed the hypothesis that older patients had a higher rate of mortality and lower rates of favorable neurological outcome. One of the causes may have been mass lesions—these were more common in the older patients and were associated with a poorer outcome. However, the outcome of the older patients was even worse in the absence of a mass lesion. The starting GCS score and the occurrence of systemic complications was not significantly different by age group, and multisystem injury was actually less common in the elderly. The authors speculate that mechanisms intrinsic to the aging brain cause the worse outcome, rather than systemic complications or other injuries.

This hypothesis appears plausible, and may be confirmed in further clinical and basic research. There is one aspect of care that may be different at different ages: the perception of medical personnel and caregivers. If the attitude toward the patients and their expected outcome changes with age, this may become a self-fulfilling prophecy. In the January 2009 issue of the same journal, Furlan and Fehlings [4] examined the question of ageist attitudes. They used a questionnaire which was mailed to all the members of the National Neurotrauma Society. Non-responders received a second mailing, for a final response rate of 137 (27.2%). Female responders had more positive attitudes toward old people compared to males. Clinicians showed fewer negative attitudes toward old people in comparison with basic and clinical neuroscientists. Within the limitations of such a small and not necessarily representative sample, the results lend credibility to the suspicion that personal attitudes toward old patients vary widely, and therefore may reflect misconceptions.

2. What does it take to develop a randomized trial of early craniectomy in traumatic brain injury? 

Refractory elevation of intracranial pressure after diffuse traumatic brain injury can potentially be treated with decompressive craniectomy. Early craniectomy makes sense in order to preserve maximal brain function, and a prospective randomized study is considered the best way to prove the efficacy of any intervention. Cooper et al [2] from Melbourne set out to evaluate how big such a study must be in order to demonstrate a significant benefit of early craniectomy. They evaluated 69 consecutive intensive care patients diagnosed over 8 months with severe TBI, found 6 patients eligible, and 5 were randomized. Six months after injury, 100% of the patients received outcome assessments; 40% had a favorable outcome. The authors indicate that a planned multicenter trial involving 18 neurotrauma centers in Australia and New Zealand would require at least 5 years in order to achieve an estimated 210-patient sample size. The authors further suggest a bigger study involving more centers around the world. This rational suggestion presents its own problems. A giant study involving many centers in many countries with different standards and different patient populations, with extensive inclusion and exclusion criteria and inevitable compromises may eventually have statistical power, but is likely to suffer from many other organizational and methodological problems. Proof of efficacy of any but the most common neurosurgical interventions is very difficult. We may need to settle for reasonable and plausible practices, rather than wait for the “proof" of a large randomized trial. Sometimes the large trials raise more questions, rather than provide answers.

3. Which method of laminoplasty is better? 

The question mainly arises outside of the United States (US). This inexpensive and efficacious procedure may be done without instrumentation, and has relatively few practitioners in the US.

Several methods exist; the basic distinction is between “open door” laminoplasty where one side is undermined completely and the other partially, serving as a hinge; a “french door" laminoplasty where the spinous process is split, both laminae are partly cut and tilted, and a spacer inserted to keep the spinous process split and the canal widened. Finally, a third method involves a complete laminectomy with reattachment of the laminae in an elevated position using plate fixation.

Okada et al [6] compared open door laminoplasty with a french door laminoplasty in a prospective randomized trial in 40 patients. In their methodology, the french door method included reattachment of the muscles and the spinous processes. Outcomes were measured mainly with the Japanese Orthopedic Association (JOA) and Short Form 36 (SF-36) scales. JOA scale outcomes were equivalent, indicating that both methods resulted in similar decompression. French door laminoplasty patients had higher SF-36 scores, indicating better overall recovery, possibly associated with better muscle reattachment and faster healing.

In a separate study, Asgari et al [1] compared open door laminoplasty with bilateral cutting of laminae (BL) with elevation and miniplating in 13 patients. Outcomes were measured with JOA and Nurick scales. In this study, open door laminoplasty appeared better. Two of the patients in the BL group had a postoperative complication caused by anterior migration of the construct. The plating was probably not strong enough. These are small studies, and the specific techniques differ among centers. It is hard to find a neurosurgical procedure more variable than a laminoplasty.

More studies of laminoplasty are needed to compare different methods, especially those that use plating and other instrumentation, and those that do without. In my experience, the construct of an open door laminoplasty is stable without the need for instrumentation if autograft is used as spacers to prevent ventral displacement of the lamina. Laminoplasty should further be compared with laminectomy, with and without instrumented fusion, in cases where either procedure may be indicated.

4. Posterior instrumentation alone may be biomechanically adequate in cervical and cervicothoracic spine 

“Junctional level” spinal injuries at the cervicothoracic junction are potentially more difficult to stabilize compared with pure cervical or thoracic injuries. A clinical comparative study is not realistic, but it is possible to understand the biomechanics in the lab. O'Brien et al [5] evaluated the stability of a posterior-only construct in a cadaver model of 2- and 3-column flexion-distraction injuries at the cervicothoracic junction. The specimens were tested for mobility in all 3 planes (flexion-extension, axial rotation and lateral bending. The procedure included C6-T2 fixation with lateral mass (C6) and pedicle (C7-T2) screws using dual diameter rods. The specimens were retested; C7-T1 was further destabilized to a 3-column injury and specimens retested once again. The motion was reduced compared to baseline by 85% in 2-column injury and by 75% in 3-column injury. Then the authors added cross-links, which further stabilized the spine in a 3-column injury but did not add anything to stabilization in the 2-column injury. Maximal effect was obtained with 2 cervical or 1 thoracic crosslink. The authors concluded that a modern posterior stabilization procedure may be adequate biomechanically even for a 3-column injury. Interestingly, another recent article [3] demonstrated a lack of additional benefit from anterior instrumentation when posterior 5 level lateral mass fixation was used. The authors tested each of 10 fresh frozen human cadaveric specimens in the intact state, after C4-6 laminectomy, with C3-7 lateral mass instrumentation, and finally with C3-7 posterior plus anterior instrumentation.

These are intriguing results, although it is difficult to translate them directly to clinical practice. One important issue is durability. Will the posterior-only construct retain its stability after multiple motion cycles? The best biomechanical test should include repetitive automatic cycling of the spine to model several months worth of neck movements.