The multicenter trial of Sygen® in acute spinal cord injury (SCI) recruited a total of 760 patients at 28 neurotrauma centers in North America in a 5-year period between 1992 and 1997. The data were carefully monitored by a centralized team that visited the sites regularly.
Results concerning the efficacy of Sygen® will be presented in the third of this series of articles. 24 The current article and the next 23 are devoted solely to using this database to characterize the patients studied: 1) their recruitment and early treatment and 2) their measurement and natural recovery patterns.
The sample recruited in any controlled clinical trial is determined first by the population at large in the community, but only as filtered through the study design. It is convenient for a study to use those centers that can rapidly recruit patients, but this biases the sample toward injury types that occur in areas of high population density. The choice of study centers is not random, and results may reflect their particular forms of neurotrauma systems organization or treatment choices.
A study’s inclusion and exclusion criteria are designed to show treatment efficacy (within ethical and practical limitations), but they inherently define the target population to which the results apply. For example, exclusion of polytrauma or of penetrating injury affects the apparent patterns of recovery and of adverse events.
As statisticians emphasize, the laws of probability and the formulas derived from them apply strictly only to samples randomly selected from a focused target population. The Sygen® Study results cannot be regarded as an epidemiologic investigation representative of the general SCI population.
Still, valid data are scarce, and perhaps with cautious understanding the Sygen® Study data can be used to supplement our knowledge of the state of acute SCI. The last major epidemiologic study was the National SCI Database collected by Stover et al, a model SCI system program that included data on 9647 patients from 19 regional centers. Their capture was estimated to range from 3% of all U.S. SCI in 1973 to 14% in 1985. 1,4,8–16,21,22,28,29
Another way in which the Sygen® Study data can be important is in making it easier to launch badly needed future studies of acute SCI by providing a record of experience on which basis treatment effect size and distribution estimates, and recruitment estimates can be based and made credible.
These are the goals of the current article and the next 23 in this series. Neither is specifically concerned with the effectiveness or ineffectiveness of Sygen®. We have furnished the third article with ammunition for use by adherents of both sides. 24 In the current and second 23 articles we are concerned only with the underlying population: with their clinical treatment and with design of further studies. 23
As Figure 1 illustrates, there are several categories of spinal column/cord injuries. Anatomic and clinical considerations would lead one to suspect that patients in different categories may have different diseases with differing treatments and prognoses.
The Sygen® trial was designed to recruit patients with an SCI that was severe but still an essentially pure SCI with potential of recovery. The study required an injury rostral to the T10 bony level resulting in at least one lower extremity with an ASIA motor score less than 15 of 25. In contrast, the NASCIS studies chose to require an SCI as “determined by physician,” which allowed a wider variety, including a large number of subjects with no or minimal motor deficit and with cauda equina and conus injuries. This difference in recruited patients makes direct comparison of the two data sets unreliable because of major differences in recovery patters with different initial severities.
At the time the Sygen® Study was designed, methylprednisolone (MPSS) was widely given. The possibly time-dependent interaction of MPSS and Sygen® was largely unknown and could not be assumed synergistic. A rat model of traumatic acute SCI had apparently demonstrated antagonism between MPSS and Sygen®. 7 Therefore, the Sygen® Study was designed to give all patients the NASCIS II regimen of MPSS (starting ≤8 hours after SCI and continuing for 24 hours) followed by study medication to be started no more than 72 hours after injury.
Before randomization patients were required to have blood pressure restored and were to have written informed consent signed by the patient or by the parent or legal guardian. Patients were to be excluded who met any of the following criteria:
- Trauma caused by ballistic or other penetrating injury that directly penetrates the spinal cord.
- Traumatic spinal cord anatomic transection.
- Presence of cauda equina damage, major brachial, lumbar plexus, significant head trauma, or other injury that was, in the opinion of the physician, sufficient to interfere with assessment of spinal cord function or otherwise compromise the validity of the patient’s data.
- Other significant systemic disease such as lung, liver, gastrointestinal, or kidney disease; or active malignancy or any other condition as determined by history or laboratory investigation that could alter the distribution, accumulation, metabolism, or excretion of the study medication, cause a neurologic deficit, or result in the patient’s life expectancy being less than 2 years.
- Any pre-existing polyneuropathy, focal or multifocal neuropathy, myelopathy, or radiculopathy by history, clinical examination, or electrophysiologic study that could be reasonably expected to interfere with any of the study assessments.
- Presence of any medical condition that could reasonably have been expected to subject the patient to unwarranted risk from participation in this study or result in a significant deterioration of the patient’s clinical course.
- History of Guillain-Barré syndrome.
- Psychoactive substance use disorder (as defined by DSM III-R) any time during the 6 months preceding study entry.
- History of major depression, schizophrenia, paranoia, or other psychotic disorder as defined by DSM III-R.
- Pregnant or nursing women.
- History of a life-threatening allergic or immune-mediated reaction.
- Not likely to be available for follow-up evaluation.
- Inability to communicate effectively with the neurologic examiner such that the validity of the patient’s data could be compromised.
- Previous use of any ganglioside preparation (e.g., Sygen®, Cronassial, or Sinassial).
The recruitment target was calculated as 720 evaluable patients, and the number to be randomized was set at about 800 to meet this.
There were 3165 patients screened, of whom 2368 (75.7%) were excluded before randomization for failure to satisfy the entry criteria, as shown in Table 1. A total of 797 were randomized in a 5-year recruitment period from April 1992 to January 1997, at an average rate of approximately 14 patients per month, with a very slight seasonal variation. Two of these patients received no study medication, and 35 were adjudicated to have gross violations of the selection criteria making them ineligible for the primary efficacy analysis, and were not followed up.
Thus, data from 760 patients were available for the primary efficacy analysis. Table 2 shows their demographic characteristics.
The median age (Figure 2) was 30 years, and the number of subjects fell almost linearly with age, from 17 to 69. Although the recruitment was greatest in the young, this tendency was not as disproportionate as expected. These results may have been skewed by the types of injury treated by the study’s university-affiliated trauma centers.
The ages were different across study strata, as shown in Figure 3. The patients with more severe injury at baseline were younger (P < 0.0001) and at any severity level patients with thoracic injury were younger than those with cervical injury (P < 0.005).
As shown in Table 3, more than half the injuries involved motor vehicles. The categories with the next numbers of patients were injuries from falls and injuries related to water. The study design did allow enrollment of concussion gunshot injuries to the vertebra without direct penetration of the spinal cord, and there were 35 of these.
How the Patients Presented
The scales used for Sygen® Study (discussed further in the second article 23 in this series) were selected because they have been standard, widely accepted tools readily administered in the acute SCI setting. 2,3,5,6,17–20,25–27Table 4 shows the distribution of the 760 eligible patients by baseline severity and by spinal level. There were many more cervical than thoracic injuries. In both the cervical and thoracic strata the majority of patients are in ASIA Impairment Scale (AIS) Group A, the most severe, and this disproportion is strongest in the thoracic stratum.
This is again apparent in Figure 4, which shows the level of bony injury and distinguishes the levels of baseline AIS severity. The bottom half of the figure shows how different the pattern is in the AIS A group compared with the B and combined C + D groups, although at some spinal levels there are too few patients to make a reliable judgment about the relative prevalence. The commonly noted peak in lower thoracic and upper lumbar injuries is not seen here because these cases were excluded by design. This exclusion maintained the study as an essentially pure SCI study by not including cauda equina injuries.
Patients with multiple trauma were not necessarily excluded unless their other injuries would hinder collection or interpretation of their data. Table 5 shows abnormalities associated with initial trauma by body systems. The most frequent associated injuries were related to the skin and soft tissue (63.4% of subjects presented abnormalities within this body system) and then those to the musculoskeletal system (25.9%), to the lungs (22.6%), and the head (22.4%). Considering that motor vehicle accidents were the leading cause of trauma, high percentages of injuries within the skin and soft tissue and musculoskeletal systems are expected.
Figure 5 shows the vital signs as assessed on arrival at the SCI acute care study facility, before randomization. Blood pressures were relatively normalized, as desired. The data do not support a picture that the majority of the patients are especially hypotensive or bradycardic. Only about the lowest quartile are hypotensive or bradycardic. Blood pressure is low, but perhaps not dangerously so. The spread of heart rate measurements in all patients is perhaps wider than one might expect and is a little low. The figure validates that heart rate is lower in cervical injuries because of their sympathectomy. Temperature is depressed in some who are unable to maintain regulation or had cold exposure.
Baseline physiologic measurements taken at the time of randomization were improved from these study center admission values, implying early treatment of hypovolemia and restoration of blood pressure. This is presumably important in preventing secondary physiologic injury early after SCI. The EMT and SCI study centers have protocols for SCI that call for early mechanical stabilization and resuscitation.
The clock time and day of the week are shown in Figure 6. As expected, more injuries happen on weekend days than on weekdays, and most happen between late afternoon and early morning. The clock time of arrival at the SCI study center follows this trend (whether or not the patient was first sent to another emergency room), as does the time traction was started. These facts support the idea that the injury was treated as an emergency, regardless of time of day.
Figure 7 shows the beginning of the sequence of times in the patients’ treatment milestones. The time until an EMT arrived at the scene of the injury was remarkably short: in 25% of cases it was less than 4 minutes (or 0.07 hours), in 50% it was less than 12 minutes, and in 75% it was less than 25 minutes. The median time for the patient’s departure by helicopter or ambulance is 39 minutes.
The sequence of times progressively widens (Figure 8) as care progresses. For those patients who were taken to another emergency room before arriving at the spinal study center, the median time of arrival was 50 minutes and the median stay there was 268.5 minutes. For the total sample the median time of arrival at the SCI study center was 200 minutes. As shown in Table 6, the centers displayed a lack of uniformity in the degree to which patients were admitted directly or stopped first at another ER, reflecting local variation in patient access to tertiary SCI care. A total of 469 of 760 (62%) patients stopped at another emergency room before the tertiary acute SCI center.
The time to MPSS therapy (Figure 9) was short. The dosage by body weight (not shown) was remarkably uniform, especially considering that MPSS administration may have started in the field and continued through one or more emergency rooms before ending possibly at the SCI center. There was minor variation in the length of MPSS infusion (not shown).
Decompression and Surgery
Figure 10 shows details of cervical traction. The spread of the weight curves suggests a mixture of low and high weight traction, following the two schools of thought on the subject and variations in bony pathology. Traction was most often discontinued perioperatively.
The time to surgery (Figure 11) was not notably short. The 25th percentile (among patients who had surgery at all) was well over 1 day (30.8 hours); half had their surgery after 3 days (median 76.4 hours).
Figure 12 shows that patients with baseline AIS severity B were operated earlier compared with those in severity Group A. Those in Group C + D were operated latest. Cervical injuries were operated earlier than thoracic ones. The difference among the severity groups was statistically significant (P < 0.025), but not the one between the spinal levels.
The differences in time to surgery among the centers (Figure 13) were wide, and a few of them are even statistically significant, despite the small samples, using Tukey’s simultaneous inference procedure that compensates for the risk of Type 1 error in multiple comparisons. This reflects differing physician philosophies on optimal surgical timing and possibly different local injury patterns.
Table 7 shows the numbers of operations according to various classes of patients and stratification variables.
Length of Acute Care and of Rehabilitation
Time from injury to acute care discharge is shown in Figure 14, by baseline AIS. The more severely injured patients stayed slightly longer. Figure 15 similarly shows the distribution of the length of the inpatient rehabilitation. The peak at 90 days appears to be driven by insurance and not by the severity of the SCI. Thus, the length of stay in rehabilitation is not a medical indicator.
The inequality of recruitment by spinal level and by severity has negative implications for any trial in spinal injury. Of the six cells in Table 4, a single one (cervical A) accounted for 332 of the 760 subjects recruited in the Sygen® Study. That row of the table shows 579 of the total 760 subjects with cervical injuries: just more than 75% of the sample. Further, that column of the table shows 482 subjects with AIS severity A, or nearly two thirds of the sample. At the other end of the scale, recruitment into the thoracic B and thoracic C + D groups seems hopelessly slow: in 5 years 28 study centers combined only recruited 18 and 13 of these, respectively.
In any clinical trial where the population is inhomogeneous, it is imperative to consider whether the experimental treatment might work more or less well in one subgroup rather than another. Even if recruitment is balanced between the subgroups, they are likely to be too small individually to have enough statistical power to demonstrate an effect unless the treatment is indeed effective in all subgroups. If the treatment is only effective in one subgroup, then pooling will dilute the result.
If recruitment into the subgroups is strongly unequal, then the results can be difficult to interpret. One suspects a priori that cervical and thoracic injuries have different recovery patterns and potentially different response to treatment, and that the severity strata A, B, and C + D are heterogeneous. The second article 23 in this series will show in detail that these suspicions are well founded.
This result in Table 4 points to a problem in any study of acute SCI: the tendency of the larger cervical and AIS A subgroups to drive the results. Any treatment that does not work in the cervical A group is unlikely to be found effective, whether it works in the other groups or not. Conversely, any treatment that is effective in that group has a better chance, rightly or wrongly, of being found to work in SCI, perhaps without qualification.
The data do seem to present a promising picture of the state of SCI treatment in North America, if our interpretation is correct. The injury is regarded as a true emergency. The EMT arrives early on the scene and treats according to local protocol, with training in the physiology and in mechanics, minimizing secondary injury. Patients are rapidly referred to specialized care. Acute management is designed to address the problems of lessening mechanical compression through traction, immobilization, surgery, and correcting physiologic abnormalities. It is impressive to see how closely the relatively complex MPSS protocol was followed in this sample.
Many aspects of acute SCI do not receive uniform treatment, even in major university-affiliated centers. Table 6 shows wide difference in the rate of direct admission to specialized centers. Another topic on which specialists differ is traction weight, which is reflected in Figure 10, although the differences there are also due to diversity among the patients. A similar issue is the wide difference in time to surgery, as shown in Figure 13.
However, the difficulties in interpreting such data are illustrated by Figure 12. One might expect that the length of time to surgery could increase (or, for a different reason, could decrease) with severity. Neither of these two theories is true. The shortest times are in AIS Group B. It may be that the patients in Group A are too severely injured so that physicians wait for stabilization and the patients in Groups C and D are mildly injured and so physicians feel more comfortable waiting, but such theories are post hoc speculation, possible but unsupported by any data in this article.
It might seem desirable to make treatment more uniform, but practice guidelines should not outstrip the state of our actual knowledge. The data in this article and the next 23 shed some light on the issues, but not nearly enough. They emphasize how difficult it is to gather prospective, statistically meaningful data specifically focused on the relevant issues, not confounded by other factors. There will always be issues for which it is not feasible to collect relevant data, and for these issues it is diversity, not uniformity, of practice that gives the best hope of innovation and of optimizing treatment in the context of a patient’s specific presentation and response. But with a sustained commitment to research more could be known, the physician-driven improvement process could accelerate, and patients could improve more.
- This study is a retrospective analysis of recruitment and early treatment in carefully monitored data from 760 patients in a clinical trial.
- There were strong recruitment imbalances in favor of cervical and of complete injuries.
- The vital signs and the time patterns suggest protocol-driven stabilization to prevent secondary physiologic injury early after SCI.
- Some features of care vary widely among centers.
- Inpatient rehabilitation appeared driven by insurance in addition to severity.
Frank Dorsey, Francesca Patarnello, and Simonetta Piva provided statistical analysis and William Taylor provided SAS software support for data analysis.
1. Stover SL, DeLisa JA, Whiteneck GG. Spinal Cord Injury. Clinical Outcomes From the Model Systems. Gaithersburg, MD: Aspen, 1995.
2. American Spinal Injury Association/IMSoP. International Standards for Neurologic and Functional Classification of Spinal Cord Injury. Atlanta: American Spinal Injury Association, 1992.
3. Bracken MB, Webb SB Jr, Wagner FC. Classification of the severity of acute spinal cord injury: implications for management. Paraplegia 1978; 15: 319–26.
4. Charles ED, Fine PR, Stover SL, et al. The costs of spinal cord injury. Paraplegia 1978; 15: 302–10.
5. Cohen ME, Ditunno JF Jr, Donovan WH, et al. A test of the 1992 International Standards for Neurological and Functional Classification of Spinal Cord Injury. Spinal Cord 1998; 36: 554–60.
6. Coleman W, Benzel E, Cahill D, et al. A critical appraisal of the reporting of the National Acute Spinal Cord Injury Studies (II and III) of Methylprednisolone in Acute Spinal Cord Injury. J Spinal Disord 1999; 13: 185–99.
7. Constantini S, Young W. The effects of methylprednisolone and the ganglioside GM1 on acute spinal cord injury in rats. J Neurosurg 1994; 80: 97–111.
8. DeVivo MJ, Black KJ, Stover SL. Causes of death during the first 12 years after spinal cord injury. Arch Phys Med Rehabil 1993; 74: 248–54.
9. DeVivo MJ, Fine PR, Maetz HM, et al. Prevalence of spinal cord injury: a reestimation employing life table techniques. Arch Neurol 1980; 37: 707–8.
10. DeVivo MJ, Kartus PL, Rutt RD, et al. The influence of age at time of spinal cord injury on rehabilitation outcome. Arch Neurol 1990; 47: 687–91.
11. DeVivo MJ, Kartus PL, Stover SL, et al. Cause of death for patients with spinal cord injuries. Arch Intern Med 1989; 149: 1761–6.
12. DeVivo MJ, Kartus PL, Stover SL, et al. Seven-year survival following spinal cord injury. Arch Neurol 1987; 44: 872–5.
13. DeVivo MJ, Rutt RD, Black KJ, et al. Trends in spinal cord injury demographics and treatment outcomes between 1973 and 1986 [published erratum appears in Arch Phys Med Rehabil 1992;73: 1146]. Arch Phys Med Rehabil 1992; 73: 424–30.
14. DeVivo MJ, Shewchuk RM, Stover SL, et al. A cross-sectional study of the relationship between age and current health status for persons with spinal cord injuries. Paraplegia 1992; 30: 820–7.
15. DeVivo MJ, Stover SL, Black KJ. Prognostic factors for 12-year survival after spinal cord injury [published erratum appears in Arch Phys Med Rehabil 1992;73: 1146]. Arch Phys Med Rehabil 1992; 73: 156–62.
16. DeVivo MJ, Stover SL, Fine PR. The relationship between sponsorship and rehabilitation outcome following spinal cord injury. Paraplegia 1989; 27: 470–9.
17. Ditunno JF, Ditunno JF Jr. American spinal injury standards for neurological and functional classification of spinal cord injury: past, present and future. 1992 Heiner Sell Lecture of the American Spinal Injury Association. J Am Paraplegia Soc 1994; 17: 7–11.
18. Ditunno JF Jr, Graziani V, Tessler A. Neurological assessment in spinal cord injury. Adv Neurol 1997; 72: 325–33.
19. Ditunno JF Jr, Young W, Donovan WH, et al. The international standards booklet for neurological and functional classification of spinal cord injury. American Spinal Injury Association. Paraplegia 1994; 32: 70–80.
20. Dvir Z. Grade 4 in manual muscle testing: the problem with submaximal strength assessment. Clin Rehabil 1997; 11: 36–41.
21. Fine PR, Kuhlemeier KV, DeVivo MJ, et al. Spinal cord injury: an epidemiologic perspective. Paraplegia 1979; 17: 237–50.
22. Fine PR, Stover SL, DeVivo MJ. A methodology for predicting lengths of stay for spinal cord injury patients. Inquiry 1987; 24: 147–56.
23. Geisler FH, Coleman WP, Grieco G, et al. Measurement and recovery patterns in a multicenter study of acute spinal cord injury. Spine 2001; 26 (suppl 1): S68–86.
24. Geisler FH, Coleman WP, Grieco G, et al. The Sygen®
multicenter acute spinal cord injury study. Spine 2001; 26 (suppl 1): S87–98.
25. Marino RJ, Rider-Foster D, Maissel G, et al. Superiority of motor level over single neurological level in categorizing tetraplegia. Paraplegia 1995; 33: 510–3.
26. Maynard FM, Bracken MB, Creasey G, et al. International Standards for Neurological and Functional Classification of Spinal Cord Injury. American Spinal Injury Association. Spinal Cord 1997; 35: 266–74.
27. Medical Research Council of the U.K. Aids to the Examination of the Peripheral Nervous System: Memorandum No. 45. London: Her Britannic Majesty’s Stationery Office, 1976.
28. Stover SL, DeVivo MJ, Go BK. History, implementation, and current status of the National Spinal Cord Injury Database. Arch Phys Med Rehabil 1999; 80: 1365–71.
29. Stover SL, Fine PR. Spinal Cord Injury: The Facts and Figures. Birmingham, AL: University of Alabama, 1986.