Neurosurgical forum Letters
to
the
editor
Grading Animals with Spinal Cord Injury To THE EDITOR: The recent exchange of letters between Drs. Tator and Thienprasit, et al. (J Neurosurg 42:739-743, June, 1975), on the treatment of experimental spinal cord injuries deserves further comment. The critical issue in my opinion is whether good experimental design and statistical analysis are enough to counteract inadequate data. I contend that the problem with much of the data in the recent literature in the field does not depend on the model of injury used (such details as weight drop, cuff compression) but upon the grading of postoperative deficits based on unrelated events. The Tarlov scale or its variants 5.e contain gradations based on factors such as presence of segmental reflexes, spasticity, bladder and bowel "voluntary" control, and automatism, which do not depend on the presence of surviving long tract axons at the site of injury. There are no physiological grounds for stating that an animal graded as 3 is any less anatomically transected than a Grade 0 animal. The recent review by Grillner' provides an excellent source of reference to this effect, and Grillner's own work suggests that fully transected cats might even be classed as Grade 4 following these criteria. Thus, if Grades 0 to 3 are equivalent in regard to the presence or absence of residual long tract fibers crossing the site of injury, there is no justification for treating them as if they reflected an increasing function along the same continuum, and calculating and comparing means of non-interval and non-ratio scale numbers is a meaningless exercise. It would seem that two things are needed for adequate evaluation of the consequences of cord injury and their treatment: one is an ob-
646
jective measure of residual motor function which is quantifiable along the same continuum, and the second is histological verification of the relative proportion of long fibers spared by the injury. I do not believe that average somatosensory evoked-response (AER) testing will be of much help in this problem, for the presence of AER's only indicates that at least some dorsal funiculus or spinothalamic fibers have been spared, but it yields no estimate of their relative quantity. 1-8 When quantitative and bias-free data become available the use of "double blind" designs will become a good deal less necessary. EDUARDOEIDELBERG,M.D.
Phoenix, Arizona
References 1. Eidelberg E, Kreinick CJ, Langescheid C: On the possible functional role of afferent pathways in skin sensation. Exp Neurol 47:419-432, 1975 2. Eidelberg E, Kriegsfeld BA, Dunn RJ, et al: Diagnostic and prognostic usefulness of averaged somatosensory evoked potential studies after spinal cord injury, in ,Chicago Conference on Neural Trauma, Harper & Row, 1975 (in press) 3. Eidelberg E, Woodbury CM: Apparent redundancy in the somatosensory system in monkeys. Exp Neurol 37:573-581, 1972 4. Grillner S: Locomotion in vertebrates: central mechanisms and reflex interaction. Physiol Rev 55:247-304, 1975 5. Tarlov IM: Spinal Cord Compression: Mechanisms of Paralysis and Treatment. Springfield Ill, Charles C Thomas, 1957 6. Thienprasit P, Bantli H, Bloedel JR, et al: Effect of delayed local cooling on experimental spinal cord injury. J Neurosurg 42: 150-154, 1975
J. Neurosurg. / Volume 43 / November, 1975
Neurosurgical forum RESPONSE: Dr. Eidelberg is to be commended on the important points he has raised in his letter about the grading of residual function in experimental animals after spinal cord injury. I agree that segmental and reflex functions of the cord must be excluded from the criteria used to judge the presence or absence of cord recovery across the injury site. This is precisely why our grading of recovery in monkeys with cord injuries is based only on somatic motor function) In lower animals, even this rigid system would give spurious results. For example, after complete spinal cord transection, rabbits can hop about, and cats and dogs can stand and walk. Dr. Eidelberg has correctly pointed this out in his reference to the paper by Grillner, 1 who has recently reviewed this subject. However, it is important to note that Grillner did not cite any evidence that monkeys with spinal cord injuries can walk. In fact, we know of no information in the literature which claims that monkeys can walk after thoracic spinal cord transection. When we first started our experiments in monkeys we studied the clinical response in three monkeys to complete removal of one segment of the spinal cord at T9-10. There was no difficulty in grading these monkeys as Grade 0 (no recovery) during the 12-week postoperative period. After repeated observation of monkeys with spinal cord injuries at T9-10, we have had little or no difficulty in differentiating purposeful movement from reflex withdrawal. Thus, in monkeys, the grading of recovery based on criteria strictly limited to somatic motor function is valid. Moreover, strict observance of the double-blind protocol and statistical analysis which we described previously should completely counteract the effect of any errors made in the grading of recovery. CHARLESH. TATOR,M.D. Toronto, Ontario
In Memory of Samuel R. Snodgrass To THE EDITOR: This is tO notify you of the death on August 8, 1975, of Samuel Robert Snodgrass, M.D., who was a member of the American Association of Neurological Surgeons. Dr. Snodgrass was Professor of Surgery and Chief of Neurosurgery at the University of Texas Medical Branch in Galveston, Texas, from 1937 to 1968. He was one of the pioneers in neurosurgery in the Southwest. A memorial lectureship was established last year in his name by his former residents at the University of Texas Medical Branch. ROBERTG. GROSSMAN,M.D. Galveston, Texas Mechanism of Transient Ischemia To THE EDITOR: This note is prompted by the interesting case report by Melamed, et aL, in the January issue of the Journal (Melamed E, Lavy S, Reches A, et al." Chronic subdural hematoma simulating transient cerebral ischemic attacks. Case report. J Neurosurg 42:101-103, January, 1975). I would like to add a further consideration of the mechanism of the attacks. The brief neurological symptoms might well correlate with transitory increase of ICP or "plateau wave." These waves, first described by Lundberg, 1 represent a sudden elevation of ICP by 50 to 100 mm Hg over a period of 6 to 15 minutes with a quick return to previous values; the occurrence repeats irregularly and is not correlated to Valsalva maneuvers or changes of systemic blood pressure. The plateau waves seem to depend on vasodilation and vasoconstriction phenomena of the cerebral arterial system. I favor this explanation since the hematoma usually exerts a compressive effect rather than an irritative one over the cortex. SALVATOREMINGRINO,M.D. Padua, Italy
References 1. Grillner S: Locomotion in vertebrates: central mechanisms and reflex interaction. Physiol Rev 55:247-304, 1975 2. Tator CH: Acute spinal cord injury in primates produced by an inflatable extradural cuff. Can J Surg 16:222-231, 1973 J. Neurosurg. / Volume 43 / November, 1975
Reference 1. Lundberg N: Continuous recording and control of ventricular fluid pressure in neurosurgical practice. Acta Psychiatr Neuroi Scand (Suppi) 149:1-193, 1960 647
to
the
editor
Grading Animals with Spinal Cord Injury To THE EDITOR: The recent exchange of letters between Drs. Tator and Thienprasit, et al. (J Neurosurg 42:739-743, June, 1975), on the treatment of experimental spinal cord injuries deserves further comment. The critical issue in my opinion is whether good experimental design and statistical analysis are enough to counteract inadequate data. I contend that the problem with much of the data in the recent literature in the field does not depend on the model of injury used (such details as weight drop, cuff compression) but upon the grading of postoperative deficits based on unrelated events. The Tarlov scale or its variants 5.e contain gradations based on factors such as presence of segmental reflexes, spasticity, bladder and bowel "voluntary" control, and automatism, which do not depend on the presence of surviving long tract axons at the site of injury. There are no physiological grounds for stating that an animal graded as 3 is any less anatomically transected than a Grade 0 animal. The recent review by Grillner' provides an excellent source of reference to this effect, and Grillner's own work suggests that fully transected cats might even be classed as Grade 4 following these criteria. Thus, if Grades 0 to 3 are equivalent in regard to the presence or absence of residual long tract fibers crossing the site of injury, there is no justification for treating them as if they reflected an increasing function along the same continuum, and calculating and comparing means of non-interval and non-ratio scale numbers is a meaningless exercise. It would seem that two things are needed for adequate evaluation of the consequences of cord injury and their treatment: one is an ob-
646
jective measure of residual motor function which is quantifiable along the same continuum, and the second is histological verification of the relative proportion of long fibers spared by the injury. I do not believe that average somatosensory evoked-response (AER) testing will be of much help in this problem, for the presence of AER's only indicates that at least some dorsal funiculus or spinothalamic fibers have been spared, but it yields no estimate of their relative quantity. 1-8 When quantitative and bias-free data become available the use of "double blind" designs will become a good deal less necessary. EDUARDOEIDELBERG,M.D.
Phoenix, Arizona
References 1. Eidelberg E, Kreinick CJ, Langescheid C: On the possible functional role of afferent pathways in skin sensation. Exp Neurol 47:419-432, 1975 2. Eidelberg E, Kriegsfeld BA, Dunn RJ, et al: Diagnostic and prognostic usefulness of averaged somatosensory evoked potential studies after spinal cord injury, in ,Chicago Conference on Neural Trauma, Harper & Row, 1975 (in press) 3. Eidelberg E, Woodbury CM: Apparent redundancy in the somatosensory system in monkeys. Exp Neurol 37:573-581, 1972 4. Grillner S: Locomotion in vertebrates: central mechanisms and reflex interaction. Physiol Rev 55:247-304, 1975 5. Tarlov IM: Spinal Cord Compression: Mechanisms of Paralysis and Treatment. Springfield Ill, Charles C Thomas, 1957 6. Thienprasit P, Bantli H, Bloedel JR, et al: Effect of delayed local cooling on experimental spinal cord injury. J Neurosurg 42: 150-154, 1975
J. Neurosurg. / Volume 43 / November, 1975
Neurosurgical forum RESPONSE: Dr. Eidelberg is to be commended on the important points he has raised in his letter about the grading of residual function in experimental animals after spinal cord injury. I agree that segmental and reflex functions of the cord must be excluded from the criteria used to judge the presence or absence of cord recovery across the injury site. This is precisely why our grading of recovery in monkeys with cord injuries is based only on somatic motor function) In lower animals, even this rigid system would give spurious results. For example, after complete spinal cord transection, rabbits can hop about, and cats and dogs can stand and walk. Dr. Eidelberg has correctly pointed this out in his reference to the paper by Grillner, 1 who has recently reviewed this subject. However, it is important to note that Grillner did not cite any evidence that monkeys with spinal cord injuries can walk. In fact, we know of no information in the literature which claims that monkeys can walk after thoracic spinal cord transection. When we first started our experiments in monkeys we studied the clinical response in three monkeys to complete removal of one segment of the spinal cord at T9-10. There was no difficulty in grading these monkeys as Grade 0 (no recovery) during the 12-week postoperative period. After repeated observation of monkeys with spinal cord injuries at T9-10, we have had little or no difficulty in differentiating purposeful movement from reflex withdrawal. Thus, in monkeys, the grading of recovery based on criteria strictly limited to somatic motor function is valid. Moreover, strict observance of the double-blind protocol and statistical analysis which we described previously should completely counteract the effect of any errors made in the grading of recovery. CHARLESH. TATOR,M.D. Toronto, Ontario
In Memory of Samuel R. Snodgrass To THE EDITOR: This is tO notify you of the death on August 8, 1975, of Samuel Robert Snodgrass, M.D., who was a member of the American Association of Neurological Surgeons. Dr. Snodgrass was Professor of Surgery and Chief of Neurosurgery at the University of Texas Medical Branch in Galveston, Texas, from 1937 to 1968. He was one of the pioneers in neurosurgery in the Southwest. A memorial lectureship was established last year in his name by his former residents at the University of Texas Medical Branch. ROBERTG. GROSSMAN,M.D. Galveston, Texas Mechanism of Transient Ischemia To THE EDITOR: This note is prompted by the interesting case report by Melamed, et aL, in the January issue of the Journal (Melamed E, Lavy S, Reches A, et al." Chronic subdural hematoma simulating transient cerebral ischemic attacks. Case report. J Neurosurg 42:101-103, January, 1975). I would like to add a further consideration of the mechanism of the attacks. The brief neurological symptoms might well correlate with transitory increase of ICP or "plateau wave." These waves, first described by Lundberg, 1 represent a sudden elevation of ICP by 50 to 100 mm Hg over a period of 6 to 15 minutes with a quick return to previous values; the occurrence repeats irregularly and is not correlated to Valsalva maneuvers or changes of systemic blood pressure. The plateau waves seem to depend on vasodilation and vasoconstriction phenomena of the cerebral arterial system. I favor this explanation since the hematoma usually exerts a compressive effect rather than an irritative one over the cortex. SALVATOREMINGRINO,M.D. Padua, Italy
References 1. Grillner S: Locomotion in vertebrates: central mechanisms and reflex interaction. Physiol Rev 55:247-304, 1975 2. Tator CH: Acute spinal cord injury in primates produced by an inflatable extradural cuff. Can J Surg 16:222-231, 1973 J. Neurosurg. / Volume 43 / November, 1975
Reference 1. Lundberg N: Continuous recording and control of ventricular fluid pressure in neurosurgical practice. Acta Psychiatr Neuroi Scand (Suppi) 149:1-193, 1960 647
