A case of Arnold-Chiari syndrome with flaccid paralysis and huge syringomyelia.

STUDY DESIGN: A case report. SETTING: Department of Orthopaedic Surgery, Shiga University of Medical Science, Japan. PATIENT: A 13-year-old woman presented progressive weakness in the lower extremities, with predominance on the right. Magnetic resonance (MR) imaging revealed a huge syrinx. The patient also showed scoliosis, cleft palate, hearing impairment, excessive sweating, hairiness, dural ectasia, and malformation of the skull. METHOD AND OBJECTIVES: We treated a very rare case of Arnold-Chiari syndrome, which presented with flaccid paralysis. Methods of differential diagnosis and suitable treatment are discussed. RESULTS AND CONCLUSION: Both the syrinx and muscle strength were quickly improved following placement of a syringo-peritoneal (S-P) shunt, after which the patient recovered the ability to walk. However, transient hypesthesia in the right hand occurred after the operation. The syrinx around the conus was thought to play a crucial role in the etiology of the patient case, which showed unique symptoms.

Effects of serotonin 1A agonist on acute spinal cord injury.

STUDY DESIGN: We evaluated the effects of serotonin (5-HT) agonists on in vitro models of spinal cord compressive injury. Evoked potentials in injured rat spinal cords (n=24) were recorded during perfusion with 5-HT agonists. OBJECTIVES: To evaluate the therapeutic effects of 5-HT agonists on the recovery of compound action potentials in injured spinal cords. METHODS: Rat dorsal columns were isolated, placed in a chamber, and injured by extradural compression with a clip. Conducting action potentials were activated by supramaximal constant current electrical stimuli and recorded during perfusion with 5-HT agonists and antagonists. RESULTS: After inducing compression injuries, mean action potential amplitudes were reduced to 33.9+/-5.4% of the pre-injury level. After 120 min of perfusion with Ringer's solution, the mean amplitudes recovered to 62.8+/-8.4% of the pre-injury level. At a concentration of 100 micro M, perfusion with tandospirone (a 5-HT1A agonist) resulted in a significantly greater recovery of mean action potential amplitudes at 2 h after the injury (86.2+/-6.9% of pre-injury value) as compared with the control Ringer's solution (62.8+/-8.4% of pre-injury value, P<0.05). In contrast, quipazine (a 5-HT2A agonist) accelerated the decrease of amplitude (54.5+/-11.7% of pre-injury value). 5-HT1A and 5-HT2A agonist did not consistently alter latencies of the action potentials. CONCLUSION: The 5-HT1A receptor agonist was effective for the recovery of spinal action potential amplitudes in a rat spinal cord injury model.

Kyphotic malalignment after anterior cervical fusion is one of the factors promoting the degenerative process in adjacent intervertebral levels.

The aim of this study was to determine whether postoperative malalignment of the cervical spine after anterior interbody fusion surgery promotes degenerative changes in the neighboring intervertebral discs. Forty-two patients who underwent anterior interbody fusion surgery for cervical spondylosis and disc herniation (34 men, 8 women) were followed for an average of 9.8 years. The average age at surgery was 50.2 years. Twenty-three patients underwent a single-level fusion, 17 underwent two-level fusion, and 2 had three levels fused. The Japanese Orthopaedic Association cervical myelopathy score, with a normal score 17 points, was 11.7 before surgery and 14.9 at follow-up. Neurological status was significantly improved postoperatively, and the improvement was preserved thereafter in most cases (paired t-test, P<0.001). Degenerative changes were evident on radiological examination in the levels adjacent to the fused segment in 21 of the 42 (50%) patients. Eight of these 21 patients demonstrated neurological deterioration caused by an adjacent disc lesion. A total of 43% of the patients with adjacent-level degeneration had malalignment of the cervical spine, such as kyphosis or sigmoid curvature. In addition, degenerative change in adjacent intervertebral levels was observed in 77% of kyphoses of the fused segment. These were statistically significant (Fisher exact method, P<0.05, P<0.04, respectively). Our findings suggest that one of the factors promoting degenerative change in adjacent intervertebral levels after anterior cervical fusion for degenerative disorders is postoperative kyphotic change in the cervical spine and the fused segment.

Prevalence of large-joint osteoarthritis in Asian and Caucasian skeletal populations.

OBJECTIVE: To determine ethnic variations of large-joint osteoarthritis (OA) in past populations. METHODS: One thousand two hundred and nine adult skeletons, excavated from archaeological sites in Japan, China and France were assessed for OA as defined by the presence of eburnation. RESULTS: Within Asian skeletal populations, elbow OA and patellofemoral joint OA were more common in hunter-gatherers than in agriculturalists. Compared with Caucasians, the Asian skeletal population had a higher prevalence of tibiofemoral joint OA. CONCLUSION: The relative frequencies of OA within and between ethnic groups at certain joint sites have changed over time from the past to the present.

Spontaneous atlantoaxial rotatory fixation in old age after cerebral infarction: case report.

STUDY DESIGN: Case report of spontaneous Fielding and Hawkins Type I atlantoaxial rotatory fixation in a 78-year-old man after hemiplegia and homonymous hemianopsia caused by cerebral infarction. OBJECTIVES: To describe a case of spontaneous atlantoaxial rotatory fixation in old age and review previous adult cases of atlantoaxial rotatory fixation without fracture. SUMMARY OF BACKGROUND DATA: Atlantoaxial rotatory fixation in adults is a relatively rare finding and is mainly caused by trauma. To the author's knowledge, there has been no previous report of spontaneous atlantoaxial rotatory fixation in old age. METHODS: The patient's head was fixed in a 40 degrees left-rotated position. Left hemiplegia and homonymous left-side hemianopsia developed due to cerebral infarction. Computed tomography of the cervical spine clearly showed rotatory fixation of the atlas on the axis. RESULTS: Successful reduction was obtained after 1 day of skull traction. CONCLUSIONS: It was hypothesized that repeated left-rotational stress due to homonymous hemianopsia loaded to the atlantoaxial joint caused abnormal laxity of the joint.

A migrated lumbar disc herniation simulating a dumbbell tumor.

We report a case of a migrated lumbar disc hemiation, which on magnetic resonance imaging (MRI) simulated a dumbbell tumor in a 44-year-old woman who had severe pain in her right buttock and leg. A large epidural mass mimicking a dumbbell tumor was detected at the L5 vertebral level by MRI and computed tomography over myelography. Surgical fenestration of the L4/L5 interlaminar space revealed a dorsolateral epidural mass connected to the L5/S1 intervertebral disc extending laterally through the right L5/S1 intervertebral foramen. Histologically, it was degenerative disc material without active inflammation. Reevaluation of the MRI suggested some clues that might be useful in differentiating such a herniated disc from an epidural tumor.

A long-term follow-up study of cervical spondylotic myelopathy treated by "French window" laminoplasty.

We investigated 30 patients with multilevel cervical spondylotic myelopathy (22 males and 8 females) treated by "French window" laminoplasties from 1979 to 1988. Patients averaged 62.6 years of age, and were followed an average of 5 years and 2 months. The average preoperative and postoperative Japanese Orthopaedic Association (JOA) scores were assessed, rising from the original 8.8 +/- 3.4 to 11.9 +/- 3.3 (p < 0.001). Patients were divided into two groups, demonstrating poor alignment and no malalignment after surgery. There was no statistical difference regarding improvement in the neurological examinations between these two groups. The "French window" laminoplasty achieved good clinical results in the management of cervical spondylotic myelopathy irrespective of whether or not alignment deteriorated in the postoperative period.

GABA increases refractoriness of adult rat dorsal column axons.

We applied randomized double pulse stimulation for assessing the effects of GABA and a GABAA antagonist on compound action potentials in dorsal column axons isolated from adult rat. We stimulated the axons with double pulses at 0.2 Hz and randomly varied interpulse intervals between 3, 4, 5, 8, 10, 20, 30, 50 and 80 ms. Action potentials were measured using glass micropipettes. The first pulse was used to condition the response activated by the second test pulse. Concentrations of GABA of 1 mM, 100 microM and 10 microM did not affect action potential amplitudes or latencies activated by conditioning pulses. In the control studies, before drug administration, test pulses induced response amplitudes that were significantly decreased at 3-, 4- and 5-ms interpulse intervals. The test action potential amplitudes were 84.6 +/- 2.5%, 89.0 +/- 3.9% and 93.3 +/- 3.6% (mean +/- S.E.M.) of conditioning pulse levels, respectively. At 3-ms interpulse intervals, test response latencies were prolonged to 104.3 +/- 1.0%, but were unchanged at the other interpulse intervals. The 10 microM, 100 microM and 1 mM concentrations of GABA affected test response amplitudes. Application of 100 microM GABA reduced the amplitudes of test responses at 3-, 4-, 5- and 8-ms interpulse intervals, to 59.2 +/- 3.0%, 70.0 +/- 3.0%, 80.2 +/- 1.1% and 88.6 +/- 3.6% of the conditioning pulse amplitudes, respectively. At both 100 microM and 1 mM concentrations, GABA significantly prolonged the latencies of test responses. Treatment with 100 microM GABA prolonged the latencies of test responses at 3-, 4- and 5-ms interpulse intervals, to 119.3 +/- 3.1%, 107.3 +/- 2.8% and 105.5 +/- 2.5% of conditioning pulse latencies, respectively. The addition of 100 microM bicuculline methochloride, a GABAA antagonist, eliminated the effects of 100 microM GABA. The combined application of GABA and bicuculline (both 100 microM) did not affect amplitudes or latencies of test responses. These results suggest that GABA(A) receptor subtypes are present on the spinal dorsal column axons of adult rat, and that they modulate the excitability of the axons. The randomized double pulse methods reveal that GABA increases refractoriness of adult rat dorsal column axons.

Clinical outcomes of cervical spinal cord injuries without radiographic evidence of trauma.

We investigated 33 cervical spinal cord injury patients (25 males and eight females) without bony injury. Patients whose neurologic recovery had reached a plateau and who had evidence on imaging of persistent spinal cord compression were considered candidates for surgical decompression. When imaging did not show spinal cord compression or patients were maintaining a good neurologic recovery from the early days after injury, we pursued conservative treatment. Age at injury varied from 20 to 76 years (mean, 55.6). Average follow-up was 31 months. Twelve patients were treated conservatively (Group 1). Groups 2 and 3 had surgery. Group 2 (14 cases) had multi-level compression of spinal cord due to pre-existing cervical spine conditions such as ossification of posterior longitudinal ligament, cervical canal stenosis, and cervical spondylosis. Group 3 (7 cases) patients existed single-level compression of spinal cord by cervical disc herniations or spondylosis. We evaluated clinical results according to the Frankel classification, the American Spinal Injury Association (ASIA) scales and Japanese Orthopaedic Association (JOA) scores. Overall improvement of JOA and ASIA scores after treatment was 56.3 +/- 35.5% and 67.1 +/- 38.0%, respectively. Patients in Group 1 showed very good recovery after conservative treatment, with improvement of JOA and ASIA scores being 70.4 +/- 40.2% and 77.4 +/- 34.2%, respectively. The average interval between injury and operation was 4.3 +/- 4.4 months. The improvement of the surgically treated patients (Groups 2 and 3) in JOA and ASIA score was 48.2 +/- 30.7% and 61.2 +/- 39.6% respectively. We obtained good neurological recovery after operation, with significantly more improvement in Group 3 than in Group 2. No significant neurologic recovery had occurred preoperatively in these groups. In such patients operative intervention is essential for neurologic recovery.

Calcium-mediated intracellular messengers modulate the serotonergic effects on axonal excitability.

We carried out experiments to investigate the mechanisms of serotonin-induced axonal excitability changes using isolated dorsal columns from young (seven to 11-day-old) Long-Evan's hooded rats. Conducting action potentials were activated by submaximal (50%) and supramaximal constant current electrical stimuli and recorded with glass micropipette electrodes. In experiment 1, to study Ca(2+)-mediated mechanisms, we superfused the preparations with Ringer solutions containing varying Ca2+ concentrations. Following superfusion with Ca(2+)-free Ringer solution for 4 h, we tested initial responses to serotonin agonists. Studies then were repeated after preparations had been washed for 1 h with Ringer solution containing 1.5 mM Ca2+ and 1.5 mM Mg2+. After 4 h superfusion of Ca(2+)-free Ringer solution, quipazine (a serotonin2A agonist, 100 microM) did not induce significant axonal excitability changes (amplitude change of 1.4 +/- 1.3%, percentage of predrug control level, +/-S.D., n = 6). A 100 microM concentration of 8-hydroxy-dipropylaminotetralin (a serotonin1A agonist) reduced response amplitudes by 36.3 +/- 4.2% (+/-S.D., P < 0.0005, n = 7) and prolonged latencies by 22.3 +/- 4.3% (+/-S.D., P < 0.0005, n = 7). Application of serotonin (100 microM) decreased amplitudes by 6.6 +/- 5.0% (+/-S.D., P < 0.05, n = 6). Extracellular calcium concentration ([Ca2+]e) was measured at various depths in the dorsal column with ion-selective microelectrodes. Four hours' superfusion with Ca(2+)-free Ringer solution reduced [Ca2+]e to less than 0.1 mM in dorsal columns. In 1.5 mM Ca2+ Ringer solution, quipazine increased the amplitudes by 38.3 +/- 5.8% (P < 0.0005, n = 6). Likewise, serotonin increased the amplitudes by 13.8 +/- 4.9% (P < 0.005, n = 6). In contrast however, 8-hydroxy-dipropylaminotetralin still reduced amplitudes by 35.0 +/- 6.4% (P < 0.0005, n = 7) and prolonged latencies by 24.1 +/- 4.5% (P < 0.0005, n = 7). In experiment 2, we investigated calcium-dependent and cAMP-mediated protein kinase signalling pathways to evaluate their role as intracellular messengers for serotonin2A receptor activation. Two protein kinase inhibitors, 50 microM H7 (an inhibitor of protein kinase C and c-AMP dependent protein kinase) and 100 microM D-sphingosine (an inhibitor of protein kinase A and C) effectively eliminated the excitatory effects of the serotonin2A agonist. 100 microM cadmium (a Ca2+ channel blocker) also blocked the effects of quipazine. Neither these protein kinase inhibitors nor cadmium alone affected action potential amplitudes. These results suggest that replacing Ca2+ with Mg2+ blocks the excitatory effects of quipazine but does not prevent the inhibitory effects of 8-hydroxy-dipropylaminotetralin, and calcium-mediated protein kinase mechanisms modulate axonal excitability changes induced by serotonin and its agonist.

Evidence for serotonin sensitivity of adult rat spinal axons: studies using randomized double pulse stimulation.

We have recently shown both inhibitory and excitatory effects of serotonin on neonatal rat dorsal column axons. While neonatal rat dorsal column axons also respond to norepinephrine and GABA, adult rat dorsal columns are insensitive to the actions of both compounds. Therefore, we studied the effects of serotonin agonists on adult rat dorsal column axons using randomized double pulse stimuli at 0.2 Hz with random interpulse intervals of 3, 4, 5, 8, 10, 20, 30, 50 and 80 ms. The serotonin(1A) agonist, 8-hydroxy-dipropylaminotetralin-hydrobromide (8-OH-DPAT), significantly modulated test response amplitudes at 3, 4, 5 and 8 ms interpulse intervals by 29.6+/-4.0%, 17.4+/-2.1%, 9.6+/-2.3%, and 12.4+/-2.2% of conditioning pulse amplitudes, respectively. The mean latencies at 3, 4 and 5 ms interpulse intervals increased by 17.0+/-5.1%, 8.6+/-2.1%, and 5.1+/-1.4%, respectively (P<0.05). However, neither 10 microM 8-OH-DPAT nor 100 microM serotonin hydrochloride affected the compound action potentials evoked by conditioning or test pulses. In contrast, treatment with 100 microM quipazine dimaleate (a serotonin(2A) agonist) decreased the refractory period. While the response amplitudes to a 3-ms double pulse were reduced by 11.0+/-1.5% during the control period, the test response fell to only 2.4+/-1.8% of the conditioning response amplitudes after exposure to 100 microM quipazine. 8-OH-DPAT decreased the amplitude, prolonged the latency and increased the refractory periods of compound action potentials in the adult rat dorsal column, although a high concentration of the agonist (100 microM) was required for these effects. In contrast, the serotonin(2A) agonist, quipazine, decreased refractory periods. These results suggest that both serotonin(1A) and serotonin(2A) receptor subtypes are present on adult spinal dorsal column axons. Further, these receptors have opposing effects on axonal excitability, despite the fact that their sensitivities are relatively low.

Alternating sciatica while jogging: an early symptom of cauda equina tumor.

Three athletic patients with cauda equina or lumbosacral cord tumor noticed, as an early symptom of the disease, alternating bilateral sciatica synchronized with each stride while jogging. Comparison with athletic patients who developed lumbar disc hernia suggested that this symptom was significant. The authors speculated that the mechanism producing this symptom is the inertial force induced while jogging, which acts on the tumor in its early stage, when it is still quite mobile in the intradural space. The diagnostic role of this symptom in cauda equina and lumbosacral cord tumor should be recognized.

The recovery of 5-HT immunoreactivity in lumbosacral spinal cord and locomotor function after thoracic hemisection.

To determine the role of serotonin (5-HT) in recovery from spinal cord injury, we examined spinal cord 5-HT immunohistologically and assessed locomotor recovery after thoracic (T8) spinal cord hemisection in 68 rats. Forty eight rats had laminectomy and hemisection, while the remaining 20 rats received laminectomy only. All rats were evaluated every other day for 4 weeks, using a 0-14 point scale open field test. Hemisection markedly reduced mean hindlimbs scores from 14 to 1.5 +/- 0.32 and 5.6 +/- 0.31 (mean +/- standard error of mean) in the ipsilateral and contralateral side, respectively. The rats all recovered apparently normal walking by 4 weeks. The 5-HT immunohistological study revealed a marked reduction of 5-HT-containing terminals in the ipsilateral but not the contralateral lumbosacral cord by 1 week after hemisection. By 4 weeks after hemisection, 5-HT immunoreactive fibers and terminals returned to the ipsilateral lumbosacral cord, with many 5-HT fibers crossing over the central canal at thoracic level. We estimated the recovery of 5-HT neural elements in lumbosacral ventral horn by ranking 5-HT staining intensity and counting 5-HT terminals. The return of 5-HT immunoreactivity of the lumbosacral ventral horn correlated with locomotor recovery. Locomotory recovery invariably occurred when the density of 5-HT terminals approached 20% of control values. These results indicate that return of 5-HT fibers and terminals predict the time course and extent of locomotory recovery after thoracic spinal cord hemisection.

Effect of mianserin on locomotory function after thoracic spinal cord hemisection in rats.

To study the role of serotonin (5-HT) in spinal cord injury, we observed the effects of mianserin (a 5-HT1c and 5-HT2 receptor antagonist) on rat locomotory function after thoracic spinal cord hemisection. Three groups of rats were studied: sham, A, and B. The sham group (n = 4) received laminectomy and a 3-day course of mianserin (5 mg/kg ip); group A (n = 12) had laminectomy, hemisection, and weekly 3-day courses of saline or mianserin; group B (n = 12) was identical to group A except that the rats received saline. The rats were evaluated every other day for 6 weeks using a 0-14 point scale. Hemisection markedly reduced mean ipsilateral hindlimb scores from 14.0 to 4.0 +/- 0.4 and 4.6 +/- 0.2 (mean +/- standard deviation) in groups A and B, respectively. The saline-treated rats recovered to scores of 9 or 10 by Day 7, 12 or 13 by Day 14, and normal by Day 21. Mianserin significantly but transiently depressed mean locomotory scores, from 12.1 +/- 0.6 to 10.0 +/- 0.4 (P < 0.05, Mann-Whitney U test) in the second week and from 14.0 +/- 0.0 to 12.1 +/- 0.6 (P < 0.05, Mann-Whitney U test) in the fourth week after hemisection. Locomotory scores of mianserin-treated rats did not differ significantly from control saline-treated rats by 7 days after treatment. Immunohistological studies of the spinal cords revealed a marked reduction of 5-HT-containing terminals in ipsilateral but not contralateral lumbosacral cord by 2 weeks after hemisection. By 4 weeks after hemisection, 5-HT-immunoreactive fibers and terminals partly returned to the ipsilateral lumbosacral cord, corresponding temporally with locomotory recovery. Thus, 5-HT may play a role in recovery after hemisection. Anti-serotonergic drugs should be cautiously administered to patients recovering from spinal cord injury.

Excitatory and inhibitory effects of serotonin on spinal axons.

We studied the effects of serotonin on compound action potentials in dorsal columns isolated from young (nine to 13 days old) rats. Conducting action potentials were activated by submaximal (50%) and supramaximal constant current electrical stimuli and recorded with glass micropipettes. At 10 microM and 100 microM concentrations, serotonin significantly increased mean action potential amplitudes by 9.6 +/- 6.5% (+/- S.D., P < 0.05) and 16.6 +/- 12.2% (+/- S.D., P < 0.005), respectively. Likewise, 10 microM and 100 microM of quipazine (a serotonin2A agonist) increased the amplitudes by 9.6 +/- 2.5% (+/- S.D., P < 0.0005) and 37.7 +/- 8.7% (+/- S.D., P < 0.0005), respectively. In contrast, 10 microM and 100 microM concentrations of 8-hydroxy-dipropylaminotetralin-hydrobromide (a serotonin 1A agonist) reduced axonal excitability by -9.4 +/- 5.5% (+/- S.D., P < 0.05) and -32.9 +/- 10.6% (+/- S.D., P < 0.0005), respectively. At 50 microM concentration, mianserin (a serotonin2A and serotonin2C antagonist) eliminated the excitatory effects of 100 microM quipazine dimaleate. The combination of 50 microM mianserin and 100 microM serotonin reduced action potential amplitudes by -5.6 +/- 4.9% (+/- S.D., P < 0.05). These results suggest that serotonin1A and serotonin2A receptor subtypes are present on spinal dorsal column axons. These two receptor subtypes have opposing effects on axonal excitability. The ratios and sensitivities of these two axonal receptor subtypes may modulate axonal excitability in rat dorsal column axons and have important implications for both development and injury of axons.

Parosteal (juxtacortical) chondrosarcoma of the humerus associated with regional lymph node metastasis. A case report.

Recurring parosteal (juxtacortical) chondrosarcoma of the humerus occurred in a 79-year-old man. Roentgenograms of the left humerus showed minimal cortical irregularity and a large soft tissue mass without calcification or periosteal reaction. Magnetic resonance imaging (MRI) showed a multilobular tumor of high intensity (T2-weighted imaging) without marrow involvement. During the left shoulder disarticulation an enlarged axillary lymph nodes was discovered and removed. Macroscopic examination of the disarticulated humerus showed a large yellow parosteal mass, which looked like cartilage, completely surrounding the humeral shaft. The marrow cavity was almost normal except for a small erosion at the site of the olecranon fossa. The histologic diagnosis of the mass was Grade 2 chondrosarcoma of parosteal (juxtacortical or periosteal) origin associated with regional lymph node metastasis. The features that made this case unique were a long history of the tumor (for 32 years), minimal cortical destruction of the bone by the tumor, and regional lymph node metastasis. This case illustrates the differential diagnosis of the parosteal chondrosarcoma from periosteal osteosarcoma and extraskeletal myxoid chondrosarcoma.

Evidence for a neural source of acute accumulation of serotonin in platelets in the injured spinal cord of rats. An experimental study using 5,6-dihydroxytryptamine treatment.

It was found previously that large numbers of platelets showing high serotonin (5-hydroxytryptamine; 5-HT) immunoreactivity appeared in hemostatic plugs at the traumatized cord segment in the acute phase of a trauma. In order to determine the origin of 5-HT in the platelets, we investigated the 5-HT immunoreactivity of platelets accumulating in hemostatic plugs at the traumatized spinal cord segment at 5 minutes after injury. This investigation was carried out by light and electron microscopic immunohistochemistry in rat spinal cord pretreated with 5,6-dihydroxytryptamine (5,6-DHT). The hemorrhagic lesion formed at the neural 5-HT-depleted spinal cord segment was completely 5-HT immunonegative, while platelets in lesions in cord segments of control animals or rostral to the injection site of 5,6-DHT in experimental animals where neural 5-HT was not depleted were 5-HT immunoreactive. The results strongly suggest that a significant amount of 5-HT is released from neural elements at the injury site and is transiently incorporated into the platelets in situ.

Acute aggregation of serotonin-immunoreactive platelets in the injured spinal cord of rat and change of serotonin content in the neural fibers.

Serotonin (5-HT) immunoreactivity was investigated at the injury site of the rat spinal cord by light and electron microscopic immunohistochemistry. The animals were perfused with fixative immediately, 1-5 min, 7-60 min, and 1-48 h after a 3-sec extradural compression of the thoracic cord with a Sugita aneurysm clip. Five minutes after injury, abundant hemostatic plugs containing platelets with high 5-HT immunoreactivity appeared at the traumatized cord segment, whereas 5-HT-containing varicose fibers became less immunoreactive. The 5-HT-immunoreactive platelets initially were localized to the periphery of the hemostatic plugs. By 30 min after injury, the platelets had decreased 5-HT, but neural fibers and terminals began to show increased 5-HT immunoreactivity, which increased and lasted as long as 48 h. These results indicate that platelets contributed to the initial 5-HT concentration increase at the injury site but that later 5-HT accumulation occurred in the neural elements.