The pathophysiology of syringomyelia - historical overview and current concept.

Various ideas and hypotheses have been brought forward to explain the development of syringomyelia in the past two centuries. None of them offers a sufficient basis to serve as a concept for the treatment of all affected patients. Apart from a discussion of the different hypotheses this paper proposes a new pathophysiological concept based on clinical, experimental and literature studies. Syringomyelia is understood as a state of chronic interstitial edema of the spinal cord due to accumulation of extracellular fluid (ECF). This accumulation is caused by a cascade of events starting with obstruction of cerebrospinal fluid (CSF) flow and/or spinal cord tethering which ultimately alter ECF flow and increase ECF volume. Treatment should be targeted against the pathological process which causes CSF flow obstruction and cord tethering to inhibit this pathophysiological process at a decisive point.

Thoracic intradural extramedullary lipomas. Report of three cases and review of the literature.

BACKGROUND: Indication for surgery and the surgical strategy for treatment of spinal intradural lipomas are still discussed controversially. Among spinal lipomas the cervicothoracic region is rarely affected. We report on our experience with spinal decompression for the treatment of three intradural thoracic lipomas. FINDINGS: Three adult patients with extramedullary intradural lipomas of the thoracic spinal cord demonstrated progressive neurological symptoms and signs without evidence of growth of the lipoma. All patients underwent surgery with decompression of the affected spinal levels. No attempt was made to reduce the size of the lipomas. Postoperatively, each patient demonstrated significant clinical improvement. Interpretation. Attempts to remove lipomas are associated with significant risks of surgical morbidity. Lipomas are hamartomas which change their size according to alterations of body fat. Therefore, decompression of the affected spinal levels is sufficient to achieve significant neurological improvement.

Spontaneous resolution of Chiari I malformation and syringomyelia: case report and review of the literature.

OBJECTIVE AND IMPORTANCE: Indications for surgery and the surgical technique of foramen magnum decompression for patients with Chiari I malformation and syringomyelia are controversial issues. This case report supports the view that observation may be adequate for patients without progressive symptoms or with mild clinical symptoms. CLINICAL PRESENTATION: A 37-year-old woman presented with a 3-month history of burning dysesthesias and hypesthesia in her right arm. A neurological examination revealed hypesthesia in the right trigeminal distribution. A magnetic resonance imaging scan revealed a Chiari I malformation with syringomyelia between C2 and T2. No hydrocephalus was observed. CLINICAL COURSE: Because the patient's symptoms regressed spontaneously, surgery was not performed. Thirty-two months after her initial examination, the patient was asymptomatic. A second magnetic resonance imaging scan was obtained, which demonstrated complete spontaneous resolution of the Chiari I malformation and syringomyelia. CONCLUSION: We attribute the regression of the patient's symptoms to spontaneous recanalization of cerebrospinal fluid pathways at the foramen magnum, which most likely was due to rupture of the arachnoid membranes that had obstructed cerebrospinal fluid flow.

Disturbances of cerebrospinal fluid flow attributable to arachnoid scarring cause interstitial edema of the cat spinal cord.

OBJECTIVE: Spinal arachnoid scarring may be caused by trauma, inflammation, surgery, spinal instability, degenerative diseases, or malformations and may lead to progressive neurological deficits and syringomyelia. We wanted to investigate the effects of focal arachnoid scarring in the cervical spinal canal of cats on pressures in the subarachnoid space and spinal cord tissue, as well as on spinal cord histological features. METHODS: Twenty-nine adult cats were used for this study. Nine animals served as control animals, whereas 20 animals received a focal arachnoid scar at C1-C2, which was produced by placement of a kaolin-soaked fibrin sponge on the posterior surface of the spinal cord. After 4 months, pressure recordings above and below the scar, in the subarachnoid space and spinal cord, were performed. Elasticity measurements were performed with small bolus injections. Morphometric analyses of brain and ventricle volumes, sizes of the central canal, and sizes of the perivascular spaces in gray and white matter were also performed. RESULTS: No animal developed clinical or neurophysiological evidence of neurological symptoms at any time. In the kaolin-treated group, pressure recordings revealed a significant increase in the subarachnoid pressure at C1, because of the cerebrospinal fluid flow obstruction. Pressure gradients tended to increase at all measuring points. A significant difference was detected between the spinal cord and subarachnoid space at C2, where the intramedullary pressure exceeded the subarachnoid pressure. Elasticity was significantly increased in the spinal cord at C2. Intracranially, no evidence of hydrocephalus was observed. In the spinal cord, perivascular spaces were significantly enlarged in the posterior white matter above the arachnoid scar and in the central gray matter below the area of scarring in the cervical cord. CONCLUSION: Arachnoid scarring at C1-C2 produces an interstitial type of edema in the central gray matter below the area of scarring in the cat cervical cord, because of altered cerebrospinal fluid and extracellular fluid flow dynamics. These changes may be interpreted as the initial stage in the development of syringomyelic cavities.

Neurenteric cyst of the craniocervical junction. Report of three cases.

Neurenteric cysts of the craniocervical junction (CCJ) are very rare lesions. Their origin is the subject of long-standing controversy, but a failure during the embryogenic phase may be responsible for their formation. Accurate histopathological diagnosis may be difficult due to the similarity they share with other cystic lesions such as colloidal cysts, Rathke cysts, and cystic teratomas. Surgical removal is the treatment of choice for intracranial neurenteric cysts, but in some cases, infiltration of the surrounding structures may hinder complete resection. Three cases of neurenteric cysts located at the CCJ are reported.

Use of high-dose corticosteroids and high-frequency oscillatory ventilation for treatment of a child with diffuse alveolar hemorrhage after bone marrow transplantation: case report and review of the literature.

BACKGROUND: Other than relapse, pulmonary complications are the most common cause of mortality in patients who undergo bone marrow transplantation (BMT). Diffuse alveolar hemorrhage (DAH) is one noninfectious pulmonary complication of BMT. Presenting clinical findings include nonproductive cough usually without hemoptysis, dyspnea, hypoxemia, a decrease in hematocrit, and diffuse infiltrates on chest radiograph. PATIENT: We report a case of DAH after allogeneic BMT in a 6-yr-old female patient. Although a chest radiograph revealed patchy bilateral alveolar densities and large volumes of bright red blood were suctioned from the endotracheal tube, there was no evidence of coagulopathy and no infectious agent was identified on examination of bronchoalveolar lavage fluid, blood, and urine. INTERVENTION: The child was treated with high-dose corticosteroids and high-frequency oscillatory ventilation and experienced a complete clinical recovery from her pulmonary disease. RESULTS: The definition, presenting symptoms, findings and timing, and associated risk factors of DAH after BMT are reviewed. Prospective hypotheses for the pathogenesis of DAH after BMT are presented. Evidence for the role of high-dose corticosteroids for treatment of DAH after BMT and the role of high-frequency oscillatory ventilation for treatment of acute hypoxemic respiratory failure in children with diffuse alveolar disease is also reviewed. CONCLUSION: This case supports the contention that early treatment with high-dose corticosteroids is warranted in children with DAH after BMT.

Cervical transfacet versus lateral mass screws: a biomechanical comparison.

The authors directly the compared biomechanical pullout strength of screws placed in the cervical lateral masses to that of screws placed across the facet joints. Posterior cervical fixation with lateral mass plates is an accepted adjunctive technique for cervical spine fusions. Altered anatomy resulting from congenital malformation, tumor, trauma, infection, or failed lateral mass fixation may limit traditional screw placement options. Transfacet screw placement, which has been studied extensively in the lumbar spine, may offer an alternative when posterior cervical fusion is required. Ten fresh human cadaveric cervical spines (postmortem age range, 69 to 91 years) were harvested. On one side, transfacet screws were placed at the C3-4, C5-6, and C7-T1 levels. On the other side, lateral mass screws were placed at the C3, C5, and C7 levels. The screw insertion technique at each level was randomized for right or left. After screw placement, each set of vertebral bodies were dissected and mounted in a custom jig for axial pullout testing using a servohydraulic testing machine. The load-displacement curves were obtained for each screw pullout. The mean pullout strength for the screws placed across the facets was 467 N (range, 192 to 1,176 N). This compares with 360 N (range, 194 to 750 N) for the lateral mass screws (p = 0.008). At each level, transfacet screws exhibited greater pullout resistance compared with the lateral mass placement, but the difference was most pronounced at the C7-T1 level (lateral mass = 373 N, transfacet = 539 N, p = 0.042). Cervical transfacet screw placement provides pullout resistance that is comparable to, if not greater than, lateral mass placement. This type of placement, although technically difficult, may be an alternative to lateral mass screws in cases with unusual anatomy, stripped screws, or when additional intermediate points of fixation are desired.

Prerequisites for fetal neurosurgery: management of central nervous system anomalies toward the 21st century.

With the dramatic technical advances in neuroimaging, it has become possible to diagnose central nervous system (CNS) anomalies in the fetus in utero with more precise morphological analysis. A new technique, high-resolution magnetic resonance (MR) imaging using heavily T2-weighted fast spin echo sequences, has been reported to solve motion artifact of the fetus. However, it has also been recognized that the morphological fetal CNS findings detected in early development are not always the final features: occasionally they may not be determined in diagnosis and may change developmentally or chronologically during the fetal life in utero. Certain factors of the fetal chronology of CNS anomalies can cause irreversible changes during fetal life. These include: (1) significant delay in the neuronal maturation process in fetal hydrocephalus developed in clinicoembryological stage II in the Perspective Classification of Congenital Hydrocephalus (PCCH), (2) secondary neural injury in the intactly developing spinal cord above the neural placode in fetus with spina bifida aperta (myeloschisis), (3) histological "evolution" of tumors or dysgenetic CNS, and (4) deformity of the normally developed intracranial or intraspinal CNS structures. Considering the current status of fetal surgery in general and technical advances promising improved outcomes, fetal neurosurgery can also be applied in the above-mentioned progressive pathology or pathophysiology in the fetal CNS. However, since the failure of the first trial of fetal neurosurgery in the 1980s, the prerequisites have still not been clarified. In order to use advanced neurosurgery techniques in the management of fetal CNS anomalies, these prerequisites have to be established.

Risk factors associated with methicillin-resistant staphylococcal wound infection after spinal surgery.

We used the data from a retrospective case controlled study to identify risk factors for methicillin-resistant staphylococcal wound infection after spinal surgery. Thirty-five cases and 35 uninfected control patients were matched for indication for initial surgery and approximate operative date. Preoperative, intraoperative, and postoperative risk factors were examined. At our institution between 1989 and 1995, 35 adult patients developed spinal wound infection requiring operative debridement; 16 infections were caused by methicillin-resistant staphylococci (MRS). Significant risk factors for MRS infection were lymphopenia, history of chronic infections, alcohol abuse, recent hospitalization, and prolonged postoperative wound drainage. Patients with MRS infections were also somewhat less likely to have received vancomycin prophylaxis. In contrast, the only factor associated with infection caused by other pathogens was alcohol abuse. A number of preoperative risk factors were significantly associated with subsequent MRS spinal wound infection. Chemoprophylaxis with vancomycin should be targeted to patients at increased risk, because overuse may promote the emergence of vancomycin-resistant pathogens.

The use of vancomycin and tobramycin in acrylic bone cement: biomechanical effects and elution kinetics for use in joint arthroplasty.

We examined the effects of vancomycin on the compressive strength and fatigue life of bone cement and the pharmacokinetics and antimicrobial activity against methicillin-resistant Staphylococcus aureus of vancomycin eluted from bone cement, both alone and in combination with tobramycin. Two cements, Palacos and Simplex, were tested. Three antibiotic preparations were tested: lyophilized vancomycin (vancomycin-L), vancomycin powder (vancomycin-P), and tobramycin powder (Lilly, Indianapolis, IN). Although antibiotics did not significantly affect compressive strength, the fatigue life of bone cement was significantly decreased with vancomycin. Thus, fatigue testing revealed effects on cement strength not apparent by compression testing. Vancomycin-P had a substantially less detrimental effect on fatigue strength than vancomycin-L. Vancomycin-P elutes less efficiently than tobramycin. Although relatively little vancomycin-P eluted from bone cement, it retained biologic activity.

No influence of large volume blood loss on serum vancomycin concentrations during orthopedic procedures.

We prospectively studied orthopedic patients with either large or small blood loss who also received vancomycin prophylaxis to determine the effect of intraoperative volume shifts on serum vancomycin concentrations. There were 6 index patients in the large blood loss group (greater than 2 L), and 7 in the control group (less than 2 L). Mean estimated blood loss for index and controls was 4.4 L and 1.0 L, respectively. Mean intraoperative fluid resuscitation, excluding blood products, was 12.4 L and 5.1 L, respectively. There was a modest inverse correlation between blood loss and intraoperative serum half-life of vancomycin. Although controls maintained slightly higher intraoperative vancomycin concentrations at each time-point, there was no statistically significant difference between the groups with regard to absolute concentrations or rate of decline. After 8 hours, the serum vancomycin concentration exceeded the MIC-90 for Staphylococcus aureus by approximately eightfold in all but one case patient. This was a morbidly obese patient with massive blood loss. Thus, blood loss during orthopedic procedures has minimal effects on intraoperative kinetics of vancomycin. Redosing is rarely indicated, although a preoperative 1.5 gram-dose should be considered for patients weighing more than 90 kg.

Extradural infections of the spine.

OBJECTIVES: We have observed a recent increase in the incidence of spinal extradural infections. To determine postoperative outcome and prognosis we have undertaken a retrospective study on patients with spinal extradural abscesses between 1978 and 1996 treated in the Department of Neurosurgery at the Nordstadt Hospital in Hannover, Germany. METHODS: Case records, outpatient files, operation notes, neuroradiological examinations and pathological reports were analysed. Neurological function was documented using a score system for each symptom. RESULTS: Twenty-two patients underwent 24 operations during the study period. Staphylococcus aureus was the organism most commonly isolated. Patients presented after a mean history of 1.8+/-2.6 months with acute development of severe para- or tetraparesis and were followed up for an average period of 6+/-7 months. Two groups were distinguished. One group was characterized by epidural collections of pus (14 patients). Two patients in this group were not operated due to their moribund state and died from uncontrollable septicemia. Two of the remaining 12 operated patients died within 30 days after surgery due to generalized septicemia or other medical problems unrelated to the spinal involvement. Of the ten surviving patients, five sustained major neurological deficits, whereas the remaining five patients made an incomplete recovery. The second group consisted of eight patients in a significantly better pre-operative health condition in whom granulomatous material was obtained during the operation and the outcome was considerably better. No patient in this group died. Six patients recovered with no or mild neurological deficits. CONCLUSION: Spinal extradural infections require immediate surgical intervention. Neurological outcome depends on the pre-operative neurological status. Survival is determined by the general health condition of the patient.

Exposure to hyperoxia decreases the expression of vascular endothelial growth factor and its receptors in adult rat lungs.

Exposure to high levels of inspired oxygen leads to respiratory failure and death in many animal models. Endothelial cell death is an early finding, before the onset of respiratory failure. Vascular endothelial growth factor (VEGF) is highly expressed in the lungs of adult animals. In the present study, adult Sprague-Dawley rats were exposed to >95% FiO2 for 24 or 48 hours. Northern blot analysis revealed a marked reduction in VEGF mRNA abundance by 24 hours, which decreased to less than 50% of control by 48 hours. In situ hybridization revealed that VEGF was highly expressed in distal airway epithelial cells in controls but disappeared in the oxygen-exposed animals. Immunohistochemistry and Western blot analyses demonstrated that VEGF protein was decreased at 48 hours. TUNEL staining demonstrated the presence of apoptotic cells coincident with the decline in VEGF. Abundance of VEGF receptor mRNAs (Flt-1 and KDR/Flk) decreased in the late time points of the study (48 hours), possibly secondary to the loss of endothelial cells. We speculate that VEGF functions as a survival factor in the normal adult rat lung, and its loss during hyperoxia contributes to the pathophysiology of oxygen-induced lung damage.

Surgical results for spinal meningiomas.

Among a series of 782 spinal tumors, 130 spinal meningiomas in 117 patients were operated in the Department of Neurosurgery at the Nordstadt Hospital in Hannover, Germany, between 1977 and 1998. Patients were followed postoperatively for up to 13 years (mean 20 +/- 33 months). Comparing the period of 1977 through 1987, before magnetic resonance imaging (MRI) was available, to the period of 1988 to 1998 revealed that the average history until diagnosis shortened by about 6 months during the second decade of this study (24 +/- 33 to 18 +/- 29 months; not significant). Consequently, the preoperative Karnofsky Score increased significantly (59 +/- 15 and 66 +/- 16; p < 0.05). The rates of complete resection and the postoperative neurological outcome, however, remained unchanged. Even though the overall prognosis of neurological deficits is favorable after complete resection of a meningioma, a subset of 18 patients had either en plaque growing or recurrent tumors that were more likely to be removed incompletely and to cause postoperative neurological problems, with a significantly worse Karnofsky Score after 1 year (57 +/- 12 and 77 +/- 12, respectively; p < 0.01) and a significantly higher recurrence rate after 5 years (86.7% and 20.4%, respectively; log rank test p = 0.0014). In conclusion, a favorable postoperative neurological outcome requires complete resection of the spinal meningioma. The advent of MRI has shortened the time until diagnosis and made it possible to perform surgery before severe deficits have occurred, but did not have a major impact on postoperative results. En plaque and recurrent meningiomas remain surgical challenges, as infiltration of surrounding structures and associated arachnoid scarring may render complete resection difficult to achieve.

Surgical results for spinal metastases.

Among a series of 740 spinal tumours treated in the Department of Neurosurgery at the Nordstadt Hospital in Hannover, Germany, between September 1977 and December 1996, 106 spinal metastases in 101 patients were operated on. After an average period of 4.0 +/- 6 months (2 days to 5 years) patients presented at an average age of 62 +/- 12 years. 79% of the tumours were operated on by a posterolateral approach, 12% by an anterior and the remaining 9% by an anterior and posterior approach. A complete resection was achieved for 43.4% of the metastases while 48.1% were removed partially, 7.5% were biopsied and one patient received an opiate pump. Operations were followed by radiotherapy, chemotherapy or hormone treatment. The overall local recurrence rates as determined by the Kaplan Meier method were 57.9% after 6 months, 69.3% after 1 year and 96% after 4 years. Multiple regression analyses revealed that an independent preoperative status of ambulation, favourable tumour histology, cervical level, complete resection, low number of affected vertebral bodies, and elective surgery were significant, independent predictors of a low rate of local metastatic recurrence. Postoperative neurological outcome was related to preoperative neurological deficits. 96% of patients walking preoperatively kept this ability for at least 3 months postoperatively. However, only 22% of patients unable to walk regained walking capacity for 3 months. Correspondingly, 89% of patients remained continent of urine postoperatively for 3 months while only 31% regained sphincter control for this amount of time postoperatively. In terms of postoperative survival, multiple regression analyses showed longer survival times for patients with a favourable tumour histology, independent ambulation, long history, male sex, cervical level, complete resection, posterior approach, no additional metastases in other organs, and no instability. The overall survival rates were 58.8% after 6 months, 48% after 1 year and 19.5% after 5 years postoperatively. In conclusion, surgery has a place in the treatment of patients with metastatic disease of the spine and neurological symptoms and/or spinal instability. The surgical strategy should be tailored according to the general health of the patient and expected time of survival. Primary radiotherapy should be administered to patients without neurological deficits or instability and to patients who cannot undergo or do not accept surgery.

Results of elective lumbar discectomy for patients involved in the workers' compensation system.

We compared the outcomes from lumbar discectomy for patients who were workers' compensation claimants and/or who were involved in active litigation with patients who underwent elective lumbar discectomy, but who were not involved with either compensation or litigation. Eighty-two consecutive patients who underwent elective lumbar discectomy by the senior author were identified from 1989 through 1994. Those patients who underwent a primary discectomy with a minimum of 6 months' follow-up were studied. Patients were excluded if a spinal fusion was performed or if a multilevel laminectomy procedure was required. Patients were classified as compensation patients if they were involved in either worker's compensation claims or active litigation at the time of the lumbar discectomy. The compensation group was further divided into three subsets of patients: those involved in active litigation without compensation, those involved in both compensation and litigation, and those pursuing workers' compensation claims without litigation. The control group was comprised of patients who were not in any way involved with compensation or litigation. Outcome assessment and ratings were determined independently by the coauthors, not the primary surgeon. Outcome was based on pain, employment status, analgesic use, and level of activity. Fifty-four patients met the inclusion criteria. Average follow-up for the compensation patients was 40 weeks. Follow-up for the noncompensation patients averaged 51 weeks. Eighty-one percent of our patients in the noncompensation group achieved a good result. Only 1 of 27 patients was categorized as having a poor outcome. Conversely, patients who were actively involved in the compensation and/or litigation process had significantly poorer outcomes, with only 29% of the patients receiving a good outcome evaluation (p = < 0.0002). Legal involvement was associated with poorer outcome in compensation patients (p = < 0.001).

A critical appraisal of syrinx cavity shunting procedures.

OBJECT: This study was conducted to evaluate the results of shunting procedures for syringomyelia. METHODS: In a follow-up analysis of 42 patients in whom shunts were placed in syringomyelic cavities, the authors have demonstrated that 21 (50%) developed recurrent cyst expansion indicative of shunt failure. Problems were encountered in patients with syringomyelia resulting from hindbrain herniation, spinal trauma, or inflammatory processes. A low-pressure cerebrospinal fluid state occurred in two of 18 patients; infection was also rare (one of 18 patients), but both are potentially devastating complications of shunt procedures. Shunt obstruction, the most common problem, was encountered in 18 patients; spinal cord tethering, seen in three cases, may account for situations in which the patient gradually deteriorated neurologically, despite a functioning shunt. CONCLUSIONS: Placement of all types of shunts (subarachnoid, syringoperitoneal, and syringopleural) may be followed by significant morbidity requiring one or more additional surgical procedures.

Surgery of spinal nerve sheath tumors with special reference to neurofibromatosis.

OBJECTIVE: We conducted a retrospective study of 87 patients with spinal nerve sheath tumors to determine the overall clinical outcome and specific features in 20 patients affected with neurofibromatosis Types 1 and 2 (NF-1 and NF-2, respectively). METHODS: Case records, operation notes, outpatient files, and radiological examinations were analyzed for all patients treated between September 1977 and August 1994. Additional follow-up data were obtained using outpatient examinations, questionnaires, and telephone calls. RESULTS: During the study period, 128 spinal neuromas (i.e., schwannomas) and 6 neurofibromas in 87 patients were treated. Fifty-seven neuromas were associated with NF-2 in 17 patients and six neurofibromas with NF-1 in 3 patients. Patients with NF-2 and symptomatic neuromas presented with more severe neurological deficits compared to patients without NF-2. Eighty-six percent of the neuromas were removed completely. On average, most preoperative deficits or symptoms improved in patients without NF-2, whereas neurological symptoms remained unchanged in patients with NF-2. Multiple regression analysis revealed that partial removal, surgery of a recurrent tumor, NF-2, and old age predisposed for tumor recurrence. No increased risk of recurrence was observed for patients with NF-1. For patients without NF-2, we observed overall recurrence rates of 10.7% after 5 years and 28.2% after 10 and 15 years, respectively, as determined by Kaplan-Meier analysis. For NF-2, the recurrence rate at 5 years was 39.2%, and all tumors had recurred by 9 years. CONCLUSION: Spinal nerve sheath tumors carry an excellent prognosis in patients with NF-1 and in patients without neurofibromatosis. Symptomatic neuromas occurring in association with NF-2 present with more severe neurological deficits, demonstrate little postoperative improvement, and have a very high recurrence rate.

Vascular endothelial growth factor is expressed in ovine pulmonary vascular smooth muscle cells in vitro and regulated by hypoxia and dexamethasone.

Chronic lung disease in neonates results from both lung injury and inadequate repair processes. Little is known about the growth factors involved in lung injury and repair, but vascular endothelial growth factor (VEGF) has recently been reported in several animal models of lung injury. VEGF is an endothelial cell-specific mitogen, which is also known as vascular permeability factor because of its ability to induce vascular leak in some tissues. Chronic lung disease is complicated by increased vascular permeability, which can be improved by avoidance of hypoxia and in some cases by dexamethasone therapy. In many cells, hypoxia stimulates VEGF expression. Also, in some cases, dexamethasone blocks VEGF expression. This study examined the role of hypoxia and dexamethasone in regulating the expression of VEGF in pulmonary artery smooth muscle cells. An ovine VEGF cDNA fragment (453 bp) was cloned and found to be highly homologous to known human sequences for VEGF165. Sheep pulmonary artery smooth muscle cells were cultured and exposed to room air, hypoxia, and dexamethasone, alone or in combination for 6 h. At baseline these cells expressed VEGF mRNA at approximately 3.9 kb. The half-life of VEGF mRNA in the smooth muscle cells was 171 min, more than 3-fold longer than previous reports for epithelial cells. Exposure to hypoxia caused a 3-fold increase in mRNA abundance, primarily through transcriptional up-regulation. Dexamethasone blocked the hypoxia-induced increase in VEGF mRNA. The results demonstrate that hypoxia and dexamethasone are regulators of VEGF expression in ovine pulmonary artery smooth muscle cells. It is not known whether VEGF derived from these cells is involved in lung injury and/or normal homeostatsis.