Glucose transport across the rat blood-brain barrier during anesthesia

The authors studied blood-brain barrier (BBB) glucose transport kinetics in awake rats and in pentobarbital- and halothane-anesthetized rats, using a 3H2O/14C-D-glucose double-indicator method corrected for cerebral blood flow at glucose concentrations from 1 to 80 mM. At normal glucose concentrations (5 mM), total brain glucose influx was unaltered by pentobarbital. In contrast, halothane attenuated glucose transport capacity from 1.9 to 0.4 mumol/g-min-1 and increased diffusional transport, Km (Michaelis constant) was decreased sixfold, from 12 to 2 mM. Halothane appears to inhibit BBB glucose transport by competing for the glucose carrier and by altering the affinity of the carrier for glucose, perhaps by altering the environment of the carrier or the carrier itself. The finding of halothane-induced increased diffusional transport of glucose across the BBB corroborates earlier reports and more recent evidence that halothane increases the permeability of the BBB to diffusional processes.     

Triglycerides induce leptin resistance at the blood-brain barrier

Obesity is associated with leptin resistance as evidenced by hyperleptinemia. Resistance arises from impaired leptin transport across the blood-brain barrier (BBB), defects in leptin receptor signaling, and blockades in downstream neuronal circuitries. The mediator of this resistance is unknown. Here, we show that milk, for which fats are 98% triglycerides, immediately inhibited leptin transport as assessed with in vivo, in vitro, and in situ models of the BBB. Fat-free milk and intralipid, a source of vegetable triglycerides, were without effect. Both starvation and diet-induced obesity elevated triglycerides and decreased the transport of leptin across the BBB, whereas short-term fasting decreased triglycerides and increased transport. Three of four triglycerides tested intravenously inhibited transport of leptin across the BBB, but their free fatty acid constituents were without effect. Treatment with gemfibrozil, a drug that specifically reduces triglyceride levels, reversed both hypertriglyceridemia and impaired leptin transport. We conclude that triglycerides are an important cause of leptin resistance as mediated by impaired transport across the BBB and suggest that triglyceride-mediated leptin resistance may have evolved as an anti-anorectic mechanism during starvation. Decreasing triglycerides may potentiate the anorectic effect of leptin by enhancing leptin transport across the BBB.     

Effect of intracerebral and subdural hematomas on energy-dependent transport across the blood-brain barrier

Although both intracerebral and subdural hematomas induce brain edema, previous studies have indicated that they may have different cerebrovascular effects. Our own investigations have demonstrated that while subdural hematomas (SDH) are associated with ischemia this is not the case following intracerebral hematomas (ICH). Previous studies have demonstrated a decrease in energy-dependent transport of glutamine across the blood-brain barrier (BBB) following focal cerebral ischemia. The present study investigates this further by examining the effects of SDH, ICH, and intracerebral thrombin injections, an agent involved in ICH-induced injury, on blood to brain glutamine transport. The injection of 200 microL of blood into the subdural space induced a marked reduction in glutamine transport (Ki, influx rate constant) into the cerebral cortex at 4 and 24 h following SDH (sham, 105+/-4% of contralateral cortex; SDH 4 h, 63+/-5%, p<0.01; SDH 24 h, 47+/-12%, p<0.05). There were no significant changes in glutamine Ki in subcortical areas following SDH. Following ICH (200-microL clot); however, there were only modest decreases in glutamine Ki in subcortical areas (sham, 98+/-2% of right cortex; ICH 4 h, 91+/-5%, p<0.01; ICH 24 h, 91+/-2%, p<0.05). Intracerebral injection of thrombin (5U) had minimal effect on glutamine Ki, in subcortical areas, at 4 h and induced a modest decrease in transport at 24 h (sham, 98+/-2% of right cortex; thrombin 4 h, 98+/-2%; thrombin 24 h, 86+/-2%, p<0.05). The present studies demonstrate marked differences in the effects of ICH and SDH on BBB function.     

Cellular inflammatory response associated with breakdown of the blood-brain barrier after closed head injury in rats

This study reports a widespread microglial response characterized by an upregulation of surface antigens, such as complement type 3 receptors (CR3) and major histocompatibility complex (MHC) class II antigens on these cells following closed head injury. Increased expression of CR3 (OX-42) and MHC class II antigens (OX-6) was observed in rats killed at 1, 3, and 5 days after injury. Intense OX-42 immunoreactivity was observed in microglial cells throughout the brain with a smaller number of them being OX-6 positive. In addition to microglial reaction, astrocytic activation reflected in cellular hypertrophy and increased immunoreactivity for glial fibrillary acidic protein (GFAP) was observed at 5 days after head injury. Together with the above, a diffuse perivascular and intraneuronal immunostaining for immunoglobulin G (IgG) was observed primarily in the cerebral cortex. This was accompanied by an enhanced expression of both endothelial nitric oxide synthase (eNOS) in blood vessels and inducible nitric oxide synthase (iNOS) in brain macrophages. In rats subjected to closed head injury followed by a single intraperitoneal (i.p.) injection of rhodamine isothiocyanate (RhIc), seepage of the fluorescent dye into the neuropil was observed. This had resulted in the labelling of the cortical neurons clearly demonstrating a breakdown of the blood-brain barrier (BBB). In the latter, it is conceivable that the ensuing leakage of plasma immunoglobulins and other serum-derived materials could induce the expression of MHC class II antigens on microglia. The mechanism causing the BBB dysfunction is not clear, although present results suggest that excessive release of nitric oxide (NO) may be a contributory factor. The widespread activation of microglia in rats after head injury suggests their involvement in increased endocytosis and immunological responses.     

Obesity induced by a high-fat diet is associated with reduced brain insulin transport in dogs

Insulin transported from plasma into the central nervous system (CNS) is hypothesized to contribute to the negative feedback regulation of body adiposity. Because CNS insulin uptake is likely mediated by insulin receptors, physiological interventions that impair insulin action in the periphery might also reduce the efficiency of CNS insulin uptake and predispose to weight gain. We hypothesized that high-fat feeding, which both reduces insulin sensitivity in peripheral tissues and favors weight gain, reduces the efficiency of insulin uptake from plasma into the CNS. To test this hypothesis, we estimated parameters for cerebrospinal fluid (CSF) insulin uptake and clearance during an intravenous insulin infusion using compartmental modeling in 10 dogs before and after 7 weeks of high-fat feeding. These parameters, together with 24-h plasma insulin levels measured during ad libitum feeding, also permitted estimates of relative CNS insulin concentrations. The percent changes of adiposity, body weight, and food intake after high-fat feeding were each inversely associated with the percent changes of the parameter k1k2, which reflects the efficiency of CNS insulin uptake from plasma (r = -0.74, -0.69, -0.63; P = 0.015, 0.03, and 0.05, respectively). These findings were supported by a non-model-based calculation of CNS insulin uptake: the CSF-to-plasma insulin ratio during the insulin infusion. This ratio changed in association with changes of k1k2 (r = 0.84, P = 0.002), body weight (r = -0.66, P = 0.04), and relative adiposity (r = -0.72, P = 0.02). By comparison, changes in insulin sensitivity, according to minimal model analysis, were not associated with changes in k1k2, suggesting that these parameters are not regulated in parallel. During high-fat feeding, there was a 60% reduction of the estimated CNS insulin level (P = 0.04), and this estimate was inversely associated with percent changes in body weight (r = -0.71, P = 0.03). These results demonstrate that increased food intake and weight gain during high-fat feeding are associated with and may be causally related to reduced insulin delivery into the CNS.     

Blood-brain barrier choline transport is reduced in diabetic rats

The kinetics of blood-brain barrier (BBB) choline transport in streptozocin-induced diabetic rats were compared with those of age-matched vehicle-injected control rats. The brain uptake index (BUI) of choline in diabetic rats (13.9 +/- 1.1%) was significantly lower than that in control rats (22.6 +/- 0.7%) (P less than .05). This alteration in brain choline uptake appeared to occur in long-standing (9 wk) diabetes. Thus, acute hyperglycemia and diabetes mellitus for shorter periods (3 wk) did not significantly alter the BUI of choline. Insulin (8 U/kg) treatment for 5 days did not alter BUI in diabetic rats (12.9 +/- 0.9%). The maximal velocity of BBB choline transport (Vmax) in diabetic rats (0.14 +/- 0.07 nmol . min-1 . g-1) was significantly lower than the Vmax in control rats (2.2 +/- 0.8 nmol . min-1 . g-1) (P less than .05). The Km of choline transport in diabetic rats (120 +/- 70 microM) was modestly but not significantly lower than that in control animals (400 +/- 160 microM). Similarly, the constant of the nonsaturable component of the transport (Kd) in diabetic animals (0.5 +/- 0.07 microliter . min-1 . g-1) was not significantly different from that in control rats (0.9 +/- 0.3 microliter . min-1 . g-1). The data indicate that diabetes mellitus in rats is associated with a decreased BBB choline transport.     

Sequelae of the osmotic blood-brain barrier opening in rats

Histopathological sequelae of the osmotic blood-brain barrier opening were studied in 69 adult Wistar rats sacrificed between 2 minutes and 6 days after infusion of 1.6 M mannitol into the unilateral internal carotid artery. The results were correlated with immunohistochemical localization of autologous albumin in the brain parenchyma on paraffin sections. Extravasation of serum albumin was evident in all rats, and the albumin immunoreactivity, commonly localized to the territories of the ipsilateral anterior, middle, and posterior cerebral arteries and contralateral anterior cerebral artery, showed maximum intensity in the rats sacrificed 30 minutes after infusion. The albumin immunoreactivity remained macroscopically visible in the brain parenchyma for 24 to 48 hours, and then gradually faded out. Serum extravasation was accompanied by widening of the perivascular space and focal edema, which largely subsided within 48 hours as the albumin immunoreactivity of the tissue diminished. Although no overt neurological sequelae were seen in the present experiment, minute but definite foci of infarction with focal accumulation of albumin were found in 23 (38%) of 61 rats surviving more than 30 minutes. In addition, ischemic neuronal death of delayed onset was encountered among neurons in the CA-1 region of the hippocampus, in the cerebellum, and in the thalamus in five (25%) of 20 rats sacrificed between Days 4 and 6. Thus, care should be exercised in the practice of this procedure.     

Obesity was associated with a decreased postoperative recurrence of rectal cancer in a Japanese population

Purpose

Obesity contributes to the technical difficulty of rectal surgery and is considered to be a risk factor for postoperative complications. The impact of obesity on the long-term outcomes of rectal cancer surgery remains unclear.

Methods

A total of 263 consecutive rectal cancer patients who underwent surgery were categorized into two groups according to the body mass index (BMI) based on the Asian BMI classification: non-obese (BMI <25 kg/m²) and obese (BMI ≥25 kg/m²). The postoperative survival and recurrence rates and oncological surgical quality indicators were compared between groups using the univariate and multivariate analyses. The differences in recurrence patterns were assessed by a competing risk regression analysis.

Results

64 (24 %) patients were included in the obese group. The number of retrieved lymph nodes was significantly greater in the non-obese group than in the obese group (22.4 vs. 16.0, P < 0.01). The 5-year disease-free survival (DFS) rates were 86.5 and 68.8 % in the obese and non-obese groups, respectively (P = 0.01). The multivariate analysis demonstrated that obesity significantly decreased the postoperative recurrence rate (P = 0.04). Moreover, the BMI was significantly associated with distant metastasis (P = 0.04).

Conclusions

Obese rectal cancer patients have high DFS rates and a decreased incidence of distant metastases compared to non-obese patients. The BMI may be a key factor for predicting the postoperative prognosis and determination of an appropriate strategy for the treatment of rectal cancer patients.     

Effects of fentanyl on alpha-aminoisobutyric acid transfer across the blood-brain barrier.

The effect of fentanyl on the transfer of small hydrophilic molecules across the blood-brain barrier was studied in rats by measuring the blood-brain transfer coefficient (Ki) and the regional cerebral blood flow (rCBF) and by calculating the capillary permeability-surface area product. In the control group (n = 14), after a femoral artery and vein were catheterized under isoflurane anesthesia, the rats were allowed to remain awake for 1.5 h before measuring Ki (n = 8) using intravenous 14C-alpha-aminoisobutyric acid or rCBF (n = 6) using 14C-iodoantipyrine. In the fentanyl groups, rats were injected with 25 micrograms/kg (low (n = 6) or 100 micrograms/kg (high dose) of fentanyl (n = 14), followed by a continuous infusion at a rate of 50 or 200 micrograms.kg-1.h-1, respectively. Their lungs were mechanically ventilated. The Ki (low dose, n = 6; high dose, n = 8) and rCBF (high dose, n = 6) were measured 1 h after fentanyl infusion. The Ki was lower in 9 of 13 brain regions in the low-dose fentanyl group and in 7 of 13 brain regions in the high-dose fentanyl group than in the control animals. The average value of Ki in all the brain regions was 8.6 +/- 4.6 microL.g-1.min-1 in the control group, 5.2 +/- 2.9 microL.g-1.min-1 in the low-dose fentanyl group, and 5.7 +/- 2.9 microL.g-1.min-1 in the high-dose fentanyl group. High-dose fentanyl did not significantly affect rCBF in any brain region studied. The value of the regional permeability-surface area product was similar to the corresponding regional Ki in the groups studied. In conclusion, fentanyl decreased the transfer of small hydrophilic molecules across the blood-brain barrier, as demonstrated by a decreased Ki and permeability-surface area product, without significant changes in rCBF.     

Optison (FS069) disrupts the blood-brain barrier in rats

Optison is a new echocardiographic contrast agent, designed for IV injection, that is very useful in delineating cardiac structures during ultrasound examination. Because Optison could be a valuable adjunct in the diagnosis and evaluation of congenital heart disease, this study was undertaken to assess its effects on the blood-brain barrier when introduced directly in the cerebral circulation, as might occur with some congenital lesions. In this study, Sprague-Dawley rats were anesthetized, and Optison, at various dosages, was injected into the carotid artery. After this, Evans blue dye, a marker for blood-brain barrier disruption, was injected at different time intervals. Gross and histologic examination of the animals' brains revealed disruption of the blood-brain barrier that appeared to be Optison-dosage-dependent. Although the mechanism for this disruption is unclear, it may be related to the use of octofluoropropane gas used in the Optison as a contrast medium. Further studies are necessary to determine the pathologic consequences of Optison's effects on the blood-brain barrier.     

Early blood-brain barrier disruption after high-dose single-fraction irradiation in rats

Summary We studied the effect of high-dose single-fraction irradiation on the permeability of the blood-brain barrier (BBB) in rat brains. Immunohistochemistry with an antibody to serum albumin was used as a sensitive method for detecting the extravasation of endogenous serum components. Extravasation of albumin was detected as early as 1 day after irradiation with 20 or 40 Gy. Immunoreactivity reached its maximum after 3 days, gradually decreased during the following few weeks and had disappeared by day 30. Extravasation was much greater after irradiation with 80 Gy and continued to increase during the whole period of the experiment (6 days). Disruption of BBB this early after irradiation has not been previously documented. The time course of observed serum albumin extravasation, however, agrees well with the previous ultrastructural evidence for increased BBB permeability after irradiation with 27 Gy in monkey brains⁴. This transient impairment of BBB may contribute to the reversible neurological symptoms after radiosurgery. It may also allow drugs that normally not pass the BBB to do so and thus disperse in the brain when administered at this time.     

Cyclosporine-A induced nephrotoxicity is associated with decreased renal bone morphogenetic protein-7 expression in rats

The aim of our study was to investigate bone morphogenetic protein-7 (BMP-7) expression in a rat model of chronic cyclosporine (CsA) toxicity compare with healthy controls, as well as the influence of treatment with the angiotensin-converting enzyme inhibitor (ACEI) quinapril. Twenty-four male Wistar rats were divided into groups of eight animals treated with CsA (15 mg/kg intraperitoneally) for 8 weeks (CsA group) without or with quinapril (10 mg/kg per day in the drinking water: CsA group + Q) for comparison with healthy controls (H group). The renal tissues were examined by light microscopy for CsA toxicity; specifically, tubulointerstitial damage and afferent arteriolopathy as well as BMP-7 expression were semiquantitatively scored by immunohistochemical staining. Mean CsA levels were 1982 ng/mL and 1968 ng/mL for the CsA and CsA + Q groups, respectively. At the end of the study period, the mean serum creatinine levels were 0.8 +/- 0.2 mg/dL, 1.6 +/- 0.8 mg/dL, and 1.4 +/- 0.8 mg/dL for the H, CsA, and CsA + Q groups, respectively. Interstitial fibrosis, tubular atrophy, and afferent arteriolar hyalinization were present in the CsA group and, to a lesser degree, in the CsA + Q group, compared with the H group. CsA-treated rats displayed significantly decreased BMP-7 expression compared with healthy controls (P <.0005). BMP-7 expression was higher among the CsA + Q group than the the group CsA group. In a rat model histologic changes characteristic of CsA-induced nephrotoxicity are associated with decreased expression of BMP-7, which seems to be at least partially restored by ACE inhibition.     

Elevated serum ubiquitin carboxy-terminal hydrolase L1 is associated with abnormal blood-brain barrier function after traumatic brain injury

Serum S100B elevations accurately reflect blood-brain barrier (BBB) damage. Because S100B is also present in peripheral tissues, release of this protein may not be specific to central nervous system (CNS) injury. Ubiquitin C-terminal hydrolase 1 (UCHL1), and phosphorylated neurofilament heavy chain (pNF-H) are found exclusively in neurons, but their relationship to BBB dysfunction has not been determined. The objective of this study was to determine the accuracy of serum UCHL1 and pNF-H as measures of BBB integrity after traumatic brain injury (TBI), to and compare them to S100B. We performed a prospective study of 16 patients with moderate to severe TBI (Glasgow Coma Scale [GCS] score ≤12) and 6 patients with non-traumatic headache who had cerebrospinal fluid (CSF) collected by ventriculostomy or lumbar puncture (LP). Serum and CSF were collected at the time of LP for headache patients and at 12, 24, and 48?h after injury for TBI patients. BBB function was determined by calculating albumin quotients (Q(A)), where Q(A)=[albumin(CSF)]/[albumin(serum)]. S100B, UCHL1, and pNF-H were measured by enzyme-linked immunosorbent assay (ELISA). Pearson's correlation coefficient and area under the receiver operator characteristic (ROC) curve were used to determine relationships between serum markers and Q(A). At 12 hours after TBI, a significant relationship was found between Q(A) and serum UCHL1 concentrations (AUC=0.76; 95% CI 0.55,1.00), and between Q(A) and serum S100B concentrations (AUC=0.794; 95% CI 0.57,1.02). There was no significant relationship found between these markers and Q(A) at other time points, or between pNF-H and Q(A) at any time point. We conclude that serum concentrations of UCHL1 are associated with abnormal BBB status 12?h after moderate to severe TBI. This relationship is similar to that observed between serum S100B and Q(A,) despite the fact that S100B may be released from peripheral tissues after multi-trauma. We conclude that peripheral release of S100B after multi-trauma is probably negligible and that UCHL1 may have some utility to monitor BBB disruption following TBI.     

Effect of hyperosmotic blood-brain barrier disruption on transcapillary transport in canine brain tumors

Whether hyperosmotic blood-brain barrier (BBB) disruption is a technique that can be used to increase permeability of brain-tumor capillaries and thereby transiently increase drug delivery to the brain tumor is controversial. Nine virally induced brain tumors were studied in seven dogs, before and after hyperosmotic BBB disruption with 1.4 osmolar mannitol. Each dog was studied with computerized tomography (CT) after administration of the water-soluble tracer meglumine iothalamate. Each study lasted 30 minutes. A baseline CT scan and 35 to 40 additional CT scans were obtained to provide a time-related measurement of the amount of meglumine iothalamate in tissue (Am(t], and 30 plasma samples were collected to provide the time-related measurement of meglumine iothalamate in plasma (Cp(t]. The data were analyzed by three different methods: 1) a two-compartment model and nonlinear curve fitting were used to calculate K1 (blood-to-tissue or influx constant), k2 (tissue-to-blood or efflux constant), and Vp (plasma vascular space); 2) K1 values were calculated with a two-compartment model, assuming no efflux, at the time point for each CT scan; and 3) a "tissue advantage ratio" was calculated that expressed the ratio of tissue uptake of meglumine iothalamate at each time point, comparing values before and after BBB disruption. Regardless of which method of data analysis was used, there was a marked and significant increase in transcapillary transport of meglumine iothalamate to tumor-free brain regions, while there was only a small, transient, and insignificant increase to the brain tumors. Although there were often marked increases in delivery to cortex in the same hemisphere as the tumors, there was no significant increase to brain immediately surrounding the tumors, perhaps due to altered circulatory dynamics in this region. These data raise serious questions as to the wisdom of using this technique to increase drug delivery to brain tumors in patients and strongly support the continued study of this technique in experimental brain tumors before it is used in patients.     

Ultrastructural changes in the blood-brain barrier in rats after treatment with Nimodipine and Flunarizine. A comparison

The idea of using induced hypertension to treat the symptomatic ischaemia resulting from vasospasm after subarachnoidal hemorrhage, and the effect of this therapy on the blood-brain barrier, is checked in animal experiments. This therapy is combined with the application of nimodipine, which is recognised as the standard medication for prophylaxis of vasospasm. The effects of the induced hypertension combination with Nimodipine and in combination with another calcium antagonist, Flunarizine are compared.Seventy-four narcotised rats, one group with 22 animals treated with Nimodipine and 22 with placebo, and a second group 20 animals treated with Flunarizine and 10 with placebo, are evaluated. The blood pressure is raised to 150–180 mmHg by i.v. application of norfenephrine and measured continuously. The standard tracer, horseradish peroxidase, is applied as indicator for the blood-brain barrier function. 15 minutes later the experimental animals are exsanguinated by perfusion with saline, then perfused with Karnovsky's solution. After removal, the brains are stained for peroxidase to visualise extravasation of the horseradish peroxidase, and after evaluation of the results each brain is assigned to its experimental group.In the Nimodipine group, a significant accumulation (p<0.001) of perivascular deposits of peroxidase reaction product were found, these were not found in the placebo group. The Flunarizine group does not differ from its placebo group in the number of extravasates, and thus, with respect to protein extravasation, appears better than the Nimodipine group.In electron micrographs of the extravasates one sees intact tight junctions and a neuroendothelial transport, and also vesicles, filled with horseradish peroxidase in the endothelium, the muscle cells, and the brain parenchyma, which arise from pinocytosis. The vesicles, which transport the high-molecular-weight protein, horseradish peroxidase, also transport other proteins and can, therefore, cause a brain edema. It follows from these morphological results that Nimodipine can disrupt the blood brain barrier function and can, therefore, also interfere with cerebral autoregulation, which depends on the resistance of vessels.     

Laparoscopic colectomy is associated with decreased postoperative gastrointestinal dysfunction

Background  Major abdominal surgery is associated with early postoperative gastrointestinal dysfunction, which may lead to abdominal distention and vomiting, requiring nasogastric (NGT) tube insertion. This study aimed to compare the rates of early postoperative NGT insertion after open and laparoscopic colorectal surgery. Methods  A retrospective chart review was performed for patients who underwent colorectal surgery with removal of the NGT at completion of surgery. Patients who required reinsertion of the NGT in the early postoperative course were identified. The reinsertion rate for patients who underwent laparoscopic surgery was compared with that for the open group. Results  There were 103 patients in the open group and 227 in the laparoscopic group. In the laparoscopic group, 42 patients underwent conversion to open surgery. Reinsertion of the NGT was required for 18.4% of the patients in the open group, compared with 8.6% of the patients for whom the procedure was completed laparoscopically (p = 0.02). Conversion to open surgery resulted in a reinsertion rate of 17%. Conclusion  Laparoscopic colorectal surgery is associated with decreased postoperative gastrointestinal dysfunction, resulting in a significantly lower NGT reinsertion rate. Presented as a poster at the annual meeting of the Society of American Gastrointestinal and Endoscopic Surgeons, (SAGES), Dallas, Texas, 26–29 April 2006, and at the annual meeting of the European Society for Endoscopic Surgeons (EAES), Berlin, Germany, 13–16 September 2006.     

Hyperosmotic blood-brain barrier disruption in brains of rats with an intracerebrally transplanted RG-C6 tumor

To test the results of blood-brain barrier (BBB) disruption in the treatment of brain tumor, RG-C6 glioma was transplanted into the brains of rats. Intracarotid infusions of normal saline and hyperosmotic mannitol were then made, followed by intravenous injection of Evans blue dye plus albumin (EB, MW 68,000), horseradish peroxidase (HRP, MW 40,000), and 5-fluorouracil (5-FU, MW 130). Uptake of the drug and the consistency of drug levels in the normal brain and tumor varied widely among these three agents. Both EB and HRP penetrated the brain tumors but did not stain the normal brain tissues. After BBB opening, penetration of EB and HRP into the normal brain was drastically increased; however, the uptake of EB and HRP in the tumor was not increased. The concentration of 5-FU in the tumor was higher than that in the serum and, although it increased 1.5-fold after BBB opening, the increase was not statistically significant. Conversely, there was a progressive increase in concentrations of 5-FU in the tumor-free brain regions (p less than 0.05). These observations suggest that an intracarotid infusion of hyperosmotic mannitol may increase neurotoxicity because it allows greater delivery of anticancer drugs into the normal brain tissue than into the tumor tissues.