The effect of graded steatosis on flow in the hepatic parenchymal microcirculation.

BACKGROUND: Steatosis is a major cause of microcirculatory impairment and graft dysfunction after liver transplantation. The mechanism of this circulatory compromise is unclear. The aim of this study was to evaluate in vivo the effect of steatosis on parenchymal microcirculation and on total hepatic blood flow in an animal model. METHODS: Four groups of New Zealand White rabbits (n=24) were investigated. Group 1 were fed on normal diet (controls). In groups 2, 3, and 4 graded steatosis was induced by feeding on a high cholesterol diet (1.5%) for 4, 8, and 12 weeks, respectively. After laparotomy and exposure of the liver, total hepatic blood flow (THBF) and the hepatic parenchymal microcirculation (HPM) were measured. These parameters were correlated with the degree of histological fat infiltration classified as mild (<30%), moderate (30-60%), or severe (>60%). RESULTS: The 4-, 8-, and 12-week cholesterol diets resulted in mild, moderate, and severe steatosis, respectively. There was an inverse correlation between the degree of fat infiltration and both HPM (Spearman r=-0.967, P<0.0001) and THBF (r=-0.893, P<0.0001). THBF was 137+/-6 ml/min in controls, which reduced to 121+/-3, 99+/-5, and 63+/-5 ml/min in steatotic livers of groups 2, 3, and 4, respectively. HPM was 226+/-5 flux units in the controls and 197+/-7, 119+/-8, and 37+/-9 flux units in steatotic livers of groups 2, 3, and 4 respectively. Comparing with controls using analysis of covariance, the fall in HPM and THBF was found to be significant (P<0.002) in the moderate and severe groups, but not significant (P>0.050) in the mild group. Parenchymal perfusion was reduced to a greater extent than total liver blood flow in moderate and severe grades of steatosis. CONCLUSIONS: Fatty infiltration reduces hepatic blood flow and parenchymal microcirculation. The latter is more markedly reduced with severe steatosis. This may explain the development of microcirculatory impairment and graft failure after transplantation of fatty livers despite adequate liver blood flow.

Reversal and protection against indomethacin-induced blood stasis and mucosal damage in the rat jejunum by a beta3-adrenoceptor agonist.

BACKGROUND: In the rat, indomethacin causes jejunal villous shortening, microvascular distortion, and blood stasis prior to ulceration. The beta3-adrenoceptor agonist CL316,243 (CL) prevents both the early histological changes and ulceration. AIM: To test the hypothesis that the beta3-adrenoceptor agonist CL316,243 exerts its protective effect by prevention and/or reversal of blood flow changes in the rat jejunum exposed to indomethacin. METHODS: In anaesthetized rats, jejunal villous blood flow was measured in surface capillaries using fluorescence microscopy. Stasis of superficial capillary blood flow was induced by combined topical and i.v. indomethacin (100 microg/mL, 2.8 x 10(-4) M). To examine the effect of CL on blood stasis, CL was applied either i.v. (1 mg/kg) or luminally (100 microg/mL, 2.5 x 10(-5)M) at the onset of stasis. Prophylactic protection was assessed by giving i.v. CL simultaneously with indomethacin. Results were compared with controls which received luminal saline applied at blood stasis. The effect of i.v. CL (1 mg/kg) alone, or luminal CL (100 microg/mL) alone on basal villous blood flow was also examined. The small intestines were perfusion-fixed with 10% formol saline, and removed for histology, n = 5 for all groups. RESULTS: Luminal CL given at stasis reversed indomethacin-induced stasis within 10 min, whereas i.v. CL did not. Pretreatment with i.v. CL prevented the onset of stasis. Basal blood flow was raised slightly only by luminal CL. CONCLUSION: The beta-adrenoceptor agonist CL316,243 can protect against indomethacin-induced blood stasis in rat jejunal villi.

Endothelial changes precede mucosal ulceration induced by indomethacin: an experimental study in the rat.

BACKGROUND: Indomethacin has been shown to damage the villous microvasculature concomitant with alterations in villous blood flow in the rat. AIM: To test the hypothesis that alterations in blood flow result from ultrastructural damage to microvasculature endothelium. METHODS: In anaesthetized rats, jejunal villi were exteriorized in a chamber and blood flow in surface capillaries visualized by fluorescence microscopy. Villi were exposed both luminally and systemically to indomethacin (100 microg/mL) for 10 min or until blood slowing or stasis had occurred in superficial capillaries (n=3 per group). Control animals received both a luminal and intravenous vehicle for 45 min (n=3). The small intestines were vascular perfusion-fixed with 1.5% glutaraldehyde and studied by transmission electron microscopy. RESULTS: All controls appeared to be ultrastructurally normal. A 10 min exposure to indomethacin had no effect upon the epithelium but resulted in mild endothelial vacuolization and the development of small finger-like projections into the lumen of villus surface microvasculature. At the point of blood slowing, villus tip epithelium was again normal but the endothelial vacuolization and finger-like projections became more obvious. The endothelial projections and vacuolization became severe at the point of blood stasis; this also coincided with epithelial degeneration. CONCLUSION: This study shows that villus surface microvasculature is the earliest site of morphological damage after indomethacin exposure.

Early histological features of small intestinal injury induced by indomethacin.

The early histological features of indomethacin-induced jejunal injury in the rat are described in tissues preserved by perfusion-fixation with 10% formol-saline. After an oral dose of indomethacin (15 mg/kg, known to cause severe multifocal ulceration of the rat jejunum), groups of rats were anaesthetized with subsequent perfusion-fixation of the gastrointestinal tract at 1, 2, 3, 6 and 48 h after dosing. Using routine light microscopic techniques, we have observed a sequence of four distinct stages, in time, of small intestinal injury. The earliest histological features were shortening of the villi, epithelial stratification, basal lamina degeneration, eosinophil degranulation and infiltration of the epithelium prior to infiltration of the mucosa by neutrophils. We consider that these earliest changes, seen at 1, 2 and 3 h, represent a distinct histological entity termed Type 1 change or villous 'tufting'. Type 2 change includes all of the features of Type 1 change plus the subsequent infiltration of the mucosa by neutrophils at 2, 3 and 6 h. Type 3 change includes necrosis of the upper-third of the villi and was mainly seen at 3 and 6 h. Type 4 change describes extreme injury to more than one-third of the mucosa with severe, acute inflammation and perforation of the bowel wall by 48 h. Although a small number of neutrophils had appeared to infiltrate the mucosa as early as 2 h after dosing, they were only significantly increased at 3, 6 and 48 h. Possible pathogenic mechanisms involved in shortening of villi as a result of smooth muscle contraction and the role of mucosal eosinophils in NSAID-induced jejunal injury in the rat are discussed.

Evidence for an infarctive pathogenesis of acute and chronic gastroduodenal ulceration.

It is clear that all mucosal defensive mechanisms acting against aggressive ulcerogenic factors depend on adequate blood flow. When defence is active, ulcers tend to heal and do so faster when luminal aggression is prevented by reduction of acidity or eradication of H. pylori. Such successful treatment is so profitable that pharmaceutical companies invest vast fortunes on research into every aspect of therapy. This may explain why research on basic aetiology has been slower. Nevertheless there have been recent advances which increasingly point towards an ischaemic pathogenesis of both acute and chronic ulcers. We have been studying those ischaemic mechanisms that may be triggered by alteration of normal physiological processes, and we now have a body of evidence supporting an infarction-like mechanism induced by abnormal motility which might explain the initiation of both acute and chronic human ulceration. In this article we review the evidence for this and show that such a pathogenesis is compatible with the features and current concepts of gastro-duodenal ulceration. Perhaps the most striking feature of chronic ulcers is their singularity, and localisation to the lesser curvature and first part of the duodenum. Within the lesser curvature there is an increasing incidence from the oesophageal end towards pylorus, with maximal incidence in the incisural area (1). Duodenal ulcers occur on the anterior or posterior walls of the first 4 cm. uncommonly on the superior "cap" and rarely on the inferior wall. Such localisation points to a primary cause which, by analogy with other localised necroses eg coronary or stroke, is usually an infarction of an end-artery system.

Non-invasive measurement of hepatic oxygenation by an oxygen electrode in human orthotopic liver transplantation.

Precise evaluation of graft reperfusion is difficult in clinical liver transplantation. The oxygen electrode (OE) is a novel technique to detect blood flow indirectly by measuring the quantity of oxygen which can diffuse from the hepatic tissue to the surface electrode. Application of the surface OE does not influence the liver blood flow or parenchymal perfusion. Adequate graft oxygenation is essential to the outcome of organ transplantation and has not previously been analysed intra-operatively in liver transplant recipients. The OE was applied to the surface of the graft intra-operatively in 22 human liver grafts after restoring portal vein and hepatic artery inflow. OE readings were compared with liver blood flow using an electromagnetic flowmeter (EMF). Intra-operative haemodynamics and donor organ parameters known to influence graft function were correlated with the OE readings. There was a significant correlation (r=0.89; p<0.001, n=14) between tissue oxygenation using the OE and total liver blood flow measured by EMF. The tissue oxygenation measurements were reproducible with a coefficient of variation of 5%. The hepatic tissue oxygenation increased significantly from baseline following venous reperfusion of the graft (282+/-23 vs 3107+/-288 (+/-SE) nA, p<0.001). Hepatic arterial revascularisation resulted in a significant (p<0.001) increase of 41+/-7% in liver oxygen perfusion. There was significant negative correlation (r=0.80, p<0.001, n=22) between cold ischaemic time and graft tissue oxygenation. The OE provides a reliable, cheap and non-invasive method of monitoring liver graft oxygenation and perfusion during transplantation.

Role of ischaemia in the initiation of peptic ulcer.

Study of intramural vascular pathways to the human gastroduodenal mucosa shows that some persons have patches of mucosa supplied by end arteries. These patches occur only in the ulcer-bearing areas. It is suggested how this may contribute to the initiation of peptic ulceration and explain some of its features.

A new method for the assessment of gut viability.

A surface polarographic oxygen electrode was modified so that on the basis of physical considerations its reading became dependent on blood flow. The preliminary results show that when applied to the serosal surface of the bowel under operative conditions, readings correlated well with total blood flow to the segment measured by electromagnetic flowmetry. Graded local ischaemic areas could also be detected. Readings from the serosal surface appeared to be independent of those from the mucosa. The instrument may therefore provide a useful surgical tool for detecting partially ischaemic areas of bowel.

First experimental results with the oxygen electrode as a local blood flow sensor in canine colon.

A modified oxygen electrode was tested against the electromagnetic flowmeter to see if its reading indicated blood flow to the canine colon. Although gross inflow was compared with local flow, the correlation was good in each of seven individuals (r ranged from 0.86 to 0.97). The closeness of the relative values gives the instrument potential for cheap, fast, repeatable and non-invasive estimation of regional blood flow, in colon and other exposed tissues.

Observations on the submucous plexus and mucosal arteries of the dog's stomach and first part of the duodenum.

Arteriolar patterns of the submucous plexus were studied in all areas of the dog's stomach and in the first inch of the duodenum. There appeared to be no poverty of plexus, although in some cases the vessels were somewhat smaller in the pyloric part of the lesser curvature than elsewhere. Mucosal arteries arose from the plexus, and none appeared to have an extramural origin. In man, on the other hand, there is a poverty of the submucous plexus in the 'ulcer region', i.e. in the incisural region of the lesser curvature and in the first inch of the duodenum, associated in some cases with mucosal end arteries of extramural origin. The absence of these features in the dog, which does not suffer from spontaneous chronic ulceration, lends further support to the view that they play a role in the aetiology of the disease in man.

Patterns of blood supply to the gastric mucosa. A comparative study revealing an end-artery model.

The form of the gastric arterial supply to the mucosa has been studied in dog, swine, ferret, cat, guinea-pig, rabbit and rhesus monkey. In all these species, the bore of vessels in the submucous plexus diminished from body to pylorus, though this was most marked in the guinea-pig and rabbit. The plexus was also continuous across the pylorus with duodenal vessels. Thus the well known poverty of vascularity in distal parts of the human stomach is shared by other species and is unlikely to be a contributory factor to the initiation of peptic ulcer, a disease limited to man. In dog, swine, ferret and cat, as in man, the primary (largest) and secondary (smaller) components of the plexus lay entirely in the submucosa. In the cat, there was a secondary plexus of much smaller vessels deep to the muscularis mucosae. In the guinea-pig, rat, rabbit and monkey, both plexuses were mostly embedded within the muscularis mucosae. As a result, mucosal arteries had two modes of origin: (a) the first, in which they did not pass through the muscularis mucosae as exemplified in the cat, and (b) the second, where they did pass through muscularis mucosae as exemplified by the dog, ferret and swine; in other species, they passed through part of the muscularis mucosae. Areas of mucosa supplied by a single mucosal artery were measured, and ranged widely from the smallest in the cat to the largest in the dog. These features do not seem to have been reported previously, and may be associated with as yet undiscovered functional mechanisms of the muscularis mucosae. Mucosal arteries of extramural origin were found to occur occasionally in the guinea-pig and rabbit, and hence these may provide an experimental model of the pattern existing in man.

Continuous intraoperative monitoring of hepatic blood perfusion using a noninvasive surface electrode.

Continuous noninvasive measurement of local blood flow at one or more chosen sites will be useful during experiments on the liver, during liver surgery, or after hepatic transplantation. We have compared a Clark-type flow-dependent oxygen electrode having a 3-mm-diameter cathode applied to the surface of rabbit liver to an electromagnetic flowmeter (EMF) on the portal vein. Reduction in portal flow (ranging from 4 to 100% and maintained over 2 min), correlated with reduction in electrode output (r = 0.944, P less than 0.001). Electrode output was independent of systemic arterial PO2 (ranging from 85 to 340 mm Hg) (P greater than 0.99) and thus of oxygen in inspired gases. These results indicate that this electrode gives a continuous indication of portal venous inflow when hepatic central inflow is undisturbed and may thus prove to be a useful tool in the clinical assessment of liver perfusion.

Spasm of gastric muscularis mucosae might play a key role in causing focal mucosal ischemia and ulceration. An experimental study in guinea pigs.

While transilluminating guinea pig gastric wall, we noticed foci of irregular compression of submucosal veins and arteries, featuring parallel bands across arteries, with complete cessation of blood flow, mucosal pallor, and shrinkage of the area. Other features demonstrated that this was not due to artifactual pressure. We attribute it to compression of vessels by spasm of the muscularis mucosae, because these vessels are embedded in the muscle. Arteries and veins returned to normal with normal mucosal perfusion, either spontaneously within 5 min, or within 5 sec following mucosal stimulation. However, in 20%, "ischemia" remained for 2-3 hr, resulting in full-thickness necrosis of the underlying mucosa but not in surrounding areas. Since in man mucosal arteries pass through the muscularis mucosae before reaching mucosa, and in view of evidence that they may be end arteries, we suggest that stress-induced spasm of muscularis mucosae may play an important role in human acute and chronic ulceration.

Ulcers produced by ligation of individual gastric mucosal arteries in the guinea pig.

It is not known whether or not individual gastric mucosal arteries are end-arteries. We therefore occluded single mucosal guinea pig arteries, resulting in necrosis of the whole thickness of mucosa in the area supplied by the artery in 25 of 33 cases (76%). However, there were no cases of necrosis in 33 control sham ligations performed in the same animals (p less than 0.001). In another group, ligation of two or three adjacent mucosal arteries resulted in necrosis in 10 of 11 cases (p less than 0.0001), against 0 necrosis in 11 sham ligations. Killing the animals at varying intervals showed that after 1 day the mucosa was necrotic, after 2 days it was shed, and later it might perforate. We conclude that many mucosal arteries in the guinea pig may behave as functional end-arteries. How these findings may pertain to ulceration is discussed in relation to (a) the existence of atypical mucosal arteries found in the ulcer-bearing areas in humans and (b) the role of the muscularis mucosae on arterioles perforating it in both humans and other species.

Vascularity at sites of gastro-jejunal anastomoses.

Recent interest in vascularity in relation to peptic ulceration poses the question of relationship between three entities: (1) site of gastro-jejunal (post-anastomotic) ulcer, (2) site of the anastomotic line and (3) site of junction and possible vascular anastomosis between gastric and jejunal circulations in the region of the anastomosis. Vascular injection studies in dogs show that the main gastric and jejunal circulations run up to, but do not communicate across, the surgical anastomosis, there being a line of least vascularity at the anastomosis. Since anastomotic ulcers occur distal to the anastomotic line, this supports the view that such ulcers are situated wholly in an area of jejunal circulation.

Cyclosporin A and doxorubicin-ifosfamide in resistant solid tumours: a phase I and an immunological study.

In order to test whether circumvention of clinical resistance can be obtained in common solid tumours by targeting different drug resistance mechanisms, a phase I clinical and immunological study was designed. The purpose of the study was to determine the dose of cyclosporin A (CsA), in combination with doxorubicin (DOX) and ifosfamide (IFX), needed to achieve steady-state whole-blood levels of 2000 ng ml-1 and the associated toxicity of this combination. Treatment consisted of CsA 5 mg kg-1 as a 2 h loading infusion, followed by a CsA 3 day continuous infusion (c.i.) (days 1-3) at doses that were escalated from 10 to 18 mg kg-1 day-1. Chemotherapy consisted of DOX 55 mg m-2 by i.v. 24 h c.i. (day 2) and IFX 2 g m-2 i.v. over 1 h on days 1 and 3. Treatments were repeated every 4 weeks. Eighteen patients with previously treated resistant solid tumours received 39 cycles. Mean steady-state CsA levels > or = 2000 ng ml-1 were reached at 5 mg kg-1 loading dose followed by a 3 day c.i. of 16 mg kg-1 day-1 or greater. Haematological toxicity was greater than expected for the same chemotherapy alone. One patient died of intracranial haemorrhage due to severe thrombopenia. Other observed toxicities were: asymptomatic hyperbilirubinaemia (46% cycles), mild nephrotoxicity (20% cycles), hypomagnesaemia (72% cycles), mild increase in body weight (100% cycles), hypertension (15% cycles) and headache (15% cycles). Overall the toxicity was acceptable and manageable. No alterations in absolute lymphocyte number, the lymphocyte subsets studied (CD3, CD4, CD8, CD19) or CD4/CD8 ratio were observed in patients receiving more than one treatment cycle, although there were significant and non-uniform variations in the values of the different lymphocyte subsets studied when pre- and post-treatment values were compared. There was also a significant increase in the CD4/CD8 ratio. Tumour regressions were observed in two patients (epidermoid carcinoma of the cervix and Ewing's sarcoma). The CsA dose recommended for phase II trials is a 5 mg kg-1 loading dose followed by a 3-day c.i. of 16 mg kg-1 day-1 simultaneously with DOX and IFX at the doses administered in this study.

Endoscopic detection of ischaemia with a new probe indicates low oxygenation of gastric epithelium in portal hypertensive gastropathy.

Changes in mucosal blood flow may be important in the pathogenesis of many conditions. Study of mucosal blood perfusion is difficult, and available methods have significant technical limitations. This study describes the development of an instrument for endoscopy, which indicates blood flow indirectly, by measuring the quantity of tissue oxygen that can diffuse from the mucosa to a luminal surface electrode. The instrument was used through an endoscope in patients with portal hypertension (n = 14), scleroderma (n = 3), disease controls (n = 7), and normal controls (n = 11). In portal hypertension readings were one quarter that in normal controls in both antrum (geometric mean (SEM) 35 (1.1)), nanoamps v 137 (1.1), and upper corpus 34 (1.1) v 125 (1.1)). Scleroderma patients showed greatly reduced oxygen readings in both antrum (18 (1.2)) and corpus (24 (1.2)), an expected but hitherto undiscovered result. These differences are highly significant (p = 0.0001), and the findings suggest that tissue hypoxia may contribute to mucosal changes in portal hypertensive gastropathy and in scleroderma.

Focal reduction of villous blood flow in early indomethacin enteropathy: a dynamic vascular study in the rat.

BACKGROUND: Oral indomethacin causes villous shortening, microvascular damage, and distortion, which might induce mucosal ischaemia and necrosis. AIMS: In order to determine the early events in indomethacin induced jejunal injury we examined the temporal relations between morphological damage and changes in villous blood flow following indomethacin. METHODS: In anaesthetised rats, mid jejunal villi were exteriorised in a chamber and observed by fluorescence microscopy. Blood flow in surface capillaries was calculated from velocities and diameters. Indomethacin was applied by both luminal and intravenous routes for 90 minutes, after which the animal was perfusion fixed and the villi were processed for histological examination. Control animals received intravenous or luminal bicarbonate (1.25%). RESULTS: Blood flow slowed in individual villi at 20 minutes, and progressed to complete stasis (in another group) by 45 minutes. Histological examination at 20 minutes revealed microvascular distortion, but no villous shortening; crypt depth:villous height ratios were 0.356 (0.02) in test and 0.386 (0.01) in surrounding villi (p > 0.05). At stasis, the villi under study showed epithelial clumping and were shortened: crypt depth:villous height ratios were 0.92 (0.2) in test and 0.42 (0.06) in surrounding villi (p < 0.02). Vehicle alone had no effect on either blood flow or histology. CONCLUSIONS: Focal slowing of villous blood flow and microvascular distortion precede villus shortening and epithelial disruption, and indicate that damage to surface microvasculature is an early event in indomethacin induced mucosal injury in this model.

A ferret model of acute multifocal gastrointestinal infarction.

Based on the demonstration of mural granulomatous vasculitis in Crohn's disease, it was hypothesized that this vasculitis may account for the discontinuous pattern of lesions in this condition. Accordingly, the present study investigated the histological changes produced by interruption of the submucosal and mucosal microcirculation in the ferret midgut. Two techniques were used. First, up to 30 adjacent vasa recta were ligated using microsurgical techniques; this produced no evidence of ischemic damage. Second, interruption of the submucosal collateral plexus by the intra-arterial injection of styrene microspheres (27-, 50-, or 90-microns diameter) produced acute intestinal mucosal damage. A combination of 27- and 90-microns spheres resulted in focal mucosal inflammation, necrosis, and ulceration. "Summit" lesions with normal adjacent mucosa were observed 48 hours after embolization, with evidence of regeneration of the mucosa overlying the occluded vessels at 72 hours. This model shows that focal gastrointestinal infarction with normal adjacent mucosa can be produced by acute occlusion of submucosal and mucosal arteries.

A vascular hypersensitivity model of acute multifocal gastrointestinal infarction.

We have investigated the hypothesis that submucosal vasculitis may account for the patchy transmural inflammation observed in Crohn's disease. Test ferrets (N = 11) were sensitized to human albumin. Five days after the last sensitization injection, human albumin microspheres (15-150 microns diameter) were injected intraarterially into the mesenteric circulation of a defined loop of mid-gut. Six control ferrets showed no histological abnormality at either 48 hr or two weeks after intraarterial injection. At 48 hr, five of six presensitized ferrets demonstrated submucosal vasculitis with fibrinoid necrosis. In two cases there was transmural inflammation and mucosal ulceration. A further five presensitized ferrets received weekly subcutaneous human albumin injections following the mesenteric intraarterial injection of albumin microspheres: after two weeks one animal demonstrated mild perivascular inflammatory changes and another demonstrated vasculitis. One of the two animals with transmural inflammation and mucosal ulceration at 48 hr, and the animal with vasculitis at two weeks, had precipitating antibodies to human serum albumin. This model demonstrates that an immune-mediated submucosal vasculitis can sometimes result in discontinuous transmural inflammation of the intestinal wall.