Three-dimensional view of the vascular structure of the lower esophagus in clinical portal hypertension

The venous anatomy of the lower esophagus and upper stomach was studied in nine patients with portal hypertension and in five without, following infusion of a silicon rubber compound into vessels of the excised organs within whole tissues made transparent with methyl salicylate. Four venous channels were identified in normal tissues: intraepithelial, subepithelial superficial, deep submucosal and adventitial veins. In portal hypertensive patients, giant esophageal varices formed 2 to 3 cm above the esophagogastric junction fused with numerous superficial and deep submucosal veins. Gastric varices were present in the submucosa where the left gastric venous branch penetrated the gastric wall 2 cm below the esophagogastric junction. The lower esophageal varices were classified into two types of vascular structure: palisading type in five and bar type in four. The palisading type had dilated intraepithelial channels and numerous small superficial collateral veins extending in a longitudinal arrangement. The bar type had triply dilated subepithelial superficial veins and deep submucosal veins which eroded the epithelium, and the gastric varices were present in the latter type. Our study provides evidence of the three-dimensional vascular structure of the lower esophageal varices without the necessity for tissue dissection.

Trial of sclerosing agents in patients with oesophageal varices

Forty-five cirrhotic patients with oesophageal varices underwent endoscopic injection sclerotherapy in a prospective randomized trial carried out to compare two sclerosing agents (5 per cent ethanolamine oleate and 2 per cent sodium tetradecyl sulphate (STD] with respect to safety, efficacy and complications. Twenty-three patients were allocated to the ethanolamine group and twenty-two to the STD group. The rate of control of acute bleeding was 100 per cent (6/6) in the ethanolamine group and 75 per cent (3/4) in the STD group. There was a significantly lower rate of postinjection bleeding after the over-tube was removed at the initial session of sclerotherapy when ethanolamine was injected 0/23 versus 7/22, 32 per cent; P less than 0.01) and at the second session there was a significantly (P less than 0.01) higher rate of jet-like bleeding from injection sites in the STD group (6/21, 29 per cent) than in the ethanolamine group (0/22). The disappearance rate of red colour signs 1 week after the initial session of sclerotherapy in the ethanolamine group was 100 per cent and 62 per cent in the STD group. Early oesophageal ulcers developed less frequently in the ethanolamine group (0 and 9 per cent) than in the STD group (24 per cent and 43 per cent both after the initial (P less than 0.05) and the second session of sclerotherapy (P less than 0.01]. Early bleeding from an oesophageal ulcer occurred only in the STD group (5/21, 24 per cent) before the third session of sclerotherapy (P less than 0.05). The rate of early mortality did not differ between the two groups. We conclude that ethanolamine seems to be safer and more efficacious than STD for sclerosing oesophageal varices.

An ultrasonic duplex system facilitates detection of portal hemodynamic changes following selective shunts for esophageal varices

We used an ultrasonic duplex system (US system) to assess portal hemodynamics in 52 patients with liver cirrhosis and esophageal varices, who underwent 2 types of distal splenorenal shunt (DSRS), conventional DSRS (group A, 8 patients) or DSRS with splenopancreatic disconnection (group B, 44 patients). The portal blood flow rate (PBF) was determined in 64 out of 70 patients (91.4%) and the shunt flow rate (SVF) in 39 out of 42 patients (92.9%) who had angiographically confirmed patent portal vein and shunt vein, during the peri- and postoperative period. In group A, a remarkably small amount of postoperative PBF (193 ml/min) and a concomitant increase in SVF (1039 ml/min) were evident. Such ultrasonic findings were compatible with a reduction in portal vein diameter, in accordance with the poor portal perfusion grade of the liver, and a transpancreatic stealing of the portal blood flow to the shunt, as evidenced by postoperative angiography. In contrast, the reduction in PBF was minimal, that is 663 ml/min preoperatively to 562 ml/min at discharge, and 536 ml/min at late follow-up, in group B patients. Significant alterations in portal circulation of the group B patients were not evident angiographically. This US system is most useful to assess portal hemodynamics in patients with a selective shunt.

Intraoperative hemostasis and wound healing in intestinal anastomoses using the ILA stapling device

Intraoperative hemostasis and wound healing of 24 side-to-side intestinal anastomoses constructed with the ILA stapling device were studied in 12 dogs by comparing the ILA-32 and ILA-52 staple cartridges. Hemostasis was evaluated by intraoperative measurement of blood loss and bleeding time at the staple line. There was no statistically significant difference in mean blood loss (p greater than 0.05) or mean bleeding time (p greater than 0.10) between the two cartridges. Wound healing was studied using bursting strength measurements and silicone rubber casting of the microvasculature at the staple line. At 3 days, 1 week, and 2 weeks postoperatively, there was no significant difference between bursting strength values achieved with the two cartridges. Microscopic examination revealed that wound healing in the ILA-52 anastomoses lagged behind healing in the ILA-32 anastomoses at each postoperative time period studied. The silicone rubber casting study showed a paucity of microvasculature at the healing staple line with the ILA-52 cartridge as compared with the ILA-32 cartridge. Our findings suggest that the ILA-52 cartridge does not offer significantly improved intraoperative hemostasis over the ILA-32 cartridge and may affect the microvasculature at the staple line in a way that delays wound healing.

Dopamine-mediated behaviours produced in naive mice by bromocriptine plus SKF 38393

The ability of bromocriptine and SKF38393 (2,3,4,5-tetrahydro-7,8-dihydroxy-1-phenyl-1H-3-benzazepine) to produce some dopamine-mediated behaviours has been assessed in mice. Soon after injection, SKF38393 produced moderate increases in grooming and sniffing which were not very intense, while bromocriptine (with or without SKF38393) inhibited all grooming behaviour. Bromocriptine alone also depressed rearing and sniffing in the first hour and SKF38393 alone was without effect on rearing, but the combination produced a marked increase in the incidence of both rearing and sniffing, both of which behaviours appeared to be stereotyped. When bromocriptine-induced locomotor stimulation was peaking about 3 h after injection, as measured in automated activity cages in previous studies, there was an increase in sniffing and rearing, the incidence of which was unaffected by the addition of SKF38393, perhaps due to the shorter duration of action of SKF38393 than of bromocriptine. The data indicate that the D-1 agonist SKF38393 can qualitatively and quantitatively alter the behavioural spectrum produced by the D-2 agonist bromocriptine.

Temporary deterioration of pulmonary functions after injection sclerotherapy for cirrhotic patients with esophageal varices

In 34 cirrhotic patients with esophageal varices, a significant but temporary deterioration in pulmonary function tests occurred 24 h after endoscopic injection sclerotherapy using 5% ethanolamine oleate. Included were vital capacity, forced expiratory volume in 1 s, closing volume/vital capacity and arterial oxygen content. Twenty-four hours after the sclerotherapy, the patients complaining of postinjection retrosternal pain had a larger fall in vital capacity and forced expiratory volume in 1 s than did the patients without pain. Before the injection sclerotherapy, 11 of 34 patients had an arterial hypoxemia (PaO2 less than 80 mm Hg). In these patients, there was a significantly (p less than 0.001) higher value of closing volume before sclerotherapy and there were larger changes in both closing volume (p less than 0.01) and arterial oxygen content (p less than 0.01) 24 h after the injection sclerotherapy than in the patients without hypoxemia. Reversion to a state before sclerotherapy was attained 7 days after the sclerotherapy. Thus, patients undergoing sclerotherapy for bleeding esophageal varices should be closely monitored with regard to pulmonary function.

Further studies on the interaction between bromocriptine and SKF38393 in reserpine and alpha methyl-para-tyrosine-treated mice

In a previous report, we showed that the relatively selective dopamine (DA) D-2 agonist bromocriptine (BRC), when combined with the selective D-1 agonist SKF38393, produced in DA-depleted mice a marked locomotor stimulation, despite BRC and SKF38393 being inactive by themselves (Jackson and Hashizume 1986). The present series of experiments was designed to further explore this interaction. In all experiments, mice were pretreated with reserpine and/or alpha methyl-p-tyrosine (AMPT). In mice pretreated with reserpine, AMPT or reserpine plus AMPT, BRC plus SKF38393 produced marked excitation whether the BRC was given 3 or 1 h prior to the SKF38393 challenge. However, while there was no absolute requirement that BRC be given a certain time before SKF38393, this factor was of some importance, with the onset of locomotor stimulation produced by the combination being much more rapid if the BRC was given 3 h rather than 1 h before the SKF38393. Interestingly, the degree of locomotor stimulation produced by the combination was always greatest in the animals premedicated with reserpine alone. If AMPT was also used (with or without reserpine), the stimulation produced by the combination was reduced, which may have resulted in part from a non-specific depressant effect of the AMPT. From these results, it seems as though endogenous DA is not required for BRC to work, provided that D-1 receptors are stimulated.(ABSTRACT TRUNCATED AT 250 WORDS)

Direct vision valvulotomy in in situ venous bypass

The autogenous saphenous venous bypass is accepted as an optimal procedure for distal arterial reconstruction. However, serious complications, including incomplete valvulotomy, laceration of the venous wall and persistent arteriovenous communications, are found in in situ bypasses. To avoid these complications, a new and simple technique using an angioscope in the in situ venous bypass has been developed. Angioscopy prevents tearing of the vein wall and avoids inaccurate incision of the valves.

Direct vision valvulotomy for nonreversed vein graft

A new and easy technique using an angioscope in a nonreversed, isolated vein bypass is suggested herein. Direct vision valvulotomy with an angioscope in nonreversed vein graft saves operating time, prevents tearing of the vein wall and inaccurate incision of the valves.

Argon laser angioplasty with a laser probe

A preliminary basic study of argon laser angioplasty with the use of a specially designed probe is presented. Arterial specimens were harvested from 10 amputated lower extremities. The studies included the evaluation of coaxial laser angioplasty in 10 partially or totally occluded arteries; the observation of the effect of perpendicularly applied laser energy on 35 thrombi, 54 soft and 10 hard atherosclerotic, and 51 normal arterial walls; the comparison of laser energy requirements for coaxial vessel lumen enlargement or recanalization vs. perpendicular penetration in 10 occluded, hard atherosclerotic arteries; and the spectrographic analysis of calcium density gradient in two specimens. The results showed that the power required for vessel lumen enlargement was 10 +/- 6 J/mm of atherosclerotic vessel (mean +/- standard deviation). The probe followed the original arterial lumen, did not perforate the vessel wall, and created a smooth, enlarged path. The power required to penetrate perpendicularly to a similar depth for thrombi, soft and hard atherosclerotic plaques, and normal arterial walls was 15 +/- 4, 30 +/- 15, 65 +/- 32, and 246 +/- 123 J/mm, respectively. In the hard calcified specimens, laser energy required for coaxial lumen enlargement or recanalization was significantly less than that for perpendicular penetration (p less than 0.05), which correlated with the calcium density map indicating an increase from inside to outside.

Intimal response of saphenous vein to intraluminal trauma by stimulated angioscopic insertion

Endothelial damage after simulated angioscopic insertion was evaluated in 22 dogs by radioimmunoassay of endogenous 6-keto-prostaglandin F1 alpha (stable product of prostacyclin produced by the endothelium) by means of a closed perfusion system and scanning electron microscopy. In 20 dogs, a 1.7 mm diameter polyethylene tube similar in size and physical characteristics to a commercially available angioscope was passed once into the distal venotomy site and out of the proximal venotomy site for a total distance of 15 cm in the left saphenous vein (group A) and 10 times in both directions in the right saphenous vein, destroying all valves (group B). The vein segments were examined immediately after simulated angioscopic insertion and at intervals of 2, 3, and 4 weeks. The remaining two dogs were used as control animals to establish baseline prostacyclin production. All veins were patent when harvested. Scanning electron microscopy revealed globular and contracted endothelial cells in some areas of group A veins, whereas partial longitudinal absence of endothelial cells was noted in group B veins. Intimal coverage was complete by 2 weeks in group A veins and between 3 and 4 weeks in group B veins. Prostacyclin synthesis of group B veins was significantly less than that of group A at day 1 and at 2 weeks after operation (p less than 0.05), reaching that of control vein synthesis by 4 weeks. This study suggests that simulated angioscopic trauma of the luminal surface of the canine saphenous vein had no negative effect on early patency.

Bromocriptine-induced locomotor stimulation in mice is modulated by dopamine D-1 receptors

Mice were pretreated with reserpine plus alpha-methyl-p-tyrosine (10 mg/kg plus 200 mg/kg). One hour later they were administered the selective dopamine D-2 agonist bromocriptine or vehicle. Three hours after the bromocriptine, mice were challenged with the selective D-1 agonist SKF 38393, and locomotor activity was measured each 5 min for three hours. Neither bromocriptine nor SKF 38393 produced significant stimulation. The combination, however, produced a dose-dependent and coordinated increase in activity. If the bromocriptine was given only one hour before the SKF 38393 challenge (i.e., three hours after the reserpine plus alpha-methyl-p-tyrosine), no interaction was seen. In naive mice, when SKF 38393 and bromocriptine were administered together, the locomotor response to bromocriptine was quantitatively and qualitatively altered. The initial depressant response to bromocriptine was shortened, producing a more rapid onset of the stimulant response. In one experiment, the maximal activity induced by bromocriptine was increased by SKF 38393. The ability of SKF 38393 to alter the locomotor stimulant effect of bromocriptine in naive mice was blocked by their pretreatment with the selective D-1 antagonist, SCH 23390. The data indicate that the locomotor stimulant effects of bromocriptine are modulated by D1 receptors.

Ultrasonic duplex system to facilitate assessment of portal blood flow velocity

The usefulness of an ultrasonic duplex system to assess portal blood flow was investigated. In a model involving a steady flow through a vinyl tube in agar, there was a significant linear correlation between the maximum blood flow velocity measured by this system (V-max) and the mean blood flow velocity calculated from the actually measured blood flow volume (V-mean), that is, V-mean = 0.53 X V-max was obtained (r = 0.994; n = 47). This equation was used to calculate the mean portal blood flow velocity by this system (V-dopp) in 10 patients with liver disease, and the findings were compared with data simultaneously obtained by cineangiographic mapping of Lipiodol droplets released into the portal vein through a catheter placed in situ at the time of surgery (V-cine). A linear correlation between V-dopp and V-cine was statistically significant (r = 0.970; n = 13), and the regression line was V-cine = 1.29 X V-dopp -2.11. The ultrasonic duplex system proved reliable for a quantitative assessment of portal hemodynamics.

Patterns of pannus growth of the expanded polytetrafluoroethylene vascular graft with special attention to the intimal hyperplasia formation

Serial histologic observations of the healing process of expanded polytetrafluoroethylene (PTFE) vascular grafts in dogs were performed. The common iliac arteries of 12 mongrel dogs were replaced with a 2-cm segment of 6 mm PTFE graft, bilaterally. The grafts were studied microscopically after the animals were killed at 4, 8, 12, 16, 20, and 24 weeks. All the grafts remained patent up to 24 weeks. Special attention was paid to the histological characteristics of pannus growth, which were divided into two types: smooth appearing and irregular. The former type was characterized by smooth, thinner connective tissue growth along the graft, and the latter was defined as irregular-shaped pannus containing large and small thrombi. Only pannus of the smooth type showed similar growth patterns and growth rates. These findings suggest essentially that pannus develops smoothly from both ends of the graft; however, thrombi disturbed this growth pattern. In 24 grafts, eight nodular lesions of the pannus (greater than 500 mu in height) were found: five distally and three proximally. All eight lesions were found within 3 mm of the anastomoses. Among these eight lesions, five were on the completely healed pannus, and the remainings were all on the irregular pannus. This suggests that thrombi organization followed by nodular lesions are major offending factors that lead to the formation of intimal hyperplasia.

Bromocriptine induces marked locomotor stimulation in dopamine-depleted mice when D-1 dopamine receptors are stimulated with SKF38393

In mice pretreated with reserpine plus alpha-methyl-p-tyrosine, neither the D-2 selective agonist bromocriptine, nor the D-1 selective agonist SKF38393, produced any measurable increase in locomotion in mice. However, the combination of the two agonists produced a marked and dose-dependent increase in co-ordinated locomotor activity. In mice with their dopamine stores and dopamine synthesis intact, SKF38393 was inactive by itself, but significantly enhanced the stimulant effect produced by bromocriptine. The data suggest that bromocriptine requires concomitant stimulation of D-1 receptors for the full expression of its behavioural stimulant effects.

The role of prostacyclin in patients with portal hypertension

The serum levels of prostacyclin in portal venous blood samples from patients with portal hypertension were measured by radioimmunoassay to assess the role of prostacyclin in this clinical entity. Portal venous prostacyclin activity exceeded peripheral venous activity 2-3 fold in both cirrhotics and non-cirrhotics. The size of esophageal varices showed an inverse relation to the concentration of portal venous prostacyclin in the cirrhotics. There was no significant difference between the portal vein pressure and the size of varices, and the portal vein pressure did not correlate with the concentration of 6 keto-PGF1 alpha. Thus, serum levels of portal venous prostacyclin may depend on the development of collateral circulation in cirrhotics and increased splanchnic blood flow probably accelerates the washout effect.

Hyperdynamic circulation in the left gastric venous area in patients with portal hypertension: angiological assessment

The time of appearance of the left gastric vein on serial celiac arteriograms in patients with portal hypertension and esophageal varices was compared with that of the portal vein to assess regional hemodynamics in the left venous portion of the stomach, an area located in close proximity to the varices. In two thirds of all the patients with cirrhosis or non-cirrhotic idiopathic portal hypertension (IPH), the left gastric vein was visualized earlier or simultaneously than or with the portal vein, while in all but one patient with prehepatic portal obstruction, there was a delayed opacification of the left gastric vein. These results suggest the presence of a hyperdynamic circulatory state which promotes venous hypertension in the left gastric venous area of the stomach of a considerable number of patients with cirrhosis or IPH. In such a hemodynamic state, selective decompression of varices can be achieved by a left gastric venous caval shunt.

Exclusion of nonisolated splenic vein in distal splenorenal shunt for prevention of portal malcirculation

In an attempt to prevent portoprival malcirculation after distal splenorenal shunt (DSRS), a splenic hilar renal shunt (HRS) with proximal flush ligation of splenic vein was designed. To accomplish this procedure, two methods were compared: HRS alone (Group A) and HRS plus proximal flush ligation of the splenic vein (Group B). In Group A, which included 20 cirrhotic patients with esophageal varices, angiographic as well as pulsed Doppler flowmetric follow-up study revealed a portal thrombosis in two patients and severe narrowing of a portal vein in another two. Considerable stealing flow was observed in these four patients. In the Group B series, which included 33 cirrhotic patients, there were no gross changes in the portal hemodynamics. Normal prograde portal flow was confirmed by Doppler flowmeter in this series including 14 patients of more than 8 months after surgery. When the amount of nonisolated splenic vein embedded in the pancreas is minimized, portal malcirculation after distal splenorenal shunt can, to a great extent, be prevented.

The natural history of non-alcoholic cirrhosis

To assess the natural history of non-alcoholic liver cirrhosis, one hundred and eighty medically treated Japanese cases, including 110 accompanied by esophageal varices were investigated retrospectively. Among those patients with varices fifty-one (46.4%) bled from the upper gastrointestinal (GI) tract and thirty-two (29.1%) from esophageal varices, while GI bleeding was found in only six out of 70 patients without varices. The GI bleeding rate was the highest in patients with varices and concomitant hepatoma (76.5%). The mortality rate of the GI bleeders was 68.6% in patients with varices and 33.3% in patients without varices. The mortality on the first variceal bleeding episode was 65.6%, and another 25.0% had rebleeding from varices, resulting in a one-year survival of 9.4%. The ten-year cumulative percentage of variceal bleeding was 61.2% in patients with varices, and that of occurrence of hepatoma was 50.7% in total of 180 patients. This study revealed that the non-alcoholic cirrhotic patients have a highly rate of complication by hepatoma and that the development of hepatoma doubles the risk of varix rupture.

Morphology of gastric microcirculation in cirrhosis

Morphologic alterations in the gastric microcirculation in cirrhosis were investigated following infusion of a silicone rubber compound into vessels of the excised stomach which was then cleared with methyl salicylate. In cirrhosis, arteriovenous anastomoses 15 to 50 micron in diameter were present in 5 of 10 patients; spiral arterioles were less than one-tenth as numerous as in stomachs from noncirrhotic patients. The arterioles mainly had a straight pattern and precapillaries, capillaries, and submucosal and subserosal veins were dilated in cirrhotics (p less than 0.05). The number of arteriovenous anastomoses was unrelated to the degree of vascular dilatation and the number of spiral arterioles. These morphological alterations are consistent with decreased arteriovenous flow resistance in the stomach of cirrhotic patients.