Transplantation of segmental rat intestinal grafts including the ileocecal value and the ascending colon.

Extrinsic denervation and lymphatic disruption impair nutrient absorption after small bowel transplantation. The present study was undertaken to determine whether adding the ileocecal valve with or without the ascending colon would improve the function of a segmental intestinal graft. Five groups of Lewis rats (n = 10/group) were studied. Group I had a sham laparotomy. Groups II, III, IV, and V had the native jejunum, ileum, and cecum replaced with a graft. Inbred Lewis rats were used as isogeneic donors for the transplants to avoid the confounding effect of graft rejection. Group II had the entire jejunum and ileum transplanted. Group III had 20 cm of terminal ileum transplanted. Group IV had 20 cm of the terminal ileum including the ileocecal valve transplanted. Group V had 20 cm of the terminal ileum, the ileocecal valve, and the ascending colon transplanted. The terminal ileum-transplanted and terminal ileum/ileocecal valve-transplanted groups lost more than 25% of their preoperative weight by the end of the second postoperative week; most of these animals were killed because of inanition. In contrast, the sham laparotomy, jejunum/ileum-transplanted, and ascending colon-transplanted groups remained healthy until completion of the study on the 28th postoperative day. The ascending colon-transplanted group had slower intestinal transit and less bacterial contamination of the terminal ileum compared with the terminal ileum-transplanted and terminal ileum/ileocecal valve-transplanted groups (P < 0.05). Transplantation of the ascending colon and the ileocecal valve significantly improves the function of segmental small bowel isografts in rats. These data suggest that adding a colonic segment may be a simple method to improve the function of short-segment cadaveric and living-related intestinal grafts in humans.

Treatment with FK506 prevents rejection of rat colon allografts.

Colon transplantation has been proposed as a method to improve the function of an intestinal allograft. The present study examined the risk of colon rejection and the effect of FK506 on colon rejection in BN-->LEW rats with orthotopic bowel transplants. The first 4 groups included rats with untreated allografts (group 1), rats with isografts treated with 0.6 mg/kg FK506 (group 2), rats with allografts treated with 0.6 mg/kg FK506 (group 3), and rats with allografts treated with 0.4 mg/kg FK506 (group 4). In each of these groups (10-12 rats), half of the animals received a small bowel graft only (SB), while the other half received a small bowel, ascending colon, and cecum graft (SBC). The animals were followed daily until they died or were killed at 4 weeks. In group 5, an additional 18 untreated rats with SBC allografts were randomly killed on the third, fifth, seventh, and tenth postoperative days to study the sequential histopathologic and immunopathologic changes of colon rejection. There was no difference in survival, body weight, nutritional parameters, or bacterial contamination after SB and SBC transplantation. Intestinal transit was slower after SBC than SB transplantation (P < 0.05). Sequential histopathologic studies revealed that (1) the severity and time course of colon rejection was similar to small intestine rejection, and (2) the features of colon rejection were similar to ulcerative colitis. There was no evidence of graft-versus-host disease after SBC transplantation. In summary, adding a segment of large bowel to a small bowel allograft does not increase the risk of rejection or surgical complications. The transplanted colon slows intestinal transit. Treatment with FK506 effectively prevents colon rejection. These data suggest that adding a colon graft may improve the outcome of clinical small bowel transplantation.

Liver transplantation without venous bypass.

Fifty consecutive orthotopic liver transplants were performed without venous bypass in 41 recipients. Seven patients were transplanted twice and one patient received 3 transplants. The average age of the recipients was 37 years. The commonest indications for transplantation were primary biliary cirrhosis and cirrhosis from chronic active hepatitis. Fifty-eight percent of the recipients had undergone previous upper abdominal surgery. During the anhepatic period systolic blood pressure decreased by 21% to an average of 98 mm. of mercury. Cardiac output decreased by 52% to a mean (+/- SEM) of 3.89 +/- 0.21 L/min., and there was a doubling of the systemic vascular resistance. The hemodynamic alterations promptly returned to preclamping levels following hepatic revascularization. The average intraoperative transfusion requirements were 13 units of packed red blood cells, 9.6 units of platelets, 14.5 units of plasma and 6.6 L of crystalloid. Patients with previous surgery and retransplants required an average of 13 and 17 units of packed red blood cells, respectively. There was no deterioration in renal function in the postoperative period and no patient required hemodialysis. The 30 day survival was 87.8%. The 90-day and one-year actuarial survival is 80.5% and 68.8%, respectively. It is concluded that venous bypass is not necessary as a routine in orthotopic liver transplantation.

Pulmonary hypertension in sepsis: measurement by the pulmonary arterial diastolic-pulmonary wedge pressure gradient and the influence of passive and active factors.

To examine the relative roles of passive factors (flow; filling pressures of left side of heart) and active factors (acidosis; arterial unsaturation) in the genesis of pulmonary hypertension when associated with sepsis, 37 patients with sepsis and 24 patients without sepsis were examined. Pulmonary hypertension was measured by the pulmonary arterial diastolic-pulmonary wedge pressure gradient (PAd-PWP gradient) and correlated reasonably with a standard formula for calculated resistance ([PA--PWP]/CI, where PA is mean pulmonary artery pressure and CI is cardiac index). In 22 of 37 patients, sepsis was associated with a significant degree of resistance to flow in the pulmonary circulation, as measured by the PAd-PWP gradient: and the higher the PAd--PWP gradient, the greater the likelihood of early death. None of the examined passive or active factors appeared to be adequate to explain pulmonary hypertension when present. By the use of previously derived formulae to estimate the compliance of the elastic pulmonary arteries, factors affecting this part of the pulmonary microcirculation could not be held accountable for apparent pulmonary hypertension. Therefore, the presence of pulmonary hypertension in sepis appears to be an active, rather than a passive, phenomenon and unrelated to arterial oxygen saturation or acid-base imbalance. Although the exact cause is unknown, pulmonary hypertension in sepis is associated with a high mortality and may be clinically followed by measurement of the PAd-PWP gradient.

Abdominal sepsis managed by leaving abdomen open.

Intra-abdominal sepsis and necrotizing infection of the abdominal wall are usually fatal unless adequate drainage and wide debridement are possible. To follow these principles, we managed 18 seriously ill patients with abdominal sepsis by leaving the abdomen completely open. All except two of the patients had severe intra-abdominal sepsis. Eight patients had full-thickness wound infections and intra-abdominal infections refractory to the usual surgical drainage techniques. Two had necrotizing wound infections only. In 12 an upper abdominal incision was managed open, and in six the open incision was lower. As part of the initiating illness, there were eight small bowel and six colon fistulas. They were managed by colostomy in five patients and ileostomy in two. More than one organism was cultured in all patients and 12 of 18 had a positive blood culture. Respiratory failure made mechanical ventilation necessary in 13 patients for an average of 44 days. Previous adhesions, usually present, or an intact greater omentum, were necessary to prevent bowel evisceration, but three patients required paralysis and mechanical ventilation until adhesions became strong enough to prevent evisceration. There were seven deaths (39%), six caused by continuing sepsis and one from hemorrhage. In those surviving, granulation tissue grew over omentum or bowel loops to eventually seal the abdominal cavity. The late management was split-skin grafting in five and secondary closure in two. Four healed by second intention. We conclude that leaving the abdomen completely open facilitates the widest possible drainage, uncompromising debridement of the abdominal wall, and is compatible with good recovery. The ultimate result in survivors is acceptable. This technique is preferable to closing an abdominal wall of questionable viability in the face of intraperitoneal sepsis.

Amino acid metabolism in dogs with E. coli bacteremic shock.

In 10 fasting dogs receiving 10(9) viable E. coli bacteria per kilogram intravenously, mean systolic blood pressure decreased from 120.6 +/- 15.1 to 82.2 +/- 12.8 mm Hg. The association of hypoglycemia and increased arterial alanine and glycine with elevated plasma glucagon implied impaired gluconeogenesis. A rapid elevation of blood urea concentration, indicating increased ureagenesis, a fall of blood glucose, and an increase of net urea synthesis relative to that of glucose suggested that an increased proportion of the carbon residues derived from glucogenic amino acids is catabolized via pathways other than gluconeogenesis. In the bacteremic dogs the absolute net release from the leg of valine, isoleucine, and leucine and their net release relative to the net rate of proteolysis were decreased, suggesting increased oxidation of these amino acids in skeletal muscle. An increased net release of alanine relative to the net rate of protein catabolism in muscle was in agreement with this contention.

Metabolism of branched-chain amino acids in dogs with Escherichia coli endotoxin shock.

The arterial-femoral venous difference of phenylalanine concentration is proportional to net proteolysis in the leg. In ten fasting dogs receiving Escherichia coli endotoxin (2 mg/kg) intravenously, the mean systolic blood pressure decreased from 141.1 +/- 25 to 71.5 +/- 17 mm Hg. The absolute net release from the leg of valine, isoleucine, and leucine and their net release relative to net proteolysis (arterial-femoral venous difference in concentration of each branched-chain amino acid relative to that of phenylalanine) were decreased, indicating increased transamination of these amino acids in skeletal muscle. However, the net release of the branched-chain alpha-keto acids formed by transamination, relative to net proteolysis (arterial-venous difference in concentration of each alpha-keto acid relative to that of phenylalanine), was not increased. The findings indicate that in dogs with E. coli endotoxin shock, there is increased oxidative decarboxylation in muscle of the alpha-keto acids derived from valine, isoleucine, and leucine.

Capillary muscle blood flow in human sepsis.

Tissue perfusion was determined by cardiac index (Cl) and skeletal muscle capillary blood flow (MBF), and arteriovenous oxygen difference (AVD) and oxygen uptake were compared in seven patients with severe spesis and eight nonseptic patients. Skeletal capillary muscle blood flow also was measured before and after a 2 day fast in 14 normal volunteers. In both septic and nonseptic patients, MBF varied directly with Cl. The average muscle blood flow was greater in septic than in nonseptic patients and, in addition, was greater per unit Cl. AVD in septic patients was narrower than in nonseptic patients. Septic patients with an AVD of less than 4 ml. of O2 had markedly higher MBF and Cl than did septic patients with an AVD greater than 4 ml. of O2. Fasting normal volunteers who, like the septic patients, would be catabolic had a significant increase in MBF during the fast. Although peripheral shunts are not ruled out ans an explanation of the hyperdynamic circulation in sepsis, the evidence is against their existence, at least in skeletal muscle, since capillary flow increases in direct proportion to cardiac output. If capillary flow is increased in fact in sepsis, then flow like blood pressure becomes less of a critical factor in explaining the demise of the septic patient. It might be postulated that the increased capillary flow seen in sepsis is secondary to the mobilization of amino acids from the body cell mass for gluconeogenesis and energy.

Amino acid metabolism in patients with severe burns.

In this study we set out to determine, if relative to net catabolism of skeletal muscle protein as measured by phenylalanine release, the transamination of branched-chain amino acids (valine, isoleucine, and leucine) was greater in nonseptic burn patients than in controls. Arterial and femoral venous amino acid concentrations and circulating liver enzyme levels were measured. When the ratio of the arterial-femoral venous difference in concentration of each branched-chain amino acid to that of phenylalanine was determined, transamination of the branched chain amino acids, relative to net proteolysis, was not occurring at a greater rate in the burn patients. The net release of alanine relative to that of phenylalanine was not significantly greater in the burn patients, consistent with the conclusion that relative to the net rat of proteolysis, transamination of branched-chain amino acids in skeletal muscle is not increased in burn patients. This finding differs from that in septic dogs and septic humans. The mean arterial-femoral venous differences in concentration of alanine, valine, isoleucine, leucine, and phenylalanine were greater in the burn patients (P less than 0.03), indicating increased proteolysis in this group.

Total exchangeable potassium in patients receiving total parenteral nutrition.

Total exchangeable potassium (Ke) measurements were done on 13 patients who received total parenteral nutrition (TPN) at 25% dextrose and 2.5% amino acids (AA) (a hypertonic dextrose solution) by central vein, and 12 patients who received TPN as 10% dextrose, 10% fat emulsion (Intralipid), and 2.5% AA with lipid supplying about 60% of the nonprotein calories. There were 13 patients with benign diseases, three with carcinoma resected for cure, four with sepsis, three with severe burns, and two receiving chemotherapy for cancer. All but four of the septic and three of the burned patients had lost more than 10% of their normal body weight. A tracer dose of 150 to 300 microCi of K42 was injected intravenously during the first few days of the TPN therapy and a 48-hour urine collection and 24-, 36-, and 48-hour urine spot samples were obtained. Measurement of Ke was repeated after 7, 14, or 21 days of TPN therapy. Calories per kilogram per day, delta K per day, and percent delta K per day were determined for each patient and the data entered into a Textronics graphics computer for selection of curves of best fit. Similar curves of calories per kilogram per day versus delta K per day were obtained for the central and peripheral TPN groups. Intercepts where the body cell mass was maintained were at 32 calories/kg/day for the central TPN group and 34 calories/kd/day for the peripheral TPN group. Patients in both groups received a minimum of 0.6 gm protein/kg/day and the majority received from 0.8 to 1.2 gm protein/kg/day. Protein intake correlated with delta K per kilogram per day in the central but not the peripheral TPN group. We conclude that in nonseptic malnourished patients, an Intralipid calorie is as efficient as a dextrose calorie and that delta K is related to caloric rather than protein intake. None of the septic or burned patients maintained his body cell mass. However, they all received less than 30 calories/kg/day for a number of reasons.

Insulin decreases muscle protein loss after operative trauma in man.

Seventy-two hours after major operative trauma, nine patients receiving a constant infusion of calories (1460 kcal/m2/day) and protein (75 gm of amino acid/m2/day) showed a negative nitrogen balance, increased muscle catabolism, as measured by 3-methylhistidine excretion, increased amino acid efflux from muscle, and decreased circulating levels of insulin. When 5 U of insulin/hr were added to the infusate, arterial insulin levels rose significantly from 39.7 +/- 4.1 microU/ml to approximately the pretrauma levels (74.6 +/- 7.7 microU/ml). Despite this normalization of insulin levels, excretion of nitrogen and 3-methylhistidine and the efflux of amino acids from forearm muscle fell but did not return to pretraumatic levels, suggesting some insulin resistance. Visceral gluconeogenesis from amino acids appeared to decrease, since insulin infusion decreased the efflux of alanine from skeletal muscle with no change in its arterial level. Insulin also significantly reduced the efflux of isoleucine, tyrosine, phenylalanine, glutamine, and total amino acid nitrogen from forearm muscle. These findings, along with the partial reduction in the excretion of 3-methylhistidine and nitrogen, suggest that insulin, in combination with infused calories and protein, decreases the loss of muscle protein after trauma.

Major operative trauma increases peripheral amino acid release during the steady-state infusion of total parenteral nutrition in man.

The effect of major operative trauma on skeletal muscle metabolism was examined in nine patients receiving a constant infusion of calories (1460 kcal/m2/day) and protein (75 gm of amino acids/m2/day) for 5 days before and 4 days after an operation. Compared with the preoperative state, 72 hours after the operation there was a significant rise in arterial levels of glucagon, cortisol, norepinephrine, and inactive triiodothyronine and a drop in concentrations of insulin, active triiodothyronine, and amino acids. Forearm blood flow increased, as well as the efflux from forearm muscle of lactate, taurine, serine, glycine, valine, methionine, isoleucine, leucine, phenylalanine, lysine, arginine, and total amino acid nitrogen (440%). This loss of muscle protein after trauma is associated with increased muscle proteolysis, as measured by increased urinary 3-methylhistidine excretion (83%), and accounts for increased nitrogen loss (54%) from the body. Increased activity of the sympathetic nervous system is manifested by increased levels of epinephrine and norepinephrine, a relative lack of insulin, and increased levels of glucagon. This hormonal milieu plays an important role in the production of hypoaminoacidemia, increased efflux of amino acids and lactate from muscle, and negative nitrogen balance observed in these traumatized patients.

Femoral arteriovenous amino acid differences in septic patients.

Femoral arteriovenous differences and flux of amino acids across the leg were measured in seven septic patients and compared with those of six nonseptic patients on days 1 and 3 following major surgery. The septic patients were seriously ill and judged clinically to be catabolic. The postoperative patients, although not septic, were expected to have a maximal catabolic response to operation during the first 3 days after operation. Both groups had increased release of phenylalanine from the leg, an index of muscle proteolysis. Septic patients had decreased femoral arteriovenous differences (--20 vs --74 and --60 mumoles/liter) and decreased flux (34 vs 169 and 128 nm/100 gm of calf muscle) of the branched-chain amino acids as compared with the nonseptic postoperative patients on days 1 and 3. The arterial plasmal levels of the branched-chain amino acids and alanine were not different, but phenylalanine was elevated in the septic patients (88 vs 49 and 55 mumoles/liter). The insulin:glucagon molar ratio was lower in the septic patients (2.4 vs 4.4 and 5.5). These findings suggest that in the catabolism of sepsis there is greater oxidation of branched-chain amino acids in muscle than in the catabolism associated with uncomplicated surgery.

Arterial plasma amino acids in patients with serious postoperative infection and in patients with major fractures.

Arterial plasma amino acids were measured in 27 patients with serious septic complications after operation, 15 patients following reduction of femoral shaft fractures and nine control patients on the first and third days following uneventful major abdominal surgery. Amino acid concentrations in the controls were similar to those which have been reported during early starvation. The amino acid patterns seen in all groups did not resemble that previously observed following glucocorticoid administration. In the patients with infection, mean phenylalanine concentration (108.0 +/- 46.9 mumoles per liter) was significantly greater than in the controls on the first (p greater than 0.001) or third (p less than 0.001) postoperative days. Four of the septic patients with hyperphenylalaninemia also had elevated arterial methionine concentrations. These observations suggest that many of the patients with sepsis had seriously impaired liver metabolism. In patients with fractures, the concentrations of ornithine (p less than 0.001), taurine (p less than 0.05), and aspartic acid (p less than 0.05) were lower than in controls. No other significant differences of amino acid concentrations were observed. It is difficult to relate these differences to a specific metabolic abnormality.

2,3-Dihydroxybenzoic acid. Effect on mortality rate in a septic rat model.

Neutrophil-derived oxygen-free radicals may play a role in organ dysfunction associated with generalized sepsis. A rat model was used to test the effects of two free radical scavengers, dimethyl sulfoxide (DMSO) and 2,3-dihydroxybenzoic acid (2,3-DHB), on mortality from intra-abdominal sepsis produced by cecal ligation and perforation. Being an iron-chelating agent, 2,3-DHB may have an additional bacteriostatic effect. Therapeutic regimens included no treatment; gentamicin sulfate (2 mg given intraperitoneally [IP] every eight hours); DMSO (2 g/24 hr given IP every eight hours in divided doses); 2,3-DHB (35 mg/kg given IP every eight hours); and combinations of gentamicin with each free radical scavenger. No statistically significant improvement in survival was obtained by therapeutic intervention with gentamicin alone, DMSO alone, 2,3-DHB alone, or gentamicin in combination with DMSO. When used in combination with gentamicin, 2,3-DHB yielded a statistically significant improvement in survival when compared with gentamicin alone or with no treatment. These results show that 2,3-DHB when used in combination with gentamicin has a beneficial effect on mortality following intra-abdominal sepsis in this model.

Cardiovascular changes in sepsis.

The hyperdynamic circulation characteristic of severe sepsis is not likely due to peripheral arteriovenous shunts since in skeletal muscle, at least, capillary blood flow is increased and varies directly with cardiac index. The finding that flow is normal in some septic patients who are severely ill and close to death suggests that blood flow can no longer be considered the critical factor explaining the death of the septic patient. Clearly, the commonly accepted sequence of low blood flow, tissue hypoxia, lactacidosis, and death does not apply to all patients dying from shock. The hyperdynamic circulatory state and the metabolic changes associated with severe sepsis may be related. Skeletal muscle capillary blood flow was increased in fasting normal subjects and septic postoperative patients, both of whom were catabolic. Therefore, elevated blood flow, which is characteristic of severe sepsis, may be a response to the catabolism of body protein required for energy production. If this concept of sepsis is accepted, it follows that treatment which heretofore has been aimed at increasing blood flow and blood pressure should be redirected to therapy which provides energy substrates and alters hormonal patterns to favor anabolism.