Deficiency of carnitine in cachectic cirrhotic patients.

Carnitine is synthesized from lysine and methionine. In the rat, inadequate intake of either of these essential amino acids causes carnitine depletion. Inasmuch as protein deficiency is common in the hospital population, we have investigated the possible occurrence of nosocomial carnitine deficiency. Fasting serum carnitine concentration was measured in 16 normal and 247 patients in 16 disease groups. Normal range of carnitine was 55-103 muM. Only the cirrhotic group showed significant (P < 0.05) hypocarnitinemia. 14 of 36 hospitalized cirrhotics had subnormal values for serum carnitine. The creatinine/height index, midarm muscle circumference, and triceps skin-fold thickness indicated protein-calorie starvation in the 14 hypocarnitinemic liver patients. In six of the hypocarnitinemic cirrhotics (average serum level 50% of normal), spontaneous dietary intakes of carnitine, lysine, and methionine were measured and found to be only 5-15% as great as in six normocarnitinemic, healthy controls. When these six cirrhotic and six normal subjects were given the same lysine-rich, methionine-rich, and carnitine-free nutritional intake, the normals maintained normal serum carnitine levels and excreted 100 mumol/day, whereas the cirrhotics' serum level fell to 25% of normal, and urinary excretion declined to 15 mumol/day. Seven hypocarnitinemic cirrhotics died. Postmortem concentrations of carnitine in liver, muscle, heart, kidney, and brain averaged only one-fourth to one-third those in corresponding tissues of eight normally nourished nonhepatic patients who died after an acute illness of a 1-3-day duration. THESE DATA SHOW THAT CARNITINE DEPLETION IS COMMON IN PATIENTS HOSPITALIZED FOR ADVANCED CIRRHOSIS, AND THAT IT RESULTS FROM THREE FACTORS: substandard intake of dietary carnitine; substandard intake of lysine and methionine, the precursors for endogenous carnitine synthesis; and loss of capacity to synthesize carnitine from lysine and methionine.

Evidence for central hypertyraminemia in hepatic encephalopathy.

In mongrel dogs, the effect of end-to-side portacaval shunt on plasma, cerebrospinal fluid (CSF) and brain tyramine, tyrosine, dopamine, norepinephrine, and epinephrine were studied. It was found that the level of tyramine in plasma, CSF, and selected brain regions increased steadily after the construction of the shunts. These elevations became more pronounced when the dogs manifested symptoms of hepatic encephalopathy. In postshunted dogs with stage II and III hepatic encephalopathy, tyramine concentration in corpus striatum (1,312 +/- 371), hypothalamus (400 +/- 67.0), and midbrain (660 +/- 78.7 ng/g) was significantly (P less than 0.05) higher than the level in dogs with stage 0 and I hepatic encephalopathy and sham-operated dogs serving as controls (corpus striatum, 831 +/- 140; hypothalamus, 167 +/- 40.0; and midbrain, 132 +/- 37.4 ng/g). This was followed by a concomitant depletion of dopamine and norepinephrine in these brain regions (postshunt: dopamine 104 +/- 20.0, 3,697 +/- 977, and 105 +/- 14.1; norepinephrine 521 +/- 71.6, 81.6 +/- 13.7, and 218 +/- 31.7 ng/g; vs. sham group: dopamine 532 +/- 83.1, 8,210 +/- 1,126, and 192 +/- 35.0; norepinephrine 1,338 +/- 425, 124 +/- 21.3, and 449 +/- 89.7 ng/g) of encephalopathic dogs with portacaval shunt. Furthermore, tyramine, tyrosine, dopamine, and norepinephrine levels in plasma and CSF increased markedly as clinical features in the dogs' behavior characteristic of hepatic encephalopathy occurred, including hypersalivation, ataxia, flapping tremor, somnolence, and coma. Cerebral hypertyraminemia and a defect in sympathetic neurotransmission may contribute to the development of hepatic encephalopathy of liver disease.

Accelerated fibrinogen and platelet destruction after peritoneovenous shunting.

Peritoneovenous shunting with the LeVeen valve is generally recognized as an effective procedure for the treatment of intractable ascites and renal failure associated with severe liver disease. We recently observed a generalized hemorrhagic diathesis in patients receiving these valves. To investigate the mechanism of this hemorrhagic complication, we prospectively performed kinetic studies with 51Cr-labelled platelets and 125I-labelled fibrinogen in ten patients. When results of studies before and after valve insertion were compared, the following reductions were noted: fibrinogen concentration, 55%; the fibrinogen survival, 49%; the platelet count, 55% and the platelet survival, 35%. No endotoxin was detectable in ascitic fluid preoperatively, and there was no apparent relationship between ascitic fluid cell counts and changes in fibrinogen and platelet survival. Until the component or components of ascitic fluid responsible for accelerated consumption can be identified and steps are taken to modify the rates of platelet and fibrinogen consumption, it would seem prudent to select patients for surgery conservatively.

Nasogastric hyperalimentation through a polyethylene catheter: an alternative to central venous hyperalimentation.

We performed nasogastric hyperalimentation with polyethylene catheters and appropriate feeding solutions in 12 cachectic patients who had been referred as candidates for central venous hyperalimentation. Most patients had primary gastrointestinal disease. The duration of hyperalimentation averaged 31 days. Seven patients obtained rapid weight gain (average 0.3 kg/day) with the nasogastric hyperalimentation alone. An additional two were successfully repleted with the addition of parenteral fluids via peripheral veins. In the nine repleted patients, serum albumin rose by average 19%, 24-hr urine creatinine by average 21%, and triceps skinfold by average 46%. The nature of the weight gain in the nine successful cases was analyzed by the metabolic balance study technique. Average composition of the increment in weight was: 50% protoplasm, 48% extracellular fluid, 19% adipose tissue, and less than 1% bone. We conclude that nasogastric hyperalimentation can replace central venous hyperalimentation in a substantial proportion of patients now receiving the latter type of treatment.

Revascularization for severe limb ischemia.

Patients with rest pain or tissue necrosis (generally correlated with a tibial-brachial pressure index of less than 0.25) have a high probability of limb loss unless revascularization is done. These problems are usually associated with extensive and multiple segments of anatomical occlusive disease. A review of 359 patients with such problems indicates that revascularization was done in 86%, with initial success in 92% of patients. The durability of successful limb preservation proved good, with 90% of patients having a comfortable, useful limb to time of death or for one year or longer after operation. Cumulative patency rates were 90% at five years for aortofemoral reconstructions, 70% at five years for femoropopliteal bypasses, and 49% at five years for femorotibial grafts. Cumulative limb salvage rates were 93%, 81%, and 67% at five years in the same categories, respectively.

Peritoneovenous shunts. Lessons learned from an eight-year experience with 70 patients.

The peritoneovenous shunt (PVS) is preferred over other treatment modalities in the treatment of the cirrhotic patient who has intractable ascites. The favorable effects on nutrition, pulmonary, and renal function, in addition to prompt control of ascites, frequently overshadow potentially life-threatening complications. We summarized our experience with the PVS in 70 patients with portal hypertension at Emory University, Atlanta, and identified the perioperative complications and operative mortalities. Late complications of sepsis and variceal hemorrhage were frequent and often were fatal. Of the multiple preoperative clinical and laboratory determinants, only the serum bilirubin level (greater than or equal to 3 mg/dL) was predictive of the operative mortality and longevity of survivors. The PVS should be reserved for patients with disabling, truly refractory ascites.

Infected descending aortic fistula.

Two patients, each with an infected descending thoracic aortic fistula, are described. The first patient had a postpneumonitic empyema. Thoracostomy tube drainage resulted in obliteration of the empyema cavity. Upon slight withdrawal of the tube, 49 days after its insertion, massive pulsating bleeding occurred through the sinus tract. The bleeding was controlled with manual pressure at the entry site of the chest tube, and the patient was operated upon immediately. A descending aortic defect, 3 cm long X 1.5 cm wide, at the site of the thoracostomy tube was primarily closed. Ten months after the surgical procedure, the patient has had no difficulty referable to her aortic erosion. In the second patient, 9 months after removal of the T-10 vertebra (which had a large cell tumor) and replacement of the vertebra with Dunn's metallic device, hemoptysis and left lower lobe consolidation developed. Aortography demonstrated a lobulated false aneurysm, 4 cm wide X 6 cm long, at the site of Dunn's device. A 16-mm graft was sutured end to side to the descending aorta just distal to the left subclavian artery and to the abdominal aorta below the renal arteries. The false aneurysm was then removed, the two ends of the aorta were sutured, and the stumps were covered with omental graft. Nine months after the repair the patient has had no difficulty referable to the aortic surgery.

Variceal hemorrhage in the veteran population. To shunt or not to shunt?

Portasystemic decompression remains the most definitive procedure in the control of portal hypertension (PHT) and bleeding gastroesophageal varices (BGEV). However, controversy prevails regarding shunt timing, type, and even propriety, especially in alcoholics. Analysis of a recent portal hypertension questionnaire submitted to 75 university-affiliated Veterans Administration Medical Centers (VAMC) reflected optimism regarding portasystemic shunts for the management of bleeding varices; disappointingly, however, on the average, only 20 to 25 per cent of variceal bleeders underwent definitive surgical management of any type. Ending in January 1980, a 14-year experience at the Atlanta VAMC with 72 portasystemic shunts was reviewed and demonstrates that shunt procedures may be extended to the veteran, predominantly alcoholic, population. Criteria for successful patient selection and operation are presented. While elective variceal decompression, preferably by the distal splenorenal shunt operation, may be performed with minimal morbidity and mortality, more efficient control of alcoholism is essential to prevent late deaths from hepatic failure.

Results with abdominal vascular trauma in the modern era.

This is a report of a 10-year experience (1989-1998) with 300 consecutive patients found to have an injury to a named abdominal vessel at the time of an exploratory laparotomy for trauma. An abdominal gunshot wound was the mechanism of injury in 78 per cent of patients, and injury to more than one named abdominal vessel was present in 42 per cent. The abdominal aorta, inferior vena cava, and external iliac artery and vein were the most commonly injured vessels. When management for the five most commonly injured arteries was grouped, exsanguination before attempts at repair occurred in 11 to 15 per cent of patients and the mean survival in the remainder was 46 per cent. When management for the five most commonly injured veins was grouped, exsanguination before attempts at repair occurred in 5 per cent of patients and the mean survival in the remainder was 64 per cent. A number of administrative and medical changes in the management of patients with abdominal trauma occurred from 1992 through 1994. Despite significantly increased Injury Severity Scores for patients treated from 1993 through 1998 as compared with those treated from 1989 through 1992 survival rates for patients with injuries to the abdominal aorta and inferior vena cava were unchanged. Survival rates for injuries to the external iliac artery and vein increased significantly. The local changes in management should be considered for prospective studies in other urban trauma centers.

Management of liver trauma in children.

The charts of 203 consecutive children undergoing emergency laparotomy for management of an acute liver injury were reviewed. Although drainage was all that was required in the majority of cases, massive hemorrhage, usually arising from disrupted hepatic veins, appeared to demand liver resection (as performed in 17 children, with a mortality of 18%) during earlier years of the survey. Control of exsanguinating bleeding from such liver wounds by packing with viable autogenous tissue (pedicled omentum) subsequently eliminated almost entirely the need for resection. Fatalities resulting from either hemorrhagic shock or loss of liver substance then became relatively rare. Additional significant problems were associated organ injuries, postoperative wound and intra-abdominal sepsis, bleeding diatheses, and late instances of hemobilia. The overall mortality was 6%.

Hyperalimentation in alcoholic hepatitis.

Enteral hyperalimentation in four patients with severe alcoholic hepatitis and anorexia increased spontaneous food intake, increased their nitrogen balance and the patients improved clinically. Seven patients with alcoholic hepatitis, who were clinically ill and able to eat only 410-1,100 calories per day, were given a 900 mosM/l. parenteral "hyperalimentation" solution by a peripheral vein (P-900). The intravenous nutrition provided daily 51.6-77.4 gm. amino acids in addition to oral intake. All patients improved. None developed detectable encephalopathy after 16-42 days of P-900 therapy. Five additional patients had ascites and alcoholic hepatitis. The daily infusion of 2,000 ml. P-900 was not associated with hyponatremia, renal failure or encephalopathy in four of these five patients who improved and continued their diuresis. P-900 therapy was discontinued in one because of progressive hyponatremia. The observations indicate that over and above the maximum tolerable oral nutrition, intravenous nutrition can be effectively utilized by clinically ill, jaundiced patients with alcoholic hepatitis without precipitating encephalopathy or interference with standard therapy of ascites.

Decarboxylation to tyramine: an important route of tyrosine metabolism in dogs with experimental hepatic encephalopathy.

Tyrosine metabolism via decarboxylation to tyramine was evaluated in dogs with functional end-to-side portacaval shunt. It was found that the endogenous plasma levels of both tyrosine and tyramine increased steadily after the construction of the shunt. These elevations became more pronounced when the dogs manifested symptoms of hepatic encephalopathy. In encephalopathic dogs, average endogenous plasma tyrosine and tyramine concentrations were 110.1 mumoles per liter and 7.6 ng per ml as compared to 55.4 and 1.2 in control dogs, respectively. The pattern of plasma concentrations of tyrosine and tyramine after an oral dose of L-tyrosine (50 mg per kg) was also investigated in control and shunted dogs. There was a progressive rise in peak levels of tyramine (to about 50-fold increase, at 6 weeks) after the construction of the shunt, as compared to levels obtained in pre- and at 1 and 4 weeks postoperatively (70.6 versus 1.20, 3.9, and 8.11 ng per ml). Similar observations were made with levels of plasma tyrosine. Six weeks after portacaval shunt, mean peak levels of plasma tyrosine, achieved at 5 hr after dose administration, were 450 as compared to 85 mumoles per liter obtained in preshunted dogs. These studies demonstrated a correlation between abnormalities in tyrosine metabolism and postshunt hepatic encephalopathy.

Impending aortoenteric hemorrhage: the effect of early recognition on improved outcome.

Aortoenteric hemorrhage is the result of enteric erosion and necrosis of aortic wall or anastomotic site. Mechanical or bacteriologic causes may occur singly or in combination. The temporal sequence is such that warning symptoms, often including back pain, fever, hemotochezia, and anemia, are present long before exsanguinating hemorrhage occurs. Vigorous diagnostic efforts, including gallium-67 citrate nuclear scan and computerized axial tomography, lead to a correct diagnosis. This allows planned semielective corrective operation before severe hemorrhage begins. The ideal operation consists of extra-anatomic revascularization, excision of the infected prosthesis, bowel repair with decompression, and sump drainage. Appropriate antimicrobial therapy should be continued until healing is complete. With aggressive diagnostic and therapeutic intervention according to this plan, marked improvement in survival and limb preservation can be anticipated in patients having this complication of aortic surgery. In this series, 15 of 18 patiets having operation recovered, though delayed limb loss occurred in two.

Vascular access for cancer chemotherapy.

A persistent problem in caring for patients receiving long-term intravenous chemotherapy for cancer is the maintenance of access to the vascular system. At Emory University Hospital between January 1975 and December 1977, 48 cancer chemotherapy patients had upper-arm bovine arteriovenous fistulas created for vascular access. The heterografts were inserted from the distal brachial artery to the proximal brachial vein under local, regional block, or general anesthesia. Grafts were functioning satisfactorily in 81% of the subjects at the time of death or at termination of treatment. The mean duration of useful graft patency for the entire group was 4.6 months, with a range of 0 to 27 months. Thrombosis of the graft was the most frequently encountered complication; infection and bleeding occurred infrequently. Reoperation was required in 35% of grafts, of which one half were salvaged. No mortality was attributable to the operative procedures or to the presence of a chronic arteriovenous fistula.

Enteral hyperalimentation: an alternative to central venous hyperalimentation.

Severe protein-energy undernutrition is a frequent finding among chronically ill patients. Its causes are anorexia, hypermetabolism, and malabsorption. Adverse consequences include impaired cell-mediated immunity increased susceptibility to infection, poor wound healing, weakness, and death. Spontaneous oral intake is inadequate in patients with this disorder, and therapeutic maintenance or repletion alimentation is needed. Enteral hyperalimentation is the method of choice, if tolerated. A successful treatment program usually requires a small-bore, flexible nasoenteral tube, appropriate feeding solution, and constant flow delivery of nutrient. If only partial dietary requirements are tolerated enterally, peripheral intravenous nutrient solutions can often supply the deficit. Although not suitable for all patients, enteral hyperalimentation is more physiologic, safer, easier, and more economical than central venous hyperalimentation. It would be well tolerated by many patients who now receive nutritional repletion by the latter method.

Technical pitfalls in measurement of venous plasma NH3 concentration.

Measurement of venous plasma NH3 in normal subjects by the ion-exchange method of Forman [Clin. Chem. 10, 497 (1964)] in a hospital clinical laboratory gave a mean value of 640 mug/liter (range, 300-1320 mug/liter; intraassay, intra-individual, and inter-individual coefficients of variation, 8, 47, and 47%, respectively). The following conditions adversely affect the reproducibility of the test: pollution of laboratory atmosphere and glassware by NH3-containing detergents; smoking by patient or analyst; delay, turbulence, or use of heparin lock in venipuncture; delay or warming of plasma above degrees C before mixing it with resin; and delay in colorimetric analysis of resin eluate. When these sources of error were eliminated, the mean value for normals was reduced to 330 mug/liter, the range was narrowed to 220-470 mug/liter, and the above-mentioned CV's were 5, 16, and 17%, respectively. With the precautions cited, furthermore, the intra-assay and intra-individual CV's for fasting NH3 concentration in cirrhotic patients were similarly reduced. An NH3 tolerance test was done by administering a standard dose of NH4Cl to patients and measuring venous plasma NH3 at 0, 15, 30, 60, and 90 min; the NH3 tolerance was quantified from the area under the curve relating concentration to time (mug - min/liter). As measured in the clinical laboratory, NH3 tolerance of cirrhotic patients showed intra-assay and intra-individual CV's of 50 to 90%. When the tolerance tests were repeated in the same subjects with the laboratory precautions listed above, these CV's were reduced to 8-15%.