The enhancing effect of bone marrow cells on the primary immune response of the isolated perfused spleen.

The addition of bone marrow cells or peripheral lymphocytes to the isolated pig spleen markedly enhanced the primary antibody response after 3-day perfusion and antigenic challenge in vitro. The splenic preparation without added cells or with the addition of marrow cells to an irradiated spleen gave a limited response. Contributory evidence is provided that at least two distinct cell types are needed for antibody production. For optimal antibody response by an isolated perfused spleen, marrow cells or peripheral lymphocytes should be added to the system.

Distribution of oxygenated blood in experimental perfusion via the femoral and carotid arteries.

The efficiency of femoral and carotid artery perfusion was compared in 10 dogs. The purpose of our study was to determine under what conditions the femoral artery can be used to perfuse the upper body satisfactorily during prolonged perfusions.

Replacement of portal vein during pancreatectomy for carcinoma.

Three patients are described in whom the portal vein, surrounded but not penetrated by pancreatic carcinoma, was resected and replaced by an expanded Teflon (Gore-Tex) tube. One patient, who has survived 32 months without recurrence of disease, had a radiographically patent graft at 2 and 17 months after operation. A second survivor has a patent graft at 8 months. The third patient died 10 days after operation with an open prosthesis. Excision of involved portal vein during pancreatoduodenectomy for carcinoma of the pancreas is justified in the absence of metastases. Vein grafting is the best means of portal vein reconstruction. Gore-Tex appears to be a suitable prosthesis when the portal vein must be sacrificed. Its clinical success in terms of patency substantiates previous experience in experimental animals.

Liver failure in the postoperative patient: the role of sepsis and immunologic deficiency.

In a 4 year experience with 7- postoperative or postinjury patients in liver failure, we found that 27 (37 percent) had associated bacterial sepsis. The mortality rate in this group of patients was 48 percent, compared with an average of 80 percent among patients with liver failure from other causes. Laboratory evaluation of cellular immune response in 20 cirrhotic patients confirmed the clinical observation of a relative failure of cellular immunity. Mitogen response of lymphocytes to phytohemagglutinin was significantly decreased in cirrhotic patients as was the number of thymus-derived lymphocytes. The association and poor prognosis of pre-existing liver damage, injury, bacterial sepsis, and liver failure is discussed. It is concluded that a postoperative or postinjury patient, regardless of his general condition, who has any possibility of having undrained intraperitoneal sepsis, deserves re-exploration since the trade-off for benefit usually far outweighs the possibility of doing harm. The decision for reoperation in such critically ill patients is not easy, but such operative aggressiveness is warranted.

Erythrocytes cultured from bone marrow.

The long-range objective of this study is in vitro tissue culture of bone marrow stem cells to produce erythroid cells of sufficient volume for clinical transfusion. Bone marrow from dogs and patients was cultured in 29 experiments lasting up to 15 weeks. Peripheral erythroid cells from dogs were cultured in three control experiments. Optimal tissue culture media was NCTC 109 augmented with vitamin B12, erythropoietin (EP), folic acid, and 9% fetal protein in 60 mm glass Petri dishes. Additional media and Step III EP was added at 2 to 4 day intervals. Peripheral erythroid cells in culture all were dead within 4 weeks. Marrow erythroic cells in culture proliferated as demonstrated by (1) Fe59 incorporation into cells during culture, (2) H3 thymidine uptake into cultured cells, (3) microscopic evidence of mitoses, and (4) total erythrocyte concentration in cultures far exceeding that of peripheral culture controls. For as yet unexplained reasons the total mature red blood cell concentration in the culture media remained essentially constant throughout the studies. This is a first step in achieving the ultimate goal of bulk erythrocyte production from tissue culture.

Immunosuppression by hyperbaric oxygen.

Reports from several laboratories have indicated that hyperbaric oxygen might be immunosuppressive in animals. We examined the effect of hyperbaric oxygen on a well-studied model of cell-mediated immunity in the mouse, contact sensitivity to dinitrofluorobenzene (DNFB). Using this model, we showed that daily 5-hour exposure of mice to 2.5 ATA hyperbaric oxygen was markedly immunosuppressive. Immunosuppression occurred when mice were exposed to hyperbaric oxygen (HBOX) for 4 days daily before DNFB sensitization or for 5 days daily after sensitization. The immunosuppression was reversed by intravenous administration of 2 x 10(7) peritoneal exudate cells from syngeneic mice, but was not reversed by 5 x 10(7) lymph node cells intravenously. We showed that daily HBOX exposure resulted in a dramatic decrease in circulating total leukocytes and lymphocytes in spleen weight, and in DNA synthesis in draining lymph nodes of sensitized mice. Serum cortisol levels were only marginally elevated in HBOX-treated mice.

Hepatic support with hepatocyte suspensions in a permeable membrane dialyzer.

Hepatic support by whole liver, liver slice, or hepatocyte perfusion is unsatisfactory. A new concept of support with circulating hepatocytes on one side of a cuprophan membrane and blood from an anhepatic pig on the other was evaluated in 19 dialysis experiments. Levels of pyruvate metabolism and oxygen utilization were similar to those obtained when liver cells in a centrifuge are perfused with oxygenated plasma. Of six anhepatic pigs given late hepatic support, five showed neurologic improvement. One had concomitant improvement in electroencephalographic results. A membrane hemodiolysis unit in which hepatocytes circulate against blood from an anhepatic animal is the most suitable extracorporeal liver support unit described so far. Its simplicity allows repeated and prolonged use.

Government-provided health care. A prediction.

Government-provided health care, provided primarily by the Department of Veterans Affairs and the Department of Defense, must prove itself competitively cost-effective for its continued survival. Cooperation and unification of health programs at each level of government are essential to minimize uneconomic duplications. Some form of a managed-care HMO will be necessary for future health care programs of the Department of Veterans Affairs and the Department of Defense. Enrollees must include the spouses and dependents of veterans. Each population segment of government-provided care should undergo a make-or-buy analysis to guide cost-effective strategies. A voucher system should be instituted on an experimental basis to explore the transition between the current system of health care and alternative use of community-based private care. Such a system has the political advantage of taking away nothing from the current rights of veterans or retirees. A variety of joint ventures between government health systems and private sector HMOs have been described. Cost cannot be the only consideration in make-or-buy government decisions. Overriding national priorities of defense, training of future physicians, and scientific advancement must also be considered.

Development of a new vascular prosthetic: lessons learned.

In 1972, polytetrafluoroethylene (PTFE) (GoreTex) wire insulation was first implanted into arteries and veins of dogs at the University of Colorado Health Sciences Center in Denver. Subsequent modifications including fibrillar arrangement, wall thickness, and pore size led to the development of a vascular prosthesis with tissue ingrowth, viable neointima, and encouraging patency rates. Tissue culture, surface charge, and platelet studies have all demonstrated the optimal biologic qualities of PTFE grafts. Replacement of one human portal vein by a PTFE graft for invasive carcinoma of the pancreas was strikingly successful. Only unexplained intimal hyperplasia at the arterial anastomotic margins discouraged us from extensive early clinical trials.

Operative choice and technique following pancreatic injury.

During a five-year period, 50 patients were operated on for pancreatic injury. Forty had open trauma and ten closed. Half of each group were treated by drainage alone. Overall mortality was 14%, with all deaths following open injuries due to gunshots. Deaths were more frequent after drainage than after resection. Substantial complications occurred in 70%, with more frequent amd more serious morbidity in drained patients. Sump drainage was associated with less morbidity than Penrose drainage. Guidelines for management of pancreatic trauma are (1) resection of sinistral gland for perforating injuries of the body or tail; (2) drainage of perforations of the pancreatic head when the major duct is intact; (3) resection of duodenum and pancreatic head for devitalizing injury of both structures.

Depression of cell-mediated immunity in diabetes.

Parameters of cell-mediated and humoral-mediated immunity were measured in ten infection-free, insulin-dependent, controlled diabetic patients and in ten similar but nondiabetic patients awaiting elective operations. Tests performed included total and differential leukocyte counts, neutrophil reduction of nitroblue tetrazolium, mitogen response of lymphocyte to phytohemagglutinin, ratio of thymus-derived to bone marrow-derived lymphocytes, serum immunoglobulins IgA, IgG, and IgM, macrophage inhibition factor, serum zinc, and reaction to skin test antigens. Diabetics had a significantly (P less than .05) decreased mean response response to phytoheagglutinin stimulation and a lowered ratio of thymus-derived to bone marrow-derived lymphocytes. These findings support the concept of depressed cell-mediated immunity in the controlled, adult diabetic and might explain the propensity of the uncontrolled diabetic to increased frequency and severity of bacterial infection.

Accuracy of ultrasound in diagnosing abdominal masses.

B-mode ultrasonography was performed in 246 patients with suspected abdominal masses over a seven-year period. In 105 (40 percent), the accuracy of ultrasonic diagnosis was evaluated surgically. Sonography was proven correct in 60 (57 percent) patients who had undergone operation. Among 141 patients who had not undergone operation and whose diagnoses were established by other means, ultrasonography agreed with the clinical diagnosis in 69 (31 percent). Ultrasound accuracy, as confirmed by operation, was highest for splenic masses (100 percent) and for aortic aneurysm (88 percent). Liver masses were correctly identified in 56 percent of patients and gallbladder lesions in 38 percent. While only a 48 percent accuracy was obtained in diagnosing pancreatic disease, 64 percent of all pseudocysts were localized. Ultrasonography correlated positively with operative findings in 56 percent of renal masses. Intraperitoneal abscess was accurately diagnosed in 61 percent of patients but retroperitoneal adenopathy in only 33 percent. Abdominal ultrasonography, while accurately diagnosing splenic and aortic masses, failed to identify approximately half of other mass lesions. Improved techniques hold promise of improving this diagnostic accuracy.

Feasibility of damage control surgery in the management of military combat casualties.

Unique and expanded applications of staged operative management are undergoing careful evaluation in many civilian level I trauma centers under the rubric of damage control surgery. Recently there have been advocates for its broad application to the early management of critically injured combat casualties. However, the enormous logistic requirements for such strategies are contrary to the demands of the usual wartime scenario. On the basis of experience in civilian trauma centers and combat casualty management, we question the suggested extensive role of damage control surgery during wartime. Each decision point in damage control surgery should be analyzed as it is altered (sensitivity analysis) by conditions of war. It is unwise to adopt such indications unchanged from current civilian trauma policy.

Reassessment of simple cholecystostomy.

Cholecystostomy was performed on 22 patients with acute cholecystitis after partial (13) or complete (9) removal of gallbladder stones. One patient had complementary common-duct drainage. Early mortality occurred in two patients. Three patients with associated cholangitis but intraoperative reflux of cysticduct bile were all treated by cholecystostomy alone and survived. For the poor-risk patient with cholecystitis, cholecystostomy is effective. When there is associated cholangitis and documented cystic-duct patency, cholecystostomy is also sufficient. When accompanying cholangitis is associated with cystic-duct occlusion, choledochotomy and T tube drainage should be added.

Extracorporeal Assist - Liver, Spleen, and Lung.

Laboratory and clinical studies have been described demonstrating the metabolic efficiency of animal livers perfused with human blood. Such a preparation has been used to support patients in liver failure. Pig and calf spleens washed free of blood will also function when perfused with human blood. Preliminary attempts have been made utilizing the cellular and circulating antibodies from such a preparation for immunotherapy on five patients with metastatic malignant melanoma. Laboratory studies are reported evaluating the use of pig and calf lungs as a biologic oxygenator for human blood.

The continuing challenge of acute and perforated appendicitis.

A two year review of 216 patients with a preoperative diagnosis of appendicitis is used to help clarify the continuing points of controversy concerning appendicitis. Morbidity of perforated appendicitis by far exceeds that of nonperforated cases. Twenty-nine per cent of patients had perforation with a postoperative complication rate of 33 per cent. Rate of wound infection after perforation was 15 per cent despite antibiotics and delayed wound closure. Pediatric patients in this study did not experience a higher incidence of perforation or complication. A decision tree for management of acute appendicitis is presented.

Reoperation after abdominal trauma.

A five year experience with 782 patients requiring laparotomy for trauma is reviewed. Specifically, the 70 patients requiring unplanned reexploration have been studied to delineate the indications for and implications of such repeat laparotomies. The major indications for such reoperation were intraabdominal abscess (45.7 per cent), bleeding (15.5 per cent), peritonitis (12.1 per cent), and small bowel obstruction (8.6 per cent). There were 16 negative reexplorations (13.8 per cent). Overall mortality in the reexplored patients was 21.4 per cent, all victims of gunshot or blunt trauma. Mortality correlated with the number of required reexplorations, being 67 per cent in those requiring four operations. Of the 31 laparotomies performed initially for diffuse or localized intraabdominal sepsis, only 15 were highly suspected, and 13 of these by simple chest x-ray findings. If after laparotomy for repair of intraabdominal trauma a patient fails to meet the anticipated norm of convalescence, a high index of suspicion for early postoperative hemorrhage, or later sepsis, should be maintained. Such patients have far more to gain than lose by reexploration.

Morbidity of colostomy closure.

An unexpectedly high morbidity (28 per cent) followed colostomy closure in 100 patients. One patient died postoperatively because of sepsis resulting from disruption of the colon anastomosis. Wound infection (10 per cent), intraperitoneal abscess (1 per cent), bowel obstruction (7 per cent), and fecal fistula (4 per cent) were other significant complications. Wound sepsis was greater after primary than after delayed wound closure. Obstruction did not correlate with the use of either an open or closed technic of anastomosis. Three patients required reoperation for complications. Temporary colostomy was constructed for colon injury in 85 per cent of patients. In view of the considerable morbidity of colostomy closure, alternate technics of managing colon trauma should be considered. Such technics include primary closure and exteriorization of repaired colon. When temporary colostomy is unavoidable, closure is best done by open, two layer anastomosis with delayed wound closure. Colostomy should be recognized as an important procedure associated with significant morbidity.

Clinical indications and accuracy of gray scale ultrasonography in the patient with suspected biliary tract disease.

One hundred patients with suspected biliary tract disease underwent gray scale cholecystosonography (GSCS) and had diagnostic confirmation by oral cholecystogram (OCG) and/or operation. Ultrasonography demonstrated the gallbladder in 94 of the 100 patients; 2 patients had had previous cholecystectomy and 3 of the 4 remaining patients had documented stones with no confirmation of a nonvisualizing OCG in the other patient. Among the 88 patients with OCG, GSCS findings correlated in 91 per cent (2 per cent false-positive; 7 per cent false-negative). Among the 43 operative patients, GSCS was proven correct in 91 per cent (no false positive; 9 per cent false-negative). Of 12 patients with jaundice GSCS correlated with operative findings in 75 per cent (no false-positive; 25 per cent false-negative). Diagnostic errors occurred in patients with very small biliary calculi, particularly when a single stone was impacted in the cystic duct. Failure to identify the gallbladder with ultrasound signifies probable cholelithiasis in the patient without previous cholecystectomy. On the basis of this experience, we conclude that (1) GSCS is most useful when jaundice or acute illness precludes conventional studies; (2) GSCS provides an inexpensive, quick, accurate means of diagnosing cholelithiasis with a very high specificity (97 per cent) and moderate sensitivity (88 per cent); and (3) GSCS is the optimal diagnostic procedure for evaluating the biliary tract in the acutely ill, jaundiced, vomiting, allergic, and/or pregnant patient.

Cholecystectomy with and without surgical drainage.

Thirty-seven patients who met specific criteria had cholecystectomy without drainage, and thirty-seven matched control patients had cholecystectomy with drainage. This study suggests that surgical drainage after every uncomplicated cholecystectomy is unnecessary and may be unwise. Such drainage may result in an increased incidence of postoperative morbidity and prolonged hospital stay.