Neutralization of meningococcal endotoxin by antibody to core glycolipid.

Antibodies to Escherichia coli J5, a uridine 5'-diphosphate-galactose epimerase-less mutant of E. coli 0111, neutralized meningococcal endotoxemia from all three major capsular serogroups. We chose the dermal necrosis of the local Shwartzman phenomenon and the renal cortical necrosis of the general Shwartzman phenomenon as assays because these are the hallmarks of meningococcemia, and because meningococcal lipopolysaccharide (LPS) is a uniquely potent cause of dermal purpura and necrosis. Meningococcal antisera raised against LPS from MGC A, B, and C also provided good protection against endotoxemia from the homologous capsular groups, but it was inconsistent against the heterologous serogroups. The superiority of J5 antibodies (purified IgG as well as antiserum) is probably due to the fact that J5 LPS contains only the endotoxin core. Consequently, immunization with this mutant stimulates production of antibodies to core LPS without interference by the "0" antigenic determinants of the side chains. These observations indicate that the endotoxin core is the toxic moiety of meningococcal LPS, that the core LPS of meningococcus (MGC) is immunologically similar to enteric LPS, and that the antigenically variable "0" side chains of MGC LPS interfere with antibody production against the common core. They also suggest that antibodies prepared against this E. coli mutant could interrupt the devastating course of meningococcal endotoxemia in man, regardless of the capsular serogroup of the infecting strain.

Human antiserum for prevention of the local shwartzman reaction and death from bacterial lipopolysaccharides.

Bacterial lipopolysaccharides from dead bacteria have been blamed for the continuing high mortality from gram-negative infections despite antibiotic treatment. Because animal antiserum against these lipopolysaccharides has been shown to protect against several of the effects of endotoxin, we undertook the development of antiserum in human subjects. 21 men were immunized with a single injection of Salmonclla typhimurium or Escherichia coli 0:111 heat-killed cells and immune serum was collected at 2 wk. Preimmune serum was obtained as a control in all animal experiments. 1 ml antiserum given intravenously protected mice against a lethal intravenous dose of homologous endotoxin (P < 0.005 for both antisera). E. coli antiserum reduced the incidence of positive local Shwartzman reactions with E. coli endotoxin from 100 to 38%; S. typhimurium antiserum reduced the incidence from 92 to 35%. (P < 0.0005 for both antisera). There was no protection against heterologous endotoxin in either animal model. These experiments demonstrate for the first time that human antiserum confers exceedingly potent passive immunity to the effects of endotoxin.

Induction of immunity against lethal Haemophilus influenzae type b infection by Escherichia coli core lipopolysaccharide.

Efforts to prevent Haemophilus influenzae type b (HIB) infections in infancy have been hampered by the low immunogenicity of capsular polysaccharide vaccines in children younger than 18 mos. In searching for alternate immunogens, we have studied the protective potential of polysaccharide-poor, lipid-rich endotoxin (LPS) core in experimental HIB infections. Because all gram-negative bacteria have similar LPS core structures, we were able to use as vaccine the J5 mutant of Escherichia coli 0111, the LPS of which consists only of core components, and thus to avoid problems in interpretation arising from vaccine contamination with non-LPS HIB immunogens. Mice were given graded inocula of HIB and developed lethal infection analogous to human HIB disease when virulence was enhanced with mucin and hemoglobin. After active immunization with heat-killed E. coli J5, 40/50 (80%) of infected mice survived, compared with 14/50 (28%) of saline-immunized controls (P less than 0.005). Passive immunization with rabbit antiserum against E. coli J5 prevented lethal HIB infection when administered 24 or 72 h before or 3 h after infection. This protection was abolished by adsorption of antiserum with purified J5 LPS, with survival reduced from 14/24 to 0/24 (P less than 0.005). Furthermore, rabbit antiserum to purified J5 LPS gave just as potent protection against death as antiserum to whole J5 cells. These studies demonstrate that immunity to core LPS confers protection against experimental murine HIB infection and provide the framework for a new approach to prevention of human disease from HIB.

Immunization against nosocomial infection.

Overwhelming infection with gram-negative bacteremia has become the most serious nosocomial infection in compromised patients. Because gram-negative bacteria share a common core lipopolysaccharide, we tried to develop a single vaccine or antiserum that might control these infections regardless of species. We used a mutant of Escherichia coli 0111 (J5) deficient in uridine diphosphate-galactose (UDP-GAL) epimerase and thus unable to attach "0" side chains, so that core lipopolysaccharide was exposed. A vaccine composed of this mutant produced antibody that gave broad protection against lethal infections by different gram-negative bacteria in immunosuppressed animals. The J5 vaccine protected against 98 percent lethal doses of Pseudomonas aeruginosa, and J5 antiserum improved survival tenfold in animals dying of Esch. coli, Klebsiella and Pseudomonas bacteremia. The protection with vaccine or prophylactic antiserum was undiminished in animals challenged six weeks after immunization. Encouraged by these results, we conducted a double-blind trial in patients with gram-negative bacteremia. In those given J5 antiserum, the mortality rate was cut in half and survival from deep shock increased from 28 percent to 82 percent. Because of these preliminary results in 136 patients, the study has been extended to 300 patients and the double blind code will be examined again to see if the early favorable results are confirmed and extended.

HA-1A. A human monoclonal antibody for the treatment of gram-negative sepsis.

HA-1A is a human monoclonal IgM antibody that binds to endotoxin. The results of the clinical trials of HA-1A demonstrate that HA-1A reduces mortality among patients with sepsis and gram-negative bacteremia. Secondary endpoints, including resolution of organ failure, discharge from intensive care unit, and discharge from the hospital, support the beneficial effects of the antibody. The antibody is well tolerated with rare side effects, including hypotension and urticarial rash. No anti-HA-1A antibodies have been detected.

Pseudomonas aeruginosa vasculitis and bacteremia following conjunctivitis: a simple model of fatal pseudomonas infection in neutropenia.

During attempts to create a realistic model of fatal bacteremia due to Pseudomonas aeruginosa during immunosuppression, it was found that the invasive as well as the disseminated phase of infection could be mimicked by gentle instillation of 10(8) colony-forming units of P. aeruginosa into the intact conjunctival sac of agranulocytic rabbits. Within 48 hr animals developed conjunctivits leading to severe necrotizing vasculitis and fatal bacteremia. Twelve of 26 strains from patients with P. aeruginosa infections were virulent, causing death in 50%--100% of animals. Nine (75%) of 12 isolates from blood but only two (15%) of 13 isolates from sputum and urine were highly lethal. Neither proteolytic enzyme production nor serum resistance alone accounted for virulence. No infection developed in animals and normal leukocyte counts or in neutropenic animals given Escherichia coli, Klebsiella pneumoniae, or non-aeruginosa pseudomonads. A rare vasculitic lesion was observed in animals inoculated with Serratia marcescens. This model, which illustrates the distinctive features of P. aeruginosa infection, is so simple and reproducible that it should be useful for evaluation of the efficacy of drugs and immunization against Pseudomonas in the compromised host.

An immunoprotective monoclonal antibody to lipopolysaccharide.

The ability of antisera against lipopolysaccharide (LPS) raised by immunization with gram-negative bacteria to prevent LPS toxicity and death from gram-negative bacteremia is well established. To demonstrate conclusively that the protective antibody is specific for LPS, we tested an anti-LPS monoclonal antibody (mAb) in three animal models. 7G is an IgG3 mAb directed against an oligosaccharide side chain determinant of LPS from E. coli 0111:B4. This anti-LPS mAb increased the LD50 of 0111:B4 LPS in mice and protected rabbits against the dermal Shwartzman reaction elicited by 0111:B4 LPS. 7G mAb also protected mice against lethal infection with mucin-enhanced E. coli 0111:B4. Pretreatment with 250 micrograms of 7G increased the LD50 by more than 1.5 logs. These studies prove that oligosaccharide side chain-specific antibody to LPS confers protection against LPS toxicity in vivo and against experimental gram-negative infection. In addition, these studies suggest the potential of anti-LPS monoclonal antibody as therapy for gram-negative infection.

Antiserum treatment of gram-negative bacteremia.

In order to lower the mortality rate from gram-negative bacteremia, 136 patients were treated with human antiserum against core glycolipid, or with control nonimmune human serum, in a double-blind clinical trial. The antiserum was prepared by immunizing healthy young men with a vaccine composed of heat killed cells of the J5 mutant of E. coli 0111 B4. Since the core glycolipid in this mutant is not encumbered with "O" side chains, it can stimulate antibody against the core glycolipid possessed in common by the different species of gram-negative bacteria responsible for lethal bacteremia in patients. No serious reactions occurred in over 300 men receiving this vaccine, and the J5 antiserum gave striking broad-spectrum protection against experimental gram-negative bacteremia and endotoxemia. When human J5 antiserum was administered to seriously ill bacteremic patients, the mortality rate was virtually cut in half, as compared to controls. The death rate from gram-negative bacteremia was 14% in patients treated with J5 antiserum, and 26% in those given nonimmune control human serum. Among patients in profound gram-negative bacteremic shock, the recovery rate rose from 29% in controls to 82% in those treated with J5 antiserum (p = 0.02). On the basis of these encouraging results we propose to treat more bacteremic patients with J5 antiserum, because larger groups are needed to establish the full significance of the initial findings.

Antiserum to endotoxin in hemorrhagic shock.

Antiserum to Escherichia coli J5, a mutant endotoxin (LPS) which contains only core determinants, has proven effective in reducing mortality from endotoxic shock due to a wide variety of gram-negative bacteria. Twenty New Zealand white rabbits with coliforms in the gut were subjected to hemorrhagic shock of 36 mm Hg for 3 hr. Treated rabbits were resuscitated with 15 cc of rabbit J5 antiserum (hemagglutinating antibody titer against J5 lipopolysaccharide of 1:1024), remaining shed blood, and lactated Ringer's to achieve a mean arterial blood pressure (MABP) within 20% of baseline. The control group was similarly resuscitated but received 15 cc normal rabbit serum (titer 1:2). Catheters were removed and rabbits were returned to their cages until death or 5 days of survival. Hemodynamic parameters (heart rate, MABP, cardiac output, and total peripheral resistance) did not differ significantly between groups. However, six treated rabbits survived 5 days (60%) and no control rabbit lived past the third postexperimental day (P less than 0.019). Our data suggest that systemic endotoxemia may contribute to morbidity and mortality in severe hemorrhagic shock.

Antibody to cell wall glycolipid of Gram-negative bacteria: induction of immunity to bacteremia and endotoxemia.

Antiserum to the core glycolipid of gram-negative bacteria was prepared by immunization of rabbits with vaccine composed of killed cells of the uridine diphosphate galactose-deficient mutant (J5) of Escherichia coli O:111. Antiserum to J5 not only prevented death of animals from endotoxin but also prevented the local and generalized Shwartzman reactions. Antiserum to endotoxin also prevented renal cortical necrosis and disseminated intravascular coagulation during the evolution of the generalized Shwartzman reaction. Antiserum to be J5 mutant was successful in the treatment of overwhelming bacteremia produced by other gram-negative bacteria; in addition to bacteremia cause by coliform organism, antiserum to J5 was dramatically effective in treatment of bacteremia due to Pseudomonas aeruginosa. One injection of rabbit antiserum to J5 improved the survival rate from 15% in controls to 59% in treated animals (P less than 0.002). Active immunization with J5 vaccine was even more effective against pseudomonas bacteremia: such immunization improved the survival rate from 13% in controls to 92% in vaccinated rabbits. Since an antiserum effective against the J5 mutant of E. coli can be prepared safely in human subjects, such immunotherapy should be considered for patients with gram-negative bacteremia.

Prevention of lethal pseudomonas bacteremia with epimerase-deficient E. coli antiserum.

All of these experiments demonstrate that despite many microbiologic differences between Pseudomonas and the enteric rods, there is a close immunologic relationship between their endotoxin cores. These studies also show that by genetic manipulation we can unmask a protective antigen for preparing antiserum with a broad range of activity, even in the face of neutropenia. Production of potent human J5 antiserum is a safe, simple procedure, making it possible to consider endotoxin core antibody treatment of all Gram negative bacteremia in patients regardless of the causative organism.

Lethal Haemophilus influenzae type b infection in mice.

Previous animal models of invasive Haemophilus influenzae type b (HITB) infection are characterized by a low mortality rate. We produced a highly lethal infection in CF1 mice using mouse passage, mucin, and hemoglobin to enhance infectivity. Infection by the intraperitoneal route was followed by progressive peritonitis and bacteremia with subsequent HITB infection of the brain and meninges, and death. Death occurred between eight and 72 hours after infection and was associated with 10(6) to 10(9) HITB per ml of blood and with 10(2) to 10(5) HITB per g of brain. Mucin-hemoglobin did not augment HITB growth, but impaired macrophage adherence to glass in vitro, without decreasing cellular viability. In vivo, mucin-hemoglobin decreased the rate of disappearance of 51Cr-labelled HITB from the blood by impairment of hepatic clearance. This technically simple and inexpensive model is useful for the study of HITB infections in which bacterial multiplication, invasion and host lethality are desired features.

Ineffectiveness of single-dose human antiserum to core glycolipid (E. coli J5) for prophylaxis of bacteremic, gram-negative infections in patients with prolonged neutropenia.

It has been recently established that serum from human volunteers immunized with E. coli J5 vaccine prevents death of patients with gram-negative shock. The present study addressed the question whether the prophylactic administration of a similar amount of J5 antiserum could protect neutropenic patients from acquiring gram-negative infections. One hundred patients, the majority of which had acute non-lymphoblastic (63%) and lymphoblastic (29%) leukemia, presented 109 episodes of neutropenia. Sixty of the 100 patients underwent bone marrow transplantation. All patients were given one unit of either pre-immune (control) or J5 antiserum serum from volunteers at the onset of neutropenia. When compared to control serum, J5 antiserum given prophylactically did not reduce the number of febrile days, the number of gram-negative bacteremic episodes, or death from these infections. This inability to demonstrate a beneficial effect of prophylaxis with a single unit of J5 antiserum in prolonged neutropenia may have several explanations that are discussed.

Demonstration of cross-reactive antibodies to smooth gram-negative bacteria in antiserum to Escherichia coli J5.

We investigated the discrepancy between the broad cross-protection against gram-negative infections afforded by antiserum to Escherichia coli J5 and its apparently narrow cross-reactivity in vitro. Rabbits immunized with J5 bacteria produced antibodies to both the J5 lipopolysaccharide (LPS; titer by ELISA, 1:60,000) and LPS from the Re mutant of Salmonella minnesota (i.e., to the ketodeoxyoctonate [KDO] and lipid A determinants; titer, 1:3,200). In highly diluted antiserum, titers of antibody to J5 LPS were reduced by 28%-41% after adsorption with seven strains of smooth gram-negative bacteria and by only 4% after adsorption with the Re mutant. Smooth gram-negative bacteria adsorbed virtually all antibody to Re LPS. Therefore, rabbit antiserum to J5 contains type-specific antibodies to core determinants distal to KDO that can obscure highly cross-reactive antibodies to lipid A-KDO in vitro. Cross-reactive antibodies are demonstrable by adsorption with whole bacteria at limiting concentrations of antibody.

Analysis of the fine specificity and cross-reactivity of monoclonal anti-lipid A antibodies.

We have investigated the fine specificity of anti-lipid A antibodies to identify conserved lipid A antigens. Because lipid A derived from many different Gram-negative bacteria has similar biologic activities, the conserved regions may be of particular importance for the immunostimulatory and toxic properties of lipid A. We found that five of nine antibodies bound to a wide variety of Gram-negative bacteria. All these widely cross-reactive antibodies bound to the same antigenic site within lipid A. Polymyxin B, an inhibitor of lipid A activity, bound to this site as well. The widely cross-reactive antibodies bound to native and base-hydrolyzed lipid A equally well, and also bound to the monosaccharide precursor lipid X. The less cross-reactive antibodies recognized base-hydrolyzed lipid A poorly, and did not recognize lipid X at all. Other investigators have shown that lipid X has some of the activities of lipid A in vitro and can inhibit the lethal toxicity of LPS in vivo. On the basis of this study, we suggest that lipid X contains a conserved lipid A epitope as well.

Inhibition of lipopolysaccharide activation of 7OZ/3 cells by anti-lipopolysaccharide antibodies.

We have investigated the ability of mAb against LPS to inhibit LPS-induced activation of 7OZ/3 pre-B cells. The fine specificity and relative affinity of these mAb for lipid A and LPS were also determined. We found that antibodies inhibited only the activity of glycolipids which they bound with relatively high affinity. However, two high affinity antibodies binding to non-lipid A epitopes did not block cellular activation. Some, but not all, relatively high affinity antibodies binding to the lipid A region of the LPS molecule inhibited biologic activity. The inhibitory antibodies bound to at least two distinct epitopes within the lipid A region. These data suggest that LPS interacts with 7OZ/3 cells in a highly specific fashion.

Amebic meningoencephalitis caused by Balamuthia mandrillaris: case report and review.

Balamuthia mandrillaris, formerly referred to as a leptomyxid ameba, is a free-living ameba that has recently been identified as a cause of meningoencephalitis. Previously, only two genera, Naegleria and Acanthamoeba, were recognized as causes of central nervous system (CNS) infections in humans. In contrast to Naegleria, Balamuthia causes a subacute-to-chronic infection of the CNS. Distinct from Acanthamoeba, which appears to favor the immunocompromised host, Balamuthia is capable of infecting both healthy and immunosuppressed hosts. Retrospective analyses as well as an accumulation of newly identified cases have demonstrated that this ameba is an increasingly important pathogen to recognize. We report the isolation, histopathologic features, and confirmation by indirect immunofluorescence of B. mandrillaris in a case of fatal amebic meningoencephalitis.

Influence of alterations in foregoing life-sustaining treatment practices on a clinical sepsis trial. The HA-1A Sepsis Study Group.

OBJECTIVES: To evaluate the timing of foregoing life-sustaining treatments in patients enrolled in a sepsis trial and to determine their influence on patient outcome and trial results. DESIGN: Subset of patients in a prospective, randomized, double-blind, placebo-controlled study. SETTING: Twenty-three academic medical centers. PATIENTS: Enrolled patients who had life-sustaining therapies withheld or withdrawn. MEASUREMENTS AND MAIN RESULTS: The number of patients, types of disorders and interventions, reasons, and timing of withholding and withdrawing life-sustaining treatments and their effect on mortality and trial results were assessed. Foregoing of life-sustaining therapies took place in 117 (22%) of 543 patients and occurred within 72 hrs of study drug administration in 38 (32%) patients. Withholding treatment (60%) was more common than withdrawing treatment (40%), but withdrawing treatment was more frequent (51%) than withholding treatment (20%) in the first 72 hrs of the trial (p < .01). Sixty-one (52%) patients had severe underlying disorders with a poor prognosis. The hospital mortality rate was 94% (of the 117 patients). The mean time (SEM) from withholding or withdrawing of treatment until death was 2.83 +/- 0.57 and 0.32 +/- 0.13 days, respectively (p < .001). Patients who had therapies foregone in the first 24, 48, and 72 hrs after receiving the study drug had higher mortality rates in the first 72 hrs (p < .01). CONCLUSIONS: A substantial number of patients enrolled in a sepsis trial had severe underlying diseases and had foregoing of therapies early in the course of the trial, which led to a higher early mortality rate. Enrollment of patients in clinical trials with severe underlying disorders with a high likelihood of having therapies foregone may bias the potential for showing the efficacy of new therapeutic modalities.

Treatment of gram-negative bacteremia and shock with human antiserum to a mutant Escherichia coli.

In an effort to decrease deaths from gram-negative bacteremia and endotoxin shock, we treated bacteremic patients with human antiserum to endotoxin (lipopolysaccharide) core. Antiserum was prepared by vaccinating healthy men with heat-killed Escherichia coli J5; this mutant lacks lipopolysaccharide oligosaccharide side chains, so that the core, which is nearly identical to that of most other gram-negative bacteria, is exposed for antibody formation. In a randomized controlled trial, patients were given either J5 antiserum or preimmune control serum intravenously, near the onset of illness. The number of deaths in the bacteremic patients was 42 of 109 (39 per cent) in controls and 23 of 103 (22 per cent) in recipients of J5 antiserum (P = 0.011). In those with profound shock, mortality was 30 of 39 (77 per cent) in controls and 18 of 41 (44 per cent) in recipients of J5 antiserum (P = 0.003). We conclude that human antiserum to the lipopolysaccharide core can substantially reduce deaths from gram-negative bacteremia.

Prevention of gram-negative shock and death in surgical patients by antibody to endotoxin core glycolipid.

The prophylactic effect of antibody to endotoxin core glycolipid was studied in surgical patients at high risk of gram-negative infection. At randomisation (on admission to intensive care unit), every 5 days thereafter, and at onset of septic shock, patients received plasma taken from donors before (control) or after immunisation with Escherichia coli J5, a mutant with only core determinants in its endotoxin. Gram-negative shock occurred in 15 of 136 controls and 6 of 126 J5 antibody recipients and related deaths in 9 of 136 and 2 of 126, respectively. J5 antibody was most effective in abdominal surgery patients, in whom shock occurred in 13 of 83 controls and 2 of 71 antibody recipients. Although antibody prophylaxis did not lower the infection rate, it prevented the serious consequences of gram-negative infections and thus improved the overall prognosis.