Geoffrey K. Richards

The influence of oral versus parenteral preoperative metronidazole on sepsis following colon surgery.

The incidence of wound infection following surgery on the colon is reduced by the preoperative administration of appropriate antibiotics. Quantitative bacteriologic studies raise the fundamental question as to whether effective antibiotic prophylaxis results from reduction of the bacterial content of the gut prior to surgery or whether effective tissue levels of the antibiotic is the key factor. Oral neomycin and metronidazole have been shown to markedly reduce the incidence of wound infection following colon surgery. A prospective randomized double-blind clinical trial was undertaken to compare the effectiveness of intravenous metronidazole (high tissue level) with oral administration (tissue and gut activity) on the incidence of wound infection. There was no difference in wound infection rates between the two groups of patients. Surprisingly, there was a significant reduction in the bacteroides content in the colon of patients who received intravenous metronidazole one hour before operation to a level almost equal to that achieved by the administration of the drug for two days by mouth before operation. Metronidazole levels in the colon at the time of surgery were comparable for both groups. The median time for recolonization of the colon was six days for the oral group, and four days for the intravenous group. Although peritoneal fluid contained significant numbers of coliforms and enterococci, clinical infection did not occur. These data suggest that systemic antibiotics effective against anerobic flora of the colon markedly reduce postoperative septic complications.

The effect of metronidazole on the incidence of postoperative wound infection in elective colon surgery

A prospective randomized clinical trial assessing the relative effectiveness of erythromycin-neomycin and metronidazole-neomycin as a preoperative bowel preparation was carried out. Bacteriologic studies of feces and colon content revealed no significant difference in the reduction of aerobic bacteria between the two groups. There was, however, a significantly greater reduction in anaerobic bacteria in the feces and colon contents of patients receiving metronidazole. Wound infection rate was 25% in the erythromycin group, and organisms recovered from the wound in all cases were fecal in nature. Two wound infections occurred in the metronidazole group (5%) and in both cases the organisms recovered were staphylococci of presumed skin origin. These studies suggest that anaerobic bacteria are the major contributors to wound infection after colon surgery and that their specific reduction is associated with a lower incidence of wound infection.

Increased incidence of experimental colon cancer associated with long-term metronidazole therapy

Using the well-established DMH model for colon neoplasia, we demonstrated that a high-fiber diet pair-fed to animals was associated both with certain changes in bacterial profile and with protection against experimental colon neoplasia. The addition of metronidazole on a long-term basis to both high- and low-fiber diets did not alter stool bacteroides counts as expected and was associated with an apparent cocarcinogenic effect. Concern exists among surgeons and gastroenterologists as to whether metronidazole places their patients at risk. The status of long-term metronidazole therapy for patients with Crohns disease is a pertinent example. In view of our findings, it is important to further elucidate the metabolism of metronidazole in both the rat and human gut.

Experimental Staphylococcus epidermidis implant infection in the mouse. Kinetics of rifampin and vancomycin action.

Staphylococcus epidermidis implant infections remain a therapeutic challenge; they frequently result in failure of conservative management and require removal of the implant. This resistance to antibiotic therapy has been ascribed to the presence of a protective bacterial biofilm at the surface of the implant. An animal model of implant associated infection has been developed in which preformed bacterial biofilm catheter segments are implanted intraperitoneally, resulting in a chronic standardized localized infection. The authors have determined the superior rapid action of rifampin (Cieba-Geigy) compared to vancomycin hydrochloride (Eli Lilly) and determined that the combination is superior to either used alone. No rifampin resistant surviving infection was noted, which indicates the significant contribution of mammalian host defenses. This animal model is an excellent vehicle for the study of Staphylococcus epidermidis implant associated infection and the evaluation of the relative efficacy of antibiotic regimens, singly and in combination.

Hemodynamic responses to sepsis: Hypodynamic versus hyperdynamic states

Abstract It has never been fully understood why the circulatory responses of some septic patients and experimental septic animal models are hyperdynamic, while others are hypodynamic, the latter often thought to reflect an “overwhelming” sepsis. This study identifies the “focus of infection” as the central factor which governs the host response to sepsis. Similar hosts (piglets) received the same bacteria (Escherichia coli strain U9-41) in comparable doses, by two different routes, one intravenous and the other intramuscular. The intravenous group did not have a focus of infection and developed hypodynamic shock (low cardiac output, hypothermia, leukopenia). The low flow state was not preceded by a high output phase and was not reversed by increasing the preload, indicating myocardial depression. The intramuscular group, with a focus of infection and inflammation, developed a hyperdynamic state (high cardiac output, fever, and leukocytosis). The important pathophysiologic role of a focus of infection should be recognized, particularly in devising experimental models to study septic shock.

Response of Chronic Renal Failure Mice to Peritoneal Staphylococcus Epidermidis Challenge: Impact of Repeated Peritoneal Instillation of Dialysis Solution

The effect of repeated instillation of peritoneal dialysis (PD) solution on the peritoneal clearance of a Staphylococcus epidermidis challenge was investigated in a mouse model of surgically induced chronic renal failure. For periods of up to 2 weeks, mice bearing peritoneal catheter implants underwent daily (3 mL) or twice daily (1.5 mL) peritoneal instillation of PD solution (4.25% dextrose) by transcutaneous injection into the catheter lumen. Peritoneal instillation of PD solution did not have a significant influence on the microbiological status of peritoneal structures of renal failure or sham-operated mice following experimental intracatheter S epidermidis inoculation with 10(6) colony-forming units (CFU) (assessment 48 hours after inoculation) or 10(8) CFU (assessment 1 week after inoculation). Microbiological and scanning electron microscopy (SEM) assessments of recovered peritoneal catheters demonstrated that S epidermidis remained associated with the catheter site after other peritoneal structures had become culture negative. SEM of the parietal peritoneum revealed striking morphologic alterations of the mesothelial surface as a consequence of daily PD solution infusion. In the absence of S epidermidis inoculation, repeated instillation of PD solution caused a marked acute peritoneal inflammation without evidence of a concomitant systemic inflammatory response. Furthermore, peritoneal inflammatory response to S epidermidis challenge was augmented by the infusion procedure. Concurrent assessments of inflammatory response and microbiological status revealed that, in spite of heightened peritoneal inflammatory response with peritoneal infusion, bacterial clearance from the catheter site was not improved. Although the animal preparation was limited to peritoneal infusion without drainage, the influence of repeated peritoneal instillation of hyperosmolar, acidic PD solution on the response of mice to S epidermidis challenge was successfully addressed.

Mechanical and Cellular Bacterial Clearance in Lung Atelectasis

A pig model with aerosolized pneumococcal bacteria was used to establish that bacterial clearance in a collapsed lung in the perioperative period was decreased compared with the opposite, aerated lung. Cannulation of the right lymphatic duct revealed a significant increase in both lymph flow and ratio of lymph to plasma protein, indicating the development of a high-permeability edema in the collapsed, infected lung. Only 22% of the efferent lymph and blood was positive for the infecting organism. Examination of the T cells obtained by bronchopulmonary lavage showed an initial fall in the numbers of alveolar macrophages at 6 hours after collapse and infection, relative to the opposite, control lung. However, at 24 hours, the collapsed lung had replenished its alveolar macrophage population to such a degree that it was greater than the control. Electron microscopy revealed that the macrophages in the collapsed lung were more activated with increased lysosomal and pseudopodial activity. The in vitro chemotactic function of the macrophages appeared depressed, but phagocytosis and intracellular bactericidal activity were increased in the atelectatic lung. We conclude that there is a decreased bacterial clearance capacity in atelectasis. This finding indicates that impaired mucociliary clearance plays the dominant role in susceptibility to infection.

An assay to measure antibiotic efficacy against Staphylococcus epidermidis biofilms on implant surfaces.

Infection is a major limitation of implantable devices. Optimal antibiotic therapeutic regimes have not yet been defined. Implant-associated infections have a number of differentiating characteristics, which include the predominance of Staphylococcus epidermidis and other skin bacteria of normally low pathogenicity as the causative agents, together with a relative resistance to host defenses and to antibiotic therapy. These properties have been ascribed to the ability of the bacteria to exist on implant surfaces in the biofilm phase, which is protective. An assay of antibiotic activity using a standardized bacterial biofilm preparation of S. epidermidis is described. The assay is used to evaluate the relative efficacy of antibiotics to sterilize the biofilm, when they are used singly, or in double or triple combinations. The modulating effects of changing antibiotic concentrations and modifying the environment with CAPD variables (fresh and spent dialysis fluid, common PD solution additives) are also measured and the data summarized. It is hoped that, by using this and similar assays, individualized optimal therapeutic regimes of implant-associated infections may be logically planned.

Role of an Intraperitoneal Catheter Implant in the Pathogenesis of Experimental Staphylococcus epidermidis Peritoneal Infection in Renal Failure Mice

The influence of a permanent peritoneal catheter implant on the response of renal failure and control mice to peritoneal inoculation with 10(6) colony-forming units (CFU) Staphylococcus epidermidis was assessed 48 h after bacterial challenge. Two weeks after the surgical induction of renal failure or sham surgery, a segment of a peritoneal dialysis catheter was implanted entirely within the confines of the peritoneal cavity of mice. One month later peritoneal S. epidermidis inoculation was performed by transcutaneous injection through the abdominal wall either directly into the peritoneal cavity (i.p.) or via the catheter lumen (i.c.). Following i.p. inoculation, minimal bacterial growth was recovered from the peritoneal structures of all mice, including the peritoneal catheter. In contrast, following i.c. S. epidermidis challenge, the catheter site remained heavily colonized while peritoneal washings and parietal peritoneum again presented minimal bacterial recoveries. S. epidermidis recovery from the catheter site of renal failure mice was significantly greater than from the respective site of sham-operated controls. Scanning electron microscopy of catheter segments recovered from mice following i.c. inoculation revealed single cocci or microcolonies associated with the catheter surface and differential leukocyte counts of fluid aspirated from the catheter lumen revealed evidence of acute inflammation. Signs of inflammatory processes in peritoneal washings and peripheral blood, however, were not observed. These results are discussed in relation to S. epidermidis peritonitis and continuous ambulatory peritoneal dialysis.