Ann. Medit. Burns Club - vol. VII - n. 2 - June 1994

SELECTIVE INTESTINAL DECONTAMINATION OF THE DIGESTIVE TRACT: A TOOL FOR INFECTION PROPHYLAXIS IN BURNS?

Public Health Laboratory and Burn Unit, Martini Ziekenhuis, Groningen Burn Unit, Martini Ziekenhuis, Groningen, The Netherlands

SUMMARY. In the last few years much attention has been paid to the role of the gastrointestinal tract (GIT) in the pathogenesis of infectious complications in severely burned patients. There is considerable evidence that micro-organisms originating from the GIT play a role in the colonization of burn wounds. There may also be a direct dissemination of microbes from the GIT by means of bacterial translocation. Promising results have been obtained from recent studies in which selective decontamination of the digestive tract was used to prevent infections in burns.

In the past, invasive microbial infection greatly hindered survival after major burns. Although survival from thermal injuries has significantly improved in the last decades, infection remains a leading cause of mortality in burn patients. It has been estimated that infections account for approximately 50-80% of all deaths occurring after the initial shock phase (2).
Necrotic tissue provides a rich culture medium for rnicro-organisms. The avascularity of the eschar reduces the capacity of the patient's host defence mechanisms to attack potential invading micro-organisms and largely prevents systemically administered antimicrobial agents from penetrating into the devitalized tissue. Early attention to the burn wound, with removal as promptly as possible of the eschar, together with other infection prevention measures, has significantly decreased the incidence of wound infections. Yet, despite all the improvements, burn wounds become colonized with micro-organisms a few days after the injury.
The surface of the burn wound becomes colonized after the second day post-burn. Gram-posifive micro-organisms, particularly Staphylococcus aureus and coagulase-negative staphylococci, will colonize the burn wounds during the first week. Probably these micro-organisms arise from the depth of the skin. Later, partly depending on the extent of the lesion, aerobic gram-negative bacteria predominate. In the beginning these are mostly enteric gram-negative bacilIi, such as Escherichia coli, Klebsiella and Proteus. After several weeks there may be a second shift in the wound microflora and burn wounds may be colonized by Pseudomonas aeruginosa (5). The incidence of Pseudomonas colonization varies from one hospital unit to another and other nosocomial pathogens (Providencia species, Serratia marcescens, Acinetobacter species) may be found, largely depending on the local situation. Yeast and fungi may also colonize the burn wounds.
Candida species are the most frequently isolated species, but other fungi like Aspergillus, Altenaria and Fusarium may occasionally be cultured from the wound surface.
Most of the above-mentioned micro-organisms are normal inhabitants in the human host. Staphylococci are found on the surface of the intact skin, and S. aureus and gram-negatives are harboured in the upper respiratory and the gastrointestinal tract (GIT). Although aerobic gramnegative rods and yeasts are isolated from the oropharynx in only a minority of healthy persons, the presence of these microbes in the oropharynx has increased in patients and may be further enhanced by antimicrobial therapy and the presence of an intratracheal cannula.
It is thus reasonable to suppose that burn wounds are initially colonized with the patient's own endogenous microflora. Several studies have considered this close relationship. Brook and Randolph investigated the aerobic and anaerobic microflora of burns in children (1). They found a close relationship between bacteria isolated from the upper part of the body and bacteria originating in the oropharynx. A similar relationship was found between the microflora of the lower body part and the faeces. Also with regard to enteric gram-negatives, wounds from the anal and oral region tended to yield micro-organisms respectively found in the stool or mouth flora (9). Candida species were more often isolated from wounds in patients harbouring these micro-organisms in the GIT (4).
Strict hygienic isolation precautions in fact suppressed colonization of the wound by exogenous micro-organisms, but had no effect on the colonization of wounds by endogenous micro-organisms.
Thus, although the relationship between the GIT and burn microflora has been established, little attention has been paid to possibilities to prevent this endogenous contamination.

The GIT, from the oropharynx to the rectum, harbours a complex and unique ecological microflora. The human faecal content is about 1011 micro-organisms per gm of stool, with about 200-400 different species. Most of these strains - more than 99% - are anaerobic bacteria, which are seldom pathogenic. Aerobic gram-negatives, and to a lesser extent yeasts, are potential pathogenic micro-organisms (PPMO) causing the majority of nosocomial infections. There are normally about 101 enteric gram-negatives per gm of fecal content. Anaerobes maintain what is known as colonization resistance (CR), by which is meant that they provide a hostile environment that suppresses the growth of colonizing and newly ingested micro-organisms. In patients with immunodeficiencies PPMO may cause infections, particularly if CR is decreased by the administration of antibiotics which suppress anaerobic microflora.
By means of selective decontamination of the digestive tract (SDD), PPMO are eliminated from the GIT by means of oral antimicrobial agents which preserve indigenous anaerobic microflora and thus maintain CR. SDD has been used in different clinical situations: patients with granulocytopenia, patients in intensive care units (ICU) and patients undergoing transplantations (bone marrow and liver transplants). With the exception of patients with granulocytopenia - in which SDD is accepted by many clinicians - the beneficial effects of SDD in other groups are debatable. Although SDD used in the ICU setting reduced the incidence of infection and mortality related to infectious complications, neither overall mortality nor the length of stay in the ICU was significantly influenced by SDD (10).
The relationship between GIT and burn wound microflora suggests the use of SDD in extensive burn victims in order to control wound colonization and to prevent additional infection.

SDD may thus be useful in severely burned patients, starting from the principle that wound infection - and subsequent bacterial infection - will be caused by micro-organisms originating in the GIT.
In the 1970s Jarrett et al. (3) investigated the use of total intestinal decontamination in burns. Using an oral antimicrobial regimen of gentamycin, erythromycin and nystatin, they showed that there was a delay of wound colonization in treated patients compared to the control group.
They also found a lower infection rate and a lower mean duration of hospital stay, although these two findings did not reach the level of statistical significance.
In an experimental study we investigated the role of SDD - by means of oral administration of azthreonam - in burned mice (7). Aztbreonarn is a non-adsorbable monolactam with an antibacterial spectrum that includes nearly all clinically relevant aerobic gram-negative rods but excludes anaerobic micro-organisms. Administration of azthreonam significantly reduced the incidence of burn wound colonization in treated mice compared to a control group over a 20-day period.
Mackie et al. (6) investigated the use of SDD in patients with extensive burns. This was a prospective study in a control group of 33 patients admitted to their unit in the two years immediately preceding the SDD period. The two groups were comparable as regards age, severity of burn injury and incidence of inhalation injury. In the study an oral regimen of polymyxin, tobramycin and amphotericin B was used, together with a short course of systemic antiniicrobials in the first few days. The results showed a significant reduction of gram-negative rods (Pseudomonas aeruginosa and Enterobacteriaceae) cultured from the wounds in SDD treated patients. They also found a decrease in lower respiratory tract infections (6.5% compared to 27%), and septicaernia (3% compared to 24%) as well as a decrease in the use of systemic administered antimicrobial agents (48% compared to 79%) in the SDD group compared to the control group. The study also showed a reduction in fever and overall mortality (3 % compared to 21 %).
In a retrospective study we compared two SDD regimens in burn patients (8). A regimen containing polymyxin B only was compared to a regimen containing polymyxin B, co-trimoxazole and amphotericin B.
All patients with burns in at least 25% TBSA and at least 10% full-thickness area were included in this study. In both groups all patients were treated according to a standard which included infusion of hypertonic salt solution, a diet high in proteins and calories (always given orally) and daily application on the burn wounds of silver sulphadiazine creams or similar ointments. Patients received the oral SDD antimicrobials on entering the burn unit, nearly always within 24 hours after the accident. SDD was continued until the last operative procedure had been performed.
The triple regimen resulted in better eradication of aerobic gram-negatives from the GIT., Proteus in particular, a micro-organism resistant to polymyxin B, was isolated from the GIT in 50% of the patients treated with polymyxin alone. Compared to the polyrnyxin-treated group the triple regimen group resulted in a reduction of wound colonization by Enterobacteriaceae from 74 to 13% (p<0.005). Colonization by Proteus was eliminated in patients treated with co-trimoxazole, compared to an incidence of 30% in the group treated with polymyxin alone (p<0.001). The addition of amphotericin B to the regimen resulted in a decrease of wound colonization by yeasts from 39 to 10%. In nearly all cases the causative micro-organisms could be detected in throat and/or rectum cultures before it colonized the wound; this was the case in 75% of the aerobic gramnegatives and in 100% of the yeasts. These results suggest that SDD is an effective method for the prevention of wound colonization.
In the last 1.2 years 142 patients with more than 25% T`BSA (mean TBSA 42%, mean full-thickness area 31%) have been treated with SDD. Sixteen patients (11%) died, but mortality was related to infectious complications in only two of them. Only 22 patients (15%) developed a bacterial infection (seven with bacteraernia).
SDD not only prevented wound colonization by potential pathogenic micro-organisms but also delayed wound colonization. This last fact is important, because in patients the immunological defence mechanisms, which are suppressed in the first few weeks after the injury, may recover from the initial suppression later in the course of treatment. In this case patients have a better defence mechanism for the control of invading micro-organisms later in the treatment period.
In the last few years attention has been paid to the role of the GIT in the promotion of bacterial dissemination by means of bacterial translocation (11).
Bacterial translocation has been defined as the passage of viable and non-viable micro-organisms and microbial products, such as endotoxin, from the bowel lumen through the intestinal mucosa to the mesenteric lymph nodes and other visceral organs.
Several animal studies indicate that the GIT may serve as a portal of entry for bacteria and bacterial toxins. In human beings altered intestinal permeability has been determined in severely burned patients. There is evidence that dissemination of micro-organisms from the intestinal lumen might occur in these patients. However the clinical significance of the increased permeability and of the ability to inhibit translocation by means of SDI) remains speculative.

Recently, attention has been focused on the role of the GIT in the development of infectious complications in burns. The few studies undertaken to investigate the effects of SDI) on the prevention of bacterial colonization and infections appear to be promising. However, the studies are not complete enough to permit definitive conclusions to be drawn.

RESUME. Depuis bien des années le rôle de l'appareil gastro-intestinal (GI) a été étudié pour ce qui concerne la pathogénie des complications infectieuses des grands brûlés. Il y a beaucoup de preuves que les micro-organismes qui dérivent de l'appareil GI jouent un rôle dans la colonisation des brûlures. Les microbes peuvent aussi se disséminer de l'appareil GI moyennant la translocation bactérienne. Des études récentes, qui ont utilisé la décontamination sélective de l'appareil GI pour prévenir les infections dans les brûlures, ont donné de bons résultats.

1. Brook I., Randolph J.G.: Aerobic and anaerobic bacterial flora of burns in children. J. Trauma, 21: 313-8, 1981.
2. Goodwin CW., Yurt R.W.: Epidemiology of burn wounds. In: Galling J.1., Fauci A.S. (Eds): "Advances in host defence mechanisms", Vol. 6, 5-18, Raven Press, New York, 1986.
3. Jarrett F., Balish E., Moylan J.A., Ellerbe S.: Clinical experience with prophylactic antibiotic bowel suppression in burn patients. Surgery, 83: 523-7, 1978.
4. Kidson A., Lowbury E.L.: Candida infections of bums. Burns, 6: 2-7, 1979
5. Lawrence J.C.: The bacteriology of burns. J. Hosp. Infect., 6 (Suppl. B): 3-17, 1985.
6. Mackie D.P., Van Herturn W.A.J., Schumburg T.: Prevention of infection in bums; preliminary experience with selective intestinal decontamination of the digestive tract in patients with extensive injuries. J. Trauma, 32: 570-3, 1992.
7. Manson W.L., Dijkema H., Klasen H.J.: Alteration of wound colonization by selective intestinal decontamination in thermally injured mice. Bums, 16: 166-8, 1990.
8. Manson W.L., Klasen H.J., Sauer E.W., Olieman A.: Selective intestinal decontamination for prevention of wound colonization in severely'burned patients: a retrospective analysis. Bums, 18: 98102,1992.
9. Van Saene H.K.F., Nicolai J.P.A.: Prevention of wound infections in bums. Scand. J. Plas. Reconstr. Surg., 13: 63-7, 1979.
10. Van Der Waaij D., Manson W.L., Arends J.P., De Vries-Hospers H.G.: Clinical use of selective intestinal decontamination: the concept. Intensive Care Med., 16 (S3): 212-5, 1990.
11. Wells L.L., Maddaus M.A., Simmons R.L.: Proposed mechanisms for the translocation of intestinal bacteria. Rev. Infect. Dis., 10: 95879,1988.