Abstract

Background. Gut ischaemia may contribute to morbidity in sepsis, but little is known about the metabolic state of the gut mucosa in such patients.

Methods. Nine patients with abdominal septic shock treated with norepinephrine, and ten healthy subjects, were subjected to equilibrium dialysis with a rectal balloon. pH, Pco2 and concentrations of l‐lactate were measured by auto‐analyser.

Results. In rectal dialysis fluid from patients with septic shock, acidosis was present (pH 7.23, 95% CI 7.11–7.36) and concentrations of l‐lactate were approximately five times greater than controls (2.5–5.8 vs 0.5–1.2 mmol litre–1). The lactate concentration was related to the dose of norepinephrine (P<0.001). In contrast, values of dialysate Pco2 did not differ significantly between patients and controls (6.4–11.0 vs 8.9–13.8 kPa).

Conclusions. The results suggest that, either lactic acidosis in rectal mucosa is related to shock severity, or that norepinephrine causes mucosal ischaemia. In any case, metabolic dysfunction is present in the rectal mucosa in patients with abdominal septic shock treated with norepinephrine.

Br J Anaesth 2002; 89: 919–22

Accepted for publication: July 21, 2002

In septic shock, gut ischaemia may contribute to morbidity and this feature may be remediable in septic patients.1 However, very little is known about the metabolism of the gut in human septic shock.

We set out to measure pH, Pco2 and concentrations of l‐lactate in the gut of patients with septic shock and healthy subjects. Luminal equilibrium dialysis is a valid, non‐invasive method for the estimation of extracellular concentrations of small molecules (<12 kDa) in rectal mucosa.2 Animal studies show that measurement of mucosal lactate by luminal microdialysis is a valid method by which to assess intestinal ischaemia.3

Methods and results

Participants

After approval by the regional ethics committee and with informed written consent, we studied nine patients with peritonitis (six had intestinal perforation and three acute pancreatitis) and established septic shock (>24 h), and ten healthy volunteers.

Septic shock was defined as the presence of a positive bacterial culture from the blood or the peritoneal cavity and the need for infusion of norepinephrine (>0.04 µg kg–1 min–1) to maintain a mean arterial pressure (MAP) >70 mm Hg. We did not include patients with any of the following conditions: (i) any changes in cardiovascular treatment in the preceding 2 h, (ii) systemic hypoxia (PaO2<8 kPa), (iii) abnormal rectum or left colon, (iv) active gastrointestinal bleeding, or (v) an intra‐abdominal pressure >20 mm Hg. All patients had been resuscitated with i.v. fluids before the study using Dextran 60 (Macrodex®; Pharmacia‐Upjohn, Uppsala, Sweden) and isotonic saline until the arterial pressure was stable, after which another 1000–2000 ml of fluid were given. Blood was given if the haemoglobin concentration was <5 g dl–1. If oliguria was present, dopamine was infused at 4 µg kg–1 min–1 (four patients). All patients received systemic antibiotics, selective digestive decontamination (except three patients) and enteral nutrition (except one). Three patients had a pulmonary artery catheter in place during the study. All of these patients had a cardiac index >3 litre min–1 m–2, a pulmonary capillary wedge pressure >16 mm Hg and a systemic vascular resistance index <1000 dyn s cm–5 m–2.

Luminal equilibrium dialysis

pH, Pco2 and concentrations of l‐lactate in rectal mucosa were measured by luminal equilibrium dialysis as previously described.2 Bags of dialysis tubing (semipermeable cellulose, cut‐off 12 kDa; Sigma, St Louis, MO, USA) were closed over 5 cm of Tygon® tube (Cole‐Parmer Instruments Company, Vernon Hills, IL, USA) with a three‐way stopcock at the distal end for airtight sampling. Bags were filled with 4 ml of Dextran 40 10% in isotonic saline (Rheomacrodex®; Pharmacia‐Upjohn) and placed in the rectal lumen for 4 h, which was the time required for 100% equilibrium of eicosanoids in vivo.2 Incubation of dialysis bags for 2 h in a saline bath containing 1 mmol litre–1 l‐lactate (Sigma) at 37°C was sufficient for 100% equilibrium of l‐lactate. pH, Pco2 and l‐lactate were measured by auto‐analyser (ABL 625, Radiometer, Copenhagen, Denmark). As d‐lactate may be produced by luminal bacteria, saline solutions containing 1, 10 and 100 mmol litre–1 d‐lactate (Sigma) were analysed in the ABL 625, but d‐lactate was undetectable (n=3).

Statistics

Normal distribution of the variables was tested with the Kolmogorow–Smirnov statistics and the Levine test was used to test for equal variance. Data were analysed by Student’s t‐test for unpaired or paired variables, or linear regression analysis where appropriate.4P‐values <0.05 (2‐tailed) were considered significant.

Results

In dialysis fluid from the patients with septic shock, acidosis and increased concentrations of l‐lactate were observed compared with values in fluid from healthy subjects (Table 1). In septic patients, the mean concentration of l‐lactate in dialysis fluid was 70% greater than in arterial blood (P=0.03, Table 1), and there was no correlation between the two values (P=0.27). In contrast, dialysate concentrations of l‐lactate correlated with the dose of nor epinephrine (Fig. 1, r2=0.89; P=0.0001). The dose of norepinephrine was not related to systemic values of l‐lactate (P=0.48).

Comment

Our results show lactic acidosis in the rectal lumen of patients with abdominal septic shock. Lactic acidosis may represent anaerobic glycolysis from hypoperfusion of rectal mucosa. Although sympatomimetics can cause systemic lactic acidosis through aerobic glycolysis, this has not been seen during infusion of norepinephrine,5 as was used in the present study. In support of this, systemic values of l‐lactate were not related to norepinephrine dose in our patients, but the regression analyses of the present study have a high risk of a type II error. In addition to being markers of metabolism, acidosis6 as well as lactate7 can cause cellular dysfunction leading to increased intestinal permeability and increased morbidity in septic shock.

Rectal equilibrium dialysis may be a simple, non‐invasive method to measure markers of metabolism in gut mucosa in critically ill patients. The 4 h of equilibration used in the present study may hamper its clinical use. However, it is possible that equilibrium can be obtained earlier or that non‐equilibrium dialysis with a shorter time of exposure can detect clinically relevant differences. Future studies should address these questions as well as the effects of age, dopamine, antibiotics, enteral nutrition and fluid management.

The high values of Pco2 observed in the rectal lumen of both patients and healthy subjects suggest that carbon dioxide may come from bacterial metabolism. Alternatively, weak acids in faeces may be buffered by HCO3, which is secreted into the lumen by rectal epithelial cells. Consequently, any change in bacterial number or metabolism or epithelial secretion of HCO3 could affect Pco2, pH, or both and complicate the interpretation of these values. In the gastric mucosa, an increased luminal–arterial Pco2 gap may indicate regional hypoperfusion. We did not determine arterial Pco2 in the healthy subjects, so comparison between the groups is not possible. Bacterial metabolism may also generate d‐lactate, but this was not detectable by the auto‐analyser used in the present study. This suggests that bacterial metabolism did not contribute to the high dialysate lactate observed in septic patients.

The observed relationship between dialysate l‐lactate and norepinephrine dose suggests that mucosal l‐lactate is related to shock severity, which may be indicated by the dose of norepinephrine. Alternatively, norepinephrine treatment may itself cause mucosal ischaemia. In any case, metabolic dysfunction is present at the rectal–mucosal barrier in patients with abdominal septic shock treated with norepinephrine.

Fig 1 Relationship between rectal dialysis fluid concentration of l‐lactate and the dose of norepinephrine in patients with abdominal septic shock. r2=0.89; P=0.0001 by linear regression analysis.

Fig 1 Relationship between rectal dialysis fluid concentration of l‐lactate and the dose of norepinephrine in patients with abdominal septic shock. r2=0.89; P=0.0001 by linear regression analysis.

Table 1

Characteristics and metabolic values in arterial blood and rectal dialysis fluid in patients with abdominal septic shock and healthy subjects. Data are mean (sd or range) or number of patients (%). *Values from seven patients, as air‐contamination occurred in dialysates from two patients

 Septic patients  Healthy subjects  t‐test 
 (n=9) (n=10) P 
Age (yr) 56 (38–72) 26 (23–32) 0.0001 
Mortality (%) 45 –  
SAPS II 43 (9) –  
MAP (mm Hg) 80 (9) –  
HR (beat min–1103 (19) –  
Arterial blood    
 pH 7.34 (0.05) –  
Pco2 (kPa) 5.8 (0.8) –  
Po2 (kPa) 11.9 (1.4) –  
l‐lactate (mmol litre–12.4 (1.3) –  
Rectal dialysate    
 pH 7.23 (0.15)* 7.46 (0.15) 0.01 
Pco2 (kPa) 8.7 (2.8)* 1.3 (3.2) 0.34 
l‐lactate (mmol litre–14.1 (2.2) 0.8 (0.5) 0.0002 
 Septic patients  Healthy subjects  t‐test 
 (n=9) (n=10) P 
Age (yr) 56 (38–72) 26 (23–32) 0.0001 
Mortality (%) 45 –  
SAPS II 43 (9) –  
MAP (mm Hg) 80 (9) –  
HR (beat min–1103 (19) –  
Arterial blood    
 pH 7.34 (0.05) –  
Pco2 (kPa) 5.8 (0.8) –  
Po2 (kPa) 11.9 (1.4) –  
l‐lactate (mmol litre–12.4 (1.3) –  
Rectal dialysate    
 pH 7.23 (0.15)* 7.46 (0.15) 0.01 
Pco2 (kPa) 8.7 (2.8)* 1.3 (3.2) 0.34 
l‐lactate (mmol litre–14.1 (2.2) 0.8 (0.5) 0.0002 

HR=heart rate; MAP=mean arterial pressure; SAPS II=simplified acute physiology score II.

References

1
Landow L, Andersen LW. Splanchnic ischaemia and its role in multiple organ failure.
Acta Anaesthesiol Scand
 
1994
;
38
:
629
–39
2
Lauritsen K, Laursen LS, Bukhave K, Rask‐Madsen J. In vivo profiles of eicosanoids in ulcerative colitis, Crohn’s colitis, and Clostridium difficile colitis.
Gastroenterology
 
1988
;
95
:
11
–17
3
Tenhunen JJ, Kosunen H, Alhava E, Tuomisto L, Takala JA. Intestinal luminal microdialysis: a new approach to assess gut mucosal ischemia.
Anesthesiology
 
1999
;
91
:
1807
–15
4
Kuo J, Fox E. Sigma Stat® Manual. Microcomputer Tools for Scientist. Revision SSD‐1.0. Augsburg: Jandel Scientific, 1992
5
Totaro RJ, Raper RF. Epinephrine‐induced lactic acidosis following cardiopulmonary bypass.
Crit Care Med
 
1997
;
25
:
1693
–9
6
Menconi MJ, Salzman AL, Unno N, et al. Acidosis induces hyperpermeability in Caco‐2BBe cultured intestinal epithelial monolayers.
Am J Physiol
 
1997
;
272
:
G1007
–21
7
Plum J, Razeghi P, Lordnejad RM, et al. Peritoneal dialysis fluids with a physiologic pH based on either lactate or bicarbonate buffer‐effects on human mesothelial cells.
Am J Kidney Dis
 
2001
;
38
:
867
–75

Author notes

1Department of Anaesthesia and Intensive Care, Herlev Hospital, University of Copenhagen, DK‐2730 Herlev, Denmark. 2Department of Anaesthesia and Intensive Care, Rigshospitalet, University of Copenhagen, DK‐2100 Copenhagen, Denmark

Comments

0 Comments