both in vitro and in vivo, in order to char-acterize the method and its limitations [1,3, 4, 5]. Furthermore, we have always usedit under appropriately controlled experi-mental conditions. It is therefore clear thatwhile we can make no claims concerningthe absolute luminal lactate concentrations,the changes observed in cardiac tamponademodel are, indeed, still valid. It is highlypossible that luminal microdialysate lactateconcentrations were lower than the ‘real’concentration in the surrounding milieu. Inother words, the ‘recovery’ was not 100%in the gut lumen as opposed to withinblood stream validation where the relativerecovery was close to 100% [1]. Low re-covery is now easily overcome with lowerdialysate flow rate (0.2 µl/min) and highlydeveloped analyzers which are specificallydesigned for microdialysis. The samplevolumes needed may be as low as 2–4 µlfor lactate and still the sampling frequencyremains reasonably short for both researchand clinical purposes (10–20 min) as op-posed to 4 h of equilibration in the paperby Due and Perner. They quite correctlydiscuss the problems related to long equili-bration time in a dynamic disease state.

We compared our results from the car-diac tamponade study to our earlier experi-ments, where selective occlusion of the su-perior mesenteric artery (SMA) was ap-plied [1, 4]. At comparably low SMAblood flow levels with the identical micro-dialysis method, intestinal luminal lactatewas markedly higher in selective intestinalischemia as opposed to low systemic bloodflow. Thereby we suggested that the smallintestine is better adapted to low bloodflow during low systemic blood flow thanisolated, selectively reduced, regionalblood flow [3]. In addition, systemic hy-perlactatemia could largely explain the lu-minal lactate release observed [1, 3]. Final-ly, generally spoken, the terminology hasto be precise: recovery should be reservedto indicate, in both microdialysis and equi-librium dialysis, the percentage of concen-tration as compared to the surrounding mi-lieu. Sensitivity and specificity for cellulardysoxia, ischemia or hypoperfusion needto be defined by formal receiver operatingcurve (ROC) analysis. While we find the

result of Due and Perner intriguing indeed,we would be interested to see the valida-tion of the method including, for example,occlusive and non-occlusive ischemia andischemic and non-ischemic hyperlactate-mia under controlled conditions.

References

1. Tenhunen JJ, Kosunen H, Heino A, Tuomisto L, Alhava E, Takala J (1999)Intestinal luminal microdialysis- a newapproach to assess mucosal dysoxia.Anesthesiology 91:1807–1815

2. Due VL, Bonde J, Espersen K, JensenTH, Perner A (2002) Lactic acidosis inthe rectal lumen of patients with septicshock measured by luminal equilibriumdialysis. Br J Anaesth: (in press)

3. Tenhunen JJ, Jakob SM, Ruokonen E,Takala J (2002) Jejunal luminal microdi-alysate lactate in cardiac tamponade –effect of low systemic blood flow on gutmucosa. Intensive Care Med28:953–962

4. Tenhunen JJ, Jakob SM, Takala JA(2001) Gut luminal lactate release dur-ing gradual intestinal ischemia. Inten-sive Care Med 27:1916–1922

5. Tenhunen JJ, Uusaro A, Ruokonen E(2001) Mucosal dysoxia in endotoxinshock can be detected by luminal micro-dialysis but not by regional L/P ratios(abstract). Intensive Care Med 27:S182

J.J. Tenhunen (✉)Division of Intensive Care, Department of Anesthesiology and Intensive Care,Kuopio University Hospital,P.O. Box 1777, 70211 Kuopio, Finlande-mail: Jyrki.Tenhunen@kuh.fi

J. TakalaDepartment of Intensive Care Medicine,University Hospital of Bern,3010 Bern, Switzerland

Intensive Care Med (2003) 29:336DOI 10.1007/s00134-002-1599-5 C O R R E S P O N D E N C E

Jyrki J. TenhunenJukka Takala

Intestinal luminal lactate in shock-marker for alteredepithelial cellular metabolism?

Received: 29 October 2002Accepted: 5 November 2002Published online: 15 January 2003© Springer-Verlag 2003

Sir: We fully agree with Drs. Due and Per-ner that new methods for monitoring (in-testinal) tissue-specific metabolism in criti-cal illness are warranted. Intestinal luminalmicrodialysis [1] or dialysis membrane [2]based methods may be such approaches.

Due and Perner suggest that the rela-tively low luminal lactate concentrationsobserved by us in experimental cardiactamponade could be related to the microdi-alysis method [3]. They base their concernon their own observation on high rectallactate concentrations in patients with sep-tic shock as measured by a different rectaldialysis method [2].

We are well aware that results obtainedby microdialysis or other dialysis tech-niques are highly variable depending onthe exact method used, since multiple fac-tors affect the actual concentration of thesubstance measured. These include, for ex-ample: membrane or capillary characteris-tics, dialysate flow rate, equilibration time,the dialysate itself etc. In addition, a highlyimportant factor is the tissue per se and itsmetabolic and perfusion dynamics. We aretherefore very reluctant to draw any com-parison to our results based on the observa-tions by Due and Perner, which were ob-tained under completely different circum-stances with completely different metho-dology. As far as our method is concerned,we have performed a careful validation,