A reply

Abstract

We thank van Eijk et al. for their interest in our work and appreciate their congratulation for the first investigation of vascular reactivity in septic patients [1]. Despite some evidence in favour of endothelial dysfunction during experimental sepsis, we showed that in septic shock, acetylcholine-induced vasodilation in the forearm vasculature is well preserved, albeit in a small number of patients [1]. Van Eijk et al. remark that prior studies demonstrate that in septic animals or during experimental endotoxaemia, endothelial function is impaired as indicated by decreased vasodilation in response to acetylcholine in several organ systems [2-4] and feel that this contrasts with our observations in septic patients [1]. We would like to suggest that observations made in animal models of ‘sepsis’ or during experimental endotoxaemia in human volunteers, may not reflect the complex changes seen in patients with sepsis. In fact, this was the very reason why we decided to study vascular reactivity in septic patients so as to overcome the deficiencies of several sepsis models. By studying patients with similar cardiovascular instability we thought to be able to minimize inter-individual variability in vascular function as much as possible. In fact, in patients with sepsis, acetylcholine induced endothelium-dependent relaxation was found to be intact in the skin compared to ICU patients without sepsis as well as to healthy volunteers [5]. Accordingly, preserved endothelium-dependent vasodilation in the forearm vasculature of our patients with septic shock parallels previous findings in skin microcirculation during sepsis. Of note, however, we did not study endothelial dysfunction during an early stage of sepsis. Van Eijk also suggests that lower vascular resistance in sepsis may be the reason for the decreased response to acetylcholine observed. If this were the case, one would expect impaired vasodilation to acetylcholine due to a decreased vasodilator reserve. In contrast, at least in the forearm vasculature, we observed increased blood flow in patients with septic shock but at the same time a similar response to the exogenous nitric oxide-donator nitroprusside [1]. Thus, vasodilator reserve appears to be maintained in forearm vasculature during human septic shock. Finally, van Eijk comments that a full dose-response relationship of acetylcholine cannot be characterized by the two doses administered (a point we have already addressed in our publication) and that the dose of L-N monomethylarginine given is rather low. We fully agree, but can report that blood flow measurements were indeed performed in the contralateral arm concurrently. These indicated that higher doses of acetylcholine or L-N monomethylarginine resulted in systemic effects in both patients and volunteers. Since we did not want to confuse measurements of regional effects in the forearm by systemic baroreflex mediated effects we limited interpretation and data to the lower dosages reported. Van Eijk speculates that endothelial dysfunction may have been demonstrated following administration of other endothelium-dependent vasodilators, like substance P or bradykinin. Since we studied endothelium-dependent and endothelium-independent vasodilation as well as vasoconstriction using three different vasoconstrictors in the same subjects/patients we had to limit the duration of the study to maintain stable baseline conditions and dispensed the evaluation of additional vasoactive drugs. Obviously, what effect other agents might have had is an area for speculation. In summary, we challenge prior work in animal models of sepsis and in human volunteers following lipopolysaccharide administration by demonstrating in ‘real sepsis’ that endothelial function is preserved in the forearm vasculature of septic patients. While endothelial dysfunction may still occur during earlier stages of sepsis and further studies are needed to address the issue there is nothing that beats data in real, septic humans.