STARTING in 1997, Hahn et al. 1–4introduced and developed the concept of volume kinetics, which describes the peak effects and clearance of intravenously infused fluids in terms similar to those used in pharmacokinetics to describe the peak effects and clearance of drugs. Stanski,5in an editorial accompanying the landmark article by Svensen and Hahn in Anesthesiology in 1997,1commented that the volume kinetic approach could “allow for more rational design of intravenous fluid paradigms.” Toward that end, Hahn et al. have examined several key questions in volunteers6–8and in experimental animals.9–12One of the least expected observations, obtained in sheep, was that isoflurane anesthesia seemed to be associated with extravascular accumulation of infused crystalloid.9However, in this issue of Anesthesiology, Ewaldsson and Hahn13convincingly demonstrate that, in humans, neither isoflurane nor propofol anesthesia is associated with extravascular fluid accumulation. The authors infer from their data that volume kinetics are powerfully influenced by hypotension,13an inference that merits examination in the context of previous volume kinetic studies.
In pharmacokinetics, an exogenous substance is introduced, blood or other fluids are repeatedly sampled, and the resulting temporal pattern is analyzed to determine important kinetic variables. In contrast, volume kinetics examines the clearance of endogenous substances, e.g. , water, that already are present in considerable quantities. For such studies, an endogenous tracer is necessary, the best being the blood concentration of hemoglobin, which is an obligatory intravascular tracer. To confidently calculate volume kinetic variables, the blood concentration of hemoglobin should be repeatedly measured before, during, and after fluid infusion in a relative steady state. High-probability solutions to the kinetic equations necessitate that changes in potentially confounding physiologic and pharmacologic influences be minimized for a sufficient time interval to construct clearance curves. In practice, time intervals of 180 min after the beginning of an intravenous infusion have provided sufficient data for reliable kinetic analyses.
Such time intervals of relative stability can be achieved easily in certain types of volunteer and animal studies. For example, in volunteers, isotonic crystalloid solutions were rapidly cleared,1colloid solutions were less rapidly cleared,1and crystalloid solutions produced higher peak volume expansion and more delayed clearance in hypovolemic than normovolemic volunteers.6During intervals of relative stability in preeclamptic parturients, crystalloid solutions were more rapidly cleared than in normal volunteers.14In experimental animals, isoflurane anesthesia was associated with similar clearance of infused crystalloids from blood but markedly delayed urinary excretion, implying greater extravascular retention.9,15As in volunteers, hemorrhage in sheep both increased peak expansion and delayed clearance from blood.12,Pseudomonas bacteremia, which in sheep mimics many characteristics of clinical sepsis, unexpectedly did not influence volume kinetics.16In sheep, continuous infusion of α-adrenergic agonists dramatically accelerated, whereas β agonists delayed, clearance of infused crystalloids.17
However, the clinical circumstances of anesthesia and surgery usually preclude 180 min of steady state conditions, the influences of surgical stress and surgically induced fluid shifts are difficult to separate from the influence of anesthesia, and blood loss confounds kinetic analyses based on measurements of the blood concentration of hemoglobin. Nevertheless, volume kinetic studies have been performed in patients undergoing surgery. During laparoscopic cholecystectomy in women undergoing sevoflurane–narcotic anesthesia, induction of anesthesia, before fluid infusion, was associated with 4.2% plasma dilution (equivalent to intravascular volume expansion); subsequent fluid infusion was associated with calculated kinetic variables that were similar to those acquired in female volunteers, despite marked inhibition by anesthesia of the infusion-associated diuresis seen in volunteers.18In contrast, in men undergoing prostatectomy during enflurane anesthesia, crystalloid fluids seemed to produce greater volume expansion than in unanesthetized volunteers.19In a heterogeneous group of patients undergoing elective surgery of variable magnitude during subarachnoid block or sevoflurane–narcotic general anesthesia, volume expansion was greater in patients undergoing general anesthesia, but urinary elimination was similarly reduced in both groups.20Men undergoing short urologic procedures during epidural anesthesia retained a relatively high proportion of infused volume intravascularly.21
Together, these studies demonstrate the difficulty of separating the effects of volume kinetics during anesthesia and surgery. The magnitude of surgery and the hemodynamic responses to anesthetic and surgical manipulations varied substantially, with the most striking effects being differences in blood pressure. In the women undergoing cholecystectomy, induction of anesthesia was associated with hypotension so that blood pressure was substantially lower than baseline when fluid infusion began as surgery started; data collection for volume kinetic analysis continued for at least 2 h after completion of surgery, during which time blood pressure returned toward preanesthetic values.18In the comparison of general and subarachnoid anesthesia, a 60-min, 20-ml/kg fluid bolus was initiated 20 min before anesthetic induction.20Data were collected only until the end of the infusion, during which time blood pressure was 30–40 mmHg below the preinduction baseline in both groups, with the greater reduction occurring in the group receiving subarachnoid blocks. In men undergoing urologic surgery during epidural anesthesia, greater reductions in blood pressure were associated with greater intravascular retention of fluid.21
The study published in this issue of Anesthesiology was well designed to minimize the influence of surgical manipulation and partially isolate the influence of anesthesia while providing sufficient time to collect samples for kinetic analysis.13Thyroid surgery, which is associated with little soft tissue manipulation, lasted a mean of 143 min—sufficient time to complete most kinetic analyses. Anesthetic management was randomized to permit comparison of the effects of isoflurane and propofol. Although no control data were collected in unanesthetized patients, published data from unanesthetized subjects were available for comparison. The greater intravascular retention of fluid, in comparison to unanesthetized subjects in previous studies,1was associated with hypotension during both propofol and isoflurane anesthesia. Fractional plasma dilution was greater than in previously studied, unanesthetized volunteers, in association with reductions of 30–40 mmHg in both anesthetized groups.
These data should encourage advocates of crystalloid fluid therapy. Intravascular volume expansion produced by crystalloid fluids was increased during anesthesia in humans, and excess interstitial accumulation of fluid did not occur. Why do these data seem to conflict with data in sheep? One possibility is that because the sheep in the previously cited studies9,15did not have development of hypotension, the effects of anesthesia per se were evident. Perhaps anesthesia is associated with intravascular fluid retention if hypotension is prominent and with extravascular fluid retention if blood pressure is maintained at a higher level. In the surgical patients in the current study, the reduction in blood pressure seems to have provided the dominant influence. Further studies are necessary to determine the influence on volume kinetics of “typical” clinical anesthetic management, in which blood pressure is maintained closer to preoperative baseline than in the current study.
University of Texas Medical Branch, Galveston, Texas. dsprough@utmb.edu
