Dichloroacetate Reduces Diaphragmatic Lactate Formation But Impairs Respiratory Performance RUSS CIUFO, ANTHONY DIMARCO, DANIEL STOFAN, DAVID NETHERY, and GERALD SUPINSKI Pulmonary Division, Department of Medicine, Case Western Reserve University and MetroHealth Medical Center, Cleveland, Ohio Previous studies have found that administration of dichloroacetate (), an agent that reduces lactic acid generation, increases limb muscle endurance. The purpose of the present study was to determine if this agent also improves respiratory muscle performance. To examine this issue, we determined the effect of administration on the response to application of a large inspiratory resistive load (32,000 cm H 2 O L s) in unanesthetized decerebrate rats. Studies were carried out in four groups of animals: saline unloaded, unloaded, saline loaded, and loaded. was administered as 100 mg kg, given intravenously over 30 min, prior to respiratory loading. We found that diaphragm lactate levels were higher in saline-treated loaded animals than in unloaded controls and that administration prevented loading-induced increases in diaphragm lactate (p 0.001). -treated animals tolerated loading poorly, however, with a more rapid reduction in diaphragm pressure generation and a shorter time to respiratory arrest (42 3 min) than for saline-treated animals (57 3 min, p 0.01). These data indicate that administration decreases the tolerance to loaded breathing despite reductions in diaphragm lactate concentrations. We speculate that suppression of lactate formation by may impair metabolic regulation within the diaphragm during resistive loaded breathing. Keywords: lactate; respiratory failure; diaphragm; respiratory muscles It has been known for decades that lactic acid concentrations rise in strenuously contracting limb muscles (1, 2). Moreover, several studies have demonstrated that lactic acid production by contracting limb muscle can be reduced by the administration of dichloroacetate (), an agent that increases pyruvate dehydrogenase enzyme activity and reduces glycogenolysis (3, 4). Schneider and coworkers have further demonstrated that administration is associated with an increase in the endurance of swimming rats, an effect attributed to the action of this agent to suppress limb muscle lactic acid formation (5). In this earlier work, was administered as a 0.8 mol intraperitoneal injection, animals were subsequently placed in a water tank, and the total time that animals could swim was recorded. Control animals swam for only 255 18 s, while treated animals swam for 354 18 s, a 40% increase. has also been reported to improve limb muscle endurance in humans (6). Specifically, Ludvik and coworkers examined the effect of on time to exhaustion during treadmill exercise in human volunteers (6). In this particular study, each volunteer underwent two treadmill trials; saline was administered before one trial, while was administered as a 50 mg kg intravenous infusion before the other trial. increased time to exhaustion, increased maximum oxygen consumption, (Received in original form February 9, 1999; accepted in final form July 3, 2001) Supported by NIH 54825 and 38926. Correspondence and requests for reprints should be addressed to G. S. Supinski, M.D., 601 Elmwood Ave., Box 692, Rochester, NY 14642-8692. This article has an online data supplement, which is accessible from this issue s table of contents online at www.atsjournals.org Am J Respir Crit Care Med Vol 164. pp 1669 1674, 2001 DOI: 10.1164/rccm9902054 Internet address: www.atsjournals.org and reduced blood lactate concentrations (i.e., lactate levels were 8.53 0.45 mmol L at the end of control runs and 9.92 0.59 mmol L for runs, p 0.05). A more recent study found that infusion reduced phosphocreatine utilization at the beginning of exercise in human subjects, improving the cellular energy state (7). Taken together, these studies suggest that administration of dichloroacetate, or other agents with similar mechanisms of action, may be a reasonable strategy to improve skeletal muscle performance. In view of the demonstrated effect of dichloroacetate to enhance limb muscle endurance, we thought it possible that administration of this agent might also increase the capacity of the respiratory muscles to tolerate an elevated workload. In particular, we postulated that dichloroacetate administration would reduce the rate of development of diaphragmatic fatigue in animals breathing against a large inspiratory resistive load, improving load tolerance and slowing the development of respiratory failure. The purpose of the present study was to test this hypothesis by comparing the effects of inspiratory loading on groups of rats treated with saline or dichloroacetate; as controls, responses in loaded animals were compared with responses in groups of unloaded animals treated with either saline or dichloroacetate. Based on our hypothesis, we postulated that administration would result in (1) a reduction in loading-induced diaphragmatic lactate accumulation and (2) a decrease in loading-induced diaphragm contractile dysfunction (assessed from in vitro measurements of diaphragm force generation) associated with either a constant or an increased time to respiratory arrest. METHODS Approval was provided by the Case Western Reserve University Institutional Review Board. Details of the methodology used are provided in the online data supplement. In brief, male adult Zivic-Miller white rats (500 600 g) were initially anesthetized (using ketamine and halothane), arterial, venous lines were placed, and a tracheostomy was performed. A mid-collicular decerebration was then done and halothane was discontinued. A Hans-Rudolf valve was then attached to the endotracheal tube and a sideport in this valve was used to monitor airway pressure. Esophageal and gastric pressures were monitored with polyethylene catheters and a pneumotachograph attached to the expiratory limb of the Hans- Rudolf valve was used to assess airflow and tidal volumes. Diaphragm electromyographic (EMG) activity was assessed using bipolar electrodes inserted transcutaneously into the right lateral costal diaphragm. Animals were then assigned to one of four experimental groups: (1) a group (n 9) treated with (100 mg kg administered as a 50 mg ml intravenous infusion over 30 min) and then placed on an inspiratory resistive load, (2) a control group (n 8) to which saline (2 ml) was administered intravenously over 30 min prior to inspiratory loading, (3) an unloaded group (n 7) given over 30 min, and (4) an unloaded group (n 7) given saline over 30 min. Integrated diaphragmatic EMG activities monitored because recent work indicates that administration can alter motor outflow during respiratory loading by affecting neural reflex pathways (8). Selection of the dose was based on previous animal studies, which found that dosages of approximately this magnitude were adequate to improve limb and cardiac muscle performance (4 6, 9). The inspiratory load used in the first two experimental groups consisted of a long piece of polyethylene tubing (internal diameter of 0.86 mm, resistance 32,000 cm H 2 O L s).
1670 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 164 2001 Figure 1. Airway pressure (i.e., the pressure measured between the inspiratory resistive load and the trachea) over time for loaded animals treated with either saline (circles) or (squares). Error bars represent 1 SEM. Because our previous work (10 12) indicated that a load of this magnitude was associated with the rapid induction of significant hypoxemia in animals breathing room air, we administered 100% oxygen to animals during the period of inspiratory loading. Arterial blood samples were obtained for blood gas analysis immediately before and 5 min after initiation of inspiratory load trials. In inspiratory-loaded experimental groups, an additional blood gas was obtained at the point of respiratory arrest, which was defined as cessation of inspiratory efforts for 10 s. After respiratory arrest, the diaphragm was excised and a portion was used for assessment of in vitro contractile performance as previously described (10 12). Samples of diaphragm, soleus, and blood were assayed for lactate concentrations (13), and a portion of the diaphragm was also assessed for glycogen concentration (14). Data Analysis Data are presented as mean 1 SE. Unpaired t tests were used to compare time to respiratory arrest, peak EMG activity, and arterial Pa O2 levels at the conclusion of loaded breathing trials between salineloaded and -loaded groups. ANOVA with SNK post hoc testing was used to compare other variables in the four experimental groups. A p value of 0.05 was taken to indicate significance. Figure 3. Peak electromyographic (EMG) activity over time for loaded breathing trials. Symbols as for Figure 1. RESULTS Response to Breathing The institution of respiratory loading in saline-treated animals was followed by an immediate increase in airway pressure, an increase in transdiaphragmatic pressure, an increase in integrated diaphragm EMG activity, and an appreciable reduction in tidal volume and minute ventilation (Figures 1 4). Airway pressure, transdiaphragmatic pressure, and EMG activity continued to increase over approximately the first 10 min of loading, with airway pressure increasing to 54 4 cm H 2 O, transdiaphragmatic pressure increasing to 69 4 cm H 2 O, and EMG activity increasing to 383 50% of baseline (Figures 1, 2, and 3). After reaching this peak, further loading resulted in a gradual fall in airway pressure generation and transdiaphragmatic pressure in all saline-treated animals (Figures 1 and 2). Breathing frequency declined (i.e., breathing cycle duration increased) over the course of loaded breathing trials, and peak integrated EMG activity declined near the very end of loaded breathing trials (Table 1, Figure 3). Loading of saline-treated animals also produced a significant reduction in arterial PO 2 levels and Figure 2. Transdiaphragmatic pressure over time for loaded animals treated with either saline (circles) or (squares). Error bars as for Figure 1. Figure 4. Tidal volume over time for loaded breathing trials. Symbols as for Figure 1.
Ciufo, DiMarco, Stofan, et al.: Alters Respiratory Function 1671 TABLE 1. BREATHING PATTERN, ARTERIAL PRESSURE, AND ARTERIAL BLOOD GASES Inspiratory duration, s Initial 0.23 0.020 0.19 0.010 0.23 0.014 0.20 0.018 Final 0.25 0.020 0.19 0.010 0.32 0.017 0.32 0.023 Breathing cycle duration, s Initial 0.46 0.030 0.39 0.010 0.49 0.014 0.47 0.040 Final 0.52 0.050 0.45 0.020 1.71 0.060 1.62 0.080 Arterial pressure, systolic/diastolic in mm Hg Initial 159 9/120 8 161 6/117 3 153 9/129 7 154 7/110 6 Final 159 8/120 8 161 6/110 8 159 19/114 10 143 19/100 16 Pa O2, mm Hg Initial 342 37 317 49 365 33 375 8 Final 400 27 389 31 68 7* 78 6* Pa CO, mm Hg 2 Initial 40 6 37 2 38 4 38 2 Final 37 2 33 1 338 24* 305 24* Definition of abbreviation: dichloroacetate. * Findings statistically different from initial values, p 0.01. an increase in arterial PCO 2 concentrations (Table 1). Arterial pressure remained relatively constant over the duration of trials (Table 1). Although the general pattern of the response to inspiratory loading for -treated animals was similar to that observed for the saline-treated group, there were several important differences. First, the trajectory of the EMG time curve for the -treated group appeared to be slightly shifted upward when compared with the same curve for saline-treated animals (Figure 3) over the early portions of loaded breathing trials. Peak EMG activity (assessed as a % of preloading values), however, was statistically similar for saline and -treated groups of loaded animals (NS). The time course of inspiratory pressure generation also differed between the two groups of animals. Over the initial portion of loaded breathing trials (i.e., over the first 20 min), inspiratory pressure generation (i.e., airway pressure and transdiaphragmatic pressure) was similar in saline- and -treated animals, but during the latter portions of loaded trials (after 20 min), inspiratory pressure decreased more rapidly in - than in saline-treated animals (Figures 1 and 2). This more rapid reduction in inspiratory pressure generation in -treated animals was associated with a concomitant fall in tidal volume over the latter portion of loaded breathing trials and a shorter time to respiratory arrest (Figure 4). On average, respiratory arrest occurred 15 min earlier in -treated loaded animals than in the saline-treated loaded group, that is, -treated animals arrested at 42 3 min, while saline-treated animals arrested at 57 3 min into loaded breathing trials, p 0.01. Arterial Pa O2 was statistically similar at the end of loaded breathing trials in both saline- and -treated groups (NS, see Table 2). As expected, ventilatory parameters did not appreciably change over time in unloaded saline- and -treated groups. Specifically, diaphragm EMG activity, tidal volumes, inspiratory duration, breathing frequency, and arterial blood pressures all remained constant over the period of study in these groups (Table 1). For nonloaded groups, arterial Pa O2 remained high and Pa CO2 remained at approximately 40 mm Hg over the 1 h of data collection (Table 1). In Vitro Assessment of Diaphragm Force Generation The lengths and weights of excised diaphragm muscle strips used for in vitro force determination were not significantly different for the four experimental groups and had average values of 1.80 0.03 cm and 38.3 2.4 mg, respectively (NS for both comparisons across the four groups). In vitro force generation was significantly reduced for diaphragm muscles from both saline- and -treated inspiratory loaded animal groups when compared with unloaded control animals (Figure 5, p 0.01 for comparison of force frequency curves for muscles taken from loaded and unloaded animals). Of note, loading-induced reductions in diaphragmatic force generation were virtually identical for muscles TABLE 2. DIAPHRAGM TWITCH KINETICS (ms) Contraction time 87 9 95 9 72 5 89 8 1/2 relaxation time 64 6 64 8 63 2 79 8 Definition of abbreviation: dichloroacetate. Figure 5. Force frequency relationships for diaphragm muscles from unloaded saline-treated animals (open circles), unloaded -treated animals (open squares), loaded saline-treated animals (filled circles), and loaded -treated animals (filled squares). *Statistical difference, with force for loaded animals lower than force for unloaded groups, p 0.01 for comparisons at 1, 20, and 50 Hz; p 0.05 for the comparison at 100 Hz.
1672 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 164 2001 from both saline- and -treated animals. For example, diaphragm twitch force averaged 7.4 1.6, 7.6 1.0, 4.6 0.6, and 4.9 0.5 N cm 2, respectively, for saline-unloaded, unloaded, saline-loaded, and -loaded groups (p 0.002 for comparison of loaded to unloaded groups, with no difference between twitch forces for - and saline-treated loaded groups). Similar findings were seen for stimulation frequencies of 20, 50, and 100 Hz (i.e., the force developed by the diaphragm muscle strips taken from loaded and unloaded animals was statistically different at 20, 50, and 100 Hz, p 0.01 for each comparison). Twitch contraction and half-relaxation times for the four experimental groups were not different (NS, Table 2). The relative fall of force over time during in vitro repetitive stimulation trials was also similar for diaphragms taken from the four experimental groups, with force decreasing over 5 min to 16 2, 24 4, 19 3, and 19 3% of its initial value in salineunloaded, -unloaded, saline-loaded, and -loaded groups, respectively (NS). Lactic Acid (Lactate) and Glycogen Measurements Diaphragmatic lactate concentrations for saline- and treated unloaded groups of animals averaged 3.1 0.3 and 2.4 0.3 mmol g, respectively; both values for unloaded groups were low when compared with lactate levels of diaphragms taken from saline-treated loaded animals, which averaged 6.4 0.8 mmol g (p 0.001 for comparison of lactate in loaded saline-treated animals with the other experimental groups). treatment prevented loading-induced increases in diaphragmatic lactate, with diaphragmatic lactate levels averaging only 2.3 0.2 mmol g in -treated loaded animals. Lactate concentrations for blood and soleus muscles were not statistically different across the four experimental groups (Table 3). Diaphragms from unloaded saline- and -treated animals had glycogen concentrations of 15.8 1.5 and 17.0 2.3 mmol g, respectively. Diaphragms from loaded saline-treated animals had glycogen levels that were significantly lower than those of unloaded animals, averaging 8.4 1.8 mmol g (p 0.001). Glycogen levels for diaphragms from loaded, treated animals were similar to those of the unloaded groups, averaging 12.9 1.8 mmol g. Blood Glucose Measurements Blood glucose levels at the end of experimental trials averaged 201 10, 133 11, 196 14, and 206 14 mg dl, respectively, for saline-unloaded, -unloaded, saline-loaded, and -loaded groups, respectively; glucose levels between saline-loaded and -loaded groups were not statistically different. DISCUSSION In this study, we found that inspiratory resistive loaded breathing in the rat was associated with an increase in diaphragm lactate concentration. As expected, we also found that administration of prevented this loading-induced increase in diaphragm TABLE 3. END-OF-STUDY LACTATE CONCENTRATIONS Diaphragm, mmol g 3.1 0.3 2.4 0.3 6.4 0.8* 2.3 0.2 Soleus, mmol g 3.8 0.8 3.1 0.4 4.7 0.6 3.0 0.8 Blood, mmol ml 1.5 0.6 1.8 0.4 2.2 0.4 2.0 0.6 Definition of abbreviation: dichloroacetate. * Statistically different from the other experimental groups, p 0.001. lactate level. Surprisingly, administration appeared to accelerate the development of respiratory failure, shortening the time to respiratory arrest. Diaphragms taken from both control and -treated animals at the conclusion of loaded breathing trials had severe reductions in force-generating capacity when compared with diaphragms from unloaded-control animals. The magnitude of this loading-induced diaphragmatic dysfunction was similar for control and animals, however, despite the fact that load trials were substantially shorter for the -treated group. These latter findings suggest that contractile dysfunction developed more rapidly in the diaphragms of -treated animals when compared with saline-treated control animals. Taken together, these data indicate that had adverse, rather than beneficial, effects on respiratory system performance during loaded breathing, that is, both time to respiratory arrest was shortened and the rate of development of diaphragm contractile dysfunction was increased following administration. These findings refute our original hypothesis, that is, that would improve diaphragm performance during loaded breathing. Moreover, these findings stand in stark contrast to the reported ability of to improve limb muscle performance. Diaphragmatic Lactate In the present study, diaphragm lactate levels rose during loaded breathing in the absence of administration, reaching levels approximately twice baseline concentrations. Soleus and arterial blood lactate levels were not statistically different across the four experimental groups in this study. These two findings indicate that loading-induced increases in diaphragm lactate levels in saline-treated animals were largely the result of increased lactate production by the diaphragm. It is not possible to explain elevations of diaphragm lactate levels solely on the basis of muscle lactate uptake from the blood, as arterial blood lactate concentrations were similar for unloaded and loaded groups. Moreover, the fact that soleus muscle lactate concentrations did not rise indicates that increases in diaphragm lactate concentrations were not the result of a nonspecific stress associated with loading (e.g., hypoxemia, increased catecholamine levels), as such a nonspecific effect should have increased lactate concentrations in both soleus and diaphragm. Moreover, the Pa O2 levels seen at the conclusion of loaded breathing trials in the present study (see Table 2) are not low enough to induce a nonspecific increase in whole body lacate production. The increased diaphragm lactate generation observed in the present study, therefore, represents a specific increase in diaphragm glycolytic pathway activity in response to the heightened metabolic demands faced by the diaphragm during inspiratory resistive loading. This increased glycolytic activity is likely to also be responsible for the reduction in diaphragmatic glycogen stores observed for loaded, salinetreated animals when compared with unloaded, saline-treated control animals (i.e., glycogen concentrations decreased by 47% as a consequence of respiratory loading). More importantly, we also found that administration suppressed loading-related increases in diaphragm lactate concentrations. This finding is consistent with previous reports examining the effects of on lactate metabolism. Specifically, is known to reduce muscle glycogenolysis and to augment pyruvate dehydrogenase enzyme activity (3). These two effects work together to diminish lactic acid generation, the first by reducing the entry of substrate into the glycolytic pathway and the second by directing pyruvate into the Krebs cycle and away from metabolism to lactic acid via the action of lactic dehydrogenase. It seems likely that both actions of were operative in preventing loading-induced increases in dia-
Ciufo, DiMarco, Stofan, et al.: Alters Respiratory Function 1673 phragmatic lactate concentrations in the current study. Specifically, the fact that prevented loading-induced reductions in diaphragm glycogen concentrations (i.e., diaphragm glycogen content was lower for saline-treated loaded animals than control animals and prevented this loading-induced reduction in glycogen) indicates that this agent reduced glycolysis in the diaphragm during loaded breathing. A number of previous studies have suggested that the diaphragm may not produce lactate under all circumstances when the respiratory workload is increased, but instead, can sometimes become a net consumer of lactate (15). For example, Ferguson and coworkers found that diaphragm lactate levels did not increase during inspiratory resistive loading in anesthetized rats despite the fact that the load applied was of sufficient magnitude to induce respiratory failure (16). The fact that diaphragm lactate concentrations rose in the present study but not in some previously published experiments employing other models of increased respiratory muscle work is probably related to differences in study design (15, 16). Our animals were unanesthetized, were subjected to very high levels of inspiratory resistive loading, and developed respiratory failure fairly rapidly. This study design may well have resulted in higher levels of respiratory drive and pressure production than in previous respiratory loading models utilizing either anesthetized animals or more modest levels of respiratory loading of unanesthetized animals. Effect of on Breathing Trial Time and Diaphragmatic Force Generation A number of previous studies examining the effects of administration have indicated that this drug favorably influences the performance of limb muscles (5, 6). For example, Schneider and coworkers found that administration of this agent to rats produced an increase in the swimming endurance of these animals (5). In addition, Ludvik and coworkers found that administration of to normal human volunteers enhanced performance during treadmill exercise (6). In this latter experiment, eight healthy male volunteers were tested by bicycle spiroergometry after infusion of either (50 mg kg) or saline. Prior infusion of reduced blood lactate levels at both submaximal levels of exercise (50% of the exercise trial) and at the point of exhaustion. Oxygen consumption increased significantly after when compared with saline at both the anaerobic threshold and at maximal exercise (e.g., at the latter point oxygen consumption was 35.6 1.7 and 30.5 1.0 ml kg min, respectively, for saline and trials, p 0.05). These authors concluded that significantly increased exercise capacity in normal volunteers, and suggested that should be tested as a therapy for patients with reduced exercise capacity. Other reports have indicated also improves cardiac muscle function (9), increasing stroke volume and enhancing myocardial efficiency in patients with coronary artery disease after administration as a 35 mg kg intravenous infusion. The authors of these previous studies have suggested that the improved performance of limb skeletal and cardiac muscle observed following administration was due to an effect of this agent to suppress lactic acid generation during strenuous contractions. Based on these previous findings, we hypothesized that administration would also improve the performance capabilities of the diaphragm and other respiratory muscles, making it possible for animals given this agent to have an increased time-to-failure during inspiratory resistive loaded breathing, and or to develop less diaphragmatic fatigue over time during loaded breathing. Instead, we found that time to respiratory arrest was reduced by administration during resistive loaded breathing and that the amount of fatigue developed in the diaphragm by the end of loaded trials was the same in control and -treated animals despite the fact that load trial time was shorter in -treated animals. Carbon dioxide levels were markedly elevated in both saline- and -treated animals during in vivo loaded breathing trials, but the magnitude of this elevation was similar in both groups and carbon dioxide elevations were likely a consequence of the development of respiratory failure in inspiratory loaded animals rather than a cause of respiratory failure. More importantly, in vitro diaphragm force measurements were done in organ baths in which carbon dioxide levels were constrained to remain at normal levels (i.e., carbon dioxide levels of 5% correspond to a tension of 38 mm Hg). These in vitro measurements indicate, again, that the same degree of muscle fatigue developed in a shorter time for -treated loaded animals than for saline-treated, loaded controls. It is worth considering possible mechanisms by which may have impaired respiratory muscle performance, and also, considering potential explanations for the apparently differential response of limb muscles and respiratory muscles to this agent. As mentioned earlier, is known to reduce muscle glycogenolysis and to augment pyruvate dehydrogenase enzyme activity, and these two effects are thought to work together to diminish lactic acid generation. It is possible that the effect of to inhibit diaphragm glycogenolysis in the present study so reduced glucose entry into the glycolytic pathway that substrate flux into the Krebs cycle was reduced and oxidative ATP generation decreased, increasing reliance of the diaphragm on energy storage pools (i.e., CPK) during repetitive contraction. can also inhibit pyruvate transport into mitochondria, and it is possible that this latter effect may have further contributed to a reduction in substrate availability for oxidative phosphorylation in the present study (17). Recent work has also suggested that lactic acid plays an important feed-forward regulatory role in increasing oxidative phosphorylation in response to muscle activity (18). Impairment of lactic acid generation as the result of administration may have altered the regulation of oxidative phosphorylation in the diaphragm during loaded breathing, further attenuating ATP generation. In addition to having potentially detrimental direct effects on diaphragm metabolism, it is conceivable that may also have altered the course of development of diaphragmatic fatigue by changing the pattern of neural drive to this muscle. Diaphragm EMG activity appeared to be greater during the early portion of loaded breathing trials in as compared with control loaded animals; it is possible that this heightened initial respiratory motor outflow may have overdriven the respiratory muscles, contributing to an acceleration in the development of fatigue. It seems plausible that differences between limb and respiratory muscles with regard to the role of reflex modulation of efferent motor outflow and or the sensitivity of cellular energy pathways to may exist and account for the fact that appears to enhance limb muscle endurance while impairing respiratory muscle performance. In addition, one might reasonably postulate that the differences between the present and past reports of the actions of this agent on muscle performance may be related to differences in dosing. This possibility seems unlikely, however, because the dosage used in the present study is lower than that used for some studies finding a beneficial effect of on muscle function (5), but higher than that used in other studies reporting benefit (6, 9). Clinical Usage of Regardless of whether the effects of on the respiratory apparatus in the present study were mediated by direct effects of this agent on respiratory muscle metabolic function or by some other, more indirect mechanism, it is clear that adminis-
1674 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 164 2001 tration of this agent produced a significant impairment in the functional capacity of the respiratory pump. This drug has been used to treat acidosis in critically ill patients (19, 20), and recent reports indicate that this drug is still employed to treat acidosis in children (21), during sepsis (21), during antiretroviral therapy (22), and in neurological disorders (23). In fact, a phase III trial to evaluate the effects of in patients with traumatic brain injury is currently being planned (24). The present data suggest, however, that therapeutic use of this agent may be associated with a previously unrecognized hazard. We would suggest that future therapeutic use of this drug be performed with caution and should include an assessment of the impact of this agent on respiratory muscle function. References 1. Hill AV, Kupalov P. Proc Roy Soc Ser B 1929;105:313 328. 2. Nakamura Y, Swartz S. Influence of hydrogen ion concentration on calcium binding and release by skeletal muscle sarcoplasmic reticulum. J Gen Physiol 1972;59:22 32. 3. Clark AS, Mitch WE, Goodman MN. Dichloroacetate inhibits glucolysis and augments insulin-stimulated glycogen synthesis in rat muscle. J Clin Invest 1987;79:588 594. 4. Stacpoole PW. The pharmacology of dichloroacetate. Metabolism 1989; 38:1124 1144. 5. Schneider SH, Komanicky PM, Goodman MN, Ruderman NB. Dichloroacetate: effects on exercise endurance in untrained rats. Metabolism 1981;30:590 595. 6. Ludvik B, Mayer G, Stifter S, Putz D, Barnas U, Graf H. Effects of dichloroacetate on exercise performance in healthy volunteers. Pflügers Arch 1993;423:251 254. 7. Howlett RA, Heigenhauser GJ, Hultman E, Hollidge-Horvat MG, Spriet LL. Effects of dichloroacetate infusion on human skeletal muscle metabolism at the onset of exercise. Am J Physiol 1999;277:E18 E25. 8. Petrozzino JJ, Scardella AT, Santiago TV, Edelman NH. Dichloroacetate blocks endogenous opioid effects during inspiratory flow-resistive loading. J Appl Physiol 1992;272:590 596. 9. Wargovich TJ, Macdonald RG, Hill JA. Myocardial metabolic and hemodynamic effects of dichloroacetate in coronary artery disease. Am J Cardiol 1988;61:65 70. 10. Supinski GS, Nethery D, Ciufo R, Renston J, DiMarco A. Effect of varying inspired oxygen concentration on diaphragm glutathione metabolism during loaded breathing. Am J Respir Crit Care Med 1995; 152:1633 1640. 11. Ciufo R, Nethery D, DiMarco A, Supinski G. Effect of varying load magnitude on diaphragm glutathione metabolism during loaded breathing. Am J Respir Crit Care Med 1995;152:1641 1647. 12. Supinski GS, Stofan D, Ciufo R, DiMarco A. Diaphragmatic glutathione levels and fatigue during loaded breathing: examination of the effect of N-acetylcysteine administration. J Appl Physiol 1995;79:340 347. 13. Noll F. Methode zur quantitativen Bestimmung von L-( )-Lactat mittels Lactat-Dehydrogenase and Glutamat-Pyruvat-Transaminase. Biochem Z 1966;346:41 49. 14. Passonneau JV, Lauderdale VR. A comparison of three methods of glycogen measurement in tissues. Anal Biochem 1974;60:405 412. 15. Mayock DE, Standaert TA, Murphy TD, Woodrum DE. Diaphragmatic force and substrate response to resistive loaded breathing in the piglet. J Appl Physiol 1991;70:70 76. 16. Ferguson GT, Irvin CG, Cherniack RM. Relationship of diaphragm glycogen, lactate, and function to respiratory failure. Am Rev Respir Dis 1990;141:926 932. 17. Halestrap AP. The mitochondrial pyruvate carrier. Kinetics and specificity for substrates and inhibitors. Biochem J 1975;148:85 96. 18. Connett RJ, Gayeski TE, Honig CR. Lactate accumulation in fully aerobic, working, dog gracilis muscle. Am J Physiol 1984;264:H120 H128. 19. Stacpoole PW, Harman EM, Curry SH. Treatment of lactic acidosis with dichloroacetate. N Engl J Med 1983;309:390 396. 20. Stacpoole PW, Wright EC, Baumgartner TC. Dichloroacetate-lactic acidosis study group: a controlled trial of dichloroacetate for treatment of lactic acidosis in adults. N Engl J Med 1992;327:1564 1569. 21. Arnon S, Litmanovits I, Regev R, Elpeleg O, Dolfin T. Dichloroacetate treatment for severe refractory metabolic acidosis during neonatal sepsis. Pediatr Infect Dis J 2001;20:218 219. 22. Shaer AJ, Rastegar A. Lactic acidosis in the setting of antiretroviral therapy for the acquired immunodeficiency syndrome. A case report and review of the literature. Am J Nephrol 2000;20:332 338. 23. Tarnopolsky MA, Beal MF. Potential for creatine and other therapies targeting cellular energy dysfunction in neurological disorders. Ann Neurol 2001;49:561 574. 24. Williams PJ, Lane JR, Turkel CC, Capparelli EV, Dziewanowska Z, Fox AW. Diclhoroacetate: population pharacokinetics with a pharacodynamic sequential link model. J Clin Pharacol 2001;41:259 267.