| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Physiology, Queen's University, Kingston, Ontario, Canada
Submitted 14 July 2006 ; accepted in final form 29 November 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
75% of the peak tracheal pressure generated during a 30-s occlusion) until pump failure (fall in tracheal pressure to half; mean 38 min). Hypercapnia and hypoxemia developed rapidly (4 min), coincident with diaphragmatic fatigue (decreased ratio of transdiaphragmatic pressure to peak integrated phrenic activity) and the detection in blood of the fast isoform of skeletal troponin I (muscle injury). At 23 min, respiratory frequency and then blood pressure fell, followed immediately by secondary diaphragmatic fatigue. Blood taken after termination of loading contained cardiac troponin T (myocardial injury). Contrary to our hypothesis, diaphragmatic fatigue and injury occurred early in loading before central failure, evident only as a change in the timing but not the drive component of the central respiratory pattern generator. Stepwise multiple regression analysis selected changes in mean arterial pressure and arterial PCO2 during loading as the principal contributing factors in load endurance time, and changes in mean arterial pressure as the principal contributing factor in peak tracheal pressure generation. In conclusion, the temporal development of respiratory failure is not stereotyped but depends on load magnitude; moreover respiratory loads induce cardiorespiratory, not just respiratory, failure. diaphragm; fatigue; heart; injury; troponin
This lack of consensus reflects, in part, a lack of understanding of the variables involved in the development of respiratory pump failure. Only two studies have tested the effects of a single variable [anesthetic level (19); PO2 and PCO2 (18)] on the development of respiratory pump failure. Moreover, studies are characterized by a wide array of load types (resistive, threshold, elastic of varying severities), experimental models (different species and ages), and conditions (level of inspired oxygen, criteria for failure). These differences complicate comparisons not just of results but also of the temporal development of respiratory pump failure and of the relative contributions of central failure, peripheral fatigue, and neurotransmission failure. Although each study provides insights into a specific experimental model and mechanisms, taken together they confound understanding of the physiological and molecular mechanisms underlying respiratory muscle pump failure.
We recently developed an anesthetized rat model in which a moderate inspiratory resistive load (IRL) [60% of the peak tracheal pressure (Ptr) developed during a prior 30-s occlusion] elicits a stereotyped sequence of hypercapnic failure, muscle injury [release of skeletal troponin I (sTnI)], diaphragmatic fatigue [decreased ratio of transdiaphragmatic pressure to integrated phrenic activity (Pdi/
Phr)], central failure (abrupt bradypnea), and respiratory pump failure (decreased Pdi and Ptr) (43). Failure occurred after 2.4 h, much longer than that reported by others in rats (6, 11, 34, 45, 47) and rabbits (39), where failure occurs in <20 min, a failure considered to be of central origin (39, 49).
The purpose of this study was to test the hypotheses that, compared with the moderate load in our previous study (43), a severe load would cause respiratory pump failure without the development of respiratory muscle injury and fatigue and do so primarily because of central failure (often called central fatigue). Contrary to the first hypothesis, injury and diaphragmatic fatigue occurred very rapidly, before pump failure. In terms of the second, respiratory drive increased during loading but respiratory frequency fell because of an increase in expiratory duration (TE), indicating an effect on only the timing component of the central respiratory pattern generator. Critically, pump failure was preceded by a marked hypotension, and cardiac troponin T (cTnT), a marker of myocardial injury, was found in samples of blood taken after the load was removed. Pearson correlation and stepwise multiple regression analysis revealed that changes in mean arterial pressure (MAP) and arterial PCO2 (PaCO2) during loading accounted for 74% of the variance in load endurance.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
37°C with a servocontrolled heating pad. Preparation of the animals was identical to that described previously (43). In brief, after making a midline incision in the neck, a tracheal cannula was inserted; one port was connected to a pressure transducer to measure Ptr, the other to a two-way valve to separate inspiratory and expiratory flows (Hans Rudolph 2300, Kansas City, MO). The right carotid artery (for measuring blood pressure; Cybersense CyQ BPM01, Nicholasville, KY) and sampling arterial blood gases (Radiometer ABL-5, Copenhagen, Denmark) and jugular vein were cannulated and maintained patent with heparinized saline. The left phrenic nerve was isolated and its activity recorded en passant, amplified, filtered (100–10,000 Hz; Grass P-511, Quincy, MA) and integrated (Paynter filter, time constant 50 ms). We recorded Pdi as the difference between abdominal (Pab) and esophageal (Pes) pressures (Millar SPR 524, Houston, TX). All signals (blood pressure, Phr, and Ptr, Pes, Pab, and Pdi) were acquired to computer (CED Spike2, Cambridge, UK).
To increase inspiratory flow resistance, we gradually tightened, in three steps at 5-min intervals, a small clamp placed on the inspiratory side of the valve until the rat generated a tidal Ptr 75% of the peak Ptr obtained during a previous 30-s occlusion. Loading was considered to have started (normalized time = 0) at the end of this 10-min period. Blood samples were taken before loading and at 5-min intervals thereafter. Loading was discontinued when Ptr declined to half of the peak loaded value (respiratory pump failure; normalized time = 1.0). All animals were given a supplemental dose of pentobarbital before tissues were harvested.
Samples were analyzed for sTnI by using Western blot-direct serum analysis (WB-DSA) as previously described (42, 43). Anti-TnI monoclonal antibodies of confirmed isoform specificity were chosen for WB-DSA: sTnI, FI-32 and FI-23 (fast only, Spectral Diagnostics, Toronto, ON, Canada); MYNT-S (preferential for slow in rat; 31); and 3I-35 (fast, slow, and cardiac; Spectral Diagnostics). Specificity of all antibodies was confirmed by Western blot analysis of cardiac and skeletal tissue from human and rat as described previously (44). To compare release of fast sTnI between loading and sham, blots were scanned at a resolution of 300 dpi, and three-dimensional volume (density) plots were generated; densitometry values of fast sTnI were normalized to levels in control lanes. Immediately after the load was removed, we obtained blood samples (n = 7) to test for the presence of cTnT, a cardiac-specific marker of myocardial injury (Roche E170, Laval, Quebec, Canada).
Because endurance on the load varied between rats, the time on the full load until pump failure was normalized by dividing the load duration into deciles, along with recovery in rats that survived for at least 15 min after removal of the load. At least 30 breaths, excluding sighs and the postsigh breath, were analyzed at each of these time points. Data are presented as means ± SD or SE, as indicated. Paired t-tests or ANOVA, Holm-Sidak corrected, for multiple comparisons were used to compare data; P < 0.05 was considered significant. To establish associations between endurance time and its contributing factors, we followed the procedures described by Lougheed et al. (28). Pearson's correlation coefficients were first calculated using the changes in endurance time as the dependent variable and the initial or concurrent changes in cardiorespiratory parameters [MAP, heart rate (HR), arterial PO2 (PaO2) and PaCO2, pH, breathing frequency (f), diaphragmatic contractility (Pdi/Phr), peak Ptr, inspiratory and expiratory durations (TI and TE,), respiratory duty cycle (TI/TTOT), respiratory drive (Phr/TI), and the product of Phr and f (minute phrenic activity, MinPhr)] as independent variables. The significant correlates of peak Ptr or endurance time were then selected for forward stepwise linear regression analysis to generate a predictive equation for endurance time. The same process was used to establish the contributing factors for peak Ptr but using the prevailing MAP, HR, PaCO2, PaO2, and pH and the change in MAP during the occlusion (
MAPoccl).
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
100 mmHg) MAP that fell much less than in rats with a higher (140 mmHg) MAP (Table 1 and Fig. 1, A and B). There were no differences in preocclusion blood gases and load endurance times between these two groups, but preocclusion MAP and HR were significantly greater in those rats in which MAP fell more (Fig. 1A and Table 1). Four variables were identified by Pearson's correlation analysis as correlates of the peak Ptr during occlusions: the PaO2, MAP, and HR before occlusion and MAPoccl. Subsequent stepwise multiple regression analysis selected MAPoccl as the strongest independently significant predictor of peak Ptr. With MAPoccl established, the model was then reevaluated by linear regression analysis; the other variables did not significantly add to the model. Finally, the model was reestimated by linear regression analysis to obtain the final equation: peak Ptr = 43.2 (cmH2O) + 0.118 (cmH2O/mmHg) x MAPoccl (mmHg); r = 0.723, r2 = 0.523, P = 0.012.
|
|
Phr during IRL in a rat. In Fig. 2, extracts of the tracing (A, control; B, load onset; C, midway; and D, during respiratory pump failure) are shown on an expanded time scale in the bottom panels. The occasional large swings in pressure and Phr are sighs. Pdi in all rats was due almost entirely to changes in Pes.
|
MAPirl), the increase in PaCO2 during loading (PaCO2), the fall in PaO2 during loading, and the final (pump failure) MAP. Endurance time did not correlate with the other [initial, delta () or final] cardiorespiratory variables. Stepwise multiple regression analysis selected PaCO2 as the strongest independently significant predictor. With PaCO2 established, the model was then reevaluated by linear regression analysis; only MAPirl added significantly to the model. Finally, the model was reestimated by linear regression analysis to obtain the final individual equations: endurance time (s) = 5,047.3 (s) – 69.4 (s/mmHg) x PaCO2 (mmHg), r2 = 0.466, P < 0.001; endurance time (s) = 1,402.8 (s) + 26.7 (s/mmHg) x MAPirl (mmHg); r2 = 0.140, P = 0.094. The final combined equation in the nine rats for which complete data were available was: endurance time (s) = 4,223 (s) – 70.1 (s/mmHg) x PaCO2 (mmHg) + 30.4 (s/mmHg) x MAPirl (mmHg); n = 9, r2 = 0.740, P = 0.017.
The normalized temporal profiles of changes in blood gases are shown in Fig. 3A. Load onset was associated with a rapid (time = 0.1, 4 min) decrease in PaO2 and increase in PaCO2 (hypercapnic ventilatory failure), indicating hypoventilation; these values persisted throughout loading. Just before respiratory pump failure, PaO2 and PaCO2 were 51 ± 8 (SD) and 82 ± 9 mmHg, respectively.
|
Diaphragmatic contractility (Pdi/Phr) during normal (not sighs) loaded breaths decreased within 4 min (time = 0.1) of full application of the load; this constitutes primary diaphragmatic fatigue (Fig. 3D). Pdi/Phr remained at this value until falling again (secondary diaphragmatic fatigue) at time = 0.95, preceding respiratory pump failure by 2 min. Unlike Pdi/Phr, respiratory drive (the rate of increase in phrenic activity during inspiration, Phr/TI) and MinPhr increased and remained elevated throughout loading until pump failure (Fig. 3, E and F, respectively).
MAP abruptly decreased at time = 0.9 and continued to fall (to 55 mmHg) until respiratory pump failure (Fig. 4A). This decrease was not due to a decrease in HR that increased throughout loading (Fig. 4B). In the 11 rats that died within 5 min of respiratory pump failure, MAP fell even after removal of the load.
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Phr), and hypercapnic failure occurred almost simultaneously (to the limit of detection based on the frequency of blood samples) and after only 4 min of full loading. This sequence of events leading to pump failure (Fig. 7) differs markedly from that in our previous study of the effect of a moderate load (43) in which hypercapnic failure and the release of sTnI preceded diaphragmatic fatigue by 30 min. Third, diaphragmatic fatigue occurred in two stages; very rapidly, coinciding with hypercapnic failure and the detection of fast sTnI in the blood, and later almost immediately (2 min) after an abrupt fall in MAP. Fourth, severe IRL consistently caused release into the blood of cTnT, an indicator of cardiac injury, in all rats tested. Fifth, stepwise linear regression analysis selected MAPoccl as the best predictor of peak Ptr generation during occlusions. A similar analysis selected MAPirl and PaCO2 as the best predictors of load endurance. Taken together, our results are the first to implicate cardiovascular function as a key factor determining endurance time when breathing against an IRL.
|
Although endurance times were, as expected, less than in our previous study using a smaller IRL (43), it was still greater than those reported by others in rats [3 min (34), 9 min (6), 7 min (11), 11.5 min (45), and < 20 min (47)] and rabbits [16 min (39)] despite the size of the load. Such rapid failure is generally considered to be mainly of central origin (e.g., 39, 49). The most notable difference between our study and others is that we applied the load in stages, each stage being accompanied by no (Fig. 2) or only a transient drop in blood pressure, unlike the situation when application of the final load in a single step resulted in death within minutes (unpublished observations). Thus too rapid application of the target load could overwhelm the cardiovascular system whereas incremental application of the load would allow it time to adapt. Nevertheless, IRL caused cTnT release. The underlying mechanism of adaptation is uncertain and merits further study. Our results suggest that monitoring MAP during weaning from mechanical ventilation may provide useful information about the success or failure of the weaning effort.
Variability of endurance is a hallmark of studies of responses to respiratory loads. For example, in one study of loaded rabbits, the endurance range was 8–44 min, a 5-fold difference (39); in another, the range was 10- 180 min (on a lower load, and 3 of the rabbits breathed oxygen), an 18-fold difference (1). The reasons for this wide variation, which accounts for load duration being normalized in most studies, have never been explained. A clue that cardiovascular factors may be involved came from the studies of Borzone et al. (12, 13) who reported that, in preliminary studies, the rats failed rapidly on an IRL because of hypoxemia and hypotension. Using Pearson correlation and stepwise multiple linear regression, we now show that, in barbiturate-anesthetized rats breathing air, 74% of the variance in load endurance can be accounted for by the fall in MAP and the increase in PaCO2 during the load. The contribution of MAP is the best, indeed only, predictor of the peak Ptr developed during the 30-s occlusion that determines the size of the load. These novel findings, along with the fact that loading caused myocardial injury, suggest that measurements of cardiovascular parameters other than just blood pressure and HR should be a focus of future studies of loading.
Central failure.
Central failure (sometimes referred to as central fatigue) is generally described as a reduction in motor output to the respiratory muscles. Reduced output could result from a decrease in the drive component, indicated by Phr/TI, a decrease in f, or both. The term central failure or fatigue suggests that drive is inadequate. However, failure implies that the objective is ventilation (prevention of hypercapnia) rather than hypoventilation (resting the respiratory muscles and allowing hypercapnia to develop in an attempt to prevent impending respiratory pump failure or to allow more time for perfusion of the respiratory muscles). In this context, a change in breathing pattern (e.g., a decrease in f due to an increase in TE) does not necessarily indicate central failure especially if Phr/TI drive and MinPhr do not fall. However, if f and MinPhr or f and Phr/TI both fall, this would certainly be consistent with central failure. How the timing and drive components of the central pattern generator can be affected separately is not understood.
Our results support a limited role for central failure in IRL-induced respiratory failure. Although TE increased (and f decreased) starting at t = 0.6 (Fig. 3, B and C), indicating alteration in the timing component of the respiratory central pattern generator, TI did not change except at the point of pump failure. In contrast, in rats breathing against a moderate IRL, f fell drastically and abruptly only at the time of pump failure (Fig. 3 in 43). In both the previous (43) and this study, Phr/TI increased and was maintained during loading (Fig. 3E), indicating no failure of the drive component of the central pattern generator. Similarly, HR increased during loading (Fig. 4B), indicating greater drive from sympathetic neurons in the brain stem; the plateau may indicate that the limits of this compensatory increase had been reached. Our results are consistent with those of previous studies in rabbits subjected to IRL (36, 39) or inspiratory threshold loads (19) (regardless of anesthetic depth), in dogs bronchoconstricted by inhalation of nebulized methacholine (51) or subjected to cardiac tamponade (7) in which frequency falls but neural drive is maintained even at task failure. On the other hand, our results differ from those in cats subjected to IRL (2) and rabbits subjected to inspiratory threshold loads (18, 19) or IRL (39) in which central drive declines while frequency remains unchanged at task failure. The reasons for these discrepancies remain unresolved.
Bradypnea, in our model, was unlikely related to the load per se as it appeared 23 min after full application of the load and was due to changes in TE, not TI, even though the load was applied only to inspiration. The mechanism(s) responsible for the bradypnea is unknown but may be part of a central reflex mechanism designed to reduce diaphragmatic activity, thereby avoiding fatigue and pump failure (2, 21, 50) at the expense of worsening hypercapnia, or to increase O2 delivery to the respiratory muscles by increasing the time available for perfusion. This strategy can be successful only if the subject can tolerate the hypercapnia or shift the load to other respiratory muscles, as occurs in human subjects voluntarily generating large diaphragmatic pressures (10). This strategy may not be available to animals in which accessory respiratory muscle activity is suppressed by anesthesia.
Diaphragmatic fatigue and injury.
We observed that diaphragmatic fatigue, indicated by the fall in Pdi/Phr, occurred in two stages: soon after the application of the load (primary fatigue; t = 0.1 or 4 min) and immediately following a decrease in MAP (secondary fatigue; t = 0.95, or 36 min) (Fig. 3D). Both occurred despite greater central drive (assuming that Phr/TI is a valid index of drive even under fatiguing conditions) and are consistent with peripheral fatigue, not central failure. Although we cannot exclude transmission failure at the neuromuscular junction as a factor contributing to fatigue (9), we feel this is unlikely because it is most often observed in in vitro models in which the diaphragm is activated by shocks.
What accounts for the load-induced decrease in diaphragmatic contractility? A decrease in diaphragmatic length due to an increase in end-expiratory lung volume is unlikely to be responsible because we observed no increases in end-expiratory Pes during loading. Hypoxia depresses in vitro diaphragmatic contractility (30), but this cannot account for our findings because in the present study, Pdi/Phr fell within 4 min while in the previous study (43) the rats experienced over 2 h of a similar degree of load-induced hypoxia without a change in contractility, suggesting that hypoxia is not responsible. Whether respiratory acidosis causes diaphragmatic fatigue is unresolved (26, 40, 41), but the generation of intracellular acidosis and inorganic phosphate as a result of increased contractile activity have been well recognized to impair force generation (for reviews, see Refs. 3, 20, 48). While intracellular changes may contribute to the development of fatigue, the detection and quantification of serum levels of sTnI are specific and sensitive markers of skeletal muscle injury. While sTnI levels may not correlate with the onset of fatigue, its appearance does indicate injury, which means contractile dysfunction (fatigue) is present.
Assessment of striated muscle injury can be direct or indirect. The former is invasive (muscle biopsy) and is not always reliable or sensitive (for review, see Ref. 16), while the latter involves assessment of muscle soreness, observing a decline in force and/or power, or relying on elevation of serum biomarkers. Muscle soreness is problematic as it does not always occur with muscle injury and usually takes days to develop (16). Detecting a decline in force and/or power requires control measurements for comparison; these are seldom available for patients. Serum biomarkers are the best tool to assess muscle injury as analysis can be done within minutes, are quantitative, require only a blood sample, and are more sensitive than other measurements of injury. Several biomarkers of injury exist for striated muscle (e.g., creatine kinase, aldolase, carbonic anhydrase, and myosin light chain), but TnI and TnT, key regulatory myofilament proteins, have emerged as the best. TnI exists as three isoforms: cardiac (cTnI) and fast and slow sTnI, which are found exclusively in cardiac and in fast- and slow-twitch skeletal muscle, respectively. As troponins are absent in the serum of healthy individuals, any detectable levels in serum constitute irrefutable evidence of a loss of cell membrane integrity (injury) whether the injury is reversible (commonly referred to as stunning in the myocardium) or irreversible (necrosis). Currently, cTnI and cTnT are the gold standards for diagnosing myocardial injury (4).
The appearance of fast sTnI in blood indicates injury of fast-twitch skeletal muscle fibers, presumably of the diaphragm because it was the major if not the only active muscle. While we cannot exclude the possibility that other muscles contributed to the released sTnI, severe hypoxemia leading to respiratory arrest in anesthetized dogs induced protein changes only in the diaphragm even though other muscles were activated by the hypoxemia (44). In rabbits, loading injured only the diaphragm, not other respiratory or limb muscles, despite their being exposed to the same asphyxic blood gases (25). This supports our contention that the diaphragm is the source of the sTnI. Injury could contribute to any loss of force-generating capacity caused by decreases in contractility due to hypoxemia and/or respiratory acidosis, unless recruitment and/or increases in discharge frequency of motor units compensate. The appearance of fast sTnI coincided with diaphragmatic fatigue and hypercapnic failure, results similar to those observed in rats subjected to a moderate IRL (43), and suggests that substrate delivery to and/or waste removal from at least some fast-twitch fibers was inadequate. Slow sTnI was not detected in serum, indicating that slow-twitch skeletal muscle fibers were not injured by the load. The absence of slow sTnI in serum even after surgery probably reflects the composition of the muscles at the site of surgery (e.g., sternohyoid), muscles that are almost exclusively fast twitch (14).
Blood pressure and cardiac injury. Blood pressure in animals subjected to IRL is seldom reported, perhaps because of an unstated assumption, based on previous work reporting large increases in perfusion of the diaphragm during loaded breathing [see, e.g., Ref. 32 and for references to earlier work], that perfusion and therefore O2 delivery increase to match increases in metabolic demands. Thus cardiac limitations to respiratory muscle O2 delivery are seldom considered unless the impairment is of cardiac origin. For example, in dogs subjected to cardiac tamponade, diaphragmatic fatigue occurs before respiratory pump failure (7), likely because of inadequate perfusion. Similarly, acute hypotension (40–50 mmHg) in dogs depresses diaphragmatic contractility before task failure (23) as its perfusion becomes pressure dependent below a MAP of 70 mmHg (24).
In rats, the onset of hypotension is associated with simultaneous decreases in diaphragmatic microvascular PO2 (38). If the diaphragm's vascular bed has already dilated maximally, O2 delivery will decrease if MAP falls and, eventually, be unable to meet metabolic demands. Load-induced increased metabolic demands combined with arterial hypoxemia may account for the initial damage to the fibers in our study; later, however, hypotension-induced hypoperfusion of the respiratory muscles may have compromised O2 delivery, precipitating the abrupt fall in diaphragmatic contractility (secondary fatigue).
Our findings indicate that at least one factor contributing to diaphragmatic dysfunction occurs at the level of the myofilament proteins, specifically sTnI. Our results are similar to those observed in patients presenting to the emergency department with respiratory-related disorders (42); they, too, release only the fast isoform. In contrast, in patients with drug-induced muscle injury or blunt trauma, both isoforms are detected in the bloodstream (42), indicating that both fiber types can be injured. Other markers of injury have been used before. During acute exacerbations of asthma, the increase in creatine kinase activity correlates with the severity of airway obstruction (15). Moreover, this increase in serum creatine kinase activity was not cardiac in origin because cTnT was not present in such patients (29).
Our results also suggest that severe inspiratory loads can impair cardiac function and thereby limit O2 delivery to the periphery, despite compensatory reflexes (increased HR). The few studies that do report arterial blood pressure during IRL are consistent with this idea. In rats subjected to 60–70% IRL, MAP fell to 25 mmHg (post-IRL) and, after removal of the load, to <10 mmHg despite 15 min of mechanical ventilation (11), a result consistent with irreversible cardiac injury. In decerebrate rats subjected to IRL, MAP decreased from approximately 150 to 39 mmHg over the final few minutes before respiratory arrest (46); smaller falls were observed in two other studies (45, 47). Borzone et al. (12, 13) noted that rats breathing air and subjected to IRL failed within 10 min because of hypoxemia and hypotension unless they were given supplementary O2. Thus the combination of load-induced hypoxemia and increased metabolic demands can injure the heart.
In the present study, cTnT was always present in blood when the load was terminated, indicating the presence of cardiac injury. The load-induced hypotension must have had a cardiac component because the death of any fibers will, since the heart is a syncitium, impair contractile function when compensatory mechanisms are exhausted. Any impairment would have been exacerbated by the prevailing hypoxemia and hypercapnia/acidosis, both of which impair cardiac function (e.g., 17), especially because the heart's O2 extraction is already high. The heart relies almost exclusively on aerobic metabolism and can only develop a small O2 debt while still maintaining normal function. The elevation of cardiac troponin unequivocally indicates injury, but the exact mechanism(s) and location (e.g., right vs. left ventricle) require further investigation. Thus the most likely cause of the hypotension is load-induced cardiac injury due to increased myocardial O2 demand in a setting of reduced O2 supply. This concept is consistent with recent clinical reports indicating that respiratory loads can induce cardiac damage. For example, cTnI was elevated in a patient with acute severe bronchospasm with no evidence of coronary artery disease (5). Some patients with chronic obstructive pulmonary disease but no evidence of myocardial ischemia present with elevations in cTnI or cTnT (8, 22). Collectively, these results suggest that respiratory loads can and do cause cellular damage sufficient to allow the release of intracellular proteins from the heart and support our contention that loading of the respiratory system can, at least under some circumstances, cause myocardial injury.
Summary statements. The ability of anesthetized rats to generate pressure against an IRL depends on the prevailing MAP. Since this peak pressure was used to determine the size of the load, the MAP, along with the degree of hypercapnia caused by loading, also determined load endurance. In our model, loading caused rapid hypercapnic failure and diaphragmatic fatigue that coincided with injury of fast-twitch, presumably diaphragmatic, fibers as indicated by the appearance of only fast sTnI in blood. Eventual respiratory pump failure occurred only after arterial hypotension that preceded a second fall in diaphragmatic contractility. The release of cTnT into the blood indicates that respiratory load-induced myocardial injury can play an important, yet underappreciated, role in the development of respiratory pump failure.
|
GRANTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. A. Simpson, K. R. Brunt, C. P. Collier, and S. Iscoe Hyperinflation-induced cardiorespiratory failure in rats J Appl Physiol, July 1, 2009; 107(1): 275 - 282. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Simpson, K. R. Brunt, and S. Iscoe Repeated inspiratory occlusions acutely impair myocardial function in rats J. Physiol., May 1, 2008; 586(9): 2345 - 2355. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
) or nominal (
) fall in mean arterial pressure (MAP) during a 30-s occlusion. Rats demonstrating a modest fall in MAP during the 30-s occlusion started with a significantly lower MAP and generated significantly greater tracheal pressures (Ptr). A: time course of changes in Ptr. B: time course of changes in MAP. *P = 0.008; 
P < 0.001, unpaired t-test.
) and IRL (
) and sham-operated (