点击显示 收起
ABSTRACT |
---|
TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES |
---|
Key Words: ventilator weaning • respiratory muscles • muscle fatigue • phrenic nerve • respiratory insufficiency
INTRODUCTION |
---|
TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES |
---|
Patients who fail a trial of weaning from mechanical ventilation are at a considerable risk for developing respiratory muscle fatigue. Much circumstantial evidence suggests that these patients do develop contractile fatigue (10–15). For example, we found that 5 of 17 weaning failure patients experienced an imbalance between mechanical load and respiratory muscle capacity, expressed as a tension–time index (TTdi), which would be expected to induce contractile muscle fatigue (13). None of the experimental techniques used to date, however, provide direct evidence of contractile fatigue, and it is still unknown whether contractile fatigue occurs in patients.
The question of whether respiratory muscle fatigue occurs in weaning failure patients is of major clinical importance. Patients who fail a trial of weaning are at a disadvantage when compared with weaning success patients because they have greater abnormalities in lung mechanics (13–15). If these patients also develop contractile fatigue of their respiratory muscles during a failed weaning trial, the superimposed structural injury is likely to set them back in their clinical course. The new injury might even become the ultimate determinant of whether some patients are ever successfully weaned from the ventilator.
The most direct method for detecting fatigue in patients is to stimulate the phrenic nerves and measure the resulting change in transdiaphragmatic pressure. We used this approach to test the hypothesis that patients who fail a trial of weaning from the mechanical ventilation develop contractile fatigue of the diaphragm, whereas successfully weaned patients do not. Some the results of these studies have been previously reported in the form of abstracts (16, 17).
METHODS |
---|
TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES |
---|
|
Compound diaphragmatic action potentials.
Compound diaphragmatic action potentials were recorded bilaterally using surface electrodes. Single bilateral phrenic nerve stimulation was performed using two magnetic stimulators with two sets of double 40-mm coils that generated a magnetic field of 3.2 Tesla at maximal output (see online supplement for further details).
Protocol
Twitch stimulation before the weaning trial.
To avoid twitch potentiation (5, 18), the patients received controlled ventilation for 20 minutes before delivering the first stimulation. Six to 15 stimulations were then delivered bilaterally at intervals of approximately 15 seconds at end exhalation while closing the inline valve. After the last stimulation, maximal voluntary inspiratory efforts were recorded during a 20-second occlusion of the airway (19) (see online supplement for details).
Trial of spontaneous breathing.
The weaning trial was conducted for up to 1 hour as tolerated. In eight weaning failure and four weaning success patients, arterial blood samples were obtained at 2 minutes and at the end of the trial. Patients who did not develop criteria of weaning failure (13, 20) (see online supplement for criteria) were extubated. Patients who sustained spontaneous breathing for more than 24 hours were deemed the weaning successes group (21). The remaining patients, the weaning failure group, required mechanical ventilation for 3 days to more than 57 days after the study.
Twitch stimulation after the trial.
After the weaning trial, all patients, irrespective of the weaning outcome, were returned to mechanical ventilation for 30 minutes. At the end of this period, twitch Pdi and maximal voluntary inspiratory efforts were measured. To determine whether maximal depolarization of the phrenic nerve was achieved, progressively increasing outputs from the stimulator were delivered in 12 patients (see online supplement for details). Thereafter, patients who had met the a priori criteria of weaning failure were maintained on mechanical ventilation, whereas the remaining patients were extubated.
Twitch interpolation.
Twitch interpolation measurements were performed in seven patients before and after the weaning trial (22) (see online supplement for details).
Physiologic Measurements
Transdiaphragmatic twitch pressure.
Twitch Pdi was measured as the difference between the maximum Pdi displacement secondary to phrenic nerve stimulation and the value immediately before stimulation (18, 23). Criteria for acceptable twitch responses are listed in the online supplement. The within-occasion coefficient of variation of twitch Pdi before and after weaning was 10% or less in 17 patients, and it was within 12 and 14% in the remaining 2 patients. Data collected after the weaning trial did not satisfy the a priori criteria for acceptable twitch responses in 3 of the 19 patients, and they were excluded from data analysis.
Maximum voluntary inspiratory pressure.
The pressure developed by the diaphragm was computed as the maximal excursion in Pdi (Pdimax) during the 20-second occlusions (19).
Respiratory mechanics and effort indices.
Inspiratory resistance of the lung, dynamic compliance of the lung, and intrinsic positive end-expiratory pressure (PEEPi) were computed according to standard formulae (24–27). Pressure-time product (PTPdi) and TTdi of the diaphragm were quantified using standard formulae (1, 26–28). The relative contribution of the different respiratory muscles to tidal breathing was assessed as the ratio of swings in Pga to swings in Pes (Pga/Pes) (see online supplement for calculation of these variables).
Data Analysis
Analysis of variance and t tests were used as needed (see online supplement). Some measurements are included in another manuscript (29).
RESULTS |
---|
TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES |
---|
At the beginning of the weaning trial, PaO2, PaCO2, and pH were not different between the groups . By the end of the trial, a small decrease in pH (p = 0.025) occurred in the failure group, and no significant change occurred in the success group.
|
|
|
|
|
|
Ribcage and Expiratory Muscle Recruitment
At trial onset, the Pga/Pes ratio was greater in the failure group than in the success group: -0.03 ± 0.04 versus -0.19 ± 0.05 (p = 0.002)
. Over the course of the trial, Pga/Pes remained greater in weaning failure patients (p = 0.004). At the end of the trial, the ratio had increased to 0.12 ± 0.07 in the failure group (p = 0.05). At the end of the trial, the ratio in the success group was -0.10 ± 0.02. Because patients with diaphragmatic paralysis can have enhanced rib cage muscle recruitment even in the absence of respiratory distress (30), Pga/Pes ratio of the failure patients was compared with that of the success patients after excluding the two patients (both weaning failure patients) with hemidiaphragmatic paralysis. The Pga/Pes was still greater in the weaning failure patients (p = 0.01)—a finding similar to the overall group.
|
At onset of the trial, total PEEPi (not corrected for expiratory rise in Pga) was not different between the failure and success groups: 4.6 ± 1.3 versus 2.1 ± 0.4 cm H2O (p = 0.13). Likewise, corrected PEEPi (corrected for expiratory rise in Pga) was not different between the failure and success groups: 3.8 ± 1.2 versus 2.1 ± 0.4 cm H2O (p = 0.59). At the end of the trial, total PEEPi had increased to 11.6 ± 4.4 cm H2O (p = 0.03) in the failure group and to 4.4 ± 1.0 cm H2O (p = 0.001) in the success group. At the end of the trial, corrected PEEPi was 6.2 ± 3.6 cm H2O in the failure group. The values of total and corrected PEEPi in the success patients were nearly identical.
DISCUSSION |
---|
TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES |
---|
Researchers have long thought that some, if not most, patients who fail a weaning trial develop respiratory muscle fatigue (10, 12–15). The techniques used in previous studies were indirect (10–14), raising doubt as to whether fatigue truly occurred (31). We used a direct test of muscle fatigability, namely stimulation of the phrenic nerve and measurement of the resulting Pdi (32, 33). Even with this technique, data can be inaccurate because of several confounding variables: changes in lung volume, variation in the degree of neural depolarization achieved by the stimulator, and the contraction history of the muscle (i.e., twitch potentiation). We took particular care to avoid these confounding factors and excluded data that did not satisfy our a priori inclusion criteria. As such, we view the absence of a fall in twitch Pdi as evidence that low-frequency fatigue is not a mechanism of weaning failure.
Factors That Can Contribute to Fatigue
Diaphragmatic fatigue occurs only when there is a critical stress on the muscle. A critical stress can result from an increase in mechanical load, which leads to an increase in respiratory center output and thus increase in respiratory muscle pressure (PTPdi). When a patient's muscle strength is small, an increase in PTPdi is more likely to exceed the diaphragmatic threshold (TTdi) for fatigue (1).
At onset of the weaning trial, the weaning failure patients had abnormalities in lung mechanics comparable to those reported in previous studies: inspiratory resistance of the lung was 13 cm H2O/L/second versus 9 (13) and 22 cm H2O/L/second (14); dynamic lung compliance was 55 ml/cm H2O versus 48 (13) and 70 ml/cm H2O (14), and total PEEPi was 4.6 cm H2O versus 2.0 (13) and 5.9 cm H2O (14). By the end of the trial, inspiratory resistance of the lung (17 cm H2O/L/second) and total PEEPi (12 cm H2O) increased to or exceeded previously reported values (16 cm H2O/L/second and 4 cm H2O, respectively [13]). Dynamic lung compliance at the end of a failed trial was within the range of previously reported values: 48 ml/cm H2O versus 29 (13) and approximately 70 ml/cm H2O (14). Accordingly, our weaning failure patients displayed abnormalities in pulmonary mechanics equivalent to patients in previous studies. Our weaning failure patients displayed greater resistive and elastic loads than did our weaning success patients—a finding similar to our previous report (13).
The increase in PTPdi over the course of the weaning trial indicates that the respiratory centers attempted to defend alveolar ventilation in the face of deteriorating lung mechanics. This finding is consistent with our previous report of an increase in overall respiratory muscle pressure (esophageal pressure-time product ) (13). In that previous study (13), the increase in PTPes was not sufficient to prevent hypercapnia in 13 of the 17 weaning failure patients. Hypercapnia was less common in this study: PaCO2 increased by 5 to 16 percent in three of eight weaning failure patients in whom arterial blood samples where obtained. Three factors may account for the difference in the two studies: baseline arterial samples were collected at approximately 2 minutes after the start of spontaneous breathing in this study, whereas they were collected during controlled mechanical ventilation in the previous study; only four weaning failure patients in this study had COPD as compared with all patients in the previous study, and respiratory muscle effort, as reflected by PTPes, was somewhat greater in this study than in the previous study (538 and 388 cm H2O x seconds/minute, respectively).
TTdi combines three key determinants of diaphragmatic fatigue: pressure generated by the diaphragm (PTPdi), muscle strength (Pdimax), and respiratory duty cycle (inspiratory time/total respiratory cycle time). In healthy subjects, a sustained increase in TTdi above 0.15 leads to diaphragmatic fatigue (1). The threshold of 0.15 was exceeded by 77% of our weaning failure patients and by 15% of our weaning success patients. However, no patient showed evidence of low-frequency fatigue. Three factors could explain why a high TTdi was not accompanied by fatigue: the heightened muscle effort was not sustained for a sufficient time; endurance of the diaphragm was greater in weaning failure patients than in healthy subjects; and the recorded value of TTdi was an overestimate.
Bellemare and Grassino (1) reported that the relationship between TTdi and time to task failure in healthy subjects follows an inverse power function: time to task failure = 0.1 (TTdi)-3.6. The average duration of weaning trials in our failure patients was 44 minutes. The average values of TTdi for the first, second, third, and fourth quintiles of the trial durations were 0.17, 0.17, 0.22, and 0.22, respectively. Based on the formula of Bellemare and Grassino (1), the expected times to task failure for the respective quintiles would be 59, 59, 28, and 28 minutes. The average value of TTdi during the last minute of the trial was 0.26, and the weaning failure patients would be predicted to sustain this effort for another 13 minutes before developing task failure. These calculations suggest that weaning failure patients did not sustain the increase in load for a duration sufficient to cause low-frequency fatigue; that is, the trial was stopped because patients developed a priori–defined clinical manifestations of respiratory distress before they developed fatigue.
Fatigue is not an all-or-none phenomenon (34). Measuring twitch pressures after forceful voluntary contractions, so-called potentiated twitches, has been suggested as a means for detecting an early decrease in muscle contractility (34, 35). In eight patients (four weaning failure and four weaning success patients), we were able to record the potentiated twitches both before and after weaning by stimulating the phrenic nerves immediately after the patients performed maximal voluntary inspiratory efforts. Potentiated twitch Pdi was 8.6 ± 3.1 cm H2O before the trial and 8.7 ± 2.9 cm H2O after the trial in the four weaning failure patients; the corresponding values for nonpotentiated twitches were 6.9 ± 2.6 and 6.8 ± 2.4 cm H2O, respectively. Potentiated twitch Pdi was 10.5 ± 1.5 cm H2O before the trial and 10.6 ± 1.1 cm H2O after the trial in the four weaning success patients; the corresponding values for nonpotentiated twitches were 8.3 ± 1.5 and 8.7 ± 1.4 cm H2O, respectively. The failure of potentiated twitch Pdi to decrease after a failed weaning trial further supports our reasoning that the inspiratory load, even if it was in the fatiguing range, was not sustained for a sufficient length of time to cause fatigue.
If endurance of the respiratory muscles is supranormal in critically ill patients, fatigue would not develop at a TTdi of 0.15. Direct measurements of diaphragmatic endurance have not been obtained in critically ill patients, but circumstantial evidence suggests that it is not supranormal. Indeed, endurance of the diaphragm is decreased in stable patients with spinal cord injury (36), probably because fatigue-sensitive, type II myosin heavy chains are increased in the diaphragm (37).
Reliable calculation of TTdi is critically dependent on an accurate measurement of diaphragmatic strength. Our data show that even carefully made measurements of Pdimax commonly underestimate maximum strength. The pressure tracings in all of our patients during the Pdimax measurements had the characteristics of a Mueller maneuver: large negative excursions in Paw and Pes with slightly positive (or, in one patient, negative) deflections in Pga . In healthy subjects, the combination of a Mueller maneuver with an expulsive maneuver results in higher values of Pdimax (38), but critically ill patients have great difficulty in performing the combined maneuver. Moreover, the finding of twitch interpolation (that is, a measurable twitch Pdi when the phrenic nerves were stimulated during maximum voluntary effort) indicates that patients were not able to activate completely the diaphragm during a "maximum" maneuver (22) . The underestimation of Pdimax will necessarily produce an overestimate of TTdi, which further explains why patients did not develop low-frequency fatigue despite recorded values of TTdi above 0.15.
Defense Mechanisms against Low-Frequency Fatigue
All of the weaning failure patients experienced severe respiratory distress, but none developed low-frequency fatigue. Three strategies may have protected the diaphragm against fatigue: increased rib cage and expiratory muscle recruitment, the early reinstitution of mechanical ventilation, and respiratory center downregulation.
Immediately after the start of the weaning trial, the weaning failure patients displayed greater recruitment of rib cage and expiratory muscles during tidal breathing than did the weaning success patients (greater Pga/Pes) (Figure 6). (Excluding the two patients with hemidiaphragmatic paralysis does not affect this finding.) The same alteration in respiratory muscle recruitment has also been reported in patients (39–42) and volunteers (34, 43) when diaphragmatic effort is increased during tidal breathing. Recruitment of the rib cage and expiratory muscles appears to contribute to the development of dyspnea (41, 44). Clinicians also take increased activity of the rib cage and abdominal muscles into account in deciding whether to continue or interrupt a weaning trial.
The increase in PTPdi during the weaning trial signifies a progressive increase in respiratory motor output, as has been previously reported (14, 45, 46). Some patients, however, developed hypercapnia, suggesting that respiratory motor output may have been downregulated. Studies in animals show that a decrease in respiratory motor output occurs as a preterminal event (47) and a decrease in drive can be accompanied by the development of diaphragmatic fatigue at the time of apnea (48). Afferent signals originating in fatiguing muscles may activate neural pathways responsible for downregulation of respiratory motor output (49–52). Downregulation of respiratory drive will decrease metabolic demands and the likelihood of contractile fatigue (53). During the usual protocol for inducing respiratory muscle fatigue in healthy volunteers (achieving a target inspiratory pressure while breathing through a resistor [1, 5, 9]), the exhortation of the investigators and the volition of the subjects may override the afferent signals that downregulate respiratory drive (9). The artificial and constrained nature of this laboratory protocol is very different from the natural evolution of respiratory distress in weaning failure patients.
Other Causes of Weaning Failure
Although low-frequency fatigue does not appear to be responsible for weaning failure, other abnormalities of the respiratory muscles may be causative. Possible mechanisms include diaphragmatic weakness, atrophy, high-frequency fatigue, and hyperinflation.
In our laboratory, the mean amplitude of nonpotentiated twitch Pdi ranges from 35.4 to 38.9 cm H2O in healthy subjects (5, 23, 34) and from 17.2 to 20.1 cm H2O in stable patients with COPD (54, 41). Most of our weaning failure and weaning success patients had twitch Pdi values lower than the values recorded in ambulatory patients. Six weaning failure patients had twitch Pdi values of less than 10 cm H2O. These results suggest that many mechanically ventilated patients have diaphragmatic weakness and that some weaning failure patients have severe weakness.
Eight weaning failure patients and six weaning success patients had infections (pneumonia or sepsis) while receiving ventilator support. Sepsis is known to cause diaphragmatic injury and weakness (55, 56). All of our patients had been ventilated with patient-triggered modes, and the involved muscle contractions may aggravate the diaphragmatic injury caused by sepsis (57). Several studies in experimental animals (58) demonstrate that mechanical ventilation can induce respiratory muscle atrophy, although it is not known whether this occurs in patients. Many ventilator-supported patients are malnourished (59), and this will further contribute to atrophy (60–62).
Our study does not directly address whether the patients developed high-frequency fatigue, as has been suggested (11). The lack of change in twitch Pdi and Pdimax does not exclude the possibility. High-frequency fatigue can resolve within 10 to 15 minutes (4, 53), and it could have disappeared by the time of testing (30 minutes after the end of the weaning trial).
Weakness of the inspiratory muscles arises when patients develop progressive hyperinflation. An increase in end-expiratory volume causes shortening of inspiratory muscles and a decrease in force generation. Total PEEPi increased over the course of the trial in the weaning failure patients, but an indirect measurement of end-expiratory volume—PEEPi corrected for expiratory muscle recruitment (26)—revealed no change. Corrected PEEPi was also not different between weaning success and weaning failure patients, suggesting that hyperinflation did not increase during the trials.
Clinical Implications
Respiratory muscle fatigue has been thought to be a common cause of weaning failure, and accordingly, clinical management has been directed toward improving the capacity of the respiratory muscles to generate (strength) and to sustain (endurance) force (63). Does the lack of low-frequency fatigue in our weaning failure patients mean that these strategies are misdirected? No. Many weaning failure patients have severe diaphragmatic weakness (66% had twitch Pdi values below 10 cm H2O), and weakness probably sets in motion the complex processes described in this study, which ultimately lead to the early reinstitution of mechanical ventilation.
Investigators have shown that respiratory muscle training can increase respiratory muscle strength and endurance in ambulatory patients, but the improvement has not been shown to achieve better clinical well-being or outcome (64, 65). The lack of benefit is not surprising because baseline maximum inspiratory pressure was not reduced to a level that hinders spontaneous breathing (66). In weaning failure patients, however, a small improvement in respiratory muscle strength and endurance could have a profound effect on clinical outcome. Based on the results of this study, it could prove useful to develop a training regimen that can achieve an improvement in twitch pressure (the noninvasive measurement of twitch airway pressure may satisfactorily substitute for twitch Pdi [29]). Such a training regimen could have a major benefit in the difficult-to-wean patients (63), although proof of this possibility requires a randomized control trial. Although the challenge of designing and undertaking such a trial will be considerable, the scientific motivation for such a study is stronger than before.
In summary, patients who fail a weaning trial displayed greater mechanical load than did weaning success patients. The increase in load caused TTdi to increase above the threshold associated with fatigue, yet twitch Pdi and Pdimax did not decrease in these patients. Factors that may have protected the diaphragm against fatigue include greater rib cage and expiratory muscle recruitment, downregulation of respiratory motor output, and early reinstitution of mechanical ventilation. In conclusion, in contrast to our hypothesis, weaning failure was not accompanied by low-frequency fatigue of the diaphragm, although many weaning failure patients displayed severe diaphragmatic weakness.
REFERENCES |
---|
TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES |
---|