INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The measurement of transdiaphragmatic pressure (Pdi) in response to bilateral supramaximal electrical phrenic nerve stimulation (E_S) is the gold standard of nonvolitional tests of diaphragm properties (1, 2). However, the E_S-Pdi combination has practical limitations that restrict its use in the clinical field. A highly expert operator is required to achieve reliable and reproducible stimulations, and even then there is a risk of false negative results. In addition, the costimulation of the brachial plexus may be impossible to avoid and may generate discomfort. Finally, there is discomfort from the stimulus intensities required to achieve supramaximal E_S, from the pressure of the electrodes on the soft tissues of the neck, and from the balloon-catheter technique. Conversely, mouth pressure (Pm) is very easy to measure and accurately reflects pleural pressure, provided that the time constant for equilibration of alveolar and mouth pressure is sufficiently short. Thus, if the diaphragm contracts independently of other inspiratory muscles (as is the case with E_S), Pm twitch (Pm,t) becomes an index of the proportion of diaphragm contraction that is transformed into inspiratory driving pressure. This information is highly relevant, and E_S-Pm,t as an index of diaphragm properties has been validated in normal control subjects (3) and in patients with chronic obstructive pulmonary disease (4).

Pm,t can also be obtained by cervical magnetic stimulation (CMS) (5, 6), which is easier to use and better tolerated by the patient (7, 8). The CMS-Pm combination is attractive for studying large numbers of subjects, patients in difficult settings, or for sequential studies. However, unlike E_S, CMS does not induce isolated diaphragm contractions. It coactivates other muscles, including inspiratory neck muscles (INM) (9). As a result, the rib cage is stabilized and better resists distortion (9, 10), which explains higher Pdi outputs with CMS (8). It is important to note that the contribution of the CMS-induced INM contraction to Pdi is an indirect one. A hypothetical esophageal pressure (Pes) response to this contraction (Pes,t-INM) should be transmitted across the diaphragm to the abdominal compartment without producing Pdi,t. In the absence of diaphragm contraction Pdi,t should thus be zero even if the response of INM to CMS does indeed produce a Pes swing. This means that what is measured by the CMS-Pdi combination actually pertains to intrinsic diaphragm function, even though the meaning of the data is not exactly the same as with the E_S-Pdi combination because of different rib cage properties at the time of diaphragm contraction. If Pm is studied in response to CMS instead of Pdi, the situation is different. Theoretically, Pm,t should be the sum of the diaphragm and an INM component (Pm,t-di and Pm,t-inm, respectively) and therefore CMS-Pm,t should be an index of global inspiratory muscle contractility not specific for the diaphragm. Although Mills and coworkers (11) have shown that CMS-Pm,t-INM is negligible in patients with recent diaphragmatic paralysis and without chronic respiratory disease, in chronic diaphragm dysfunction compensatory neck muscle hypertrophy could lead to overestimation of diaphragm action by CMS-Pm,t.

To address this question, the present study was conducted in patients with amyotrophic lateral sclerosis (ALS), a fatal neurodegenerative disorder altering spinal motor neurones, motor nuclei of the lower brainstem, and upper motor neurones of the motor cortex. Patients with ALS combine amyotrophy and signs of corticospinal tract dysfunction. No treatment is effective except for the recent demonstration of activity of an antiglutamate agent, riluzole (12). The respiratory muscles are often involved at various stages of ALS, leading to respiratory insufficiency and death. To specifically address the possible contribution of inspiratory neck muscles to the mouth pressure response to CMS, we selected patients with ALS having signs of major diaphragm dysfunction and apparent hypertrophy of inspiratory neck muscles for study. Our results indicate that negative CMS-Pm,t can be observed despite diaphragm paralysis.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patients

Nine patients (seven men and two women, 47 to 74 yr of age) with definite ALS (13), dyspnea, and signs suggestive of diaphragm paralysis on respiratory clinical examination (14), were enrolled in the study (Table 1). At the time of the study, the mean duration of disease was 25.3 ± 21.7 mo (range, 4 to 75 mo). All patients were severely disabled and presented signs of bulbar involvement (limb and bulbar functional score, 34.8 ± 16.7 and 28.5 ± 6.6, respectively; maximal values, 63 and 39, respectively) (15). Their body mass index was low (22.4  ± 2.5; range, 18.2 to 25.7). All complained of dyspnea. In addition, three patients with ALS (two men, one woman) with similar disability and bulbar signs were studied while mechanically ventilated through a cuffed endotracheal tube or tracheostomy cannula. These patients were 52, 63, and 68 yr of age, and disease duration at the time of the study was 18, 22, and 26 mo.

To study CMS-Pm,t in patients with diaphragm paralysis but in the absence of neck muscle hypertrophy, we also studied two men (34 and 55 yr of age) and 1 woman (49 yr of age) referred to the laboratory for suspected acute bilateral diaphragm paralysis (stretch trauma of the neck in one case, recent cardiac surgery in the second, and no particular clinical context in the third). In these patients, there was no evidence of underlying respiratory or neurologic disease.

Our laboratory has been granted Ethics Committee approval for the use of the noninvasive techniques involved by the study in patients with neurologic disease for physiopathologic purposes. All the patients were informed of the purpose of the study and the methods used, and they gave their consent.

Measurements

Clinical examination and pulmonary function tests. A detailed clinical respiratory examination (14), a semiquantitative evaluation of the intensity of dyspnea using the modified Borg scale (16), and standard pulmonary function tests (Pulmonet III^®; Gould, Cleveland, OH) were performed during the week preceding the respiratory muscle tests.

Mouth pressure (Pm). Pm was measured using a linear differential pressure transducer (Validyne, Northridge, CA) (range 0 to 150 cm H₂O for static pressures; range 0 to 50 cm H₂O for twitch pressures) attached to the side tap of a cylindrical rubber mouthpiece.

Transdiaphragmatic pressure (Pdi). Pdi was not measured in the nine spontaneously breathing patients with ALS because they had major deglutition impairment resulting from bulbar involvement. In the three subjects referred for suspected acute diaphragm paralysis and in the three mechanically ventilated patients with ALS, Pdi was obtained using two polyethylene catheters (1.7 mm ID) with distal side holes. One catheter was placed in the midesophagus to measure Pes and the other in the stomach to obtain gastric pressure (Pga). Both were connected to Validyne differential pressure transducers (0 to 150 cm H₂O).

Rib cage (RC) and abdominal (AB) displacements. Qualitative RC and AB displacements were obtained using mechanical strain gauges (Nihon Kohden, Tokyo, Japan), which measured RC and AB circumferences. These consisted of two piezo-electric sensors encased in a molded box 3 × 3 cm and attached to elastic belts. One belt was placed as high as possible around the upper part of the rib cage and the other was placed around the abdomen at the level of the umbilicus (Figure 1). They were not calibrated in absolute terms but were cross-calibrated (isovolume maneuver). Similar devices have been used in other studies to examine changes in RC and AB dimensions in response to phrenic nerve stimulation (9, 10).

View larger version (25K): [in this window] [in a new window]   Figure 1.   Schematic representation of the experimental setup used to study the patients. Diaphragmatic electromyogram (EMGdi), mouth pressure (Pm), and rib cage (RC) and abdominal (AB) displacements were measured in response to cervical magnetic stimulation and to transcutaneous electrical stimulation of the phrenic nerve in the neck.

Electromyograms (EMG). Surface recordings of right and left costal diaphragm EMG were obtained using disposable silver cup electrodes taped to the skin on the anterior axillary line in the sixth to eighth right and left intercostal spaces and connected to a electromyograph (Neuropack Sigma^®; Nihon Kohden). The motor response to phrenic stimulation is henceforth termed M-wave and abbreviated M-w.

In three patients with ALS, similar surface electrodes were placed over the muscle mass of the sternomastoid to record a possible phasic activity during spontaneous breathing.

Stimulations. E_S was attempted by a trained operator using the classic technique (1, 17). Bipolar electrodes with saline-soaked felt tips 5 mm in diameter were used. The interelectrode distance was 2 cm, and 0.1 ms square-wave pulses of adjustable intensity were used for stimulation. The operator attempted to stimulate the phrenic nerve, beginning from a starting point at the posterior border of the sternomastoid muscle at the level of the cricoid cartilage. The maximal trial-and-error time allowed on each side was 30 min.

CMS was performed using an upgraded Magstim 200 stimulator equipped with a doughnut-shaped 90-mm coil (Magstim; Whitland, Dyfed, UK) and employed a previously reported technique (7) (Figure 1). Briefly, the stimulating coil was centered on the spinous process of the seventh cervical vertebra. The magnetic field generated by the coil is currently thought to stimulate the phrenic nerves anteriorly, in the very initial segment of their intrathoracic path (18). The maximal output of the stimulator was always used. All patients, except the three who were mechanically ventilated, were studied seated on a chair equipped with headrests and with their abdomens unbound. The three ICU patients were studied in a semirecumbent position. All stimuli were delivered at end-expiration, after rapid occlusion of the mouthpiece. The reported results are an average of three to five stimulations.

Procedures

(1) Dynamic RC and AB displacements during quiet breathing were observed first; the subjects were then asked to perform deep inspirations and to sniff sharply. These maneuvers permitted us to assess diaphragm movement over a range of neural outputs.

(2) Maximal inspiratory mouth pressures (PImax) were measured at RV and maximal expiratory mouth pressures (PEmax) at TLC.

(3) In order to prevent voluntary maneuvers from affecting the results (twitch potentiation), the patients were instructed to relax and breathe quietly for 15 to 30 min before the subsequent phrenic nerve stimulation (19).

Data Analysis

Phrenic nerve conduction time (PNCT) was defined as the time elapsed between the stimulus and the onset of M-w. M-w amplitudes were measured from peak to trough. Pdi was reconstructed off-line using electronic subtraction of Pes from Pga. The amplitude of Pdi and Pm twitches was measured from baseline to peak. Values presented in RESULTS are mean ± SD.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Pattern of Breathing and Respiratory Physical Examination

AB paradox during tidal breathing was observed in all spontaneously breathing patients. In six, it was constant, whereas in the remaining three, it alternated with periods of normal AB motion. Forced inspiratory maneuvers and sniffs were consistently associated with AB paradox in these three patients. On inspection, neck muscles were hypertrophic in all the patients and exhibited strong phasic inspiratory contractions when palpated (respiratory pulse) that were often associated with retraction of the clavicle. Surface electrodes placed over the sternomastoid muscles confirmed the phasic inspiratory activity (Figure 2).

View larger version (16K): [in this window] [in a new window]   Figure 2.   Representative tracings from one patient with ALS showing EMG activity recorded by surface electrodes placed over the muscle mass of the sternomastoid (EMGsm) (top trace). The two bottom traces depict changes in rib cage (RC) and abdominal (AB) circumferences, respectively, during spontaneous breathing (left) and during a sniff maneuver (right). Inspiratory AB paradox and an important phasic inspiratory EMGsm activity are visible.

Pulmonary Function Tests and Static Pressures

On average, VC was 62.0 ± 20.32% pred (range, 27 to 98%), and it was associated with a normal FRC and a slightly increased RV (117.2 ± 27.3% pred). The overall spirometric pattern was that of a restrictive syndrome of neuromuscular origin, and it showed a marked heterogeneity between patients (Table 2). Predicted values for spirometric data are from the European Economic Community recommendations (20).

PImax and PEmax were dramatically reduced (Table 2). At RV, PImax was 38.7 ± 9.1% pred (range, 20 to 50%), whereas at TLC, PEmax was 41.3 ± 18.5% pred (range, 15 to 70%). Predicted values for static pressure data are taken from reference (21).

Phrenic Nerve Stimulation

In seven patients, it was impossible to locate the phrenic nerve with E_S within the preset limit of 30 min. In two patients, a small M-w was occasionally observed on either side. These responses were difficult to reproduce.

With CMS, M-w was present bilaterally in three patients and on the left side alone in two. CMS-M-w was of very small amplitude in all patients, being hardly discernible in the two patients where it was unilateral. Despite this low amplitude, CMS-M-w proved easy to reproduce. The average PNCT was 9.4 ± 1.4 ms (range, 7.8 to 10.9), at least 50% longer than expected in normal subjects with this technique (18). CMS was consistently associated with inward AB and outward RC motions of comparable amplitudes (Figure 3). A negative, twitch-shaped mouth pressure swing was consistently observed in response to CMS (Pm,t, 2.6 ± 1.1 cm H₂O; range, 1 to 4.5 cm H₂O (Table 2 and Figure 3).

View larger version (12K): [in this window] [in a new window]   Figure 3.   Representative tracings obtained in a patient with ALS. The two upper traces correspond to changes in rib cage (RC) and abdominal (AB) circumferences in response to cervical magnetic stimulation, whereas the lower trace corresponds to the mouth pressure (Pm) response. A negative Pm swing was observed despite an RC-AB pattern of motion suggestive of diaphragm paralysis.

Additional Patients

In the three mechanically ventilated patients with ALS, CMS also induced RC expansion and AB paradox. This was associated with a negative Pes twitch (2, 3, and 3.5 cm H₂O). In all cases, this Pes twitch was matched to a Pga twitch of similar time course and amplitude; hence, the absence of change in Pdi. In two patients, CMS did not evoke any visible diaphragmatic EMG response, whereas in one, a small M-w was present but markedly delayed on both sides (right and left PNCT, 9.9 and 11.2 ms, respectively).

The three patients with suspected acute bilateral diaphragm paralysis exhibited AB paradox during quiet breathing. Their VC was reduced by about 30%, and it decreased further in the supine position. They did not have visible neck muscle hypertrophy on inspection. E_S and CMS failed to elicit M-w in all patients and CMS-Pdi,t and CMS-Pm,t were both undetectable. A minor increase in RC diameter was observed in response to CMS, but there was no visible change in AB diameter.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The major finding of this study was the observation that CMS can produce Pm twitches despite diaphragm paralysis. This indicates that under certain circumstances the contraction of extradiaphragmatic muscles coactivated with the diaphragm by CMS can directly produce inspiratory pressure in addition to stabilizing the rib cage.

Diaphragm Paralysis

For this assumption to be valid, it is crucial to rule out any contribution of the diaphragm to the negative mouth pressure twitch observed in response to CMS in our patients with ALS. In the main group of nine patients with ALS, severe diaphragm dysfunction was suggested by the presence of a characteristic AB paradox during both spontaneous breathing and voluntary inspiratory maneuvers. RC expansion and AB paradox in response to phrenic stimulation attested to the fact that the spontaneous breathing pattern was a result of intrinsic diaphragm abnormalities rather than a lack of transmission of respiratory commands. In this case, AB paradox would have been expected during automatic or volitional breathing but phrenic stimulation would have elicited AB expansion. Although CMS at times produced some EMG activity, several lines of evidence support the fact that the patients had either complete functional diaphragm paralysis or diaphragm dysfunction severe enough to reasonably exclude a diaphragmatic contribution to the negative CMS-Pm,t. First, in four of the nine patients with ALS constituting the main group the lack of any diaphragmatic EMG response to CMS combined with the observation of stimulation-induced abdominal paradox seems to make the diagnosis of diaphragm paralysis unquestionable. In these four patients, CMS-Pm,t was not different from the average value observed in the study group (2 cm H₂O in Patient 1, 3 cm H₂O in Patient 2, and 5, and 2.5 cm H₂O in Patient 9) (see Table 2). Similarly, two of the three mechanically ventilated patients with ALS showed AB paradox in response to CMS, an absence of EMG response, and a negative CMS-Pes,t with a magnitude of 2 and 3.5 cm H₂O. Second, in the patients with ALS in whom CMS was associated with an electromyographic response (five patients in the main group of nine, one patient among the three mechanically ventilated ones), there were elements suggestive of diaphragm denervation atrophy. The EMG response was always of very small amplitude and often hardly discernible. PNCTs were abnormally prolonged, with values 50 to 100% greater than the normal values commonly found in our laboratory with the same technique (18). Such an increase in PNCT probably denotes extremely severe denervation atrophy (22). In this context, the abdominal paradox elicited by CMS attests to the fact that the action of neck muscles was paramount in producing negative intrathoracic pressure and makes it very probable that the electrical activation of the diaphragm, which could be termed "residual," did not correspond to an adequate inspiratory action. Finally, we are aware that systematic Pdi measurements would have provided an unambivalent certitude about the reality of diaphragm paralysis. However, this could not be done in the patients with ALS who were spontaneously breathing, because they had major deglutition impairment resulting from bulbar involvement of the disease, which made use of the balloon-catheter technique both difficult and unethical. Pdi measurements were performed in three other patients with ALS in whom the airway protection provided by the cuffed endotracheal prosthesis used for mechanical ventilation alleviated this concern. The data, although limited in number, supports our hypothesis. Indeed, Pdi was zero in all cases despite a negative esophageal pressure (Pes) swing, comparable in amplitude to the CMS-Pm,t described in the nonintubated patients. In one patient, this was true despite a persistent EMG response to CMS, substantiating the possibility of diaphragmatic "electromechanical dissociation" secondary to denervation atrophy. Therefore, we feel confident that RC expansion and AB paradox in response to CMS are sufficient arguments overall to consider that the diaphragm did not contribute to the corresponding inspiratory pressure.

Neck Muscle Hypertrophy

With regard to neck muscle hypertrophy, clinical inspection can be misleading (23), but this seems to be particularly true of patients with hyperinflation and abnormally large pleural pressure swings (23). Our patients were not hyperinflated, and suffered from a disease that tends to render muscles amyotrophic; therefore, there is no reason to think that clinical inspection could have overdiagnosed neck muscle hypertrophy. In addition, several additional arguments can be made to support neck muscle hypertrophy and hyperactivity in our patients. First, the clinical impression of hypertrophy corresponded to a palpable contraction with clavicular retraction. Second, in three patients, surface EMG electrodes placed over the sternomastoid muscles clearly picked out a phasic inspiratory activity (Figure 2), which is unusual (24). Third, neck muscle hypertrophy is the most likely explanation for the relative preservation of VC in our patients despite signs of diaphragm paralysis and reduced expiratory pressures (Table 2). This preservation of VC is in contrast to observations made by Laroche and coworkers (25) in patients with recent, isolated bilateral diaphragm paralysis. In these patients, VC was approximately 50% of predicted values, compared with an average 62% in our series. Admittedly the range of VC values observed in our patients (27 to 98% pred) was wide, but three patients had values that were in the normal range (Patient 1; VC, 98% pred; Patient 6; VC, 71% pred; Patient 8; VC, 78% pred) (Table 2). We interpret this finding as indicating a somewhat successful compensation for slow onset diaphragm paralysis by neck muscle recruitment and hypertrophy. This is supported by the fact that VC tended to be better preserved in the patients with the most negative Pm,t values (Figure 4). Patient 6, who had the longest disease duration and the most severe motor deficit but the most negative CMS-Pm,t value and a rather preserved VC (71% pred), can also be viewed as a good illustration of this concept.

View larger version (19K): [in this window] [in a new window]   Figure 4.   Relationship between mouth pressure response to cervical magnetic stimulation (CMS-Pm,t) and vital capacity, with 95% confidence intervals. When forced through zero, the relationship was highly significant (p < 0.01) (see text for details).

Practical Implications

Our investigation confirms that negative Pm,t can be recorded in response to cervical magnetic stimulation (5). Although glottis closure has consistently been mentioned as a limiting factor for this technique (3), it was not a problem in the present study. All the Pm twitches observed in our patients were regular over time. We did not observe abrupt interruptions of the Pm trace or a flat Pm trace despite RC-AB displacements, patterns that would have been suggestive of glottis closure. We propose that this lack of evidence for CMS-related glottis closure in our patients is, in some way, related to bulbar involvement by ALS, but this remains very speculative, and the corresponding mechanisms are not clear.

The Pm twitches we observed, despite diaphragm paralysis, are far from negligible (1 to 4.5 cm H₂O). These values overlap with the purely diaphragmatic Pm,t values obtained with specific E_S in stable COPD (4) and represent up to 25% of the standard set by Hamnegard and coworkers (5) as the lower limit of normal CMS-Pm,t. However, we do not wish to conclude that as much as 25% of CMS-Pm,t can be the result of contraction of muscles other than the diaphragm, whatever the clinical setting. In the patients with ALS we studied, progressive disappearance of diaphragm function over a long period of time probably led to the development of adaptive mechanisms. In the three patients with recent diaphragm paralysis and no INM hypertrophy, CMS elicited some signs of RC expansion. Nevertheless, there was no CMS-Pm,t and no CMS-induced AB paradox. These findings support the notion that the differences between CMS and E_S twitch Pdi are due to stiffening of the RC by CMS related contraction of extradiaphragmatic muscles (8). Our study also suggests that for CMS to produce inspiratory pressures through contraction of INM, sufficient time must have elapsed from the onset of diaphragm dysfunction for INM to become hypertrophic.

In conclusion, our aim in this study was to draw attention to the fact that it can be misleading to interpret CMS-Pm,t data as a pure index of diaphragm function in patients where INM hypertrophy is plausible. This further supports the notion that, as there is more than one tool for studying human diaphragm function, the choice of test and the results that it yields should be considered in the context of clinical history, clinical examination, and the results of other tests.