© 2002 American Thoracic Society
Ventilator-associated Diaphragmatic DysfunctionVeterans Affairs Long Beach Health Care System and University of California Irvine, California During the poliomyelitis epidemic in the 1950s, the introduction of artificial positive pressure ventilation saved the lives of many patients with respiratory muscle paralysis. Since then, mechanical ventilation has been essential to treating patients with ventilatory pump or gas exchange failure. However, its adverse effects are disconcerting. In this issue of AJRCCM (pp. 1135–1140), Yang and coworkers demonstrated in sedated, paralyzed rats a profound reduction (approximately 50%) in diaphragm force-generating capacity after 58 hours of controlled mechanical ventilation, compared with nonparalyzed control animals (1). Most importantly, they found that short-term controlled mechanical ventilation was associated with reduced diaphragm muscle mass, reduced cross-sectional area of all fiber types, reductions in muscle fibers expressing slow myosin heavy chain (MHC) isoform, and increases in fibers expressing both slow and fast MHC isoform (hybrid fibers). These findings were confined to the diaphragm. These observations have important clinical implications for patients receiving mechanical ventilation. We should, however, ask the following questions: Is controlled mechanical ventilation solely responsible for diaphragm muscle dysfunction? To what extent do diaphragm remodeling and structural changes contribute to reduced force-generating capacity? To answer the first question, the effect of controlled mechanical ventilation on diaphragm function should be separated from the confounding effects of anesthesia, sedative, or neuromuscular junction–blocking drugs. Although Yang and coworkers (1) included a spontaneously breathing control group, the "control" did not receive neuromuscular junction–blocking agent and positive-end expiratory airway pressure (PEEP), as did the experimental animals. The findings reflect the complex interactive effects of diaphragm inactivity associated with (1) controlled mechanical ventilation, in which the diaphragm is phasically shortened; (2) neuromuscular junction–blocking agent, in which neuromuscular activation to the diaphragm is disrupted; and (3) PEEP, in which the diaphragm is tonically shortened. Thus, the findings are not necessarily attributable to controlled mechanical ventilation alone. The effect of inactivity on the diaphragm depends on the type (2) and duration of inactivity (3, 4). Phrenic denervation, blockage of phrenic motoneuron action potential conduction by tetrodotoxin, C2 spinal cord section, and controlled mechanical ventilation are models of diaphragmatic inactivity. All result in significant but varying degrees of reduced force-generating capacity of the diaphragm and transformation of muscle fiber types, with the phrenic denervation and tetrodotoxin models having the most pronounced effects on the diaphragm (2). This may explain the observation of Yang and coworkers of reduced diaphragmatic force but does not explain the muscle fiber transformation and structural changes. In a short-term (3 days) unilateral phrenic denervation model of diaphragmatic inactivity in rats, the force-generating capacity of the denervated diaphragm decreased to 38%, and the cross-sectional area of Types I, IIa, and IIb fibers increased significantly by 33%, 35%, and 28%, respectively (3). In contrast to these short-term effects, Lewis and coworkers (4) demonstrated that with prolonged (6 weeks) phrenic denervation, the proportions of Type I fibers decreased and of Type II fibers increased, with the majority comprised of hybrid fibers (Type IIc). The cross-sectional area of Types I and IIa fibers did not change, but that of Types IIx and IIb decreased (4). The acute effects of C2 spinal cord section on the diaphragm are unknown; 2 weeks later, however, the proportion of Type I fibers was unchanged, and mild hypertrophy was observed, whereas the proportion of Type IIa fibers decreased, with no change in the other Type II fibers (2). In rabbits, after 3 days of controlled mechanical ventilation, fibers expressing MHCslow isoform were unchanged. Fibers expressing MHC2X were not identified in three of the five animals studied, and fibers expressing MHC2A isoform tended to increase when compared with animals breathing spontaneously for the same duration (5). Unlike rats, the diaphragm of rabbits does not contain fibers expressing the MHC2B isoform. In the experimental group of Yang and coworkers (1), diaphragmatic fiber compositions were similar to the prolonged denervation model of Lewis and coworkers (4). The reason for this finding is not apparent (1) but may be related to the complex interaction of the factors mentioned previously here. At the transcriptional level, mean MHC2A mRNA was elevated (170%), with no change in the MHCslow mRNA (100%). It would have been useful if Yang and coworkers had determined the MHC isoform expressions of both diaphragm and hind limb muscles. In the study of Yang and coworkers, changes to the hind limb muscles were dissimilar to those of the diaphragm. Differences include diaphragmatic activation history and exposure to both tonic (effect of PEEP) and cyclic passive (effect of mechanical ventilation) shortening. The rhythmic and prolonged duty cycle of diaphragmatic activity predisposes it to damage with inactivity. The effects of PEEP on contractile properties of the diaphragm or structural changes have not been systematically studied. In rabbits, 2 days of immobilization in a cast with the soleus in a shortened position (plantar flexion) resulted in muscle fiber atrophy compared with the contralateral muscle without cast (6). Nuclear DNA fragmentation, ultrastructural changes indicating myonuclear apoptosis, and myofibrillar disruption were observed. Except for the atrophy, myofibril disruptions also appeared when hind-limb muscle was immobilized at its resting length (7) and in the diaphragm after 3 days of controlled mechanical ventilation without PEEP (5). Inactivity in a shortened position may have accelerated protein degradation leading to muscle fiber atrophy. To what extent do fiber atrophy and fiber types transformation contribute to the profound reduction in diaphragmatic force-generating capacity? In the experimental animals, the proportion of fibers expressing MHCslow isoform decreased by 5% (see Figure 1 of Yang and coworkers [1]) and mean diaphragmatic muscle mass decreased by 13% (see Table 2 of Yang and coworkers [1]). These findings were associated with a 50% reduction in twitch and maximum tetanic force (see Figure 4 of Yang and coworkers [1]). Another factor(s), including potential diaphragmatic damage, must also contribute to the disproportionate reduction in force-generating capacity of the diaphragm. Several researchers have demonstrated the detrimental effects of controlled mechanical ventilation on diaphragm force-generating capacity (5, 8–11), with only one study (5) isolating these effects from the confounding factors of anesthesia, neuromuscular junction–blocking agent, and PEEP. If the changes observed in animals occur in the human diaphragm during controlled mechanical ventilation, reduced force-generating capacity will influence the weaning process and the length of hospital stay. Future research should address the following questions: What is the mechanism of diaphragmatic damage caused by controlled mechanical ventilation? Does PEEP impair diaphragmatic function? Could maintenance of partial diaphragmatic activity or intermittent loading of the diaphragm prevent the harm done to diaphragmatic function? REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
