148: Overview of Anatomy, Physiology, and Pathophysiology of the Diaphragm

The diaphragm is generally referred to as the main “respiratory muscle”—which is to some extent a misnomer. A better description would be that the diaphragm is the primary muscle of ventilation. Its critical physiological role is to serve as the main muscle which moves air into the lungs, where this air, of course, oxygenates the blood. Since delivery of oxygenated blood to all parts of the body is critical to life, and since intact cardiac and respiratory systems are the essential elements to assure that this oxygenated blood is manufactured and delivered, failure of any of the components that makes up either one of these systems results in death or severe disability. One could thus make a very strong argument that along with the heart and lungs, the diaphragm is one of the three most important organs in the body.

Just as the heart is the blood pump, the diaphragm is the air pump. Brief consideration will convince the reader that the heart and diaphragm are the only two muscles in the body that are continuously active throughout the life of an individual. If either one of these muscles fails, the individual fails. As the only skeletal muscle that is continuously active (the heart is made up of cardiac muscle), the diaphragm presents a number of interesting physiological adaptations. Diaphragm muscle fibers demonstrate unique characteristics that permit it to function adequately given this unusual role of constant activity. Further, its special role seems to influence the diaphragm muscle's response to various disease states (e.g., emphysema) and medical interventions (e.g., mechanical ventilation [MV]), such that it often does not respond in the same manner as skeletal muscles in the limbs do under similar circumstances.

Beyond its physiological role, the diaphragm also serves an important anatomic role—separating the mobile abdominal contents from the compressible lung. Defects in this anatomic barrier result in herniation of abdominal contents into the chest, which can lead to both malfunction/failure of the abdominal organs, and dysfunction of the lung. This anatomic role of the diaphragm, like its physiological role in ventilation, serves at bottom to preserve a fully functioning respiratory system such that oxygenated blood can efficiently reach the heart.

Basic Anatomy and Embryology

The diaphragm is made up of two muscular parts and a central tendon. The muscular portions are the costal diaphragm and the crural diaphragm. The costal diaphragm is its main functional component. It is formed by muscle fibers which originate on the lower six ribs and the xiphoid process and attach on the central tendon (Fig. 148-1). The crural diaphragm consists of bundles of muscle fibers that arise from the first three lumbar vertebral bodies and the medial and lateral arcuate ligaments on each side. Although some mistakenly think of the costal diaphragm as being made up of two separate muscles – right and left hemidiaphragms each forming their own separate dome – in fact the diaphragm is a single, continuous organ that spans the midline as one broad dome.

The importance of the diaphragm's role in separating the abdominal from the intrathoracic contents is highlighted by the apertures in the diaphragm, through which a number of structures that span each of these cavities must pass. These apertures for the vena cava, the esophagus, and the aorta also allow passage of smaller structures. The vagi and sympathetic trunks pass through the esophageal aperture; the thoracic duct and azygos vein pass through the aortic aperture. Although the apertures are sealed around each of these structures by soft tissue including the mesothelial layer that lines the abdominal surface of the diaphragm and the pleura that lines its thoracic surface, the apertures are nevertheless sites that can allow herniation of visceral contents into the chest. By far the most common acquired diaphragmatic hernias, as discussed in Chapter 46, are hiatal hernias through the esophageal aperture.

Congenital diaphragmatic hernias occur at sites that are compromised by failure of the normal embryological development of the diaphragm. In the embryo, the diaphragm is formed by the fusion of the septum transversum, the pleuriperitoneal membrane, the dorsal mesentery, and the lateral body wall mesoderm (Fig. 148-2). There is also a contribution of myoblasts from the third through fifth cervical somites which migrate to contribute to the formation of muscle fibers. The various congenital diaphragmatic defects can be attributed to failure of one of these processes. For example, posterolateral Bochdalek hernias result from failure of the fusion of the pleuriperitoneal membrane. Morgagni hernias, which occur in the anterior, medial diaphragm posterior to the xiphoid, result from failure of the myoblasts from the cervical somites to appear. The congenital hernias are discussed in Chapter 51.

It is also important to note that the barrier function of the diaphragm fails to a lesser extent in other clinical syndromes that are perhaps less obvious than the diaphragmatic hernias. Some patients appear to have “pores” of varying sizes in the diaphragm which may allow intraabdominal fluid and other materials to pass into the chest. These syndromes were first categorized by Kirchner.¹ The most obvious of the Porous Diaphragm Syndromes result from abnormal intraabdominal fluid passing into the chest. The fluid may consist of benign ascites (as in hepatic hydrothorax in cirrhotic patients), malignant ascites (as in Meigs syndrome with ovarian cancer), or peritoneal dialysis fluid (in patients receiving PD), as just a few of the more common examples. All of these tend to occur in the right hemithorax, perhaps because peritoneal fluid circulates such that it pools in the right subphrenic space, or because the presence of the liver on the right creates a sumping action upon the right hemidiaphragm. These problems can often be solved by: (1) at least temporarily minimizing as much as possible the abdominal fluid component by drainage or medical therapy, and then (2) creating a basilar pleurodesis with talc or mesh that will allow the adherent lung to occlude the pores. Depending upon the chronicity of the pleural effusion, a formal decortication may be required as part of this procedure, and if the pores are large, success may be improved by actual pore suture or stapling rather than relying upon a pleurodesis alone. The thoracic procedure is generally carried out by video-assisted thoracoscopic surgery (VATS). Examples of this type of approach are described in publications by Cerfolio and Bryant² for hepatic hydrothorax and Mak et al.³ for effusions related to PD.

Catamenial pneumothorax refers to pneumothorax that occurs cyclically with menses. This is thought to result in some cases from implants of endometriosis on the lung (which may lead to direct leakage of air from the visceral pleural surface during the hemorrhagic phase within the implants), and in other cases from implants on the diaphragm which may in a similar fashion lead to diaphragmatic defects through which air may enter the chest from the peritoneal space. Implants in both of these locations have been identified during thoracoscopic exploration in a high percentage of patients with this syndrome. It is becoming clear that creating a particularly effective pleurodesis at the level of the diaphragm is an important component of successful management of these patients (as with other “porous diaphragm syndromes”) in addition to the more standard pleurodesis or pleurectomy covering the remainder of the chest.⁴ Resection of any blebs and visceral pleural implants that are seen are of course also performed during these procedures. Hormonal manipulation to prevent menses for several weeks after the time of the surgical procedure has also been suggested.

Blood Supply

The diaphragm's arterial supply derives largely from the right and left inferior phrenic arteries, the intercostal arteries, and the musculophrenic branches of the internal mammary arteries. There is also a minor contribution from the pericardiophrenic arteries which run with the phrenic nerves. Drainage is through the inferior phrenic veins, which drain into the inferior vena cava and through the azygos and hemiazygos systems. There is clearly substantial collateral circulation across the diaphragm, since it does not appear that diaphragm muscle infarcts or suffers necrosis when individual elements of the arterial supply become occluded or are surgically divided.

Lymphatics

Diaphragmatic lymphatics form a specialized system which appears to be important in draining fluid from the peritoneal cavity, and perhaps the pleural cavity, and returning it to the vascular system. This fluid appears to enter subperitoneal lymphatic lacunae, which sit between muscle fibers of the diaphragm. From the lacunae, fluid traverses the diaphragm via intrinsic lymphatics to reach collecting lymphatics beneath the diaphragmatic pleura. Both the intrinsic and collecting lymphatics contain valves. The collecting lymphatics drain principally into retrosternal (parasternal) lymphatic trunks that carry lymph to the great veins after it filters through mediastinal lymph nodes.⁵ Subpleural lymphatic channels along the diaphragm have been demonstrated to drain into lymph vessels that ascend along the inferior pulmonary ligaments, along the esophagus, or between the pulmonary veins, before connecting to the intertracheobronchial lymph nodes and sometimes toas high as the upper mediastinal lymph nodes.⁶ Thus, lung cancers invading the diaphragm may drain by this alternate route to involve the same mediastinal lymph node chains that receive lymph from tumors within the pulmonary parenchyma.

It is thought that the contraction of the diaphragm creates a sumping action that facilitates absorption of fluid and lymph into all of these channels, making the diaphragm an important contributor to the total volume of lymph flow.

Innervation

The motor supply to the diaphragm is via the phrenic nerves. These nerves run in very reproducible positions along the anterior scalene muscle and down along the mediastinum on each side. It is critical for thoracic surgeons to be intimately familiar with the anatomy of the nerves along their entire course in order to avoid injury to them (likely the most common cause of diaphragm paralysis) during all thoracic procedures. The costal diaphragm's innervation derives from the third and fourth cervical segments, while the crural diaphragm's innervation is from the fifth segment. Thus, spinal cord injury at C4 or higher leads to partial or complete diaphragm paralysis.

The right phrenic nerve passes through the central tendon along the vena cava through the caval aperture. It divides there into anterior, lateral, and posterior branches (Figs. 148-3 and 148-4). On the left, the nerve enters the diaphragm just anterior to the central tendon and sends off similar branches. The locations of these branches have implications for the placement of incisions in the diaphragm that will be least disruptive to innervation (see separate chapter on diaphragm incisions).

In normal individuals there is very little or no cross-innervation between the left and right sides of the diaphragm. The right phrenic nerve supplies only the right hemidiaphragm and the left nerve supplies only the left hemidiaphragm. As a result, injury to one nerve only causes major dysfunction to the ipsilateral diaphragm. In experimental animals, there seems to be some ability of the contralateral, healthy phrenic to eventually, partially reinnervate a denervated diaphragm, but it is unclear if this occurs to any clinically significant extent in humans.

Role in Ventilation

As a result of its circumferential attachments to the ribs and spine, its attachment to a central tendon, and its sheet-like, domed configuration, when the muscle fibers of the diaphragm contract, the sheet-like muscle descends caudally. Wade measured the displacement of the diaphragm's dome with inspiration at 9.5 cm on average.⁷ The diaphragm essentially acts as a piston, descending caudally with contraction. This diaphragmatic descent increases the intrathoracic volume, thereby reducing the intrapleural pressure. A lesser component (approximately 10%) of the reduction in intrapleural pressure with diaphragm contraction results from what is called the “insertional component” – or the diaphragm fibers' insertion on the superior margin of the lower ribs – which during contraction cause the lower ribs to expand outward (the increase in intraabdominal pressure may also contribute to this action). The resulting reduced intrapleural pressure expands the lung, drawing air into the lungs via the airways. As the diaphragm relaxes, it moves passively in a cranial direction, allowing air to passively exit the lung as a result of the intrinsic elastic recoil of the lung tissue and chest wall.

When the diaphragm is dysfunctional or ventilatory requirements are increased by exercise, other inspiratory muscles – which do not contribute much during quiet breathing – may become active. These include mainly the scalene muscles, the parasternal intercostal muscles, the sternocleidomastoid muscles, the transversus abdominis muscle, and possibly the external intercostal muscles.

As in limb muscle, shifts in the distribution of the various muscle fiber types within the diaphragm (e.g., slow or fast fibers) occur in response to certain stressors. These shifts may be adaptive, or in some cases perhaps maladaptive (see below discussion regarding changes in fiber types in emphysema).

Pathophysiology of Dyspnea

The sensation of dyspnea remains poorly understood, but appears to result from a very complicated set of neurophysiological events. Interestingly, it has been clearly shown that dyspnea is far more closely related to respiratory muscle function than it is to FEV₁. To summarize a large body of research in a few sentences, one might say that one of the most important contributor to the sense of dyspnea in many patients is the work of breathing (WOB).

WOB is work performed by the respiratory muscles—chiefly the diaphragm in the absence of expiratory airflow limitation. Its level is determined by the two main types of resistances against which the respiratory muscles must work—the elastic resistance of the lung or chest wall tissues and the frictional resistance of the airways to gas flow (e.g., increased COPD). When WOB is increased, dyspnea results.

One common causes of increased WOB is diaphragm dysfunction. When the diaphragm fails to carry out its normal role of pulling air into the lungs by descending – whether because of paralysis, emphysema, or other causes – the secondary muscles of respiration must become activated. This increased ventilatory muscle activation appears to be sensed by the patient as dyspnea.

Role in Respiratory Failure and Weaning from Mechanical Ventilation

Since the diaphragm is the primary ventilatory muscle, its function is of course central to the pathophysiology of respiratory failure and to weaning from MV once MV has been instituted. While many patients who require MV suffer from primary pulmonary failure (e.g., failure of gas exchange on the basis of parenchymal lung disease such as pneumonia, interstitial lung disease, pulmonary edema), some patients will require MV primarily because of failure to ventilate (i.e., diaphragmatic failure). This situation is perhaps most common in surgical patients following prolonged anesthetics, often including paralytic agents, and in emphysema patients. Diaphragmatic function in emphysema will be discussed further below, but simply stated, at some point in the evolution of hyperexpansion, the diaphragm simply cannot move enough air to maintain an acceptable pCO₂, and at that point the patient will either die or require MV.

Similarly, weaning from MV, once pulmonary parenchymal disease has been improved such that oxygenation is adequate, is primarily a function of the adequacy of the diaphragm to move air. The ability to inspire adequately without support would seem to be the sine qua non of separation from a ventilator. Thus, it is critical to do everything possible to preserve diaphragm function while a patient is subjected to MV (see below section on MV-associated diaphragm dysfunction). And measurements of diaphragm function serve as useful determinants of a patient's readiness for discontinuation of MV (see next section on measuring diaphragm function).

Physical Examination and Radiographic Studies

On physical examination in a spontaneously breathing patient, one can often recognize diaphragm dysfunction by reduced auscultation of breath sounds on the side of the dysfunction. Even with diaphragm paresis in the absence of true paralysis, one can generally hear reduced breath sounds. When diaphragm elevation is marked in the setting of paralysis, there are often nearly absent breath sounds at the base on the involved side.

Auscultation, however, provides on a very crude estimate of diaphragm dysfunction, and there are of course many other disease processes that can cause reduced breath sounds from which diaphragm dysfunction must be distinguished. In clinical practice, then, the main means of establishing diaphragm dysfunction in spontaneously breathing patients (i.e., nonventilated patients) are (1) the position of the diaphragm on radiographic examinations and (2) diaphragm fluoroscopy.

Essentially any study that images the chest and upper abdomen can suggest the possibility of diaphragm dysfunction by demonstrating elevation of the diaphragm above its normal position. It is important to remember, however, that different radiographic studies are performed with the patient in different body positions, and that the dysfunctional diaphragm will move to markedly different apparent degrees of elevation depending upon the body position of the patient. For example, a patient with diaphragm paralysis will usually demonstrate far less elevation of the involved side of the diaphragm on a PA and lateral chest radiogram (which is obtained with a patient standing upright) than on the scout films from a computed tomogram (CT) of the chest (which is obtained with the patient fully supine) (Fig. 148-5). I suspect that many dyspneic patients have been misdiagnosed to have normal diaphragm function on the basis of PA/lateral chest radiogram showing near-normal diaphragm position.

Diaphragm fluoroscopy (often termed the “sniff test”) examines the movement of the diaphragm in real time as the patient is asked to breathe deeply. A diaphragm that is paretic but not paralyzed may show some, but a reduced degree of, descent with deep inspiration, or it may show no motion at all. A fully paralyzed diaphragm will generally show paradoxical elevation during inspiration. This is because the involved side of the diaphragm not only fails to contract and thus fails to descend, but also actually rises in response to the increased negative pressure created in the ipsilateral pleural space as a result of contraction of the uninvolved, fully contractile, contralateral side of the diaphragm. When the intact side of the diaphragm contracts, it conveys a strong negative pressure to its ipsilateral pleural space, which is then conveyed to a lesser degree through the mediastinal structures to the contralateral pleural space, which will therefore result in mild paradoxical elevation of the paralyzed diaphragm.

It is important to remember that there are degrees of diaphragm dysfunction, any of which may cause a patient to be dyspneic, and which may result in less than classic findings on fluoroscopy. In fact, a patient may be dyspneic when the diaphragm is elevated but fully contractile (termed an eventration). In this situation, the phrenic nerve may be fully intact, but a structural weakness of the diaphragm muscle leads it to assume a more elevated position. An eventrated diaphragm may descend nearly normally on diaphragm fluoroscopy, but more commonly it appears to descend incompletely, or there may be portions of the diaphragm that descend normally while others do not. Some portions may actually show paradoxical motion. With an eventration, even though the diaphragm may descend, thereby reducing intrapleural pressure and causing inspiration as in normal diaphragm function, if the muscle is significantly elevated it may cause atelectasis of the lung, ventilation/perfusion mismatch, and shunting, which may themselves be causes of dyspnea.

Electromyography

Changes in diaphragmatic electrical activity reflect various aspects of diaphragm contraction. Although perhaps less clinically useful than measurement of maximal inspiratory pressure (MIP) (see below), the “compound motor action potential” (electromyography [EMG] signal) measured at the skin level does correlate with force of contraction. Other information, for example, on whether the diaphragm is fatiguing, can also be learned from more detailed evaluation of the EMG signal characteristics. Phrenic nerve conduction time can be measured with electrodes placed over the rib cage and stimulation of the nerve in the neck. Although not a primary mode of evaluating diaphragm paralysis – a diagnosis which is more easily established by fluoroscopy – nerve conduction time can be useful in the evaluation of phrenic nerve dysfunction when the diagnosis is in doubt.

Maximal Inspiratory Pressure

The most commonly used and clinically useful means of measuring the overall strength of diaphragm contraction are measurements of the maximal amount of negative inspiratory pressure that can be created by a patient at the level of the mouth. Although rarely ordered preoperatively in the evaluation of surgical patients, most pulmonary function laboratories are equipped to measure the MIP. This is done by having the patient inspire maximally from residual volume against an obstructed mouthpiece with only a small leak. It has been suggested that in cardiac surgery, preoperative MIP and MEP are associated with the need for prolonged postoperative ventilation.⁸ To my knowledge, a study of whether preoperative MIP is an independent determinant of postoperative pulmonary complications in general thoracic surgical patients has never been carried out.

In the intensive care unit, MIP is commonly used as one of the determinants of whether a ventilated patient is “ready to wean.” There are a number of studies that correlate successful weaning with higher levels of inspiratory pressure that are able to be generated. A MIP of less than 30-cm water is generally considered insufficient to support successful weaning. However, more complex formulas combining MIP, respiratory load, and/or minute ventilation have also been proposed to predict successful weaning.⁹

Transdiaphragmatic Pressure

Although used primarily in research applications, the force of diaphragmatic contraction can be more precisely measured by measuring the transdiaphragmatic pressure (Pdi) that is generated and plugging this into a formula that allows calculation of the force of contraction. The Pdi is measured by placing a catheter perorally that has both intraesophageal and intragastric balloons that can measure the intrathoracic and intraabdominal pressures, respectively. The difference between intrathoracic and intraabdominal pressures is the Pdi. Pdimax can be measured while having awake patients inspire maximally against a nearly occluded mouthpiece, or while maximally stimulating the phrenic nerves in the neck in sedated, intubated patients.

Since a variety of factors affect the force of contraction from breath to breath, clearly the more objective measure of diaphragm function is the Pdimax measured during maximal bilateral phrenic nerve stimulation. Since this value is designed to measure the force the diaphragm can create when all of its muscle fibers are maximally recruited, it is a measure that is relatively reproducible and can be used to follow diaphragm function over time far more precisely than with measures such as MIP.

The diaphragm may rarely give rise to primary diaphragmatic tumors. Since the diaphragm is intimately associated with vital organs on both its thoracic and abdominal sides, it is more common for neoplasms of these organs to spread to involve the diaphragm, requiring en bloc resection. We review diaphragm tumors here, since they are not covered elsewhere in this volume.

Primary tumors of the diaphragm are more often benign than malignant, and as one would expect, they arise from muscle or other mesenchymal tissues.¹⁰ The benign tumors most commonly include mesenchymal cysts, lipomas, and fibromas. The malignant tumors are most commonly rhabdomyosarcomas or leiomyosarcomas. These tumors are often asymptomatic at the time of presentation, although pain, cough, or dyspnea is not uncommon.¹¹

Secondary tumors that involve the diaphragm most commonly arise from the lungs, the liver, or the gastroesophageal junction. One might also consider tumors of the pleura, such as mesothelioma, that commonly penetrate into the diaphragm to also be secondary tumors of the diaphragm. In each of these cases, diaphragmatic involvement does not preclude cure by diaphragmatic resection along with resection of the primary tumor, but the prognosis is generally worse than without diaphragmatic involvement.^12–14 This may be because of the extensive lymphatic and venous drainage of the diaphragm.

Metastases to the pleural or peritoneal spaces from a variety of tumors may also involve the diaphragm secondarily, but these cases have a very poor prognosis. Cytoreductive peritonectomy¹⁵ and pleurectomy,¹⁶ including diaphragm resection plus adjuvant therapies, have been described for these situations, with some suggestion of benefit. However, there is currently no indication for surgical resection in these cases outside of clinical trials.

It can be quite difficult to determine whether a tumor in the vicinity of the diaphragm is a primary diaphragmatic tumor versus a primary tumor of an adjacent organ secondarily involving the diaphragm. The thin structure of the diaphragm, and its constant movement, render it difficult to accurately image. Fortunately, this determination is often not critical for therapy. It can, however, be helpful for preoperative planning to know if the diaphragm will need to be resected. While CT can suggest loss of normal tissue planes and thus diaphragmatic invasion, its volume averaging effect on curved surfaces can render this less than reliable. Magnetic resonance imaging (MRI) does not have this technical limitation, and a number of studies have suggested that it is more accurate than CT in predicting diaphragmatic involvement.

Paralysis/Paresis

This topic is covered deeply elsewhere in the text. Unilateral diaphragm paralysis is a relatively common condition – whether idiopathic or iatrogenic – and has substantial impact on dyspnea and quality of life. Diaphragm plication is a very effective therapy, substantially improving symptoms and objective measures of pulmonary function. Particularly since it can now reliably be performed by VATS,¹⁷ it should be considered in all patients with paralysis that has persisted for over 6 months. Still, because the diaphragm remains noncontractile after plication, exercise capacity does not generally return to the full preoperative level. Although one might consider direct diaphragm pacing as a more physiological solution, the problem of timing the paced contraction to the normal phrenic nerve impulse and contraction on the contralateral side has limited exploration of this possibility.

Post Thoracotomy Diaphragm Dysfunction

It has been clearly demonstrated in sheep that diaphragm function is less robust following thoracotomy,¹⁸ and that this dysfunction extends at least to 28 days postoperatively. Recovery from this postoperative deficit in diaphragm contractility is somewhat faster following VATS than following thoracotomy.¹⁹ Although one might infer that this compromised diaphragm function following thoracic surgery is a source of pulmonary complications, it would be very difficult to demonstrate this experimentally.

Diaphragm Dysfunction in Emphysema

In emphysema, the progressively increasing volume of the lung gradually pushes the diaphragm caudally until, in severe disease, it has attained a nearly flat, horizontal position quite different from its usual domed, piston-like configuration (Fig. 148-6). In this flatter configuration, the piston-like descent of the diaphragm is obviously severely compromised. Also, the diaphragm's secondary action of expanding the lower rib cage may actually be reversed such that the ribs are drawn inward with contraction, favoring exhalation. The diaphragm is thus an extremely ineffective inspiratory muscle in severe emphysema. This diaphragm dysfunction increases the WOB and forces emphysema patients to recruit their accessory muscles of inspiration to draw air in. This is a major contributor to the severe sensation of dyspnea in these patients.

It is interesting to note that although the contraction of the diaphragm in emphysema is less effective at creating inspiration than in the normal state, it is unlikely that the diaphragm is receiving any less descending neural stimulation. In fact, diaphragm muscle fibers of emphysema patients shift towards a fiber type (slow, oxidative type expressing type I myosin heavy chains [MHCs]) – similar to the adaptations that occur within the leg muscles in a long-distance runner – suggesting that the muscle is in emphysema under greater rather than lesser load.²⁰ Muscle with more type I fibers may be “adaptive” to the extent that it will be more resistant to fatigue, but on the other hand it cannot generate the same maximal force as a muscle with a greater (normal) proportion of fast myofibers.

After lung volume reduction surgery (LVRS), the diaphragm progressively attains, over a 3- to 6-month period, a more normal, domed configuration and improved function.²¹ New sarcomeres are actually laid down within the diaphragm muscle fibers, allowing the now longer fibers to function at their optimal point on their length–tension curve.²² Along of course with changes within the lung itself, such as improved elastic recoil and V/Q matching, these changes within the diaphragm certainly contribute in a major way to the objective symptomatic improvements obtained by appropriately selected LVRS patients. Remember, dyspnea is more closely related to respiratory muscle function than to FEV₁.

Obesity

As in emphysema, obesity places an increased load on the respiratory muscles including the diaphragm, and they function less effectively. The load in this situation appears to result from reduced chest wall compliance. This results in similar shifts in the diaphragm as seen in emphysema—to a greater proportion of slow fiber types, MHCs, and greater oxidative enzyme capacity (at least in experimental models).²³

Aging

At age 70 to 80, the maximal contractile force able to be generated by the diaphragm is 20% to 30% lower than at age 20.²⁴ This seems to result from shifts in the proportions of the various fiber types with age.²⁵

Mechanical Ventilation

A very important recent discovery is that even brief periods of full mechanical ventilator support appear to result in rapid and severe (~50%) reduction in diaphragm muscle fiber cross-sectional area in humans (Fig. 148-7).²⁶ This finding confirmed a substantial amount of data showing this phenomenon in experimental animals.²⁷ The loss of diaphragm muscle mass with MV is far more rapid than that seen in casted limb muscles, a finding that may be somehow related to the diaphragm's unique, continuous function when compared with limb muscles. The atrophy seems to occur via rapid activation of muscle proteolytic pathways.

Obviously, reduction of muscle fiber size by 50% will translate to a dramatic reduction in the ability of the diaphragm to create inspiratory force. Although it has not been proven with certainty, strong circumstantial evidence exists that this reduction in diaphragm force is a central cause of “failure to wean” and thus the need for prolonged MV in many patients.^28,29 If ventilator-induced diaphragm atrophy (VIDD) could be prevented, it might prevent the major morbidity and mortality rates that are associated with prolonged MV as well as saving millions of health care dollars.

One way to prevent VIDD might be to avoid, where possible, full ventilator support. Using pressure support, or other modes that require an ongoing contribution by the patients diaphragm to inspiration, may prevent VIDD. This has in fact been demonstrated to be effective in rats.³⁰ However, there are many disease processes and severities in which pressure support modes are impossible to employ. Another approach might be electrical stimulation of the diaphragm when patients require full MV support, although this has not been shown to be effective even in animals as of this writing. Lastly, and perhaps most conveniently, a drug might be developed that would prevent diaphragm atrophy. Such a drug could be administered to all patients at the time they are placed on the ventilator. Work delineating the molecular basis of VIDD is ongoing,³¹ and has already yielded drugs showing effectiveness in animal models.^32,33

Dr. Shrager presents a clear and concise description of the important anatomy of the diaphragm. I appreciate his comparison of the diaphragm to an “air pump” or “piston.” Readers should note the clear explanation of porous diseases of the diaphragm in this chapter. This chapter also addresses the role of the diaphragm in respiratory failure and the feeling of dyspnea. A number of ways of measuring diaphragm function are presented. This is an area of growing understanding in thoracic surgery, and this chapter is a valuable contribution to those wishing to further this field.