Posteroanterior Chest Roentgenograms Research Articles

Demonstration of the pericardial shadow on the routine chest roentgenogram: a new roentgen finding; preliminary report.

Attention has not previously been drawn in the radiologic literature to a visualization on the routine postero-anterior chest roentgenogram of the pericardial shadow along the left cardiac contour. The silhouette of the pericardium, normal or abnormal, becomes demonstrable in those instances in which there is sufficient deposition of subepicardial fat to afford a linear or curvilinear contrasting shadow of increased radiolucency between the density of the myocardial mass and the adjacent parietal pericardium. It must be understood that the explanation for the roentgen finding herein described is presented on a theoretical basis. Postmortem studies and roentgen-anatomic correlation will be the subject of a future report if results so warrant. Anatomico-Pathological Aspects The pericardium is described as a fibroserous sac in which the heart and the roots of the great vessels are contained (1). The parietal pericardium consists of a serous membrane composed of a single layer of flattened mesothelial cells resting on loose connective tissue, which connects it to an outer fibrous layer. The visceral pericardium, or epicardium, is lined by a single layer of polygonal cells which may be flat or cuboidal, depending on the state of contraction of the heart. Below this is a layer of connective tissue, the subepicardial layer of areolar tissue, connecting the epicardium to the myocardium and containing blood vessels, nerves, and varying amounts of fat (2). The pericardial cavity is merely a potential space, and in the normal state the serous layers of the visceral and parietal tunics are everywhere in contact and the adjacent surfaces are moistened by a slight amount of serous fluid. The mediastinal pleura lies in close apposition to the parietal pericardium and these structures are, in fact, fused above but separate at the level of the left lower pole of the heart. The separated pleural and pericardial layers, along with the diaphragm, invest a space filled with loose connective tissue and a variable amount of fat, a space variously referred to as the apical, or pericardial, or epicardial fat pad (3, 4). This extrapericardial collection of adipose and loose connective tissue so frequently seen on the roentgenogram in the left (and right) cardiophrenic angle is not to be confused with the subepicardial fat layer being described (Figs. 1 and 2). Gross inspection of the fresh human heart after removal from the pericardial sac, at the postmortem table, readily reveals the yellowish fatty tissues of the subepicardial layer upon its surface as seen through the closely applied and transparent overlying mesothelial cells. The amount of adipose tissue seen on the gross specimen shows considerable variation in different individuals, depending in part upon the general nutritional status preceding demise.

Determination of total lung capacity in disease from routine chest roentgenograms

In a group of seventy-six patients with various types of pulmonary disease, values for total lung capacity determined by volumetric measurement of routine postero-anterior and lateral chest roentgenograms were compared with the total lung capacities determined by spirometry and an open circuit method of determining functional residual capacity. Comparisons and correlations were analyzed statistically. The method used to estimate radiological chest volume (RCV) was that of Hurtado and Fray with two modifications: (1) the anteroposterior chest diameter was measured between bony landmarks on a lateral chest film, rather than externally at an inconstant level with calipers; (2) routine rather than special films were used. The radiological total lung capacity may be calculated from the modified RCV by the regression equation derived. High degrees of correlation were obtained, and evidence suggested that accuracy of radiological total lung capacity determination was increased by subtracting the radiological volume of regions of disease on the x-ray films, and decreased by subtracting the radiological volume of the heart and mediastinum from the total RCV. Lung field area on routine chest roentgenograms seems more constant than was previously supposed. When the subjects were separated into two groups, one consisting of thirty-three patients in whom the disease present did not include replacement of ventilating lung tissue, the other comprising forty-three patients with types of disease often replacing lung tissue with masses of non-functioning material, a large and statistically significant difference in correlation between RCV and total lung capacity was found for the two groups. It is believed that this difference enables degree of replacement of functioning pulmonary tissue in the single subject to be estimated more quantitatively than by previously available methods. The closeness of correlation between radiological and gaseous methods of lung capacity determination in the group of subjects possessing volumetrically intact pulmonary tissue was sufficient to suggest soundness and feasibility in screening large numbers of individuals for abnormal changes in pulmonary compartment ratios by a combination of simple spirometry and radiological measurement of routine chest roentgenograms. In seventeen patients with pulmonary emphysema the residual volume total lung capacity × 100 ratios determined using radiological and lung washout total capacities agreed within ±10 units in 94.1 per cent of cases, and within ±5 units in 70.6 per cent. Agreement was poor in diseases with mass replacement of ventilating parenchyma. Accuracy in RCV calculation is believed to be increased by measuring the anteroposterior chest diameter radiologically rather than externally with calipers. As a standard routine chest x-ray technic was used, the regression equation derived here for radiological determination of total lung capacity may be used in other hospitals employing the same roentgenographic technic.

Diaphragmatico-pericardial tear with intrapericardial herniation of the transverse colon.

Herniation of the abdominal contents into the chest through a diaphragmatic hernia or through a traumatic diaphragmatic tear is not uncommon. The case to be reported here is of special interest because of the rarity of the type of herniation observed, the long delay which preceded the diagnosis, and the unusual fashion in which the clinical and radiological picture simulated cardiac disease. Case Report A 45-year-old white male physician had felt well until May 1946, when he experienced tachycardia, fatigue, and slight dyspnea following moderate exertion. He was overweight and attributed the symptoms to obesity. In 1947 he had an episode of chest pain during exertion in cold weather, which he interpreted as an anginal attack. In 1948, frontal headaches developed. At this time, an enlargement of the cardiac contour was demonstrated roentgenologically, and the patient's past history was obtained. On June 20, 1943, while on service with the Armed Forces in North Africa, he had been in a jeep accident which caused mild cerebral concussion and a fracture of the right clavicle. He also received at that time a heavy blow to the precordial area, producing substernal pain and tenderness which lasted for about six weeks. An electrocardiogram in 1944 was normal. Further roentgenologic study in 1948 and 1949 indicated cardiac enlargement of unknown etiology and the patient was advised by one physician to give up his medical practice and to restrict his activities. From May to September 1950 he was hospitalized elsewhere for extensive observation. Roentgenkymography showed diminished pulsations along the right and left borders of the heart. Venous pressure and circulation times were normal. Bromsulfalein retention was 13 to 15 per cent. No other abnormal findings were elicited. Digitalization was without effect. No definitive diagnosis was reached. Angiocardiography, pericardial paracentesis, and exploratory thoracotomy were recommended but refused by the patient. Admission to the Medical Service of the Veterans Administration Hospital (Bronx, N. Y.) was on Dec. 30, 1950. The pertinent abnormal physical findings included moderate emphysema of the chest, distant heart sounds, a mild diastolic hypertension in both arms (130/95), enlargement of the heart manifested by extension of the cardiac dullness to the anterior axillary line, some enlargement of the liver (two fingers below the costal margin), and a vague “fullness” in the left upper quadrant of the abdomen. A hemogram was normal, as were the sedimentation rate, urinalysis, and serology. Liver function studies showed a thymol turbidity of 8.0 units, and a bromsulfalein retention of 16 per cent, without other abnormalities. An electrocardiogram was normal. A postero-anterior chest roentgenogram (Fig. 1A) showed generalized enlargement of the cardiac contour, while a lateral view (Fig. 1B) showed several radiolucent zones overlying the anterior aspect of the heart just posterior to the sternum.

Segmental localization of pulmonary disease on the postero-anterior chest roentgenogram.

With the advent of bronchoscopy and bronchography and especially under the influence of modern thoracic surgery, the sublobar divisions of the lung have come to be recognized as important units. The term “bronchopulmonary segment,” although applicable to any area of lung supplied by a bronchus, has through common usage come to designate that portion supplied by a tertiary bronchus. The importance of these pulmonary segments in pathogenesis, diagnosis, and treatment of many lung diseases is now widely recognized. To the specialist in pulmonary diseases, the distribution of certain segments and the direction of their draining bronchi explain the effects of gravity on the localization of lung abscess (4) and suggest more effective methods of postural drainage for lungs and bronchi (12). To the thoracic surgeon, segmental resection for bronchiectasis (7), tuberculosis (5, 6), and other diseases, presents an opportunity for conservation of pulmonary tissue and pulmonary function. The bronchopulmonary segments are consequently of interest to all clinicians having occasion to study chest roentgenograms. Attention has been called to the radiographic appearance of segmental pulmonary disease (16) and to the influence of retrocardiac processes on the contour of the cardiac silhouette (9, 15). The extent, however, to which segmental localization of pulmonary disease can be recognized from the conventional posteroanterior chest roentgenogram, is not fully appreciated. By applying the principles of x-ray projection to the three-dimensional anatomy of the bronchopulmonary segments, the writer has found it possible, in most instances, to determine from the routine chest film the anterior or posterior location of lung processes, as well as the segment involved. The following schematic presentation is offered as a guide to segmental localization on the postero-anterior chest roentgenogram. It is not intended to exclude the use of oblique or lateral films, or to minimize either the importance of lateral chest planigraphy for accurate preoperative localization of tuberculosis or the need for bronchography in mapping out the extent of surgery for bronchiectasis. Lobes For orientation, it may be well to review the five lobes into which the two lungs divide, as well as the commonly recognized accessory lobes. Although the lobe formations are not always separated by fully developed fissures, the appearance of their total involvement as seen on the posteroanterior chest film must be distinguished from that of their component segments. The distribution of the three lobes of the right lung and the two lobes of the left lung, as well as the extent to which they overlap the contours of heart, diaphragm, and adjacent lobes, is shown in the three-dimensional drawing in Figure 1.

Segmental anatomy of the lung.

For the past decade physicians who are concerned with diseases of the chest have exhibited a renewed interest in the details of the segmental anatomy of the lungs. The interest of the radiologist in this field is exemplified in the article, “Segmental Localization of Pulmonary Disease on the Postero-Anterior Chest Roentgenogram” by Irving Kane appearing in this issue of Radiology. The spectacular advances of thoracic surgery during the past two decades in large part date from the successful lobectomies, chiefly for bronchiectasis, done during the early thirties, and from Graham's successful pneumonectomy for carcinoma of the lung in 1933. In time the ability to remove large portions of the lung without undue risk became fully established. As long as the surgeon's efforts were confined to resection of a lobe or of a whole lung, the importance of segmental anatomy was not fully appreciated. But when Churchill, in 1939, successfully removed a segment of the lower lobe for bronchiectasis, a new emphasis was created. For it now appeared possible to remove, surgically, local areas of inflammation or small, non-malignant tumors, without a serious loss of functioning lung. Further development of the technic of segmental pneumonectomy by Blades, Overholt and his associates, Clagett, and many other thoracic surgeons in this country, has finally made local resection of the lung a relatively safe procedure, carrying a small risk, with few complications, and little abnormal residue. There followed the demonstration that surgical excision of tuberculous foci in the lung was a practical, safe procedure for the permanent eradication of tuberculosis. Here is a situation in which it is frequently necessary to remove multiple areas while, at the same time, it is desirable to save as much normal lung as possible, so that the technics of segmental pneumonectomy become of the greatest importance. Finally, it has become possible for thoracic surgeons to remove very small areas of the lung without regard to their segmental distribution, merely excising the actual tissue involved. The latter technical development has permitted the removal of large numbers of nodules from the lungs, some of which have proved to be malignant, others benign, and still others of an inflammatory nature. Complete extirpation of such a nodule is the surest guarantee against any eventual harmful effects which it may produce. In this same way the removal of relatively quiescent, or even well defined, but active, tuberculous foci, when they are accessible, and when the patient is fully protected by antibiotics, seems to make more certain a complete cure of this disease. Concomitant with the development of the procedures for the surgical removal of segments of the lung has come a development of technics for the accurate localization of such segments.

Encapsulated pleural effusion simulating mediastinal tumor; report of two cases.

The purpose of this report is to describe two cases of encapsulated paramediastinal pleural effusion which strongly simulated mediastinal tumors. In so far as it has been possible to determine, similar cases have not heretofore appeared in the literature. Since failure to recognize this condition might well lead to the performance of an unnecessary thoracotomy, it would appear important to add this disorder to the many others which must be considered in the differential diagnosis of apparent mediastinal masses. Case Reports Case I: S. S., a 38-year-old white man, was admitted to the U. S. Naval Hospital, St. Albans, N. Y., on Feb. 23, 1950, with a history of frequent attacks of mild precordial pain during the two years preceding admission. These attacks were precipitated by fatigue or excitement but not by exertion. The diagnosis of arterial hypertension had been made four years previously and the patient was receiving disability compensation for that disorder. His blood pressure was said to average about 190/120. There was no history of cardiac decompensation. Frequent attacks of headache and vertigo had occurred during the three or four years preceding the onset of his present illness. Ten days before admission there had been a severe attack of precordial pain. The pain originated along the left border of the sternum in the region of the fifth anterior interspace and radiated to the left shoulder, down the left arm, to the fourth and fifth fingers of the left hand. The family physician, after making an examination and taking an electrocardiogram, told the patient that he had a “heart strain” and prescribed bed rest and mild sedation but the symptoms failed to subside. A slight nonproductive cough and a low-grade fever developed. Malaise and weakness became noteworthy and the patient sought hospitalization. The patient had undergone yearly physical examinations because of hypertension and during the course of the routine study carried out ten days before the onset of his present illness a chest film had been taken. It was reported as negative. Admission physical examination showed few remarkable findings. The blood pressure was 195/140. The heart was enlarged to the left. The aortic second sound was accentuated. There was dullness to percussion and breath sounds were absent over the left base posteriorly. No râles or friction rubs were heard. The right chest showed no abnormal findings. Both testes were atrophic. A serologic test was negative and routine laboratory findings were within normal limits except for a leukocytosis of 17,200. The differential white blood count showed no significant departure from normal. A postero-anterior chest roentgenogram (Fig. 1) showed a sharply circumscribed, smoothly marginated, rounded homogeneous area of increased density projecting to the left at the level of the pulmonary conus. Aneurysm of the left pulmonary artery was among the diagnostic possibilities prominently considered at that time.

Innominate artery: angiocardiographic study.

The Radiologist is familiar with the problems involved in the differential diagnosis of roentgenographically demonstrated right superior mediastinal prominences. Similar shadows may represent a variety of neoplastic, infectious, granulomatous, vascular, and other lesions (Fig. 1). Aneurysm, dilatation, or buckling of the innominate artery may produce roent-genographic findings which simulate those of retrosternal thyroid enlargement, mediastinal lymph node enlargement, mediastinitis, or disease within the apex of the right lung. In the past it has occasionally been necessary to resort to exploratory thoracotomy in order to distinguish between innominate artery aneurysm and superior mediastinal tumor; this distinction may be made with ease through the use of angiocardiography (1). The present report summarizes our experience with contrast visualization as it bears upon the innominate artery. Normal Innominate Artery The normal innominate artery is the largest branch of the arch of the aorta, having been found in postmortem studies to measure from 3.7 to 5.0 cm. in length. It arises as the first of the three great brachiocephalic arteries from the ascending portion of the aortic arch; its point of origin is approximately on a level with the upper border of the second right costal cartilage. It takes an oblique course upward, backward, and to the right, to the level of the upper border of the right sternoclavicular junction, where it bifurcates to form the right common carotid and subclavian arteries. Anteriorly, the innominate artery is related to the left innominate and right inferior thyroid veins and the right vagus nerve. Posterior to it is the trachea, which it crosses obliquely. To the right are the right innominate vein, the superior vena cava, the right phrenic nerve, and the pleura. To the left are the thymic remnant, the left common carotid artery, the inferior thyroid veins, and the trachea (2, 3). The conventional postero-anterior chest roentgenogram does not reveal the normal innominate artery, since it does not project to the right of the spine and since the superior vena cava and the right innominate vein form the right superior mediastinal border in this projection. In the left anterior oblique projection, the innominate artery forms the anterior, superior portion of the cardiovascular silhouette, above the curve of the ascending aorta. Its posterior border is invisible. Elongation and dilatation of the artery may result in its becoming border-forming in the frontal projection (4). Angiocardiographic findings are in conformity with the anatomical and roentgenographic descriptions given above (Fig. 2). Due to its relatively small caliber (in comparison to that of the aorta), the innominate artery is not invariably well demonstrated by angiocardiography. Serial studies are of special value in this connection.

An Aortic Deformity Simulating Mediastinal Tumor: A Subclinical Form of Coarctation

Three examples of an unusual aortic deformity have been described. All were found in adults without clinical signs or symptoms referable to the chest. In all three, the anomaly was an incidental finding on routine roentgenologic examination of the chest. The deformity appeared in the postero-anterior chest roentgenogram as a mediastinal soft tissue density overlying a normal aortic arch. The true nature of the aortic abnormality was best seen in the left anterior oblique chest position. In one case in which operation was performed there was a kinking deformity of the aorta toward the pulmonary artery at the attachment of a short ductus arterious. Minimal narrowing of the aorta was present. This buckling deformity has been interpreted as a subclinical coarctation of the aorta in which the external traction or pulling deformity predominated and the constricting element was minimal or absent. The scanty literature on the subject is discussed.

Dilatation and pulsation of the left subclavian artery in the roentgen-ray diagnosis of coarctation of the aorta; roentgenkymographic studies in 13 cases.

Postero-anterior chest roentgenograms in 17 of 20 patients with coarctation of the aorta showed widening of the left upper mediastinal shadow attributed to dilatation of the left subclavian artery; pulsations of this vessel were evident in 12 of 13 cases with roentgenkymograms. In 3 of the latter cases rib notching was absent or equivocal. The subclavian pulsations may be observed fluoroscopically. Recognition of dilatation and pulsation of the left subclavian artery may provide roentgenologic corroboration of a diagnosis of coarctation of the aorta even in the absence of rib notching.

The Significance of the Roentgen Visibility of the Horizontal Fissure

THE appearance in the roentgenogram of the chest of linear shadows corresponding to the interlobar fissures of the lung is not unusual, and by comparison of such shadows with anatomic specimens a better understanding of the anatomy of the lungs has been attained (10, 5, 24). This has permitted the more accurate localization of pathologic processes seen in the chest roentgenogram in terms of the lobes and fissures which may be involved. Particularly important has been the identification of certain previously puzzling shadows sometimes included among the “unresolved pneumonias,” and which have, in some instances, been shown to be due to effusions limited to the pulmonary fissures. The unusual and, at times, bizarre outlines of these shadows are thus easily understood. By means of its normal position in the horizontal plane, the fissure between the right upper and middle lobes has attracted primary attention in this extension of x-ray critique. Thus, it is obvious that the plane of a fissure which becomes partially radiopaque would more easily be demonstrable when its greatest diameter lay in the axis of the roentgen ray, whereas, if the ray should penetrate it perpendicularly, the result might be only a faint haziness in the area of its projection, difficult of evaluation, if even detectable. In the customary postero-anterior chest roentgenogram, the horizontal fissure is, therefore, in a position, despite its slight superior convexity, to cast its densest shadow, whereas the long or oblique fissures in front of the lower lobes, being intercepted perpendicularly by the x-ray in the greater part of their extent, are not usually noticed on the film unless they contain a considerable amount of fluid, or their walls are grossly thickened by pathologic induration. Such pathology might give rise to the unusual shadows referred to previously, correctly identified through an adequate understanding of pulmonary lobar anatomy, and by the study of the interlobar spaces, in the case in question, by unusual or special technic. At times the lateral extremity of the oblique fissure attains a plane passing anterolaterally from the hilus which permits it to be demonstrated in the usual chest radiogram as an almost vertical line or one passing slightly laterally and cephalad from the diaphragm, approximately one-fourth the distance from the lateral costal margin to the mid-line (Fig. 1). This shadow may be either unilateral or bilateral, and visualization is made possible by the occasional concurrence of the plane of the fissure and the more divergent rays from an x-ray target comparatively close to the film. A lateral projection, however, permits better demonstration of the oblique fissure of the hemithorax nearest the film, and, on the right, its relationship with the horizontal fissure (11). These demonstrations have been presented by several authors (2, 4, 10).