Vol. 136 REPRINTED FROM ANNALS OF SURGERY 227 South Sixth Street, Philadelphia, Penna. DECEMBER, 1952 Copyright, 1952, by J. B. Lippincott Company. No. 6 TRANSTHORACIC REPAIR OF DIAPHRAGMATIC HERNIA IN INFANTS* C. Everett Koop, M.D., Sc. D. (Mp) anp Jutian Jonnson, M.D., Sc. D. (Mp) PHILADELPHIA, Pa, FROM THE SURGICAL CLINIC, CHILDREN’S HOSPITAL OF PHILADELPHIA AND THE DEPARTMENT OF SURGERY, SCHOOL OF MEDICINE UNIVERSITY OF PENNSYLVANIA, PHILADELPHIA Congenital diaphragmatic hernia produc- ing symptoms in the newborn period should be dealt with as a surgical emergency. In spite of the fact that untreated diaphrag- matic hernias are not uniformly fatal, the mortality is so high during the neo-natal period in untreated patients as to place this lesion, in our opinion, in the category of congenital anomalies incompatible with life, but amenable to surgical correction. Many infants with a diaphagmatic hernia are in great distress, and the decision to operate upon them is not a difficult one. However, some newborn infants have a period of temporary improvement with the administration of oxygen or after aspiration of mucus from the nasopharynx, and the temptation to postpone operation may be- come great. One never knows when the delicate balance in cardiorespiratory re- serve may be upset by further crowding of abdominal viscera into the thorax, with consequent shift of the mediastinum. There would therefore seem to be no excuse for postponement of surgery in infants with diaphragmatic hernias, either to await growth of the child, or improvement of the symptoms and signs caused by the lesion itself. The subject of congenital diaphragmatic hernia has been well covered in the surgical literature.14 These lesions seem to be re- * Read in part before the Society of University Surgeons, Philadelphia, February 9, 1950. Sub- titted for publication March, 1952. ported more frequently than formerly and this probably reflects earlier diagnosis and more prompt treatment. A satisfactory classification of congenital diaphragmatic hernias is given by most authors and includes the usual sites of the foramen of Bochdalek, the esophageal hiatus, the foramen of Morgagni and pos- terior defects or congenital absence of the diaphragm. DIAGNOSIS When the diagnosis of diaphragmatic hernia can be made in the delivery room, the infant is usually in severe distress. Cyanosis is a prominent finding and dia- phragmatic hernia should be considered in any newborn infant whose cyanosis per- sists after the usual measures directed at clearing the airway and establishing res- pirations have been carried out. Most congenital diaphragmatic hernias that give symptoms at birth are on the left side. Physical examination of the infant usually reveals a relatively large thorax, the left side of which lags behind the right in respiratory effort. The abdomen is small and frequently scaphoid. The heart is dis- placed to the right and often to an extreme degree. Breath sounds are absent over the left chest and are heard only over the upper portion of the right lung field, where they are usually of harsh character. Later, as gas fills the bowel, the mediastinal shift increases and symptoms become more se- 1007 PRINTED IN U.S.A. KOOP AND JOHNSON vere. In right-sided lesions, the thoracic signs are reversed and usually less severe because the liver serves as a barrier below the diaphragmatic defect. The diagnosis can be, and usually is made on physical findings alone, but roentgeno- grams of the chest are almost universally taken to confirm the clinical impression. The use of a contrast medium in the bowel in most instances is unnecessary and is con- traindicated except in unusual circum- stances. Should the necessity arise for using a radiopaque substance in the upper gastro-intestinal tract, an iodized oil prep- gration is preferable to barium because of the likelihood of vomiting and the danger of aspiration. In ascertaining the nature of a small mass of abdominal viscera in the chest, which may be colon only, the injection of a radiopaque medium into the colon might be desirable. Here, again, a radiopaque material other than barium is preferable because of the desirability of using fluid which can be evacuated readily. Fortu- nately, those patients in whom the diag- nosis is not clear on the basis of physical signs alone, are usually not severely embar- rassed from the cardiorespiratory stand- point, so that studies can be undertaken without undue risk. COMPARISON OF ABDOMINAL AND TRANS- THORACIC APPROACHES Most authors have advocated the repair of diaphragmatic hernia in infants by the abdominal approach’® even when the thoracic approach is considered to be su- perior for adults.* Our experience leads us to recommend the thoracic approach for infants as well. The difficulties of an abdominal approach in the repair of diaphragmatic defects are several, and even though means may be found of circumventing them, there are some trying moments for the surgeon dur- ing the procedure. Perhaps the major problem lies in the Annals of Surgery December, 195 2 fact that anesthesia seldom improves the cardiorespiratory status of these patients, and hence it is the desire of the surgeon to remove the abdominal viscera from the thorax as rapidly as possible. In this en- deavor he is frequently hindered by the difficulty of extracting gas-filled loops of bowel through the defect in the diaphragm from the abdominal side. Secondly, the repair of the defect itself must be carried out with difficult exposure caused by the overhanging costal cage. Finally, there exist the difficulties asso- ciated with enclosing the viscera in a peri- toneal cavity not large enough to contain them. These difficulties are reflected in the technics of closure which have been de- scribed, such as delayed closure of fascia and the use of through-and-through sutures instead of the conventional closure in layers.’ Because of the problems presented in repairing congenital diaphragmatic hernia in the neo-natal period by the abdominal approach, we have used a transthoracic approach to the diaphragm in 15 infants with diaphragmatic hernia or eventration of the diaphragm. The first advantage of the trans-thoracic approach is encountered when in a cya- notic, distressed infant, with a materially diminished cardiovascular reserve, it is pos- sible to empty the pleural cavity of ab- dominal viscera and relieve the respiratory distress by making a rapid intercostal inci- sion. As soon as the chest is opened, the viscera deliver themselves, the mediastinun is free to return toward the midline, the compressed lung is permitted to expand. and the patient becomes a relatively normal cardiovascular risk under anesthesia. The remainder of the procedure can then be undertaken without the previous urgency: and with appreciably diminished risk. In the second place, we have not found it difficult to replace the abdominal viscer4 through the diaphragmatic opening, excep! in one instance, when it was necessary to 1008 volume 136 Number snlarge the defect to free the incarcerated bowel. It has proved less difficult to accom- plish this than to close the peritoneum in patients with a small peritoneal cavity and , large mass of abdominal viscera. A third advantage that becomes obvious is the ease with which the closure of the diaphragmatic defect is possible under direct vision. We have used an imbrication of the anterior lip of the defect over the | posterior in most patients. DIAPHRAGMATIC HERNIA IN INFANTS OPERATIVE FINDINGS Three of the infants in this series had eventration of the diaphragm which was suspected preoperatively, but not diagnosed without question (Table I). Of the re- maining 12 hernias, six were through the foramen of Bochdalek, two through the esophageal hiatus, one through the fora- men of Morgagni, two through a defect which seemed to combine the foramina of Bochdalek and Morgagni, and there was Taste I. Age at Associated Pt. Sex Age Symptoms Onset of | Operative Findings Side Anomalies Complications Symptoms LJ M 7 Cyanosis; hematemesis Birth Eventration L wks Dyspnea cs. F 3 Cyanosis 8 hrs. Esophageal hiatus Malrotation of colon Intestinal obstruction dys Defect Duodenal bands E.H. M 17. Cyanosis; convulsions 15m. Foramen of Morgagni Atelectasis m with sac P.D. F 13. Deformity of costal Eventration R m cage RL. M 4 Cyanosis; difficulty 2dys Foramen of Boch- L dys with feeding dalek with sac B.B. F 6 Cyanosis; dyspnea Birth Absence of L Agenesis of left lung: Died; subarachnoid hrs diaphragm Premature ossifi- hemorrhage cation of skull EC. M_ 11 Recurrent U.R.L Birth Foramen of Boch- L m Dyspnea; melena dalek D.H. 2 Cyanosis; dyspnea Birth Anterior defect L dys ww. M 9 Cyanosis; cough; Birth Esophageal hiatus m vomiting; hematemesis R.D. F 7m Dyspnea Birth Foramen of Boch- R_ Pleural sac encased dalek right lower lobe JS. M 20. Cyanosis; dyspnea 19 dys Foramen of Boch- L dys dalek BM. M 1d Cyanosis; periods of Birth Anterior defect L Died; congenital apnea heart lesion LS. M 6m Vomiting 2 mos. Foramen of Boch- R Partial situs inversus, Cardiac arrest, died dalek with sac duodenal obstruc- tion by portal vein and common bile duct N.G. M 5m Fever; dx pneumonia 2 mos. Eventration L SW. M 4m Vomiting 3 mos. Foramen of Boch- R dalek There would seem to be two additional physiologic advantages to the trans-thoracic approach. The thoracotomy closure is se- cure and safe by standard layer approxi- mation. In addition, there seems to be less ileus when there is no opportunity for the bowel to come in contact with a raw sur- face as in those instances when the fascia is left open and the skin and subcutaneous tissue alone are closed. one congenital absence of the left hemidia- phragm. Additional anomalies found in the thorax were agenesis of the lung in B.B., who died a few hours postoperatively, and a pleural sac about the right lower lobe in R. D. Mal- rotation of the colon in C. J. was corrected a month after thoracotomy when vomiting led to a diagnosis. L. S., who died three days after cardiac arrest and resuscitation 1009 KOOP AND JOHNSON on the operating table, had situs inversus of stomach and duodenum with both the por- tal vein and the common bile duct crossing the duodenum anteriorly. Splenectomy was carried out in two pa- tients. One infant with an eventration re- Cc quired a right nephropexy in the retro-pleu- ral space when the repaired diaphragm depressed the kidney sufficiently to kink the ureter. Crushing of the phrenic nerve was not found necessary. ANESTHESIA Positive pressure endotracheal anesthesia is most desirable if not essential. The anes- thetist has better control of the respiratory Annals of Surgery December, 19 5 2 situation under these circumstances than with any other variety of administration. This is especially obvious when the patient is turned on the side opposite that of the hernia prior to making the intercostal in- cision. The added encroachment on the B Fic. 1.—AP and lateral roentgenograms preoperatively and an AP view on the sec- ond postoperative day. available space for functioning lung by the shift of the thoracic contents with a change in position could produce serious sequellic without a method of controlling respiration under this circumstance. POSTOPERATIVE COURSE All thoracotomy wounds were closed primarily without drainage. No patient hat . i pneumothorax postoperatively of suffi 1010 Volume 136 Number 6 cient size to warrant aspiration. The return of the heart and mediastinum toward the midline was evident by physical examina- tion immediately in the postoperative period, and roentgen examinations taken about eight hours postoperatively showed about 80 per cent of the total return to nor- mal to be accomplished in that interval. Several patients showed roentgenologic evidence of fluid in the costophrenic sulcus, but none required thoracentesis. A definite haziness was discernible over the dome of the diaphragin postoperatively, but was uni- formly clear in two weeks (Fig. 1). Postoperatively all patients were main- tained on gastric suction drainage and in- travenous fluids until peristalsis was heard. Transfusions of blood or plasma were given for the first four to five days. Feedings were begun by gavage tube rather than by nursing bottle in most in- fants to prevent dilatation of the stomach with air, usually on the second or third postoperative day. Feedings were given by bottle to all by the end of the first postop- erative week. Special attention was given to keeping the air-ways clear of secretions by aspirat- ing the nasopharynx when indicated. Films of the chest were taken daily for three days, if indicated, and then at inter- vals of about four days until discharge. One-half of the incisional sutures were re- moved on the fourth or fifth day, and the remainder on the sixth or seventh day. RESULTS Three of the 15 patients in this series succumbed postoperatively. In each of them, death was attributable to other causes than the diaphragmatic hernia or its treatment. One four-pound infant (B. B.) had agenesis of the left lung, premature ossification of the skull, and subarachnoid hemorrhages, as well as complete absence of the left diaphragm. Another infant had been cyanotic for so long as to raise reason- able doubt concerning her cerebral func- tion. Autopsy revealed, in addition to an DIAPHRAGMATIC HERNIA IN INFANTS anterior diaphragmatic hernia, a congenital cardiac anomaly with large interventricular and interatrial septal defects. L. C. died with decerebrate rigidity after cardiac ar- rest and had the additional anomalies noted above. Of the remaining 12 patients, serious postoperative problems were encountered in only one (C. S.) where a subsequent operation was necessary for correction of malrotation of the colon causing duodenal obstruction. One infant (E. H.) had a par- tial atelectasis of the right lung relieved by aspirating a mucous plug from his trachea by catheter. The postoperative course of the remaining ten patients was consistently benign. No wound complications occurred. Discharge from the hospital was usually accomplished in less than ten days follow- ing operation. None of the successfully treated patients have had recurrences of symptoms or re- turn of the hernia. The period of followup varies from two months to two years. SUMMARY Fifteen congenital diaphragmatic defects occurring in infancy have been repaired by a trans-thoracic approach. Three infants who died postoperatively had associated congenital anomalies, two of them incom- patible with life. The advantages of the trans-thoracic over the abdominal approach in the repair of diaphragmatic defects in infants are discussed. BIBLIOGRAPHY 1 Donovan, E. J.: Congenital Diaphragmatic Hernia. Ann. Surg., 122: 569, 1945. 2 Harrington, S. W.: Congenital Diaphragmatic Hernias in Children. Ann. Surg., 115: 705. 1942, 3 Hartzell, J. B.: Congenital Diaphragmatic Her- nias in Children: Resume of 68 Cases Treated by Operation. Am. J. Surg., 48: 582, 1940. 4 Ladd, W. E., and R. E. Gross: Congenital Dia- phragmatic Hernia. New England J. Med., 223: 917, 1940. 5 Thorek, P.: Congenital Hernia, Anatomic and Surgical Importance of the Left Triangular Ligament of the Liver. Arch. Surg., 56: 238, 1948, 1011