Interruption of Patent Ductus Arteriosus in Premature Infants with Respiratory Distress Syndrome

Transcription

1 Interruption of Patent Ductus Arteriosus in Premature Infants with Respiratory Distress Syndrome Sidney Levitsky, M.D., Elizabeth Fisher, M.D., DharmapuriVidyasagar, M.D., Alois R. Hastreiter, M.D., Edward J. Bennett, M.D., Tonse N. K. Raju, M.D., and Keith Roper, M.D. ABSTRACT In infants with respiratory distress syndrome (RDS) hypoxemia inhibits closure of the patent ductus arteriosus (PDA), resulting in increased pulmonary blood flow with subsequent increased hypoxemia. In an attempt to interrupt this cycle 42 consecutive premature infants with RDS and PDA, weighing between 550 and 2,000 gm (average,,383 gm) and with an average gestational age of 3 weeks, were arbitrarily treated either medically (3 patients) or by interruption of the PDA (20 patients). Eleven patients who were initially treated medically could not be weaned from the respirator and later underwent operation. There were no operative or anesthetic deaths; late survival was 65% (20 patients). The last 3 patients were randomly divided into operative and nonoperative groups. Preliminary results revealed no significant differences in late survival between the two groups. Since the operative risk is minimal, further investigative efforts are indicated to settle this issue. The combination of prematurity, respiratory distress syndrome (RDS), and the presence of a patent ductus arteriosus (PDA) is frequently found in neonatal distress. In addition, hypoxemia inhibits closure of the PDA in premature infants with RDS, causing an increase in pulmonary blood flow, which-assuming the presence of a left-to-right shunt-results in further hypoxemia. If this state is allowed to progress, congestive heart failure may ensue, resulting in further deterioration of pulmonary function and continuation of this vicious cycle leading to death. From the Departments of Surgery, Pediatrics, and Anesthesiology, The Abraham Lincoln School of Medicine of the University of Illinois College of Medicine, Chicago, IL. Presented at the Twelfth Annual Meeting of the Society of Thoracic Surgeons, Jan 26-28, 976, Washington, DC. Address reprint requests to Dr. Levitsky, Department of Surgery, University of Illinois College of Medicine, PO Box 6998, Chicago, IL Several conflicting reports have appeared in the literature indicating the salutary effects of PDA interruption to treat this spectrum of events [9, 0, 7, 2, 23, 27. However, most of the studies either have involved too few patients or have examined mixed groups of patients operated upon for congestive heart failure following unsuccessful pharmacological therapy. There have been no previous studies limited to the relationship of PDA and RDS before the onset of congestive heart failure, nor have any randomized controlled studies been reported. The purpose of this article is to present our entire medical and surgical experience in 42 consecutive premature infants with severe RDS and PDA in an attempt to determine whether early operative intervention will decrease the morbidity and mortality of RDS. The last 3 patients in this series were treated as part of a prospective randomized study, and this report analyzes our preliminary findings. Methods Patient Selection All the patients were admitted to the high-risk nursery at the University of Illinois Medical Center between August 4, 973, and December 8, 975; two-thirds of the patients had been transferred from other institutions. The 42 patients in this series were divided into two groups: Group, consisting of infants of whom 0 were arbitrarily subjected to operative intervention because of deterioration following standard medical treatment, and Group 2, 3 patients who were randomly divided into either surgical or medical categories and are part of an ongoing clinical study (Table). Since the results achieved in Group patients suggested that surgical intervention involves minimal risk, the Clinical Research Review Board at our institution suggested that patients admitted t9 the study and assigned to medical therapy should 3

2 32 The Annals of Thoracic Surgery Vol 22 No 2 August 976 Clinical Data and Results in 42 Infants Having PDA lnterruption for RDS Gesta- Survival Age (days) when: No. of Birth tional Age Respirator Support Survival FDA Patients Treatment Weight (gm) (wk) (days) Number % Detected Extubated Discharged Group 0 Surgical,259 f f f f f 8 Medical, Group 2 8 Surgical,42 f f f f 22.5f 2a 90.5 f 27 2 Medical,476 f f 3 0 f f a 9.3 f 27 Medical,397f f 3 3.9f a 5.9 f f 3a 204 f 26 failure, operation p < undergo operation if, after ten days, medical treatment was unsuccessful. Thus there were three subgroups: surgical, medical, and medical failures who subsequently underwent operation (Fig ). Before a patient was entered into the Group 2 study, informed consent was obtained from the parents. Diagnosis of RDS The clinical features of RDS have been clearly defined by Silverman [26], Downes [8], and their co-workers. The most significant feature of the disease is the infant s inability to maintain adequate oxygenation while breathing room air. Although there may be many other causes of respiratory distress such as congenital anomalies, upper airway obstruction, congenital heart disease, infections, and aspiration, the Fig I. Protocol for prospective randomized study of RDS and PDA. most common form of respiratory distress in the newborn is hyaline membrane disease. The diagnostic criteria for RDS in this study were:. Onset of respiratory distress within 6 hours following birth and persisting for more than 4 hours. 2. Prematurity, defined as a gestational age of less than 37 weeks. 3. Presence of historical risk factors such as cesarean section delivery or a diabetic mother. 4. Abnormal chest roentgenograms, consisting of a diffuse reticulogranular pattern associated with poor aeration of the lung parenchyma. 5. Abnormal arterial blood gases, with the Pacq greater than 60 torr and Pao, below 40 torr when breathing room air. 6. Need for nasotracheal intubation and mechanical respiratory assistance. RDS and Respiratory Assistance Digitalize 24 hours) \ Improved / No lmprpvement (Random Distribution) Continued Improvement /Observe\ No Improvement Surgery No Surgery (Random Distribution) /\ Surgery No Surgery 2urgery No Improvement after 0 Days Surgery

3 33 Levitsky et al: PDA in Infants with Respiratory Distress Diagnosis of PDA All the infants admitted to the hospital with a diagnosis of RDS were evaluated daily for clinical signs of a PDA. A diagnosis of PDA was made if two or more of the following findings were present: Hyperactive precordium. Bounding peripheral pulses. Heart murmur at the upper left sternal border. Third heart sound or middiastolic rumble at the apex. Hepatomegaly and a typical continuous murmur were present in some patients. After the diagnosis of PDA was made, the infants were digitalized. Patients weighing less than,000 gm received 20 pglkg of digoxin; those over,000 gm were given 40 pglkg. Diuretics were added as required. Blood transfusions were given when necessary to keep the hematocrit above 40%. Within 24 hours after digitalization was begun, the infants were reevaluated for evidence of improvement in respiratory status as judged by improvement in blood gases, decrease in inspiratory oxygen concentration, or change to a lesser degree or type of respiratory assistance. Infants who showed improvement continued to receive digoxin and were observed; they form part of the medically treated group. Those who did not improve or whose condition deteriorated after initial improvement were subjected to random treatment selection (see Fig ). Treatment of RDS The basic principles of management for infants with hyaline membrane disease are based on providing an adequate thermal environment and sufficient calories, treating acidosis, correcting hypovolemia, and ensuring adequate oxygenation through respiratory support [4, 27, 29. The respiratory and circulatory status were evaluated every 24 hours by two pediatricians, a neonatologist and a cardiologist. If surgical therapy was selected, the operative procedure was performed within 2 to 24 hours. Opera tion All infants were transported to the operating room in a warmed incubator by an anesthesiologist. The basic anesthetic agent was 50% nitrous oxide in oxygen with controlled rebreathing utilizing a Rees variation of the Ayre's T-tube, so that the Paco, was limited to a range of 35 to 40 torr. This was achieved by limiting the fresh gas flow from the anesthesia machine, and the type of ventilation was that achieved with a T-piece, at a low volume and rapid rate and with a moderate amount of positive end-expiratory pressure [4]. The ancillary agent was tubocurarine or pancuronium at doses of 250 and 30 pg/kg, respectively [5. Because of the obviously limited surface area, the neonate was not placed on a warm-water mattress, but rather the operating room temperature was raised to 32 C. During the procedure the temporal pulse, temperature, and electrocardiogram were monitored. Blood pressure, blood gases, and other indicators, though helpful, were used only if available and easy to obtain. The intravenous fluid administered, 5% dextrose in lactated Ringer's solution, was limited to the amount required to ensure that drugs given through the intravenous tubing would reach the patient, and it was not allowed to exceed 3 mllkg/hour [3]. An oral endotracheal tube was used during the operation and was replaced at the end of the procedure by a nasotracheal tube, which was checked for position and length by a chest roentgenogram when the patient was returned to the high-risk nursery. The ductus arteriosus was exposed through a posterolateral thoracotomy in the fourth intercostal space [2. On observation, the ductus varied from two-thirds to one and one-third the size of the aorta. In several instances the ductus was noted to constrict during the dissection. The ductus arteriosus was singly or doubly ligated, depending on its length, with -gauge silk. There were no untoward events during any of the operations that necessitated administration of blood or inotropic agents, nor did we observe the increased friability of the ductus noted by previous authors [9. Open thoracotomy was limited to 5 minutes. There were no anesthetic or early postoperative deaths.

4 34 The Annals of Thoracic Surgery Vol 22 No 2 August 976 Results Clinical data for the 42 infants studied in the entire series are summarized in the Table. There were no significant differences between subgroups with respect to birth weight, gestational age, and age at PDA detection. Similarly, there were no differences in total time of mechanical ventilation except for the medically treated patients in Group 2. These latter differences are to be expected in view of the fact that ventilated infants in this group who were not improving after ten days of medical therapy were shifted, by virtue of the protocol, to the surgical group (ie, medical failure and operation). In order to compare the effectiveness of each of the treatment modalities, success was measured in terms of survival at discharge. When viewed in their totality (Groups and 2) there were no significant differences between the medically treated and operated groups-62% versus 55 O/O survival, respectively. Neverthe- Fig 2. Chest roentgenograms obtained from a patient in Group. (A) Preoperative roentgenogram shows diffuse reticulogranular pattern of the lung parenchyma, with cardiac enlargement. (B) Roentgenogram obtained 4 hours postoperatively reveals marked clearing of the lung fields and a decrease in cardiac size. less, when late operative survival was achieved, the change in clinical condition with rapid clearing of pulmonary congestion was often dramatic (Fig 2). There were no differences in survival between the medical and surgical subgroups of the randomized study (Group 2), with 58 and 50% survival, respectively. However, there was a significantly increased survival of 82% in the medical failure-surgical subgroup (p < 0.05). These differences were anticipated, since this subgroup was preselected and admittance to it required at least ten days' survival. As would be expected, the ten days of mechanical ventilation for the medical failure-surgical group delayed extubation significantly compared with the surgical group (p < 0.05) (see the Table). The greatest salutary effect of early operative intervention appeared to be early discharge, in that patients in the surgical subgroup were discharged at 90.5 k 27 days of age compared with 9.3 f 27 and 204 f 26 days, respectively, in the medical and the medical failure-surgery subgroups. Since the number of patients in each of the subgroups is small, these differences must be viewed only as preliminary findings until the total study, numbering in excess of 75 patients, is completed. A B

5 35 Levitsky et al: PDA in Infants with Respiratory Distress Comment There is ample evidence to show that the incidence of PDA is higher in premature as compared with term infants and is even higher in premature infants with RDS [7,. The mechanismsfor persistence of the PDA are thought to relate to the reduced cholinergic innervation of the ductus and the decreased ductal response to oxygen, resulting in a loss of vasoconstrictive ability [l, 20. The rationale for interruption of the PDA in infants with RDS is based on the assumption that a left-to-right shunt is present and significantly increases pulmonary blood flow. Thus, the relationship of pulmonary vascular resistance (PVR) to systemic vascular resistance is of paramount importance. Since the ratio of pulmonary arterial muscular tissue to lung parenchyma is low in the 2-week-old fetus and gradually increases with age, the PVR will be lower in the premature than in the term infant [29]. However, acidosis, hypercapnia, hypoxemia, and positive-pressure ventilation tend to raise PVR [9,2. It has been shown that increased pulmonary blood flow without heart failure has a negligible effect on gas exchange and that substantial intrapulmonary shunting with transudation of fluid does not occur until there is left atrial pressure and volume overload [7, 8. On the other hand, lung compliance declines with small increases in pulmonary blood flow [5]. Hence, several observers have thought that PDA interruption is not indicated unless congestive heart failure is present or a significant left-to-right shunt is demonstrated by aortography [9, 27. Others, however, have suggested that prompt PDA ligation may prevent the development of severe bronchopulmonary dysplasia [lo, 22. It is obvious from all the conflicting data that a prospective randomized study of the various therapeutic options is necessary. Before embarking on a study of this type, in which there may be small differences, it is obvious that surgical morbidity and mortality must be negligible. Our data, as well as those of others, indicate that this is the case [lo, 6. In addition, survival following medical therapy in an institution undertaking this study should be comparable to that observed in other institutions. Neal and associates [2] have described 396 infants with RDS admitted from January, 970, to December, 972, with a 29% survival in patients who had assisted ventilation, which is comparable to the 58% survival in our medical subgroup. The classic signs and symptoms of PDA are frequently lacking in infants, and in premature neonates with RDS the signs may be evanescent and altered by pulmonary disease and assisted ventilation [2, 23. In addition, recurrent bradycardia and apnea, tachypnea, pulmonary rales, rising Pcoz, increasing opacification of the lungs, and decreasing respiratory function after initial improvement have been found to be signs of a PDA complicating RDS. Like others [2, 9, we found that the electrocardiogram does not have great diagnostic value, though it may serve to exclude associated congenital heart defects. Similarly, the chest roentgenogram is of little diagnostic value during the acute phase of RDS, as the cardiac and pulmonary vascular shadows are obscured by opacification of the lungs. Although the signs are atypical, a clinical diagnosis of the presence of a PDA can usually be made. Evaluation of the magnitude of the left-to-right shunt is more difficult. Cardiac catheterization, advocated by some authors, carries significant risk in a sick premature infant and yields little more useful information than an aortogram obtained through an umbilical artery catheter [9, 6, 27. However, shunt size can be evaluated by an even safer, noninvasive technique+chocardiography. Carter and Bowman [6] have shown echocardiographic left atrial dimension to correlate well with the magnitude of left-to-right shunts in patients with ventricular septa defect. The same principle has been applied to infants with RDS and PDA. Investigators have used increases in the left atrial/ aortic root ratio to identify premature infants with RDS who had a large left-to-right shunt P4. Goldberg and colleagues [3], in a preliminary report of a prospective study using echocardiographic criteria, indicated that the ratio was helpful in selecting 3 out of 83 premature infants with RDS and PDA for operative intervention. This preliminary report of a prospective, randomized study evaluating the effects of PDA in premature infants with RDS suggests no evi-

6 36 The Annals of Thoracic Surgery Vol 22 No 2 August 976 dence of increased survival with early operative interruption of the PDA. Since the operative procedure appears to carry minimal risk, further investigative efforts are indicated and necessary to settle this issue. References. Aronson S, Gennser A, Owman CH, et al: Innervation and contractile response of the human ductus arteriosus. Eur J Pharmacol :78, Baylen BG, Meyer RA, Kaplan S, et al: The critically ill premature infant with patent ductus arteriosus and pulmonary disease-an echocardiographic assessment. J Pediatr 86:423, Bennett EJ: Fluids for Anesthesia and Surgery in the Newborn and the Infant. Springfield, IL, Thomas, Bennett EJ, Ignacio A, Pate K, et al: The Rees system in infants: fresh gas flow and Paco,. Middle East J Anesth (in press) 5. Bennett EJ, Ramamurthy S, Dalal FY, et al: Pancuronium and the neonate. Br J Anaesth 47:75, Carter WH, Bowman R: Estimation of shunt flow in isolated ventricular septa defect by echocardiogram (abstract). Circulation 48:Suppl 4:64, Danilowicz D, Rudolph AM, Hoffman JIE: Delayed closure of the ductus arteriosus in premature infants. Pediatrics 37:74, Downes JJ, Vidyasagar D, Morrow GM, et al: Respiratory distress syndrome of newborn infants: I. New clinical scoring system with acid base and blood gas correlation. Clin Pediatr 9:325, Edmunds LH, Gregory GA, Heymann MA, et al: Surgical closure of the ductus arteriosus in premature infants. Circulation 48:856, Gay JH, Daily WJR, Meyer BHP, et al: Ligation of the patent ductus arteriosus in premature infants: report of 45 cases. J Pediatr Surg 8:677, 973. Girling DJ, Hallidie-Smith KA: Persistent ductus arteriosus in ill and premature infants. Arch Dis Child 46:77, Glenn WWL, Bloomer WE, Spear HC: Operative closure of the patent ductus arteriosus: a report of 0 operations without mortality. Ann Surg 43:47, Goldberg SJ, Allen HD, Sahn DJ, et al: A prospective 2/2 year experience with echocardiographic evaluation of prematures with patent ductus arteriosus and respiratory distress syndrome (abstract). Am J Cardiol35:39, Gregory GA, Kitterman JA, Phibbs RH, et al: Treatment of IRDS with CPAP. N Engl J Med 284: 333, Griffin AJ, Cassels DE: Pulmonary compliancean index of cardiac status in infants with heart disease. Am J Cardiol 25:99, KilmanJW, KakosGS, WilliamsTE, et al: Ligation of patent ductus arteriosus for persistent respiratory distress syndrome in premature infants. J Pediatr Surg 9:277, Lees MH: Patent ductus arteriosus in premature infants-a diagnostic and therapeutic dilemma (commentary). J Pediatr 86:32, Lees MH, May RC, Ross 88: Ventilation and respiratory gas transfer of infants with increased pulmonary blood flow. Pediatrics 40:259, Malik AB, Kidd BSL: Independent effects of changes in H+ and CO, concentrations on hypoxic pulmonary vasoconstriction. J Appl Physiol 34:38, McMurphy DM, Heymann MA, Rudolph AM, et al: Developmental changes in construction of the ductus arteriosus: responses to oxygen and vasoactive agents in the isolated ductus arteriosus of the fetal lamb. Pediatr Res 6:23, Neal WA, Bessinger FB, Hunt CE, et al: Patent ductus arteriosus complicating respiratory distress syndrome. J Pediatr 86:27, Northway WH, Rosen RC, Porter DY: Pulmonary disease following respirator therapy of hyaline membrane disease. N Engl J Med 276:357, Rudolph AM, Mayer FE, Nadas AS, et al: Patent ductus arteriosus: a clinical and hemodynamic study of 23 patients in the first year of life. Pediatrics 22:892, Silverman NH, Lewis AB, Heymann MA, et al: Echocardiographic assessment of ductus arteriosus shunt in premature infants. Circulation 50:82, Silverman WA, Andersen DH: A controlled trial of effects of water mist on obstructive respiratory signs, death rate and necropsy findings among premature infants. Pediatrics 7:L Silverman WA, Fertig JW, Berger AP: The influence of the thermal environment upon the survival of newly born premature infants. Pediatrics 22:876, Thibeault DW, Emmanouilides GC, Nelson RJ, et al: Patent ductus arteriosus complicating the respiratory distress syndrome in preterm infants. J Pediatr 86:20, Vidyasagar D, Chernick V: Continuous positive transpulmonary pressure in hyaline membrane disease. Pediatrics 48:296, Wagenvoort CA, Neufeld HN, Edwards JE: The structure of the pulmonary arterial tree in fetal and early postnatal life. Lab Invest 0:75, 96 Discussion DR. GERARD A. KAISER (Miami, FL): I congratulate Dr. Levitsky and his associates on this very fine paper and commend them for the use of a randomization protocol in this study. We have recently reviewed a similar group of 20 infants operated upon between May, 972, and July,

7 37 Levitsky et al: PDA in Infants with Respiratory Distress 975. Our results support their views, first, that no formal cardiac catheterization is required for physical diagnosis and that noninvasive techniques will establish the diagnosis, and second, that within the proper setting an operation can be performed with minimal operative mortality. We too have been fortunate in having no operative deaths in this group of patients. I do believe, however, that there are two groups of patients who fall into this general category. Eleven of the 20 infants we operated upon had severe associated hyaline membrane disease that necessitated mechanical ventilatory.support from the first day of life. Only 3 of these infants ultimately survived to leave the hospital; the others died from various associated anomalies as well as their severe pulmonary disease. In contrast, 8 of the 9 infants in the other subgroup, with severe congestive heart failure and RDS but without hyaline membrane disease, are alive even though these infants required ventilatory support prior to operation. I think surgical ligation of the PDA in a premature infant is an appropriate and successful mode of therapy when congestive heart failure is refractory to medical management. Proof is lacking, however-as Dr. Levitsky states-to establish its efficacy in the management of premature infants with hyaline membrane disease. I look forward to future reports from his group on the randomized study of this subset. DR. ROSS KYGER (Houston, TX): I would also like to congratulate Dr. Levitsky and his colleagues on a very well done study. At the Texas Heart Institute and Texas Children s Hospital, we too have struggled with the dilemma of when to operate on these tiny infants with a patent ductus. As experience has accumulated, our neonatologists have tended to favor earlier surgical therapy. They now believe that any premature infant with a ductus who requires five days of mechanical ventilatory support is a candidate for ligation. The echocardiogram has proved to be a very useful tool in helping us differentiate the relative contribution of congestive heart failure and hyaline membrane disease to the infant s pulmonary problem. With the echocardiogram one can determine left atrial size and compare it to the size of the aortic root. By following this ratio, one can then gauge the effectiveness of medical therapy for congestive heart failure. If the left atrium should enlarge despite adequate medical therapy, the child then has the ductus ligated. We use an extrapleural approach and perform the procedure with the infant in an Ohio warmer in order to maintain the body temperature. We have operated on 35 such infants, with 30 survivors leaving the hospital. Our neonatologists assure us that modern ventilatory methods are preserving the essential nervous system function of these babies, but they think prolonged respiratory support contributes to the problem of pulmonary dysplasia. Since the procedure has proved to be safe even in infants weighing less than,000 gm, we think it should be used early in an effort to shorten respirator time and preserve pulmonary function. DR. LEVITSKY: I thank the discussants for their kind supporting remarks. I want to be certain that we have differentiated the types of patients we operated upon, because the literature is quite obscure on this point. The indication for operation in this group of patients is not congestive heart failure secondary to a large PDA with a huge left-to-right shunt; in fact, we have not measured the shunt size in our patients. Our indication for operation in this study was hyaline membrane disease plus the presence of a PDA. It has been shown, as we indicated in the paper, that even patients with a small shunt that is not necessarily large enough to increase the size of the left atrium will exhibit a decrease in pulmonary compliance. In conclusion, I wish to emphasize that the group of patients we have reported is a small subgroup of all the infants with respiratory distress syndrome.