- Email: [email protected]
Aggressive Treatment of Acquired Phrenic Nerve Paralysis in Infants and Small Children
Aggressive Treatment of Acquired Phrenic Nerve Paralysis in Infants and Small Children Robert Shoemaker, M.D., George Palmer, B.S., John W. Brown, M.D., a n d Harold King, M.D.
ABSTRACT Transthoracic diaphragmatic plication is a simple and effective means of treatment for phrenic nerve injury and resultant diaphragmatic paralysis, and should be undertaken as soon as the diagnosis of diaphragmatic paralysis is established in a child with substantial respiratory difficulty. Seven of more than 1,500 patients sustained phrenic nerve injury during operation for congenital heart disease at our institution over the past five years. All but 1 patient were less than 5 months old, and the average weight was 3.8 kg. Five of the 7 could not be weaned from the ventilator, and the other 2 had persistent postoperative tachypnea and stridor. Before 1980, 3 patients were intubated for an average of 16 days prior to diagnosis by fluoroscopy. Since then, 4 patients have been intubated for an average of 7.8 days before diagnosis. All 7 underwent transthoracic diaphragmatic plication and were extubated by 6 days after operation (mean, 3.7 days). Six patients survived and are doing well. The 1 death occurred in 1976 in a premature infant with multiple congenital cardiac lesions; diagnosis was delayed until the twenty-eighth day after intubation and transthoracic diaphragmatic plication, until the sixty-second day. This infant died of pneumonia one month after extubation. These data support our policy of establishing an early diagnosis of phrenic nerve injury and early treatment utilizing transthoracic diaphragmatic plication in symptomatic children.
Phrenic nerve injury and resultant diaphragmatic paralysis often cause life-threatening respiratory embarrassment in infants and young children [l-31. Prior to the advent of palliative or corrective operation for congeniFrom the Department of Surgery, Indiana University School of Medicine, Indianapolis, IN. Presented at the Seventeenth Annual Meeting of The Society of Thoracic Surgeons, Jan 26-28,1981, Los Angeles, CA. Address reprint requests to Dr. Brown, Department of Surgery, Room 212, Emerson Hall, Indiana University School of Medicine, 1100 W Michigan St, Indianapolis, IN 46223.
tal heart disease, most patients sustained phrenic nerve injury and diaphragmatic paralysis secondary to birth trauma [I, 2, 41. An increasing number of phrenic nerve injuries associated with thoracic surgical operations have shown up in recent studies [3, 5-10]. Traditional treatment, regardless of etiology, has been prolonged ventilatory support [l-111. With this method of nonoperative therapy, mortality in both groups of patients is 20 to 25% [l-111. One effective means of surgical treatment for this complication is diaphragmatic plication, and the mortality for diaphragmatic paralysis should be essentially zero if plication is done early [2, 5, 7,12, 131. Plication has been effective for those sustaining birth trauma [ 2 , 41, but only a small number of symptomatic patients have undergone plication after diaphragmatic paralysis secondary to thoracic surgery [5-9, 13, 141. This study describes 7 patients, all of whom were plicated for symptomatic diaphragmatic paralysis after operation for congenital heart disease.
Material and Methods From 1976 through 1980, more than 1,500 patients underwent palliative or corrective operation or both for congenital heart disease at the Riley Hospital for Children at the Indiana University Medical Center. Seven of them (5 boys and 2 girls) sustained phrenic nerve injury resulting in symptomatic diaphragmatic paralysis and underwent transthoracic diaphragmatic plication. Excluding 1 older child, each of these patients weighed less than 10 kg (average, 3.8 kg) and was less than 5 months old (average, 36 days). The oldest child was 3 years of age at the time of operation and had had two previous thoracotomies. Of the 7 patients, 5 could not be weaned from the ventilator, and the other 2 , after extubation, had persistent tachypnea, stridor, and difficulty feeding. The diagnosis of phrenic nerve paralysis was confirmed by
251 0003-4975/81/090251-09$01.25@ 1981 by The Society of Thoracic Surgeons
252 The Annals of Thoracic Surgery Vol 32 No 3 September 1981
Table 1 . Data on 7 Patients Undergoing Cardiac Surgical Procedures Resulting in Diaphragmatic Paralysis
Patient No.
Sex
Date of Operation
1
M
12/75
Age 8d
6d 5 wk 1d 4.5 mo
Congenital Weight Heart (kg) Disease 2.7
2 3 4 5 6
M F M M M
2/77 3/78 3/80 5/80 7/80
4 wk
3.7 4.1 3.4 4.8 4.1
7
F
7/80
3 yr
10.0
Procedure
PDA, coarctation
+
PDA pericardial window," coarctation repair PDA, VSD PDA, PA band PDA, PA band PDA, VSD Blalock-Taussig TOF TGV, PS, ASD Blalock-Hanlon TGV, VSD PDA + ASD, PA band coarctation, Coarctation reTA + PDA pair ASD, VSD, SIP Total repair PA band 2, PDA
Incision
Anticipated Damage to Phrenic Nerve
L thoracotomy
No
x 2
L L R R L
thoracotomy thoracotomy thoracotomy thoracotomy thoracotomy
Sternotomy
Yes No Yes No No
No
=Atrial perforation at heart catheterization. PDA = patent ductus arteriosus; VSD = ventricular septal defect; PA = pulmonary artery; TOF = tetralogy of Fallot; TGV = transposition of great vessels; PS = pulmonary stenosis; ASD = atrial septal defect; TA = tricuspid atresia; SIP = status post.
fluoroscopic demonstration of immobility or by paradoxical motion of the diaphragm to spontaneous respiration. The cardiac surgical procedures that resulted in diaphragmatic paralysis in these patients are listed in Table 1. All 7 children had multiple congenital cardiac defects, for a total of nineteen cardiac lesions. Five infants had ventricular septal defects, and 3 had atrial septal defects. An associated patent ductus arteriosus and coarctation affected 5 and 2 of the patients, respectively. At the time the phrenic nerve was injured, 5 of the 7 patients were undergoing multiple concomitant procedures. One patient underwent a Blalock-Taussig shunt for tetralogy of Fallot and 1, a Blalock-Hanlon procedure for transposition of the great arteries. Five of 6 infants undergoing palliative procedures had pericardiotomy, and 4 of these 6 had dissections performed both anterior and posterior to the phrenic nerve. Closure of an associated patent ductus in 4 infants was the most common operation. Pulmonary artery banding and ductus closure were done in 3 and were the most common combined procedures.
Prior to the diagnosis of diaphragmatic paralysis on the ipsilateral side, 4 patients had a left thoracotomy and 2 , a right thoracotomy. One patient experienced diaphragmatic paralysis following a median sternotomy. Although older, she had had a previous left thoracotomy for ductus ligation and pulmonary artery banding in infancy with a second thoracotomy for pulmonary artery band adjustment. A median sternotomy incision was performed for the correction. The pulmonary artery band had migrated distally, resulting in a difficult dissection in the region of the left pulmonary artery and left phrenic nerve. An outflow patch extending far out onto the left main pulmonary artery was required. In 2 patients, it was suspected that phrenic nerve injury had been done at the time of the cardiac operation. One was an obvious laceration during a Blalock-Taussig shunt, and at the time of the shunt procedure, a primary anastomosis of the nerve was performed. Postoperatively, the patient could not be weaned from the ventilator and demonstrated paradoxical motion of the diaphragm; plication was
253
Shoemaker et al: Acquired Phrenic Nerve Paralysis in Children
A third approach was used in 4 patients. done later. The other patient had no obvious injury to the phrenic nerve, but damage was Central plication (central pleats in the diaconsidered likely from stretch of the nerve phragm) were created as described by Schwartz during clamping of the aorta and the in- and Filler [7] (Fig 5). In this procedure, the sutrapericardial left pulmonary artery during tures are placed in the diaphragm so as to avoid injury to the branches of the phrenic nerve. ductus division. They are placed as partial-thickness sutures excluding only the peritoneum in order to avoid Operative Technique Plication to flatten the diaphragm in its inspi- injury to the abdominal organs. This is done by ratory position was performed through a post- grasping the diaphragm with a pair of vascular erior lateral thoracotomy incision in all pa- forceps between phrenic nerve branches, thus tients. Three different surgical techniques were tenting the diaphragm up away from the utilized. The classic "pants-over-vest" repair peritoneum as the needle is passed through it. was used in 1 patient (Fig 1). The abdominal Either of the latter two methods are preferable cavity was entered, but the repair was un- to the pants-over-vest technique. usually complicated by bowel prolapse through the diaphragmatic incision. Keeping the bowel Results safely retracted away from the closure also Our 7 patients fall into two groups: an early made the procedure difficult. group (operated on before 1980) and a late A circumferential incision through the dia- group (after 1980). The 3 patients in the early phragm near the chest wall avoided entry into group were intubated for an average of 16 days the peritoneal cavity and was the technique prior to the diagnosis of diaphragmatic pautilized in 2 patients (Fig 2). This approach was ralysis by fluoroscopy, whereas the 4 patients devised and used by one of us (H. K.) and, to in the late group were intubated for 7.8 days our knowledge, has not been described previ- prior to diagnosis. The largest difference beously. Interrupted sutures are placed in a radial tween the early and late groups was the number fashion from the thoracic wall diaphragmatic of days between diagnosis and plication-31.7 attachments so as to flatten the diaphragm and, days and 4.3 days, respectively. The average at the same time, avoid major phrenic nerve number of days spent in the intensive care unit reflected the pulmonary status of these patients. fibers (Figs 3, 4). The early group had an average of 59 days and the late group, an average of 15.5 days. The avFig 1 . Pants-over-vest technique: (A) In this repair, an erage number of days of mechanical ventilatory incision is made through the diaphragm and then closed support after plication was less than 4 in both with the outer edge sutured more centrally under the groups. medial portion. ( B ) The edge of the medial portion is Prior to 1980, 1 patient received a tracheosthen tacked laterally.
251 The Annals of Thoracic Surgery Vol 32 No 3 September 1981
Fig 2 . Circumferential pleat technique: ( A ) An incision is made circumferentially near the chest wall without incising the peritoneum. Sutures are then placed from the thoracic wall diaphragmatic attachments centrally to shorten the diaphragm. ( B ) Completed repair.
Fig 4 . (Patient 2.) After plication, the left hemidiaphragm is in its normal location.
bation. This infant died of pneumonia one month after discontinuing mechanical ventiFig 3 . (Patient 2 . ) Chest roentgenogram showing elevalator support. Also before 1980, plication was tion of the left hemidiaphragm following phrenic nerve delayed in 1 patient due to a wound infection paralysis. and in another, for congestive heart failure. To date, there have been no recurrences with tomy after 36 days of intubation, and another any of the plication techniques. Diaphragmatic patient died. The latter was a premature infant function has been assessed by fluoroscopy in 3 with multiple congenital cardiac lesions in of the 6 survivors. Function was demonstrated whom the diagnosis of diaphragmatic paralysis in all 3. One of these patients was studied sevwas delayed the longest (28 days) and in whom eral years following use of the circumferential plication was delayed until 59 days after intu- pleating technique (see Fig 2), and the other 2
255
Shoemaker et al: Acquired Phrenic Nerve Paralysis in Children
Fig 5. Central pleating technique: (A) Mattress sutures are placed, with care taken to avoid major nerve fibers and entry into the abdominal cavity. (B) Completed repair. (See text for details.)
were studied several months after use of the central pleating technique (see Fig 3). All 6 survivors are doing well with a mean follow-up of 21 months (range, 9 to 50 months) (Table 2). Comment Symptomatic unilateral diaphragmatic paralysis is an uncommon cause of respiratory difficulty in neonates and infants [l-4, 11-14]. However, recent reports indicate a rising number of patients with diaphragmatic paralysis secondary to thoracic and cardiac operations [5-lo]. In their large series of children undergoing operation for congenital heart disease, Mickell and associates [8] found the incidence of symptomatic 'diaphragmatic paralysis to be 1.7%. Our data indicate that about 0.5% of patients who undergo operation for palliation or correction of congenital heart disease will sustain symptomatic diaphragmatic paralysis. Some authors suggest that infants are more likely to become symptomatic from diaphragmatic paralysis than older children [6, 8, 91.
This point is well illustrated in the series of Mickell and co-workers [8]: 14 of 32 patients with documented unilateral diaphragmatic paralysis were less than 8 months old. Of these, 10 required prolonged ventilatory support for an average of 24 days with an average stay in the intensive care unit of 26 days. Five required tracheostomy, and 3 died. One patient underwent plication and is alive. In the group of patients older than 18 months, there was no mortality and no plications were necessary. Moreover, the number of days in the intensive care unit and on the ventilator were not different from those patients having an uneventful recovery from cardiac operation. Data from other authors corroborate this finding [6]. Our experience with infants seen prior to 1980, when plication was delayed, was similar to that of Mickell and colleagues [8]. One patient required tracheostomy, and another died after an average of 42 days of ventilatory support. Both required prolonged stays in the intensive care unit. Among the infants seen in 1980, there were no deaths and no tracheostomies were required. An average of 13 days was necessary for intubation, and an averge of 15.5 days in the intensive care unit. Although both of our groups had dia-
256 The Annals of Thoracic Surgery Vol 32
No 3 September 1981
Table 2 . Data on 7 Patients Undergoing Operation for Diaphragmatic Paralysis Intubation Patient No.
Days from Injury to DX a
Ventilatory Support
Days to Plication after DX
Days before DX
Days after Plication
Days in ICU
1
14 and 28"
Yes
29
28
5
82
2 3
21 11
No Yes
23" 43'1
9 11
0 6
34 60
4 5 6 7
1 10 4 14
Yes Yes No Yes
4 2 10 1
0"
3 2 2 4
8 18 13 23
10 7
14
Techn:ique/Results l/Died of bronchopneumonia 26 d after plication 2iGood; motion returned 2IGood; tracheostomy before plication; n o late fluol-oscopy 3/GOOd; no late fluoroscopy 3/Good; no late fluoroscopy 3/Good; motion returned 3/Good.; motion returned
aDiagnosis of diaphragmatic paralysis. I'Fatient had two operations before diagnosis. "Delayed due to wound infection. "Delayed due to congestive heart failure. "Obvious injury. DX = diagnosis; ICU
=
intensive care unit; 1 = pants-over-vest; 2 = circumferential pleat; 3
=
central pleating.
position of infants further reduces vital caphragmatic plication, we attribute the lower pacity [61. morbidity and mortality in the late group to aggressive diagnosis and treatment. Prior to 1980, 3. Pulmonary compliance is, decreased on the ipsilateral side [8] in that the diaphragm is patients were intubated an average of 16 days unable to resist the negative intrapleural prior to fluoroscopic diagnosis and waited 31.7 pressure and thus moves paradoxically with days before plication, whereas in 1980, they each respiration. were diagnosed at 7.8 days and the diaphragm was plicated 4.3 days later. These data indicate 4. Infants have relatively small intrabronchial calibers, a condition facilitating obstruction that plication reduces the number of days of and atelectasis by retained secretions [6]. ventilatory support. Other authors [3, 6-81 have discussed the Diaphragmatic paralysis should be suspected pathophysiology of diaphragmatic paralysis when the patient cannot be weaned from the and stressed that infants tolerate diaphragmatic ventilator or the patient's arterial blood gases paralysis less well than older children for four show retention of carbon dioxide despite apessential reasons. parent adequacy of respiratory efforts. Standard Infants are dependent on diaphragmatic chest roentgenogram made when the infant is contraction for adequate gas exchange and if on mechanical ventilation usually fails to show there is unilateral diaphragmatic paralysis, elevation of the hemidiaphragm, because posup to 50% of pulmonary function is lost. The itive airway pressure tends to keep the flail diaintercostal muscles of infants are weaker and phragm depressed. Opacification of the pleural at a mechanical disadvantage 181. space by atelectasis, effusion,. or both may also Infants have increased mediastinal mobility. make diagnosis by roentgencigraphy elusive. This means that in the presence of diaFluoroscopy is the diagnostic procedure of choice, and it can be performed at the bedside phragmatic paralysis, the mediastinal contents are shifted to the contralateral side on in most instances. Once the diagnosis of diainspiration. This decreases pulmonary func- phragmatic paralysis is suspected, the patient tion on the contralateral side as well as the should undergo fluoroscopy while breathipsilateral side [3]. The usual recumbent ing spontaneously. If diaphragmatic paralysis
257
Shoemaker et al: Acquired Phrenic Nerve Paralysis in Children
Table 3 . Mortality Following Acquired Phrenic Nerve Paralysis with Congenital Heart Disease Plication
Author
Year
City
Deathsa
Percent
Unplicated
Weldon [lo] Mickell et a1 [8] Greene et a1 [6] Othersen and Lorenzo [14] Schwartz a n d Filler [7] Mickell et a1 [81 Shoemaker et a1 [this report] Jewett and Thomson [131 Sethi and Reed [5] Shoemaker et a1 [this report]
1978 1978 1975 1977
St. Louis Pittsburgh Chicago Charleston
1 (4) 3h ( 1 1 ) C 1 (4) 0 (1)
25
1978
Boston
1“(2)
50
1978 1981
Pittsburgh Indianapolis
0 (1) 1 (3)
33
1964
Buffalo
I f (3)
33
1971 1981
Kansas City Indianapolis
lP (2)
50
0 (4)
0
Plicated lated
Plicated early
27 25
0
0
“Number in parentheses is total number of patients. bLate mortality in 1 child due to congenital heart disease. “Includes five tracheostomies. dMore than two weeks after diagnosis. ‘Cardiac death one month postoperatively. ‘Secondary to recurrence after plication; not treated. gSecondaryto aspiration pneumonia two weeks after plication.
is confirmed, surgical plication should be promptly carried out to reduce the morbidity and mortality associated with prolonged mechanical ventilation. Transthoracic plication can be performed easily by several techniques with minimal morbidity. Return of diaphragmatic function has been documented in 3 of our severely ill patients. If there is obvious phrenic nerve injury at the time of the initial operation in infants, concurrent plication using a technique that would allow later return of function should be considered. This would avoid the necessity of a later thoracotomy if the infant did not tolerate diaphragmatic paralysis. Plication does not preclude return of function, as shown by 3 of our patients who had documented return of function. Our data support previous studies suggesting age-dependent management of patients with diaphragmatic paralysis [6]. The 20% mortality with diaphragmatic paralysis due to birth trauma has now been demonstrated in most series of patients with unplicated diaphragmatic paralysis [l-4, 6, 111. Series of patients with acquired diaphragmatic paralyses, if unplicated, show a similar mortality [8, 101, as do results of late plications for diaphragmatic
paralysis secondary to operation for congenital heart disease [6]. Our early experience is similar. Recent experience with 4 patients, however, suggests that the number of days of mechanical ventilation, the length of stay in the intensive care unit, and mortality can be greatly reduced by aggressive diagnosis and treatment (Table 3). A retrospective analysis of 7 infants with diaphragmatic paralysis demonstrates a reduced morbidity and mortality for this complication when the diagnosis and treatment are approached aggressively. We recommend fluoroscopic evaluation of infants at risk for this complication who cannot be weaned from mechanical ventilation following operation for congenital heart disease. Prompt diaphragmatic plication in affected patients will reduce the hazards of prolonged mechnical ventilation and will not adversely affect return of phrenic nerve function. References 1. Richard J, Chevalier V, Capelle R, et al: La paralysie diaphragmatique obstetricale. Arch Fr Pediatr 14:563-598, 1957 2. Stauffer UG, Rickham PP: Acquired eventration
258 The Annals of Thoracic Surgery Vol 32 No 3 September 1981
of the diaphragm in the newborn. J Pediatr Surg comment about when they elect to use fluoroscopy, 716355640,1972 whether they do it in their intensive care unit, and 3. Wayne ER, Campbell JB, Burrington JD, Davis how they carry it out in a relatively sick infant. WS: Eventration of the diaphragm. J Pediatr Surg Certainly not every infant who has a phrenic nerve 9:643-651, 1974 paralysis requires diaphragmatic plication. In my 4. Bishop HC, Koop CE: Acquired eventration of experience it sometimes is an incidental finding obthe diaphragm in infancy. Pediatrics 22:1088- served relatively late after the operation. I think that 1096, 1958 infants with stiff lungs, pulmonary edema, high 5. Sethi G, Reed WA: Diaphragmatic malfunction pulmonary arterial pressure, or nutritional depletion in neonates and infants. J Thorac Cardiovasc are certainly at greater risk to become respirator deSurg 62:138-143, 1971 pendent following phrenic nerve dysfunction. I was 6. Greene W, L’Hereux P, Hunt CE: Paralysis of the interested that the authors found a greater incidence diaphragm. Am J Dis Child 129:1402-1405, 1975 of this problem in 1980 and wonder if they have a 7. Schwartz MZ, Filler RM: Plication of the dia- comment about that. In reading the manuscript, I noted that one inphragm for symptomatic phrenic nerve pastance of phrenic nerve dysfunction occurred folralysis. J Pediatr Surg 13:259-263, 1978 8. Mickell JJ, Oh KS, Siewers RD, et al: Clinical im- lowing ductus division. It has b’een our policy not to plications of postoperative unilateral phrenic perform ductus division in premature infants, innerve paralysis. J Thorac Cardiovasc Surg 76: fants, and younger children but to ligate the ductus. 297-304, 1978 Particularly in premature infants, we frequently use a 9. Mearns AJ: Iatrogenic injury to the phrenic nerve hemoclip to close the ductus. This can be done with in infants and young children. Br J Surg 64:558- minimal dissection and might perhaps decrease the risk of phrenic nerve problems with that procedure. 560, 1977 Topical hypothermia in the pericardium has been 10. Weldon CS: Discussion of Mickell JJ, et a1 [8] 11. France NE: Unilateral diaphragmatic paralysis indicated as a factor in the genesis of phrenic nerve and Erb’s palsy in the newborn. Arch Dis Child dysfunction postoperatively. I would be interested in the authors’ comments about the possible effect of 29:357-359, 1954 12. Symbas PN, Hatcher CR Jr, Waldo W: Dia- topical hypothermia in their patients, although I phragmatic eventration in infancy and child- noted from the manuscript that most of them underwent closed procedures. hood. Ann Thorac Surg 24:113-119, 1977 13. Jewett TC Jr, Thomson NB Jr: Iatrogenic eventration of the diaphragm in infancy. J Thorac Car- DR. AVRAHAM DOV MERAV (Bronx, NY): I congratulate the authors on their interesting paper and will diovasc Surg 48:801-806, 1964 14. Othersen HB Jr, Lorenzo RL: Diaphragmatic report the case of a patient who had successful repair paralysis eventration: newer approaches to diag- of a phrenic nerve transected inadvertently. A 20nosis and operative correction. J Pediatr Surg year-old man was stabbed in the left chest and had no palpable pulse or blood pressure on arrival in the 12:309-315, 1977 emergency room. A junior surgical resident performed an anterior left thoracotomy, opened the Discussion pericardium, plugged a hole in the right ventricle DR. JOE R . UTLEY (San Diego, CA): I congratulate Dr. with his finger, and then called for me. After repairShoemaker and his colleagues on such a fine presen- ing the ventricular hole, we noted that the left tation of a very important problem. When phrenic phrenic nerve was cleanly severed. The resident had nerve paralysis has occurred postoperatively in my performed a vertical pericardiotolmy. In the following days, the patient recovered uneexperience, i t has always been a relatively unexpected finding and I wonder why it should have oc- ventfully, but we noted persistent left lower lobe curred in that patient. Certainly surgeons take great collapse despite all attempts to reexpand that lobe. The patient was transferred to Montefiore Hospicare to avoid injury to the phrenic nerve during any thoracic procedure, but there must be times when tal, and on the twelfth day after his initial operation very subtle retraction on the nerve can result in major we reopened the thoracotomy incision. With the help disruption. One of the major contributions of this of our plastic surgeons, we reconstructed the phrenic paper is that it brings to our attention the necessity to nerve with 10-0 sutures and a microsurgical techalways think of phrenic nerve paralysis in any infant nique. The nerve edges had retracted and had to be whom we have difficulty weaning from a respirator freshened up; a free sural nerve graft was required to postoperatively. However, making an early diag- reconstitute the nerve. Follow-up serial inspiratory and expiratory roentnosis is sometimes a problem, and diagnosis by fluoroscopy in an infant off a respirator presents genograms demonstrated that between June, 1980, some obvious logistical problems. Would the authors which was six months after repair, and August, there
259
Shoemaker et al: Acquired Phrenic Nerve Paralysis in Children
was already some return of function. The left lower lobe was reexpanded. A cine of the diaphragm of this patient was made thirteen months after the nerve repair. The right hemidiaphragm, the uninvolved side, moved nicely. The left hemidiaphragm moved, but with some dyskinesia. It looked as though the two parts were not moving synchronously, but there was good inspiratory depression of the left hemidiaphragm. In view of this successful result, I suggest that surgical repair of the phrenic nerve be considered in appropriate patients after either accidental or deliberate transection. It is certainly more applicable in adult patients than in children. An intercostal nerve is readily available at the time of thoracotomy if a nerve graft is required, and is suitable for this purpose. To my knowledge, successful phrenic nerve repair has not been reported to date. DR . J . KENT TRINKLE (San Antonio, TX): I am delighted to see this paper on the program because it discusses a problem that occurs more frequently than is recognized. Also, I am sure that this problem is very seldom treated appropriately or early enough. A number of years ago we worked on this problem in the laboratory and devised what is probably the most simple of all plications-in fact, it is so simple, it’s virtually idiotproof. A stitch is put through the muscular portion of the diaphragm on either side of the central tendon going from medial to lateral. After a whole series of these stitches have been placed, they are tied. We tested the efficacy of this technique in dogs by sectioning a phrenic nerve and then flipping a coin to plicate or not plicate the diaphragm. We obtained preoperative and serial postoperative ventilatory functions in conscious dogs. This was no mean feat! We accomplished it by creating a tracheostomy, hooking the dogs up to a spirometer, and removing the calcium carbonate so that carbon dioxide was not absorbed. First we got a resting tidal volume and then, as the animal became hypercarbic, we actually obtained a maximum breathing capacity, forced vital capacity, timed vital capacity in one second, and so forth. I won’t bore you with all the data that we presented, just the resting tidal volume. The two groups of dogs were almost equal on the average of the control value, but there was a significant depression in the animals that did not undergo the plication. We subsequently used this method clinically on several occasions. One patient had acute lead poisoning. He had been shot through the phrenic nerve, and the lead was visible in the chest. The diaphragm was elevated, and we plicated it. A month later, it was in the normal position. The preoperative
and postoperative ventilatory functions were almost identical to those in the dogs. Therefore, I believe that this technique is applicable to not only infants and dogs, but perhaps to adults as well. DR. SHOEMAKER: I thank the discussants for their information and Dr. Trinkle for the presentation of some of his experimental studies. We would be somewhat bothered about whether or not function returns with the repair described by Dr. Trinkle. We obviously do not favor the pants-over-vest repair because of the placement of the incision and the fear that function does not return because of further injury to the branches of the nerve. As for the other two techniques we described, and maybe Dr. Trinkle’s technique, we think that as long as there is a good chance for return of function, then the technique is probably all right. Dr. Merav, we appreciate your presentation of an interesting case report. We did have 1 child who sustained transection of the nerve. It was repaired; however, we have not done fluoroscopy to know whether or not that nerve has regenerated. Fluoroscopy in all children was done in the intensive care unit using a portable image intensifier, which made things fairly simple when dealing with these sick infants. Dr. Utley commented on our apparent increased incidence of diaphragmatic paralysis in 1980. We have operated on a large number of newborn infants with complex congenital lesions in the past year or two. Most of those operations were palliative procedures, which are associated with a higher incidence of diaphragmatic paralysis. It also may be that we are more aware of the diagnosis and are picking up on all patients who do have phrenic nerve paralysis. Dr. Utley mentioned the advisability of ligating the ductus as opposed to dividing it. In no infant in our series did we divide the ductus intentionally. There was one instance in which the ductus tore, and division for vascular control was necessary. In all other infants the ductus was closed by a ligature or hemoclip. The important thing is that the closure of the patent ductus arteriosus was associated with other procedures in all these children. The majority had banding and ductus problems or they had coarctation repairs with ductus ligation. We were not dealing with isolated ductus problems. In adults we have seen instances of phrenic nerve paralysis with topical hypothermia. They had problems with persistent atelectasis in the postoperative period, although we have not seen any whose atelectasis posed major problems that necessitated prolonged hospitalization.
ABSTRACT Transthoracic diaphragmatic plication is a simple and effective means of treatment for phrenic nerve injury and resultant diaphragmatic paralysis, and should be undertaken as soon as the diagnosis of diaphragmatic paralysis is established in a child with substantial respiratory difficulty. Seven of more than 1,500 patients sustained phrenic nerve injury during operation for congenital heart disease at our institution over the past five years. All but 1 patient were less than 5 months old, and the average weight was 3.8 kg. Five of the 7 could not be weaned from the ventilator, and the other 2 had persistent postoperative tachypnea and stridor. Before 1980, 3 patients were intubated for an average of 16 days prior to diagnosis by fluoroscopy. Since then, 4 patients have been intubated for an average of 7.8 days before diagnosis. All 7 underwent transthoracic diaphragmatic plication and were extubated by 6 days after operation (mean, 3.7 days). Six patients survived and are doing well. The 1 death occurred in 1976 in a premature infant with multiple congenital cardiac lesions; diagnosis was delayed until the twenty-eighth day after intubation and transthoracic diaphragmatic plication, until the sixty-second day. This infant died of pneumonia one month after extubation. These data support our policy of establishing an early diagnosis of phrenic nerve injury and early treatment utilizing transthoracic diaphragmatic plication in symptomatic children.
Phrenic nerve injury and resultant diaphragmatic paralysis often cause life-threatening respiratory embarrassment in infants and young children [l-31. Prior to the advent of palliative or corrective operation for congeniFrom the Department of Surgery, Indiana University School of Medicine, Indianapolis, IN. Presented at the Seventeenth Annual Meeting of The Society of Thoracic Surgeons, Jan 26-28,1981, Los Angeles, CA. Address reprint requests to Dr. Brown, Department of Surgery, Room 212, Emerson Hall, Indiana University School of Medicine, 1100 W Michigan St, Indianapolis, IN 46223.
tal heart disease, most patients sustained phrenic nerve injury and diaphragmatic paralysis secondary to birth trauma [I, 2, 41. An increasing number of phrenic nerve injuries associated with thoracic surgical operations have shown up in recent studies [3, 5-10]. Traditional treatment, regardless of etiology, has been prolonged ventilatory support [l-111. With this method of nonoperative therapy, mortality in both groups of patients is 20 to 25% [l-111. One effective means of surgical treatment for this complication is diaphragmatic plication, and the mortality for diaphragmatic paralysis should be essentially zero if plication is done early [2, 5, 7,12, 131. Plication has been effective for those sustaining birth trauma [ 2 , 41, but only a small number of symptomatic patients have undergone plication after diaphragmatic paralysis secondary to thoracic surgery [5-9, 13, 141. This study describes 7 patients, all of whom were plicated for symptomatic diaphragmatic paralysis after operation for congenital heart disease.
Material and Methods From 1976 through 1980, more than 1,500 patients underwent palliative or corrective operation or both for congenital heart disease at the Riley Hospital for Children at the Indiana University Medical Center. Seven of them (5 boys and 2 girls) sustained phrenic nerve injury resulting in symptomatic diaphragmatic paralysis and underwent transthoracic diaphragmatic plication. Excluding 1 older child, each of these patients weighed less than 10 kg (average, 3.8 kg) and was less than 5 months old (average, 36 days). The oldest child was 3 years of age at the time of operation and had had two previous thoracotomies. Of the 7 patients, 5 could not be weaned from the ventilator, and the other 2 , after extubation, had persistent tachypnea, stridor, and difficulty feeding. The diagnosis of phrenic nerve paralysis was confirmed by
251 0003-4975/81/090251-09$01.25@ 1981 by The Society of Thoracic Surgeons
252 The Annals of Thoracic Surgery Vol 32 No 3 September 1981
Table 1 . Data on 7 Patients Undergoing Cardiac Surgical Procedures Resulting in Diaphragmatic Paralysis
Patient No.
Sex
Date of Operation
1
M
12/75
Age 8d
6d 5 wk 1d 4.5 mo
Congenital Weight Heart (kg) Disease 2.7
2 3 4 5 6
M F M M M
2/77 3/78 3/80 5/80 7/80
4 wk
3.7 4.1 3.4 4.8 4.1
7
F
7/80
3 yr
10.0
Procedure
PDA, coarctation
+
PDA pericardial window," coarctation repair PDA, VSD PDA, PA band PDA, PA band PDA, VSD Blalock-Taussig TOF TGV, PS, ASD Blalock-Hanlon TGV, VSD PDA + ASD, PA band coarctation, Coarctation reTA + PDA pair ASD, VSD, SIP Total repair PA band 2, PDA
Incision
Anticipated Damage to Phrenic Nerve
L thoracotomy
No
x 2
L L R R L
thoracotomy thoracotomy thoracotomy thoracotomy thoracotomy
Sternotomy
Yes No Yes No No
No
=Atrial perforation at heart catheterization. PDA = patent ductus arteriosus; VSD = ventricular septal defect; PA = pulmonary artery; TOF = tetralogy of Fallot; TGV = transposition of great vessels; PS = pulmonary stenosis; ASD = atrial septal defect; TA = tricuspid atresia; SIP = status post.
fluoroscopic demonstration of immobility or by paradoxical motion of the diaphragm to spontaneous respiration. The cardiac surgical procedures that resulted in diaphragmatic paralysis in these patients are listed in Table 1. All 7 children had multiple congenital cardiac defects, for a total of nineteen cardiac lesions. Five infants had ventricular septal defects, and 3 had atrial septal defects. An associated patent ductus arteriosus and coarctation affected 5 and 2 of the patients, respectively. At the time the phrenic nerve was injured, 5 of the 7 patients were undergoing multiple concomitant procedures. One patient underwent a Blalock-Taussig shunt for tetralogy of Fallot and 1, a Blalock-Hanlon procedure for transposition of the great arteries. Five of 6 infants undergoing palliative procedures had pericardiotomy, and 4 of these 6 had dissections performed both anterior and posterior to the phrenic nerve. Closure of an associated patent ductus in 4 infants was the most common operation. Pulmonary artery banding and ductus closure were done in 3 and were the most common combined procedures.
Prior to the diagnosis of diaphragmatic paralysis on the ipsilateral side, 4 patients had a left thoracotomy and 2 , a right thoracotomy. One patient experienced diaphragmatic paralysis following a median sternotomy. Although older, she had had a previous left thoracotomy for ductus ligation and pulmonary artery banding in infancy with a second thoracotomy for pulmonary artery band adjustment. A median sternotomy incision was performed for the correction. The pulmonary artery band had migrated distally, resulting in a difficult dissection in the region of the left pulmonary artery and left phrenic nerve. An outflow patch extending far out onto the left main pulmonary artery was required. In 2 patients, it was suspected that phrenic nerve injury had been done at the time of the cardiac operation. One was an obvious laceration during a Blalock-Taussig shunt, and at the time of the shunt procedure, a primary anastomosis of the nerve was performed. Postoperatively, the patient could not be weaned from the ventilator and demonstrated paradoxical motion of the diaphragm; plication was
253
Shoemaker et al: Acquired Phrenic Nerve Paralysis in Children
A third approach was used in 4 patients. done later. The other patient had no obvious injury to the phrenic nerve, but damage was Central plication (central pleats in the diaconsidered likely from stretch of the nerve phragm) were created as described by Schwartz during clamping of the aorta and the in- and Filler [7] (Fig 5). In this procedure, the sutrapericardial left pulmonary artery during tures are placed in the diaphragm so as to avoid injury to the branches of the phrenic nerve. ductus division. They are placed as partial-thickness sutures excluding only the peritoneum in order to avoid Operative Technique Plication to flatten the diaphragm in its inspi- injury to the abdominal organs. This is done by ratory position was performed through a post- grasping the diaphragm with a pair of vascular erior lateral thoracotomy incision in all pa- forceps between phrenic nerve branches, thus tients. Three different surgical techniques were tenting the diaphragm up away from the utilized. The classic "pants-over-vest" repair peritoneum as the needle is passed through it. was used in 1 patient (Fig 1). The abdominal Either of the latter two methods are preferable cavity was entered, but the repair was un- to the pants-over-vest technique. usually complicated by bowel prolapse through the diaphragmatic incision. Keeping the bowel Results safely retracted away from the closure also Our 7 patients fall into two groups: an early made the procedure difficult. group (operated on before 1980) and a late A circumferential incision through the dia- group (after 1980). The 3 patients in the early phragm near the chest wall avoided entry into group were intubated for an average of 16 days the peritoneal cavity and was the technique prior to the diagnosis of diaphragmatic pautilized in 2 patients (Fig 2). This approach was ralysis by fluoroscopy, whereas the 4 patients devised and used by one of us (H. K.) and, to in the late group were intubated for 7.8 days our knowledge, has not been described previ- prior to diagnosis. The largest difference beously. Interrupted sutures are placed in a radial tween the early and late groups was the number fashion from the thoracic wall diaphragmatic of days between diagnosis and plication-31.7 attachments so as to flatten the diaphragm and, days and 4.3 days, respectively. The average at the same time, avoid major phrenic nerve number of days spent in the intensive care unit reflected the pulmonary status of these patients. fibers (Figs 3, 4). The early group had an average of 59 days and the late group, an average of 15.5 days. The avFig 1 . Pants-over-vest technique: (A) In this repair, an erage number of days of mechanical ventilatory incision is made through the diaphragm and then closed support after plication was less than 4 in both with the outer edge sutured more centrally under the groups. medial portion. ( B ) The edge of the medial portion is Prior to 1980, 1 patient received a tracheosthen tacked laterally.
251 The Annals of Thoracic Surgery Vol 32 No 3 September 1981
Fig 2 . Circumferential pleat technique: ( A ) An incision is made circumferentially near the chest wall without incising the peritoneum. Sutures are then placed from the thoracic wall diaphragmatic attachments centrally to shorten the diaphragm. ( B ) Completed repair.
Fig 4 . (Patient 2.) After plication, the left hemidiaphragm is in its normal location.
bation. This infant died of pneumonia one month after discontinuing mechanical ventiFig 3 . (Patient 2 . ) Chest roentgenogram showing elevalator support. Also before 1980, plication was tion of the left hemidiaphragm following phrenic nerve delayed in 1 patient due to a wound infection paralysis. and in another, for congestive heart failure. To date, there have been no recurrences with tomy after 36 days of intubation, and another any of the plication techniques. Diaphragmatic patient died. The latter was a premature infant function has been assessed by fluoroscopy in 3 with multiple congenital cardiac lesions in of the 6 survivors. Function was demonstrated whom the diagnosis of diaphragmatic paralysis in all 3. One of these patients was studied sevwas delayed the longest (28 days) and in whom eral years following use of the circumferential plication was delayed until 59 days after intu- pleating technique (see Fig 2), and the other 2
255
Shoemaker et al: Acquired Phrenic Nerve Paralysis in Children
Fig 5. Central pleating technique: (A) Mattress sutures are placed, with care taken to avoid major nerve fibers and entry into the abdominal cavity. (B) Completed repair. (See text for details.)
were studied several months after use of the central pleating technique (see Fig 3). All 6 survivors are doing well with a mean follow-up of 21 months (range, 9 to 50 months) (Table 2). Comment Symptomatic unilateral diaphragmatic paralysis is an uncommon cause of respiratory difficulty in neonates and infants [l-4, 11-14]. However, recent reports indicate a rising number of patients with diaphragmatic paralysis secondary to thoracic and cardiac operations [5-lo]. In their large series of children undergoing operation for congenital heart disease, Mickell and associates [8] found the incidence of symptomatic 'diaphragmatic paralysis to be 1.7%. Our data indicate that about 0.5% of patients who undergo operation for palliation or correction of congenital heart disease will sustain symptomatic diaphragmatic paralysis. Some authors suggest that infants are more likely to become symptomatic from diaphragmatic paralysis than older children [6, 8, 91.
This point is well illustrated in the series of Mickell and co-workers [8]: 14 of 32 patients with documented unilateral diaphragmatic paralysis were less than 8 months old. Of these, 10 required prolonged ventilatory support for an average of 24 days with an average stay in the intensive care unit of 26 days. Five required tracheostomy, and 3 died. One patient underwent plication and is alive. In the group of patients older than 18 months, there was no mortality and no plications were necessary. Moreover, the number of days in the intensive care unit and on the ventilator were not different from those patients having an uneventful recovery from cardiac operation. Data from other authors corroborate this finding [6]. Our experience with infants seen prior to 1980, when plication was delayed, was similar to that of Mickell and colleagues [8]. One patient required tracheostomy, and another died after an average of 42 days of ventilatory support. Both required prolonged stays in the intensive care unit. Among the infants seen in 1980, there were no deaths and no tracheostomies were required. An average of 13 days was necessary for intubation, and an averge of 15.5 days in the intensive care unit. Although both of our groups had dia-
256 The Annals of Thoracic Surgery Vol 32
No 3 September 1981
Table 2 . Data on 7 Patients Undergoing Operation for Diaphragmatic Paralysis Intubation Patient No.
Days from Injury to DX a
Ventilatory Support
Days to Plication after DX
Days before DX
Days after Plication
Days in ICU
1
14 and 28"
Yes
29
28
5
82
2 3
21 11
No Yes
23" 43'1
9 11
0 6
34 60
4 5 6 7
1 10 4 14
Yes Yes No Yes
4 2 10 1
0"
3 2 2 4
8 18 13 23
10 7
14
Techn:ique/Results l/Died of bronchopneumonia 26 d after plication 2iGood; motion returned 2IGood; tracheostomy before plication; n o late fluol-oscopy 3/GOOd; no late fluoroscopy 3/Good; no late fluoroscopy 3/Good; motion returned 3/Good.; motion returned
aDiagnosis of diaphragmatic paralysis. I'Fatient had two operations before diagnosis. "Delayed due to wound infection. "Delayed due to congestive heart failure. "Obvious injury. DX = diagnosis; ICU
=
intensive care unit; 1 = pants-over-vest; 2 = circumferential pleat; 3
=
central pleating.
position of infants further reduces vital caphragmatic plication, we attribute the lower pacity [61. morbidity and mortality in the late group to aggressive diagnosis and treatment. Prior to 1980, 3. Pulmonary compliance is, decreased on the ipsilateral side [8] in that the diaphragm is patients were intubated an average of 16 days unable to resist the negative intrapleural prior to fluoroscopic diagnosis and waited 31.7 pressure and thus moves paradoxically with days before plication, whereas in 1980, they each respiration. were diagnosed at 7.8 days and the diaphragm was plicated 4.3 days later. These data indicate 4. Infants have relatively small intrabronchial calibers, a condition facilitating obstruction that plication reduces the number of days of and atelectasis by retained secretions [6]. ventilatory support. Other authors [3, 6-81 have discussed the Diaphragmatic paralysis should be suspected pathophysiology of diaphragmatic paralysis when the patient cannot be weaned from the and stressed that infants tolerate diaphragmatic ventilator or the patient's arterial blood gases paralysis less well than older children for four show retention of carbon dioxide despite apessential reasons. parent adequacy of respiratory efforts. Standard Infants are dependent on diaphragmatic chest roentgenogram made when the infant is contraction for adequate gas exchange and if on mechanical ventilation usually fails to show there is unilateral diaphragmatic paralysis, elevation of the hemidiaphragm, because posup to 50% of pulmonary function is lost. The itive airway pressure tends to keep the flail diaintercostal muscles of infants are weaker and phragm depressed. Opacification of the pleural at a mechanical disadvantage 181. space by atelectasis, effusion,. or both may also Infants have increased mediastinal mobility. make diagnosis by roentgencigraphy elusive. This means that in the presence of diaFluoroscopy is the diagnostic procedure of choice, and it can be performed at the bedside phragmatic paralysis, the mediastinal contents are shifted to the contralateral side on in most instances. Once the diagnosis of diainspiration. This decreases pulmonary func- phragmatic paralysis is suspected, the patient tion on the contralateral side as well as the should undergo fluoroscopy while breathipsilateral side [3]. The usual recumbent ing spontaneously. If diaphragmatic paralysis
257
Shoemaker et al: Acquired Phrenic Nerve Paralysis in Children
Table 3 . Mortality Following Acquired Phrenic Nerve Paralysis with Congenital Heart Disease Plication
Author
Year
City
Deathsa
Percent
Unplicated
Weldon [lo] Mickell et a1 [8] Greene et a1 [6] Othersen and Lorenzo [14] Schwartz a n d Filler [7] Mickell et a1 [81 Shoemaker et a1 [this report] Jewett and Thomson [131 Sethi and Reed [5] Shoemaker et a1 [this report]
1978 1978 1975 1977
St. Louis Pittsburgh Chicago Charleston
1 (4) 3h ( 1 1 ) C 1 (4) 0 (1)
25
1978
Boston
1“(2)
50
1978 1981
Pittsburgh Indianapolis
0 (1) 1 (3)
33
1964
Buffalo
I f (3)
33
1971 1981
Kansas City Indianapolis
lP (2)
50
0 (4)
0
Plicated lated
Plicated early
27 25
0
0
“Number in parentheses is total number of patients. bLate mortality in 1 child due to congenital heart disease. “Includes five tracheostomies. dMore than two weeks after diagnosis. ‘Cardiac death one month postoperatively. ‘Secondary to recurrence after plication; not treated. gSecondaryto aspiration pneumonia two weeks after plication.
is confirmed, surgical plication should be promptly carried out to reduce the morbidity and mortality associated with prolonged mechanical ventilation. Transthoracic plication can be performed easily by several techniques with minimal morbidity. Return of diaphragmatic function has been documented in 3 of our severely ill patients. If there is obvious phrenic nerve injury at the time of the initial operation in infants, concurrent plication using a technique that would allow later return of function should be considered. This would avoid the necessity of a later thoracotomy if the infant did not tolerate diaphragmatic paralysis. Plication does not preclude return of function, as shown by 3 of our patients who had documented return of function. Our data support previous studies suggesting age-dependent management of patients with diaphragmatic paralysis [6]. The 20% mortality with diaphragmatic paralysis due to birth trauma has now been demonstrated in most series of patients with unplicated diaphragmatic paralysis [l-4, 6, 111. Series of patients with acquired diaphragmatic paralyses, if unplicated, show a similar mortality [8, 101, as do results of late plications for diaphragmatic
paralysis secondary to operation for congenital heart disease [6]. Our early experience is similar. Recent experience with 4 patients, however, suggests that the number of days of mechanical ventilation, the length of stay in the intensive care unit, and mortality can be greatly reduced by aggressive diagnosis and treatment (Table 3). A retrospective analysis of 7 infants with diaphragmatic paralysis demonstrates a reduced morbidity and mortality for this complication when the diagnosis and treatment are approached aggressively. We recommend fluoroscopic evaluation of infants at risk for this complication who cannot be weaned from mechanical ventilation following operation for congenital heart disease. Prompt diaphragmatic plication in affected patients will reduce the hazards of prolonged mechnical ventilation and will not adversely affect return of phrenic nerve function. References 1. Richard J, Chevalier V, Capelle R, et al: La paralysie diaphragmatique obstetricale. Arch Fr Pediatr 14:563-598, 1957 2. Stauffer UG, Rickham PP: Acquired eventration
258 The Annals of Thoracic Surgery Vol 32 No 3 September 1981
of the diaphragm in the newborn. J Pediatr Surg comment about when they elect to use fluoroscopy, 716355640,1972 whether they do it in their intensive care unit, and 3. Wayne ER, Campbell JB, Burrington JD, Davis how they carry it out in a relatively sick infant. WS: Eventration of the diaphragm. J Pediatr Surg Certainly not every infant who has a phrenic nerve 9:643-651, 1974 paralysis requires diaphragmatic plication. In my 4. Bishop HC, Koop CE: Acquired eventration of experience it sometimes is an incidental finding obthe diaphragm in infancy. Pediatrics 22:1088- served relatively late after the operation. I think that 1096, 1958 infants with stiff lungs, pulmonary edema, high 5. Sethi G, Reed WA: Diaphragmatic malfunction pulmonary arterial pressure, or nutritional depletion in neonates and infants. J Thorac Cardiovasc are certainly at greater risk to become respirator deSurg 62:138-143, 1971 pendent following phrenic nerve dysfunction. I was 6. Greene W, L’Hereux P, Hunt CE: Paralysis of the interested that the authors found a greater incidence diaphragm. Am J Dis Child 129:1402-1405, 1975 of this problem in 1980 and wonder if they have a 7. Schwartz MZ, Filler RM: Plication of the dia- comment about that. In reading the manuscript, I noted that one inphragm for symptomatic phrenic nerve pastance of phrenic nerve dysfunction occurred folralysis. J Pediatr Surg 13:259-263, 1978 8. Mickell JJ, Oh KS, Siewers RD, et al: Clinical im- lowing ductus division. It has b’een our policy not to plications of postoperative unilateral phrenic perform ductus division in premature infants, innerve paralysis. J Thorac Cardiovasc Surg 76: fants, and younger children but to ligate the ductus. 297-304, 1978 Particularly in premature infants, we frequently use a 9. Mearns AJ: Iatrogenic injury to the phrenic nerve hemoclip to close the ductus. This can be done with in infants and young children. Br J Surg 64:558- minimal dissection and might perhaps decrease the risk of phrenic nerve problems with that procedure. 560, 1977 Topical hypothermia in the pericardium has been 10. Weldon CS: Discussion of Mickell JJ, et a1 [8] 11. France NE: Unilateral diaphragmatic paralysis indicated as a factor in the genesis of phrenic nerve and Erb’s palsy in the newborn. Arch Dis Child dysfunction postoperatively. I would be interested in the authors’ comments about the possible effect of 29:357-359, 1954 12. Symbas PN, Hatcher CR Jr, Waldo W: Dia- topical hypothermia in their patients, although I phragmatic eventration in infancy and child- noted from the manuscript that most of them underwent closed procedures. hood. Ann Thorac Surg 24:113-119, 1977 13. Jewett TC Jr, Thomson NB Jr: Iatrogenic eventration of the diaphragm in infancy. J Thorac Car- DR. AVRAHAM DOV MERAV (Bronx, NY): I congratulate the authors on their interesting paper and will diovasc Surg 48:801-806, 1964 14. Othersen HB Jr, Lorenzo RL: Diaphragmatic report the case of a patient who had successful repair paralysis eventration: newer approaches to diag- of a phrenic nerve transected inadvertently. A 20nosis and operative correction. J Pediatr Surg year-old man was stabbed in the left chest and had no palpable pulse or blood pressure on arrival in the 12:309-315, 1977 emergency room. A junior surgical resident performed an anterior left thoracotomy, opened the Discussion pericardium, plugged a hole in the right ventricle DR. JOE R . UTLEY (San Diego, CA): I congratulate Dr. with his finger, and then called for me. After repairShoemaker and his colleagues on such a fine presen- ing the ventricular hole, we noted that the left tation of a very important problem. When phrenic phrenic nerve was cleanly severed. The resident had nerve paralysis has occurred postoperatively in my performed a vertical pericardiotolmy. In the following days, the patient recovered uneexperience, i t has always been a relatively unexpected finding and I wonder why it should have oc- ventfully, but we noted persistent left lower lobe curred in that patient. Certainly surgeons take great collapse despite all attempts to reexpand that lobe. The patient was transferred to Montefiore Hospicare to avoid injury to the phrenic nerve during any thoracic procedure, but there must be times when tal, and on the twelfth day after his initial operation very subtle retraction on the nerve can result in major we reopened the thoracotomy incision. With the help disruption. One of the major contributions of this of our plastic surgeons, we reconstructed the phrenic paper is that it brings to our attention the necessity to nerve with 10-0 sutures and a microsurgical techalways think of phrenic nerve paralysis in any infant nique. The nerve edges had retracted and had to be whom we have difficulty weaning from a respirator freshened up; a free sural nerve graft was required to postoperatively. However, making an early diag- reconstitute the nerve. Follow-up serial inspiratory and expiratory roentnosis is sometimes a problem, and diagnosis by fluoroscopy in an infant off a respirator presents genograms demonstrated that between June, 1980, some obvious logistical problems. Would the authors which was six months after repair, and August, there
259
Shoemaker et al: Acquired Phrenic Nerve Paralysis in Children
was already some return of function. The left lower lobe was reexpanded. A cine of the diaphragm of this patient was made thirteen months after the nerve repair. The right hemidiaphragm, the uninvolved side, moved nicely. The left hemidiaphragm moved, but with some dyskinesia. It looked as though the two parts were not moving synchronously, but there was good inspiratory depression of the left hemidiaphragm. In view of this successful result, I suggest that surgical repair of the phrenic nerve be considered in appropriate patients after either accidental or deliberate transection. It is certainly more applicable in adult patients than in children. An intercostal nerve is readily available at the time of thoracotomy if a nerve graft is required, and is suitable for this purpose. To my knowledge, successful phrenic nerve repair has not been reported to date. DR . J . KENT TRINKLE (San Antonio, TX): I am delighted to see this paper on the program because it discusses a problem that occurs more frequently than is recognized. Also, I am sure that this problem is very seldom treated appropriately or early enough. A number of years ago we worked on this problem in the laboratory and devised what is probably the most simple of all plications-in fact, it is so simple, it’s virtually idiotproof. A stitch is put through the muscular portion of the diaphragm on either side of the central tendon going from medial to lateral. After a whole series of these stitches have been placed, they are tied. We tested the efficacy of this technique in dogs by sectioning a phrenic nerve and then flipping a coin to plicate or not plicate the diaphragm. We obtained preoperative and serial postoperative ventilatory functions in conscious dogs. This was no mean feat! We accomplished it by creating a tracheostomy, hooking the dogs up to a spirometer, and removing the calcium carbonate so that carbon dioxide was not absorbed. First we got a resting tidal volume and then, as the animal became hypercarbic, we actually obtained a maximum breathing capacity, forced vital capacity, timed vital capacity in one second, and so forth. I won’t bore you with all the data that we presented, just the resting tidal volume. The two groups of dogs were almost equal on the average of the control value, but there was a significant depression in the animals that did not undergo the plication. We subsequently used this method clinically on several occasions. One patient had acute lead poisoning. He had been shot through the phrenic nerve, and the lead was visible in the chest. The diaphragm was elevated, and we plicated it. A month later, it was in the normal position. The preoperative
and postoperative ventilatory functions were almost identical to those in the dogs. Therefore, I believe that this technique is applicable to not only infants and dogs, but perhaps to adults as well. DR. SHOEMAKER: I thank the discussants for their information and Dr. Trinkle for the presentation of some of his experimental studies. We would be somewhat bothered about whether or not function returns with the repair described by Dr. Trinkle. We obviously do not favor the pants-over-vest repair because of the placement of the incision and the fear that function does not return because of further injury to the branches of the nerve. As for the other two techniques we described, and maybe Dr. Trinkle’s technique, we think that as long as there is a good chance for return of function, then the technique is probably all right. Dr. Merav, we appreciate your presentation of an interesting case report. We did have 1 child who sustained transection of the nerve. It was repaired; however, we have not done fluoroscopy to know whether or not that nerve has regenerated. Fluoroscopy in all children was done in the intensive care unit using a portable image intensifier, which made things fairly simple when dealing with these sick infants. Dr. Utley commented on our apparent increased incidence of diaphragmatic paralysis in 1980. We have operated on a large number of newborn infants with complex congenital lesions in the past year or two. Most of those operations were palliative procedures, which are associated with a higher incidence of diaphragmatic paralysis. It also may be that we are more aware of the diagnosis and are picking up on all patients who do have phrenic nerve paralysis. Dr. Utley mentioned the advisability of ligating the ductus as opposed to dividing it. In no infant in our series did we divide the ductus intentionally. There was one instance in which the ductus tore, and division for vascular control was necessary. In all other infants the ductus was closed by a ligature or hemoclip. The important thing is that the closure of the patent ductus arteriosus was associated with other procedures in all these children. The majority had banding and ductus problems or they had coarctation repairs with ductus ligation. We were not dealing with isolated ductus problems. In adults we have seen instances of phrenic nerve paralysis with topical hypothermia. They had problems with persistent atelectasis in the postoperative period, although we have not seen any whose atelectasis posed major problems that necessitated prolonged hospitalization.