Although various complications in the respiratory system after pediatric liver transplantation (LT)1 are observed, paralysis in the phrenic nerve after cadaveric LT has been reported to be high.2,3 On the other hand, this complication has rarely been reported after living-donor liver transplantation (LDLT).4 This report describes an infant who experienced repeated respiratory failure with paralysis in the left phrenic nerve after LDLT; the infant had biliary atresia (BA) with situs inversus (SI) and an absent inferior vena cava (IVC).
A 7-month-old boy with biliary atresia accompanied by situs inversus and absent inferior vena cava (IVC) underwent living-donor liver transplantation (LDLT). Because a constriction in the recipient hepatic vein (HV) was detected during the preparation of the HV in LDLT, a dissection in the cranial direction and a total clamp of the suprahepatic IVC was performed, and the suprahepatic IVC and the graft HV were anastomosed end-to-end. Postoperatively, atelectasis in the left upper lobe and ventilator failure accompanied by an elevation of the left hemidiaphragm were observed and mechanical ventilation was repetitively required. Paralysis in the left phrenic nerve was diagnosed by chest radiograph and ultrasonography. In our patient, conservative treatment was administrated, because weaning him from mechanical ventilation was possible a few days after intubation and the ventilator function was expected to be improved with growth. The disease course was good, and he was discharged from the hospital at 78 days after LDLT. Complications of paralysis in the phrenic nerve after cadaveric liver transplantation have been reported to be high. Although using a conventional technique during the reconstruction of the HV may injure the phrenic nerve directly, use of the piggyback technique with preservation of the IVC is rare. Even if LDLT was undertaken, a dissection of the HV or a total clamp of the suprahepatic IVC as a conventional technique can directly injure the phrenic nerve. Therefore, a dissection of the HV or a total clamp of the suprahepatic IVC at the reconstruction of the HV in LDLT should be carefully performed, and the possibility of paralysis in the phrenic nerve should be considered in patients with a relapse of respiratory symptoms and an elevation of the hemidiaphragm after LDLT. Liver Transpl 14:1659–1663, 2008. © 2008 AASLD.
The patient, a 7-month-old boy with a history of BA, had undergone a Kasai operation at 93 days after birth, at which time he was noted to have SI with an absent IVC. However, because there was no improvement of cholestatic jaundice, ascites, and malnutrition, he was admitted to the hospital to undergo LDLT. On admission, he was conscious, and abdominal distention was observed. His liver was easily palpated 5 finger-widths below the left hypochondrium, and the spleen was palpated 4 finger-widths below the right hypochondrium. No cyanosis was observed, but paroxysmal stridor was occasionally heard. A hematological examination demonstrated hyperbilirubinemia, hypoalbuminemia, and abnormal coagulation (Table 1), and the Pediatric End-Stage Liver Disease score was 13.4. A chest radiograph, abdominal computed tomography, and echocardiography revealed SI with an absent IVC, but neither polysplenia and asplenia nor cardiac malformation was detected. Mild consolidation was observed on the dorsal side of the bilateral lungs, but no atelectasis or elevation of the hemidiaphragm was detected (Fig. 1). Although we suspected intrapulmonary shunting by his respiratory conditions and his need for oxygen, a lung perfusion scintigraphy demonstrated a shunt rate of 3.1%–3.4% and no intrapulmonary shunting was observed.
|WBC||6,600/μL||T-Bil||16.77 mg/dL||CRP||2.44 mg/dL|
|RBC||354 × 104/ μL||D-Bil||11.58 mg/dL||T-Cho||151 mg/dL|
|Hb||10.8 g/dL||BA||124.0 μmol/L||FBS||64 mg/dL|
|Ht||32.4 %||TP||6.7 g/dL||PT %||81.8 %|
|PLT||15.7 × 104/ μL||ALB||2.8 g/dL||HPT||44.0 %|
|AST||166 IU/L||Na||138 mEq/L||APTT||48.7 seconds|
|ALT||65 IU/L||K||4.6 mEq/L||AT III||59.7 %|
|LDH||348 IU/L||Cl||104 mEq/L||Fe||59 μ g/dL|
|ALP||1849 IU/L||BUN||10 mg/dL||UIBC||202 μg/dL|
|γ-GTP||811 IU/L||CRE||0.14 mg/dL||Ferritin||62.2 μg/dL|
|CHE||111 IU/L||NH3||69 μg/dL|
In April 2005, LDLT for progressive hepatic failure was performed using a lateral segment graft from his father, who does not have SI. The procedure took 20 hours and 43 minutes. The amount of bleeding was 828 mL, and the graft liver/standard liver volume ratio was 116%. The anatomy of the organs in the abdominal cavity were a complete mirror-image (Fig. 2), and the infrahepatic IVC was absent. The recipient hepatic vein (HV) flowed into the right atrium (RA) via the suprahepatic IVC. However, because the HV was thought to be insufficient for anastomosis because it narrowed to 10 mm at the diaphragm, the suprahepatic IVC was dissected up to the RA. The suprahepatic IVC was exposed by blunt dissection from the surrounding thick connective tissues, and separated from the diaphragm. A total clamp of the suprahepatic IVC using Satinsky forceps was performed (Fig. 3), and the suprahepatic IVC was anastomosed using an oblique incision and the graft HV (diameter, 25 mm) in end-to-end fashion. The anatomy of the portal vein (PV) was a complete mirror-image, but the major trunk had atrophied to a diameter of 3.0 mm and was sclerotic. Therefore, the PV was separated at the confluent site of the splenic vein (SV) and the superior mesenteric vein, and part of the donor SV was interposed as a vascular graft. The vascular graft was harvested from the distal SV at the confluence of the SV and the inferior mesenteric vein (diameter, 10 mm; length, 30 mm). Although the donor did not undergo a splenectomy, the postoperative course was uneventful. The hepatic artery anatomy was a complete mirror-image, and a general end-to-end anastomosis could be performed under a microscope.
The graft function was excellent on the postoperative course, but atelectasis in the left upper lobe, which was not detected before LDLT, was detected on postoperative day (POD) 2, and extubation could not be performed until POD 11. After extubation, tachypnea and tachycardia remained, and an elevation of the left hemidiaphragm was detected on POD 12 (Fig. 4A). On POD 24, his respiratory conditions were aggravated by a fever, and because severe ventilator failure (pH 7.081, PaCO2 96.9 mm Hg, PaO2 67.1 mm Hg, base excess [BE] −3.8) occurred, mechanical ventilation to maintain continuous positive airway pressure under intubation was resumed on POD 25. After intubation, the elevation of the left hemidiaphragm was improved and his respiratory conditions were stable. Extubation was performed on POD 29, but atelectasis in the left upper lobe and elevation of the left hemidiaphragm were progressive, and tachypnea and tachycardia persisted. On POD 46, because his respiratory conditions were aggravated by vomiting, mechanical ventilation under intubation was again performed. The elevation of the left hemidiaphragm improved during mechanical ventilation of continuous positive airway pressure but, as shown on chest radiographs, it remained during spontaneous respiration, thus suggesting paralysis of the left phrenic nerve. Furthermore, because hypokinesia of the left hemidiaphragm was demonstrated by ultrasonography during spontaneous respiration on POD 53, paralysis in the left phrenic nerve was diagnosed. On POD 54, extubation was performed, but the elevation of the left hemidiaphragm remained (Fig. 4B). Although surgical plication of the diaphragm was considered, conservative treatment was administrated because his respiratory conditions were stable by administration of oxygen via a transnasal cannula and his ventilator function was expected to improve with growth. Although home oxygen therapy was required, he was discharged from the hospital on POD 78 because of his good general condition and graft function. Six months after LDLT, the elevation of the left hemidiaphragm remains, but his respiratory conditions are stable.
SI is a rare anomaly, with a frequency reported to be between 0.002% and 0.1%,5 and, interestingly, its association with BA occurs in up to 28% of children with SI.6 Conventionally, LT is absolutely contraindicated in patients demonstrating SI because it is often accompanied by vascular abnormalities,7 Mattei et al.8 performed LT in 26 patients with SI, and reported various vascular abnormalities detected during LT, including an absent infrahepatic IVC (50%), a preduodenal PV (69%), and an aberrant hepatic artery distribution (35%). Kamei et al.9 reported that LT could be applied even to patients with these vascular abnormalities. In an absent infrahepatic IVC, reconstruction of the HV is not considered difficult and a general end-to-end anastomosis could be performed. In a preduodenal PV, a vascular graft is sometimes required, depending on the length or tension in cases requiring an end-to-end anastomosis, but generally, the anastomosis in front of the duodenum can be performed without tension.
In our patient, where there was an absent infrahepatic IVC, the suprahepatic IVC was dissected up to the RA because of a constriction of the HV, and the graft HV was anastomosed with the suprahepatic IVC. The postoperative blood flow was good, without any complications. The PV anatomy was a complete mirror-image, but the major trunk was atrophied to a diameter of 3.0 mm and was hardened. Therefore, a part of the donor SV was interposed as a vascular graft. With the recent improvement in imaging and surgical techniques, reports on successful LT even in patients with SI, as in our patient, have been increasing.9–11 To perform LT in patients with SI, preoperative evaluation of the vascular system by imaging examination and refinement of the anastomotic methods, including the use of a vascular graft and the site of the vascular clamp, are important.
Generally, paralysis in the phrenic nerve after celiotomy is considered to occur on the right side. Immediately in front of the diaphragm, the phrenic nerve is aligned along the IVC on the right side and along the left boundary of the heart on the left side. Therefore, a dissection of the HV toward the deep site in the cranial direction and a total clamp of the suprahepatic IVC could damage the right phrenic nerve. In our patient, the damage to the left phrenic nerve may have occurred anatomically because of SI, leading to paralysis in the left phrenic nerve. Although we thought that the elevation of the hemidiaphragm could have been due to atelectasis, it was actually caused by paralysis in the phrenic nerve; which delayed the diagnosis of paralysis in the phrenic nerve. The differential diagnosis of an elevation of the hemidiaphragm includes ipsilateral atelectasis, lobar collapse, and diaphragmatic hernia.12, 13 In cases of cadaveric LT using a conventional technique, paralysis in the phrenic nerve is thought to be a relatively frequent complication. Starzl2 first reported paralysis in the phrenic nerve as a complication of, and McAlister et al.3 indicated that the incidence was 79% in LT. Although a conventional technique may possibly injury the phrenic nerve directly, use of the piggyback technique with preservation of the IVC is rare. At the moment, a diagnostic criteria of paralysis in the phrenic nerve is uncertain, but the following criteria are reported: (1) an elevation of the hemidiaphragm detected by chest radiograph; (2) a paradoxical diaphragmatic movement in response to the sniff test; and (3) an electrophysiological evaluation of the diaphragm.14 In our patient, because an elevation of the hemidiaphragm, which was not detected before LT, was detected by chest radiograph after LT and a paradoxical diaphragmatic movement was detected by ultrasonography, not fluoroscopy, during spontaneous respiration, paralysis in the phrenic nerve was diagnosed. An electrophysiological evaluation of the diaphragm was not done for our patient, but diaphragm function can be obtained by magnetic phrenic nerve stimulation.15 Rafferty et al.16 reported that magnetic phrenic nerve stimulation could be effective in assessing diaphragm function in children following LT. Magnetic phrenic nerve stimulation is thought to be effective not only in diagnosis but also in follow up of paralysis. In the future, magnetic phrenic nerve stimulation will be examined as a diagnostic criteria of paralysis in the phrenic nerve. McAlister et al.3 reported diaphragmatic function that spontaneously recovered completely by 9 months after surgery, but ventilator hypofunction caused by paralysis in the phrenic nerve sometimes leads to severe respiratory failure in children, in particular in infants. Hasegawa et al.4 reported a patient in whom an elevation of the hemidiaphragm, which had occurred 3 weeks after LDLT, was alleviated by surgical plication of the diaphragm. When weaning from mechanical ventilation was impossible, surgical plication has been indicated.17 Although early surgical plication of the diaphragm in severe cases has been effective,18 conservative treatment was administrated in our patient because, due to his respiratory conditions, weaning from mechanical ventilation was possible and his ventilator function was expected to improve with growth. If respiratory conditions are stable, as in our patient, conservative treatment for paralysis in the phrenic nerve can be performed. Then, because long-term results of the phrenic nerve injury may result in diaphragmatic eventration in the future,12, 13 we should recognize this as a possible complication and surgical treatment should be administrated immediately.
In LDLT using the common piggyback technique, a complication of paralysis in the phrenic nerve is rare because of preservation of the IVC. However, when a dissection of the HV or a total clamp of the suprahepatic IVC is required during LDLT, it should be recognized that paralysis of the phrenic nerve is an occasional complication due to direct injury to the phrenic nerve. In a dissection of the HV at LDLT, it is important to sufficiently dissect the HV in the cranial direction with careful observation of the phrenic nerve. In a total clamp of the suprahepatic IVC at LDLT, it is important not to clamp the diaphragm together with the IVC. Furthermore, after LDLT using conventional techniques, clinical care should consider the possibility of respiratory failure caused by paralysis in the phrenic nerve; early diagnosis and appropriate treatment must be performed if paralysis in the phrenic nerve is suspected.