Biliary atresia (BA) is a rare disease with an occurrence rate of 1 in 8000 to 18,000 live births (1). The clinical symptoms of BA characteristically include jaundice and acholic stools at 1 or 2 months after birth. Although extremely rare, the presentation of a bleeding disorder as the first symptom of BA has also been reported (2). Intracranial hemorrhage (ICH) is one of the complications found in patients with BA caused by late-type vitamin K–deficiency bleeding (VKDB), which occurs most commonly at the age of 1 to 2 months (3,4). This complication causes not only mortality but also significant long-term morbidity in the survival BA before and even after liver transplantation (2,3,5,6).
Vitamin K is required for the synthesis of coagulation factors II, VII, IX, and X by the liver (7); however, the vitamin K level in the newborns is usually low because of the insufficient vitamin K stores of the newborn and low placental transfer of vitamin K (8,9). In Japan, prophylactic oral administration of vitamin K to newborns at birth, on the sixth day after birth, and 1 month after birth beginning in 1981, has markedly reduced the incidence of idiopathic vitamin K deficiency (VKD) (5). Unfortunately, this prophylaxis has no effect on secondary VKD caused by malabsorption of vitamin K caused by cholestatic disorders such as BA (6).
Although BA had been reported to be one of the major causes of secondary VKD, there are only a few reports on patients with BA presenting with ICH, especially regarding their long-term outcome. In the present study, we describe 7 patients with BA presenting with ICH. We investigated the incidence of ICH in patients with BA who previously received prophylactic oral administration of vitamin K in the neonatal period. Moreover, we also describe the management of BA-associated VKDB complicated by ICH and also focus on the long-term outcome of these patients.
Between 1979 and 2009, 88 consecutive infants with BA underwent surgery for BA and were followed up in the Department of Pediatric Surgery of Kyushu University Hospital. Infants who presented with ICH were enrolled in the present study. Their clinical records and imaging studies were retrospectively reviewed. ICH caused by VKDB was diagnosed based on clinical and neurologic signs and symptoms, hematologic examination, and findings on computed tomography (CT) scans of the head. A confirmed case of VKDB should fulfill the diagnostic criteria of at least 2 of the following: hepaplastin level <10% and/or thrombo test <10%; prothrombin time (PT) percentage <10%; activated partial thromboplastin time (APTT) >120 seconds; protein induced by vitamin K absence (PIVKA)-II level exceeding normal controls; improvement of bleeding tendency and PT after 24 hours of vitamin K administration, and by a normal or raised platelet count (7,10).
BA was diagnosed based on clinical presentation including jaundice and acholic stools, hematologic examination, and ultrasonographic findings, and was confirmed by surgical cholangiography. BA types are classified using anatomical classification by Morio Kasai (11).
Long-term outcomes focusing on neurologic sequelae, such as developmental delay, mental retardation, epilepsy, and hemiparesis, were observed until the end of the follow-up period. Developmental delay was defined as a subset of developmental disabilities in 1 or more developmental domains: gross/fine motor, speech/language, cognition, social/personal, and activities of daily living, on age-appropriate, standardized norm-reference testing. Mental retardation was defined as a condition characterized by intellectual ability that is significantly below average (specifically intelligence quotient [IQ] ≤70) combined with deficits in adaptive abilities. Developmental delay assessment or developmental quotient if available was done by a pediatrician in our hospital, whereas IQ and mental retardation assessment was made by a psychologist/psychiatrist. The development was evaluated in 3 aspects: motor, social, and speech development.
Of the 88 consecutive infants with BA, 7 infants (all girls) presented with ICH (7.95% of patients with BA). The age of onset of ICH ranged from 47 to 76 days (mean 62.0 days), with vomiting, consciousness disorders, seizures, dyspnea, anisocoria, and conjugate deviation as the major symptoms at presentation (Table 1).
All of the patients were born full-term and received oral prophylactic vitamin K during the neonatal period. Neither perinatal complications nor any history of head trauma were found for these infants. During the first 6 months of life, 4 patients received exclusive breast-feeding (BF), 1 patient formula milk (FM) containing vitamin K supplementation, and 1 patient mixed BF and FM (Table 1). Although all of the patients were screened at 1 month of age by an obstetrician, VKDB still occurred before the diagnosis of BA was established. After stabilization of their ICH, 5 patients underwent the Kasai operation and 2 patients underwent hepaticojejunostomy for definitive diagnosis of BA. The timing of the Kasai procedure was 9 to 37 days (mean 22.3) after the onset of ICH (Table 1). The types of BA found during surgery were III-b1-ν in 3 patients, III-b2-o in 1 patient, III-a1-ν in 1 patient, and I-cyst in 2 patients.
CT scans of the head demonstrated that intraparenchymal hemorrhage was found in 4 cases, subarachnoid hemorrhage in 2 cases, and subdural hemorrhage (SDH) in 3 cases. A midline shift (MS) caused by massive hemorrhage was found in 5 cases; however, only 2 of these patients underwent urgent surgical evacuation of an intracranial hematoma before laparatomy because of anisocoria. Consistent with CT evaluation upon admission, magnetic resonance imaging/magnetic resonance angiography (MRI/MRA) during hospitalization (first–fifth weeks after admission) showed a more detailed description of ICH and its complication to the brain such as the development of encephalomalacia in middle cerebral artery territory in 5 patients, cerebral atrophy in 2 patients, cerebral ventricle enlargement in 2 patients, and atrophic of corpus callosum in 1 patient. During follow-up, 2 patients underwent living-related donor liver transplantation (LDLT) and 3 patients died at the ages of 33, 36, and 278 months of liver failure because liver transplantation could not be performed (Table 2).
Upon admission, routine laboratory examination showed platelet levels within the normal range (range 20.8–61.1 × 104 cells/μL), whereas severe anemia was found in 2 patients (range 4.1–4.6 g/dL). Elevated direct bilirubin was found in all of the patients, and elevated alanine aminotransferase and aspartate aminotransferase were found in 6 of 7 cases (Table 3).
Significant prolonged PT and APTT were found in coagulation examination during admission in the patients for whom data were available, with a range of 36.2 to >200 seconds and 67.3 to >200 seconds, respectively. Right after the diagnosis of VKDB was made, intravenous (IV) vitamin K was administered to all 7 patients and fresh frozen plasma (FFP) was given to 5 patients. Follow-up examinations showed improvement in PT and APTT levels from 11.7 to 18.5 seconds and 26.4 to 40.3 seconds, respectively. An elevated PIVKA-II level was also noted in 4 cases (Table 3).
At follow-up examination at 22 to 24 months, developmental delay in all aspects of evaluation (motor, social, speech) was seen in 4 patients, cognitive impairment in 1 patient, and normal development in 2 patients. In 1 patient who had normal developmental evaluation, a mild hemiplegia of the right extremities was observed with upper extremity stronger than the lower 1; however, overall evaluation was normal development. At this point of age, an electroencephalography (EEG) examination was also performed in 4 patients, 3 of whom showed abnormal EEG. One of 3 patients who did not have an EEG evaluation at this time point (case 3) showed clinical epilepsy during follow-up. At this time point, evaluation of the liver function showed elevation of aspartate aminotransferase, alanine aminotransferase, and direct bilirubin in 5 patients (Table 4).
During the 22 to 278 months (mean 90.3) of follow-up, some neurologic sequelae persisted in 5 of the 7 cases. The types of neurologic sequelae found were mental retardation in 2 patients, epilepsy in 1 patient, hemiparesis in 2 patients, and developmental delay in 1 patient who showed bilateral cerebral atrophy. In 2 patients, no neurologic deficits were observed until the end of the follow-up. Follow-up CT scans of the head demonstrated significant ischemic changes shown as a low-density area (LDA) consistent with encephalomalacia in the left hemisphere were found in 5 patients (Table 5). In the present study, a comparison of head CT scans at admission and latest CT at follow-up showed that in case 6, the head CT showed SDH and MS at admission. During follow-up, head CT did not show any new hemorrhage; however, LDA was found in the left hemisphere. In addition, neurologic sequelae such as right hemiparesis were observed. In contrast, head CT in case 7 also showed SDH and MS at admission; however, during follow-up, head CT showed a complete resolution and no LDA was found. Clinically, case 7 also did not show any neurologic sequelae (Fig. 1).
Several recommendations for vitamin K administration had been proposed for all of the newborn infants (12); however, the prophylactic administration of vitamin K is not sufficient to prevent bleeding disorders in infants with BA (4,6). As described in our study, all of our cases received 2 mg prophylactic oral administration of vitamin K during the neonatal period; nevertheless, ICH still occurred in 7.95% of our patients with BA. The prophylactic failure is probably the result of their increased bleeding tendency caused by secondary late-type VKD associated with cholestasis-induced malabsorption of vitamin K in the digestive tract, as mentioned in previous studies (2,6). This is supported also by a study by Van Hasselt et al (13), which described the risk of VKDB in breast-fed infants with BA receiving 25 μg oral vitamin K from the second week until the end of the 13th week was 8 to 10 times higher than in breast-fed infants receiving either a higher weekly oral prophylaxis dose or intramuscular (IM) prophylaxis at birth. More than 80% of the infants in that study developed VKDB by the time that cholestasis was diagnosed, and 43% presented with an ICH. A comparison study of VKD and VKDB risk in cholestatic jaundice infants who received regular infant formula and hypoallergenic formula by Van Hasselt et al (14) also reported that an increased risk is predominant in cholestasis infants receiving (whey-based) hydrolyzed formula. In the present study, we also describe that of 7 patients, 4 patients received BF only, whereas 1 patient received only regular FM and 1 patient received both BF and FM; however, in the present study we could not overestimate that BF is related to the development of ICH because of a lack of data about infant nutrition from the rest of the patients with BA who did not have ICH complication. Moreover, regarding the administration route of vitamin K, Komatsu et al also described a case of ICH in 2 infants with the late type of VKD despite 3 administrations of 2 mg oral vitamin K after birth, at 5 days old and 1 month old (15). Although the underlying diseases of these patients until hospital discharge remain undetermined, it is suggested that prophylactic oral vitamin K failed to prevent ICH in late-type VKD.
In the present study, all of the patients fulfilled the criteria of VKDB. Elevation of the PT at admission (range 36.2 to >200 seconds) was found in all of the patients with available data and was rapidly improved after administration of IV vitamin K and FFP (range 11.7–18.5 seconds). One patient (case 1) did not have exact PT data upon admission (referral information from the previous hospital was not complete); only PT and APTT were written as elevated and the patient had been administered with 2 mg vitamin K intravenously; however, hepaplastin test upon admission was <10% and PT after 24 hours’ administration of vitamin K returned to the normal level, combined with improvement of bleeding tendency. Therefore, we concluded that case 1 fulfilled the VKDB criteria, and counted as ICH-associated VKDB. In addition, an elevated PIVKA-II level was also found in all 3 patients whose PIVKA-II levels were measured and 1 patient just written elevation of PIVKA-II in the referral letter. Grossly abnormal PT with a rapid normalization of coagulation test and bleeding tendency after vitamin K administration combined with an elevation of PIVKA-II level, despite a normal platelet count, were characteristic findings of laboratory examinations in patients with VKD (7,10,16,17). The severity of the coagulation disorder was not inversely proportional to the value of hemoglobin. Nevertheless, the severity of anemia was probably related to the symptoms at admission, as shown in our 2 patients with severe anemia, both of whom had dyspnea as a presenting symptom. Nevertheless, dyspnea could also occur as a cause of brain swelling or compression.
In the present study, intraparenchymal hemorrhage with MS was the most common lesion found in CT images. These findings are different from other reports, which found SDH and subarachnoid hemorrhage as the most common types of ICH (8,16). In addition, in 2 of 7 of our cases, >1 lesion was found on CT. This finding is similar to Majeed et al (8), who reported that most of the children presenting with VDKB had hemorrhage in multiple sites, and pallor was seen in all of the infants (100%). Recently, magnetic resonance (MR) imaging was widely used in clinical practice for neonates with brain injury because of its high sensitivity for depicting a developing brain. MR can provide greatly detailed images of brain structures without exposing infants to ionizing radiation (18). CT is frequently used in workup in children with sudden onset of neurologic symptoms, especially to rule out acute hemorrhage. In addition, MR demands particular care with regard to patient transport, monitoring of vital signs, and optimization of acquisition techniques (use of appropriate coils, sequences, and protocols) (18–20). In our institution, we use head CT as a diagnostic tool for emergency brain injury, especially because CT was faster as compared with MR (<5 minutes), did not require any sedation, and allows close monitoring of vital signs during evaluation of unstable neonates after ICH. During hospitalization, except for 2 patients who were born before 1995, MR imaging and MR angiography were done after patients’ conditions were stabilized, to obtain an accurate evaluation, which revealed a more clear description of late brain complications as early as the first week after ICH (Table 2).
In spite of the wide variety of lesions in our cases, craniotomy was done only in 2 cases, without correlation to the lesion(s) itself, and was done before laparatomy. The indications of craniotomy of our patients were increased intracranial pressure with worsening of consciousness and the presence of anisocoria suggesting severe brain compression caused by large hemorrhage. The other cases were treated supportively or were treated for increased intracranial pressure or cerebral edema, if suspected. With or without surgical intervention for hematomas, a laparatomy was performed in all of our patients on an average of 22.3 days after the onset of evacuation of ICH and 9 days at the earliest. This was based on the fact that the Kasai portoenterostomy has become the primary surgical treatment for uncorrectable BA (21–23). In patients with BA with ICH as a presenting symptom, early Kasai portoenterostomy is also recommended. In the patients presenting with ICH and in those who have undergone craniotomy, early Kasai portoenterostomy is possible only when both hemorrhage and the bleeding tendency are well controlled. IV administration of vitamin K at 0.5 to 1.0 mg/kg, with or without addition of FFP given concurrently for more rapid restoration, is effective in patients with VKDB (16,24). Vitamin K administration had been reported to improve bleeding tendency within 1 hour and normalizes the coagulation within several hours, to facilitate even an urgent craniotomy (6). In our patients with or without craniotomy, no difficulties in hemostasis or recurrent hemorrhage were noted after initial administration of vitamin K IV and FFP. These findings suggest that surgical intervention for ICH has no adverse effects on the subsequent surgical management of BA if coagulation disorders are managed properly.
Studies on the long-term follow-up reported that some neurologic sequelae, such as developmental delays, mental retardation, and epilepsy, were observed in more than half of the patients with ICH caused by VKD (25,26). In the present study, developmental delay in all of the evaluation aspects showed in 4 patients at 2 years of age, which in 3 patients remains until the end of study follow-up. At further follow-up until the end of the present study (at 22–278 months), none of the patients died of ICH. Causes of death of our patients were end-stage liver diseases in those unable to undergo liver transplantation, in which the progressivity of the disease already could be seen at 2-year follow-up (Table 4). Follow-up imaging studies obtained using head CT scans showed an LDA consistent with encephalomalacia in the left hemisphere in 5 cases. Until the end of our observation, some neurologic sequelae persisted in 5 of 7 cases. The type of neurologic sequelae found were mental retardation in 2 cases, epilepsy in 1 case, hemiparesis in 2 cases, and developmental disorder in 1 case, with no correlation to the substantial ischemic brain damage demonstrated on CT. Mental retardation in case 1 was determined by an IQ measurement of 47 combined with a marked decreased ability to interact with individuals at 20 years of age without mental retardation. These findings are in agreement with the latest head CT evaluation, which showed a marked cerebral atrophy, cerebral ventricle enlargement, and encephalomalacia. Unfortunately, the MRI/MRA evaluation is not available in this patient; however in case 5, mental retardation was determined by an IQ measurement of 70 at 6 years of age, combined with a decreased ability to speak, write, and understand conversation. In this patient, in addition to the MRI/MRA findings at 3 months of age (Table 2), a follow-up head CT showed brain atrophy, ventricle enlargement, and encephalomalacia. Furthermore, MRI evaluation also showed atrophy at the splenium and the posterior part of the corpus callosum, which explains the low IQ measurement (figure not shown). In spite of mental retardation, case 5 was enrolled in mainstream elementary school because the mental retardation was on the borderline. Compared with Akiyama et al (3), who found only 2 of 15 patients with long-term neurologic sequelae, the present study showed a higher incidence; however, the long-term outcomes observed in our cases were similar to those for other cases of ICH, which were not associated with BA. In addition, the neurologic sequelae found in our patients did not correlate with the severity of the clinical findings, the laboratory findings, or the need for craniotomy. These findings suggest that the long-term outcomes cannot be predicted from examinations during admission, and even a mild lesion should be managed properly.
In conclusion, although vitamin K prophylaxis had been given during the neonatal period, VKDB-associated ICH was still found in 7.95% of patients with BA, of whom 4 received exclusive BF. In regard to the low concentration of vitamin K in breast milk compared with regular FM, apparently healthy infants who received exclusive BF during 6 months of life and oral prophylactic vitamin K should be monitored closely. Based on our long-term observation, even though the rate of ICH in patients with BA is low, it should be managed properly to prevent long-term neurologic disability, as shown in 5 of our patients. Based on our study and other studies from Japan, we suggested that 2 mg of oral vitamin K 3 times during the neonatal period should be reevaluated for their protective dose against ICH. A study comparing routes of administration of neonatal vitamin K should be conducted to investigate the effectiveness of oral and IM administration of vitamin K in preventing the occurrence of ICH in infants with BA.
1. Kahn E. Biliary atresia revisited. Pediatr Dev Pathol
2. Okada T, Sasaki F, Itoh T, et al. Bleeding disorder as the first symptom of biliary atresia. Eur J Pediatr Surg
3. Akiyama H, Okamura Y, Nagashima T, et al. Intracranial hemorrhage and vitamin K deficiency associated with biliary atresia: summary of 15 cases and review of the literature. Pediatr Neurosurg
4. Cekinmez M, Cemil T, Cekinmez EK, et al. Intracranial hemorrhages due to late-type vitamin K deficiency bleeding. Childs Nerv Syst
5. Miyasaka M, Nosaka S, Sakai H, et al. Vitamin K deficiency bleeding with intracranial hemorrhage: focus on secondary form. Emerg Radiol
6. Sato H, Node Y, Araki T, et al. Intracranial hemorrhage due to vitamin K deficiency associated with congenital biliary atresia: a case report. Neurosurg Emerg
7. Shearer MJ. Vitamin K deficiency bleeding (VKDB) in early infancy. Blood Rev
8. Majeed R, Memon Y, Majeed F. Clinical presentation of late hemorrhagic disease of newborn. Pak J Med Sci
9. Pichler E, Pichler L. The neonatal coagulation system and the vitamin K deficiency bleeding—a mini review. Wien Med Wochenschr
10. Sakai M. Vitamin K deficiency. Shounika Rinshou
11. Hays DM, Kimura K. Biliary atresia: new concepts of management. Curr Probl Surg
12. American Academy of Pediatrics Vitamin K Ad Hoc Task Force. Controversies concerning vitamin K and the newborn. Pediatrics
13. Van Hasselt PM, De Koning TJ, Kvist N, et al. Prevention of vitamin K deficiency bleeding in breastfed infants: lessons from biliary atresia registry. Pediatrics
14. Van Hasselt PM, de Vries W, de Vries E, et al. Hydrolysed formula is a risk factor for vitamin K deficiency in infants with unrecognized cholestasis. J Pediatr Gastroenterol Nutr
15. Komatsu M, Komatsu F, Tsugu H, et al. Intracerebral hemorrhage despite prophylactic administration of vitamin K in infants: two case reports. Neurol Med Chir (Tokyo)
16. Van Winckel M, De Bruyne R, Van de Velde S, et al. Vitamin K, an update for paediatrician. Eur J Pediatr
17. Ijland MM, Pereira RR, Cornelissen EAM. Incidence of late vitamin K deficiency bleeding in newborns in the Netherlands in 2005: evaluation of current guideline. Eur J Pediatr
18. Shroff MM, Soares-Fernandes JP, Whyte H, et al. MR imaging for diagnostic evaluation of encephalopathy in the newborn. Radiographics
19. Huisman TAGM. Intracranial hemorrhage: ultrasound, CT and MRI findings. Eur Radiol
20. Huisman TAGM, Singhi S, Pinto PS. Non-invasive imaging of intracranial pediatric vascular lesions. Childs Nerv Sust
21. Matsuo S, Suita S, Kubota M, et al. Hazards of hepatic portocholecystostomy in biliary atresia. Eur J Pediatr Surg
22. Lykavieris P, Chardot C, Sokhn M, et al. Outcome in adulthood of biliary atresia: a study of 63 patients who survived for over 20 years with their native liver. Hepatology
23. Hussein A, Wyatt J, Guthrie A, et al. Kasai portoenterostomy: new insights from hepatic morphology. J Pediatr Surg
24. Behrmann BA, Chan WK, Finer NN. Resurgence of hemorrhagic disease of the newborn: a report of three cases. Can Med Assoc J
25. Fukuda T, Akimoto J, Chin M, et al. Intracranial hematoma accompanying bleeding tendency: therapeutic practice and analysis of literature. No Shinkei Geka
26. Hanawa Y, Maki M, Murata B, et al. The second nation-wide survey in Japan of vitamin K deficiency in infancy. Eur J Pediatr
Keywords:Copyright 2012 by ESPGHAN and NASPGHAN
biliary atresia; intracranial hemorrhage; vitamin K–deficiency bleeding