Presentation of Case: Intraparenchymal and intracranial hemorrhage. Cause?
Dr. Madeline L. Keyes (Pediatrics): A 32-day-old male infant was transferred to the emergency department and admitted to this hospital for evaluation of suspected abusive head trauma.
On the evening of admission, the patient’s mother called emergency medical services and reported that the patient had fallen. The mother reported that the infant had been well until that evening, when she placed him on her bed before briefly stepping a short distance away to warm up a bottle of formula in the microwave. She heard a thud and then returned to find that the baby had fallen less than 1 m (2 to 3 ft) from the bed to the hardwood floor, where he was lying face down and crying. She picked him up immediately to console him. After continuing to cry for several minutes, the patient suddenly became quiet; this behavior was worrisome to the mother and prompted her to call for emergency medical assistance. During the phone call, the baby resumed crying. On arrival of emergency medical services personnel several minutes later, the infant appeared well and was bottle-feeding without difficulty; there were no signs of injury. He was transported by ambulance to the emergency department of another hospital for evaluation.
At the other hospital, the patient’s mother reported that the baby had not lost consciousness or vomited during the event and was behaving normally. On examination, the patient was alert, regarded the examiner, cried when he was undressed, and calmed when he was swaddled. The vital signs were normal. The head was normocephalic and atraumatic. There were no bruises or subconjunctival hemorrhages, and results of abdominal, musculoskeletal, and neurologic examinations were normal, as was the remainder of the physical examination.
Dr. Paul A. Caruso: Computed tomography (CT) of the head, performed without the administration of contrast material, revealed scattered punctate densities in the right and left superior occipital lobes, a density (3 mm in diameter) extending along the left tentorial leaflet, mild prominence of both opercula, and mild diffuse prominence of the extraaxial spaces (Figure 1).
Dr. Keyes: A report of suspected abusive head trauma was filed with the Massachusetts Department of Children and Families, and the patient was transferred to the emergency department of this hospital for further evaluation.
On arrival at this hospital, a history was obtained from the patient’s mother. The patient was born at another hospital after 37 weeks of gestation; labor was induced because of maternal preeclampsia, and the infant was delivered vaginally. At delivery, a nuchal cord was unwrapped from the baby’s neck. He received routine newborn care and was feeding and growing well. Two weeks before the current evaluation, he was seen in the emergency department of the first hospital because of three episodes of vomiting; abdominal ultrasonography performed during that visit did not reveal findings consistent with pyloric stenosis. During the 10 to 14 days before the current evaluation, the patient was fussy. He had a diaper rash, for which his pediatrician prescribed topical nystatin; in addition, his mother was applying a zinc oxide–containing cream to the rash. He took no other medications and had no known allergies. His mother had a history of asthma, post-traumatic stress disorder, and anxiety; she had placed frequent calls to emergency medical services because of concerns about her health but was often reassured by the responding personnel and not transported to a hospital. Both of the patient’s maternal grandparents had a history of substance use disorder, and his maternal grandmother had died of a heroin overdose.
The patient lived with his mother in housing funded by the Department of Mental Health. They were visited each week by a case manager and received social support from an aunt. The patient’s mother was a high school graduate and was certified as an emergency medical technician; she had been in the custody of the Department of Children and Families during adolescence and had a 2-year-old daughter who had been placed with the aunt at the recommendation of the Department of Children and Families because of concerns about the mother’s mental health issues. Although the case involving the daughter was closed, the mother reported that she still had contact with her former social worker, who continued to provide support. The pregnancy with the patient was unplanned and occurred near the beginning of the mother’s relationship with the father of the baby; this relationship was characterized by young age and intimate-partner violence, and the couple separated before the birth of the infant. The mother was the patient’s primary caregiver, although on two occasions (the most recent one occurring 6 days before this evaluation), the baby had spent time with his father.
On examination, the patient was alert and vigorous and appeared well. The temperature was 36.6°C, the pulse 161 beats per minute, the blood pressure 86/50 mm Hg, the respiratory rate 56 breaths per minute, and the oxygen saturation 99% while he was breathing ambient air. The weight was 3.99 kg (18th percentile), the length 53 cm (17th percentile), and the head circumference 35 cm (2nd percentile). There were no exterior signs of facial or scalp trauma, and the anterior fontanelle was soft and flat. An intraoral examination did not reveal lesions; the lingual frenulum was intact. The abdomen was soft, without distention, tenderness, or masses. Examination of the skin revealed scattered skin-colored papules (1 mm in diameter) on the chest and face, a crescent-shaped erythematous patch above the right nipple, a blanching erythematous macule (3 mm in diameter) on the left buttock, erythema of the intergluteal cleft, and a scratch (3 mm by 2 mm) behind the left knee. The remainder of the physical examination was normal. Laboratory test results are shown in Table 1. The patient was admitted to the hospital, and additional imaging studies were obtained.
Dr. Caruso: Susceptibility-weighted magnetic resonance imaging (MRI), performed 16 hours after the head CT, revealed a lobular focus of hemorrhage (15 mm by 2 mm) in the right superior occipital gyrus that was probably chronic and corresponded to the 2-mm punctate density seen on non–contrast-enhanced CT. A similar focus of hemorrhage in the left superior occipital gyrus was also probably chronic. A focus of hemorrhage in the right caudothalamic groove and two foci of hemorrhage or mineralization (3 mm in diameter) in the bilateral choroid plexuses were also seen (Figure 2).
Dr. Keyes: Additional evaluation was performed, and management decisions were made.
Differential Diagnosis
Dr. Alice W. Newton: I am aware of the diagnosis in this case and participated in the care of this patient. This 1-month-old infant reportedly had a short fall at home. Although he appeared to be neurologically normal, a CT scan of the head showed multiple areas of intraparenchymal bleeding as well as enlarged extraaxial spaces, findings suggestive of subdural hemorrhage. In this case, it is critical to rule out possible abusive head injury. On the basis of this concern, I will construct a differential diagnosis to consider nonaccidental head injury as well as other medical conditions that may result in subdural bleeding that mimics trauma due to child abuse.
ACCIDENTAL TRAUMA
This infant reportedly had a short fall with no signs of bruising or head injury and no loss of consciousness. The vast majority of household falls in infants result in no injury, and in most cases, parents do not seek medical attention.1 In this case, the mother reported that she did not see the fall but was nearby when it happened, and she said that the infant cried immediately and was able to be consoled. This minor household fall would not explain intracranial hemorrhage (possibly of a chronic nature) in multiple areas of the brain.
ARTERIOVENOUS MALFORMATION
The presence of intracranial hemorrhage leads to the consideration of possible ruptured vascular malformation. The most common vascular malformation in infants, albeit rare, is a vein of Galen abnormality,2 which may be diagnosed on the basis of magnetic resonance angiographic findings such as dilated venous structures in the posterior fossa that are likely to displace cerebellar tissue. Other forms of arteriovenous malformation — such as those seen in PHACE syndrome, defined by posterior fossa anomalies, hemangiomas, arterial and cardiac anomalies, and eye and endocrine abnormalities — are exceedingly rare. This patient’s clinical presentation and neuroimaging findings are not consistent with rupture or leaking of a vascular malformation.
METABOLIC DISEASE
Metabolic disease is a consideration, with glutaric acidemia type I being the metabolic disorder most commonly associated with unexplained subdural hemorrhage in infants. This condition rarely leads to clinical signs in infancy and often comes to medical attention with an episode of metabolic decompensation associated with illness. In untreated infants, microencephalic macrocephaly may develop over time, possibly leading to movement disorder and intellectual disability. Because there are a few published cases of glutaric acidemia type I that resulted in bleeding in the subdural space, this rare metabolic condition is classically considered as a possible alternative diagnosis to abusive head trauma.3 However, studies show that subdural hemorrhage in glutaric acidemia type I is exceedingly uncommon and is associated with other brain abnormalities, including cerebral atrophy.4 This infant did not have any of the head imaging findings that would commonly suggest glutaric acidemia type I, and routine newborn screening for this disorder had been negative.
BENIGN HEMORRHAGE
Many infants have prominent subarachnoid spaces on imaging. Some have benign enlargement of the subarachnoid space (BESS), a clinical syndrome that leads to mild macrocephaly.5 These infants usually receive this diagnosis when their head circumference increases disproportionately to the rest of their growth measurements; the increased head circumference is noted at approximately 3 to 4 months of age, with the accelerated growth slowing during the second year of life. In rare cases of BESS in infants, imaging reveals an incidental small collection of localized asymptomatic subdural hemorrhage that is thought to result from minor trauma with associated rupture of bridging veins, which have been stretched owing to the presence of excessive subarachnoid fluid collection. Because this phenomenon is rare and inflicted injury is much more common, the presence of unexplained subdural hemorrhage in an infant, even one with BESS, leads to an evaluation for signs of possible child abuse.
CEREBRAL SINUS VENOUS THROMBOSIS
Although cerebral sinus venous thrombosis is rare, with an estimated incidence of 0.67 in 100,000 children per year, the development of this condition can result in severe neurologic complications and death in infants.6,7 Initial cranial imaging may show signs of intracranial hemorrhage with hemorrhagic infarction of brain tissue. It is important to note that cerebral sinus venous thrombosis does not occur without an inciting event, such as dehydration, infection, a prothrombotic disorder, or an underlying chronic disease, such as cancer or sickle-cell disease. In addition, infants with this condition usually appear ill, often presenting with seizure, focal neurologic signs, vomiting and lethargy, or a decreased level of consciousness, findings not consistent with this patient’s presentation.
BIRTH TRAUMA
Asymptomatic intracranial hemorrhage in the subdural, intraventricular, and parenchymal regions has been documented in full-term infants after birth. The most common type is subdural hemorrhage, which has been detected in 26 to 50% of full-term infants.8,9 Subdural hemorrhage is most often found in the posterior supratentorial region, although it may also be seen in the posterior fossa. It is usually related to the application of direct forces on the head — either compression during vaginal birth with associated stretching of the tentorium as the head is distorted, or external forces related to vacuum or forceps delivery. In addition, a small but significant number of infants delivered by uncomplicated cesarean section are found to have asymptomatic subdural hemorrhage. Most cases of subdural hemorrhage resolve within 1 month, with all cases resolving in less than 3 months.9 A significant number of infants have hemorrhage not only in the subdural region but also in other compartments of the brain, including parenchymal, periventricular, and subarachnoid hemorrhage.10 This patient’s clinical presentation is consistent with a diagnosis of intracranial bleeding related to birth trauma, provided that we can rule out abusive head trauma in this case.
ABUSIVE HEAD TRAUMA
In cases of suspected child abuse, a complete and detailed medical history should be taken, with attention paid to adherence to proper well-child care, nutritional status, medications and vitamins, previous episodes of trauma or unexplained bruising or bleeding, and unexplained neurologic symptoms that may represent past episodes of abuse, including a history of apnea, altered mental status, or unexplained vomiting. Symptoms of abusive head trauma occur on a spectrum, with mild injury leading to minimal symptoms that may be mistaken for viral illness or normal newborn irritability. Among patients with a diagnosis of abusive head trauma, the incidence of unrecognized past episodes of abuse has been estimated to be 25 to 30%.11,12
This infant had a household fall without any neurologic consequences, and the physical examination was reassuring. However, two features of this case — the scratch behind his knee and the history of unexplained vomiting that was severe enough for the child to undergo testing to rule out pyloric stenosis — are worrisome. The interpretation of imaging findings from the other hospital also raised concerns, with possible intraparenchymal and “chronic” subdural hemorrhage noted.
On the one hand, there are several social risk factors for child abuse in this case, including the parents’ history of intimate-partner violence, young age, and poverty status, as well as the mother’s history of mental illness. On the other hand, there are clear family strengths and supports: the mother had ongoing housing assistance and mental health treatment from the Department of Mental Health, as well as active help from her aunt, who was raising her other child. Initial evaluation of the infant for abusive head trauma was driven by the social risk factors, since the child did not have clinical signs or symptoms that would prompt imaging evaluation for head injury. Although some findings on the head CT were interpreted correctly, the initial interpretation of the CT findings as possibly chronic subdural hemorrhage increased the level of concern about abusive injury.
MRI evaluation of the infant provided important information in that the pattern of hemorrhage and normal brain findings in the context of the healthy clinical status of the infant did not fit classic findings of abusive head trauma. Subdural hemorrhage along the tentorium is a finding commonly associated with birth trauma. Intraventricular hemorrhage in small amounts has also been described with birth trauma, as has the presence of subpial hemorrhage. Although it is possible for birth-related hemorrhage to occur in multiple compartments, such extensive areas of hemorrhage should prompt a hematologic evaluation for possible bleeding disorders, since signs and symptoms of bleeding dyscrasia may mimic child abuse.13 Making the determination of whether a child has been abused is often not clear-cut. Given the ramifications to the child and the family, the determination must be based not only on descriptions in the medical literature but also on review of all the features of the case in combination with experienced clinical judgment. In this case, evaluation for other signs of physical abuse was reassuring, with normal results of a retinal examination, initial laboratory testing, and a skeletal survey. Taken together, these features suggest that the most likely diagnosis in this case is hemorrhage due to birth trauma rather than abusive head trauma.
Dr. Virginia M. Pierce (Pathology): Dr. Ebb, what was your clinical impression when you evaluated this infant?
Dr. David H. Ebb: In cases of suspected child abuse, the process begins with an effort to elicit any history of abnormal bleeding or bruising in the child or family as well as a thorough physical examination to look for evidence of acute or subacute bleeding or bruising. Diagnostic studies include imaging of the head, trunk, arms, and legs to identify bleeding that may not be immediately apparent on examination.14 Imaging evaluation must be undertaken in concert with blood tests to look for evidence of an inherited or acquired bleeding disorder, which may guide emergency therapeutic interventions.15 In cases in which a rare bleeding disorder is discovered, it is the hematologist’s role to assist the multidisciplinary team in ensuring that the child is not inappropriately removed from the family when no clear history of abuse is elicited and a robust network of social supports can be documented.
In reviewing the history of the child and the family, it is important to learn whether there is a history of mucocutaneous bleeding (e.g., bruising, epistaxis, or menorrhagia), which could suggest a quantitative or functional platelet defect. A history of joint bleeding or deep-muscle hematoma is most frequently elicited in the context of a clotting-factor deficiency. Although the complete blood count, prothrombin time, partial-thromboplastin time, and fibrinogen level can be used to identify most inherited or acquired coagulation defects that cause bleeding complications in infants or older children, these measurements cannot be used to identify deficiencies in factor XIII, which may be associated with delayed separation of the umbilical cord stump. Of note, easy bruising or mucocutaneous bleeding may be caused by von Willebrand’s disease, the most common inherited bleeding disorder. This disease causes impaired platelet binding to injured vessel walls, and a specific panel of tests is needed to confirm or rule out its contribution to unexpected bruising or bleeding.
On further review, this patient’s family history was notable for easy bruising and epistaxis in maternal relatives of both sexes, a feature that suggests the possibility of von Willebrand’s disease. Laboratory studies revealed a normal platelet count and fibrinogen level and normal results of screening tests for clotting-factor deficiencies (i.e., measurements of prothrombin time and partial-thromboplastin time). A subsequent panel of tests for von Willebrand’s disease revealed a relatively low factor VIII level and normal levels of ristocetin cofactor activity and von Willebrand factor antigen. This patient’s constellation of family history and laboratory findings was thought to be most consistent with either mild factor VIII deficiency (hemophilia A) or von Willebrand’s disease type 2N, which is a rare variant; further laboratory investigation was needed. Although von Willebrand’s disease is relatively common in its mildest form, it rarely causes intracranial hemorrhage.13,15 The possibility that this child’s bleeding occurred in the absence of any identifiable predisposing factors was also strongly considered.
Clinical Diagnosis
Possible intracranial hemorrhage at parturition due to a bleeding disorder, most likely von Willebrand’s disease.
Dr. Alice W. Newton’s Diagnosis
Birth trauma resulting in subdural and intraventricular hemorrhage.
Pathological Discussion
Dr. Grace Linder: Interpretation of neonatal coagulation test results can be challenging because of developmental hemostasis. At birth, full-term newborns can have prothrombin-time values that are higher than the upper limit of the normal range for adults. In addition, the mean activated partial-thromboplastin time for newborns is higher than the upper limit of the normal range for adults.16 Although factor XIII and fibrinogen levels at birth are similar to normal adult levels, factor VIII and von Willebrand factor levels can be elevated at birth and then fall to normal adult levels over a period of 3 to 6 months.17 In contrast, the blood levels of procoagulant factors II, VII, IX, X, XI, and XII and anticoagulant proteins antithrombin, protein C, and protein S are physiologically low in newborns and increase with age, reaching normal adult levels by approximately 6 months of age (except the protein C level, which can remain slightly below the lower limit of the normal range for adults throughout childhood).16-19
It is critical to optimize preanalytic variables to obtain accurate coagulation test results. Traumatic venipuncture, hemolysis, and inadequate filling of the sodium citrate tube can alter the prothrombin time and activated partial-thromboplastin time. Factors V and VIII are labile, and sample clotting and delays in specimen processing can lead to factor degradation.19 When abnormal coagulation test results are present in infants, the tests should be repeated to confirm persistence with age and to rule out false results caused by preanalytic issues.
The initial screening tests in this patient did not reveal any abnormalities. Given the patient’s maternal history, a workup for von Willebrand’s disease was pursued. There are several variants of von Willebrand’s disease, which are classified as quantitative (types 1 and 3) or qualitative (types 2A, 2B, 2M, and 2N).20 Laboratory testing for von Willebrand’s disease includes quantitative assays for factor VIII and von Willebrand factor as well as a functional assay, such as a ristocetin cofactor activity assay, which assesses the ability of von Willebrand factor to bind platelet glycoprotein 1b. The level of von Willebrand factor varies with age and ABO blood type and can increase during acute-phase reactions.21 Definitive diagnosis of von Willebrand’s disease may require repeat testing.
In this patient, results of a repeat panel of tests for von Willebrand’s disease were unremarkable, with normal levels of factor VIII (120%), von Willebrand factor (122%), and ristocetin cofactor activity (116%) for his blood type (blood type O) and age. Results of additional testing for rare causes of bleeding disorders — including studies of α2-antiplasmin activity, reptilase time, thrombin time, and platelet aggregation — were normal. Overall, there is no laboratory evidence to support a diagnosis of a bleeding disorder.
Discussion of Management
Dr. Ebb: This infant was discharged in his mother’s custody, with close supervision by the Department of Children and Families. He has been brought back to the hematology clinic on multiple occasions for follow-up examinations and further laboratory studies. He is a thriving infant who has appropriate growth and development, without evidence of bruising. The results of his neurologic examinations are normal for his age. Repeat MRI has shown expected evolution of the previously noted parenchymal bleeding on imaging, with no evidence of new intracranial bleeding. Genetic testing for von Willebrand’s disease type 2N was negative. In the patient’s mother, the panel of tests for von Willebrand’s disease revealed factor levels in the low normal range. Ultimately, our extensive laboratory studies have not shown any evidence of an underlying bleeding disorder.
It is important to note that confirmation of a suspected bleeding disorder in an infant or child does not rule out the possibility of nonaccidental trauma. Decisions regarding the child’s welfare and appropriate disposition must be based on careful consideration of all features of the history, examinations, and imaging and laboratory studies, as well as the integrity of the child’s and parents’ social supports. Involvement of a pediatrician familiar with the Child Protection Program and child abuse is a critical component of this very complex endeavor.
Final Diagnosis
Intraparenchymal and intracranial hemorrhage presumably due to birth trauma.
This case was presented at Pediatric Grand Rounds.
Disclosure forms provided by the authors are available with the full text of this article at NEJM.org.
No potential conflict of interest relevant to this article was reported.