Vitamin K Deficiency

Vitamin K is a fat-soluble vitamin required for the carboxylation of certain clotting factors (II, VII, IX, X) in humans. Vitamin K deficiency bleeding (VKDB) is one of the most commonly reported mimics of AHT. It was initially described by Townsend in 1894. At birth, the newborn pool of vitamin K is less than that required for robust production of vitamin K–dependent clotting factors. In the vast majority of infants, this relative lack of vitamin K stores remains clinically imperceptible, although most symptomatic infants present within the first week of life. In 1939, administration of vitamin K was demonstrated to dramatically reduce the incidence of this potentially fatal condition. The clinical manifestation of VKDB traditionally has been characterized by the timing of presentation: early (<24 hours of life), classic (1–7 days), and late (2–12 weeks). Typical symptoms of VKDB include rectal bleeding, bruising, hematemasis, and hemoptysis. The major source of vitamin K for the neonate is dietary. Maternal vitamin K deficiency can contribute to the development of VKDB in the fetus and newborn. ^, Medical conditions that impair the infant’s absorption of vitamin K can increase the risk of VKDB in an infant. These include biliary atresia, liver dysfunction, cystic fibrosis, α-1-antitrypsin deficiency, malabsorption, and Alagille’s syndrome. ^20-23 Importantly, medications or exposure to toxins also need to be considered. ^,

In most developed nations, the vast majority of infants receive oral or intramuscular vitamin K supplementation. Vitamin K replacement has been shown to lower the incidence of ICH caused by VKDB. The intramuscular route of administration is the preferred route because of reports of infants receiving only oral supplementation who have developed ICH from VKDB. The administration of vitamin K does not completely prevent the development of VKDB, ^, or ICH from VKDB, but the rates are dramatically lower.

Breast milk provides inadequate vitamin K for the nursing infant. Infants fed cows’ milk-based formula have a much lower risk of VKDB than those who are strictly breast fed. Breastfeeding increases the risk of VKDB in infants who have not received perinatal vitamin K supplementation. Infants who have received parenteral vitamin K and are breast fed will have coagulation profiles similar to those infants who have received vitamin K and are bottle fed. Although common in European countries, oral administration of vitamin K also has less of a protective effect than does parenteral administration.

The most common presentation of the early form of VKDB is profound and potentially fatal intracranial bleeding. Recent data support ICH (most commonly, subarachnoid hemorrhage) from VKDB as being limited to the first 12 weeks of life, unless an underlying medical condition or hepatic disease is present. In infants with VKDB-related ICH, RHs are rarely seen. ^, Whereas RHs have been noted in children with ICH from VKDB, no cases of retinoschisis or retinal folds have been reported. Infants with ICH from VKDB will have abnormal coagulation studies (PT, aPTT) including an elevated level of protein induced in vitamin K absence (PIVKA-II). There have been no reported cases of ICH from VKDB in infants with normal PT, aPTT and PIVKA.

Factor XIII Deficiency

Factor XIII (FXIII) is found both in plasma and on platelets, and it is the last enzyme in the coagulation/thrombosis cascade. It has two major functions: to cross link fibrin monomers (essential in clot formation) and to cross-link α ₂ -antiplasmin to fibrin (essential in clot stability). FXIII deficiency was first described in 1960 and is another recognized cause of bleeding in children to be mistaken for AHT. The “classic” presentation of FXIII deficiency is delayed or prolonged bleeding from the umbilical stump (seen in >80% of infants with FXIII deficiency), along with delayed umbilical cord separation and bleeding during circumcision. Although the majority of affected infants will have umbilical stump bleeding, the absence of this historical finding does not exclude the diagnosis. As with most inherited disorders, consanguinity increases the risk of FXIII deficiency. Although most bleeding in FXIII deficiency involves the skin and soft tissues, the incidence of ICH is higher than that seen in other factor deficiencies. ICH is the cause of death in 30% of those who die from hemorrhagic complications of their FXIII deficiency. Since FXIII is integral to clot stability by protecting against dissolution, affected individuals are subject to prolonged and recurrent bleeding from a single event. Although ICH has been described in children and adults with mildly suppressed levels (10-30% FXIII activity), _, it is generally accepted that FXIII deficiency usually becomes clinically apparent with levels of less than 5% activity. In the reported cases of infants and children with ICH in the presence of FXIII deficiency, RHs are either not mentioned, or described as not present. FXIII deficiency cannot be identified by routine coagulation testing (i.e., PT and aPTT); specific testing is required.

Hemophilia A (Factor VIII Deficiency)

The most common inherited coagulopathy that causes ICH is classic hemophilia A. Birth-related ICH in hemophilia A occurs in 3% to 8% of those infants later found to have the condition, with mortality reported to be as high as 33%. Early reports of the presence of ICH caused by birth in infants without an identified bleeding disorders stated that it occurred in approximately 0.11% of births, although recent MRI studies of nonsymptomatic, normal newborns have shown much higher prevalences. ^, ICH in newborns can be the presenting symptom of infants with hemophilia A. More importantly, 1.9% of newborns with hemophilia A had only ICH without other clinical co-morbidities. In older infants with hemophilia A, ICH presents without a history of trauma approximately 10% of the time, with those who have more severe hemophilia accounting for the majority. The diagnosis is usually not in question since these children all have profoundly abnormal coagulation studies (i.e., PT and aPTT).

Von Willebrand Disease

Von Willebrand disease (vWD) is the most common coagulopathy worldwide. The majority with this condition have mild disease and are identified on asymptomatic routine screening (e.g., preoperative evaluation of coagulation). Broadly, vWD is classified into types 1, 2, and 3, with 80% of those affected having type 1. The diagnosis of vWD can be complex or nuanced, often involving relative levels of von Willebrand factor (vWF), von Willebrand factor ristocetin cofactor (vWF:RCo), and Factor VIII (FVIII). Given the importance of vWF in platelet function, testing for platelet defects can often reveal vWD. ^,

Compared with hemophilia A, vWD is not a commonly identified cause of ICH or RH in newborns and infants. One survey of the medical literature revealed only 23 reported cases of intracranial bleeding associated with vWD. In a series of 42 boys with vWD, one ICH was identified, a posttraumatic SDH in a 15-month-old who fell while running. Wetzstein and colleagues reported a term newborn who presented with SDH and transverse sinus thrombosis who was subsequently diagnosed with vWD type 3. On presentation, the infant had a prolonged aPTT. Of note, the neonate did not receive vitamin K at birth. Only three reported cases of retinal bleeding associated with vWD exist. ^, These were in individuals 13, 19, and 33 years of age, none of whom had a history of trauma. There are no published reports of vWD being associated with RHs, retinoschisis, or retinal folds in infants or children.

Trauma-Related Coagulopathy

The presence of abnormal coagulation studies in children and adults with traumatic brain injury (TBI) is well described. A meta-analysis of recently published reports identified an overall prevalence of coagulation abnormalities of 32.7% in patients with TBI, with reports of incidence as high as 60% in those with severe TBI. In children specifically, the rate of coagulopathy associated with TBI has been reported to be as high as 77%. Exposure of brain parenchyma to the circulation can initiate a cascade of both fibrinolysis and thrombosis. Systemic hypoperfusion likely plays an important role in augmenting the coagulopathy associated with TBI. This cascade can be local and self limited, or it can lead to disseminated intravascular coagulation (DIC). The extent of resulting DIC is correlated to the severity of the brain injury. The presence of other systemic injuries does not affect the likelihood of TBI resulting in a coagulopathy. A prospective study of adults with TBI found a significant correlation between the serum level of fibrin degradation products (FDPs) and the severity of TBI. As the FDP level increased, the prognosis worsened. Similar data were reported in a population of children diagnosed with AHT. Hypoxia alone with normal perfusion (and without TBI) would not be expected to cause DIC (e.g., near-drowning). Prolonged hypoxia with systemic hypoperfusion, however, can itself trigger DIC. DIC can result in a nonspecific pattern of RHs. However, retinal folds and retinoschisis do not occur in DIC alone without trauma as a component.

Coagulopathy associated with TBI has been described in children with AHT. In the presence of concerning skeletal findings or historical data, brain injury with coagulopathy could be due to trauma. A normal FDP in a child with ICH would indicate the blood not being due to TBI-associated coagulopathy.

Platelet Disorders

Defects of platelet function or number can be responsible for bruising or bleeding with routine activities. The most common clinical findings seen in children with abnormalities of platelets are mucocutaneous bleeding (e.g., epistaxis), ecchymoses, purpurae, and hemorrhagic surgical complications. ^, It is important to consider both the number and function of platelets in the evaluation of a platelet disorder. A normal platelet count is between 150,000 and 400,000 per mm ³ . Size and shape of the platelets can be assessed by microscopic evaluation of the peripheral smear. In addition, there are functional tests available to investigate qualitative platelet defects (see below). An excellent guide outlining all platelet disorders and their clinical spectra is available.

The most common pathological cause of bruising in childhood is idiopathic thrombocytopenic purpura (ITP). ITP is a self-limited autoimmune disease that results in destruction of platelets and is the most common coagulopathy mistaken for abuse. Serious bleeding is very uncommon in children with ITP, although ICH can occur, specifically when platelet counts drop below 10,000 per mm ³ . RHs have been very rarely reported in children with ITP. ^, ITP should not be mistaken for abuse, since the definitive diagnostic test—a platelet count—is readily available and should be part of the very basic evaluation of a child with bleeding or bruising.

The two most common platelet dysfunction conditions in children are Glanzmann thrombasthenia (GT) and Hermansky-Pudlak syndrome (HPS). Although there are case reports of these conditions being mistaken for child abuse, ^, coagulation studies were abnormal in both instances. In GT, a surface platelet fibrinogen receptor defect results in poor platelet-platelet clumping (and clot formation). Most affected patients present with epistaxis, bruising, gingival bleeding, and petechiae; ICH is rare. Ocular findings are also quite rare in GT, and RHs have not been reported.

HPS is often cited as a potential cause of retinal hemorrhaging in children. Only one report exists of an infant with HPS who had retinal and subdural hemorrhages. In this report, a 7-week-old presented with seizure activity and was found to have an occipital SDH. The infant was found to have fewer than a dozen posterior pole hemorrhages on the right and a single macular hemorrhage on the left. The ocular findings (strabismus and nystagmus) in HPS are dependent upon the retinal pigment and macular function; the vascular integrity is not affected. HPS is usually not a diagnostic dilemma, since affected individuals have albinism.

What Evaluation Should Be Performed to Rule Out Coagulopathy in Suspected Abuse Cases?

Inherited coagulopathies are a small but important cause of ICH in children. It is worth noting that the presence of a bleeding disorder does not preclude the diagnosis of AHT. One study of 50 children referred to a hematology service in Edinburgh, Scotland with suspected nonaccidental injury found eight (16%) with coagulation abnormalities. The authors highlight that of the eight children, seven were confirmed to have been abused, with four of those having normalization of their screening tests upon repeat examination. Many general reviews of appropriate coagulopathy testing in suspected cases of abusive injuries have been published. ^{, .} A detailed history and physical examination is the cornerstone for identification of a coagulopathy. An extensive history should include family history, medication history, prior hospitalization or surgeries, consanguinity, dietary history, and recurrent miscarriages in family members. For neonates and infants, a specific history of vitamin K administration and route is crucial. A history of umbilical stump bleeding and prolonged period to cord separation should be noted. A meticulous review of systems should include unexplained bruising, prolonged bleeding, epistaxis, gingival bleeding, and menorrhagia. The laboratory evaluation of a potential bleeding disorder should be carried out after all of this information is gathered.

An initial screening panel of labs should include the following:

• •
  

  Complete blood count (CBC), with a platelet count

  
  
• •
  

  Review of a peripheral smear (for assessment of platelet size and form)

  
  
• •
  

  Activated partial thromboplastin time (aPTT)

  
  
• •
  

  Prothrombin time (PT)

  
  
• •
  

  Protein induced in vitamin K absence (PIVKA-II) (for infants under 3 months)

This initial battery of tests should exclude any significant predisposition to bleeding or severe factor deficiency. However, some conditions might have a normal initial screening evaluation but may still be clinically relevant. These include mild hemophilia (A or B), FXIII deficiency, mild vWD, or functional platelet disorders. ^, If a bleeding predisposition is still a clinical concern, a secondary panel of labs should include:

• •
  

  Specific Factor levels (specifically FXI and FVIII levels)

  
  
• •
  

  Factor XIII activity level

  
  
• •
  

  vWF antigen level

  
  
• •
  

  vWF ristocetin cofactor assay (vWF:RCo)

  
  
• •
  

  Platelet function testing (i.e., PFA-100)

  
  
• •
  

  Thrombin time (TT)

  
  
• •
  

  Fibrinogen assay

PIVKA-II

Historically, vitamin K deficiency was identified by correction of an elevated PT after administration of vitamin K. Currently, testing for vitamin K deficiency can be done in two ways: measuring serum vitamin K levels, and measuring levels of proteins induced in vitamin K absence (PIVKA-II). ^, Vitamin K levels do not correlate to the state of coagulopathy and can be misleading. PIVKA-II testing identifies an undercarboxylated prothrombin protein present in the serum of children who are vitamin K deficient. PIVKA-II can be detectable in approximately 25% of all newborns, term or preterm. PIVKA-II can be detected in up to 50% of cord blood samples while vitamin K levels in cord blood are negligible. After parenteral vitamin K administration, PIVKA-II levels will be universally undetectable within 2 weeks, but vitamin K might not result in complete PIVKA-II suppression. In vitamin K deficiency, PIVKA-II will be elevated prior to elevation of the PT or aPTT. Medications (e.g., warfarin) or malabsorption (e.g., cystic fibrosis or liver disease) can result in elevated PIVKA-II despite a normal dietary vitamin K intake.

Since there are other causes of PT elevation, it is not a sensitive screen for vitamin K deficiency, but it can be used to screen for coagulopathy resulting from vitamin K deficiency. There have been no reported cases of VKDB in children with normal PIVKA-II.

Platelet Function Testing

Rarely, screening tests can miss a clinically relevant bleeding predisposition caused by platelet dysfunction. An excellent review of specific platelet conditions and their testing has recently been published. Platelet functioning has historically been tested by an Ivy bleeding time (BT). The BT is performed by a technician using a standardized template creating a 10-mm long and 1-mm deep incision, typically on the forearm. As an operator-dependent test, BT results are difficult to standardize. The sensitivity and specificity of BT for identifying a bleeding disorder are unacceptably low. ^, For these reasons, in addition to the risk of scaring and infection, BT should no longer be part of routine coagulation testing in children. Platelet aggregometry testing is a series of challenge tests to patient whole-blood or serum. The patient sample is placed with aggregation agonists (i.e., ristocetin, thrombin), and time to platelet clumping is measured. Platelet response to ex vivo vascular shear injury is measured by the platelet function analyzer (PFA-100) (Dade Behring, Marburg, Germany). This test uses collagen membranes over which the patient’s citrated blood flows as an attempt to duplicate vascular injury. Time to cessation of flow is measured, with a prolonged time being abnormal. Since this test measures occlusion of flow through a column, thrombocytopenia and anemia can confound the test results. Because of significant overlap of reported normal ranges, true pathology (i.e., subtle qualitative platelet disorders) can be missed. Any abnormal PFA-100 test needs to be clarified as to the precise abnormality. As noted earlier, platelet function testing can identify children with vWD. In addition, testing with platelet flow cytometry and platelet nucleotide content is available but requires the guidance of an experienced pediatric hematologist.

Testing for Factor XIII Deficiency

Since FXIII is involved in clot formation and stability and is terminal in the coagulation cascade, PT and aPTT testing are normal in patients with FXIII deficiency, even with a profound predisposition to bleeding. Patients with FXIII deficiency will have an unstable clot formed when placed in the 5 M urea solution. The clot then stabilizes with the addition of normal serum. ⁴¹ Patients with normal PT, aPTT, and TT who have an abnormal clot solubility test are presumptively diagnosed with FXIII deficiency. The clot solubility test is a qualitative test and might only be positive in patients who have complete absence of FXIII. Although most people with inherited FXIII deficiency will be identified via this method, a quantitative FXIII assay might be beneficial. ^, In evaluating children with catastrophic ICH from an unknown etiology, testing for FXIII deficiency by clot solubility testing, FXIII assay, or both might identify rare cases of this deficiency. Of note, patients with FXIII do not form fibrin degradation products, since fibrin linkages are not formed. Hence, the presence of fibrin degredation products rules out significant FXIII deficiency.

Accidents

One of the most common alternative explanations provided for AHT is that of a household accident, most typically a fall. To fully appreciate the difference between the findings in AHT and those occurring in common household accidents, it is helpful to understand both the frequency of serious or fatal injuries from falls, as well as the specific injuries that usually occur with household accidents.

Serious and fatal injuries from household falls can and do occur, albeit rarely. Household accidents result in serious or fatal injuries at a rate that is markedly less than nonfatal or trivial injuries. This can be best appreciated by examining national data on fatalities in the United States. In 2004, the Centers for Disease Control and Prevention published the “Surveillance for Fatal and Nonfatal Injuries—United States, 2001 Report.” The data show that in 2001, there were 55 fatal falls in children 0 to 4 years of age in a population of 20 million children. In contrast, there were 1,039,275 emergency department (ED) visits for nonfatal falls. If it is estimated that the average toddler falls four times per week (with infants falling markedly less and school-aged children falling more than four times a week), then these 20 million children sustain over 1 billion falls per year. These falls resulted in 55 fatalities and more than 1 million ED visits. This rarity of fatal injuries from falls is consistent with another large study of over 11,000 infants.

Studies of witnessed falls in children have indicated that short falls can occasionally result in skull and clavicle fractures, but serious or fatal injuries are unexpected. It is important to recognize that the term “fall” is nonspecific and complex falls that can indeed injure children do occur. The two most common examples of complex falls are falls that involve stairs and falls that involve infant walkers. Falls involving stairways increase the risk for extremity injuries. ^, Falls involving walkers increase the likelihood of a severe brain injury. ^,

Denton and Mileusnic reported a case of a 9-month-old who was noted to have a fatal short fall from a bed and who had a lucid interval of approximately 72 hours. Although this is an intriguing report, attributing the fatal intracranial injuries to the only witnessed fall is speculation. Given the long period between the witnessed event and death, there might have been other injuries (unwitnessed or inflicted) that could have exacerbated, or even been solely responsible for, the terminal injuries.

Plunkett published a report on 18 deaths reported to the U.S. Consumer Product Safety Commission National Injury Information Clearinghouse. This database included information on reports of 75,000 injuries to children on playground equipment. Of the 18 fatalities reviewed, none were infants. Of the five deaths under 2 years of age, only one was witnessed. A 23-month-old who was videotaped falling off playground equipment was reported to be normal for approximately 5 minutes and then became obtunded. She was reported to have bilateral RHs (not further described but noted in the presence of papilledema) and an acute SDH. Most of the falls reported were from greater than 1.5 meters and were onto hard surfaces. Several involved children on swings or children with complicating preexisting conditions.

The entirety of the medical literature on injuries from falls in otherwise normal children indicates that fatal or devastating intracranial injuries from short falls are a profoundly rare event. This is confirmed by Chadwick et al in which the authors identified six such instances in a population of 2.5 million children.

Birth-Related Head Injuries

Injuries from the birthing process have been described since antiquity. Birth-related fatalities occur in 3.1 per every 1000 births. Traumatic birth-related head injuries have been estimated to affect 3.1% of pregnancies. Recognition of birth-related clavicular fractures and cervical nerve root palsies occasionally occur. Although the vast majority of neonates remain unscathed by the birthing process, significant injuries can rarely occur. Many more will have transient birth-related findings or injuries that resolve. The history in conjunction with the physical examination would readily exclude AHT from the differential. One series of 57,600 deliveries identified 17 cases of intracranial bleeding. In only two of these cases was the delivery described as uneventful and free of instrumentation. Neonates with underlying medical (i.e., VKDB) or anatomical conditions (e.g., arachnoid cysts ) that could predispose to ICH would typically present in the newborn period with imaging or neurological findings supporting these conditions.

Whitby et al published a report of 111 term babies who had brain MRI performed within 48 hours of birth. Nine of the 111 babies (8.1%) were found to have SDHs, all of which were clinically silent. Repeat MRI scans at 4 weeks of life demonstrated complete resolution of the intracranial blood without any residual collections or parenchymal injury. The authors conclude that SDHs that result from uneventful births are not clinically significant. These findings were confirmed by Vinchon et al. Similarly, Looney et al evaluated 88 infants with a 3.0T MRI at an average age of 3 weeks. Seventeen infants (26%) had ICH; 16 SDH, 2 SAH, and 6 parenchymal hemorrhage. The authors conclude that the birth-related ICHs are of a different pattern, such that mistaking them for AHT-associated intracranial bleeding is unlikely. Rooks et al recently reported on 101 asymptomatic infants who received cranial MRIs within 72 hours of birth. Sixty-one percent had SDH, all in the posterior portion of the head. When followed up to 2 years of age, all had no physical signs or symptoms of head injury. All SDH resolved by 3 months of age.

RHs occur in up to one third of births ^, and have been described in multiple retinal layers. ^, RHs were most commonly seen in normal vaginal deliveries. All of the hemorrhages noted by Hughes et al were intraretinal, and in 51 of the 53 neonates, the hemorrhages resolved within 16 days. In the other two neonates the RHs were noted at 31 and 58 days; both had been delivered with vacuum assistance. Emerson et al found that 86% of all RHs resolved within 2 weeks of birth, and all intraretinal hemorrhages by 1 month. The only hemorrhage noted at 4 weeks was a single subretinal hemorrhage that was not present at 6 weeks. RHs that are present after the first month of life are unlikely to be from birth.

Intracranial Fluid Collections

Physicians sometimes encounter a child with an enlarging head circumference who, upon head imaging, is noted to have large anterior hypodense, or mixed-density, collections. The differential diagnosis of large anterior hypodense collections includes chronic subdural hematomas, subdural hygroma, benign extraaxial fluid collection of infancy, or frontal cortical atrophy. Given the imprecision of the language used in the medical literature, limited information exists on accurately discriminating amongst these conditions in infants. What is clear from the literature is that in certain infants, the arachnoid space is larger than typical.

Benign extraaxial fluid collection of infancy (BEAF) is asymptomatic and is often identified as an incidental finding on CT imaging. It likely represents a temporary imbalance between CSF production and absorption. CT scanning alone is not able to reliably distinguish BEAF from a chronic subdural collection. Since BEAF is not traumatic in origin, a subdural neomembrane will not form. Subdural neomembranes are seen in infants and children with chronic SDH from either trauma or infection. ^, The neomembrane can be seen on autopsy, intraoperatively during surgery for ICH draining, or on contrast-augmented CT imaging. Because of the finer details on MRI scanning, neomembranes can also be identified on noncontrast MRI images. The presence of a neomembrane would indicate that a hypodense collection is not uncomplicated BEAF. Much controversy exists about whether BEAF predisposes an infant to hemorrhage with routine care or after a trivial injury. ^, A recent finite element study demonstrated that increased extraaxial fluid in the subarachnoid space could actually increase the stability of bridging vessels and decrease the likelihood of subdural bleeding after injury. This issue has not currently been resolved. Parenchymal, intraventricular, and retinal hemorrhages have not been reported with BEAF.

Infants can have mixed-density subdural collection identified on head imaging. Mixed-density collections have historically been described as being indicative of evidence of two distinct episodes of injury or bleeding (“acute on chronic”). Mixed-density collections, however, can appear either after a single significant traumatic event or with rebleeding of a chronic SDH. Single isolated traumatic events can result in an admixture of CSF and acute blood, giving the appearance of either rebleeding into a chronic collection, or multiple episodes of trauma. The admixture of CSF and blood likely occurs via a tear in the arachnoid. ^, Hyperacute SDH can also appear to have multiple layers, perhaps representing the settling out of red blood cells, separating solid components of blood from serum.

Chronic SDH in children can develop either as a maturation of an acute SDH ^, or as a result of extreme prematurity. Chronic SDHs have also been described in neonates, likely the result of an intrauterine insult. ^, Acute intracranial bleeding in an infant presenting with a preexisiting chronic SDH raises the concern of reinjury as the cause of the fresh blood.

Trivial rebleeding into preexisting chronic SDH has been well described in animal models and adults. This rebleeding occurs as a result of oozing from the outer subdural neomembrane and can be exacerbated by a new injury. Because of open cranial sutures, bleeding from trivial oozing would not typically cause neurological symptoms. Since the bleeding from a chronic SDH is intrahematoma , intraparenchymal, intraventricular, or bleeding remote from the chronic SDH (i.e., RHs) could not be attributable to the preexisting collection.

Scurvy

Scurvy is the clinical manifestation of vitamin C deficiency. The typical findings of scurvy are generalized perifollicular petechial bleeding, commonly involving the lower extremities, associated with gingival hypertrophy and bleeding, with 80% of the affected having anemia. Pseudoparalysis caused by lower extremity pain and edema is a common feature. Because of poor collagen formation, affected individuals often have poor wound healing with hair that breaks easily and is described as “corkscrew.” The abnormal hair is similar to that found in Menke’s disease, which is discussed below. Bone demineralization with preservation of the zones of provisional calcification (Frankel lines) along with Pelkan spurs (healing fractures through the zone of provisional calcification) are characteristic radiographic findings. Additional radiographic findings include periosteal elevation from subperiosteal bleeding and linear epiphyseal lines (Wimberger ring). ^, ICHs in scurvy are quite rare. The review of the literature by Gilman and Tanzer of ICH associated with scurvy identifies 13 definitive cases. Seven of these were SDHs and five of the seven were in an infant or child. Of the reports described, all of the infants and children had additional unmistakable findings associated with scurvy.

Ophthalmological findings in scurvy are usually hemorrhagic in nature. Hemorrhage can occur in or around the orbit or in the globe. Conjunctival hemorrhage is the most common ophthalmological finding in scurvy. Rare RHs have been described in adults with scurvy. ^, The adult patients reported in the literature had prolonged abnormal dietary history and exhibited extensive nonophthalmological manifestations of scurvy as well, such as cutaneous hemorrhage, cork-screw hairs, and gingival hypertrophy with bleeding. Only one report exists of RHs noted in a child with scurvy. This 3-year-old child had induced dietary scurvy and presented with unilateral proptosis from a periosteal orbital hematoma. He had ipsilateral SRH and retinal detachment as well as other physical and radiographic stigmata of scurvy.

The diagnosis of scurvy in infants and children is largely clinical since testing for functional vitamin C deficiency is imprecise and clinically difficult. The most reasonable criterion for diagnosing symptomatic scurvy is rapid clinical response to appropriate vitamin C replacement. ^,

Severe and fatal cases of scurvy in infants and children have been described, including subdural and retinal hemorrhages, but clinical features characteristic of scurvy are always apparent. ^, Notably, there have been no reports in the medical literature of an infant whose only manifestation of scurvy was subdural and retinal hemorrhages, without cutaneous bruising, gingival changes, or bone and joint findings on examination or radiograph. Absent other findings, the diagnosis of scurvy would be purely speculative.

Glutaric Aciduria, Type 1

Glutaric acidemia type 1 (or “glutaric aciduria type 1,” [GA1]) is an autosomal recessive metabolic disorder resulting from a mutation in the gene encoding the enzyme glutaryl-CoA dehydrogenase. The enzymatic defect results in a significant movement disorder caused by basal ganglia neuronal loss. The neuronal loss can present with sudden encephalopathy after, or in conjunction with, a mild illness. This is often seen in the first 2 years of life. Prior to neurological decompensation, initial symptoms are minimal, with an increased head circumference in infancy (but not at birth) being a real but subtle finding. ^, Although the genetic mutation can be seen in many ethnicities, it has been noted to occur most prominently in those with northern European lineage.

The brain loss seen in GA1 can result in significant cerebral atrophy. The pattern of atrophy is quite typical and classically involves widening of the Sylvian fissures, regression of the temporal lobes, and characteristic lesions noted in the basal ganglia. The cerebral atrophy results in expanded arachnoid and subdural spaces with the potential for hemorrhaging. Although SDHs are a described feature of GA1, they have not been described in the absence of frontal atrophy. Retinal hemorrhaging (but not retinal folds or retinoschisis) has also been noted in up to 20% of children affected by GA1.

GA1 has been reported as being mistaken for AHT. Since affected children present with sudden, catastrophic, and seemingly unexplained collapse and can have acute and chronic subdural blood collections and RHs, inflicted injury would be an obvious concern. Since fractures are not a feature of GA1, the presence of any fractures would be of great importance in excluding this diagnosis. GA1 is diagnosed by testing urine for quantitative organic acids. High urinary levels of glutaric acid or 3-hydroxyglutaric acid are diagnostic in affected children. If urine testing is abnormal, specific enzyme assays can be required to identify the precise enzymatic defect.

Menkes Disease

Menkes disease (MD) has also been reported as a potential mimic of AHT. ^, Menkes disease, often called Menkes kinky hair syndrome, is an X-linked recessive disease resulting from a mutation that codes a copper transport enzyme. The defect results in a systemic deficit of copper and subsequent global dysfunction of copper-dependent enzymes. This most notably results in poor collagen and elastin formation. Clinical hallmarks of MD are rapid and early neurological degeneration (within months of birth), poor growth, skeletal findings, and characteristic hair (pili torti). The hair in MD is short, friable, and twisted and has poor pigment, although fetal hair is unaffected. Blood vessels are tortuous and friable and are susceptible to rupture and bleeding with routine activities. This can result in intraabdominal or intracranial hemorrhage. ^, The skeletal manifestations include Wormian bones and metaphyseal defects (similar to the classic metaphyseal lesions associated with child physical abuse). Long bone metaphyseal findings can also resemble those found in scurvy. The combination of ICH and metaphyseal fractures can pose as findings very similar to those of AHT.

The most common ophthalmologic findings described in MD include poor visual acuity and decreased retinal and iris pigmentation. RHs have not been a described feature of “classic” MD and thus may be crucial in distinguishing this condition from AHT. Testing for MD can be done by simply microscopically examining scalp hairs for the characteristic pili torti. In addition, serum copper ceruloplasmin levels will be profoundly depressed. Infants with MD also have characteristic facies, including a high-arched palate, flat central face, and hypoplastic mandibles.