2. Case Report
The patient’s parent provided consent for publication of the following clinical data. The patient was a female of Ecuadorian heritage. She was spontaneously conceived and subsequently delivered at a community hospital in New Jersey at 36 2/7 weeks gestation to a 31-year-old gravida 2 para 2 via cesarean section. The pregnancy was complicated by a failed oral glucose tolerance test, though it was unclear from records if gestational diabetes management was initiated. Her Apgar scores were 9 and 9 at 1 and 5 min of life, respectively. Her weight and length were both in the 75th percentile for gestational age. Shortly after birth on her first day of life (DOL), hypoglycemia was noted with blood glucose measuring 12 mg/dL. Her hypoglycemia persisted and was managed with oral, then intravenous dextrose. Endocrinology was consulted and diagnostic laboratory testing demonstrated hypoketotic hypoglycemia with an inappropriately high insulin level given hypoglycemia (6.4 µIU/L in the setting of blood glucose < 50 mg/dL), low β-hydroxybutyrate (1.4 mmol/L; reference < 2.0 mmol/L) and a positive response to glucagon stimulation test, consistent with a diagnosis of hyperinsulinism. On DOL 14, diazoxide therapy of 10 mg/kg/day was initiated. On DOL 20, care was transferred to our center in Pennsylvania for additional evaluation and management of her hyperinsulinism. At that time, her diazoxide dose was 15 mg/kg/day and she required a continuous glucose infusion rate of 5 mg/kg/min.
With regards to NBS, the infant had a routine dried blood spot sent to the New Jersey state NBS laboratory on DOL 2 (Table 1
), with normal results reported 2 days following sample receipt. This sample was analyzed for all conditions on that state’s NBS panel, including TT1. At the time, the screening protocol for TT1 was comprised of tyrosine MS/MS measurement. The patient’s tyrosine levels on the first NBS were within acceptable range at 151 µmol/L (reference cut-off < 280 µmol/L). On DOL 6, 13, and 18, repeat dried blood spot samples were sent due to the patient’s prematurity and only analyzed for congenital hypothyroidism and congenital adrenal hyperplasia (Table 1
). Results from all screening were available to the patient’s care team at the time of her transfer to Pennsylvania. Following the patient’s TT1 diagnosis, her retained dried blood spot samples were retrospectively reanalyzed via MS/MS for both SUAC and tyrosine, with results shown in Table 1
. SUAC from the first sample, sent on DOL 2 and subsequently stored under temperature-controlled conditions, was elevated at 5.23 µmol/L. New Jersey did not employ SUAC quantitation for TT1 screening, therefore there was no validated reference threshold at the time of this analysis. On adoption of SUAC-based screening in January 2020, the threshold was < 0.5 µmol/L. Finally, the New Jersey NBS laboratory was not blinded to the patient diagnosis at the time of reanalysis.
On DOL 23, the patient developed progressive abdominal distention and irritability with feeds. Abdominal imaging revealed a large stool burden, which partially improved following a glycerin suppository. Her clinical examination was otherwise unremarkable, and discharge was planned for the following day. However, on DOL 24, routine electrolyte assessment revealed severe hyponatremia of 112 mmol/L (reference: 135–145 mmol/L). Her mother commented that the patient had been sleepier than usual, prompting physical examination. She was somnolent but arousable to tactile stimulation, with mild facial edema, respiratory distress evidenced by intermittent intercostal retractions, and continued abdominal distention. Intravenous normal saline was initiated for gradual sodium replacement. When she subsequently developed hypotension requiring initiation of more aggressive fluid resuscitation, she was transferred to the Neonatal Intensive Care Unit for further management.
Following fluid resuscitation, hepatomegaly was noted. An abdominal ultrasound was obtained and was concerning for hepatic and renal echogenicity abnormalities. However, her critical illness prevented better characterization of these abnormalities with additional imaging. Her α-1-fetoprotein level was elevated at 164,000 ng/mL (reference: 0.6–77.0 ng/mL). Although her hypotension stabilized with multiple vasopressors, the infant’s course was further complicated by the development of fluid overload, pulmonary hypertension, acute kidney injury, and disseminated intravascular coagulation with bleeding from multiple sites. A head ultrasound revealed bilateral subdural hematomas. Her respiratory failure required intubation, and worsening ascites necessitated peritoneal drain placement.
The patient’s acute decompensation and liver failure lead her intensive care team to consider an inborn error of metabolism and the Metabolism team was consulted on DOL 26. Oliguria with significant gross hematuria complicated collection of urine for organic acids and SUAC quantitation. Meanwhile, hypotension, need for blood replacement products, and need for other laboratory evaluation important in stabilizing the patient delayed collection of the blood sample for plasma amino acids. Plasma amino acid testing run in-house was obtained on DOL 30 and reported on DOL 31. It was concerning for elevated plasma tyrosine (586.4 µmol/L; reference 22–102 µmol/L; 10.6 mg%) and phenylalanine (369.6 µmol/L; reference 23–95 µmol/L; 6.1 mg%). However, interpretation of plasma amino acid values was complicated by a generalized amino academia, consistent with known liver and renal dysfunction.
Diagnosis of TT1 was made when semi-quantitative urine organic acid analysis revealed high levels of tyrosine metabolites 4-hydroxyphenylacetate (2037 mg/g creatinine) and 4-hydroxyphenylpyruvate (161 mg/g creatinine) as well as the presence of SUAC (48 mg/g creatinine; reference undetectable). Protein was removed from her parenteral nutrition, and instead she received intravenous dextrose-containing fluids and lipids. NTBC pharmacotherapy (1 mg/kg/day) was initiated the following day (DOL 32). Soon after, rapid whole exome sequencing, obtained in the setting of multi-organ failure of unknown etiology, confirmed the diagnosis with a novel homozygous pathogenic nonsense mutation in FAH
(c.1014 delC, p.Cys 338 Ter). With initiation of NTBC, the infant’s SUAC level normalized to undetectable on urine organic acid measurement within 3 days. Plasma SUAC measured via blood spot was within normal limits (1.2 nmol/mL; reference < 3 nmol/mL) 3 weeks following NTBC initiation and not quantifiable 2 weeks after that. As selective elimination of tyrosine and phenylalanine from parenteral nutrition was not possible, complete amino acids were gradually reintroduced to her nutrition on DOL 33. Adjustments were made to her nutritional management based on the sum of tyrosine and phenylalanine content as tyrosine levels rose, expected in the setting of NTBC (Figure 1)
. Variations in plasma NTBC without changes in weight-based dosing were the result of challenges in weight-based dosing for a patient with profound ascites and fluid shifts who is also presumably undergoing some degree of growth. Routine ophthalmologic exams, initiated given risk for tyrosine corneal deposits, have remained normal to date.
Gastroenterology was consulted to consider the utility of hepatic transplantation in this critically ill infant. Magnetic resonance imaging of the patient’s liver obtained once the patient had stabilized on DOL 68 revealed hepatomegaly and a nodular liver consistent with cirrhosis with multiple regenerative nodules but no masses suspicious for HCC (Figure 2
). With stabilization, α-1-fetoprotein levels declined to 76,759 ng/mL 13 weeks post-diagnosis. As her respiratory status and fluid balance improved, the patient was extubated to non-invasive positive pressure ventilation on DOL 45. Despite adequate metabolic control of her TT1 with low phenylalanine and tyrosine levels, and NTBC levels within therapeutic range (Figure 1
), her liver synthetic function failed to recover. The patient continued to struggle with ascites requiring continued peritoneal drainage, coagulopathy requiring frequent cryoprecipitate transfusions, thrombocytopenia, hyperbilirubinemia, and mild hyperammonemia. As a result, evaluation for hepatic transplant was initiated and the infant received an orthotopic LT from a deceased donor at 5 months of age. Her post-operative course was complicated by respiratory failure, temperature instability concerning for sepsis, and pancytopenia. She remained hospitalized in a critical care setting at the time of this report’s publication at 6 months of age due to ventilation needs. At that time, laboratory studies demonstrated normal liver parenchymal function, including coagulation studies, albumin, bilirubin, aspartate transaminase, alanine transaminase, and γ-glutamyl transferase. She has never been discharged from the hospital.
This case underscores the importance of adherence to the NBS consensus statements for inclusion of SUAC as a primary marker for hereditary TT1 [16
]. Screening for TT1 using tyrosine levels alone is known to result in false-negative cases, with at least two others in the literature [29
]. Interestingly, reanalysis of this patient’s repeated newborn dried blood spot samples (Table 1
) demonstrated tyrosine levels that surpassed the screening threshold by DOL 6 yet did not continue to rise. However, those dried blood spots, sent after the initial comprehensive NBS sample, were only analyzed for endocrine abnormalities and not amino acids. Although SUAC was consistently detectable on retrospective reanalysis (Table 1
), its quantitation was not part of the New Jersey NBS protocol at that time.
Capturing TT1 patients during their pre-symptomatic phase and instituting early pharmacologic and dietary interventions is the goal of NBS [16
]. There is evidence that this improves survival [13
] and cognitive outcomes [21
] while reducing the need for LT [11
]. Although the retrospective nature of this case prevents certainty in what this infant’s outcome might have been in the case of a NBS-facilitated TT1 diagnosis, it remains likely that morbidity could have been prevented in the avoidance of an acute decompensation. The Quebec NTBC Study showed that during 5731 months of NTBC treatment, there were no hospitalizations for acute TT1 complications [13
], suggesting that earlier initiation of NTBC could have prevented this infant’s multiorgan failure. Her false-negative NBS also delayed diagnosis; once TT1 was suspected, the patient’s critical illness delayed obtaining appropriate diagnostic samples.
Since the identification of this patient, beginning in mid-January 2020, New Jersey has implemented a SUAC-based TT1 screening program. Three states (Maryland, Oklahoma, and West Virginia) and many countries around the world do not utilize SUAC as a primary screening marker for TT1. With an incidence of the disease in the general population of about 1:100,000 [3
], such policies likely result in rare but missed opportunities to attenuate morbidity and mortality for patients and their families. Furthermore, advances in molecular and genetic technologies over the past two decades may present new and exciting potential therapies for TT1, including enzyme replacement therapy, gene therapy, and genome editing [34
]. These options, while not yet a reality, may mean that babies born and diagnosed through NBS today may have therapeutic modalities available to them that continue to improve outcomes, making early, accurate, pre-symptomatic recognition of the disease more important.
Here, we report an infant with molecularly and biochemically verified TT1 whose NBS was reported by the reported by the state’s NBS laboratory as “normal.” “False negative” results are results in which the condition being tested is present, but not detected. False negatives can occur for analytic reasons (e.g., the analyte is present, but not detected), but can also occur when the analyte under study is not sensitive for the condition. In this case, the New Jersey NBS correctly reported a normal tyrosine level, and one may argue that to practitioners who are knowledgeable about the negative predictive value of tyrosine as a sole analyte for the diagnosis of TT1, this is the expected result. While this is true, such information regarding such limitations is not disseminated to healthcare providers outside of the metabolism specialty. This is problematic: a screen reported as normal will have results sent to an infant’s primary care provider/team, who may not fully understand the methodology of the test. If no distinction is made on the report between use of tyrosine and SUAC for TT1 screening, practitioners in catchment areas that include multiple states employing multiple NBS modalities may be tempted to consider universal screening for TT1 across the United States as equivalent. Labeling the case as a false negative reminds the practitioner of the need for vigilance, in verbiage familiar to clinical practice.
It is unrealistic to expect non-metabolic physicians to be aware of the sometimes-nuanced differences between NBS methodologies and diagnostic yields that can exist both between municipalities and between different screening laboratories. For these clinicians, the case underscores the importance of early consultation with a metabolic specialist rather than reliance on a negative NBS result. For metabolic specialists, particularly those in multi-state catchment areas, it underscores the importance of familiarity with methodologic screening differences.
Although SUAC quantitation is an important part of screening for TT1 with high sensitivity [33
], it is imperative to note that no screening tool can be perfect. Indeed, Blackburn et al. reported a family with a homozygous variant in FAH
affecting the catalytic pocket of the enzyme without affecting tyrosine or SUAC levels [35
]. Three affected individuals presented in infancy with hepatosplenomegaly and cirrhosis, progressing to HCC in childhood. The family was ultimately diagnosed via whole exome sequencing, after urine organic acid levels were normal and SUAC levels were undetectable. One sibling had tyrosine levels within normal range, and another had levels only mildly elevated (212 nmol/mg Cr, where the upper limit of normal was 208 nmol/mg Cr). Thus, while a SUAC-based screening program likely would have avoided morbidity for our patient given the high SUAC levels present on reanalysis of her filter paper (Table 1
), it cannot do so for all patients. As a result, maintaining familiarity with the TT1 phenotype and suspicion for inborn errors of metabolism in infants with liver pathology, remain important tools in any clinician’s toolbox.
Finally, hyperinsulinism can be an early symptom of TT1. Children with TT1 presenting in the neonatal period commonly have hypoglycemia [8
] and pancreatic islet cell hyperplasia has been identified in a number of patients [36
]. In one retrospective review of 25 infants, three met clinical criteria for hyperinsulinism with hypoglycemia and inappropriately elevated insulin or C-peptide [38
]. The patients either presented in liver failure (2/3) or were diagnosed pre-symptomatically via NBS [38
]. They were successfully treated with diazoxide and chlorothiazide [38
]. With resolution of their hypoglycemia, they were able to be weaned off of treatment, suggesting a transient nature to their hyperinsulinism [38
]. However, the described patients carried a known TT1 diagnosis and were also treated with NTBC [38
], making the time course of this feature as a function of disease versus a function of treatment unclear. To our knowledge, this is the first patient whose initial presenting symptom of TT1 was hyperinsulinemic hypoglycemia, suggesting it occurs as a product of the disease rather than NTBC treatment. It is thus important to recognize that hyperinsulinism can be a feature of TT1, particularly in infants presenting with liver dysfunction and hypoglycemia.