Adrenoleukodystrophy in the Differential Diagnosis of Boys Presenting with Primary Adrenal Insufficiency without Adrenal Antibodies

Adrenoleukodystrophy (ALD) is an X-linked, metabolic disorder caused by deficiency of peroxisomal ALD protein resulting in accumulation of very-long chain fatty acids (VLCFA), primarily in the adrenal cortex and central nervous system. Approximately 35-40% of boys with ALD develop cerebral ALD (CALD), which causes rapidly progressive cerebral demyelination, loss of neurologic function, and death. Approximately 70-80% of boys with ALD have impaired adrenal function prior to the onset of neurologic symptoms. We present a boy who had recurrent episodes of hypoglycaemia from age two years and was diagnosed with adrenal insufficiency without adrenal antibodies at age 5.5 years. Following initial normal VLCFA levels, subsequent VLCFA analysis demonstrated elevated C26 fatty acids consistent with peroxisomal dysfunction and suggestive of ALD, which was confirmed via molecular genetic analysis of the ABCD1 gene. Brain imaging at age 7 suggested cerebral involvement and the child underwent successful allogeneic hematopoietic stem cell transplantation. At last assessment (11.5 years old), he was performing as expected for age. This case highlights the importance of pursuing a diagnosis when clinical suspicion remains, and the significance of VLCFA analysis for patients with adrenal insufficiency without adrenal antibodies in securing an ALD diagnosis. Subsequent brain imaging surveillance can detect early, pre-symptomatic cerebral disease, allowing for timely treatment and successful arrest of cerebral disease progression.

What is already known on this topic?

The cerebral adrenoleukodystrophy form (CALD) of ALD is characterized by progressive inflammatory demyelination resulting in loss of neurological function and early death. Approximately 70-80% of boys have impaired adrenal function at the time of ALD diagnosis.

What this study adds?

Despite appropriate screening, some ALD cases may still be missed when relying on very-long chain fatty acid (VLCFA) levels alone, since some mutations may result in initial normal levels of VLCFA. The need for early CALD diagnosis in order to initiate timely treatment mandates awareness of the potential need for repetitive VLCFA screening of boys with unexplained, antibody-negative, primary adrenal insufficiency.

Introduction

Adrenoleukodystrophy (ALD) is a rare, X-linked, metabolic disorder caused by mutations in the ABCD1 gene that result in deficiency of peroxisomal membrane ALD protein and accumulation of very-long chain fatty acids (VLCFA) in tissues and plasma (1,2). The most severe phenotype is the cerebral form (CALD), which develops in 35-40% of at-risk male children typically between the ages of 3 and 12 years (1). CALD rapidly progresses with inflammatory cerebral demyelination, loss of neurological function, and early death.

ALD was added to the recommended uniform screening panel in the United States in 2016, but newborn screening (NBS) remains limited, both within the US and worldwide (3). Therefore, early identification of ALD remains challenging due to nonspecific initial symptoms that overlap with other medical and developmental issues. Mutations in the ABCD1 gene are associated with genetic forms of primary adrenal insufficiency (PAI) (4,5,6); therefore PAI in boys in the absence of adrenal antibodies can be an early “red flag” for a potential ALD diagnosis. Unexplained PAI may be the only clinical sign prior to the onset of neurologic symptoms in boys ultimately diagnosed with CALD (7,8).

Early detection of CALD is critical since allogeneic hematopoietic stem cell transplantation (HSCT) has the potential to stabilize cerebral disease progression, but is only effective if performed in the early stages of cerebral disease (9,10). While not possible to determine which children with ALD will develop CALD based on genotype, plasma VLCFA levels, or family history (11,12), monitoring at-risk boys for magnetic resonance imaging (MRI) abnormalities can successfully detect changes indicative of CALD prior to the onset of neurological symptoms (13,14). HSCT is only indicated once imaging detects white matter changes indicative of progression to CALD, and current management guidelines recommend repeating MRIs in pre-symptomatic boys with ALD every six months between the ages of 3 and 12 years (15). Given this need for early ALD diagnosis and monitoring for CALD, ALD should be included in the differential diagnosis of boys presenting with PAI without adrenal antibodies.

The following case presentation highlights the clinical course in a boy who was diagnosed with ALD through presentation with PAI and who, upon progression to CALD, received a timely and successful allogeneic stem cell transplant.

Case Report

A white male infant was born at term to non-consanguineous parents. There were no antenatal or perinatal concerns and early childhood was reported to be normal with normal growth parameters. No hyperpigmentation was ever noted. At 2 years of age he was admitted to a paediatric resuscitation bay via the local GP walk-in service following a 24 hour vomiting illness with subsequent collapse, convulsion and fever with acidosis (pH 7.279, base deficit 9.2 mmol/L), cardiac compensation (pulse 161/minute with mean peripheral blood pressure 84 mm Hg) requiring 20 mL/kg saline bolus, and hypoglycaemia (glucose 1.1 mmol/L; 19.82 mg/dL) which responded rapidly to 5 mL/kg 10% dextrose. He also had elevated urine ketones (+++ on urine dipstick test), and his plasma cortisol measured after normalisation of glucose, was 320 nmol/L. Bedside sodium was 131 mmol/L and potassium 3.2 mmol/L. He recovered promptly with rehydration and dextrose infusions with intravenous antibiotics. He was discharged 48 hours later after a full infection screen, including lumbar puncture, was negative, and was tolerating oral food and fluids, with a presumptive diagnosis of hypoglycaemia following prolonged vomiting with a febrile convulsion.

At 2.5 years, he had an episode of multiple febrile convulsions, and again presented with hypoglycaemia (glucose 1.2 mmol/L; 21.62 mg/dL; sodium 137 mmol/L; potassium 3.8 mmol/L) and elevated urine ketones (+++ on urine dipstick test). Plasma insulin (46 pmol/L) and C-peptide (315 pmol/L) were found to be inappropriately elevated at the time of hypoglycaemia. Plasma cortisol was 470 nmol/L. The hypoglycaemic episodes were initially ascribed to idiopathic ketotic hypoglycaemia. In view of the multiple episodes of hypoglycaemia and hyperinsulinaemia during this most recent admission, investigations included a 24-hour glucose profile and controlled 17-hour fast, both of which did not demonstrate any hypoglycaemia but normal suppression of insulin and mobilisation of ketone bodies at the end of the fast. Random plasma cortisol concentrations were 281 and 226 nmol/L at 12:00 and 18:00 hours respectively.

A subsequent hypoglycaemic event (glucose 1.4 mmol/L; sodium 139 mmol/L; potassium 3.9 mmol/L) with vomiting at age 5.2 years was associated with a plasma cortisol of 553 nmol/L. The supervising clinician changed at this time, and because of recurrent unexplained hypoglycaemia, a standard synacthen (tetracosactide) test was performed and showed adrenal insufficiency [Table 1 cortisol and adrenocorticotropin (ACTH) test 1]. A repeat synacthen test was performed at age 5.5 years in view of the relatively high cortisol at the time of the last hypoglycaemia event and confirmed these findings (Table 1, cortisol and ACTH test 2). Adrenal antibodies were negative, and the boy was commenced on standard hydrocortisone replacement at 10.0 mg/m2/day in divided doses.

Table 1

Results of tests evaluating cortisol, adrenocorticotropin, and very-long chain fatty acid levels

In view of the adrenal failure and negative adrenal antibodies, an analysis of VLCFA levels was ordered (using UKAS accredited laboratories), although due to family reasons there was a delay of several months in performance of these tests. VLCFA testing typically assesses abnormalities in three parameters: the level of hexacosanoic acid (C26:0), and the ratio of hexacosanoic acid to tetracosanoic acid (C26:0/C24:0) and to docosanoic acid (C26:0/C22:0) (2). Initial VLCFA tests at age 6.5 years showed normal C26/C22 and C24/C22 levels (Table 1, VLCFA Laboratory A test 1). Continued clinical suspicion and negative adrenal antibodies prompted repeat VLCFA tests using a different laboratory at 7 and 7.5 years of age (Table 1, Laboratory B, test 2 and 3) that showed elevated C26 fatty acids with moderately raised C26/C22 ratio, mildly raised C24/C22 ratio, and normal phytanic and pristanic acid levels (pristanic and phytanic acids are branched chain fatty acids that undergo oxidation in peroxisomes, and are typically elevated in peroxisome biogenesis disorders in the Zellweger spectrum, but not in ALD). These findings were consistent with a disturbance in peroxisomal function suggestive of ALD rather than a peroxisomal biogenesis disorder. Subsequent molecular genetic analysis of the ABCD1 gene identified a previously reported pathogenic hemizygous mutation (c.1849C>T, [p.Arg617Cys]), which on family testing was confirmed to be maternally inherited. The child was commenced on Lorenzo’s oil (oleic acid and erucic acid 4:1, such that 20% of energy requirements in diet came from Lorenzo’s oil) available for use in UK with provision for monitoring on a non-research basis, and a low fat diet at 7.5 years of age following the second confirmatory VLCFA assessment.

At 7.5 years of age, baseline MRI brain scan identified features possibly consistent with cerebral ALD (Figure 1), with signal abnormality in the splenium of the corpus callosum and subtle changes in the deep parietal white matter, but no convincing gadolinium enhancement (an indicator of active inflammation and hallmark of cerebral disease). The MRI Loes score, a 34-point scale used to measure the extent of demyelinating brain lesions (16), was assessed as 2 with bilateral splenium of corpus callosum and bilateral parietal white matter central change. A repeat MRI after three months showed some progression of the lesion in the splenium, with patchy signal change in the parietal lobes and mild prominence of cerebellar folia and sulci. Loes score was now assessed as 3, with additional mild symmetrical cerebellar atrophy. Although the splenium lesion had progressed in size, this did not alter the Loes scoring of the already bilateral splenium lesion. Neurological examination at this stage was normal. The child had a history of temper tantrums but no other significant behavioural or psychological problems. In view of the MRI findings suggestive of progressing cerebral ALD, allogeneic HSCT was indicated.

Figure 1

Brain magnetic resonance imaging, T2-weighted axial images. Top row: pre-transplant, bottom row post-transplant. Interval of scan in years (years before [-] or after [+] transplant). Arrow indicates increased signal lesion in splenium of corpus callosum that progressively enlarged pre-transplant, further progressed at 1.1 years post-transplant, but subsequent stabilisation

At age eight years, he underwent 10/10 human leukocyte antigens-matched, unrelated, allogeneic HSCT with pre-transplant conditioning consisting of a combination of busulfan (1.9 mg/kg to a target cumulative area under the curve of 80 mg/L/hr), fludarabine (40 mg/m2 x 4 doses) and alemtuzumab (0.2 mg/kg x 5 doses). The total white cell dose was 3.68x108/kg with a CD34 count of 7.03x106/kg and CD3 count 0.36x108/kg. There were no significant peri-transplant complications. Transplant outcome was excellent with 100% engraftment in myeloid lineage and stable mixed chimerism in T-cells (82% donor). Brain MRI imaging following transplant showed initial progression of the lesion in the splenium at one year post-transplant, but subsequent stabilisation with no further progression through three years post-treatment (Figure 1). At this time the Loes score remained 3.

At last assessment at 11.5 years of age, the boy was attending mainstream secondary school where he was performing very well academically. He was on the 25th percentile for weight and 50th for height, and was being assessed for delayed puberty as possible late sequelae of either the transplant conditioning or primary testicular dysfunction due to ALD. He continued to receive hydrocortisone replacement but has not required mineralocorticoid replacement. Neurological examination revealed moderately brisk reflexes in the lower limbs but normal tone and power, and no gait disturbance.

Discussion

Recognition of PAI without adrenal antibodies in this boy with recurrent hypoglycaemic episodes prompted repeated analysis of VLCFAs and identification of the underlying ALD diagnosis before the onset of neurological symptoms. MRI surveillance detected early pre-symptomatic cerebral disease and permitted timely allogeneic HSCT, which was successful in arresting the progression of the cerebral disease. Glucocorticoid replacement requirements, stress-dosing guidelines, and monitoring for mineralocorticoid deficiency for patients with ALD are generally the same as those in other forms of PAI (17), and the patient continues on hydrocortisone replacement therapy without the need for mineralocorticoid supplementation currently. Mineralocorticoid deficiency is less common in ALD (7,18), in part due to accumulation of VLCFA preferentially in the zona fasciculata and zona reticularis, with a relative sparing of the zona glomerulosa.

Since there is no ability to predict which boys diagnosed with ALD will develop CALD, vigilant MRI monitoring must be implemented in order to detect brain changes indicative of progression to CALD and ensure early HSCT (15). Allogeneic HSCT has been shown to have a beneficial effect on clinical indices of disease and long-term survival, but outcomes are more favourable if transplant is performed in the early stages of cerebral disease (10,19,20).

PAI can result from genetic or acquired diseases that affect adrenal function. Adrenal insufficiency in ALD may arise as a result of abnormal VLCFA accumulation that alters the viscosity of adrenocortical cell membranes, and prevents the stimulatory effects of ACTH on adrenocortical cells by inhibiting ACTH receptor binding (21). Impaired activity of ACTH leads to primary atrophy of the adrenal cortex with resulting cortisol deficiency. ALD is estimated to account for up to 5% of unexplained PAI cases (5,6). Among boys with ALD, approximately 70-80% have impaired adrenal function at the time of diagnosis (7,8,18).

In retrospective analyses of boys with ALD, PAI was the presenting and only sign in 25-37% of patients (8,18). There were delays of between 1 and 10 years between diagnosis of PAI and diagnosis of ALD for approximately half of the boys in the study. Delays in ALD diagnosis were associated with higher Loes scores at the time of evaluation for HSCT, progression of disease after HSCT, and decreased survival (8). These data emphasize that all boys with unexplained PAI should be screened for ALD.

Published diagnostic guidelines for PAI include recommended VLCFA testing for males with PAI who are negative for 21-hydroxylase (21-OH) antibodies, and concurrent testing for adrenal antibodies and VLCFA through accredited laboratories in preadolescent boys (22). However, despite these guidelines and older cases or case series highlighting the need to screen children and adult males with unexplained PAI for ALD (23,24,25), a recent study highlights the continued need for education that PAI in the absence of adrenal antibodies should be a “red flag” for a potential ALD diagnosis. Age appropriate ranges for VLCFA need to be used. Furthermore, other confounding factors, such as diets rich in rapeseed or mustard seed oils, which can potentially be associated with false negative results, need to be considered when performing VLCFA tests. Results from a pilot educational program intended to encourage reflex VLCFA testing in cases of PAI with negative 21-OH antibodies suggest that gaps in reflex testing remain (26).

Although NBS programs will assist in early identification and surveillance of boys with ALD (27), the potential for delayed diagnosis persists where NBS is not available and where absence of family history precludes family screening.

Conclusion

This case demonstrates that some ALD cases may still be missed when relying on VLCFA levels alone, since some mutations may not only result in initial normal levels of VLCFA but varying levels of adrenal insufficiency requiring clinical vigilance and suspicion for a potential underlying VLCFA abnormality. The need for early diagnosis in order to initiate current and future treatments mandates pursuit of a clear diagnosis and awareness of the need to screen boys with unexplained PAI for ALD.