Natural history of ROHHAD syndrome: development of severe insulin resistance and fatty liver disease over time

A previously healthy, normally developing girl without familial obesity was evaluated at age 4 years for a sudden weight gain of 15 pounds in less than 1 year (25th to 95th percentile for weight) and decreased linear growth (25th to 10th percentile for height; Fig. 1b). She displayed no features of lipodystrophy. The medical history of the patient revealed that she was born to non-consanguineous healthy parents with a birth weight of 4 pounds 13 oz (2.14 Kg) by c-section at 37 weeks (3 weeks preterm) due to pre-eclampsia. Her growth in the first year of life included a height at the 75th percentile and weight at the 25th percentile. She never had any concerns about developmental delays or cognitive impairment. However, her parents noted fear and anxiety about routine daily activities, entering crowded spaces as well as decision making, starting around age 4 years. Ophthalmic exam did not show any vision abnormalities other than myopia and astigmatism. Given her mother’s history of hearing impairment, hearing test was done but did not show any abnormality.

She was originally started on subcutaneous DDAVP, and then transitioned to intranasal spray. She was very closely monitored with weekly blood draws, tight fluid instructions and close weight tracking. Despite this tight monitoring, she still required hospitalizations for hypernatremia at the time of infections. Given these difficulties, she never tried oral DDAVP. She was simultaneously treated with levothyroxine replacement.

After age 10, she was diagnosed with nonalcoholic fatty liver disease (NAFLD) with evidence of progressive fibrosis over time, resulting in hypersplenism and pancytopenia. Her liver function tests (LFTs) were noted to be elevated starting at age 10. At age 11, she underwent a liver US showing the presence of increased fat as well as splenomegaly. During the same period, she was noted to have autonomic dysfunction, abnormal thermoregulation, and profuse sweating.

She was treated with growth hormone from age 10 to age 14 years to assist with linear growth and it was discontinued due to increased swelling and weight gain. At age 10, she was diagnosed with sleep apnea and insulin resistance. She was started on metformin, but developed clinical diabetes by age 14. She did not achieve spontaneous menarche despite the development of oily skin and acne and was started on female hormone replacement with cyclic estrogen and progesterone therapy at the age of 14.

At age 15, she developed respiratory failure and shock from an upper respiratory tract infection, requiring intubation, oscillator, and vasopressors. She was able to recover after a prolonged hospitalization. Shortly after at age 16, she underwent a thorough evaluation at the Lurie Children’s Hospital that documented her autonomic dysfunction and hypoventilation (Additional file 1: Table S1). Management with overnight ventilator, oxygen supplementation and oscillator use were recommended. Home continuous overnight monitoring for oxygen saturation and heart rate were also initiated. At that time, she underwent PHOX2B gene testing due to the presence of autonomic dysfunction and this testing was negative.

Between the ages of 18–23, she was transitioned to adult endocrine care where her diabetes control was noted to become more challenging. She was started on insulin, with insulin requirement increasing to > 310 international units (IU) while her HbA1c increased to numbers above 9%. Poor diabetes control made management of sodium balance very difficult due to polyuria and absence of thirst. She developed severe acanthosis nigricans and lymphedema. Her weight increased to BMI over 40 kg/m2. Over time, she was started on U500 insulin. She also developed hypertriglyceridemia at this time with the highest value at 1062 mg/dL. Of note, given the hypertriglyceridemia, low adiponectin levels and the lack of a commercial test, she was never tested for presence of the insulin receptor auto-antibody.

She was started on liraglutide 0.6 mg daily with the hope of improving control and sparing insulin to improve acanthosis nigricans, hyperandrogenism, weight and fatty liver. Her insulin requirements dropped precipitously to less than 30% of baseline doses within 8 to 12 h of initiation (Fig. 1d). She was managed with liraglutide 1.2 mg daily on and off for about 2 years. Initiation of liraglutide did not cause worsening of her diabetes insipidus, hypernatremia, or pancytopenia. She did, however, experience worsened diarrhea due to the use of liraglutide leading to treatment interruption. Her acanthosis nigricans, facial acne and coarsening of facial features improved due to dose reduction in insulin.

With increased weight gain, swelling, worsening acanthosis nigricans, a weight reduction diet was attempted with meal replacement and caloric restriction. She lost 15 lbs. with this attempt. At the age of 26, she was noted to have a 1.6 cm Barcelona Clinic Liver Cancer (BCLC) Stage 0 (T1N0M0 Stage 1) liver lesion on routine semi-annual liver surveillance because of the presence of clinical cirrhosis and hypersplenism. The biopsy of the lesion favored the diagnosis of a well differentiated hepatocellular carcinoma (Fig. 2). A radiofrequency ablation procedure for this was successfully completed on 8/30/2018.

In addition, other comorbidities that developed over the years include cholelithiasis, chronic kidney disease with nephrotic range proteinuria (no biopsy performed to date to determine the underlying pathological mechanisms but being contemplated), hypertension, anemia and thrombocytopenia, splenomegaly, irritable bowel syndrome, vitamin D deficiency, and mild tricuspid regurgitation.

Our patient is one of the oldest described with ROHHAD who has developed severe insulin resistance while under observation. This case expands the spectrum of clinical findings in ROHHAD, to include severe insulin resistance progressing to diabetes mellitus with complications, progressive NAFLD with hypersplenism, and cirrhosis as well as hepatocellular carcinoma.

ROHHAD syndrome is a relatively novel disease. It was first described by Fishman et al. in 1965 [8], but the syndrome acquired the name in 2007 by Ize-Ludlow et al. when the diagnostic criteria were described [9]. To date, more than 100 cases (n = 117) based on our review) have been reported in the literature [5]. Recently Lee et al. reported a systematic review of the reported cases independent of our efforts, but this review did not focus on insulin resistance, dyslipidemia or fatty liver disease. In fact, the summary tables found in that review do not comment specifically on any of these novel features [10]. The syndrome is quite distinctive from Prader-Willi and other complex syndromes of obesity due to its sudden onset acquired nature, absence of symptoms at birth, predominance of autonomic features, and respiratory abnormalities as well as diabetes insipidus and hyperprolactinemia.

The condition is characterized by high morbidity and mortality rates [3, 11]. Forty per cent of patients were noted to develop neural crest tumors (ganglioneuroma, ganglioneuroblastoma); it was suggested that the acronym ROHHAD could be extended to ROHHADNET (ROHHAD-neuroendocrine tumors) [11]. Previously recognized features of the syndrome include obesity, alveolar hypoventilation, autonomic dysregulation, sympathetic nervous system tumors. While at this point, we cannot rule out the existence of two separate disorders in our patient leading to a blended phenotype, we suggest that severe insulin resistance and progressive fatty liver disease may be features to the syndrome. We also want to draw attention to this case as a cause for the development of atypical diabetes. Evaluation of patients with ROHHAD shoudl include evaluation for these potential complications and with time, we may learn the frequency of these features in ROHHAD. 

The etiology of the syndrome is not known. Molecular diagnostic studies including PHOX2B gene testing and whole exome interrogation in conjunction with her parents did not reveal any causative genetic alterations in our patient including monogenic or syndromic obesity or known syndromes of insulin resistance (including the insulin receptor). In 2011, Patwari et al. reported monozygotic twins, one of whom had ROHHAD syndrome while the other twin did not, [12] suggesting that the disease may not develop due to genomic changes. Even though there is no definitive understanding, it is postulated that ROHHAD may be an immune-mediated central nervous system (CNS) disorder [6, 13–15] or epigenetic factors may have a role in the pathogenesis of the syndrome [12]. In this regard, it was interesting that our patient did not demonstrate any laboratory features suggesting other autoimmune diseases at any point. Periodically performed imaging studies did not reveal any neural crest tumors to date either.

Other than the well described pituitary hormonal deficiencies, diabetes insipidus and hyperprolactinemia, other hormonal features of this syndrome are not described. Our patient had a high circulating leptin level. We are not sure if this can be generalized to other ROHHAD patients. We did not study the incretin secretion or other circulating mediators of appetite control or energy balance in our patient and are not aware of these investigations in other patients.

Our case presented several challenges and lessons to us that are worth sharing. She was prepubertal when she presented with the earliest signs of insulin resistance. She was treated with metformin almost immediately, but this did not prevent the development of severe acanthosis nigricans, oily skin, hirsutism and coarsening of features, despite the presence of central hypogonadism, suggesting that insulin resistance alone can drive these aspects of clinical presentation without any central stimulation of the pituitary. She was treated with cyclic estrogen and progesterone replacement for bone health. Our patient was also treated with growth hormone for several years to assist with linear growth, but this was kept short due to worsening weight gain and swelling. Management of her case from childhood through early adulthood remained quite complicated and she required frequent lab monitoring and clinical visits. Both her parents and herself remained very vigilant about her care. Long term survivors with this syndrome need to be closely monitored and care transitions from pediatric to adult endocrinology should be undertaken in a planned manner.

Insulin resistance and adult growth hormone deficiency are associated with non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatohepatosis (NASH); the latter can progress to liver cirrhosis and hepatocellular carcinoma in some cases [16]. High doses of insulin, growth hormone, and even estrogen may have also contributed to the development of hepatocellular cancer [17]. In addition, prolactin receptors are sometimes expressed in hepatocellular cancers [18] and our patient has had mild, but untreated hyperprolactinemia since the onset of her condition. Our patient’s hyperprolactinemia was never adequately addressed; dopamine agonists were never initiated due to the presence of autonomic dysfunction, and fear of worsening orthostasis. We recommend screening for hepatic lesions using abdominal ultrasound in other patients presenting with the clinical features of NAFLD in the setting of ROHHAD as they grow older, particularly if they have signs of advanced hepatic fibrosis or cirrhosis. In addition, presence of clinical diabetes and insulin resistance should also be screened in addition to various endocrine manifestations.

The immediate improvement in glycemic control following liraglutide initiation indicates that the insulin resistance seen in this patient can be partially overcome with glucagon-like peptide-1 (GLP-1) agonism and adds to the growing evidence that GLP-1 receptor agonists may exert their metabolic actions via neural circuitry located outside of the hypothalamus, or via peripheral mechanisms [19]. Treatment experience in a larger number of patients is needed to validate these early observations.

Finally, our patient’s sodium and fluid balance management were markedly tenuous on account of poorly controlled diabetes, presence of diabetes insipidus and absence of thirst regulation. The patient required weekly sodium monitoring and constant adjustment of her fluid intake recommendation and her DDAVP dose. Inter-current illnesses (such as upper respiratory tract, urinary tract infections, or cellulitis episodes) and natural temperature fluctuations impacted her sodium levels. Our patient had multiple hospitalizations for dehydration and hypernatremia despite being compliant with instructions and carrying out frequent laboratory testing. Ideal glucose control (also a challenging goal for this patient) was essential to maintain her sodium balance closer to the normal range.

ROHHAD is an interesting syndrome presenting with a characteristic cluster of abnormalities. There appears to be heterogeneity among patients diagnosed with ROHHAD and careful clinical characterization is required to more precisely define this disease entity. Our case expands on the clinical features of this syndrome to include severe and progressive insulin resistance and fatty liver disease. It should be noted that the fatty liver disease in this condition can be aggressive and result in development of cirrhosis that can be complicated by hepatocellular carcinoma.