Diagnostic criteria for acquired hypothalamic obesity – international expert guidance document

Abstract

Acquired hypothalamic obesity (aHO) presents as rapid, clinically relevant, and persistent weight gain due to hypothalamic damage, and leads to significant morbidity/mortality and decreased quality of life. Causes include craniopharyngioma and other space-occupying lesions, neurosurgical intervention, irradiation, and traumatic brain injury. This review summarizes the evidence and provides expert opinion on diagnostic criteria for aHO. Eight experts in neuroendocrinology and neurosurgery from Japan and Europe participated in a multidisciplinary meeting at the 57^th Annual Meeting of the Japanese Society for Pediatric Endocrinology, Yokohama, Japan, 2024. Thereafter, three experts from Korea joined the discussion. Data were sourced from a search of the databases Web of Science, MEDLINE/PubMed, and Embase for reports published since 2000. Expert opinion of the authors was used substantially when no published data were available. The consensus on diagnostic criteria for aHO included: a.) traumatic event or (oncological) disease leading to hypothalamic lesions/damage detectable on magnetic resonance imaging (MRI); b.) rapid (occurring during the first 12 months after surgery/diagnosis), persisting (for 24 months after surgery), and clinically significant increase in body mass index (BMI) (≥5% BMI increase in adult; ≥1.0 standard deviation score (SDS) BMI increase in pediatric patients) starting during the first 12 months following the onset of hypothalamic damage under clinical and anthropometric monitoring at 3 months intervals; c.) obesity of a certain level (BMI SDS ≥+2.0 standard deviations (SD) in pediatric; BMI ≥25 kg/m² or BMI ≥30 kg/m² in adult patients), depending on racial and ethnic characteristics.

Graphical Abstract

Introduction

The hypothalamus is the key player in human metabolic balance, acting as the central integrator for endocrine, autonomic, and higher brain functions [1]. This delicate neuroendocrine organ, located in the suprasellar region of the brain, does not only regulate the pituitary gland via its releasing hormones, but also regulates hunger and satiety feelings [2], energy expenditure [3], temperature stability [4], salt and water balance [5], and circadian rhythms [6]. In addition, through its connective circuits, the hypothalamus also plays a central role in psychosocial behavior [7]. Hypothalamic dysfunction may lead to a variety of symptoms, from disorders of sleeping and satiety (hyperphagia and cachexia), to severe behavioral problems with morbid obesity and panhypopituitarism with arginine vasopressin deficiency [1, 8].

Causes of hypothalamic dysregulation in childhood may be monogenic variants and syndromic constellations, such as in pro-opiomelanocortin or leptin receptor deficiency, Prader–Willi syndrome and Bardet–Biedl syndrome [9], or acquired, such as in patients with suprasellar brain tumors (e.g., chiasmatic hypothalamic glioma, germ cell tumors, craniopharyngioma [CP]) [10-12]. Also, autoimmune endocrine diseases (e.g., hypophysitis or hypothalamitis) [13] or those of unknown etiologies (e.g., rapid-onset obesity with hypoventilation, hypothalamic, autonomic dysregulation, neural endocrine tumor (ROHHADNET syndrome)) [14] and septo-optic dysplasia [15] may result in hypothalamic dysregulation. All these entities are very rare with incidences ranging from 0.05 to 10 per 100,000 persons per year. Currently, published epidemiological estimates focus on tumor- and treatment-related acquired hypothalamic obesity (aHO). The most recent data - based on the analysis of 5.2 million claims data from Germany - estimate the incidence of tumor- and treatment-related aHO to 0.7–1.7 cases per million annually, corresponding to a prevalence of 1,260 patients [16]. In addition, hypothalamic dysfunction has also been reported to occur following neurosurgical interventions and cranial irradiation for brain tumors [17] and after traumatic brain injury [18]. Symptoms of hypothalamic dysfunction are often not adequately recognized, making delayed diagnosis very common [19]. This may cause further aggravation of hypothalamic damage or damage to surrounding structures, such as the optic nerve, or may cause aggravation of the consequences of hypothalamic damage, such as hyperglycemia and severe aHO [9, 20].

For these reasons, our goal was a.) to review published data on aHO since 2000 and b.) to reach an expert consensus on diagnostic criteria for aHO to help healthcare providers recognize signs and symptoms that may be present in patients with aHO. There are no published studies demonstrating that early treatment for instance with GLP-1 receptor agonists or early bariatric surgery is effective in this population. However, we would like to speculate that once efficient pharmaceutical agents for treatment of aHO become available, early diagnosis and initiation of therapy might have advantages in management of patients with aHO.

Definition and diagnosis of hypothalamic syndrome

Hypothalamic syndrome is an umbrella term describing multiple clinical symptoms and complaints that frequently occur in patients with damage to or dysfunction of the hypothalamic structures [1, 21]. Hypothalamic syndrome is not synonymous with aHO, but surveys in patients and caregivers have shown that aHO is the major contributor to hypothalamic syndrome. By self-assessment, patients with CP who developed aHO reported ‘most severe impairment of daily life and quality of life’ [22, 23]. In another analysis of 212 patients with CP, high body mass index (BMI) negatively correlated with functional capacity as a measure of quality of life [24].

Hypothalamic syndrome consists of several clinical symptoms, and its clinical manifestation may differ between patients [12]. van Iersel et al. described five clinical domains for the definition and diagnosis of hypothalamic syndrome: eating disorders, behavioral disorders, sleep disorders, temperature dysregulation, and endocrine dysfunction [21, 25]. Eating disorders include hyper- or hypophagia, clinically presenting as extensive hunger, binge-eating, overweight, obesity or failure to thrive, and underweight (which is more prone to appear at the time of CP diagnosis and before surgical treatment) [25, 26]. Behavioral disorders include obsessive compulsive symptoms, rage, and hoarding. Sleep disorders can include sleep apnea, hyposomnia, daytime sleepiness, and fatigue [25]. Temperature regulation disorders include hypothermia, hyperthermia, and temperature dysregulation (i.e., cold or warm hands, feet, and/or face at unusual moments) [25]. Central precocious puberty, delay in pubertal development, and hormonal deficits are endocrine dysfunctions described for hypothalamic syndrome [25, 27]. Longitudinally, patients with hypothalamic syndrome and obesity face an increased risk of cardiovascular and metabolic disorders. All the above-mentioned domains of hypothalamic syndrome clearly affect health-related quality of life in patients with CP [28, 29].

Proposed diagnostic criteria for hypothalamic dysfunction have been recently published [25]. The spectrum of diagnostic clinical symptoms is subdivided into hyperphagia, hypophagia, BMI, behavioral problems, sleep disorders, temperature-regulation disorders, pituitary dysfunction, gradations of hypothalamic involvement/damage [21], and presence or suspicion of a hypothalamic genetic syndrome. A diagnostic score for hypothalamic syndrome was developed. Using this new hypothalamic scoring system, 52.5% (63/120) of pediatric patients with CP and risk for hypothalamic dysfunction were scored as having hypothalamic syndrome at last follow-up; 76.7% (92/120) were diagnosed with pituitary dysfunction, 40.0% (48/120) with sleep disorders, 14.2% (17/120) with temperature dysregulation, and 32.5% (39/120) with hyperphagia. Hyperphagia was diagnosed based on patient self-assessment of extensive hunger and binge-eating. The Dykens Hyperphagia questionnaire [30] was only used in rare cases to objectivate the presence of hyperphagia. For several criteria, including temperature dysregulation and behavioral and sleep disorders, clinical data were missing in more than 50% of cases [25]. It became also obvious that the diagnostic criteria for hypothalamic syndrome were not sufficient for the specific diagnosis of aHO.

Weight gain following hypothalamic damage or dysfunction

Physiologically, the homeostasis of satiety and weight development is maintained through anorexigenic (e.g., leptin, insulin) and orexigenic (e.g., ghrelin, neuropeptide Y) pathways [2] (Fig. 1). In patients with aHO, these pathways are impaired due to tumor- and/or treatment-related damage of the hypothalamic ventromedial/arcuate nuclei [2]. This can result in insulin and leptin resistance. An imbalance of appetite and satiety (i.e., inappropriate food intake) can be observed as a clinical consequence [21]. In patients with CP, parasympathetic dominance of the autonomic nervous system due to vagal hyperactivation can lead to aHO. The decreased sympathetic activity reduces total energy expenditure. Fat accumulation and hyperinsulinemia are a consequence of the compensatory parasympathetic dominance [21]. An aHO study with functional magnetic resonance imaging (MRI) and computerized task elucidated that treatment-related hypothalamic damage can lead to dysfunction in brain regions associated with food reward processing, leading to restrained eating behaviors [31]. Once morbid obesity levels are reached, weight reduction is difficult to achieve [21].

Fig. 1. Schematic overview of the human hypothalamus.

The hypothalamus consists of many hypothalamic nuclei, which are all highly connected through neural pathways. The connection between the arcuate nucleus and paraventricular nucleus is emphasized. Afferent and efferent blood vessels provide a pipeline for pituitary hormones, as well as hunger and satiety hormones that stimulate hypothalamic neuron orexigenic and anorexigenic responses, respectively. ARC, arcuate nucleus; CRH, corticotropin-releasing hormone; DM, dorsomedial hypothalamic nucleus; PA, preoptic area; PH, posterior hypothalamic nucleus; SC, suprachiasmatic nucleus; SO, supraoptic nucleus; VMH, ventromedial hypothalamus (reproduced from reference [21] van Iersel et al., 2019 Endocr Rev, with kind permission of Illustration Presentation ENDOCRINE SOCIETY).

Approximately 50% of patients with CP and hypothalamic lesions develop aHO [32-36]. Analyzing anthropometric data before and after a diagnosis of CP in 90 patients, Müller et al. showed that BMI increased slightly before the diagnosis of CP and the major weight gain resulting in the development of obesity occurred during the first 12 months following diagnosis/surgery [27]. Other studies also showed that the major and clinically most significant increase in BMI (mean BMI standard deviation score (SDS) increase: +2 standard deviation (SD)) occurred during the first 12 months following neurosurgical intervention that resulted in hypothalamic lesions [33, 37, 38]. Patients with CP and damage to the posterior hypothalamic structures due to neurosurgical interventions develop a mean BMI increase of +3.2 SD during the 36 months following neurosurgery, which approximates a 50–80% weight gain [39] (Fig. 2). During long-term follow-up, BMI SDS of patients with CP and aHO stabilized at a high plateau (BMI SDS increase of approximately +5.0 SD) without a trend towards further BMI increase [40]. Bogusz et al. indicated that, in patients with CP initially involving anterior and posterior hypothalamic structures, hypothalamus-sparing surgical interventions respecting the integrity of posterior hypothalamic nuclei result in best outcomes regarding postoperative weight gain; patients with CP with severe involvement and surgical damage of posterior hypothalamic areas presented with BMI +5.21 SD at 12 months after surgery, whereas patients without hypothalamic lesions showed a BMI +1.72 SD [37]. Rovani et al. described the risk factors associated with the long-term development of obesity that included female sex, hypothalamic involvement, and a BMI SDS >+2.0 SD at baseline [36]. Long-term sequelae of aHO were also associated with more radical surgical treatment approaches, performed more frequently in centers with lower patient load, indicating that therapeutic expertise might be a relevant prognostic factor [39, 41]. Beckhaus et al. suggested that familial disposition for obesity (maternal and paternal BMI at CP diagnosis) is associated with the development of morbid aHO in patients with CP [42].

Fig. 2. Hypothalamic lesions (HL) in craniopharyngioma: postoperative grading on MRI and outcome/body mass index (BMI).

Definition of different grades of hypothalamic lesions based on postsurgical magnetic resonance imaging (MRI) as suggested by Puget et al. [89], de Vile et al. [90], and Müller et al. [39]. Outcome (BMI) SDS according to de Vile et al. [90] in terms of weight gain and the development of hypothalamic obesity with regard to grade of hypothalamic lesion. Median BMI and interquartile ranges are shown for BMI SDS. White color indicates absence of undamaged tissue or anatomical structure.

Aside from reports on aHO following CP, literature on aHO due to other diseases such as traumatic brain injury [43-46] are scarce, and mainly limited to case reports. To date, no “one-size-fits-all” therapeutic solution is available in these clinically complicated and challenging situations [2]. Current approaches include prevention of hypothalamic damage via hypothalamus-sparing surgical strategies, novel irradiation techniques (e.g., proton beam therapy), and novel pharmacological agents for management of aHO [47, 48].

Consensus on diagnostic criteria for acquired hypothalamic obesity

Eight experts in neuroendocrinology and neurosurgery from Japan and Europe participated in a multidisciplinary meeting at the 57^th Annual Meeting of the Japanese Society for Pediatric Endocrinology, Yokohama, Japan, October 12^th, 2024. The goal of the meeting was to summarize the evidence and provide expert opinion on diagnostic criteria for aHO, and especially the validity of BMI changes as a diagnostic parameter for aHO. Thereafter, three experts from Korea joined the discussion.

Beforehand, a narrative review was performed after search of the Web of Science, MEDLINE/PubMed, and Embase, databases for initial identification of articles. The search terms craniopharyngioma, obesity, hypothalamic obesity, quality of life, and hypothalamus were used. Date of last search was January 2025. Selected English language papers published since 2020 were included in this narrative review. Notably, some studies used the same population cohort and were included only once in this review. The study with the largest sample size was prioritized for inclusion (Fig. 3).

Fig. 3. PRISMA flow diagram of the literature review.

All unreferenced recommendations are based on amalgamated expert opinion. A writing group, headed by H.L.M., was formed and tasked with the authorship of this document, which is intended to inform decisions and planning by local teams, but not to be prescriptive.

Because of limited published data on aHO, only low-quality evidence from 24 smaller series (Table 1) [48-72] and 13 case reports (Table 2) [73-84] was available to support the recommendations made. Diagnostic criteria for aHO reported in these series were hypothalamic lesions in 17 of 18 evaluable studies, rapid postoperative weight gain of various degree (weight gain 4.7–52 kg/year; >75% weight gain; BMI increase +2.0 to +11.5 SDS) in 11 of 21 studies, and a BMI above normal (BMI ≥+2 SDS) in all studies. In case reports, hypothalamic lesions were a diagnostic criterion in 11 of 11 evaluable case reports, postoperative weight gain in 4 of 11 reports, and hyperphagia in 8 of 11 case reports.

Table 1. Studies of acquired hypothalamic obesity and treatment in humans.

Etiology	#	Age in years (range), (mean age ± SD)	Definition of HO	BMI/weight (range) at intervention	Intervention	Weight change during/after intervention	Authors
Craniopharyngioma	5	6.0, 8.2, 8.5, 9.0, 9.8	HL, weight gain >75% during 10 months after surgery	Mean BMI ± SD: 32 kg/m2 ± 2.8	Dextroamphetamine	ΔBMI –3 to +0.3 kg/m2	Mason et al. [49]
	16	26 ± 12	HL based on MRI and/or surgical report	Mean BMI ± SD: 46 kg/m2 ± 8	RYGB (n = 12), SG (n = 4)	Mean Δweight: –22% after 5 years	van Santen et al. [50]
	23	35 (25–43)	HL, BMI >40 kg/m2	Mean BMI: 44.2 (40.7–51.0) kg/m2; 125.3 kg	SG (39%), RYGB (61%)	ΔTWL (%) –39.0% (14.0–53.3)	Faucher P et al. [51]
	5	38 (27–47)	HL, uncontrolled weight gain and hypopituitarism	Mean BMI: 41.3 (37.9–46.3) kg/m2; 123.9 kg	SG (n = 2), RYGB (n = 3)	ΔTWL (%) –14.7% (23.7–5.8)	Garrez et al. [52]
	21	24 (12–54)	HL	Mean BMI: 49.6 kg/m2	LAGB (n = 6); SG (n = 8); RYGB (n = 6); BPD (n = 1)	TWL (%) LAGB: 10.5%; SG: 20.7%; RYGB: 20.2%; BPD: 24.8%	Bretault et al. [53]
	4	13, 17, 21, 23	HL, weight gain since diagnosis: +5.3 to +11.5 SD	BMI SDS: +7.3, +10.3, +11.4, +13.9 SD	LAGB (n = 4)	ΔBMI +1.7 to +8.7 kg/m2	Müller et al. [54, 55]
	4	22, 44, 57, 69	Continuous increase in body weight after surgery, despite adhering to a weight loss program	Mean BMI: 48.0 (35.0–55.5) kg/m2	Semaglutide	ΔBMI 7.9 BMI (range: 6.7–10.1) weight loss of 17.0% (11.3–22.4%)	Gjersdal et al. [56]
	9	15.4 ± 2.9	HL; BMI gain >2 SD during 12 months after surgery; >6 months diet + standard exercise	BMI SDS: +1.8 SD to +2.96 SD	Diazoxide and metformin	Mean ΔBMI –0.3 ± 2.3 kg/m2	Hamilton et al. [57]
	9	17 (12–30)	HL; annual weight gain: 15 (5–52) kg/year; BMI: >40 kg/m2	Mean BMI: 44.7 (40.2–61.6) kg/m2	LAGB (n = 6); SG (n = 4); RYGB (n = 2)	LAGB (n = 6): No change; SG (n = 4): No change; RYGB (n = 2): mean Δweight –30%	Weismann et al. [58]
	8	33.4 ± 13.6	HL; BMI >36 kg/m2	Mean BMI ± SD: 43.3 kg/m2 ± 4.1	SG (n = 3); RYGB (n = 5)	SG (n = 3): mean Δweight –10%; RYGB (n = 5): mean Δweight –25%	Wijnen et al. [59]
	3	21, 22, 24	Eating disorders, hyperphagia	Mean BMI: 49.2 (41.6–58.1) kg/m2	SG	Mean ΔBMI –13.9 kg/m2; Δweight –17.6%, –25.0%, –41.1%	Trotta et al. [60]
Craniopharyngioma + other diagnosis	7	0.5, 11.1, 11.8, 12.5, 14.7, 14.8, 21.0	BMI >90th percentile following various CNS insults	BMI SDS: +3.17 ± 0.9 (+1.9 to +4.4)	Dextroamphetamine	Mean BMI SDS decelerated to –0.18 ± 0.12/year during the 1st year of treatment and stabilized at +0.05 ± 0.32/year during the 2nd year of treatment.	Denzer et al. [61]
	19	12.3 ± 4.0	HL on MRI	Mean BMI ± SD: 3.58 kg/m2 ± 0.85	Dextroamphetamine	ΔBMI SDS –0.14 (11 declined, 3 stabilized)	van Schaik et al. [62]
	26	52 (18–65)	HL, BMI ≥30 kg/m2	Mean BMI: 38 (28–58) kg/m2	Semaglutide	Mean weight loss of 13.4 kg (95% CI 10.3–16.5 kg)	Svendstrup et al. [63]
	9	46 (22–49)	HL on MRI	Mean BMI ± SD: 37.6 kg/m2 ± 7.2	Exenatide/Liraglutide	Exenatide: ΔBMI –6.1 to –2.8 kg/m2; Liraglutide: Δweight –22 to –9 kg	Zoicas et al. [64]
	8	27.5 ± 7.8	HL, BMI >30 kg/m2	Mean BMI ± SD: 47.5 kg/m2 ± 10.8	Exenatide	Mean Δweight –1.4 kg	Lomenick et al. [65]
	18	45.4 ± 13.3	HL, BMI ≥27 kg/m2 hypothalamic injury-related weight gain	Mean BMI ± SD: 37.3 kg/m2 ± 5.6	Tesomet (tesofensine + metoprolol)	Mean additional Δweight change –6.3% (Tesomet –6.6% vs. placebo –0.3%); % of patients achieving >5% weight reduction Tesomet 61.5% vs. placebo 12.5%	Huynh et al. [66]
	18	15.0 ± 5.3	HL	Mean BMI ± SD: 38.0 kg/m2 ± 6.5	Setmelanotide	Mean ΔBMI: –15% (SDS 10%) after 4 months Extension 12 months (12 patients): –26% (12 SDS)	Roth et al. [48]
Hypothalamic tumor ± surgery	8	10–18	HL, annual weight gain >2 SD for age	Mean BMI ± SD: 36.0 kg/m2 ± 2.5	Octreotide	Mean Δweight –6.0 ± 0.7 kg Mean ΔBMI –2.1 ± 0.3 kg/m2	Lustig et al. [67]
	10	13.8 ± 1.2	Weight gain/year >2 SD above the mean for age + at least one endocrinopathy	Mean BMI ± SD: 37.1 ± 1.3 kg/m2	Octreotide (RCT)	Mean ΔBMI –0.2 ± 0.2 kg/m2 (vs. placebo +2.2 ± 0.5 kg/m2)	Lustig et al. [68]
	3	24, 10, 12	HL, weight gain: 4.7–8.3 kg/year for 3–4 years	BMI: 25.2, 25.7, 28.2 kg/m2	Triiodothyronine	ΔBMI range –6.4 to –5 kg/m2	Fernandes et al. [69]
	3	15, 16, 25	HL	BMI: 27.7, 34.7, 41.4 kg/m2	Caffeine + ephedrine	Δweight range –18.8% to –8%	Greenway et al. [70]
	39	13.4 (4.3–18.2)	n. a.	Mean BMI SDS: 1.93 (0–3.2)	Lifestyle modification	Median ΔBMI rate +4.5 kg/m2/year (range –17.8 to +8.4); median ΔBMI SDS rate 0.0/year (–5.2 to +0.5)	Rakhshani et al. [71]
	4	24, 30, 43, 51	HL, hyperphagia, weight gain	BMI: 37.6, 37.7, 43.7, 51.0 kg/m2	SG (n = 2); RYGB (n = 2)	SG (n = 2): ΔBMI –10 and –3.6 kg/m2; RYGB (n = 2): ΔBMI –6.2 and +11.3 kg/m2	Gatta et al. [72]

Abbreviations: #, cohort size; BMI, body mass index; BPD, biliopancreatic diversion; CNS, central nervous system; HL, hypothalamic lesion; HO, hypothalamic obesity; LAGB, laparoscopic adjusted gastric banding; MRI, magnetic resonance imaging; n. a., data not available; RCT, randomized controlled trial; RYGB, Roux-en-Y gastric bypass; SDS, standard deviation score; SG, sleeve gastrectomy.

Table 2. Single case reports on acquired hypothalamic obesity and treatment in humans.

Etiology	Age (years)	Definition of HO	BMI/weight at intervention	Intervention	Weight change during/after intervention	Authors
Craniopharyngioma	n. a.	HL	BMI: 51.9 kg/m2, +5.3 SDS	Triiodothyronine (T3)	No change	van Santen et al. [73]
	19	HL, hyperphagia	BMI: 52.9 kg/m2	Bilateral nucleus accumbens deep brain stimulation	ΔBMI –5.2 kg/m2	Harat et al. [74]
	13	HL, hyperphagia	BMI: +1.77 SDS	OXT and naltrexone	OXT: ΔBMI SDS –0.28; OXT + naltrexone: ΔBMI SDS –0.67	Hsu et al. [75]
	19	HL to VMH, hyperphagia, weight gain	BMI: 43.0 kg/m2	Truncal vagotomy	Δweight –7.0 kg	Smith et al. [76]
	n. a.	HL, eating disorder	BMI: >60 kg/m2	Octreotide and RYGB + truncal vagotomy	Octreotide: Δweight no change RYGB: Δweight –49 kg	Inge et al. [77]
	29	HL, eating disorder	BMI: 52.0 kg/m2	Distal gastric bypass surgery	ΔBMI –16.6 kg/m2	Schultes et al. [78]
	25	HL; eating disorder	BMI: 51.6 kg/m2	RYGB	ΔBMI –12.6 kg/m2 (–24%)	Page-Wilson et al. [79]
	12.7	HL, weight gain 20 kg in 6 months after surgery	BMI: 65 kg/m2	RYGB	ΔBMI –22.0 kg/m2	Rottembourg et al. [80]
	16.0	HL, weight gain 11 kg in 6 months after surgery	BMI: 45 kg/m2	Biliopancreatic diversion with duodenal switch procedure	ΔBMI –13.0 kg/m2	Rottembourg et al. [80]
	14	HL, rapid weight gain, AVP-D without thirst feeling	BMI: 47.0 kg/m2 Weight: 109 kg	PSMF: 567 kcal/day, 27 g fat, 71 g protein, 10 g carbohydrates, supplementation of vitamins and minerals	Weight decreased to 71.0 kg (from 109 kg) after 9 months of PSMF treatment	Lee et al. [81]
Optic glioma	36	HL, hyperphagia, weight gain 84 to 141 kg in 3 years	Weight: 141 kg	Fluoxetine and fenfluramine	Fluoxetine: Δweight +3.0 kg Fenfluramine: Δweight 0 kg	Jordaan et al. [82]
Hypothalamic tumor and XRT	9	HL, hyperphagia, weight gain	BMI: 65 kg/m2	Exenatide and liraglutide	Δweight –10 kg (6.7%)	Thondam et al. [83]
Germ cell tumor	17	HL	BMI: 37.1 kg/m2	Exenatide	ΔBMI –8.0 kg/m2	Simmons et al. [84]

Abbreviations: AVP-D, arginine vasopressin deficiency; BMI, body mass index; HL, hypothalamic lesion; HO, hypothalamic obesity; OXT, oxytocin; PSMF, protein-sparing modified fast; RYGB, Roux-en-Y gastric bypass; SDS, Standard deviation score; VMH, ventromedial hypothalamus; XRT, irradiation.

Higher BMI at diagnosis has been observed in patients with CP, who developed long-term obesity [33] and in patients with hypothalamic involvement [26] when compared with patients who kept normal weight or presented without presurgical hypothalamic involvement. Muller et al. [26, 27] observed that the BMI SDS at CP diagnosis was statistically different (p = 0.002) between patients with hypothalamic involvement (n = 40) and patients without hypothalamic involvement (n = 47). However, the differences (BMI at diagnosis of CP without HI: 0.3 (–3.7–3.9); BMI at diagnosis of CP with HI: 1.1 (–1.9–9.1)) demonstrate that BMI at the time of CP diagnosis is not a practicable clinical marker. Most BMIs in both groups are in the normal range with a broad overlap. Other reported findings [33] are similar: BMI SDS at CP diagnosis in long-term normal weight patients (n = 72; 0.1 (–3.7–3.8)) and in patients who developed severe obesity (n = 78; 1.5 (–1.8–7.6)). Statistically significant difference (p < 0.001), but no practicable clinical marker.

Surgical damage of the anterior and posterior parts of the hypothalamus including the mammillary bodies are relevant risk factors for the development of severe aHO [35, 36]. In our consensus, we chose the assessment of general hypothalamic involvement/damage without differentiation of anterior and posterior lesions as a criterion for aHO, because patients with isolated anterior hypothalamic damage also developed severe obesity [35]. Furthermore, assessment of hypothalamic damage in publications always refers to reference-confirmed imaging analyses by an experienced neuroradiologist. Such reference-assessment of imaging is not practicable in daily routine.

Considering that not all patients with hypothalamic lesions/damage develop obesity, a consensus on the diagnostic criteria for aHO was reached as below:

• Traumatic event or (oncological) disease leading to hypothalamic lesions/damage detectable on MRI.

• Rapid (occurring during the first 12 months after surgery/diagnosis), persisting (for 24 months after surgery), and clinically significant increase in BMI (at least ≥5% BMI increase in adult patients; ≥1.0 SDS BMI increase in paediatric patients) starting during the first 12 months following onset of hypothalamic damage under clinical and anthropometric monitoring at 3 months intervals.

• Obesity of a certain level (BMI SDS ≥+2.0 SD in pediatric patients; BMI ≥25 kg/m² or ≥30 kg/m² in adult patients, depending on racial and ethnic characteristics) (Graphical Abstract).

Graphical Abstract.

Although abnormal MRI findings are included as part of the above-mentioned diagnostic criteria, there may be patients in whom abnormalities cannot be detected on MRI. In individuals with obesity along with other signs of hypothalamic dysfunction and/or hypopituitarism, the possibility of hypothalamic obesity should be considered even in the absence of abnormal MRI findings.

In cases of hypothalamic damage due to radiation therapy, the onset of weight gain may occur in a delayed manner rather than showing the typical pattern of rapid onset within 12 months after injury.

This list of diagnostic criteria for aHO should be reevaluated after 2–3 years, and the newly diagnosed cases should be examined to see if the diagnostic criteria are working or need updating.

For the minimum diagnostic criteria, there should be obesity of a certain level and observation of a hypothalamic abnormality on diagnostic imaging. As for East Asian adults, BMI ≥25 kg/m² is more appropriate as criteria of obesity.

Supporting parameters could be hormone deficiency or decrease of serum concentration of at least one pituitary hormone, and body temperature dysregulation (a rough criteria as this is difficult to measure) [4]. Hyperphagia could also be included as supporting parameters, although these are harder to define, and it is not a symptom for all patients (although they may still experience weight gain) [85]. Attention should also be paid to major weight gain starting after damage occurs to the hypothalamus (i.e., the onset of weight gain should be considered, rather than the onset of obesity).

A list of exclusion criteria is also needed, as there are many causes of rapid weight gain and obesity; lifestyle changes, steroid treatments, antidepressant medication etc. should be considered.

Patients with aHO often experience depressive symptoms; psychologists should be aware that weight gain in these patients could be related to aHO and is not necessarily linked to the antidepressive medication. Collaboration with psychologists is needed to better understand both the currently available instruments, and how to differentiate between true depression and aHO-related depressive symptoms.

Treatment of aHO is challenging [8, 86-88] and should be performed by experienced multidisciplinary teams. aHO should be diagnosed and treated only in centers with expertise in managing this condition. Collaborative research efforts are needed to confirm diagnostic criteria and improve treatment approaches in aHO.

Conclusions

Recognizing aHO may be difficult due to its rareness and variety of symptoms. Standardized diagnostic criteria are missing. Establishing diagnostic criteria for signs and symptoms of aHO may aid in understanding the etiology of the disease and in predicting the course of hypothalamic obesity. In addition, they may support early recognition, diagnoses and treatment of the condition. Furthermore, a diagnostic score for aHO based on defined and standardized diagnostic criteria may be useful for research.

Conflict of interest

H.L.M. has received reimbursement of participation fees for scientific meetings and continuing medical education events, reimbursement of travel expenses and lecture honoraria from Rhythm Pharmaceuticals. T.T. received remuneration for consultancy services to Rhythm Pharmaceuticals and reimbursement of participation fees for scientific meetings from Rhythm Pharmaceuticals. H.A. has received lecture honoraria and reimbursement of participation fees for scientific meetings from Rhythm Pharmaceuticals. T.I. has received speaker’s fees and reimbursement of participation fees for scientific meetings from Novo Nordisk, Pfizer, Roche Diagnostics, JCR pharmaceuticals, and Rhythm Pharmaceuticals. T.I. has been a principal investigator of the ongoing RM-493-040 study conducted by Rhythm Pharmaceuticals. J.M. has received lecture honoraria from Novo Nordisk, JCR Pharmaceuticals, Pfizer, Kyowa Kirin. T.H., M.K., H.N., H.J.C., A.R.H., and J.H.K. have no conflicts of interest to be disclosed.