Loss of Pons-To-Hypothalamic White Matter Tracks in Brainstem Obesity

SHORT COMMUNICATION Loss of pons-to-hypothalamic white matter tracks in brainstem obesity

JQ Purnell1, DL Lahna2, MH Samuels1, WD Rooney3,4 and WF Hoffman2,4,5

Hyperphagia and obesity have been reported following damage to the hypothalamus in humans. Other brain sites are also postulated to be involved in the control of food intake and body weight regulation, such as the amygdala and brainstem. The brainstem, however, is thought to primarily integrate short-term meal-related signals but not affect long-term alterations in body weight, which is controlled by higher centers. The objective of this study was to identify structural pathways damaged in a patient with a brainstem cavernoma who experienced sudden onset of hyperphagia and >50 kg weight gain in o1 year following surgical drainage via a midline suboccipital craniotomy. Diffusion tensor imaging revealed loss of nerve fiber connections between her brainstem, hypothalamus and higher brain centers with preservation of motor tracks. Imaging and endocrine testing confirmed normal hypothalamic structure and function. Gastric bypass surgery restored normal appetite and body weight to baseline. This is the first report of ‘brainstem obesity’ and adds to the brain regions that can determine the long-term body weight set point in humans.

International Journal of Obesity (2014) 38, 1573–1577; doi:10.1038/ijo.2014.57 Keywords: brainstem; magnetic resonance imaging; diffusion tensor imaging; bariatric surgery

INTRODUCTION presentation, she was at her lifetime maximal weight and had Unwanted weight gain and obesity result from dysregulation of been weight stable for several years with no complaint of systems that control food intake and energy expenditure.1 hyperphagia. She was a one-pack-per-day smoker beginning at Damage to the ventromedial hypothalamus, typically from tumor age 18 years. She denied excessive alcohol intake or illicit drug growth or surgical trauma in humans, leads to a well-described use. On exam, she was alert and oriented with mild dysarthria, a obesity syndrome characterized by sudden onset of hyperphagia right sixth cranial nerve deficit, reduced sensory perception of her and rapid weight gain.2 Experimental hyperphagia and obesity right face and mild dysmetria with left finger-to-nose testing. Her can also be induced in rodents and primates through lesions of muscle strength and reflexes were normal. Brain magnetic the amygdala.3 The brainstem is also involved in the regulation of resonance imaging (MRI) showed a 5 × 1.7 × 2.3-cm heteroge- food intake, in part through receiving vagal input from satiety neous lesion in the right posterior pons consistent with degrading factors and ghrelin, as well as direct binding of gut-derived blood (Figure 1). The remainder of her brain scan, including the hormones to their receptors in this area.4 Studies of decerebrate hypothalamus, was normal. A cerebral angiogram revealed a rodents have shown that the isolated brainstem is sufficient for brainstem cavernous malformation and hemorrhage. determination of meal size5 but have not clarified the brainstem’s Two-weeks after presentation, she underwent cavernoma role in long-term control of body weight. In the present report, we resection by a midline suboccipital craniotomy. Post surgery she describe a woman who experienced a brainstem lesion at the level was documented to have a new right-peripheral seventh cranial of the pons that resulted in unregulated hyperphagia and obesity, nerve palsy with new mild left-sided weakness and, beginning in which subsequently resolved with restoration of baseline body the hospital during recovery, she reported a marked increase in weight following gastric bypass. To our knowledge, this is the first hunger and failure to feel completely full when she ate a meal. reported case of ‘brainstem obesity’ in a human. Following discharge, incessant food craving resulted in near continuous eating and rapid weight gain (Supplementary Figure 1). After 5 months, her weight had increased by 30 kg and she began CASE HISTORY to experience fatigue, irregular menses, heat intolerance, snoring A 30-year-old woman presented with 1-week history of progres- and daytime somnolence and depression. Because of her rapid sive numbness of her right face, left body weakness, a right frontal weight gain, she was referred to an endocrinologist. She had no headache and double vision. Her past medical history was symptoms or biochemical evidence of hypothyroidism or hypo- unremarkable except for being obese (body mass index 32 cortisolism. Pulse and blood pressure were normal and pupils kg m−2) and she was taking no medications. At the time of her were equal, round and reactive to light and accommodation.

1Department of Medicine and Center for the Study of Weight Regulation, Oregon Health & Science University, Portland, OR, USA; 2Department of Neurology, Oregon Health & Science University, Portland, OR, USA; 3The Advanced Imaging Research Center, Oregon Health & Science University, Portland, OR, USA; 4Department of Behavioral Neuroscience, Oregon Health & Science University, Portland, OR, USA and 5Mental Health and Clinical Neurosciences Division, Veterans Affairs Medical Center, Oregon Health & Science University, Portland, OR, USA. Correspondence: Dr JQ Purnell, Division of Endocrinology, Diabetes, and Clinical Nutrition, Oregon Health & Science University, Mailstop L607, 3181 SW Sam Jackson Park Road, Portland, OR 97239, USA. E-mail: [email protected] Received 5 August 2013; revised 11 March 2014; accepted 17 March 2014; accepted article preview online 14 April 2014; advance online publication, 6 May 2014 Brainstem obesity JQ Purnell et al 1574

Figure 1. Anatomical scans of patient in radiological convention. White arrows indicate the lesion. Upper T2-weighted scans (a–c) (axial, sagittal and coronal, respectively) taken before cavernoma resection show a 1.5 × 1.7 × 2.3-cm heterogeneous lesion in the right posterior pons that is hypointense centrally and hyperintense peripherally consistent with blood products in various stages of degradation. There is a small amount of associated edema in the adjacent structures and mass effect on the adjacent fourth ventricle. Hypothalamus and pituitary anatomy appear normal. Lower T1-weighted scans (d–f) (axial, sagittal and coronal, respectively) taken 6 months after cavernoma resection show hypointensity in the right posterior pons at site of previous cavernoma resection. Hypothalamus and pituitary anatomy appear normal.

A sleep study was negative for sleep apnea. Repeat brain MRI Diffusion tensor imaging (DTI) and tractography showed a posterior defect at the site of the hematoma drainage, Imaging data were acquired on a 3T Siemens TIM Trio MRI instrument ipsilateral hypertrophic olivary degeneration and normal hypotha- (Erlangen, Germany). A localizer scan was used to guide slice align- lamus, and pituitary anatomy. ment for subsequent scans. A high-resolution T1-weighted anatomical She was subsequently given trials of phentermine and magnetization-prepared rapid gradient echo (144 slices 1-mm thick, sibutramine without successful control of her appetite or weight repetition time/echo time/TI/α = 2300 ms/4.38 ms/1200 ms/12°, field of gain (Supplementary Figure 1). Within 1 year of her initial injury, view = 208 × 256 mm) was acquired for anatomic co-registration. A single she had gained over 40 kg and by 21 months after surgery she DTI scan was acquired that consisted of 6 non-diffusion weighted (B0) images followed by 30 non-collinear directions of diffusion weighted had gained 53.2 kg, reaching a maximal body mass index of α −2 images (72 axial slices, 2-mm thick, repetition time/echo time/ = 9300 ms/ 53.7 kg m . Owing to no response to medical therapy, unabated 90 ms/90°, field of view = 256 × 256 mm). appetite and weight gain, she was referred for bariatric surgery Diffusion weighted data were corrected for head movement and eddy and underwent Roux-n-Y gastric bypass 2 years after her initial currents using the Oxford Centre for FMRI of the Brain (FMRIB) Software presentation. Immediately following this surgery, she noted a Library (FSL)’s ‘eddy correct’ tool.6 Non-brain voxels were excluded using marked decline in her hunger and restoration of meal-related the ‘Brain Extraction Tool’ (FSL).6 Fractional anisotropy was calculated using satiety. Within 12 months, she had returned to her baseline weight the FSL tool ‘dtifit’ (part of the FMRIB Diffusion toolbox (FDT) v2.0). The magnetization-prepared rapid gradient echo was coregistered to the (Supplementary Figure 1). Six years after her initial presentation fi she felt well, remained weight stable at her usual weight, had average B0 image using a 12-parameter af ne registration algorithm from FSL’s ‘Linear Image Registration Tool’ (FLIRT). The hypothalamus and the stopped all prescription medications and agreed to return to amygdala were manually drawn as regions of interest (ROI) for each Oregon Health & Science University for retesting. All study subject by an experienced research assistant using the magnetization- procedures were approved by the institutional review boards at prepared rapid gradient echo as an anatomical guide. These ROIs were PVAMC and Oregon Health & Science University, and informed then imported into MedINRIA for deterministic tractography to allow for consent was completed before undergoing study procedures. inspection of all tracts passing through the hypothalamus and amygdala.7 The brainstem was also divided into four quadrant ROIs in axial slices directly inferior and superior to the pons. These ROIs were used for MATERIALS AND METHODS deterministic tractography to determine the trajectory of white matter Control subject tracts inferior and superior to the lesion. A single sex- and handedness-matched control subject was selected retrospectively from an imaging database of normal volunteers. At the RESULTS time of study, she was 26 years old, had a body mass index of 31 kg m− 2, and smoked one pack per day of cigarettes. She had no significant past or Routine laboratory studies performed on the patient at the same present medical illnesses, was not taking centrally acting medications and time as the diffusion tensor imaging, including a complete had no other substance abuse or dependence issues. chemistry panel, lipid panel, complete blood count and levels

International Journal of Obesity (2014) 1573 – 1577 © 2014 Macmillan Publishers Limited Brainstem obesity JQ Purnell et al 1575 of vitamin D, vitamin B12, iron and folate were all normal. Normal pituitary function was confirmed by baseline and dynamic testing (data not shown). Brain imaging showed a persistent pontine defect, hypertrophic olivary degeneration and no structural changes in the hypothalamus or pituitary (Figure 1). These findings were unchanged from MRI taken at the time of her post-cavernoma resection scan 4 years earlier (data not shown). Tractography utilized two ROIs to explore the effect of the brainstem damage on white matter tracks from this area. The first ROI was placed in the brainstem at the level of the pons where the defect is apparent on MRI. This ROI was divided into color-coded quadrants corresponding to anterior and posterior, right and left lateral regions (Figures 2a–d). In the images from the control patient (Figures 2a and c), white matter tracks can be seen extending posteriorly to the cerebellum and superiorly to the motor and other cortical regions. In the patient (Figures 2b and d), white matter tracks extend from these ROIs to the inferior cerebellum and motor cortex (red and green), but the posterior-connecting tracks (yellow and blue) to the hypothalamus are nearly absent. A second ROI was drawn for the hypothalamus. In the control subject (Figures 2e and g), prominent fiber tracts from the hypothalamus extend inferiorly to the brainstem and cerebellum and several tracts that extend to the cortex. On the other hand, in the patient, tracks from the hypothalamus to the cerebellum and cortex are preserved but a majority of tracts connecting the hypothalamus with the brainstem are absent as are connecting tracks to the premotor cortex area (Figures 2f and h). Tractography from the brainstem above the lesion was chosen as a control region (Supplementary Figure 2). These images show similar connections from these regions in the patient and control, demonstrating the isolated nature of the tract disruption between the brainstem and hypothalamus.

DISCUSSION The brain area best known for inducing obesity when damaged in both animal models and in humans is the ventromedial hypothalamus.2 Selective lesioning of the posterodorsal amygdala also induces hyperphagia and obesity in animal models3,8 and may contribute to increased food-seeking behavior following temporal lobectomy in humans.9 In animal models, the hind brain has also been identified as a key center in the regulation of food intake. The brainstem receives the initial input from sensory organs via the cranial nerves,10 contains abundant glucose- sensing neurons that alter food intake in response to fluctuations in blood glucose levels11 and receives afferent signals from meal- Figure 2. Deterministic fiber tractography of the brain of the patient related gut hormones via the vagus nerve, including those that 6 years after drainage and resection of a brainstem cavernoma (b, d, induce satiety, such as cholecystekinin,12 and others that elicit f, h) and of an age, sex and body mass index-matched control (a, c, orexigenic responses, such as ghrelin.13,14 In addition, the dorsal e, g). Regions of interest were drawn on the axial plane at the lower area postrema is a circumventricular brainstem region that readily pontine level to color four brainstem quadrants (anterior: red and green; posterior: blue and yellow) (a–d). White matter tracts binds blood-derived hormones, such as amylin, to influence 15 generated using a manually drawn hypothalamus region of interest feeding behavior. Neurological and humoral signals from the (in red) indicated by the red arrows (e–h). White arrows indicate brainstem are then conveyed to higher centers by one of two disrupted white matter tracts in the patient. The patient (f and h) major white matter tracks in humans: the dorsal longitudinal exhibits fewer white matter brainstem connections both to (b and fasciculus runs posteriorly and connects to the ventral hypotha- d) and from the hypothalamus (f and h) compared with the control lamus, whereas fibers of the median forebrain bundle extend (a, c, e, g). Panels c and d are posterior views, whereas g and h are through the lateral hypothalamus into ventromedial olfactory anterior views. areas.16 This patient’s initial presentation was not unlike other reports of brainstem lesions where mild sensory deficits, cranial nerve post-op MRI contains a number of pontine structures, including abnormalities or varying degrees of motor strength and coordina- the nuclei for the cranial nerves demonstrating defects on physical tion loss are common depending on the location and extent of the exam (sixth and seventh), the medial longitudinal fasiculus lesion.17 Although she was obese before her initial presentation, it carrying descending tectospinal tract and medial vestibulospinal was not until after she underwent surgical drainage through the tracts into the cervical spinal cord, the reticular formation and floor of the fourth ventricle 2 weeks after presentation that she nucleus raphe. experienced marked hyperphagia and a progressive 50-kg weight It is likely that fiber tracks running in the dorsal longitudinal gain. The anatomic location of the lesion demonstrated on fasciculus and median forebrain bundle, from the nucleus tractus

© 2014 Macmillan Publishers Limited International Journal of Obesity (2014) 1573 – 1577 Brainstem obesity JQ Purnell et al 1576 solitarius, the area postrema and the dorsal vagal nucleus,16 were melanocortin 4 receptor, a key hypothalamic regulator of injured during the surgery to drain the cavernoma. This is appetite.27–29 Taken together, these findings and the results from supported by the DTI images (Figure 2) showing selective dropout our patient support a model in which a diffuse neural network of fiber tracks between the brainstem, hypothalamus and what involving the brain and brainstem retains the capacity to respond appear to be projections to premotor cortex (Figures 2e vs f), with to appetitive signals even after irreversible damage occurs to one fi preservation of most of the motor long- ber tracks to the of these sites. cerebrum. Damage to these areas would potentially impair Several limitations of this report are important to acknowledge. anorectic meal-related signals mediated by amylin, cholecysteki- Differences in white matter track density between the case and nin or other gut hormones. On the basis of rodent literature, control subjects other than those between the hypothalamus and however, loss of these satiety signals should not be sufficient to brain stem are also apparent. For example, the case patient disrupt weight stability long term. For example, studies in appears to have more fibers extending from the hypothalamus decerebrate rodents demonstrate that the caudal brainstem neural circuits regulate the size of individual meals short term, ROI to the posterior cortex, but less in the direction of the but are insufficient to control meal size during long-term prefrontal cortex than the control patient (Figures 2f vs e). These perturbations of body weight.18 On the other hand, the immediate differences could represent alterations in secondary connections hyperphagia and rapid, sustained weight gain experienced by our from the hypothalamus after loss of large tract input from the patient is consistent with an animal model of obesity resulting brainstem, or disruption of median forebrain bundle tract fibers from a brainstem lesion first described by Ahlskog and Hoebel.19 traversing through the lateral hypothalamus to ventromedial brain Injecting 6-hydroxydopamine into the hindbrain of rodents regions. It is possible that loss of these brainstem-to-cortical destroys ascending catecholaminergic neurons of the ventral connections, or of their secondary connections, could be noradrenergic bundle (VNAB), which connect to the important contributors to the hyperphagia experienced by this 19,20 hypothalamus. These rodents exhibit increased food intake patient. On the other hand, these differences may simply 19 and sustained increases in body weight. In addition, VNAB represent normal variations in the density of suprahypothalamic lesions are additive to the obesity phenotype when the fi fi 20 ber projections between subjects. The signi cance, if any, of ventromedial hypothalamus is damaged, indicating that these differences in non-brainstem hypothalamic projections on the VNAB has independent effects on food intake and body food intake noted in our case patient, however, cannot be tested weight. Analogous catecholaminergic pathways in humans would or verified. Interestingly, some fiber track connections between be farther anterior than the extent of damage shown in Figure 1, the hypothalamus and the brainstem are detected in our case so would likely not explain the hyperphagia and progressive ’ weight gain in our patient. Recent reports have also identified the (Figures 2f and h). Because our patient did not manifest Horner s lateral parabrachial nucleus in the brainstem as an integrative syndrome (ptosis plus miosis, chemosis and ipsilateral anhidrosis), center that modulates feeding behavior21 and that inhibition of these may represent hypothalamospinal tracts that carry outgoing neurons in this brain area increased food intake in rodents.22 sympathetic fibers that were preserved because they run The lateral parabrachial nucleus is located a bit more rostral (closer anterolaterally in the brainstem. Finally, because our patient did to the junction of pons and the midbrain) and lateral to the lesion not undergo pre-surgical physiological studies, we could not in our patient.23 It receives input from gustatory sensory fibers determine the impact of the brainstem lesion or bariatric surgery traveling in the nucleus tractus solitaries and, in turn, sends on her metabolism, changes in levels of adipocyte or gut projections to the hypothalamus and amygdala,22,23 so it is hormones involved in weight regulation or changes in her possible that ascending tracks to this nucleus were damaged at hypothalamic-pituitary function. However, besides a slight drift the time of surgery. Although neither direct damage to the VNAB downward in her insulin-like growth factor-1 levels during and at or lateral parabrachial nucleus likely occurred in our patient, we the time of her DTI studies, all the rest of her hypothalamic- would instead suggest that these rodent models establish pituitary testing were normal, including her thyroid, cortisol and fi the plausibility that a selective brainstem lesion is suf cient to gonadal axis’, making a hypothalamic cause of her obesity induce hyperphagia and obesity through disruptions of ascending unlikely. neuronal hypothalamic circuits (Supplementary Figure 3). The DTI In summary, we describe the first reported case of unregulated findings in our patient suggest similar effects on appetite and appetite and severe obesity following a brainstem lesion. From the body weight can be elicited by a very selective lesion connecting imaging data, we propose that selective lesioning of connecting more dorsal pontine regions to the hypothalamus and higher fi brain centers in humans. bers running between the dorsal vagal complex, hypothalamus The resolution of her hyperphagia and restoration of her and higher brain centers, including the dorsal longitudinal fi baseline body weight following gastric bypass potentially offers fasciculus and median forebrain bundle, are suf cient for novel insight into the mechanism of weight loss after this bariatric disrupting long-term body weight regulation. Rodent studies procedure. The gastric bypass is known to alter secretion of the and the clinical course of our patient support the hypothesis that orexigenic gut hormone ghrelin and the anorexigenic gut disruption of ascending fiber tracts to the hypothalamus from the hormones glucagon-like peptide-1 and peptide YY in directions brainstem mediate these behavioral changes, but are not that favor sustained weight loss compared with dieting.24 These necessary for sustained-weight loss to occur after gastric bypass hormones have receptors on both the vagus nerve and in surgery. hypothalamic centers that regulate food intake.26 The clinical resolution of this patient’s excessive weight gain after the gastric bypass suggests that either the remaining connections between CONFLICT OF INTEREST the brainstem and hypothalamus were sufficient to mediate The authors declare no conflict of interest. signals to or from the hypothalamus, or brain centers other than the brainstem (such as the hypothalamus) are capable of receiving inputs from these signals and regulating both meal-related ACKNOWLEDGEMENTS hunger and long-term body weight in the human. Interestingly, We would like to acknowledge the assistance of Drs Madhavi Gaddam and Diana gastric bypass has also been found to be effective for weight loss Negreanu with the endocrine workup during their fellowship training period. This in a limited number of patients with hypothalamic obesity owing work was supported in part by pilot funds through the OHSU Advanced Imaging to craniopharyngioma and functional mutations in the Research Center and NIH grants R21 DK073729 and R56 DK 88207 (JQP).

International Journal of Obesity (2014) 1573 – 1577 © 2014 Macmillan Publishers Limited Brainstem obesity JQ Purnell et al 1577 REFERENCES 15 Boyle CN, Lutz TA. Amylinergic control of food intake in lean and obese rodents. 1 Morton GJ, Cummings DE, Baskin DG, Barsh GS, Schwartz MW. Central Physiology & behavior 2011; 105:129–137. nervous system control of food intake and body weight. Nature 2006; 443: 16 Netter FH. Ciba Pharmaceutical Products inc., CIBA-GEIGY Corporation. The Ciba 289–295. Collection of Medical IIlustrations: a Compilation of Pathological and Anatomical 2 Bray GA, Gallagher Jr TF. Manifestations of hypothalamic obesity in man: Paintings vol. 1. Ciba Pharmaceutical Products: Summit: NJ, USA, 1959. a comprehensive investigation of eight patients and a reveiw of the literature. 17 Chen X, Weigel D, Ganslandt O, Buchfelder M, Nimsky C. Diffusion tensor imaging Medicine (Baltimore) 1975; 54: 301–330. and white matter tractography in patients with brainstem lesions. Acta Neurochir 3 King BM. Amygdaloid lesion-induced obesity: relation to sexual behavior, (Wien) 2007; 149: 1117–1131. olfaction, and the ventromedial hypothalamus. Am J Physiol Regul Integr Comp 18 Grill HJ. Distributed neural control of energy balance: contributions from hind- 14 – Physiol 2006; 291: R1201–R1214. brain and hypothalamus. Obesity 2006; : 216S 221S. 4 Wren AM, Bloom SR. Gut hormones and appetite control. Gastroenterology 2007; 19 Ahlskog JE, Hoebel BG. Overeating and obesity from damage to a noradrenergic 132: 2116–2130. system in the brain. Science 1973; 182:166–169. 5 Grill HJ, Kaplan JM. Interoceptive and integrative contributions of forebrain and 20 Ahlskog JE, Randall PK, Hoebel BG. Hupothalamic hyperphagia: dissociation from brainstem to energy balance control. Int J Obes Relat Metab Disord 2001; 25: hyperphagia following destruction of noradrenergic neurons. Science 1975; 190: S73–S77. 399–401. 6 Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, 21 Wu Q, Clark MS, Palmiter RD. Deciphering a neuronal circuit that mediates Johansen-Berg H et al. Advances in functional and structural MR image analysis appetite. Nature 2012; 483:594–597. and implementation as FSL. NeuroImage 2004; 23:S208–S219. 22 Carter ME, Soden ME, Zweifel LS, Palmiter RD. Genetic identification of a neural 7 Toussaint N, Souplet JC, Fillard P. MedINRIA: medical image navigation and circuit that suppresses appetite. Nature 2013; 503:111–114. research tool by INRIA In: Proc of MICCAI'07 Workshop on Interaction in Medical 23 Naidich TP, Duvernoy HM. Duvernoy's atlas of the human brain stem and cere- Image Analysis and Visualization. Asclepios: Brisbane, Australia, 2007. bellum: high-field MRI: surface anatomy, internal structure, vascularization and 3D 8 Dicks D, Myers RE, Kling A. Uncus and amygdala lesions: effects on social behavior sectional anatomy. Springer: Wien; NY, USA, 2009. in the free-ranging rhesus monkey. Science 1968; 165:69–71. 24 Cummings DE, Weigle DS, Frayo RS, Breen PA, Ma MK, Dellinger EP et al. 9 Terzian H, Ore GD. Syndrome of Kluver and Bucy; reproduced in man by bilateral Plasma ghrelin levels after diet-induced weight loss or gastric bypass surgery. removal of the temporal lobes. Neurology 1955; 5:373–380. N Engl J Med 2002; 346: 1623–1630. 10 Netter, FH. Ciba Pharmaceutical Products inc., CIBA-GEIGY Corporation. The Ciba 25 Olivan B, Teixeira J, Bose M, Bawa B, Chang T, Summe H et al. Effect of weight loss Collection of Medical Illustrations: a Compilation of Pathological and Anatomical by diet or gastric bypass surgery on peptide YY3-36 levels. Ann Surg 2009; 249: Paintings. Ciba Pharmaceutical Products: Summit: NJ, USA, 1959. 948–953. 11 Ritter S, Dinh TT, Zhang Y. Localization of hindbrain glucoreceptive sites 26 Neary MT, Batterham RL. Gut hormones: implications for the treatment of obesity. controlling food intake and blood glucose. Brain Res 2000; 856:37–47. Pharmacol Ther 2009; 124:44–56. 12 Smith GP, Jerome C, Cushin BJ, Eterno R, Simansky KJ. Abdominal vagotomy 27 Aslan IR, Campos GM, Calton MA, Evans DS, Merriman RB, Vaisse C. Weight loss blocks the satiety effect of cholecystokinin in the rat. Science 1981; 213: after Roux-en-Y gastric bypass in obese patients heterozygous for MC4R 1036–1037. mutations. Obes Surg 2011; 21:930–934. 13 Asakawa A, Inui A, Kaga T, Yuzuriha H, Nagata T, Ueno N et al. Ghrelin is an 28 Aslan IR, Ranadive SA, Ersoy BA, Rogers SJ, Lustig RH, Vaisse C. Bariatric surgery in appetite-stimulatory signal from stomach with structural resemblance to motilin. a patient with complete MC4R deficiency. Int J Obes (Lond) 2011; 35:457–461. Gastroenterology 2001; 120:337–345. 29 Bretault M, Boillot A, Muzard L, Poitou C, Oppert JM, Barsamian C et al. Clinical 14 le Roux CW, Neary NM, Halsey TJ, Small CJ, Martinez-Isla AM, Ghatei MA et al. review: bariatric surgery following treatment for craniopharyngioma: a systematic Ghrelin does not stimulate food intake in patients with surgical procedures review and individual-level data meta-analysis. J Clin Endocrinol Metab 2013; 98: involving vagotomy. J Clin Endocrinol Metab 2005; 90: 4521–4524. 2239–2246.

Supplementary Information accompanies this paper on International Journal of Obesity website (http://www.nature.com/ijo)