Patients with thalassemia major (TM) require lifelong transfusion therapy to survive, leading to toxic iron overload in the endocrine glands and heart. The pituitary gland is one of the most vulnerable targets, leading to irreversible hypogonadotropic hypogonadism in approximately half of patients. Improvements in magnetic resonance imaging (MRI) technology and understanding have allowed earlier recognition of preclinical iron deposition in the heart, pancreas, and liver; prior work also supports a similar role for the pituitary. The purpose of this study is threefold, (1) to develop age-specific nomograms for pituitary iron and volume metrics; (2) to determine the prevalence, severity, and age of onset of pituitary iron deposition and volume loss in TM patients, and (3) to determine whether deferasirox monotherapy can modify the trajectory of pituitary iron accumulation and damage over a two-year period. This article outlines relevant background studies and methodological details as well as providing preliminary results from our first two aims.

2017 Poster Board I-1039 Introduction: Hypogonadotropic hypogonadism (HH) is the most common endocrine problem in chronically transfused patients. Unfortunately, gonadotrophe iron toxicity is not readily detectable until puberty and stimulation tests can be difficult to interpret in adolescents. MRI can image preclinical pituitary iron deposition, similar to its use in the heart, liver and pancreas. Increased pituitary R2, a surrogate for iron, and decreased pituitary volume have both been shown to predict clinical and biochemical HH in iron overloaded adults 1,2. However, data regarding pituitary iron deposition and toxicity is lacking in younger patients with iron overload. We present baseline results from a two year observational trial of changes in pituitary R2 and volume in response to deferasirox therapy in patients with transfusional siderosis. Methods: We studied 22 chronically-transfused patients with Thalassemia Major and 2 patients with Diamond Blackfan Anemia. The average age was 13.1 ± 5.2 years (range: 3.7-23.6 years). All studies were performed on a 1.5 T Philips Achieva. Anterior pituitary R2 was assessed in the sagittal and coronal planes using multiple spin echoes from 15 to 120 ms. Pituitary volume was assessed using a 3D spoiled gradient echo sequence with 1 mm3 isotropic voxels. Pituitary R2 was calculated by pixelwise monoexponential fit, with median values used to represent the overall gland R2; boundaries were confirmed by a board-certified neuroradiologist. Normative data for pituitary iron and volume was drawn from another study in 49 normal volunteers. MRI estimates of hepatic iron concentration (HIC), cardiac iron (T2*), and pancreatic iron (R2*) were obtained as clinical standard of care. All statistics were performed using JMP5.1 (SAS, Cary, NC). Results: Patients were mild to moderately iron loaded, with a HIC of 10.2 ± 12.0 mg/g dry wt (median 4.5 mg/g, nl < 1.5 mg/g), cardiac T2* of 29.4 ± 9.2 ms (median 31.5 ms, nl > 20ms), and pancreas R2* of 136 ± 156 Hz (median 73 Hz, nl < 27 Hz). Fourteen patients were below the 10th percentile for height including 7 below the 5th percentile. One patient had been diagnosed with delayed puberty and was on estrogen replacement. Pituitary R2 was elevated in 13/24 patients, beginning in the first decade of life and worsening in patients older than 13 years of age (Figure 1, left). Mean Z-score was 3.2 ± 3.7 (median 2.5, range -0.5 to12.8). Pituitary R2 was correlated with HIC (r2=0.68) and pancreatic R2* (r2=0.40), but not cardiac R2*. Area under the receiver operator characteristic curve was 0.76 for both HIC and pancreatic iron for the prediction of pituitary iron loading. Anterior pituitary volumes were low-normal in the first decade of life but were below the 1st percentile in 4/16 patients older than 13 years of age (Figure 1, right); all 4 patients with severe pituitary shrinkage had significant pituitary iron loading (Z-scores ranging from 2.8 to 12.8). The patient having documented HH had an iron Z-score of 12.8 and a volume Z-score of –5.0. Discussion: HH remains one of the most difficult endocrinopathies to recognize and prevent in transfusional siderosis. The present data demonstrate that iron accumulation occurs early in these patients and worsens dramatically in the second decade of life. Pituitary R2 was closely correlated with liver and pancreatic iron suggesting more rapid iron transport kinetics than for the heart. The low-normal pituitary volumes in childhood may reflect ethnic mismatches between the control and study populations, early iron toxicity, or hypothalamic dysfunction from increased erythropoiesis and metabolic demand. However, severe pituitary volume loss was limited to the second decade of life and associated with markedly increased pituitary iron (R2). Thus, there appears to be a broad, preclinical window to reduce pituitary toxicity. MRI screening of pituitary size and iron concentration needs to be expanded, clinically, to explore this hypothesis and hopefully prevent the significant physiological and psychological sequelae associated with hypogonadism. Disclosures: Coates: Novartis: Consultancy, Honoraria, Research Funding, Speakers Bureau. Wood:Novartis: Research Funding.

4073 Poster Board III-1008 Introduction Despite continuing advances in iron chelation therapy, endocrine dysfunction is common in chronically transfused patients with iron overload. In thalassemia major, hypogonadotropic hypogonadism (HH), impaired glucose tolerance, delayed growth, and hypothyroidism are common, but gonadotrophic function is the most vulnerable and least reversible. Unfortunately, clinical effects of gonadotrophe iron toxicity are not detectable until puberty. Provocative GnRH stimulation can be used to detect preclinical HH, but it can be difficult to interpret in adolescents. Thus, surrogates for preclinical pituitary iron deposition and damage are imperative. MRI can be used to measure pituitary iron (R2) and volume, analogous to its use for the heart, liver, and pancreas. Increased pituitary R2 and decreased pituitary volume predict clinical and biochemical HH in adults with thalassemia major 1,2. However, to apply these techniques to prepubertal patients with iron overload, it is imperative to develop age-appropriate normative data for pituitary iron and volumes. The goal of this study was to determine trends/limits of pituitary R2 and volume in normal subjects between the ages of 2 and 25 years. Methods We studied 49 normal volunteers between the ages of 2 and 25 years old; informed consent was obtained in all examinations. Patients as young as four years of age were studied without anesthesia by using video goggles and appropriate coaching. Seven patients were studied under general anesthesia by adding the research MRI (< 10” duration) on to a clinically indicated sedated head MRI. All studies were performed on a 1.5 T Philips Achieva. Anterior pituitary R2 was assessed in the sagittal and coronal planes using a multi-echo spin echo sequence using eight echoes from 15 to 120 ms, five mm thick slices, 1 mm2 inplane resolution, and repetition time (TR) of 500 ms. Pituitary volume was assessed using a 3D spoiled gradient echo sequence with 1 mm3 isotropic voxels, TR 7.2 ms, echo time 3.3 ms, and flip angle of eight degrees. Pituitary R2 was calculated by pixelwise fitting to a monoexponential without a background offset. Regions of interest incorporating the anterior pituitary were manually traced, with mean and median values were used to represent the overall gland R2. All contours were confirmed by a board certified neuroradiologist. Piecewise linear regression was used to describe anterior pituitary volume, while R2 values were well described by simple linear regression. All statistics were performed using JMP5.1 (SAS, Cary, NC). Results Anterior pituitary volume increased linearly with age (Figure 1, left) and plateaued at 18 years of age. Normalization with respect to age produced lower variability than normalization with respect to height, weight, and body surface area. No gender differences were observed. Posterior pituitary volume also increased with age but appeared larger in mid puberty than in adulthood (not shown). Sagittal and coronary pituitary R2 increased linearly (r2 = 0.10 and 0.38, p = 0.03 and < 0.001) with age throughout the study interval (Figure 1, right). Mean and median R2 values were unbiased and highly correlated (r2 = 0.93) with one another but median values were more robust to outliers. Pituitary R2 estimates in the sagittal and coronal planes were not statistically equivalent. Sagittal R2 values exhibited a bias of -3% that increased with R2 value. Limits of agreement were ± 20%. With respect to age-normalization, coronal R2 values exhibited 31% lower variability than sagittal images. Discussion Pituitary R2 values and pituitary volumes are age-dependent. Coronal R2 estimates were slightly higher and had lower variability than values calculated in the sagittal plane. This likely reflects the larger and more homogeneous sampling area in coronal pituitary views. Nomograms from these data will improve recognition of preclinical pituitary iron loading and volume-loss in iron overloaded patients. Additional subjects will be necessary to resolve peripubertal changes in posterior pituitary volume, trends in early adulthood, and gender or ethnicity differences. Disclosures: Coates: Novartis: Consultancy, Honoraria, Research Funding, Speakers Bureau. Wood:Novartis: Research Funding.