Hoyeraal-Hreidarsson syndrome (HHS), is a rare severe variant of dyskeratosis congenita (DKC) characterized by cerebellar hypoplasia, microcephaly, myelodysplasia and immunodeficiency [12]. The multi-systemic nature of this genetic disorder often precludes the application of hematopoietic stem cell transplantation among those patients with bone marrow failure. In these patients, androgen therapy has provided an evidenced base therapeutic option [345].

Oxymetholone is a synthetic anabolic steroid that stimulates erythropoiesis through the increased production and secretion of erythropoietin. It has proven efficacy in treating patients with myelodysplastic syndromes, including Fanconi's anemia [5]. Other studies have described a 70% success rate of oxymetholone therapy in patients with DKC [3]. Adverse effects associated with use of oxymetholone include: hepatotoxicity, cholestatic jaundice which is dose and duration dependent, pelosis hepatitis, and benign and malignant hepatic tumors [6]. Oxymetholone has also been linked to decreased anticoagulant tolerance and hyperlipidemia. Other common side effects associated with the long-term use of anabolic steroids include: virilization in females, gynecomastia, amenorrhea, changes in libido and impaired thyroid function [46]. To our knowledge, this is the first case report in the literature attributing colonic angioectasia secondary to oxymetholone in a pediatric patient with HHS. Written informed consent was obtained from the patient for publication of this case report and accompanying images.

Our patient is a 13-year-old Caucasian male with cerebellar hypoplasia, skin dyskeratosis with hyperpigmentation, nail atrophy, esophageal strictures, urethral meatal stricture, developmental delay, and bone marrow failure who presented with hematochezia. He was initially diagnosed with developmental delay and failure to thrive as an infant. Features of DKC, and ultimate diagnosis with HHS was not made until he was about 8 years old when he presented with notable anemia, mild thrombocytopenia, and mild leukopenia. He was diagnosed with HHS based on his classical clinical features. Genetic testing was negative for mutations associated with DKC but revealed a telomere length less than the first percentile. When he was diagnosed with bone marrow failure syndrome, he was placed on a trial of androgens that notably improved his counts so that he was transfusion independent.

At age 11 years, a national shortage of androgens led to an interruption in oxymetholone therapy lasting two years. When the treatment course was resumed with 50 mg of daily oxymethalone, the patient failed to respond to therapy, at which point he was evaluated for bone marrow transplantation. He presented to an outlying facility with low grade fever, neutropenia and severe hematochezia. A flexible sigmoidoscopy showed friable colonic tissue, and two small rectal ulcers with an overlying eschar. No biopsies were obtained secondary to the elevated risk of bleeding. In addition, no esophagogastroduodenoscopy was performed due to an esophageal stricture. Stool cultures were negative for all infectious causes of hemorrhagic colitis, including cytomegalovirus. The suspicion for inflammatory bowel disease was low.

One week thereafter, the patient developed another episode of bright red blood per rectum associated with neutropenia and fever. Upon admission to Johns Hopkins All Children's Hospital, the patient had a platelet count of 22/µL, hemoglobin was 8.1 g/dL, white blood cell count was 0.72×10³/µL, and the international normalized ratio was 1.2. Physical examination of the abdomen was benign. Once again, an infectious workup, including stool adenovirus, cytomegalovirus, norovirus, enterovirus, cryptosporidium, giardia, clostridium difficile, Escherichia coli and other bacterial stool pathogens were negative. The patient continued to have profuse hematochezia that required multiple transfusions of packed red blood cells. Bleeding was bright red would suggesting a colonic source for the bleeding.

An abdominal ultrasonography and doppler was normal. A sedated abdominal and pelvic magnetic resonance enterography was unremarkable. On colonoscopy, several large angioectasias were identified at the hepatic flexure and cecum (Fig. 1). No lesions were actively bleeding and biopsies were not obtained due to the high risk of re-bleeding. In addition, the patient had no evidence of liver cirrhosis.

Since the patient had been evaluated for bone marrow transplantation, the oxymetholone was discontinued. Interestingly, the hematochezia spontaneously resolved and the neutropenia improved. The patient was then discharged home. A follow-up colonoscopy, performed 40 days post colonoscopy showed complete resolution of the colonic angioectasias. The planned laser ablation of these angioectasias was thus aborted. Alternative therapeutic options to laser ablation include placement of endoscopic hemoclips and thermal therapy. The patient did not have any further episode of hematochezia and no further intervention was indicated.

To the best of our knowledge, this is the first report of a case of colonic angioectasia secondary to oxymetholone in a pediatric patient with HHS. This genetic syndrome is unlikely to be the source of these abnormalities since the lesions disappeared spontaneously once the anabolic steroid was discontinued. A review of the literature revealed one case of colitis associated with DKC. In this case, the patient was reported to have idiopathic colitis, and portal hypertension [7]. No similar cases have been identified in patients with HHS. In our patient, the abdominal ultrasonography with Doppler did not show portal hypertension.

One study described an association between the androgen danazol and hemorrhagic cystitis. In this case series of 69 patients with angioneurotic edema, 13 had developed hematuria while on danazol therapy. In these patients, cystoscopic evaluation showed a nonspecific pattern of erythema, submucosal telangiectasia and neovascularity. Interestingly, as in our patient, these vascular lesions also spontaneously resolved once the anabolic steroid was discontinued [8]. Using the Naranjo adverse drug reaction probability scale, the yield shows a score of 7 which is indicative of a probable association between oxymetholone and the development of angioectasia.

With the increased use of anabolic steroids to treat patients with bone marrow failure, the potential risk in developing angioectasias must be underscored. Moreover, pediatric gastroenterologists should also familiarize themselves with these vascular lesions since they are perceived as uncommon causes of intestinal bleeding in children. Further studies are needed to elucidate the pathogenesis of anabolic steroid induced angioectasia.