Hemoptysis and Anabolic Steroids: A Case Report of Oxandrolone-Induced Bleeding

Introduction

Anabolic androgenic steroids (AAS) are a class of natural and synthetic hormones that produce anabolic and androgenic effects [1]. Oxandrolone is a synthetic AAS that used to be approved by the FDA for treatment of malnutrition and weight loss following severe burn injury, trauma, HIV-related wasting, and neuromuscular disorders [2]. It was inappropriately used by athletes for augmenting muscle build-up, which briefly led to its withdrawal from 1989 to 1995, with subsequent reintroduction as a schedule III-controlled substance. Recently, the FDA withdrew approval due to concerns about safety and effectiveness [3]. AAS use has been associated with multiple adverse effects like hepatotoxicity, hypertension, testicular hypertrophy, prostrate hypertrophy, and cancer [1]. Regarding hematologic manifestations, AAS can cause erythrocytosis, which can lead to increased hematocrit and viscosity, increasing the risk of thrombosis [4], as well as causing a wide array of effects on coagulation factors that cause both pro- and anticoagulant effects [5,6].

We present a case of episodic hemoptysis related to bleeding from the nasopharynx and adenoids in an otherwise healthy young man following oxandrolone use, which resolved upon cessation of use. To the best of our knowledge, this is the first reported case in the literature of such an association.

Case Report

A 20-year-old otherwise healthy man presented to the emergency room with a 4-day history of daily hemoptysis. He indicated that every morning on waking up for the past 4 days he was spitting up dark blood with occasional clots. He did not have any further episodes during the rest of the day. He denied any past personal or family history of unexplained bleeding. He noticed a progressive increase in the volume of hemoptysis, up to approximately 250 ml, as measured by spitting into an empty water bottle, on the day of presentation. Physical exam on admission revealed a muscular male in no apparent distress, with a heart rate 64 beats/minute, blood pressure 126/66, respiratory rate 17 breaths/minute, temperature 36.7°C. There were no significant abnormal findings on cardiac, pulmonary, and abdominal examination. A visual inspection of the oral cavity utilizing a flashlight was unrevealing; the patient could not tolerate the use of a tongue depressor due to a pronounced gag reflex. Initial labs revealed white blood cell count 5.76×103/µL (4.4–10.5×103/µL) with an unremarkable differential, hemoglobin 14.2 g/dL (12.6–16.7 g/dL), hematocrit 42% (36.9–48.5%), platelet count 153×103/µL (139–361×103/µL); serum chemistries were significant for a mildly elevated aspartate aminotransferase (AST) level at 76 U/L (5–46 U/L), and the other liver chemistries were within normal limits, with alanine aminotransferase 14 U/L (4–51 U/L), alkaline phosphatase 111 U/L (10–129 U/L), and bilirubin 0.4 mg/dL (0.1–1.5 mg/dL). Prothrombin time was 13 seconds (11.5–14.9 seconds) and INR 0.97 (0.8–1.2). A urine toxicology screen was negative. Given his symptoms, potential pulmonary causes of bleeding were evaluated first. A CT chest with IV contrast performed in the emergency room did not reveal any pulmonary infiltrates or nodules. Pulmonary causes of hemoptysis such as pneumonia, tuberculosis, or malignancy were thus considered to have been ruled out. Pulmonary embolism as a potential cause was unlikely given normal vitals, including heart rate, respiratory rate, oxygen saturation, and lack of symptoms of shortness of breath or chest pain. Similarly, lack of any other symptoms, and a normal pulmonary physical examination, ruled out bronchitis as a cause. Also, the reported time at which the hemoptysis occurred – only in the morning upon waking up – suggested expectoration of blood and bloody secretions from another source. Coughing up of aspirated blood from hematemesis was considered as a differential diagnosis. Given the mild elevation of AST, abdominal imaging was obtained to rule out hepatic disease, in case some etiology of chronic liver disease and associated complications such as esophageal varices could be playing a role. A CT abdomen and pelvis with IV contrast did not reveal any hepatic parenchymal or duct abnormalities.

Upon admission, he underwent rapid-sequence endotracheal intubation for periprocedural airway protection, followed by bronchoscopy with bronchoalveolar lavage and an upper-GI endoscopy, but there was no evidence of a source of bleeding or any other abnormality such as diffuse alveolar hemorrhage on the former, or peptic ulcer disease or esophageal varices on the latter. With these investigations, the lower respiratory tract and the upper GI tract were considered to have been fully excluded as the sources of bleeding. He was subsequently extubated. An episode of hemoptysis was witnessed the following morning by multiple hospital staff. This ruled out malingering as a potential differential diagnosis. Upon further detailed history-taking, he stated he had started using oxandrolone 10 mg by mouth twice a day for about 14 days, but stopped around 4 days prior to admission upon first noticing the bleeding. He did not have a prescription for it, and was using it for enhancing the effects of bodybuilding exercises. He underwent a bedside flexible fiberoptic nasopharyngolaryngoscopy that revealed easy bleeding from nasal turbinates, nasopharynx, and adenoids upon instrument contact, but without a clearly localized lesion. Hematology consultation was obtained due to a suspicion of inheritable coagulopathy as a cause of bleeding from upper-airway microcapillary mucosal bleeding. An initial workup including von Willebrand factor antigen (VWF) level – 102% (55–160%) and activity – 88% (40–160%) showed normal results. Factor VIII activity was also normal at 124% (70–150%), as was platelet function analysis with epinephrine (PFA-100) at 148 seconds (84–164 seconds). This ruled out common bleeding diatheses like von Willebrand disease, hemophilia, and platelet function disorders that could first present with mild bleeding in adulthood.

Over the next 2 days, the patient and staff noted that the volume of hemoptysis progressively decreased until it was a small amount of blood-tinged mucus. Due to resolution of symptoms, further laboratory workup was not obtained in the inpatient setting. Taken together, the temporal association of oxandrolone use leading up to bleeding episodes, the radiologic imaging that did not show any abnormalities in the chest or abdomen, direct visualization of the lower respiratory tract and the upper GI tract further ruling out bleeding from those locations, and coagulation studies that were performed, the resolution of bleeding upon cessation of oxandrolone use, the patient was considered to have probably had oxandrolone-spontaneous microcapillary bleeding from the upper airway (involving the nose, nasopharynx, and adenoids as evidenced by easy bleeding on contact during the nasopharyngolaryngoscopy). Had the bleeding not already resolved, further investigation of the coagulation cascade in the form of levels of each coagulation factor would have been the next step. The patient was discharged with outpatient follow-up with hematology and recommendation to avoid resuming oxandrolone. The patient was contacted by telephone for follow-up 4 weeks after discharge, at which point he indicated he had not experienced any further bleeding episodes.

Discussion

There are a few cases of AAS-induced hemorrhagic manifestations that have been reported, such as diffuse alveolar hemorrhage [7], post-tonsillectomy bleeding [8], and development of subdural hematoma [9]. Thrombotic manifestations, such as extensive arterial thrombosis [10] and ischemic stroke (with subsequent hemorrhagic conversion) [11], have also been noted.

Hvid-Jensen et al [7] reported diffuse alveolar bleeding in 2 young male bodybuilders (aged 26 years and 19 years), both of whom had taken multiple AAS preparations, some oral and others via intramuscular injection. One patient had resumed use within the past 6 weeks, while the other had escalated dosing within 2 weeks of presentation. Both presented with cough with bloody sputum, and had CT imaging consistent with alveolar hemorrhage, and the latter patient also had evidence of recent bleeding on bronchoscopy. Both patients stopped AAS use, and follow-up imaging a few months later showed resolution of previous CT findings.

Fox et al [8] reported a patient who required multiple returns to the operating room for management of post-tonsillectomy hemorrhage. Their patient was a 30-year-old man with ongoing use of AAS for bodybuilding. They noted brief spikes of BP intraoperatively but normal vitals before and after surgical procedures. They hypothesized that as a coagulation screen was unremarkable, AAS-related hypertension could have had a role in the bleeding episodes, and concluded that AAS use should be considered a risk factor for postoperative bleeding in those otherwise thought to be healthy.

Alaraj et al [9] reported 2 cases of nontraumatic subdural hematoma in young bodybuilders taking AAS. They hypothesized that increased blood pressure during weight lifting may have led to episodic increased intracranial pressure, leading to the hemorrhage. Crucially, they also noted normal lab values on coagulation screens.

There is still much uncertainty about the precise mechanism by which AAS use may have caused bleeding in these reported cases. While drug-induced hypertension coupled with spikes in BP related to exercise may certainly have been a contributing factor to the cases reported by Alaraj et al [9], such a correlation is less clear in the case reported by Fox et al [8], as the elevated BP readings were for the most part limited to intra-operative monitoring. Hvid-Jensen et al [7] did not make specific mention of hypertension being an abnormality in their patients, and we did not note elevated BP readings during our patient’s hospitalization. This leads us to speculate that the effect of AAS on coagulation cascade is likely playing a role.

The effect of short-term use of oxandrolone on the coagulation cascade is complex (Figure 1). In a study by Kahn et al, upon administration to healthy volunteers, oxandrolone stimulated both anticoagulant (via increase in plasminogen, and decrease in plasminogen activator inhibitor-1) and pro-coagulant activity (activation of intrinsic and extrinsic pathways, and increased factors II, V, and X) leading to maintenance of homeostasis [5]. A more recent study by Camilleri et al [6] enrolled 100 amateur male athletes voluntarily starting AAS. The authors conducted blood investigations to elucidate the effects of AAS use on various aspects of the coagulation pathway before starting, during the use of, 3 months after discontinuing, and 1 year after discontinuing the use of AAS. They noted an increase in levels of factors II, IX, and XI, increase in d-dimer, unchanged factor VIII, and decline in VWF. Taken together, these findings suggested an increase in prothrombotic activity. However, decline in thrombin generation parameters such as peak thrombin concentration, rate of thrombin generation between lag time of thrombin generation and peak, and total thrombin formed indicated the anticoagulant effect of AAS. These abnormalities resolved completely after discontinuation of use. These studies indicate that AAS have a complex effect on the coagulation system, and may explain why both bleeding and thrombotic adverse effects have been reported in case reports.

Our patient underwent an extensive and rapid workup in an attempt to find a source of bleeding. Once the lower airways and the GI tract were excluded as a site of bleeding, by exclusion, only the upper airway remained. Direct visualization revealed easy bleeding from contact throughout the nose, nasopharynx, and the adenoids, without a focal lesion that could be treated. Coagulation studies readily available in the inpatient clinical setting did not reveal an underlying bleeding diathesis, and the bleeding progressively decreased to resolution. There was a temporal association between use of AAS, in our case, oxandrolone, and the development of bleeding episodes a few days later, followed by improvement a few days after cessation. On application of the Naranjo Adverse Drug Reaction Probability Scale [12], our patient scored 6, classified as “probable”. We believe this is the first reported case in the literature of spontaneous bleeding from the nasopharynx and adenoids in association with AAS – specifically, oxandrolone use. While we cannot be certain, our hypothesis is that the patient likely had microcapillary bleeding all day from the nasopharynx area, but only noticed it in the morning on waking up and clearing secretions and blood built up overnight, at which time the process of doing so likely caused further bleeding.

Conclusions

AAS have been widely used as image and performance enhancement drugs. Development of dependency on AAS is a public health concern. Although the FDA has recently withdrawn approval for oxandrolone, it might still be used by patients such as our own, by obtaining it surreptitiously without a prescription. Also, other AAS such as danazol and nandrolone continue to be approved, and it is possible that their use could lead to similar adverse effects. There are many adverse effects associated with the use of AAS. While some are common, like hirsutism, acne, libido change, atrophy of testes, and infertility, others may be relatively rare. Bleeding is an uncommon adverse effect associated with AAS. It is important to be aware of the possibility of transient bleeding manifestations as a possible adverse effect of oxandrolone or other AAS, especially in young, otherwise healthy adults. A thorough evaluation of all medications, prescribed or otherwise, is particularly important when dealing with unusual presenting symptoms. AAS might not appear on standard medication reconciliation sections in electronic medical records (especially if being taken without a prescription) or on basic toxicology screen results. Obtaining a good history, with time course, of use of such drugs in relation to appearance of symptoms can be the key to making the diagnosis.

AAS-associated hemorrhagic manifestations are variable and unpredictable. More research is needed to find various factors that could influence AAS-associated bleeding versus thrombotic complications.