Spontaneous Adrenal Hemorrhage with Mild Hypoadrenalism in a Patient Anticoagulated with Apixaban for Antiphospholipid Syndrome: A Case Report and Literature Review

Background Adrenal hemorrhage (AH) is a serious endocrine complication of antiphospholipid syndrome (APLS). Case Presentation. We report a 45-year-old man who presented with several deep venous thromboses and was initially treated with apixaban, who later developed bilateral AH. Laboratory findings were consistent with cortisol deficiency yet preserved aldosterone physiology. He was diagnosed with APLS and treated with warfarin. After 8 months of follow-up, he remained on cortisol replacement with no evidence of recovery. We reviewed PubMed/MEDLINE indexed articles from 1950 to 2022 for cases of AH in APLS patients on anticoagulation. Six cases of patients on direct oral anticoagulants (DOACs) were reported. Discussion. The unique vasculature of the adrenal glands creates a “functional vascular dam” in the zona reticularis, which is susceptible to thrombosis in situ and hemorrhage. DOACs may further increase the risk of AH. Conclusion Depending on the degree of adrenal involvement in AH, patients can present with partial or complete primary adrenal insufficiency. More data are needed to characterize adrenal function after AH, and the safety of DOAC versus warfarin in patients with APLS warrants further studies.


Introduction
Primary adrenal insufciency (PAI) is a rare and lifethreatening complication of antiphospholipid syndrome (APLS). Te pathophysiology of PAI is considered due to thrombosis from the underlying coagulopathy, with possible transformation into hemorrhagic infarction [1]. Te destruction of the adrenal cortex secondary to thrombosis and/ or hemorrhage can progress rapidly; however, the degree to which it may afect adrenal function can vary.
Te use of a new generation of direct-acting oral anticoagulants (DOACs) has largely replaced warfarin in the management of hypercoagulable state. DOACs have predictable anticoagulation efects and do not require dose monitoring, unlike warfarin. Te convenience of DOACs leads to a surge in their use. However, to date, there is not an adequately powered randomized control trial establishing DOAC safety in APLS.
Our aim is to present a case of AH in APLS on DOAC and review the existing literature to inform subsequent management.

Case Presentation.
A 45-year-old man without signifcant past medical history presented with a left lower extremity deep vein thrombosis (DVT) and pulmonary embolism (PE). He had no personal or family history of previous trauma, bleeding, or clotting events. He was discharged on apixaban.
Tree days later, he was readmitted with bilateral fank pain. Abdominal CT (Figure 1) showed bilateral adrenal gland edema. He was hemodynamically stable. Due to persistent right fank pain and concern for ileus, an abdominal CT was repeated revealing enlarging adrenal glands concerning for AH ( Figure 1  Warfarin was started with an international normalized ratio (INR) target range of 2-3. Hydrocortisone was continued at a total daily dose of 30 mg. Since discharge, with up to 8 months of follow-up, his morning cortisol levels prior to taking hydrocortisone have been low at 4.4-4.9 μg/dL and ACTH has remained high at 92-137 pg/mL, but aldosterone has been variable at <1 to 9 ng/dL, and renin has been normal at 0.35 ng/mL/h, while DHEA-S remains low. Tis suggests that what transpired was a microthrombiassociated intra-adrenal hemorrhage involving the zona fasciculata and possibly reticularis with preservation of the glomerulosa.

Results.
At the time of this writing, we found 8 cases [2][3][4][5][6][7][8][9] of AH in patients with APLS on anticoagulation (Table 1). Six out of 8 patients were on DOACs. Te remaining 2 cases of AH were patients on warfarin, and their presenting INRs were supratherapeutic. All of the 8 patients were discharged on warfarin. Six out of 8 patients (75%) of these patients required corticosteroids and aldosterone at discharge. Long-term adrenal function is unknown.
Te incidence of adrenal insufciency in APLS is 0.4% [11]. In a series of 69 APLS patients with adrenal involvement who underwent CT and/or MRI, 57% had AH, whereas 10% had adrenal infarction, and 10% had adrenal enlargement. Adrenal involvement was bilateral in 77% of the patients [1].
Although APLS is a hypercoagulable state, biopsy (autopsy in 16 patients) in 22 APLS patients with adrenal insufciency revealed hemorrhagic infarction with vessel  thrombosis as the most frequent fnding, present in 55% of patients. Tis is followed by adrenal hemorrhage in 27%, adrenal infarction in 5%, and normal appearance in 9% of patients [1]. Tese fndings support the notion that in most of the cases of PAI in APLS, the primary event is adrenal vein thrombosis followed by hemorrhagic transformation. Te central pathophysiology of adrenal vein thrombosis is based on its unique vascular anatomy (Figure 2). Te adrenal gland is highly vascular, with three main arteries but only one vein. Te adrenal arteries branch into capillaries that form a vascular plexus around the zona reticularis. Te abrupt transition of the artery-capillary plexus is referred to as a "vascular dam." Venous stasis and hypercoagulability in this region can promote thrombosis, infarction, and subsequent hemorrhage. Te capillary plexus eventually empties into medullary sinusoids that form a single central vein.
Additionally, the musculature of the vein is arranged eccentrically and composed of thick longitudinal muscle bundles. Tis favors the formation of thrombi in pockets of local stasis emerging from their contraction. Te zona reticularis may be more severely afected compared to zona glomerulosa and fasciculata that are located further away from the vascular dam.
In our case, the diagnosis of adrenal hemorrhage was discovered on CT imaging incidentally showing adrenal congestion with evidence of mild hemorrhage. Te difuse thickening of the glands and adrenal enlargement has been reported as a precursor of adrenal hemorrhage [10,[12][13][14]. A high index of suspicion is needed, and any clinical indicators of acute adrenal insufciency should prompt immediate measurement of cortisol and ACTH. Additional laboratory tests may be helpful to support the diagnosis of adrenal insufciency and include DHEAS, aldosterone, and renin activity. It is not clear what the role of catecholamine measurement would be in these circumstances, but epinephrine production may be most noticeable as it is secreted from the medulla and depends on intra-adrenal cortisol for its synthesis. Norepinephrine may be of less value since it is also produced extra adrenally. What remains uncertain is whether recovery of adrenal function is possible in cases such as ours where there is evidence of partially preserved adrenal function.
Te choice of anticoagulant in APLS may be relevant to the risk of developing AH, yet only a few case reports are available and thus one can only hypothesize that warfarin may confer less risk of AH in comparison to DOACs. It has also been reported that APLS patients developed catastrophic APLS (CAPS) after switching to DOAC from warfarin [15,16]. Notably, they were all triple-positive APLS. AH also occurred in non-APLS patients taking DOACs [17][18][19].
Currently, the evidence for secondary prevention of thrombotic events in patients with APLS using DOACs is mixed [20]. Our patient had triple-positive APLS developed VTE during DOAC treatment. Te risk of thrombosis with DOACs use may be stratifed by the patient's APLS antibody status. In a recent retrospective study, the rate of venous thromboembolism or bleeding events with DOAC (apixaban or rivaroxaban) use among single-positive APLS patients is not signifcantly diferent from those treated with warfarin [21]. Te European League Against Rheumatism (EULAR) suggests against DOAC use in triple-positive APLS [22]. Warfarin should remain the as the reference anticoagulation until further evidence regarding the safety of DOACs in APLS is available. Tere is a lack of information on the course of adrenal recovery when AH occurs, particularly when there is evidence of partial preservation of function, when aldosterone appears preserved with mild to moderate defciencies in cortisol, adrenal androgens, and possibly catecholamines. Terefore, close follow-up with periodic assessment of the HPA axis and empiric steroid replacement therapy is reasonable, and long-term outcome reporting should be encouraged to obtain prognostic information.

. Conclusion
Tis case highlights that hypoadrenalism can be one of the complications of APLS, and the risk of AH might be heightened with DOAC use. Patients may have partially preserved adrenal function depending on the extent of adrenal involvement in AH, and biochemical and imaging evaluation is important for diagnostic and perhaps prognostic purposes. Hormonal replacement should be guided by clinical and biochemical fndings as is typically done for the management of adrenal insufciency with periodic assessment for recovery. Until further evidence is available for DOAC use in APLS, warfarin should be the frst-line anticoagulation treatment.

Data Availability
No data were used to support this study.

Disclosure
Tis case was presented as an abstract and poster presentation at the Endocrine Society meeting in Atlanta, GA, 2022.