Refractory vasospastic angina in a patient with autoimmune hepatitis managed by using combined non-invasive provocation tests: A case report

Abstract

Corticosteroids have emerged as a feasible treatment option for patients with refractory vasospastic angina (VSA) who have an allergic disease or hypereosinophilia. However, reports on VSA following the administration of corticosteroids are scarce. Herein, we present the rare case of a 49-year-old Japanese man who developed VSA while receiving corticosteroids for treatment of autoimmune hepatitis. Invasive coronary angiography and pharmacological spasm provocation tests under oral prednisolone (20 mg) revealed multi-vessel spasms with ST elevation but no obstructed coronary arteries. To assess the effectiveness of the pharmacotherapy following a Ca²⁺ antagonist (benidipine) administration, we adopted non-invasive and non-pharmacological coronary vasospasm provocation tests (cold-pressor stress, isometric handgrip exercise, and hyperventilation). These tests revealed transient ST elevations with angina, which prompted us to add other vasodilators to control the high disease activity. These tests can be particularly valuable in cases where spasmogenicity fluctuates with changes in corticosteroid dosage during prolonged treatment of hepatitis.

Learning objective

Corticosteroids can induce vasospastic angina (VSA) and corticosteroid-induced VSA may relapse in a dose-dependent manner. Discontinuation of corticosteroids is often difficult, and long-term administration may be inevitable. Evaluation of the fluctuated spasmogenicity during long-term observation of autoimmune hepatitis may be crucial.

Introduction

Corticosteroids have emerged as a potentially effective treatment option for patients with refractory vasospastic angina (VSA) who have an allergic disease or hypereosinophilia [1]. However, corticosteroids can induce coronary spasms, and only a few reports of VSA occurring after corticosteroid administration exist [2,3]. In many cases, corticosteroid continuation may be deemed necessary, even if it induces coronary spasms. In such cases, ongoing administration of multiple vasodilators is necessary to manage corticosteroid-induced VSA.

In this report, we present a rare case of a middle-aged man with corticosteroid-induced VSA, who underwent combined noninvasive coronary vasospasm provocation tests [cold-pressor stress, isometric handgrip exercise (IHE), and hyperventilation] for disease management and evaluation of the effects of pharmacotherapy.

Case report

A 49-year-old Japanese man (height: 153 cm, weight: 61 kg, and body mass index: 26.1 kg/m²) with no prior medical history visited our hospital because of general malaise; liver dysfunction was detected in his annual check-up. He had no history of tobacco, alcohol, drug, or supplement use or family history of premature coronary artery disease.

Upon admission, the vital signs were as follows: body temperature, 36.7 °C; pulse, 75 beats/min; blood pressure, 157/102 mmHg; and oxygen saturation, 98 % on ambient air. On physical examination, cardiac, pulmonary, and abdominal findings were normal, aside from apparent jaundice. Electrocardiography (ECG) revealed a sinus rhythm with unchanged ST segments. Initial laboratory findings were consistent with severe liver injury: aspartate aminotransferase, 1328 U/L (normal range: 13–30 U/L); alanine aminotransferase, 2388 U/L (normal range: 10–42); and total bilirubin, 3.5 mg/dL (normal range: 0.4–1.5 mg/dL). Serological examination for hepatitis-associated viruses (e.g. hepatitis viruses A, B, C, and E, Epstein–Barr virus, and cytomegalovirus) were negative. Autoimmune disease-associated autoantibodies (e.g. anti-nuclear, anti-mitochondrial, anti-liver-kidney microsomal, and anti-smooth muscle antibodies) were also negative. Computed tomography revealed no apparent abnormalities in the liver or biliary tract. The patient was admitted to the hospital with acute hepatitis.

Owing to the persistent elevation of liver enzyme levels following admission, intravenous administration of a corticosteroid (methylprednisolone, 1000 mg/day, 3 days) was initiated on hospital day 3 and gradually tapered. On day 7, percutaneous liver biopsy was performed. Based on the histological and laboratory findings, seronegative autoimmune hepatitis was diagnosed. On day 12, we replaced intravenous methylprednisolone with oral prednisolone (PSL) (25 mg/day).

In the early morning of day 13, the patient complained of transient chest pain after walking to the toilet. The chest pain lasted for approximately 20 min and disappeared after administration of isosorbide dinitrate (ISDN). Twelve‑lead-ECG demonstrated marked ST elevation in leads V2–V6, I, and aV_L during the chest pain, which returned to baseline after administration of ISDN (Fig. 1A, B). Laboratory tests 4 h after the attack revealed elevated high-sensitivity cardiac troponin I (1595 pg/mL, normal range: <46.4 pg/mL) and creatine kinase (26 U/L, normal range: 59–248 U/L) levels. Computed tomography of the coronary artery revealed no significant stenosis or thrombotic occlusion of the epicardial coronary arteries, suggesting ischemia with non-obstructive coronary arteries. Therefore, treatment with benidipine (8 mg/day) and isosorbide mononitrate (40 mg/day) was immediately initiated, after which the ischemic attack subsided.

Fig. 1.

Electrocardiography. (A) Transient ST elevation with peaked T wave in leads V2–V6, I, and aV_L during the first chest pain. (B) The ST segment returned to baseline after intravenous administration of isosorbide dinitrate. (C) Non-invasive coronary spasm provocation tests under benidipine monotherapy; ST elevation in leads V2–V6, I, and aV_L following the cold-pressor test and isometric handgrip exercise.

In the early morning of day 20, we performed combined noninvasive and non-pharmacological coronary vasospasm provocation tests (cold-pressor stress, IHE, and hyperventilation) under Ca²⁺-antagonist monotherapy with benidipine (8 mg) and ongoing oral PSL (25 mg/day) to evaluate the residual spasmogenicity. The patient first underwent cold-pressor stress by submerging both hands in ice water for 3 min. Immediately after the cold-pressor test, the IHE was performed for 3 min. Finally, hyperventilation was induced by rapid and deep breathing at a rate of 30 breaths/min for 6 min. Transthoracic echocardiography (TTE) was simultaneously performed during the IHE and hyperventilation tests. During the cold-pressor test, gradual ST changes in leads V2–V6, I, and aV_L occurred, resulting in marked ST elevation with chest pain during the IHE (Fig. 1C). The ST segments returned to baseline after spraying with nitroglycerin. During the IHE, TTE revealed severe anteroseptal and apical hypokinesis. Based on these findings, isosorbide mononitrate was resumed, and diltiazem (100 mg/day) was added.

On day 36, we performed invasive and pharmacological spasm provocation tests under oral PSL (20 mg/day) without oral vasodilators. Coronary angiography revealed no significant stenosis or thrombotic occlusion of the coronary arteries, except for 50 % stenosis of the middle segment of the left circumflex artery. Injection of acetylcholine (Ach) (50 μg) into the left coronary artery caused almost total occlusion of the proximal left anterior descending artery (Fig. 2A–C), accompanied by chest pain and ST elevation in leads V2–V6. Injection of Ach (20 μg) into the right coronary artery caused 90 % stenosis in the middle and distal right coronary arteries (Fig. 2D–F), along with chest pain and ST depression in leads II, III, and aV_F, indicating multivessel coronary spasms. Following intracoronary administration of ISDN, the coronary vasospasm was relieved, and the ECG findings normalized.

Fig. 2.

Coronary angiography and acetylcholine (Ach) provocation test. (A–C) Left coronary angiography. The coronary spasm was provoked in the proximal left anterior descending artery (red arrowheads). (D-F) Right coronary angiography. The coronary spasm was provoked in the middle and distal right coronary arteries (red arrowheads). (A, D) Control angiograms. (B, E) Post-Ach provocation test. (C, F) Post-intracoronary injection of isosorbide dinitrate. The dose of Ach was 50 μg in panel B and 20 μg in panel E.

CAU, caudal view; ISDN, isosorbide dinitrate; LAO, left anterior oblique view; RAO, right anterior oblique view.

Because the patient had no chest pain attacks before corticosteroid administration, he was diagnosed with VSA, probably due to the corticosteroids. On day 39, he was discharged without symptoms and while receiving oral PSL (20 mg/day). Following discharge, the angina did not relapse, and the PSL dosage was gradually tapered to 10 mg/day.

Four months after discharge, the PSL dosage was increased to 30 mg/day owing to worsening of the autoimmune hepatitis. The patient was readmitted to the hospital, and oral PSL (30 mg/day), intravenous methylprednisolone (1000 mg/day, 3 days), and subsequent oral azathioprine (100 mg/day) were administered. During the second hospitalization, symptomatic relapse of angina with ST elevation occurred. The patient was diagnosed with relapsed corticosteroid-induced VSA, which was attributed to the increased dosage of corticosteroids. Nicorandil (15 mg/day) was added, and there was no recurrence. After re-tapering oral PSL to 10 mg/day while liver dysfunction was stabilized and subsequently replacing with benidipine monotherapy, chest pain with ST elevation recurred. His anginal symptoms were resolved after re-administration of isosorbide mononitrate, nicorandil, and diltiazem (Fig. 3).

Fig. 3.

The clinical course. The red arrows indicate spontaneous chest pain attacks with ST elevation.

ALT, alanine aminotransferase; mPSL, methylprednisolone; PSL, prednisolone; iv, intravenous.

Discussion

We report a unique case of corticosteroid-induced VSA, in which spasmogenicity fluctuates during prolonged treatment of autoimmune hepatitis and combined noninvasive coronary vasospasm provocation tests were performed to optimize the pharmacotherapy regimen. The central mechanisms underlying the spasms involve endothelial dysfunction and the hypercontraction of vascular smooth muscle cells (VSMCs), where activation of Rho-kinase, a molecular switch in VSMC contraction, plays a major role [4]. Based on previous basic research, corticosteroid-induced VSA is hypothetically caused by the pharmacological effects of corticosteroids on both endothelial cells and VSMCs. Corticosteroids inhibit endothelial release of nitric oxide and its metabolites (e.g. nitrite and nitrate) in a dose-dependent manner [5]. Sustained elevation of the serum cortisol level sensitizes pig coronary vasoconstricting responses through Rho-kinase activation [6]. Furthermore, a previous study suggests that stress-induced increases in steroid levels may enhance coronary spasms, as evidenced by an increase in Rho-kinase activity in humans after the Great East Japan Earthquake of 2011 [7].

An invasive pharmacological provocation test with intracoronary injection of Ach or ergonovine is generally used to identify coronary vasospasms [8]. As an alternative approach, noninvasive provocation tests with various stimulations, such as cold-pressor, IHE, and hyperventilation, can be performed. Although not yet established, these tests may be useful for evaluating the effects of pharmacotherapy given their repeatability. Nakao et al. highlighted the utility of the hyperventilation test, emphasizing its high specificity (100 %), but acknowledging its relatively low sensitivity (61.7 %) [9]. Using both hyperventilation and treadmill exercises among various combinations to induce VSA, Sueda et al. reported a positive response to this combination of 57 % even in patients with low disease activity, while 93 % in patients with high disease activity (one or more attacks per week) [10]. In the present case, provocation procedures even under a Ca²⁺ antagonist provoked ST elevations and necessitated more intensive medications, reflecting high disease activity. Of note was that the angina was well-controlled during low-dose corticosteroid treatment; however, it recurred after the steroid dosage was increase with exacerbation of hepatitis, demonstrating the dose-dependency of corticosteroid-induced VSA. Furthermore, the spontaneous relapse of the angina even in controlled hepatitis after withdrawal of vasodilators to benidipine indicated the need for intensive treatment of VSA induced by corticosteroid itself with multiple calcium antagonists and nitrates.

In conclusion, we experienced a rare case of corticosteroid-induced VSA. Our case highlights the importance of recognizing that corticosteroids can induce VSA and that corticosteroid-induced VSA may relapse in a dose-dependent manner. Discontinuation of corticosteroids is often difficult, and long-term administration may be inevitable. Noninvasive provocation tests may be adopted to evaluate residual spasmogenicity during long-term observation of autoimmune hepatitis.

Declaration of competing interest

The authors declare that there is no conflict of interest.