Myopericarditis in a pregnant woman with acute promyelocytic leukemia

Abstract

Acute promyelocytic leukemia (APL) is a form of acute leukemia with a characteristic translocation, t(15;17), and is considered a hematologic emergency, typically treated with all-trans retinoic acid and an anthracycline. We present the case of a young, gravid woman who was diagnosed with APL in the third trimester, initiated typical treatment, and suffered uncommon cardiac complications.

<Learning objective: Myopericarditis is not a side effect often encountered in the management of acute promyelocytic leukemia with alpha t-retinoic acid, and its mechanism is incompletely understood but possibly related to the differentiation syndrome. This complication can be effectively treated with systemic glucocorticoids, supportive care, and withdrawal of the offending agent, even in the pregnant population.>

Introduction

Acute promyelocytic leukemia (APL) is an M3 form of acute leukemia with a characteristic translocation t(15;17) and is considered a hematologic emergency. Leukemia generally affects 1 in 70,000 pregnancies, a subset at higher risk of morbidity and mortality by virtue of underlying physiologic stress and potential placental transfer of chemotherapeutics. APL presents more commonly relative to other myeloid leukemias in pregnancy due to its higher prevalence in younger populations 1, 2. Among treated, non-pregnant APL patients in the general population, survival rates exceed 70% [3]. Similar outcome data in pregnant patients with treated APL are not available, but case reports of successful therapy have been published. Management of APL in pregnancy varies by trimester at diagnosis 1, 4. The most commonly used regimens involve high-dose all-trans retinoic acid (ATRA) along with an anthracycline both of which can have significant cardiotoxic effects. The case that follows is of a young, gravid woman who was diagnosed with APL in the third trimester with ensuing complications.

Case report

A 24-year-old obese woman with history of a prior pregnancy complicated by preeclampsia was admitted to an outside hospital for easy bruising and headaches at 27 and 1/7 weeks pregnancy. On presentation, she was found to be pancytopenic, with an initial white blood cell (WBC) count of 1.0 × 10⁹/L. Bone marrow biopsy revealed an abnormally elevated number of myeloid blasts. Cytogenetic studies indicated t(15;17) consistent with APL. ATRA was initiated on day 4 of admission at a dose of 45 mg/m²/day, with dexamethasone to induce fetal lung maturity. Cell counts were supported with intermittent packed red blood cell and platelet transfusions. At day 7 she was transferred to a tertiary center for further management of this complicated pregnancy.

ATRA therapy was continued after transfer. Idarubicin at a dose of 12 mg/m² and prednisone were also initiated (for four total doses on days 21, 23, 25, and 27). On the evening of day 26, the 22nd day of ATRA therapy, she developed substernal, sharp chest pain radiating to the back, worse with deep inspiration, and positional changes. She had been tachycardic throughout with a heart rate ranging from 120 to 140 bpm, but she was afebrile. Systolic BP had ranged from 110 to 140 mmHg over the previous 24 h. Physical examination revealed tachycardia with a regular rhythm. S1 and S2 were normal. A ventricular gallop was present. There was a Grade I pericardial rub, with the patient leaning forward at end expiration. Hemogram was remarkable for WBC count of 9.8 × 10⁹/L with a neutrophilic predominance. Serum chemistry revealed normal electrolytes and baseline creatinine of 0.56 mg/dL. B-type natriuretic peptide was elevated at 2877 pg/mL. Troponin was elevated at 3.66 ng/mL. See Table 1 for timeline of laboratory abnormalities. An electrocardiogram (ECG) revealed ST elevations with associated PR depressions (Fig. 1). A chest X-ray was significant for hydrostatic pulmonary edema and normal heart size. Computed tomography angiography of the chest did not reveal evidence of pulmonary embolism or consolidation. A trans-thoracic echocardiogram showed a small, circumferential pericardial effusion (Fig. 2) with multiple segmental wall motion abnormalities in a non-coronary distribution. The estimated ejection fraction was mildly reduced, at 45%. The constellation of findings was felt to be consistent with acute myopericarditis in the setting of ATRA and idarubicin therapy.

Table 1.

Timeline of laboratory abnormalities.

Day	0	8	26	33
Event	Day of admission	Day of transfer	Onset of symptoms	Resolution of symptoms
WBC (×109/L)	1.2	2.2	9.7	0.26
Hct (%)	24.2	23.2	25.3	29.8
Plt (×103/mm3)	8	26	41	11
Sodium (mmol/L)	136	139	135	133
Potassium (mmol/L)	3.8	3.8	4.1	3.7
Chloride (mmol/L)	105	109	105	101
Bicarbonate (mmol/L)	23	22	22	24
BUN (mg/dL)	6	8	8	7
Cr (mg/dL)	0.5	0.5	0.4	0.51
Troponin (ng/mL)			3.66	1.26a
NT-proBNP (pg/mL)			2877

WBC, white blood cells; Hct, hematocrit; Plt, platelets; BUN, blood urea nitrogen; Cr, creatinine; BNP, B-type natriuretic peptide.

^aSpecimen collected on day 27 and was the last known troponin prior to resolution of symptoms.

Fig. 1.

12-lead electrocardiogram from day 26 with chest pain. Notice ST segment elevations in leads I, II, aVL, and V3–V6 with PR depressions in the associated leads indicative of both ventricular and atrial epicardial inflammation.

Fig. 2.

Parasternal long-axis view demonstrating a posterior pericardial effusion (see arrow).

ATRA and idarubicin were immediately discontinued due to concern for acute cardiotoxic effects. Prednisone 60 mg daily was initiated to treat the myopericarditis. Chest pain improved only after an increase in prednisone dosage to 80 mg daily. Metoprolol was added for arrhythmia prophylaxis and treatment of cardiomyopathy. Hydralazine and isosorbide dinitrate were used for afterload reduction. Low-dose furosemide was given for diuresis as needed. The patient developed a brief, symptomatic episode of atrial fibrillation with rapid ventricular response on hospital day 31 with spontaneous conversion to sinus rhythm. The metoprolol dose was subsequently increased. No further atrial or ventricular arrhythmias occurred.

Chest pain resolved completely 7 days after initiation of prednisone. Troponin down trended. A repeat transthoracic echocardiogram at admission day 40 showed normalization of left ventricular function and segmental wall motion abnormalities. The pericardial effusion was no longer present. There was no evidence of pericardial constriction.

The patient's course was further complicated by Gram-negative bacteremia and hyperglycemia secondary to steroid administration. Labor was induced, and she delivered vaginally at 34 weeks with an estimated blood loss of 200 mL. The baby was healthy with only a brief hospital stay for hyperbilirubinemia. The patient was discharged shortly after delivery with plans for outpatient consolidation chemotherapy. A slow prednisone taper of 10 mg per week was recommended. Metoprolol was continued. Therapy with an angiotensin-converting enzyme inhibitor was initiated, and isosorbide dinitrate and hydralazine were discontinued.

Discussion

This patient developed myopericarditis in the setting of third trimester pregnancy, APL, and administration of ATRA and idarubicin. The acute cardiotoxic effects were resolved with the discontinuation of ATRA and idarubicin, initiation of high-dose prednisone, and standard heart failure therapy. The differential diagnosis for her syndrome included: takotsubo cardiomyopathy, anthracycline-induced cardiomyopathy, ATRA-induced myopericarditis, viral myocarditis, and peripartum cardiomyopathy.

Acute myopericarditis is a known, rare complication of ATRA therapy in the setting of APL with a total of 6 reported cases 5, 6, 7, 8, 9, but it has never been reported in pregnancy. One case reported histological findings on autopsy of APL cells infiltrating the myocardium in this setting, suggesting pathophysiologic correlation with differentiation syndrome [7]. Differentiation syndrome (DS) is a complication of APL therapy with ATRA due to rapid therapeutic differentiation of undifferentiated peripheral promyelocytes leading to tissue infiltration of cells. Myopericarditis traditionally lies outside the recognized pattern of DS, typically presenting clinically as unexplained fever, weight gain, peripheral edema, dyspnea with interstitial infiltrates, pleuropericardial effusion, hypotension, and acute kidney injury [10]. Differentiation syndrome occurs in 2–31% of patients receiving induction therapy depending on adjunctive agents administered with ATRA (anthracyclines diminish this risk) and the presence of other risk factors, including elevated WBC count at start of therapy, peak WBC count during therapy, elevated body mass index (BMI), and creatinine level 10, 11, 12. Our patient did have a significantly elevated BMI, although WBC count at onset of therapy was unremarkable. The syndrome carries a mortality rate in non-pregnant populations from 2% to 10% [13]. Management typically includes stoppage of ATRA and initiation of steroid immunosuppression.

The patient also recently received idarubicin, so the question arises about the contribution of anthracycline toxicity. There are three known time frames of toxicity: early, chronic, and late onset. Early onset toxicity can occur anytime in the acute phase following initiation of idarubicin therapy, is dose-dependent, and characteristically involves high rates of atrial and ventricular arrhythmias 14, 15. Our patient received a relatively low total dose of idarubicin, 48 mg/m², well under the prescribed toxic point of 350 mg/m² [16]. In addition, she had only one self-limited episode of an atrial arrhythmia. In the setting of receiving ATRA therapy, acute anthracycline-induced cardiotoxicity was felt to be less likely.

Takotsubo and peripartum cardiomyopathy (CM) were also entertained in the differential; however her echocardiographic wall motion pattern was not consistent with takotsubo (i.e. basal segment hypokinesis with apical ballooning). Peripartum CM can present in the last trimester of pregnancy, but typically does not involve the pericardium [17]. A viral myopericarditis was felt to be unlikely given a more proximal culprit, the ATRA therapy, and the patient's lack of a viral prodrome [18].

Classic findings of myopericarditis on transthoracic echocardiogram often include a pericardial effusion, with segmental left ventricular wall motion abnormalities and depressed left ventricular ejection fraction. Cardiac biomarkers are also typically elevated. Prognosis varies by etiology, but reported mortality rates for biopsy-verified myocarditis are upwards of 20% and 56% at 1 and 4 years [19]. Treatment of myopericarditis is largely supportive, namely with the initiation of standard heart failure therapy [20]. Few rigorous comparisons have been drawn between therapies for various etiologies of the syndrome. Therapy for isolated pericarditis focuses specifically on symptom control with anti-inflammatory medications including aspirin, non-steroidal anti-inflammatory drugs (NSAIDs), colchicine, and prednisone. Non-aspirin NSAIDs have not traditionally been used in treatment of pericarditis when myocardium is involved due to the risk of poor scar formation and wound healing [21]. Therapeutic modalities available in the pregnant patient with myocarditis, pericarditis, or myopericarditis are limited (Table 2) by fetal concerns [22].

Table 2.

Treatment modalities for acute pericarditis in the gravid female.

Medication	FDA Pregnancy Risk Classificationa
All NSAIDs	Class D
Aspirin <100 mg/day	Class B
Prednisone	Class B
Colchicine	Class C

FDA, Food and Drug Administration; NSAIDs, non-steroidal anti-inflammatory drugs.

^aClass A: Adequate studies fail to demonstrate risk to fetus in first trimester. Class B: No adequate studies, but animal studies do not demonstrate risk. Class C: No adequate studies, but animal studies have shown an adverse effect in the fetus. Class D: Positive evidence of human fetal risk based on adverse reaction data, but benefit may outweigh risk. Class X: Positive evidence of human fetal risk based on adverse reaction data, and risk clearly outweighs benefit.

ATRA-associated myopericarditis in the pregnant patient has never been reported. In this setting, it can be difficult to identify the true inciting agent amongst a plethora of potential toxic agents. ATRA is the most likely culprit for reasons described above, although the mechanism of insult, whether by direct toxicity or leukemic infiltration, remains unclear. A management strategy similar to the non-pregnant patient was used. Future studies will be limited by the low prevalence of this treatment complication, but may focus on duration of therapy, as the side effects of prednisone in the pregnant population are accentuated by fetal concerns.

Contributors

We would like to thank the Knight Cancer Institute and OHSU Obstetrics & Gynecology for their aid in caring for this patient. We would also like to thank Drs Alan Hunter and Joseph Chiovaro for their aid in the project.

Funding

None.

Conflict of interest

Andrew Oehler and Shimoli Shah have no conflicts of interest to report.