Irreversible End-Stage Cardiomyopathy Following a Single Dose of Cyclophosphamide

Abstract

Cyclophosphamide (CPA) is an alkylating agent used in a variety of chemotherapeutic regimens as well as for autoimmune disorders, the latter owing to its immunosuppressive effects. Its primary dose-limiting toxicities are bone marrow suppression and cardiomyopathy. The clinical course of CPA-induced cardiomyopathy varies widely. Most reported cases have described a transient, reversible course, similar to acute anthracycline toxicity.¹ However, there have been a few cases of fulminant restrictive cardiomyopathy characterized by biventricular pump failure and a rapidly fatal course.² Unlike anthracycline toxicity, in which the cumulative dose over time is the main determinant predicting the development of cardiomyopathy, the primary risk factor for CPA-induced cardiotoxicity is the total dose administered per course.³ We report the case of a pulmonary hypertension patient taking long-term ambrisentan therapy who developed irreversible end-stage cardiomyopathy following a single dose of CPA at a dose well below that which has been previously reported.

A 61-year-old woman with a history of scleroderma with lung involvement and pulmonary hypertension presented with a 4-month history of worsening renal function. Her creatinine level had increased from 1.2 mg/dL to 4.5 mg/dL during this period and she experienced a gradual diminution in urine output. A renal biopsy was performed, which showed rapidly progressive crescenteric glomerulonephritis (RPGN), and she was admitted for initiation of dialysis. A routine echocardiogram done for follow-up of her pulmonary hypertension (PH) 3 weeks prior to admission showed an ejection fraction (EF) of 60%, normal diastolic function, normal chamber sizes, and a right ventricular systolic pressure (RVSP) of 27 mm Hg. The patient’s PH had been well-controlled on a regimen of sildenafil and ambrisentan and she had not experienced any signs or symptoms of right heart failure. She had no prior cardiac history nor was there a family history of cardiomyopathy. She was a nonsmoker and a nondrinker and denied any drug use.

Following the diagnosis of RPGN, the patient was given 3 days of intravenous pulse methylprednisolone (1 gm/d) and a single dose of CPA 870 mg (0.5 g/m²). Renal function did not improve despite treatment. Outpatient dialysis was initiated and the patient was subsequently discharged home on a course of prednisone.

Three weeks later, the patient was readmitted to the hospital with seizures and altered mental status secondary to steroid psychosis. Steroids were discontinued and the patient was started on phenytoin. On day 8 of admission, the patient suddenly developed respiratory distress, hypoxia (oxygen saturation 82% on room air), and metabolic acidosis (pH 7.21, bicarbonate 17). She was placed on bilevel positive airway pressure ventilation (BiPAP). Her blood pressure remained stable (130/88 mm Hg) but the patient was tachycardic (heart rate, 100–115 beats per minute). Electrocardiography showed sinus tachycardia. Chest x-ray showed mild-moderate pulmonary edema. An echocardiogram was immediately performed and showed severe biventricular failure with an EF of 15% to 20%, severe left ventricular (LV) global hypokinesis, right ventricular (RV) hypokinesis, moderate-severe mitral regurgitation (MR), and an RVSP of 60 mm Hg. In addition, there was evidence of grade 2 to 3 diastolic dysfunction (Table I).

At this point, Swan-Ganz catheterization was performed and demonstrated severely elevated filling pressures and low output cardiac failure (Table II). In response to the severely diminished cardiac output, inotropic support was initiated with dobutamine at 5 μg/kg/min. The patient was also started on a regimen of losartan and spironolactone and fluid removal was achieved with hemodialysis. In addition, sildenafil and ambrisentan were discontinued, as pulmonary vasodilator therapy is contraindicated in patients with left heart failure and elevated filling pressures. Within 24 hours, the patient had clinically stabilized and was weaned off BiPAP and onto nasal canula. Oxygen saturations remained stable (>96%).

During the next several days, dobutamine was gradually uptitrated to 7.5 μg/kg/min. Cardiac index improved to 2.1 L/min/m² to 2.5 L/min/m² and pulmonary capillary wedge pressure decreased to 18 mm Hg to 20 mm Hg with continued dialysis. The patient remained hemodynamically stable throughout her hospital course. Once clinically compensated, the patient underwent left heart catheterization (LHC) to rule out ischemic cardiomyopathy. LHC demonstrated nonobstructive disease in all coronary arteries. Other etiologies of the cardiomyopathy, including viral and autoimmune, were not suspected in our patient. Her heart failure was not preceded by a distinct viral illness to suggest a viral myocarditis. Moreover, she did not present with clinical features that would suggest an autoimmune myocarditis such as giant cell or eosinophilic myocarditis. These forms of heart failure often manifest with malignant arrhythmias or heart block and have a fulminant course leading to early death. For this reason, endomyocardial biopsy was not performed. Based on the temporal association between the administration of CPA and the development of heart failure, our patient was diagnosed with CPA-induced cardiomyopathy with low-output cardiac failure and discharged home on a continuous infusion of dobutamine 7.5 μg/kg/min.

The patient remained clinically stable and well compensated over the next several weeks. She was readmitted to the hospital approximately 1 month later for an unrelated reason. Due to her stable clinical status, we elected to wean her off dobutamine at this point and assess her response. The patient remained normotensive off of inotropic support and did not decompensate. Repeat echocardiography showed a slight improvement in EF to 25% to 30% and a decrease in tricuspid regurgitant jet and RVSP to 32 mm Hg. There was persistence of RV systolic dysfunction, and the degree of MR was unchanged.

Approximately 10 days later, the patient returned to the emergency department complaining of severe shortness of breath and was found to be in cardiogenic shock. Her blood pressure was 56/31 mm Hg and oxygen saturation was 82% on room air. Chest x-ray showed moderate pulmonary vascular congestion and small bilateral effusions. Electrocardiography showed sinus tachycardia. Arterial blood gas demonstrated respiratory acidosis (pH 7.28, partial pressure of carbon dioxide 14). The patient was initially placed on a 100% nonrebreather mask and eventually transitioned to BiPAP. Dobutamine was immediately restarted at 7.5 μg/kg/min. Within 1 hour of dobutamine initiation, blood pressure had stabilized and the patient was off BiPAP and down to 2 L of oxygen via nasal canula. Arterial pH had normalized and the patient reported significant clinical improvement. A right heart catheterization was performed (while on dobutamine), and as expected revealed elevated filling pressures and severe PH but with normal cardiac output and index (Table III). Fluid removal was again achieved with hemodialysis as the patient continued to be dialysis-dependant. Repeat echocardiography this time was unchanged (EF 20%–25%). A cardioverter-defibrillator was implanted for primary prevention of sudden cardiac death and for prophylaxis against dobutamine-induced tachyarrhythmias. The patient was discharged home on 7.5 μg/kg/min of dobutamine and with New York Heart Association class II symptoms.

Since her most recent discharge from hospital, our patient has not had any further episodes of decompensated heart failure or shock and she remains clinically stable and active. Given her failure to wean from dobutamine, she will most likely exhibit long-term inotrope dependence.

CPA is among a handful of chemotherapeutic drugs that have cardiotoxic side effects. The spectrum of CPA-induced cardiac complications includes myocarditis, pericarditis, cardiomyopathy, isolated arrhythmias, and, rarely, fulminant cardiac failure.⁴ In a study of 61 patients receiving CPA as part of a sequential high-dose regimen for metastatic breast cancer, 10% developed symptomatic left ventricular dysfunction, with a median EF of 31%.⁴ Nearly all of these patients received a previous anthracycline-based regimen, and all but one had normalization of EF within 5 months.⁴ This study and others show that in the majority of cases, CPA-induced systolic dysfunction is acute and reversible. We report the case of a patient who developed irreversible, end-stage cardiomyopathy and who subsequently became dependent on chronic inotropic therapy following a single dose of CPA.

The only risk factor thus far identified that appears to predict the development of CPA cardiotoxicity is total cumulative dose per course. This is in contrast to anthracyclines, in which toxicity follows a dose-response relationship and depends on the total cumulative lifetime dose.⁵ Moreover, anthracycline cardiotoxicity has been associated with preexisting cardiac disease, hypertension, history of irradiation, and prior anthracycline exposure.⁴ No association between these risk factors and CPA has been found.⁴

The incidence of CPA cardiotoxicity seems to depend on a certain critical dose per course of administration.⁶ As reported by Goldberg and colleagues,⁶ this threshold critical dose is approximately 1.55 g/m². In several case series, the incidence of symptomatic cardiotoxicity was 25% when CPA was dosed at or above 1.55 g/m²; when dosed below 1.55 g/m², the incidence was only 3%.⁶ Our patient received one dose of CPA dosed at 0.5 g/m², which is well below the previously published threshold critical dose. Not only did she develop symptomatic left ventricular dysfunction at this dose, she went on to develop irreversibility. Reports of irreversible cardiomyopathy at such doses of CPA are extremely rare and have not been well documented in the literature.

The cytotoxicity of CPA is related primarily to 4 of its biologically active metabolites: 4-hydroxy-CPA, phosphoramide mustard, nornitrogen mustard, and acrolein.⁷ The parent drug has no alkylating or cytotoxic effects.⁷ These metabolites, in turn, are eliminated by the kidneys. Thus, in renal insufficiency, CPA may exhibit potentially greater cytotoxicity/cardiotoxicity, which suggests that the dose of CPA may need to be adjusted in renal failure. However, Bramwell and colleagues⁸ found significant individual variation in the metabolism and breakdown of CPA, so much so that they were unable to establish a correlation between the active metabolites of CPA and renal insufficiency.^1,8 Moreover, they showed that further metabolism of CPA occurred prior to renal elimination and that this was an important route of removal of alkylating activity.^1,8

A small study by Ayash and colleagues⁹ showed that women receiving a preparative regimen of CPA for metastatic breast cancer and who subsequently developed cardiomyopathy had significantly lower total CPA levels than women who did not develop cardiomyopathy. This study suggests that enhanced CPA activation by the liver can lead to increased cardiotoxicity. Thus, liver metabolism of the parent compound may, in addition to renal elimination of the active metabolites, be a key determinant and strong predictor of the development of cardiomyopathy in patients receiving CPA. This may explain why, when compared with our patient who had end-stage renal disease, other patients with lesser degrees of renal impairment do not develop cardiotoxicity after receiving CPA.

Another key determinant in our patient’s irreversible clinical course may have been ambrisentan, an endothelin receptor antagonist. Endothelin-1 (ET-1) is among a number of biomarkers that have been postulated to play a role in predicting the development of cardiomyopathy associated with CPA. In some small studies, a decrease in ET-1 levels has been shown to correlate well with a decrease in EF after anthracycline administration.¹⁰ This suggests that the reduction in ET-1 is a result of the direct cytotoxic effect of anthracyclines. If this is indeed the case, CPA may play a similar role in reducing ET-1 levels. In the setting of long-term ambrisentan therapy in which ET-1 levels may already be reduced or the activity of ET-1 nevertheless attenuated, it is therefore conceivable that CPA may have a synergistic effect on the development of fulminant cardiomyopathy at doses far lower than the previously reported threshold doses for cardiotoxicity.

The precise mechanism by which CPA causes cardiomyopathy is not fully understood but most likely involves oxidative stress caused by generation of oxygen-free radicals as well as inflammatory cytokines, with resultant production of neurohormones.⁷ These free radicals can induce apoptosis. Cardiac myocytes are, in particular, susceptible to oxidative stress due to low levels of antioxidants in the heart.⁷ Recent studies in rats who developed congestive heart failure after receiving bevacizumab, which is a monoclonal antibody against vascular endothelial growth factor, showed overexpression of inflammatory cytokines as well as monocyte chemoattractant protein, both in plasma and in the myocardium.¹¹ These inflammatory cytokines, chemoattractants, and neurohormones may ultimately be markers of impending cardiotoxicity.¹¹

This case demonstrates that a greater awareness of the potential cardiotoxic side effects of CPA is warranted. The clinical course of our patient shows that CPA can lead to an irreversible, end-stage cardiomyopathy and at doses far lower than previously reported. Therefore, just as for anthracycline-based chemotherapy, it is imperative that baseline assessments of cardiac function (including echocardiography and/or multi-gated acquisition scan) be obtained prior to initiation of any CPA regimen. In fact, this should be standard of care. Furthermore, dose adjustments may be needed in patients with renal impairment. Whether PH patients who are taking long-term ET receptor antagonists are at greater risk for a more fulminant course remains unclear and deserves further investigation. Although biomarkers such as ET-1 may prove to be useful, more work will be needed in the future in order to identify those who are at greatest risk for developing CPA-induced cardiomyopathy.