Article Text


Capecitabine-associated coronary vasospasm: a case report
  1. C A Papadopoulos,
  2. H Wilson
  1. Norfolk and Norwich University Hospital, Norwich, UK
  1. Dr C A Papadopoulos, Abbotswood Gardens, Clayhall, Essex IG5 0BG, UK; Norfolk and Norwich University Hospital, Norwich, UK; christina.papadopoulos{at}


Capecitabine (Xeloda) is an oral 5-fluorouracil pro-drug used in the treatment of two of the commonest cancers: breast and colorectal. This report concerns a 43-year-old woman with metastatic cancer of the sigmoid colon who developed cardiac chest pain 5 days after starting capecitabine therapy. Capecitabine-induced cardiac symptoms have previously been reported but infrequently. In the main they have documented pain and electrocardiogram (ECG) changes associated with exercise. This case report is of a patient with minimal cardiac risk factors, who had ischaemic cardiac pain with widespread ECG changes at rest that resolved with a nitrate infusion. Coronary vasospasm is proposed as the probable mechanism for the cardiac ischaemia and dramatic ECG changes. Capecitabine is now in widespread use and so physicians will encounter an increasing number of patients using this therapy. In the light of this, it is important that doctors in emergency and acute medicine are aware of its treatable cardiac side effects.

Statistics from

Capecitabine is an oral cytotoxic drug used in the treatment of metastatic breast and colon cancer. Treatment is on an ambulatory basis and is highly tumour specific. Absorbed across the gut mucosa, capecitabine is metabolised to 5-fluorouracil (5-FU) by a process of three enzymatic reactions. In the final step, 5-FU is only produced intracellularly by thymidine phosphorylase and even then, preferentially in tumour cells that have a higher concentration of thymidine phosphorylase than neighbouring non-neoplastic cells. Plasma levels of 5-FU are thus very low and capecitabine is regarded as highly specific for tumour tissue. Cardiac symptoms and toxicity caused by capecitabine have rarely been reported in the literature; less than the 1.5% to 2% documented for intravenous 5-FU. Previous cases mostly describe the onset of chest pain with ST segment changes on exercise, which resolve at rest and are not reproduced after the withdrawal of capecitabine. Echocardiography and coronary artery angiography are normal in all these cases. In this report, we discuss an extremely rare presentation of cardiac symptoms associated with capecitabine treatment.


The patient, a 43-year-old white woman, had been diagnosed with sigmoid carcinoma 6 weeks previously after an episode of rectal bleeding. Biopsies taken on flexible sigmoidoscopy showed a moderately differentiated adenocarcinoma of the colorectal type. A staging computed tomography scan revealed multiple hepatic metastases distributed throughout both lobes (radiological staging T3, N1, M1). There was a strong family history of colorectal cancer, with her father currently in the terminal stages of the disease. The patient was commenced on palliative chemotherapy: bevacizumab and the capecitabine/oxaliplatin regime (1000 mg/m2 twice a day of capecitabine for 2 weeks on and then one week off, with oxaliplatin 130 mg/m2 three weekly).

On day 5 of cycle 1, the patient presented to the oncology ward having developed mild chest pain at rest lasting 20 minutes. As the electrocardiogram (ECG) while she was pain free was normal, the patient was reassured and discharged. Overnight, however, she was woken by two further episodes of retrosternal chest pain. The patient presented to the A&E department, where the chest pain resolved within minutes of sublingual glyceryl trinitrate administration. An ECG while she was pain free again showed no ischaemic changes (fig 1). Her medications at the time were lisinopril, the oral contraceptive pill, as well as capecitabine. She had two cardiac risk factors: hypertension and a family history. Vital observations, clinical examination, chest x ray and cardiac markers (troponin I and creatine kinase) were all normal.

While on the Emergency Admission Unit she developed further chest pain at rest, 10 minutes after capecitabine ingestion. An ECG recorded at this time showed ST elevation in all the chest and limb leads (fig 2). Buccal suscard, aspirin 300 mg and a glyceryl trinitrate infusion were started. Subsequently the pain resolved and the ECG returned to normal. Capecitabine-induced cardiac pain was suspected, therefore the drug was withdrawn. The patient remained on a nitrate infusion on the cardiac care unit for 48 h, where she remained pain free. Further ECG and an echocardiogram were all normal, as were subsequent troponin I measurements.

Figure 2 ECG during episode of chest pain.


This case report clearly demonstrates the development of cardiac chest pain occurring at rest in a premenopausal woman with minimal cardiac risk factors on capecitabine therapy.

5-FU, the active compound to which capecitabine is converted in vivo, is known to have an adverse effect on the heart: hence 5-FU or its metabolites are likely to be responsible for the cardiotoxic effects of capecitabine. The mechanisms of 5-FU-induced cardiotoxicity that have been proposed are coronary artery vasospasm,1 direct myocardial injury and thrombogenesis.

Coronary vasospasm associated with capecitabine has previously been reported in the literature.24 The symptomatic relief and resolution of the ST segment changes with vasodilator therapy are highly suggestive of vasospasm. In addition, these symptoms did not return after the withdrawal of capecitabine. By comparison with previous cases, however, our patient, with no previous ischaemic heart disease, presented with chest pain at rest and not on exertion.3 4 The exact pathogenesis behind coronary artery vasospasm is yet to be discovered but it has been hypothesised that endothelin-1, a potent vasoconstrictor, may act as a mediator.4

Capecitabine causing an acute thrombosis within the coronary vasculature is improbable given that the clinical features were reversible with nitrates, cardiac enzymes were normal and echocardiogram was normal; there was no irreversible damage to the myocardium. Furthermore, the unique ECG finding of ST elevation in all leads, which to our knowledge has not previously been reported, would imply thromboses in all three major vessels. This scenario is more than unlikely and again supports the hypothesis of coronary vasospasm. Although no coronary angiogram was performed, earlier cases of capecitabine-associated chest pain have documented normal angiograms.3

Direct damage to the myocardium or its blood vessels caused by 5-FU has also been suggested.2 It could be that 5-FU metabolites induce metabolic changes that produce hypoxia and thus create a transient, reversible myocardial ischaemia.


As capecitabine is used in the treatment of the two commonest cancers, medical practitioners are likely to encounter the toxic effects on the cardiovascular system in their patients and so should be aware of these and how to manage them appropriately. Coronary vasospasm remains a feasible explanation for the mechanism of capecitabine-induced cardiotoxicity.


The authors would like to thank the Medical Illustrations and Imaging Department at the Norfolk and Norwich University Hospital.


View Abstract


  • Competing interests: None declared.

  • Patient consent: Written consent for publication of this case was obtained from the patient.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.