Atherosclerosis Vindicated: A Case of Chest Pain Due to Capecitabine-Induced Coronary Artery Spasm

Background

Capecitabine is a chemotherapy prodrug that, when converted to its active form 5-fluorouracil (5-FU), effectively interferes with DNA synthesis and cell division leading to cancer cell death [1]. 5’-deoxy-5-fluorouridine, a metabolite of capecitabine, is converted to 5-FU by thymidine phosphor-ylase, which is expressed in higher concentrations in tumor cells, leading to improved tumor-targeting specificity and decreased systemic exposure [2]. Capecitabine is a widely prescribed agent for solid tumors, with recent studies suggesting that capecitabine is projected to expand to a global market size of two billion dollars [3]. It is considered a first-line chemotherapy option for advanced breast cancer, and a first-line treatment option for advanced colorectal cancer [2]. Its convenient oral formulation has aided in making this drug increasingly popular while retaining efficacy comparable to intravenous fluorouracil. Despite its many benefits, capecitabine has extensive list of adverse effects. Coronary artery vasospasm and other cardiac adverse effects of capecitabine are uncommon and under-recognized [4,5]. The incidence of coronary vasospasm in patients being treated with 5-FU varies between 1.2% to 18% depending on the mode of medication administration and pre-existing cardiac disease [6]. However, for capecitabine, this adverse effect has only been reported in case reports, so its frequency is not well established. Thus, it is critical to draw attention to this potentially fatal adverse effect of a very commonly used cytostatic agent.

Case Report

A 45-year-old man presented to our hospital’s emergency department with new-onset exertional chest pain. He described it as substernal and crushing, noting that it started during an outdoor walk and improved with rest. It had resolved by the time he was seen. Before this episode, he was able to walk without limitation and complete his activities of daily living.

He did not have a history of coronary disease or other cardiac conditions. Notably, he was a prior cigarette smoker, reportedly quitting 4 years ago with a 20-pack-year smoking history. He had an active diagnosis of a well-differentiated, high-grade pancreatic neuroendocrine tumor. He was previously treated with pancreatectomy and splenectomy followed by lanreo-tide. The disease was noted to have recently progressed despite these therapies, and the patient was started on a regimen of capecitabine and temozolomide (CAPTEM) only 7 days before the onset of his chest pain. He was not on any other new or long-term medications or supplements that would predispose to angina.

Initial evaluation included a 12-lead electrocardiogram (ECG), cardiac enzymes, other basic labs, and a contrasted chest computed tomography scan, which was performed to rule out a pulmonary embolism. His ECG was normal, high-sensitivity troponin T was 17.0 ng/L and 17.8 ng/L, and all other evaluations were within normal limits. With his recent initiation of capecitabine, there was rising concern that this pain was an-gina resulting from coronary vasospasm. For further risk stratification and as an attempt to reproduce this patient’s symptoms, a graded exercise test (GXT) was performed. The patient’s resting ECG demonstrated normal sinus rhythm with baseline nonspecific ST abnormalities most consistent with early repolarization (Figure 1A). After 9 min of the standard Bruce protocol, the patient developed this presenting substernal, crushing chest pain. The exercise ECG demonstrated corresponding hyperacute T waves (Figure 1B). The patient was immediately placed in recovery and subsequently developed ST segment elevation in his inferior and lateral leads, which quickly resolved (Figure 1C, 1D). His symptoms rapidly resolved with rest, and no medication was required. An accelerated idioventricular rhythm was captured further into recovery. Due to these findings, a coronary angiogram was performed to rule out obstructive coronary artery disease. This demonstrated angio-graphically normal vessels of normal caliber without luminal irregularities (Figure 2).

Further provocative testing for coronary reactivity and spasm was considered but was deferred given the high pretest likelihood of vasospasm. Additionally, provocative testing can rarely lead to clinically significant ischemia and ventricular arrhythmias despite prompt administration of intracoronary nitroglycerin and should therefore be performed judiciously.

Overall, the patient’s symptoms, GXT findings, and the timing between capecitabine initiation and symptom onset solidified the diagnosis of capecitabine-induced coronary vaso-spasm. After a review of his presentation, cardiac evaluation, and discussion with the patient and his oncology team regarding the risks for potential irreversible cardiac injury, the decision was made to discontinue capecitabine therapy completely. Multidisciplinary discussion and patient-centered decision-making were crucial in determining whether to continue capecitabine despite the cardiovascular risk, given its potential benefit in his cancer treatment. He continued with temozolomide monotherapy. A GXT was repeated 4 weeks after capecitabine discontinuation, and the patient was able to exercise for 10 min on the standard Bruce protocol without pain or electrical changes. At that visit, he noted that he had not experienced any further episodes of chest pain and had been able to restart exercise without limitation. Unfortunately, the patient had a progression of disease on single-agent temozolomide and was subsequently evaluated for alternative chemotherapy treatment options.

Discussion

This case report highlights the potential for coronary artery vasospasm induced by capecitabine and emphasizes the importance of promptly discontinuing this medication when this complication occurs. Coronary artery vasospasm can occur within 24 h of initiating treatment, and patients remain at increased risk for coronary artery vasospasm throughout the entire treatment course [7]. If left untreated, there is a rare but serious risk of myocardial ischemia and, potentially, infarction. Given the prevalence of coronary vasospasm, it may be considered to not pursue invasive testing in patients who are unlikely to have atherosclerotic coronary disease. Given the increased risk for coronary disease due to prior tobacco abuse, our patient underwent both a noninvasive and invasive assessment to fully characterize his chest pain [8,9]. Ultimately, through shared clinical decision-making, our team decided to discontinue capecitabine due to our high index of suspicion for vasospasm, to prevent irreversible harm.

This patient had a significant cigarette smoking history, possibly contributing to the risk for vasospasm along with capecitabine use [10]. Cigarette smoke contains toxins that can damage the vascular endothelium, including the coronary arteries. This damage is a nidus for vasospastic angina and the cascade of ischemic events it may cause. Smoking is also associated with thicker arterial walls and an increased plaque burden, exacerbating the effects of vasospasm. Patients should universally be counseled to stop smoking, especially individuals who are receiving agents like capecitabine that are linked to vasospasm [10].

During exercise testing, our patient developed hyperacute T waves. Unlike the T waves of a healthy individual, these are tall relative to the size of the preceding QRS complex. They are often the first sign of acute myocardial injury such as seen in the early stage of an ST-elevation myocardial infarction, and often precede the development of ST elevation or Q waves. Our exercise tracings demonstrate this exact phenomenon; the onset of hyperacute T waves precedes any ST segment changes and serves as a harbinger of injury coupled with the patient’s chest pain symptoms. It is important to note that similar ECG changes can result from acute plaque rupture or vasospasm. Since the GXT reproduced the patient’s symptoms and showed typical ECG changes, and since the subsequent coronary angiogram was normal, additional diagnostic testing would not have altered medical management. Sufficient evidence existed to confidently attribute the chest pain to capecitabine.

Though not fully understood, capecitabine is postulated to induce vasospasm through interactions with vascular endothelium and smooth muscle. In previous studies, capecitabine has been shown to impair nitric oxide synthesis and inhibit vaso-dilation, thus leading to endothelial dysfunction and smooth muscle constriction [11]. Another possible mechanism of vasospasm is that capecitabine has been shown to increase the levels of endothelin-1, a known vasoconstrictor [12]. This combination of impaired nitric oxide synthesis and overproduction of endothelin-1 can reduce myocardial blood flow and lead to the classic symptoms of ischemic chest pain.

Capecitabine and its active metabolites have short half-lives and do not accumulate significantly after repeat administrations, which may suggest that toxicities such as coronary vaso-spasm can be expected to resolve quickly with the discontinuation of the medications [13]. Regardless, patients experiencing drug-induced coronary vasospasms should have their exercise restricted and be closely monitored after drug discontinuation, though no exact duration has been recommended.

There are some reports in the literature of patients experiencing 5-FU-induced coronary vasospasm who were successfully managed with a reduced dose of 5-FU with nitrates or a calcium channel blocker, or with a transition to capecitabine with nitrates or a calcium channel blocker [14]. It is unclear if the addition of a nitrate or calcium channel blocker is a safe alternative to discontinuation in cases of vasospasm. This treatment approach would require very close monitoring. More recently, patients have been successfully transitioned to Teysuno, a similar fluoropyrimidine analog, with notably reduced rates of vasospasm. This has resulted in the European Society of Medical Oncology recommending that patients who experience cardiovascular toxicity during therapy with capecitabine or 5-FU be switched to Teysuno without any attempt to rechallenge patients with capecitabine or 5-FU [15]. However, this medication has not yet been approved in the United States for treatment of pancreatic neuroendocrine tumors as seen with our patient [16].

Conclusions

This case serves to remind clinicians of the rare but clinically significant adverse reaction to a medication that is being more frequently used. Capecitabine-induced coronary vasospasm is rare, but if left untreated can lead to myocardial ischemia and possible infarction. In our case, the coupling of noninvasive and invasive evaluations along with multidisciplinary discussions was paramount in deciding the most appropriate management for our patient’s neuroendocrine tumor and the ultimate cessation of capecitabine. For this class of medication, a high index of suspicion and prompt intervention by the non-oncologist can mitigate serious cardiovascular complications and improve patient outcomes.