Fatal Heart Failure After a 26-Month Combination of Tyrosine Kinase Inhibitors in a Papillary Thyroid Cancer

Patient

A62-year-old Polynesian man was found to have chronic myeloid leukemia in June 2007. He was moderately obese with a body mass index of 34, but otherwise he did not have any cardiovascular risk factors (no past or present smoking, diabetes, hypertension, hyperlidemia, or relevant familial cardiovascular disease), or history of receiving cardiotoxic agents. He was treated with imatinib (Gleevec, Novartis, Switzerland), 400 mg daily, and a durable complete cytogenetic response was obtained. Concomitantly, he was discovered to have papillary thyroid carcinoma with lung and cervical lymph node metastasis (Fig. 1a). A total thyroidectomy with ipsilateral cervical lymph node dissection [pT3N1b (17N+/27N)M1 according to the last TNM classification system (1)] was performed. Lung metastasis did not take up ¹³¹I on whole-body scans. Biopsies of lung metastasis were not done because the plasmatic thyroglobulin level was very high before radioiodine treatment (4100 μg/L). In November 2007 the patient was started on sorafenib (Nexavar, Bayer, Germany) at 400 mg twice a day based on the promising effects of vascular endothelial growth factor (VEGF) receptor inhibitors in thyroid carcinoma (2). Six months later, the dose was decreased to 600 mg daily because he was experiencing grade II diarrhea and grade II hand-foot syndroma. After 24 months of treatment, he achieved a durable partial response (Fig. 1b) according to Response Evaluation Criteria in Solid Tumors (RECIST) (3).

OPEN IN VIEWER

FIG. 1.

Computed tomography (CT) scan evaluation of lung metastasis of the thyroid cancer. The sequential CT scans performed at the beginning of the combination of imatinib and sorafenib (a), and after 24 months of treatment (b) showed a partial response according to Response Evaluation Criteria in Solid Tumors (RECIST) criteria.

In July 2009, 21 months after he started this combination of imatinib and sorafenib, the patient developed a de novo acute coronary syndrome without ST elevation on electrocardiogram (EKG). EKG showed left anterior hemiblock, without repolarization abnormalities. Laboratory findings revealed elevated levels of troponin I (56 μg/L), and creatine-kinase MB (660 UI/L), and the coronary angiography disclosed a tritruncal stenotic coronary atheroma with subcoronary occlusion of the distal left circumflex. Echocardiography showed a moderate left ventricular hypertrophy, without segmental kinetic changes. It was decided not to do angioplasty or surgery because of the tritruncal stenosis with a poor vascular bed distal to the lesions and his poor prognosis related to thyroid cancer. His median progression-free survival was considered to range from 14 to 21 months based on phase II studies (4 –6). He was started on daily doses of clopidogrel 75 mg, atenolol 50 mg, perindopril 2 mg, atorvastatin 80 mg, and acetylsalicylate 75 mg for his cardiac disease. In the next 7 months coronary angiography was not repeated because of absence of new cardiac symptoms.

Seven months later, in February 2010, while the patient was undergoing a 18-fludeoxyglucose positron emission tomography scan (18-FDG PET scan) for evaluation of the thyroid cancer, he suffered acute thoracic pain. Despite an immediate transfer into the intensive care unit, he developed an extensive anterior acute myocardial infarction with cardiogenic shock and died 1 hour later.

His last 18-FDG PET scan indicated major left ventricular dilatation and lack of 18-FDG captation (Fig. 2c). Retrospectively, we found on the related 18-FDG PET images obtained in November 2008 (12 months after the beginning of the combined therapy and 8 months before the first clinical acute coronary syndrome) a de novo lack of 18-FDG captation in the apical and med-septal regions (arrows) corresponding to the territory of the left anterior descending artery (Fig. 2b), which was not seen on the normal imaging before tyrosine kinase inhibitor (TKI) therapy (Fig. 2a).

OPEN IN VIEWER

FIG. 2.

Short axis, vertical long axis, and horizontal long axis sequential images of cardiac 18-FDG positron emission tomography (18-FDG PET) scan. Normal myocardial metabolism before combined imatinib and sorafenib treatment (a). The cardiac PET scan performed after 12 months of the combined therapy, and 8 months before the first clinical acute coronary syndrome evidenced in the relatives static images a de novo lack of 18-FDG captation in the apical and med-septal regions (arrows) corresponding to the territory of the left anterior descending artery (b). The last one performed after 28 months of the combined treatment, and few hours before the lethal cardiogenic shock showed a major left ventricular dilatation and lack of 18-FDG capitation (c).

Discussion

Imatinib and sorafenib are orally administered inhibitors of receptor tyrosine-kinases, mainly platelet-derived growth factor receptor, and c-kit receptor for imatinib, BRAF, VEGF receptor 1,2, and c-kit receptor for sorafenib. In three recent phase II studies of patients with metastatic refractory thyroid cancers, sorafenib demonstrated promising clinical activity (4 –6). TKIs have been associated with cardiac toxicity, usually reversible and not severe, as the incidence of cardiotoxic mortality ranges between 1% and 3% (7). However, with expanded clinical experience, the incidence of cardiac events could be higher than previously reported (8). Imatinib may lead to severe left ventricular dysfunction with congestive heart failure, due to alterations in mitochondria function and cardiomyocytes apoptosis (9). In the TARGET study, acute coronary syndromes, including myocardial infarction, were observed in 3% of patients in the sorafenib arm compared to <1% in the placebo arm (10). A central role of RAF1 and BRAF inhibition has been proposed (7), but recent data suggest that mechanisms of sorafenib-induced cardiotoxicity might be due to inhibition of combinations of kinases different than RAF1 or BRAF (11). Interestingly, targeting c-kit with both imatinib and sorafenib might interfere with cardiac repair after myocardial infarction (12). There has been recent interest in using combinations of TKIs to minimize the possibility of resistance to their anti-neoplastic effects. However, the use of new combinations of TKIs is challenging because of the possibility of synergism as far as adverse events are concerned (13). Here we reported a first case of possible cumulative and lethal cardiac toxicity in a patient treated for a prolonged period with imatinib and sorafenib. To our knowledge, there is no other reported case of severe cardiac toxicity with a combination of TKIs.

18-FDG PET scan is a promising tool to evaluate tumor response to new drugs such as those that target the VEGF pathway (14). However, cardiac PET scan analysis is not usually performed in patients being treated with TKIs for advanced disease in oncology since cardiac uptake with FDG imaging is considered as physiological and therefore without clinical relevance.

In retrospect, cardiac PET scan analysis should have been performed at the same time as that the whole-body PET scan was interpreted in our patient. Indeed, in our patient, cardiac PET scan was probably predictive of cardiac toxicity since an 18-FDG defect in the territory of left anterior descending artery appeared 8 months before the first coronary symptoms.

There are new imaging strategies to assess cardiac status, including quantitative gated 18-FDG PET, when ischemic heart disease is suspected. These procedures can be used in patients with left ventricle dilatation to assess global impairment of myocardial metabolism, but require specific cardiac protocols (15). To date, there is no study assessing the predictive value of sequential cardiac PET scans in patients receiving cardiotoxic chemotherapeutic agents. Potentially useful information is likely to be that related to vascular uptake. In patients being treated for cancer it has been shown that increased 18-FDG uptake in major arteries as well as calcified plaques in the same vessels could be predictive of a subsequent vascular event (16). We did not observe any similar images in our patient.

Conclusion

Systematic cardiac follow-up should be considered for patients receiving long-term treatment with TKI monotherapy or combinations of TKIs. If a whole-body 18-FDG PET scan is performed, cardiac PET scan analysis could be taken into account to help in the prediction and management of TKI-related cardiotoxicity.

There is a need for developing a protocol that may be applied to those patients undergoing an oncology PET scan study; this would permit a systematic observation of the myocardial metabolic status. A prospective study is necessary for evaluating the role of cardiac reading when performing oncology FDG-PET/computed tomography, especially in patients at high risk of developing cardiac disease secondary to oncologic drug use. Strict follow-up is required in patients receiving TKIs for oncological diseases.