Low Risk of CHF from Imatinib

Abstract & Commentary

By Andrew Artz, MD, MS, Division of Hematology/Oncology, University of Chicago, Chicago, IL. Dr. Artz reports no financial relationship to this field of study.

Synopsis: Imatinib has recently been reported to cause CHF, but the clinical incidence is unclear. The investigators reviewed 1276 imatinib treated patients on clinical trials at a single institution and found 22 cases (1.7% incidence) of CHF symptoms and/or low ejection fraction (EF). Most cases (18/22) occurred in those having pre-existing cardiac risk factors. CHF was more common in older patients. Imatinib dose did not correlate with CHF. In summary, the incidence of CHF in imatinib treated patients was similar to the expected population incidence. Prospective trials are warranted to precisely determine the incidence and attribution of imatinib to CHF.

Source: Atallah E, et al. Congestive heart failure is a rare event in patients receiving imatinib therapy. Blood. 2007; 110:1233-1237.

The Philadelphia chromosome creates the Bcr-Abl fusion protein, which characterizes chronic myeloid leukemia (CML). The tyrosine kinase inhibitor imatinib (Gleevec™, originally known as STI-571) competitively inhibits bcr-abl activity as well as other signal transduction pathways. The high activity and low toxicity lead to rapid incorporation into clinical practice.1 Concern about serious cardiac toxicity recently emerged when a series of 10 cases of CHF during imatinib treatment were reported.2 Animal models supported a pathophysiologic link. The Abl protein plays a role in cardiac cell maintenance, which is perturbed when exposed to imatinib. The lack of confirmation and estimates of incidence or risk factors have left a major void for this now commonly used treatment. In this report, Atallah and colleagues from the MD Anderson Cancer Center report on the incidence of cardiac toxicity during imatinib treatment.

Records from patients receiving imatinib from 1998 to 2006 on clinical trials were reviewed. The analysis focused on cardiac symptoms, testing (left ventricular ejection fraction), cardiac risk factors, and serious and/or unexpected adverse events. CHF was defined using the Framingham criteria or the development of asymptomatic low ejection fraction (<50%).

There were 1276 patients receiving imatinib on clinical trials. The median age was 52 years (range 15-84) with a median follow-up of almost 4 years. Twenty-two patients developed CHF during therapy. Five of these cases were previously reported in the initial publication on imatinib cardiac toxicity.2 The overall incidence was 1.7% (22/1276). The ejection fraction was objectively evaluated in 15 cases, 9 of whom showed an EF below 50%. CHF incidence increased with advancing age: 0% if under 45 years, 0.3% (01/322) between 45-55 years, 1.7% (5/291) for ages 56-65 years, 2.8% (6/211) in the 66-75 cohort, and 9.3% (4/43) when aged 76 to 85 years. Imatinib dose showed no relationship to developing CHF. Prior cardiac risk factors were present in 18/22 patients in whom CHF developed.

COMMENTARY

Much concern has been generated since investigators reported a series of 10 CHF cases in imatinib treated patients. The postulated mechanism of CHF was perturbation of the Abl protein's normal role in cardiac maintenance.2

In this retrospective review of over 1000 thousand imatinib treated patients at a single institution, the authors suggest imatinib induced congestive heart failure is uncommon. Only 1.7% of patients developed CHF as assessed by symptoms and/or low EF. Older age and pre-existing cardiac conditions were the major risk factors. The median age for cases was 70 years vs 52 years for those without CHF (P= 0.001). Most patients who developed CHF (18/22) had at least one cardiac risk factor.

As an observational study, patients were not prospectively followed. The 1.7% incidence may be an underestimate as ascertainment may be incomplete or subtle cardiac symptoms may be missed. Further, as a cohort of patients enrolled in clinical trials, one can assume better health than the average patient treated in community clinical practice. The relatively young age (median 52 years) supports this notion. Alternatively, imatinib induced cardiac damage may be rare to non-existent. The incidence of CHF was similar to the population incidence of CHF and imatinib symptoms of edema or fatigue could be mistaken for CHF, leading to misclassifying cases as CHF. Other large series of imatinib treated patients mirror the results reported by Atallah.3,4

How do we reconcile the seemingly conflicting results? The true incidence requires a prospective analysis of imatinib-treated patients for CHF. Although the exact incidence remains unknown, the data are reassuring that serious CHF occurs infrequently in imatinib treated patients. These findings translate into clinical practice. The potential risk of CHF should not prevent imatinib treatment when indicated. For those who develop CHF symptoms during imatinib therapy, a cardiac evaluation is necessary irrespective of imatinib treatment. The role of cardiac biopsy is unknown although a specific imatinib associated pattern of damage has been suggested. In most cases, discontinuation of imatinib is not necessarily required, although severe CHF and/or lack of cardiac risk factors would prompt considering alternative medications. We must keep in mind that other treatments have toxicities as well and may not necessarily be safer. For example, dasatinib has been approved for CML; however, this tyrosine kinase inhibitor may lead to pericardial or pleural effusions, concerning side effects when switching because of CHF. Moreover, the new generation of tyrosine kinase inhibitors (eg, dasatinib, nilotinib) have a theoretically risk of CHF as they also inhibit the Abl protein.

In conclusion, imatinib induced CHF is uncommon but prospective trials will be required to define the precise incidence and attribution.

References:

1. Druker BJ, et al. N Engl J Med. 2001;344:1031-1037.

2. Kerkela R, et al. Nat Med. 2006;12:908-916.

3. Verweij J, et al. Eur J Cancer. 2007;43:974-978.

4. Hatfield A, et al. In reply to 'Cardiotoxicity of the cancer therapeutic agent imatinib mesylate'. Nat Med. 13:13; author reply 15-6, 2007.