Cardiotoxicity in Cancer Patients: Often More Malignant Than Cancer—Part 1

[Hopeful]

Interview by L Scott Zoeller. 2012 Aug 24, Joerg Herrmann, MD


In Part 1 of this interview, Dr. Herrmann discusses the need for specialized experience in cardio-oncology and optimal use of imaging and biomarkers in monitoring cardiac damage associated with cancer therapies. In Part 2, Dr. Herrmann will discuss the evidence behind cardio-protective agents and patient lifestyle.

The emerging field of cardio-oncology

Dr. Herrmann:Every so often, you encounter some skepticism regarding the emerging discipline of cardio-oncology. The very reason, however, this discipline is emerging is the impressive outcome improvement for a number of malignancies over the past decades. In fact, survival outcome of a number of cancers is now better than that of heart failure. In this sense, heart failure is, indeed, more malignant than cancer.

Years ago, we would have not had this conversation; but now, with a number of malignancies being turned into chronic diseases, similar to, for instance, rheumatoid arthritis, collateral damage matters, and this cannot be ignored anymore. So, I’m happy that this topic is receiving more and more attention to improve the overall survivorship of our patients.

OncologySTAT:Would you discuss some of the challenges of treating cancer patients with cardiovascular comorbidities?

Dr. Herrmann: There is a great article which illustrates these challenges very nicely. It was published in the Journal of the National Cancer Institute in January 2010, authored by colleagues from the oncology unit in Milan, Italy. They coined it the “sliding door” concept, and it is as follows.¹

Let’s consider a patient, 75 years of age, with occult colorectal carcinoma and ischemic heart disease, both of which not known at the time of the initial visit. If the patient slides through the door of cancer screening first, then he or she will undergo fecal occult blood testing, and then, if that’s positive, a colonoscopy. At the end, the patient will be diagnosed with colorectal cancer—hopefully early—and undergo surgery and even chemotherapy, which can lead to profound cardiovascular side effects to the point of heart failure, especially if there is known or unknown pre-existing ischemic heart disease. At that point, the patient gets referred to the cardiologist, which is less than ideal.

If, on the other hand, the patient, for whatever reason, gets seen by a cardiologist or a cardiology-minded internist early on—before any cancer screening—he or she might end up with a stress test. If that test is positive, the patient is sent to the catheterization lab, and, before we know it, that patient has multiple drug-eluting stents in place, is on dual antiplatelet therapy, and, after a couple of months, develops a gastrointestinal bleed. After that delay, the patient is eventually diagnosed with colon cancer, but, possibly, due to its later stage, it is found that he or she has metastatic disease, with a potentially less optimal outcome.

So, based on these illustrating scenarios, a move has been made to view the patient more in his or her entirety, rather than through the lens of a clinician’s own subspecialty, and this has been coined “cardio-oncology” and evolved toward cardio-oncology as a new discipline. The two particular challenges that this discipline is dealing with are: number one, to help patients with known cardiovascular disease undergo the most aggressive forms of cancer therapy safely; and, number two, to optimally manage patients who develop cardiovascular disease as a consequence of cancer therapy.

OncologySTAT: What treatment modality and setting do you think is most responsible for preventable cardiotoxicity in cancer patients?

Dr. Herrmann: That’s a challenging question. If we take one aspect, and just look at the relative frequency of use and the incidence of cardiotoxicity, anthracyclines would still be at the top of the list, followed by cyclophosphamide and Herceptin. If, on the other hand, we consider myocardial ischemia, then cisplatin and 5-FU must be mentioned, and the two together have additive effects in this regard.

With regard to the setting, one can easily imagine that anything that stresses the cardiovascular system before the initiation of these chemotherapeutic agents would raise the risk for any of these side effects. So, there would be a greater risk for cisplatin- and 5-FU–induced myocardial ischemia for patients with diseased coronary arteries. There is also a greater risk for cardiomyopathic cardiotoxicity with coronary artery disease (CAD), pre-existing cardiomyopathy, and even cardiovascular risk factors, in particular, hypertension. Furthermore, the risk for cardiotoxicity is a higher at either end of the age spectrum and for women. Obviously, combination therapies, especially radiation therapy in combination with chemotherapeutic agents, pose a higher risk for cardiotoxicity.

Now, regarding the modality and setting most often associated with preventable cardiotoxicity, this is not so easy to address. These patients need treatment from an oncologic or hematologic standpoint, and, hence, sometimes cardiovascular side effects cannot be completely avoided; however, they can be minimized. There are scenarios where a different regimen can be given; hence, risk stratification beforehand is very important. Otherwise, strategies to reduce cardiotoxicity risk with anthracyclines, as many know, entails dose reduction, continuous infusion, and liposomal delivery. With regard to preventive drugs, dexrazoxane can be considered. However, carvedilol, nebivolol, an ACE inhibitor, and an angiotensin receptor blocker may be considered for prevention as well.

Beyond clinical risk factors—predictive biomarkers and imaging

OncologySTAT: What is the potential for molecular risk profiling to identify patients at a high risk for treatment-induced cardiotoxicity?

Dr. Herrmann: I think this is really at the core of what is needed. I would love to do this study because there are inter-individual differences in the incidence of cardiotoxicities: some patients seem to tolerate the highest dosages and others cannot even tolerate their first round of chemotherapy. So, there must be something beyond the clinical risk factors mentioned earlier, and it points to some molecular, and, very likely, genetic makeup. If we were to decode this makeup, we would be ahead of the curve in determining the cardiovascular risk related to cancer therapy and might guide therapy in a much better way.

So far, there are only a few studies that have looked at this, but what they found is really fascinating. The genetic modifications that have been found to influence the individual risk for the development of anthracycline-induced cardiotoxicity encode for variants of anthracycline metabolism and cellular transport, as well as free radical metabolism, namely NADPH oxidase.
There was an article published in the Journal of Clinical Oncology last year from the Canadian Pharmacogenomics Network for Drug Safety, which showed that testing for various single nucleotide polymorphisms (SNPs) can substantially improve cardiotoxicity risk prediction beyond clinical risk factors.² In statistical terms, the area under the curve (AUC), or c-statistic, improved from 0.68 in a model using clinical factors alone vs 0.87 using clinical factors plus SNPs. This translates to 75% accuracy in predicting the occurrence of cardiotoxicity. The greatest value is in identifying which patient will not develop cardiotoxicity, with an impressive negative predictive value of 96%.

So, in the era of individualized medicine, I think that pharmacogenomics is clearly the way to go; but, the crucial point will be affordability. Cost is where things will have to fall into place. The question is whether we can finance further studies, because these are trials that require a large volume of patients. And, then, at the end, if we go to the bedside, we have to consider the cost for the patient or the medical societies and agencies to cover these gene chips as well.

OncologySTAT: What is the optimal use of imaging and/or biomarkers in monitoring cardiac damage associated with various chemo- and targeted therapies? For instance, breast cancer patients receiving trastuzumab?

Dr. Herrmann: Well, various institutions have used various approaches, and, for sure, there’s no consensus on the optimal use of biomarkers; it’s an evolving field. With the highly sensitive cardiac troponin assays coming out, this may all be redefined; but, we don’t really have a large volume of data on it quite yet.

But, putting the studies that we have so far, and the pathomechanisms that we think are responsible, into perspective, I think it would seem logical to see cardiac troponin elevation early on, at the time of acute injury, and this may persist for some time. For instance, studies from Italy³ have shown that, even within the first 72 hours of aggressive high-dose chemotherapy, one can see troponin elevation. This early elevation predicts a reduction in left ventricular (LV) function 12 months down the road, particularly if these cardiac troponin elevations persist by 1 month after the end of chemotherapy.⁴ And this is something that we see in heart failure patients, in general: at the time of decompensation, for instance, they have small biomarker rises, or they even have persistent “troponemia,” as some call it, as a reflection of an ongoing cardiac injury. So, this is where cardiac troponin, as a biomarker, might be helpful, in detecting this, but it would be either early on, or, if the process persists.

NT-proBNP is a different biomarker altogether. Studies have shown that, with just the volume changes induced by chemotherapy, levels can be increased in nearly all of these patients. Hence, substantially higher levels are necessary to detect early on if someone might develop cardiomyopathy in the future. By any means, data on this are not as robust. Later on, brain natriuretic peptides (BNPs) might be helpful to reflect the cardiomyopathic process and determine if the patient’s shortness of breath is due to some cardiomyopathy process or pulmonary process, and so forth.

When it comes to imaging modalities, and going back to the one aspect of your question concerning breast cancer patients receiving Herceptin, Davinder Jassal’s group from Winnipeg looked at this, specifically. They followed such patients at 3-month intervals over a period of 12 months for the occurrence of cardiomyopathy (defined as a declining in LVEF to <50%)⁵. What they found was that, at the time points stated, biomarkers were not predictive of cardiomyopathy, but imaging was. Importantly, it was tissue Doppler-derived indices and strain imaging that emerged as the earliest indicators of an abnormality. Even more, a study by Stoodley from Australia indicated that abnormalities on strain imaging can be seen as early as within 1 week of completion of chemotherapy.⁶

These studies actually led the American Society of Echocardiography to develop guidelines on how patients are to be followed and likely will put some emphasis on strain imaging. These guidelines are being written at the moment, and, when they eventually are released, we’ll have more direction with regard to formal recommendations.

What we have presently is a number of clinical management algorithms, specifically for Herceptin, as reflected in your question. In essence though, the consensus of these documents is that all patients are required to have an initial assessment prior to the start of Herceptin, including an echocardiogram, and, depending on their ejection fraction, their monitoring interval will change. For instance, if the ejection fraction is less than 50%, the patient is followed with a physical exam as well as an echocardiogram every 6 weeks rather than 12 weeks. And, if further decline in the ejection fraction is noted, or if heart failure symptoms occur, Herceptin would be stopped for 3 to 4 weeks and heart failure therapy would be initiated. And then, depending on the dynamics thereafter, the patient resumes Herceptin, or not; but, either way, would be followed monthly.

When it comes to anthracyclines or chemotherapy in general, the consensus approach becomes a little bit more nebulous, and we actually have to rely on seasoned recommendations from the pediatric community. Hence, a lot remains to be done to give an answer to your questions when it comes to non–Herceptin based therapies.

References
1. Albini A, Pennesi G, Donatelli F, et al. Cardiotoxicity of Anticancer Drugs: The Need for Cardio-Oncology and Cardio-Oncological Prevention. J Natl Cancer Inst. 2012;102(1): 14–25.
2. Visscher H, Ross CJD, Rassekh SR, et al. Pharmacogenomic Prediction of Anthracycline-Induced Cardiotoxicity in Children. J Clin Oncol. 2012;30(13):1422-1428.
3. Cardinale D, Sandri MT, Martinoni A, et al. Left ventricular dysfunction predicted by early troponin I release after high-dose chemotherapy. J Am Coll Cardiol. 2000;36(2):517-522.
4. Cardinale D, Sandri MT, Colombo A, et al. Prognostic value of troponin I in cardiac risk stratification of cancer patients undergoing high-dose chemotherapy. Circulation. 2004;109(22):2749-2754.
5. Fallah-Rad N, Walker JR, Wassef A, et al. The utility of cardiac biomarkers, tissue velocity and strain imaging, and cardiac magnetic resonance imaging in predicting early left ventricular dysfunction in patients with human epidermal growth factor receptor II-positive breast cancer treated with adjuvant trastuzumab therapy. J Am Coll Cardiol. 2011;57(22):2263-2270.
6. Stoodley PW, Richards DA, Hui R. Two-dimensional myocardial strain imaging detects changes in left ventricular systolic function immediately after anthracycline chemotherapy. Eur J Echocardiogr. 2011;12(12):945-952.


Hopeful