Looking down the road towards heart health

From the Cancer Epidemiology Research Cooperative, Division of Public Health Sciences

With improvements in long-term survival rate of childhood cancers in recent decades, increasing attention has focused on understanding cancer treatment-related health effects that may not appear until years or decades later. This research revealed that the cumulative exposure to some antineoplastic agents commonly used to treat these cancers are associated with morbidity and mortality later in life. For example, several anthracyclines and the anthraquinone mitoxantrone are associated with adult cardiomyopathy. Previous work by Fred Hutch researchers demonstrated that the anthracycline daunorubicin is only half as cardiotoxic as doxorubicin, one of the most widely used anthracyclines. However, differential cardiomyopathy risk among a wider panel of these agents is not yet fully understood. In a new study recently published in the journal JAMA Oncology, researchers from the Public Health Sciences and Clinical Research Divisions report associations of five cancer treatment agents with late-onset cardiotoxicity in childhood cancer survivors. This study is from Dr. Eric Chow of the Cancer Epidemiology Research Cooperative, with Dr. Wendy Leisenring and Ms. Kayla Stratton of the Childhood Cancer Survivor Study (CCSS) leading the analysis efforts along with Drs. Lieke Feijen and Leontien Kremer, their collaborators in the Netherlands.

Clinical studies with very long follow-up times present unique challenges for researchers. One of the most important breakthroughs that enabled the conduct of this study, as described by Dr. Chow, “was really the long-term commitment by many groups in the US and elsewhere, to form large cohorts of childhood cancer survivors and be able to follow them for long-term outcomes, i.e., outcomes that occur far beyond the typical 5-year time frame most cancer protocols cover. For this study, we examined heart failure by age 40, well after these individuals were treated for their cancer as children.” Collaboration among several groups is also frequently critical for success of trials of this nature. “And as important, the willingness of multiple groups, like CCSS, the Dutch DCOG-LATER cohort, and St Jude Lifetime Cohort, to collaborate and pool their data in this fashion. As otherwise, the signal within any single cohort would likely have been too weak to have been detected,” added Dr. Chow.

Graphical representation of the relative risk of cardiomyopathy associated with epirubicin and mitoxantrone compared to treatment with doxorubicin in survivors of childhood cancer.
Relative risk of cardiomyopathy associated with epirubicin (panel A) and mitoxantrone (panel B) compared to treatment with doxorubicin in survivors of childhood cancer. Image provided by Dr. Eric Chow

These collaborative efforts allowed the researchers to study a cohort of nearly 30,000 childhood cancer survivors who were younger than twenty-three years of age at time of diagnosis. The authors assessed total chest radiotherapy exposure and cumulative doses of four anthracyclines (doxorubicin, daunorubicin, epirubicin, and idarubicin) and mitoxantrone. The median follow-up time after diagnosis was twenty years. The authors aimed to determine cardiotoxicity equivalence ratios for each of these agents as compared to doxorubicin.

The overall cumulative incidence of severe, life-threatening, or fatal cardiomyopathy by the age of 40 was 3.4%. The authors assessed hazard ratios for cardiomyopathy of each drug across dose categories. The mean equivalence ratio between daunorubicin and doxorubicin was 0.6, indicating the cardiotoxicity of daunorubicin is nearly half that of doxorubicin. The mean equivalence ratio for epirubicin was 0.8, suggesting its cardiotoxicity is similar to doxorubicin (see figure, top panel). There were few participants that received idarubicin and experienced cardiomyopathy, but results from the middle dose category indicated a hazard ratio of 0.9. Surprisingly, the mean equivalence ratio for mitoxantrone was 10.5, indicating it is more than ten times as toxic as doxorubicin (see figure, bottom panel).

Dr. Chow summarized the results, “The most significant finding is these results challenge the conventional wisdom that hematologic toxicity is equivalent to cardiac toxicity, and our results would suggest that the cardiotoxicity equivalence may be quite different, at least in childhood cancer patients. In relation to doxorubicin, we found that daunorubicin appears to be half as (i.e., less) cardiotoxic than previously thought, whereas for mitoxantrone, twice as (i.e., more) cardiotoxic than previously thought.”

 “This could have implications for various treatment protocols that use these agents, for example, AML, as they now need to balance the potential oncologic efficacy with short-term hematologic toxicity as well as differential long-term cardiac toxicity,” emphasized Dr. Chow. While this study clearly moves the field forward in understanding long-term risks associated with the use of common cancer therapies, additional questions remain. “Our other work is examining the long-term efficacy of cardio-protectants given concurrently with chemotherapy. Since many of these anthracyclines remain in common use today, an effective cardio-protectant would enable more children to receive these treatments while mitigating long-term cardiac toxicity/heart failure,” said Dr. Chow. To gain mechanistic insight into anthracycline-mediated cardiotoxicity, the authors also plan to collaborate with other research groups to study these questions in animal models.

This work was supported by the National Cancer Institute.

Fred Hutch/UW Cancer Consortium member Dr. Eric Chow contributed to this research.

Feijen EAM, Leisenring WM, Stratton KL, Ness KK, van der Pal HJH, van Dalen EC, Armstrong GT, Aune GJ, Green DM, Hudson MM, Loonen J, Oeffinger KC, Robison LL, Yasui Y, Kremer LCM, Chow EJ. 2019. Derivation of anthracycline and anthraquinone equivalence ratios to doxorubicin for late-onset cardiotoxicity. JAMA Oncology. doi: 10.1001/jamaoncol.2018.6634.