Author + information
- Received December 4, 2006
- Revision received January 24, 2007
- Accepted February 6, 2007
- Published online June 26, 2007.
- Gurusher S. Panjrath, MD⁎,
- Virender Patel, MD⁎,
- Carolina I. Valdiviezo, MD‡,
- Navneet Narula, MD†,
- Jagat Narula, MD, PhD, FACC⁎ and
- Diwakar Jain, MD, FACC⁎,⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Diwakar Jain, Drexel University College of Medicine, 245 North 15 Street, MS 470, Philadelphia, Pennsylvania 19102.
Objectives The role of iron toward doxorubicin (DOX) cardiotoxicity was studied using a rodent model of dietary carbonyl iron loading.
Background Doxorubicin, a commonly used anticancer drug, is known to cause serious and potentially life-threatening cardiotoxicity. Doxorubicin cardiotoxicity is thought to be mediated through free-radical injury.
Methods Male Sprague Dawley rats fed iron-rich chow (n = 8) and regular chow (n = 8) were treated with DOX or saline (4 animals in each arm). Cardiotoxicity was assessed using mortality, weight changes, Tc-99m annexin-V imaging, histopathology, and immunohistochemistry.
Results Animals fed iron-rich chow showed significantly higher DOX cardiotoxicity as evidenced by greater weight loss (107 ± 14 g vs. 55 ± 10 g weight loss, p < 0.05), higher annexin uptake (0.14 ± 0.01% vs. 0.08 ± 0.01% injected dose/g of myocardium, p < 0.05), more severe myocyte injury on electron microscopy, and significantly higher cleaved caspase-3 staining compared with regular chow fed rats given DOX. Feeding iron-rich chow alone did not result in any cardiotoxicity.
Conclusions Dietary iron loading resulted in a substantially increased DOX cardiotoxicity in rats. Body iron stores as well as its bioavailability in tissue may be important independent predictors of susceptibility to DOX cardiotoxicity in man. Further clinical studies are warranted.
Doxorubicin (DOX), a potent anthracycline antineoplastic agent, is used widely in a variety of malignancies and tumors (1–3). Despite being one of the most effective anticancer agents, its use is limited by serious and sometimes life-threatening cardiotoxicity (4,5). Despite extensive studies in animals and cell culture models, the exact cellular, biochemical, molecular, and genetic mechanism of DOX cardiotoxicity is not fully known. Advanced age, concomitant treatment with radiation therapy, or high-dose cyclophosphamide and pre-existing heart disease have been identified as predisposing factors for DOX cardiotoxicity (4,6). However, these factors alone cannot explain the wide variation in individual susceptibility to DOX cardiotoxicity.
Redox injury and interference with protein synthesis are considered to play a significant role in DOX cardiotoxicity (7–9). Conditions that exacerbate free-radical formation may enhance DOX cardiotoxicity. Various studies have proposed a role of metallic ions in DOX-mediated damage (10–12). Doxorubicin interacts with metallic ions, especially iron, which results in the formation of DOX-iron (III) complex (13–15). Iron independently plays an important role in generation of harmful free radicals with potential deleterious effects on the myocardium. Iron chelator dexrazoxane is recommended for use in patients receiving higher doses of DOX to prevent heart failure (16,17). Despite these measures, cardiotoxicity continues to be an important limiting factor for DOX therapy.
We have aimed to elucidate the role of iron in potentiation of DOX-induced cardiotoxicity. We hypothesize that elevated body iron content enhances the cardiotoxic effects of DOX. We used an in-vivo rodent model of nutritional iron loading, which may simulate patterns of higher body iron stores observed in patients undergoing cancer chemotherapy.
Male Sprague Dawley rats (Charles River Laboratories, Wilmington, Massachusetts) were used in accordance with guidelines on animal care. The study was approved by the institutional animal care committee. The animals received a single intravenous dose of DOX. Further division was based on those fed regular rat chow or those fed iron-rich chow. Each group had a corresponding saline-treated control. Figure 1demonstrates the various study groups. There were 4 animals in each of the 4 study groups. A pilot group of 5 animals was initially tested for dosage and was not included for hypothesis testing.
Dietary iron loading
Animals age 6 weeks were fed with modified AIN-76A purified rodent chow with 10,000 ppm (1%) Carbonyl Iron (Dyets Inc., Bethlehem, Pennsylvania) for 10 weeks at Gwathmey Inc. (Cambridge, Massachusetts) before being shipped to our animal facility. Serum iron levels were periodically checked. Before drug administration, serum levels of iron-loaded animals were >2-fold higher (423 μg/dl vs. 187 μg/dl) compared with those in regular chow group. Hepatic iron content was measured in 3 animals in both the regular chow as well as the iron-rich chow fed group. Up to 7-fold elevation (2,397 μg/g vs. 312 μg/g dry weight) in hepatic iron level was observed in iron-loaded animals. This model of dietary carbonyl iron overload is commonly used in experimental work and is known to result in up to 2-fold increase in serum iron and 5- to 8-fold increase in hepatic tissue iron over a period of 12 to 16 weeks (18–20). Increased iron content is seen in liver, spleen, heart, intestinal mucosa, and pancreas, but this does not result in liver cirrhosis (18).
After acclimatization, animals were randomized into treatment groups. Iron-rich diet was maintained throughout the course of the study.
A pilot group (n = 5) of regular chow fed rats was given a single 12 mg/kg body weight intravenous dose of DOX. All animals died within 9 days. Therefore, the dose was scaled down to a single dose of 6 mg/kg. Iron-rich chow fed rats (16 ± 2 weeks old) and regular chow fed rats (14 ± 2 weeks old) were given a single intravenous dose of DOX (6 mg/kg) or saline in their tail veins. Animals were anesthetized 2 to 3 weeks later for imaging, blood sampling, and tissue harvesting (Fig. 1). A mixture of xylazine (10 mg/kg) and ketamine (80 mg/kg) was used at 0.1 ml/100 g (1:10 dilution) of body weight intraperitoneally for anesthesia.
Technetium-99m annexin imaging
Human recombinant annexin-V produced by expression in Escherichia coli(Apomate, Theseus, Boston, Massachusetts), derivatized with hydrazinonicotinamide (HYNIC, Anor Med, Langley, British Columbia, Canada), and radiolabeled with technetium-99m (99mTc) (1.0 to 1.2 mCi) (21) was injected in tail veins. Rats were anesthetized 3 h after injection of 99mTc annexin-V and placed directly on the collimator of a gamma camera (Vertex, Philips Inc., Milipitas, California). Planar whole body images were obtained in prone position using an energy window of 140 ± 20% KeV for 20 min. The images were analyzed visually for the intensity of radiotracer uptake in the hearts. After imaging, the rats were euthanized. Excised hearts were imaged by laying directly on the gamma camera. The hearts were then segmented into 4 regions: apical, subapical, midventricular, and basal. These segments were weighed and counted in an automatic well-type gamma counter for determination of the percent injected dose of 99mTc annexin-V.
Serial body weight, survival, and gross appearance
Body weight was monitored at baseline and twice a week until the end of the study. General appearances of animals, movement, feeding, social interaction, posture, dehydration, and fur quality were monitored.
The animals that became moribund, lost ≥20% of body weight compared with baseline body weight, and stopped feeding during the course of experiment and were unlikely to survive for more than 1 day had to be euthanized early using a large intraperitoneal dose of pentobarbital.
Determination of myocardial iron content
Myocardial iron content was measured for all animals using acid digestion of myocardial tissue samples followed by iron quantification with atomic spectroscopy (Kansas State University Veterinary Laboratory).
Histopathology and immunohistochemistry
Cross sections of heart were cut, fixed in buffered 4% formaldehyde, paraffin embedded, sectioned at 3 to 4 μm, and stained with hematoxylin and eosin and Prussian blue.
Sections were stained for activated caspase-3 (Biocare Medical, 1:50, Walnut Creek, California) (22). Sections were deparaffinized, blocked with 3% hydrogen peroxide, treated with 1% bovine serum albumin in phosphate-buffered saline, and incubated with the primary antibody at 37°C for 2 h. The specimens were then incubated with horseradish peroxidase-cojugated streptavidin (Biogenex, San Ramon, California) for 10 min and developed with 3,3’diaminobenzidine (Biogenex), and counterstained with hematoxylin. Grading was performed by a blinded reader based on distribution. We created a scoring system ranging from 0 to 2. Scoring based on distribution was graded on extent of area involved. Less than 30% area involvement was given a score of 0, 30% to 60% a score of 1, and more than 60% was scored as 2.
Small pieces of heart were fixed in half Karnovsky’s fixative containing 2.5% glutaraldehyde and 2% paraformaldehyde in 0.1 M phosphate buffer and embedded in epoxy resin, and 40- and 90-nm thick sections were cut with an ultramicrotome, double stained with uranyl acetate and lead citrate, and examined using a transmission electron microscope (Hitachi H-600, Gaithersburg, Maryland).
SPSS/PC 11.0 software was used (SPSS Inc., Chicago, Illinois). Results are expressed in mean ± SEM. Nonparametric tests including Kruskal-Wallis 1-way analysis of variance and Mann-Whitney Utest were used to compare continuous variables and chi-square for discrete variables. P values <0.05 were considered significant.
All saline-treated control animals survived with normal growth and weight gain with no signs of functional impairment.
Iron-rich chow fed, saline-treated group
Iron-rich chow alone did not result in any deleterious effects with no signs of distress, normal growth, and 100% survival.
DOX treatment in regular chow fed group
Single large intravenous dose of DOX (12 mg/kg) resulted in death within 10 days in all animals (n = 5). Animals showed signs of distress, decreased movement, hunched posture, fuzzy fur, appearance of dark halos around their eyes, and cessation of feeding. Based upon this, 6-mg/kg dose was used in all subsequent single-dose groups. This resulted in a 16% weight loss over 3 weeks compared with 22% weight gain (p < 0.05) in corresponding saline-treated control animals over 3 weeks (Fig. 1).
DOX treatment in rats fed iron-rich chow
Six-mg/kg intravenous dose of DOX resulted in a significant and rapid weight loss (18%) over 2 weeks whereas the corresponding DOX-treated animals in regular chow fed group lost 16% weight while both saline-treated control groups fed regular and iron-rich chow gained weight (Fig. 2).Since these animals lost 18% body weight compared with baseline within 2 weeks of receiving DOX and developed severe functional impairment, they had to be sacrificed 1 week earlier than planned.
99mTc annexin-V imaging
No 99mTc annexin-V uptake was observed in cardiac regions of regular chow or iron-rich chow fed control groups. Intense cardiac uptake was observed in all DOX-treated groups (Fig. 3).Quantitative cardiac uptake analysis from ex-vivo gamma counting revealed only a minimal uptake in regular chow and iron-rich chow fed animals (Fig. 4).Doxorubicin-treated groups had a significantly higher myocardial uptake. Iron-rich chow fed animals treated with DOX had the highest uptake of 99mTc annexin-V in cardiac tissue.
Myocardial iron content
Feeding iron-rich chow resulted in only a slight increase in myocardial iron content compared with the regular chow fed animals in saline-treated groups (59.00 ± 4.56 μg/g vs. 48.33 ± 7.88 μg/g, p = NS). Myocardial iron contents in the corresponding DOX-treated groups were 70.00 ± 3.08 μg/g vs. 52.25 ± 3.63 μg/g, p < 0.02) (23) (Fig. 5).
On light microscopy, no changes were noticed in any group (Fig. 6).Hematoxylin and eosin was remarkably negative for any abnormality. Only rare nonspecific focal inflammatory infiltrates were seen in both DOX-treated regular chow fed and iron-rich chow fed groups. On Prussian blue staining, no stainable iron was seen in any group indicating that increase in myocardial iron content with feeding iron-rich chow was relatively small and well below the levels seen in pathological iron storage conditions.
Electron microscopy showed no cardiac injury in regular chow or iron-rich chow fed control animals (Figs. 7Aand 7B). Doxorubicin-treated animal groups had varying grades of injury. Regular chow fed animals showed lesser degree of myofilament loss with DOX treatment (Fig. 7C). More severe changes were seen in iron-rich chow fed animals treated with DOX. Iron-rich chow animals treated with DOX dose had a greater extent of myocyte injury (Fig. 7D). Frequent areas of myofibrillar degeneration and loss, destruction of myocytes, along with mitochondrial and sarcoplasmic swelling and destruction were seen in this group.
Immunostaining using antibodies specific for activated caspase-3 (Fig. 8)revealed distinct positive reaction in all DOX-treated animals. In the group fed regular chow, positive reaction for caspase-3 was weaker and focal (grade 1). Animals fed iron-rich chow had a more intense and widespread positive reaction for activated caspase-3 (grade 2). No staining was observed in control animals that received saline, with and without iron-rich chow (grade 0).
We employed a rodent model for studying DOX cardiotoxicity and its interaction with iron loading by feeding iron-rich chow for 10 to 14 weeks. Feeding iron-rich chow resulted in a profound increase in DOX cardiotoxicity. These animals lost greater body weight. There was greater myocardial 99mTc annexin uptake in iron-rich chow fed DOX-treated animals compared with that in regular chow fed animals treated with DOX. These findings were confirmed by diffuse and more intense staining of activated caspase-3 on myocardial specimens of iron-rich chow fed animals on DOX treatment. Activated caspase-3 is a sensitive early marker for apoptosis (24). Iron-rich chow alone did not result in any toxicity as seen by effect on body weight, 99mTc annexin uptake, and absence of signs of injury on histopathological analysis. Light microscopy did not reveal any obvious abnormality among the various groups. This is consistent with previous observations where routine light microscopy was not found to be helpful in detecting or grading the severity of DOX cardiotoxicity (25). Electron microscopy showed most intense degenerative changes in iron-fed DOX-treated animals (Fig 7).
We used 99mTc annexin-V imaging and activated caspase-3 for studying the intensity of myocardial toxicity in rats. 99mTc annexin-V imaging has been used for studying other forms of myocardial injury including myocardial infarctions, cardiac transplant rejections (26), and inflammatory myocardial diseases (27). Radiolabeled annexin-V binds to phosphatidylserine, a phospholipid normally present only in inner leaflet of cell membrane that gets exposed on outer cell membrane leaflet in cells undergoing apoptosis. Higher uptake of annexin-V was seen in iron-rich chow fed animals treated with DOX, compared with that seen in regular chow fed animals given DOX.
Mammalian hearts share a susceptibility to DOX cardiotoxicity due to limited protective mechanism from oxidative injury. Heart tissue is relatively deficient in catalase, which, along with glutathione peroxidase, forms the defense against redox injury. The heart primarily depends on glutathione peroxidase to protect itself from free-radical injury. Doxorubicin interferes with glutathione peroxidase making the heart more susceptible to redox injury (28). Doxorubicin-iron complex can catalyze the transfer of electrons from reduced glutathione to molecular oxygen to generate highly reactive superoxide, hydrogen peroxide, and hydroxyl radicals (29).
A number of animal models including mice, rats, dogs, swine, hamsters, and rabbits have been used for studying DOX cardiotoxicity (30,31). We used a rat model because of its suitability for imaging studies and for ease with which nutritional iron overload can be induced in this species. As mentioned earlier, iron overload can be produced by parenteral administration of iron or by feeding chow fortified with iron salts or elemental iron (e.g., carbonyl iron). Parenteral iron loading results in massive tissue iron overload, with multiorgan damage and toxicity, while oral feeding results only in a relatively modest increase in body iron content without overt signs of end-organ damage. Carbonyl iron is better absorbed and results in less gastrointestinal and systemic toxicity compared with ionic iron (32,33). Dietary carbonyl iron may result in pattern of iron overload similar to that of subclinical iron overload in man. Whereas this mild increase in myocardial iron in saline-treated animals did not result in any obvious cardiotoxicity, this was associated with marked increase in myocyte injury in DOX-treated animals.
To the best of our knowledge, this is the first study to employ a rat model of dietary carbonyl iron overload to assess DOX cardiotoxicity. Doxorubicin cardiotoxicity has been under investigation in humans for well over 3 decades. A number of variables have been considered as contributors to DOX cardiotoxicity (6). Even while iron has been reported to play a role in DOX cardiotoxicity in experimental studies (13,34,35), importance of iron as an independent variable has not been studied in humans. A wide variation can occur in body iron stores in patients undergoing cancer chemotherapy, depending upon abnormal blood losses, blood transfusions, iron supplementation, and nutritional status. Adult and pediatric patients undergoing treatment for leukemia and other malignancies can develop a significant level of iron overload during and as a result of chemotherapy and bone marrow transplantation (36). It has also been observed that iron level may predict outcomes in cancer survivors. Level of iron overload observed varies between moderate to severe levels in 15% to 20% of long-term survivors (37–39). Several common genetic disorders may also influence body iron stores. Hemochromatosis, the most common form of genetic disorder of iron overload, has a frequency of 1:250 to 400 in Americans of Northern European descent. This is characterized by homozygous mutation (C282Y+/+) of iron regulatory gene (HFE gene) resulting in unregulated intestinal absorption of iron. However, heterozygous prevalence of this gene (C282Y+/−) is very common in the same population (7% to 8%) (40). Another mutation of the HFE gene associated with milder degree of iron overload (H63D+/−) has a prevalence of up to 20% (40). Heterozygous carriers of these mutations may have subclinical and otherwise asymptomatic increase in body iron stores. Dexrazoxane, the only chelator used clinically in patients, is recommended for use beyond a cumulative DOX dose of 300 mg/m2(17). As toxicity may begin much earlier, in the presence of elevated body iron, perhaps this agent may be considered earlier during the course of DOX or even before DOX therapy in those with elevated body iron stores.
Our findings are similar to those of Miranda et al. (41) who recently observed increased susceptibility to DOX cardiotoxicity in HFE knockout mice. They observed higher mortality, greater increase in creatine kinase and aminotransferase enzymes, and more severe histologic change in the HFE knockout mice compared with that in wild type. Both of these studies emphasize the critical role of iron in DOX cardiotoxicity. However, Corna et al. (42) reported a paradoxical benefit between anthracycline-derived reactive oxygen species, increased ferritin synthesis, and resistance to iron-mediated damage. Authors reported protection of H9c2 cells from ferric ammonium citrate-induced damage on pretreatment with DOX by down-regulating iron regulatory protein-2 and subsequent increase in ferritin synthesis. The role of iron and reactive oxygen species in anthracycline-induced cardiotoxicity may, therefore, be more complex than previously believed (42). However, these findings need further study and are contrary to our findings of increased cardiac injury in iron-supplemented animals receiving DOX.
We speculate that higher body iron stores in patients who have received multiple blood transfusions, prolonged iron supplementation, or those with unsuspected HFE mutations may be predisposed to DOX cardiotoxicity. Prospective clinical studies are needed to test our hypothesis. If these results are confirmed by clinical studies, simple and relatively inexpensive screening tests for total body iron stores and presence of HFE mutation can be used to screen patients before initiating DOX chemotherapy.
The authors thank Mr. Gunner Snyder for his help with nuclear imaging.
Supported by the American Society of Nuclear Cardiology/Fujisawa Healthcare Award, Bethesda, Maryland.
- Abbreviations and Acronyms
- iron regulatory gene
- Received December 4, 2006.
- Revision received January 24, 2007.
- Accepted February 6, 2007.
- American College of Cardiology Foundation
- Gianni L.,
- Myers C.E.
- Schuchter L.M.,
- Hensley M.L.,
- Meropol N.J.,
- Winer E.P.
- Blankenberg F.G.,
- Katsikis P.D.,
- Tait J.F.,
- et al.
- Gown A.M.,
- Willingham M.C.
- Edwards C.Q.
- Duan W.R.,
- Garner D.S.,
- Williams S.D.,
- Funckes-Shippy C.L.,
- Spath I.S.,
- Blomme E.A.
- Imondi A.R.,
- Della Torre P.,
- Mazue G.,
- et al.
- Hasinoff B.B.,
- Davey J.P.
- Harrison P.,
- Neilson J.R.,
- Marwah S.S.,
- Madden L.,
- Bareford D.,
- Milligan D.W.
- Halonen P.,
- Mattila J.,
- Suominen P.,
- Ruuska T.,
- Salo M.K.,
- Makipernaa A.
- Miranda C.J.,
- Makui H.,
- Soares R.J.,
- et al.
- Corna G.,
- Santambrogio P.,
- Minotti G.,
- Cairo G.