This review summarizes methods to reduce anthracycline-induced cardiotoxicity, including modified dosing schedules, cardioprotective agents such as ICRF-187, and liposomal delivery systems.
Do modified administration schedules, cardioprotective agents, or liposomal delivery reduce anthracycline-induced cardiotoxicity in patients receiving chemotherapy?
Modifying anthracycline administration to reduce peak plasma levels, such as through split dosing or prolonged infusions, is an effective strategy to mitigate cardiotoxicity.
The treatment of malignant disease has to be finely balanced to obtain maximal cure rates at the least possible clinical cost to the patient. Cardiotoxicity can be a consequence of radiation treatment ( 31; 14), chemotherapeutic agents and biological response modifiers ( 39). However, anthracyclines are by far the most important group clinically and this review will focus on methods of reducing anthracycline-induced cardiotoxicity. Anthracyclines are active agents in the treatment armamenterium for both haematological ( 38; 54; 46; 90) and oncological malignancies ( 77; 7), with good evidence of a dose–response relationship. Dose-limiting cardiotoxicity was observed in early clinical trials ( 50), and the risk factors for toxicity are now well elucidated ( 96). Total cumulative dose is the most predictive ( 95; 64); other factors are dose intensity ( 78; 47; 49), age, both young age and elderly (>70 years) ( 56; 95), gender ( 57; 49), pre-existing heart disease and hypertension ( 95), cardiac radiation, and length of follow-up ( 81). The natural history of anthracycline cardiotoxicity is unclear. Acute toxicity can occur on-treatment or arising within weeks or months of finishing treatment. This may present with heart failure due to dilating cardiomyopathy, abnormalities of cardiac rhythm, or sudden death. Congestive cardiac failure (CCF) can be amenable to medical treatment, and some paediatric patients regain good compensatory cardiac function and can discontinue all anti-failure drugs. Late cardiac dysfunction (4–20 years) has been reported in the paediatric population ( 81; 29), ranging from subclinical dysfunction ( 56; 78) to irreversible failure requiring cardiac transplantation ( 52). Studies evaluating cardioprotection in the paediatric population must take into account this long latent period. The mechanism of cardiac damage is probably multifactorial and any hypothesis needs to incorporate the clinical characteristics, such as the escalating effect of increasing cumulative dose, idiosyncratic effects, histological changes, partial early reversibility, and the late progressive pattern ( 9). Histopathological studies highlight the focal nature of the lesions which include loss of myofibrils, vacuolation within myocyte cytoplasm, focal membrane thickening, swelling of the sarcoplasmic reticulum and the mitochondria, and interstial fibrosis without an inflammatory response ( 6). Studies of cardiac damage usually use the Billingham scoring system, where the grading (scoring 1–3) depends on the percentage of myoctes which show damage by electron microscopy. Myocyte damage cannot be repaired by increase in myocyte numbers after a postnatal age of 6 months ( 86), so the damage is compensated by myocyte hypertrophy. This may explain the sudden expotential fall in function ( 81) that occurs in late cardiotoxicity when the remaining hypertrophied cells can no longer compensate. Remodelling after damage may be prevented by the inhibitory effect of anthracycline on cardiac gene expression, and may also, in part, explain this clinical observation ( 9). The mechanism by which the heart overcomes myocyte damage makes non-invasive detection of early myocardial damage problematical and a poor predictor of outcome ( 85; 93; 11). The poor predictability of non-invasive monitoring is a real problem in childhood studies, where long-term cardiac well-being is the ultimate end point. This is particularly so as the majority of childhood cancers can be cured with moderate doses of anthracyclines. This differs from adults, where the treatment objective in many cases is to give escalating doses of anthracyclines without precipitating cardiac failure in the short term. One paediatric study showed that normal fractional shortening measurements taken within 6 months from the end of treatment predicted continued normality in the readings in 88% but that abnormal values remained in 71% of patients at long-term follow-up ( 81). Histopathological changes, both on light and electron microscopy in endomyocardial biopsies, are the only truly reliable method of assessing damage and the changes correlate directly with the total doses received. 23) compared the changes in serial endomyocardial biopsies with ejection fractions calculated from echocardiographic or radioisotope techniques in adults and found no correlation between these methods but good correlation between increasing doses of adriamycin and biopsy grading. Only a few adult studies use this invasive monitoring, so assessing the value of differing cardioprotection methods is difficult. A number of toxic biochemical changes have been identified which cause significant myocyte damage ( 74). These include cellular toxicity from metabolites, generation of oxygen free radicals ( 4), release of vasoactive amines and selective inhibition of cardiac muscle gene expression for α-actin, troponin, myosin light chain 2, and the M isoform of creatine kinase ( 44); impaired calcium homeostasis causing intracellular calcium overload ( 8; 40; 48), disturbance of myocardial adrenergic function, interaction with cell membranes, and effects on nucleic acids. The free radical mediated damage is probably the most significant action causing cardiotoxicity. Considerable tissue damage is caused, in part by perioxidative injury to intracellular membranes, particularly the mitochondria. The protective enzyme systems such as superoxide dismutase, catalase and selenium-dependent glutathione peroxidase are scanty in cardiac muscle and are 10–30% lower than in the liver. The free oxygen radicals are formed by two processes, firstly by the reduction of the anthracycline molecule from a quinone to a semiquinone free radical ( 19). Regeneration of quinone releases an electron which reacts with an oxygen molecule producing the highly active hydroxyl radical. The second method is by the formation of an iron complex which participates in the formation of superoxide and hydroxl radicals from molecular oxygen ( 21; 22). Anthracyclines are metabolized to a 13-dihydro derivative which may be more cardiotoxic than the parent compound. Doxorubicinol in vitro is twice as cardiotoxic as doxorubicin and daunorubicinol is 6 times more cardiotoxic than daunorubicin ( 72). The rate at which the parent compounds are metabolized differs. Daunorubicin is more rapidly metabolized than doxorubicin and epirubicin. In mice the area under the cuve (AUC) for doxorubicinol is significantly lower than that of doxorubicin, but the AUC for daunorubicinol is only half that of daunorubicin. 72) suggested that after bolus injections doxorubicin contributes to nearly all the cardiotoxicity but daunorubicin only causes 25% of the damage, the remainder is caused by daunorubicinol. This work has not been performed using infusions of anthracyclines, but makes the point that regimens using anthracyclines are not necessarily interchangeable. The use of infusions for all anthracyclines may not be appropriate as there may be an increase of the AUC of a more cardiotoxic metabolite. Marked interpatient variability has also been reported ( 88; 67). The clearance of anthracyclines is mainly through bile, with urinary excretion accounting for 10%. Interestingly, studies with doxorubicin in patients with hepatic dysfunction show no increase in cardiac toxicity but increased mucositis and myelosuppression ( 46). This review will discuss anthracycline cardioprotection under three headings: mode of administration, cardioprotective agents, and the use of liposomal delivery of anthracycline. Modifiying scheduling of anthracyclines must take into account the pharmokinetics of the different anthracyclines. In reducing cardiotoxicity, efficacy must not be compromised nor should potentiation of other side-effects cause dose limitation. The modification of treatment schedules occurred because of the premise that peak plasma anthracycline levels were responsible for the cardiotoxic effect whereas the AUC was related to efficacy ( 70; 79), but 72) suggested that the AUC of the metabolites might be important for cardiotoxicity with certain anthracyclines. The early dosing was by single bolus doses 30–75 mg/m2 with the dose intensity limited by the acute side-effects such as myelosuppression, nausea and mucositis. Simple reduction of dose intensity might reduce the incidence of late cardiotoxicity. 78) noted dose intensity as a risk factor in children treated for Wilms' tumour when the dose intensity varied between 30 and 40 mg/m2 3–6-weekly. Multivariate analysis of left ventricular end systolic wall stress showed dose intensity as a determinant (P = 0.02). In paediatric malignancies dose intensity reduction may be a realistic option because of the high survival rate using moderate doses. Split dosing (fractionation) and lengthening infusion times (spreading the load) reduces peak plasma dose levels and lengthens exposure time but maintains the AUC ( 79; 88; 60). Studies of intracellular concentrations are conflicting. 60) showed a 43% peak dose reduction in chronic lymphatic leukaemia cells after a 96 h infusion compared with bolus dose but 79) showed that the final cellular concentration in melanoma and leukaemic blasts was similar regardless of the method of administration. They noted a 40% loss of cellular doxorubicin within 1 h of injection in the bolus group compared with minimal loss in the long-term infusion group. Studies on cardiac tissue in the rat and rabbit imply that there is an accumulation of doxorubicin and doxorubicinal after multiple dosing regardless of infusion rates, but peak left ventricular values were lower in the prolonged infusion group, although this was not sustained when the animals were killed at 7 d ( 18; 9). Clinical studies show that decreasing the peak plasma levels has a beneficial effect on the incidence and severity of cardiotoxity. Von Hoff's historical review of patients charts revealed a reduction of congestive cardiac failure after doxorubicin from 2.8% to 0.8% when split dosing was used ( 95). 85) retrieved endomyocardial biopsy material from patients treated for various malignancies who had either received anthracyclines as bolus or weekly doses of equal dose intensity. Multivariate analysis was used to predict cardiac damage, only the type of schedule was found to be significant and cardiac radiation of borderline significance.They estimated that higher cumulative doses of doxorubicin (168 mg/m2) could be given weekly to achieve the same endomyocardial biopsy score. A randomized multi-agent study was performed in patients treated for non-small round cell lung cancer with the only variable being a weekly schedule of doxorubicin (20 mg/m2) compared with a bolus 3-weekly dosing regimen (60 mg/m2). To study the cardiotoxicity all patients underwent cardiac evaluation; 38% in the single dose group and 26% in the split dose group had endomyocardial biopsy at cumulative doxorubicin doses of 250–300 mg/m2, 450–550 mg/m2 and at every 180 mg/m2 increment thereafter. The overall biopsy score was significantly less in the split dose group, but more patients in the single-dose group had received mediastinal radiation ( 93). These results were supported by 89), both studies showing no loss in efficacy and less myelosuppression, emesis and alopecia in the split dose regimens. In children, a retrospective study showed no advantage of consecutive daily dose administration (33% of single dose given daily for 3 d) compared with bolus dose every 3 weeks at a mean total dose 324 mg/m2 and 299 mg/m2 respectively. The study overall had an unexplained high incidence of cardiac dysfunction and the groups had disproportionate patient numbers, 96 versus 17 respectively, but nevertheless no difference was detected ( 24). A number of clinical studies have shown a significant reduction in early cardiotoxicity with increased infusion times ( 100; 41) ( Table I). Legha et al (1982 ) demonstrated, in a prospective non-randomized study in adult patients receiving high-dose doxorubicin, that long infusion rates of 24–96 h protected for cardiotoxicity and did not affect efficacy. Between total doses of 500 and 800 mg/m2 no patients in the infusion arm showed signs of heart failure compared with 24% in the bolus dose cohort (bolus dose 60 mg/m2). At the same total doses in excess of 800 mg/m2 the longer the infusion rate the lower the incidence of heart failure; a comparison between 24–48 h and 96 h infusion gave incidences of 14% and 9% respectively and these differences were backed by endomyocardial biopsy grades. This early study has been supported by adult randomized studies in poor prognostic groups with infusion times ranging from 6 to 96 h, but cardiac toxicity has been documented only from clinical symptoms or non-invasive assessment in the short term ( 75; 100; 15). In the U.K. many childhood malignancy treatment regimens incorporate anthracyclines given over 6–48 h. There are no studies in children apart from an observational study on five patients with hepatoblastoma who received total doses of doxorubicin ranging from 130 to 720 mg/m2 by 96 h infusion with no cardiac effects ( 82). Long infusions can produce unpredictable myelosuppression and severe stomatitis. Anti-tumour efficacy for long infusions still has to be proven ( 5). In vitro experiments have suggested long cellular exposure may lead to an increase in drug resistance ( 42; 16). There are no reports as yet on the effect of varying schedules on late-onset cardiac dysfunction. All the above studies used doxorubicin as the anthracycline. The only data on infusional daunorubicin is a small paediatric study which primarily investigated the use of varying induction protocols on reduction of the leukaemic burden. No abnormal echocardiograms were noted in 18 patients in the infusional group (median daunorubicin dose 400 mg/m2), but 4/18 in the bolus cohort (median dose 360 mg/m2) had abnormal studies ( 82). 55) raised concerns that long infusion times in children may still result in myocyte loss resulting in thinner ventricular walls, causing chronic increase in afterload and late decrease in function. Epirubicin has been shown to be active against a number of tumours with response rates varying between 3% and 82% (Arcamone, 1987). Compared with doxorubicin, epirubicin, mg for mg, is less cardiotoxic. The median dose at which cardiac failure occurs is 1134 mg/m2 compared with 492 mg/m2 of doxorubicin after a bolus dose ( 45). This may be attributed to the more rapid plasma clearance of epirubicin due to a different metabolic pathway ( 20). 71) studied a group of patients with breast cancer who had never received anthracyclines and showed that, at cumulative epirubicin doses of 900 mg/m2, 4% developed CCF which increased to 15% at doses of 1000 mg/m2. Scheduling a single dose versus day 1 and 8 split dose did not affect the risk of cardiotoxicity but the higher the mean single dose level (i.e. cumulative dose divided by the number of injections), the greater was the risk of CCF. 3) suggested idarubicin as an active anthracycline for oral use in adult non-lymphoblastic leukaemia. Phase III randomized studies have been carried out in adults with acute myeloid leukaemia and it has been found to be as effective as daunorubicin. The oral formulation may be of use for NHL and myeloma ( 46). Idarubicin is thought to be less cardiotoxic at equimyelotoxic doses. In one study, relapsed/new patients with MDS/AML who received idarubicin were assessed for cardiotoxicity. Of 127 patients, four developed clinical CCF and of whom three had received prior treatment with anthracyclines or mitroxantrone, 65 patients were assessed for subclinical cardiac damage and 10 showed abnormalities; four had received prior cardiotoxic drugs. The median dose was 138 mg/m2. In view of the previous treatment, little can be drawn from this study regarding the actual cardiotoxicity except that the effect is synergistic with other cardiotoxic agents ( 2). The search for agents to be used alongside anthracyclines to provide cardiac protection became a reality when research showed the anti-neoplastic effects may be independent of the cardiotoxic effect. Anthracyclines are intercalating agents which block DNA synthesis and effect repair processes ( 65); this is probably a more important antineoplastic effect than free radical production. As reviewed above, oxygen free radicals are thought to be the main contributors to cardiotoxicity. This is supported by the studies on animals showing modification of cardiotoxicity by antioxidants such as N-acetylcysteine (NAC), oxypurinol and selenium ( 37). Small animal (mice, rats, rabbits) studies with NAC, vitamin E and vitamin A ( 76; 59) showed effectiveness but this was not borne out in larger animals (dogs, swine) who received clinically more relevant anthracycline doses. NAC ( 61) and vitamin E ( 10) have been tested for their clinical specificity in humans and found to be ineffective. Probucol, a lipid-lowering agent and antioxident, has been shown in the rat model to protect against cardiac damage and to maintain efficacy in a mouse tumour model ( 76). Coenzyme Q10 has shown some benefit in children treated for acute lymphoblastic leukaemia in a small study ( 43). Calcium antagonists such as verapamil, propranolol and prenyamine have also been considered ( 59). Work on isolated rat hearts suggested an effect on the acute depressive affect of anthracyclines, but no trial in vivo showed evidence of cardioprotection ( 98). Amifostine, an organic thiophosphate, has cytoprotective effects for nephrotoxicity, neurotoxicity and myelosuppression. It is thought to act by scavenging free radicals and to form mixed disulphides to protect normal tissues ( 12). In vivo and in vitro studies in mice have shown a degree of cardioprotection with doxorubicin and daunorubicin but not with mitroxantrone. No human trials have been conducted ( 22). More interesting is the work on iron chelators. The theory behind the use of chelating agents is that by depletion of intracellular iron there would be less available to form anthracycline–iron complexes and so reducing free radical formation. The full explanation of the effectivness of chelating agents has not been established. 36) demonstrated the cardioprotective properties of EDTA in isolated dog heart muscle. Subsequently EDTA, desferrioxamine and related compounds, although effective metal chelators in other clinical settings, were not shown to be effective in cardioprotection, which may be due to their poor cell membrane penetration ( 33). At the same time, work was in progress to assess the antitumour activity of an EDTA-related compound (ICRF-159 Razoxane) and the cardioprotective properties were then observed ( 17). The more soluble ICRF-187 (dexrazoxane) showed a similar effect and has been the agent used in clinical trials. Initial studies were carried out in rabbits treated with daunorubicin with good effect. The myocardial lesions seen in human hearts were identified in the rabbit hearts in all those treated with daunorubicin alone compared with 33% in those treated with ICRF-187 in addition to daunorubicin. This study was extended to dogs using doxorubicin as the anthracycline and the incidence and severity of myocyte damage was significantly reduced ( 33). Epirubicin cardiotoxicity has also been shown to be reduced by the use of ICRF-187, but no effect was observed with mitoxantrone. The dosing regimes used in humans were mimicked with encouraging results in swine and dogs ( 34, 35). These studies were terminated at the end of the drug treatment, but studies in protection showed of the clinical studies on adults have been on patients with breast cancer and a small study of patients with tissue ( The study showed more anthracycline could be given ICRF-187 was used in studied a number of patients with breast cancer in a randomized in with and were given at 3-weekly The study used a of ICRF-187 to doxorubicin of with an excess of in the ICRF-187 the was reduced to The groups were for previous treatment and cardiac risk The for a cardiac of CCF or in left ventricular ejection as by a in patients was times greater in the group than the the ICRF-187 group. occurred in 15% respectively with CCF only in and This was significant in the but the with the this was due to differences with doxorubicin, as the effect of cardioprotection in at doses above mg/m2. also showed this effect with epirubicin doses mg/m2. The tumour response rate was less in the ICRF-187 group although overall time to and survival was toxicity analysis demonstrated the greater effect of ICRF-187 on myelosuppression, with a significant difference in toxicity for At present there are no trials for haematological The the use of ICRF-187 in the paediatric age group showed cardioprotection in a small number of patients with either or disease who had been to anthracyclines. The ICRF-187 group had no cardiac dysfunction at the end of treatment, as by after anthracycline doses mg/m2. in the ICRF-187 group by a mean of compared with in the ICRF-187 group ( To there is one small randomized study which has shown a similar result to the adult studies ( The dose and method of administration has not been It is that ICRF-187 needs to be given prior to or anthracycline administration because of short half it is probably not so effective when given prior to long All the clinical studies use ICRF-187 prior to a bolus anthracycline The of ICRF-187 to anthracycline dose between effectiveness has been shown with between and Table The dose has been shown to be directly related to the degree of myelosuppression ( et ) and at the et al ) showed excess More work is to the effective There are a number of particularly in paediatric ( No randomized study has shown of reduced It is to ICRF-187 the effect of anthracyclines. The effectiveness at anthracycline doses and against the of late cardiotoxicity in children has not been has been shown to be in the inhibition of and so may to the risk of second malignant ( To these a study needs to be performed with a to long-term The of drug delivery systems is a of to be to the tissue and reducing exposure to of chemotherapeutic agents may increase tumour dose with reduction of The use of for drug delivery was developed 30 and has been but only have the clinical trial are and can both or soluble ( properties are on and the of the The main have been with the of by the and poor of the drug to the of the has in the formulation of long more to of have been studied in animal and one which on to the and doxorubicin as the drug and using with high daunorubicin ( The in tumours is on a liposomal of liposomal and abnormal with increased in the tumours ( and drug release in tumours and has been studied using with ( Work was performed on patients with and it was observed that the was higher than in other tumours ( lesions showed significant with ( has been demonstrated for with partial response rates of with daunorubicin and with doxorubicin ( Phase trials on adult tumours have shown some objective response in and and ( muscle has a of and has an system, so with the release properties of anthracyclines their use should reduce cardiotoxicity. Studies on rabbits and dogs have this ( Clinical studies on patients with have not reported significant cardiotoxicity to cumulative doses of liposomal daunorubicin to 1000 mg/m2 ( In a study of liposomal doxorubicin in adults with a of tumour 17 of whom had received mg/m2, there were no cases of clinical heart in those patients receiving mg/m2 showed one of left ventricular ejection of ( There are no studies in a study is in progress in on patients in from a of There are no studies at late cardiac dysfunction. The are myelosuppression, severe mucositis with and ( The is seen when using and is related to long reducing the of doses reduces the incidence ( At liposomal anthracycline is not used for haematological malignancies or the of malignant cells in of may be by the of for tumour cell Work is in progress on animal for mice with human using with an to the doxorubicin ( anthracyclines are important chemotherapeutic the for effective cardioprotection is and may increase due to anthracycline cardiac damage in patients who have their young are being in their have received anthracyclines, for of cardiac This review the of data on cardioprotection for haematological and paediatric Clinical research with numbers of patients needs to be performed in randomized particularly in patient groups with good so can be will to be conducted with the different anthracyclines as results cannot necessarily be between drugs. non-invasive methods of cardiac damage at the time of administration which will correlate with late outcome to be developed so that the time for paediatric studies can be The to take the to or anthracyclines with less and similar efficacy. for and
Gill Levitt (Wed,) conducted a review in Anthracycline-induced cardiotoxicity. Cardioprotective strategies (scheduling, cardioprotective agents, liposomal delivery) was evaluated. This review summarizes methods to reduce anthracycline-induced cardiotoxicity, including modified dosing schedules, cardioprotective agents such as ICRF-187, and liposomal delivery systems.