Pixantrone beyond Anthracyclines - American Chemical Society

Perspective pubs.acs.org/crt

Rethinking Drugs from Chemistry to Therapeutic Opportunities: Pixantrone beyond Anthracyclines Pierantonio Menna,† Emanuela Salvatorelli,† and Giorgio Minotti* Unit of Drug Sciences, Department of Medicine, University Campus Bio-Medico, Via Alvaro del Portillo, 21, 00128 Rome, Italy

ABSTRACT: Pixantrone (6,9-bis[(2-aminoethyl)amino]benzo[g]isoquinoline-5,10-dione) has been approved by the European Medicines Agency for the treatment of refractory or relapsed non-Hodgkin’s lymphoma (NHL). It is popularly referred to as a novel aza-anthracenedione, and as such it is grouped with anthracycline-like drugs. Preclinical development of pixantrone was in fact tailored to retain the same antitumor activity as that of anthracyclines or other anthracenediones while also avoiding cardiotoxicity that dose-limits clinical use of anthracycline-like drugs. Preliminary data in laboratory animals showed that pixantrone was active, primarily in hematologic malignancies, but caused significantly less cardiotoxicity than doxorubicin or mitoxantrone. Pixantrone was cardiac tolerable also in animals pretreated with doxorubicin, which anticipated a therapeutic niche for pixantrone to treat patients with a history of prior exposure to anthracyclines. This is the case for patients with refractory/ relapsed NHL. Pixantrone clinical development, regulatory approval, and penetration in clinical practice were nonetheless laborious if not similar to a rocky road. Structural and nominal similarities with mitoxantrone and anthracyclines may have caused a negative influence, possibly leading to a general perception that pixantrone is a “me-too” anthracycline. Recent insights suggest this is not the case. Pixantrone shows pharmacological and toxicological mechanisms of action that are difficult to reconcile with anthracycline-like drugs. Pixantrone is a new drug with its own characteristics. For example, pixantrone causes mis-segregation of genomic material in cancer cells and inhibits formation of toxic anthracycline metabolites in cardiac cells. Understanding the differences between pixantrone and anthracyclines or mitoxantrone may help one to appreciate how it worked in the phase 3 study that led to its approval in Europe and how it might work in many more patients in everyday clinical practice, were it properly perceived as a drug with its own characteristics and therapeutic potential. The road is rocky but not a dead-end.

■

CONTENTS

1. Introduction 2. Why Was Pixantrone Developed? The Iron and Hydroquinone Story 3. Pixantrone in Cancer Cells: From topoisomerase IIα to Something Else 4. Mechanisms of Pixantrone Disruption of Mitotic Fidelity 5. Pixantrone and DNA Alkylation 6. Putting Things Together: From Cancer Cells Back to Cardiomyocytes 7. Regulatory Issues, Current Problems, and Future Developments Author Information Corresponding Author Author Contributions

© XXXX American Chemical Society

Funding Notes Biographies Abbreviations References

A B

F F F G G

C

1. INTRODUCTION Pixantrone (PIX, 6,9-bis[(2-aminoethyl)amino]benzo[g]isoquinoline-5,10-dione) is a novel aza-anthracenedione that shows activity in refractory/multiply relapsed aggressive B-cell non-Hodgkin’s lymphoma (NHL).1 Anthracenediones are quinone−hydroquinone drugs that look similar to doxorubicin (DOX) and other antitumor anthracyclines; however, structural

D D D F F F F

Received: May 30, 2016

A

DOI: 10.1021/acs.chemrestox.6b00190 Chem. Res. Toxicol. XXXX, XXX, XXX−XXX

Chemical Research in Toxicology

Perspective

Figure 1. Structures of DOX, MITOX, and PIX. DOX is composed of a tetracyclic quinone−hydroquinone chromophore, a carbonyl-containing side chain, and an aminosugar (daunosamine) that is attached by a glycosidic bond to the ring system. MITOX is a three-ring quinone−hydroquinone anthracenedione; it lacks daunosamine and its side chains lack carbonyl groups. Pixantrone differs from MITOX in the lack of the hydroquinone, insertion of a nitrogen heteroatom in the same ring, and substitution of (ethylamino)diethylamino for (hydroxyethylamino)ethylamino side chains (all indicated by arrows).

According to the prevailing hypothesis anthracyclines or anthracycline-like drugs may become cardiotoxic after a continuous one-electron reduction and oxidation of their quinone moiety, which increases cellular levels of reactive oxygen species (ROS) like superoxide anion (O2•−) and hydrogen peroxide (H2O2). Because cardiomyocytes are illequipped with O2•− and H2O2 detoxifying enzymes, overproduction of O2•− and H2O2 may result in formation of hydroxyl radicals (•OH) that eventually cause oxidative stress.2 Iron must be available to catalyze Haber−Weiss or Fenton reactions that convert the O2•−/H2O2 couple into •OH. Redox cycling of the quinone moiety can fulfill this requirement by reductively releasing iron from ferritin. This may occur directly (through an electron transfer from transient semiquinones to the ferric oxohydroxide core of ferritin) or indirectly (through O2•− penetration in ferritin).2−4 Thus, anthracyclines can do everything that is needed to inflict oxidative stress to the vulnerable cardiomyocyte. Further evidence for cause-and-effect relations between iron, oxidative stress, and cardiotoxicity originates from the ability of anthracyclines and MITOX to bind iron. Both drugs show a higher affinity for Fe(III) than Fe(II), but formation of relatively unstable Fe(II) complexes is nonetheless sufficient for these drugs to prevent ferritin from reincorporating Fe(II).5 Moreover, drug−Fe(II) complexes react with oxygen and are converted into more stable Fe(III) complexes, and this process is accompanied by generation of ROS.6 In DOX−Fe(IlI) complexes, three anthracycline molecules bind one Fe(III), with each anthracycline molecule coordinating iron through one carbonyl and one phenolate (hydroquinone) oxygen. This results in formation of a six-membered ring system, but other DOX:Fe(III) stoichiometries may occur via anthracycline selfassociation or coordination of one more iron through electron resonance of the quinone−hydroquinone system.7 Drug− Fe(III) complexes are redox active. Cellular reductants, like O2•− and thiols, reduce Fe(III) to Fe(II), and the latter engages in further reactions with molecular oxygen.8 Complexes of anthracyclines or MITOX with Fe(III) can also undergo selfreduction to the ferrous form, which probably occurs through an electron transfer from side-chain primary alcohols to Fe(III). Self-reduction cannot occur with anthracycline analogues that

differences between anthracenediones and anthracyclines are noticeable. The planar ring system of anthracyclines is composed of four rings while that of anthracenediones contains three rings only; moreover, anthracyclines contain an aminosugar, daunosamine, that is absent in anthracenediones (Figure 1). In spite of these major structural differences, anthracenediones show an anthracycline-like antitumor activity as well as a dose-related cardiotoxicity that may progress toward a lifethreatening heart failure (HF). Anthracenediones are therefore considered as anthracycline-like drugs, and this terminology is widely adopted in pharmacologic and regulatory settings. Here we briefly scrutinize whether PIX too is an anthracycline-like drug. We thought this was important for positioning PIX in the pharmacologic armamentarium of oncohematologists. We thought this was important also for avoiding misconceptions that could limit the clinical use of PIX. Anthracyclines have been around for so many years that oncohematologists might perceive them as too old to be good and compete with the newer targeted drugs. This would be a wrong assumption. Anthracyclines are important in many oncologic and hematologic settings. This having been acknowledged, a “novel anthracenedione” like PIX might be perceived as an old drug that has little to say in modern oncohematology. Because patients count more than words and definitions, a critical reappraisal of PIX is much needed at this point in time.

2. WHY WAS PIXANTRONE DEVELOPED? THE IRON AND HYDROQUINONE STORY In comparison with the prototypic anthracenedione, mitoxantrone (MITOX), PIX is characterized by the insertion of a nitrogen heteroatom in the tricyclic chromophore, substitution of (ethylamino)diethylamino for (hydroxyethylamino)ethylamino side chains, and removal of the hydroquinone in juxtaposition to the quinone group (see also Figure 1). Why was PIX developed this way? The main rationale was offered by the clinical need for an anthracenedione that retained the same antitumor activity as that of MITOX but caused less cardiotoxicity. Much attention was paid to removing the hydroquinone moiety and eliminating interactions with iron. In fact, iron has long been suspected to play a role in anthracycline-related cardiotoxicity.2 Was this strategy correct? B



Perspective

given its apparent cardiac tolerability vis-à-vis structural differences from MITOX, can we still consider PIX an anthracycline-like drug?

lack side-chain primary alcohols (e.g., daunorubicin or idarubicin).9 The “iron and free radical hypothesis” of cardiotoxicity is attractive in many respects. For example, hereditary hemochromatosis leads to iron overload and aggravates myocardial injury in anthracycline-treated survivors of childhood acute lymphoblastic leukemia.10 The apparent association and stability constants of DOX−Fe(III) complexes are approximated to 1033.1 M−3 and 1028−1034 M−3, respectively, which rank among the highest reported for any Fe(III) complex.8,11 In clinical settings the bis-ketopiperazine and iron chelator, dexrazoxane, remains the only agent approved for prevention of cardiomyopathy and HF in patients exposed to anthracycline-based anticancer treatment.4 The mechanisms by which dexrazoxane diffuses in cells and hydrolyzes to give a diacid diamide with EDTA-like metal-chelating properties have been well characterized.12 How does PIX fit in this scenario? With regard to oneelectron redox cycling of the quinone moiety, recent studies suggest that PIX generates its semiquinone even better than MITOX does; however, this observation is confined to cell-free systems.13 Regardless of whether PIX underwent one-electron redox cycling in a cellular environment, other reactions attributed to DOX or MITOX would nonetheless be precluded to PIX. The lack of a hydroquinone, which is crucial for iron coordination, negates formation of PIX−iron complexes.13,14 It follows that PIX should neither bind Fe(II) released from ferritin nor generate Fe(III) complexes liable to reduction or self-reduction. Preclinical studies seemed to support a possible link between failure to bind iron and lack of cardiotoxicity. Pixantrone caused little or no cardiotoxicity in rodents;15−17 moreover, PIX did not exacerbate cardiotoxicity in animals pretreated with DOX, while MITOX severely aggravated it.15 This latter observation raised hopes that PIX could be used also in anthracycline-pretreated patients, which later proved to be the case for patients with refractory/relapsed aggressive B-cell NHL. Whereas several lines of evidence recapitulate the cardiac safety of PIX under the umbrella of “no hydroquinone−no iron−no cardiotoxicity”, other reports argue against a key role for iron and oxidative stress in anthracycline-related cardiotoxicity. Iron overload did not always potentiate DOX-induced cardiotoxicity in vitro in cardiomyocytes and in vivo in mice.18 Iron chelators like deferasirox or aroylhydrazones were shown to chelate and redox-inactivate iron even more specifically than dexrazoxane did, and yet they failed to provide equal or better cardiac protection.19,20 On a different note, dexrazoxane was shown to protect by mechanisms other than iron chelation, e.g., inhibition of topoisomerase IIβ-mediated formation of DNA double-strand breaks (DSB).21 It was in keeping with this notion that the dexrazoxane analogue, ICRF-161 (3-carbon linker bisdioxopiperazine) was able to chelate iron but lacked protective efficacy due to its inability to inhibit topoisomerases.22 Ligand-binding interactions and stability of DOX−Fe(III) complexes in vivo have been questioned.23 And finally, clinical attempts to prevent or mitigate anthracycline-related cardiomyopathy with high-dose antioxidants were largely unsuccessful.24 In the light of these controversial issues, a few basic questions should be addressed. Can we still assume that the apparent cardiac tolerability of PIX is in fact due to its inability to bind iron? Does PIX lack other pharmacokinetic or pharmacodynamic determinants of cardiotoxicity? And more importantly,

3. PIXANTRONE IN CANCER CELLS: FROM TOPOISOMERASE IIα TO SOMETHING ELSE The main mechanism by which anthracycline-like drugs kill tumor cells rests with intercalation into DNA, inhibition of the α isomer of topoisomerase II, and formation of DSB. The latter triggers a number of proapoptotic signals, usually relayed by p53.2 Formation and stability of anthracycline-DNA-topoisomerase IIα complexes are governed by precise structural determinants. In particular, the planar ring system is important for anthracyclines to intercalate into DNA, as the rings B and C with adjacent quinone−hydroquinone moieties overlap with adjacent base pairs of DNA. Removal of the hydroquinone moiety should alter sterical and electrochemical patterns of formation of PIX−DNA−topoisomerase IIα complexes. Accordingly, early preclinical studies showed that PIX was able to intercalate into DNA but caused only moderate levels of DSB and toxicity in human tumor cells.25 In comparison with MITOX PIX exhibited also a reduced cellular uptake,25 which was confirmed in recent studies.13 Such a reduced cellular uptake may well be caused by the lower hydrophobicity of PIX as compared to DOX or MITOX (log P values: 0, 0.8, and 1.4, respectively).13−26 This having been acknowledged, it is worth noting that PIX retained a significant activity in tumor-bearing animals,27,28 as if the in vivo pharmacokinetics and pharmacodynamics of PIX were different from those observed in cellular systems. This might well be the case for the uptake of PIX in cancer cells. In vivo, cellular uptake of PIX should be favored by a moderate-to-weak binding of PIX to plasma proteins, such that the fraction of unbound PIX is high enough to overcome hydrophobicity barriers and to diffuse in cancer cells.29 Recent studies seem to confirm that PIX does not induce a canonical anthracycline-like DNA damage but kills cancer cells by disrupting mitotic fidelity and segregation of genomic material. This was shown by concentration- and timedependent formation of lagging chromosomes, chromosomal bridges, micronuclei, and multinucleated cells. γH2AX foci, markers of DNA damage, were observed in some micronuclei but not in main nuclei. Moreover, apoptosis occurred only after three or four waves of aberrant mitoses and showed a limited dependence on p53.30,31 Other elegant studies maintained that PIX was a good inhibitor of topoisomerase IIα and caused formation of DSB in K562 cells from human chronic myeloid leukemia.13 Interestingly, however, such effects surfaced when PIX was used at ≥5 μM, which was appreciably higher than the transient plasma Cmax of PIX in clinical studies.29,32 We believe this is an important issue that needs to be put into context. The antitumor activity of anthracyclines and congeners correlates much better with plasma exposure over time rather than Cmax.2 In the case of PIX, topoisomerase IIα-independent aberrant mitoses were observed in cells exposed to as low as 100 nM PIX,30,31 a concentration level that reproduced steady-state plasma levels of PIX after Cmax and throughout primary and terminal half-lives.32 It is safe to conclude that higher than pharmacologically relevant concentrations of PIX do inhibit topoisomerase IIα, while lower and clinically more relevant concentrations spare topoisomerase IIα and/or act primarily by altering mitotic fidelity. C



Perspective

4. MECHANISMS OF PIXANTRONE DISRUPTION OF MITOTIC FIDELITY The available data suggest that PIX induces merotelic kinetochore attachments that are responsible for chromosome non-disjunction.31 Approximately 50% of cells with chromosome bridges exhibit paired foci of centromere staining; therefore, these chromosomes fail to segregate but sequester into micronuclei that separate from the nucleus. Some micronuclei contain random pieces of broken chromosomes but these do not contain centromeres.31 Of note, PIX would not induce chromosomal bridges when added to mitotic cells; it only acts when added during interphase. This latter observation suggests that mis-segregation is caused by events that occur during interphase and perturb centromere and kinetochore functions.31 The first round of formation of chromosomal bridges does not commit cells to death. As it was said earlier, cell death occurred only after multiple waves of aberrant mitoses, presumably because mitotic checkpoints do not adequately intercept cells carrying dysfunctional centromeres and kinetochores.30 Repeat rounds of chromosome nondisjunction and micronuclei formation eventually lead to the appearance of highly multinucleated cells, a widely anticipated outcome of mis-segregation events.33,34

6. PUTTING THINGS TOGETHER: FROM CANCER CELLS BACK TO CARDIOMYOCYTES Preclinical studies did not only reveal that PIX was well tolerated in DOX-treated rodents but denoted also its remarkable activity in experimental models of leukemia and lymphomas.29 These two observations offered a solid rationale to develop PIX as monotherapy for patients with NHL that progressed or relapsed after first line therapy with DOX. In a phase 3 study that compared PIX to investigator’s choice chemotherapeutics, PIX offered significantly higher response rates and progression-free survival.1,41 In the light of the unique mode(s) of action of PIX in cancer cells, these clinical findings raise the possibility that PIX is a “new drug” that probably overcomes or bypasses cellular mechanisms of resistance or limited sensitivity to DOX. Moreover, the efficacy of PIX did not come at a cost of significant cardiotoxicity, which only manifested as transient or asymptomatic decrements of left ventricle ejection fraction (LVEF).41 Cardiotoxicity was independent of the cumulative dose of PIX, and again, this denoted a remarkable difference between PIX and anthracycline-like drugs that usually suppress LVEF in a dosedependent manner.42 How is it that PIX did not overlap with and did not exacerbate DOX cardiotoxicity? Patients who would be candidates for PIX usually received ∼300 mg of DOX/m2 as frontline therapy, which is not too far from the cumulative dose of DOX that causes 5% risk of HF (400 mg/m2).42 Few doses of an anthracycline-like drug would easily precipitate cardiotoxicity in patients with a prior exposure to 300 mg of DOX/ m2, but this was not the case for PIX. We believe that the time is mature to look at PIX as a non-anthracycline drug. Should this be the case, as the evidence suggests, molecular mechanisms of the cardiac safety of PIX should be deciphered by looking at mechanisms other than iron or oxidative stress. We suggest that our perception of toxic drug bioactivation should move from reduction of the quinone moiety to oxidation of one or more nucleophiles. In the case of anthracyclines, peroxidase-catalyzed oxidation of the hydroquinone promotes chromophore degradation and inactivation,43 which probably serves a salvage pathway against cardiotoxicity.44,45 In contrast, peroxidase-catalyzed oxidation of amino groups in MITOX and other anthracenediones serves a prevailing mechanism of toxic activation in normal tissues.46 In a translational model of human heart that reproduced sequential cardiac exposure to DOX and anthracenediones, MITOX underwent robust peroxidative metabolism. This led to the apparent formation of highly oxidized diquinone− quinoxaline−peroxyl metabolites that contributed to inactivating the labile Fe−S cluster of mitochondrial aconitase, a sensitive marker of cellular stress.47 Pixantrone, too, was converted to peroxidative metabolites, but their chemical structure was suggestive of a less robust degree of oxidation. This may have been one reason PIX did not inactivate mitochondrial aconitase.47 Other pharmacokinetics findings need to be discussed. The experimental model of human heart was designed to incorporate processes of DOX elimination from the cardiac tissue. This was important to simulate conditions when patients received PIX months or years after the last administration of DOX. Under such defined conditions, PIX uptake was ̈ cardiac significantly lower than that observed in DOX-naive samples, as if transmembrane elimination of DOX from

5. PIXANTRONE AND DNA ALKYLATION In exploring differences between PIX and anthracycline-like drugs, we would next focus on the mechanisms by which DOX, MITOX, or PIX interacts with DNA. Although topoisomerase IIα inhibition remains the most attractive mechanism of cell killing by anthracycline-like drugs, all of them can, in fact, alkylate DNA. Anthracyclines have long been known to form cross-links or adducts to DNA, but these occur only at anthracycline concentrations that exceed those measured in patients.35 Things may change if the nucleophilic amino group of daunosamine conjugates with HCHO (formaldehyde, FORM), a known byproduct of oxidation of lipids, spermine, and other carbon sources. A monomeric DOX−FORM complex uses FORM to covalently bind with a 2-amino group of a G-base on one strand of DNA, and uses hydrogen bonds to bind with a G-base on the opposing strand. This is the so-called virtual cross-linking.2,36 DOX−FORM slows DNA strand exchange by 640-fold relative to anthracycline-free DNA, and by 160-fold relative to DNA bearing intercalated unchanged anthracycline, which denotes the importance of the covalent linkage between drugs and DNA.37 Formaldehyde may also conjugate with nucleophilic secondary amines in the side chain of MITOX, but intrinsic chemical reactivity suggests that PIX would be many times more susceptible to conjugation and formation of stable and longer-lived adducts to DNA.38−40 The primary amino groups of both side chains of PIX are, in fact, characterized by the highest nucleophilic nature and weakest steric hindrance; moreover, the linearity of the side chains renders PIX an authentic bifunctional alkylator that covalently binds to both DNA strands. Possible cause-and-effect relations between such unique mechanisms of DNA alkylation and mitotic aberrations remain unknown at this point in time. Pharmacologic inhibition of pChk1 enhanced growth inhibition effects of PIX, but we do not know whether this was caused by a unopposed mitotic entry of cells that carried centromere−kinetochore abnormalities or durable DNA adducts or both kinds of lesions.30 D



Perspective

Figure 2. Mechanisms of action of PIX in DOX-pretreated cardiomyocytes or cancer cells. Upper panel shows some mechanisms of cardiac tolerability of PIX in DOX-pretreated patients (A, DOX clearance diminishes the cardiac uptake of PIX; B, DOX clearance is accompanied by a diminished availability of PIX for oxidative formation of N-dealkylated or N-cyclized metabolites; C, N-dealkyl PIX inhibits conversion of residual DOX to its secondary alcohol metabolite, DOXOL). The bottom panel shows four possible fates for sister chromatides of dividing cancer cells exposed to PIX (from left to right: normal segregation, normal segregation in micronucleus, mis-segregation in micronucleus, mis-segregation). Following repeat mitoses, mis-segregation events accumulate and apoptosis occurs. Modified with permission from refs 34 and 47.

cardiomyocytes to extracellular fluids introduced barriers to the diffusion of PIX in the opposite direction. The effects of DOX clearance on reducing PIX uptake were not due to a facilitated interaction of PIX with drug efflux pumps. Verapamil, an inhibitor of P-glycoprotein and multidrug resistance protein 1, caused increased myocardial accumulation of DOX but not of PIX.47 The reduced uptake of PIX was likely caused by chemico-physical modifications of the plasma membrane associated with the clearance of DOX and its “flip-flop” reorientation in the lipid bilayer, leading to a decreased membrane penetration by other unrelated compounds.48 This was not observed for MITOX, which diffused equally well in ̈ and DOX-pretreated samples, likely due to its DOX-naive higher lipophilicity. Limited partitioning in DOX-pretreated cardiac tissue is an additional important determinant of the safety of PIX in patients with refractory/relapsed NHL, especially if one appreciates that this was accompanied by a decreased availability of PIX for peroxidative metabolism.47 There may be other reasons PIX did not precipitate cardiotoxicity in anthracycline-pretreated patients. Anthracycline-related HF may occur any time after the last anthracycline dose. Some patients develop HF decades after cancer diagnosis

and treatment. The lifelong character of anthracycline cardiotoxicity is caused by factors that range from mutations in mitochondrial DNA to accumulation of C-13 dihydroxy anthracycline derivatives, also referred to as secondary alcohol metabolites.49 The latter originate from a two-electron reduction of a side-chain carbonyl group and are too polar to be cleared from the heart. Anthracycline secondary alcohol metabolites therefore accumulate and form a sort of long-lived toxic reservoir that primes the heart to sequential stress by chronic health conditions or chemical injuries, including rechallenge by anthracycline-like drugs.49 Secondary alcohol metabolites are formed by a hetereogeneous family of carbonyl or aldehyde reductases that are liable to regulation by other agents. Drugs that occupy a positive regulatory site of the reductases increase alcohol metabolite formation and aggravate the risk of cardiotoxicity, while drugs that compete with DOX for the active site of the reductases decrease alcohol metabolite formation and diminish the risk cardiotoxity.47,50 In the translational model that simulated sequential exposure of human heart to DOX and anthracenediones, PIX inhibited conversion of residual traces of DOX to its alcohol metabolite, doxorubicinol (DOXOL). Inhibition was caused by residual E



Perspective

levels of the most abundant PIX metabolite, N-dealkyl-PIX, which displaced DOX from the active site of reductases.47 Mitoxantrone did not inhibit DOXOL formation. These findings highlighted a paradoxical but beneficial loop in which DOX clearance diminished PIX uptake and metabolization, while remnants of PIX metabolites suppressed conversion of DOX to its cardiotoxic metabolite.47 This scenario might prove to be even more complex. We mentioned that cardiomyocytes exposed to anthracyclines may accumulate DSB that are caused by inhibition of topoisomerase IIβ, which is constitutively expressed in many quiescent tissues. Reports suggest that topoisomerase IIβ inhibition and DNA damage might represent the prevailing mechanism of anthracycline cardiotoxicity.51 In a similar manner, the ability of dexrazoxane to keep topoisomerase IIβ in a closed clamp and inactive configuration might represent the prevailing mechanism of cardioprotection afforded by this compound.52 Unpublished experiments of post-purification crystallization of topoisomerase IIβ-drug complexes suggest that MITOX engages in multiple interactions with topoisomerase IIβ, while PIX does not.53 If confirmed, these findings would unravel an additional determinant of the cardiac safety of PIX.

number of pharmacokinetic and pharmacodynamic characteristics that make it different from anthracyclines or anthracenediones. Pharmaco-economic evaluation by the UK National Institute for Health and Care Excellence54 and by an independent expert panel55 demonstrates that PIX is also cost-effective. There is no reason to refrain from using PIX according to its approved indication. This having been said, we would get a look at possible future developments for PIX. Ex novo or refractory/relapsed aggressive B-cell NHL probably consists of a family of patho-biologically distinct malignancies. Understanding which NHL subtype would most benefit from PIX is of obvious importance. In a real-life study, patients with relapsed NHL were better responders than patients with refractory disease.56 This observation might serve a starting point to optimize benefit from and cost effectiveness of clinical use of PIX. On the other hand, the observation that PIX-induced apoptosis was only in part relayed by p53 suggests a key role for PIX in NHL subtypes that harbor loss-of-function p53 mutations and show resistance to anthracyclines.57 Pixantrone should also be considered for combination with other chemotherapeutics. As clinically relevant concentrations of PIX seem to act by mechanisms that are not confined to inhibition of topoisomerase IIα, one might design combinations of PIX with non-anthracycline topoisomerase IIα inhibitors like etoposide. The two agents would likely cause greater than additive effects in tumor cells. Multiagent therapies might also include investigational inhibitors of mitotic checkpoints or drugs that were known by other mechanisms of action but actually cause effects reminiscent of checkpoint inhibitors. For example, the mixed antimetabolite alkylator, bendamustine, induces a stress response that downregulates the activity of numerous mitotic checkpoints.58 Following promising results in a limited number of patients, a trial of PIX in combination with etoposide and bendamustine has been designed, and accrual will start soon.59 It goes without saying that post-registration development of PIX might offer novel and exciting opportunities. For the time being, a wider use of PIX in its approved indication should be recommended. Pharmacologic and toxicologic reasonings go in that direction.

7. REGULATORY ISSUES, CURRENT PROBLEMS, AND FUTURE DEVELOPMENTS In scrutinizing pharmacological and toxicological features of PIX, we went through a full circle from cardiotoxicity to antitumor activity and back. Everything seems to concur and define PIX as a new entity with unique modes of action in cancer cells and cardiomyocytes (Figure 2). Unfortunately, however, not all that glitters is gold, and the progression of PIX from bench to everyday patients proved to be a difficult one. After evaluating data from the phase 3 study of PIX versus investigator-choice chemotherapeutics for the treatment of patients with refractory/relapsed NHL, the U.S. Food and Drug Administration (FDA) rejected PIX, while the European Medicines Agency (EMA) granted a conditional approval. An in-depth analysis of reasons that led the two agencies to release opposite opinions extends beyond the aims of this Perspective. In brief, the phase 3 study had limitations that the authors themselves recognized.41 The trial recruited fewer patients than planned, most of whom were recruited from outside the U.S., and only half of them had received DOX in combination with an anti-CD 20 antibody (rituximab) that became a standard of care while the trial was already ongoing. Furthermore, cardiac events seemed to be more frequent in patients exposed to PIX, which was intuitively taken to dispel the cardiac safety of PIX but actually reflected that more patients with pre-existing cardiac disease were randomized to PIX.41 Concerns about sample size, population indirectness, and risk of cardiotoxicity may have been pillars of the FDA decision. On the other hand, EMA maintained that refractory/relapsed NHL is a difficult-totreat disease for which no standard therapy exists, and hence, the effects of PIX on prolonging progression-free survival were significant and fulfilled an unmet medical need.1 Following approval by EMA, PIX became available in many European countries, yet its popularity with oncohematologists remains low to moderate. Negative recommendation by FDA may count but perceiving PIX as a “me-too anthracycline” may count even more. We suggest that things merit another perspective. Pixantrone is active in a very unfavorable clinical setting and proves reasonably tolerable in otherwise high risk and heavily pretreated patients. Pixantrone does so due to a

■

AUTHOR INFORMATION

Corresponding Author

*Phone: 011-39-06-225419109. Fax: 011-39-06-22541456. Email: [email protected]. Author Contributions †

P.M. and E.S. contributed equally to this work

Funding

Work in the authors’ laboratory was supported by University Campus Bio-Medico (Special Project “Cardio-Oncology”). Additional research funds were provided by Cell Therapeutics Inc. (Seattle, WA). Notes

The authors declare the following competing financial interest(s): G.M. received lecturing honoraria from Cell Therapeutics Biopharma, USA, and Servier Italy. Biographies Pierantonio Menna is Adjunct Professor of Pharmacology and Laboratory Manager at the Department of Medicine and University Hospital of Campus Bio-Medico of Rome. He earned his degree in Medicinal Chemistry and Ph.D. in Sciences of Aging at the G. d’Annunzio University of Chieti (2000, 2005). He trained in F



Perspective

The myth and the reality of a chemical sphinx. J. Inorg. Biochem. 75, 105−115. (8) Powis, G. (1989) Free radical formation by antitumor quinones. Free Radical Biol. Med. 6, 63−101. (9) Zweier, J. L., Gianni, L., Muindi, J., and Myers, C. E. (1986) Differences in O2 reduction by the iron complexes of adriamycin and daunomycin: the importance of the sidechain hydroxyl group. Biochim. Biophys. Acta, Gen. Subj. 884, 326−336. (10) Lipshultz, S. E., Lipsitz, S. R., Kutok, J. L., Miller, T. L., Colan, S. D., Neuberg, D. S., Stevenson, K. E., Fleming, M. D., Sallan, S. E., Franco, V. I., Henkel, J. M., Asselin, B. L., Athale, U. H., Clavell, L. A., Michon, B., Laverdiere, C., Larsen, E., Kelly, K. M., and Silverman, L. B. (2013) Impact of hemochromatosis gene mutations on cardiac status in doxorubicin-treated survivors of childhood high-risk leukemia. Cancer 119, 3555−3562. (11) Mjos, K. D., Cawthray, J. F., Jamieson, G., Fox, J. A., and Orvig, C. (2015) Iron(III)-binding of the anticancer agents doxorubicin and vosaroxin. Dalton Trans. 44, 2348−2358. (12) Schroeder, P. E., Davidson, J. N., and Hasinoff, B. B. (2002) Dihydroorotase catalyzes the ring opening of the hydrolysis intermediates of the cardioprotective drug dexrazoxane (ICRF-187). Drug. Metab. Dispos. 30, 1431−1435. (13) Hasinoff, B. B., Wu, X., Patel, D., Kanagasabai, R., Karmahapatra, S., and Yalowich, J. C. (2016) Mechanisms of action and reduced cardiotoxicity of Pixantrone; a topoisomerase II targeting agent with cellular selectivity for the topoisomerase IIα isoform. J. Pharmacol. Exp. Ther. 356, 397−409. (14) Hacker, M., McKennon, M., and Singer, J. W. (2009) Lack of iron binding by Pixantrone is associated with reduced production of reactive oxygen species and myoctye cytotoxicity in vitro. Blood 114, 4806. (15) Cavalletti, E., Crippa, L., Mainardi, P., Oggioni, N., Cavagnoli, R., Bellini, O., and Sala, F. (2007) Pixantrone (BBR 2778) has reduced cardiotoxic potential in mice pretreated with doxorubicin: comparative studies against doxorubicin and mitoxantrone. Invest. New Drugs 25, 187−195. (16) Beggiolin, G., Crippa, L., Menta, E., Manzotti, C., Cavalletti, E., Pezzoni, G., Torriani, D., Randisi, E., Cavagnoli, R., Sala, F., Giuliani, F. C., and Spinelli, S. (2001) BBR 2778, an aza-anthracenedione endowed with preclinical anticancer activity and lack of delayed cardiotoxicity. Tumori 87, 407−416. (17) Longo, M., Della Torre, P., Allievi, C., Morisetti, A., Al-Fayoumi, S., and Singer, J. W. (2014) Tolerability and toxicological profile of pixantrone (Pixuvri®) in juvenile mice. Comparative study with doxorubicin. Reprod. Toxicol. 46, 20−30. (18) Guenancia, C., Li, N., Hachet, O., Rigal, E., Cottin, Y., Dutartre, P., Rochette, L., and Vergely, C. (2015) Paradoxically, iron overload does not potentiate doxorubicin-induced cardiotoxicity in vitro in cardiomyocytes and in vivo in mice. Toxicol. Appl. Pharmacol. 284, 152−162. (19) Sterba, M., Popelova, O., Simunek, T., Mazurová, Y., Potácová, A., Adamcová, M., Kaiserová, H., Ponka, P., and Gersl, V. (2006) Cardioprotective effects of a novel iron chelator, pyridoxal 2chlorobenzoyl hydrazone, in the rabbit model of daunorubicin-induced cardiotoxicity. J. Pharmacol. Exp. Ther. 319, 1336−1347. (20) Popelova, O., Sterba, M., Simunek, T., et al. (2008) Deferiprone does not protect against chronic anthracycline cardiotoxicity in vivo. J. Pharmacol. Exp. Ther. 326, 259−269. (21) Lyu, Y. L., Kerrigan, J. E., Lin, C. P., Azarova, A. M., Tsai, Y. C., Ban, Y., and Liu, L. F. (2007) Topoisomerase IIbeta mediated DNA double-strand breaks: implications in doxorubicin cardiotoxicity and prevention by dexrazoxane. Cancer Res. 67, 8839−8846. (22) Martin, E., Thougaard, A. V., Grauslund, M., Jensen, P. B., Bjorkling, F., Hasinoff, B. B., Tjørnelund, J., Sehested, M., and Jensen, L. H. (2009) Evaluation of the topoisomerase II-inactive bisdioxopiperazine ICRF-161 as a protectant against doxorubicin-induced cardiomyopathy. Toxicology 255, 72−79. (23) Gelvan, D., and Samuni, A. (1988) Reappraisal of the association between adriamycin and iron. Cancer Res. 48, 5645−5649.

preclinical pharmacology at the G. d’Annunzio University Center of Excellence in Aging Research (2001−2007) and at the Cancer Pharmacology Center-Mass Spectrometry Unit of the University of Pennsylvania School of Medicine, Philadelphia, PA (2004−2005). In 2015, he qualified in Clinical Pharmacology at La Sapienza University School of Medicine and Pharmacy of Rome. His research interests focus on mass spectroscopy-assisted therapeutic drug monitoring and pharmacokinetics of approved or investigational drugs. Emanuela Salvatorelli is Assistant Professor of Pharmacology at the Department of Medicine of University Campus Bio-Medico of Rome. She received her degree in Pharmacy and Ph.D. in Sciences of Aging at the G. d’Annunzio University of Chieti (2000, 2005). She next underwent research training at the G. d’Annunzio University School of Medicine Center of Excellence in Aging Research (2001−2007). Dr. Salvatorelli’s research aims at optimizing translational models of the human heart to probe cardiac liability of anthracyclines as single agents or in combination with other chemotherapeutics. Giorgio Minotti is Professor of Pharmacology and Head of Clinical Pharmacology at the Department of Medicine and University Hospital of Campus Bio-Medico of Rome. He earned his degrees in Medicine and Oncology at the Catholic University School of Medicine in Rome (1981, 1984). He was fellow of the Italian Association for Cancer Research, Fulbright Scholar, and Fulbright Visiting Professor at the Departments of Biochemistry, Michigan State University (1985− 1987), and Physiology and Biophysics, Case Western Reserve University (1990−1991). His research interests focus on pharmacokinetics, pharmacodynamics, and clinical correlates of cardiovascular toxicity from anticancer drugs.

■

ABBREVIATIONS PIX, pixantrone; NHL, non-Hodgkin’s lymphoma; DOX, doxorubicin; HF, heart failure; MITOX, mitoxantrone; ROS, reactive oxygen species; O2•−, superoxide anion; H2O2, hydrogen peroxide; •OH, hydroxyl radical; DSB, DNA double strand breaks; FORM, formaldehyde; LVEF, left ventricle ejection fraction; DOXOL, C-13 dihydroxy doxorubicin (doxorubicinol); FDA, Food and Drug Administration; EMA, European Medicines Agency

■

REFERENCES

(1) Péan, E., Flores, B., Hudson, I., Sjöberg, J., Dunder, K., Salmonson, T., Gisselbrecht, C., Laane, E., and Pignatti, F. (2013) The European Medicines Agency review of pixantrone for the treatment of adult patients with multiply relapsed or refractory aggressive nonHodgkin’s B-cell lymphomas: summary of the scientific assessment of the committee for medicinal products for human use. Oncologist 18, 625−633. (2) Minotti, G., Menna, P., Salvatorelli, E., Cairo, G., and Gianni, L. (2004) Anthracyclines: molecular advances and pharmacologic developments in antitumor activity and cardiotoxicity. Pharmacol. Rev. 56, 185−229. (3) Thomas, C. E., Morehouse, L. A., and Aust, S. D. (1985) Ferritin and superoxide dependent lipid peroxidation. J. Biol. Chem. 260, 3275−3280. (4) Menna, P., Gonzalez Paz, O., Chello, M., Covino, E., Salvatorelli, E., and Minotti, G. (2012) Anthracycline cardiotoxicity. Expert Opin. Drug Saf. 11, S21−36. (5) Minotti, G. (1993) Sources and role of iron in lipid peroxidation. Chem. Res. Toxicol. 6, 134−146. (6) Gianni, L., Zweier, J. L., Levy, A., and Myers, C. E. (1958) Characterization of the cycle of iron-mediated electron transfer from Adriamycin to molecular oxygen. J. Biol. Chem. 260, 6820−6826. (7) Fiallo, M. M., Garnier-Suillerot, A., Matzanke, B., and Kozlowski, H. (1999) How Fe3+ binds anthracycline antitumour compounds. G



Perspective

(24) Minotti, G., Cairo, G., and Monti, E. (1999) Role of iron in anthracycline cardiotoxicity: new tunes for an old song? FASEB J. 13, 199−212. (25) De Isabella, P., Palumbo, M., Sissi, C., Capranico, G., Carenini, N., Menta, E., Oliva, A., Spinelli, S., Krapcho, A. P., and Giuliani, F. C. (1995) Topoisomerase II DNA cleavage stimulation, DNA binding activity, cytotoxicity, and physico-chemical properties of 2-aza- and 2aza-oxide-anthracenedione derivatives. Mol. Pharmacol. 48, 30−38. (26) Wielinga, P. R., Westerhoff, H. V., and Lankelma, J. (2000) The relative importance of passive and P-glycoprotein mediated anthracycline efflux from multidrug-resistant cells. Eur. J. Biochem. 267, 649− 657. (27) Krapcho, A. P., Petry, M. E., Getahun, Z., Landi, J. J., Jr, Stallman, J., Polsenberg, J. F., Gallagher, C. E., Maresch, M. J., Hacker, M. P., and Giuliani, F. C. (1994) 6,9-Bis[(aminoalkyl)amino]benzo[g]isoquinoline-5,10-diones. A novel class of chromophore-modified antitumor anthracene-9,10-diones: synthesis and antitumor evaluations. J. Med. Chem. 37, 828−837. (28) Pezzoni, G., Beggiolin, G., Manzotti, C., Spinelli, S., Tognella, S., and Giuliani, F. C. (1993) BBR 2778 (a novel aza-analog of anthracenediones endowed with preclinical anticancer activity). Proc. Annu. Meet. Am. Assoc. Cancer Res. 34, A2226. (29) Cell Therapeutics Inc. (2012), Pixantrone dimaleate (BBR 2778), Investigator Brochure Version 12. (30) Beeharry, N., Zhu, X., Murali, V., Smith, M. R., and Yen, T. (2013) Pixantrone induces cell death through mitotic perturbations and subsequent aberrant cell divisions in solid tumor cell lines. Mol. Cancer Ther. 12, A147. (31) Beeharry, N., Di Rora, A. G., Smith, M. R., and Yen, T. J. (2015) Pixantrone induces cell death through mitotic perturbations and subsequent aberrant cell divisions. Cancer Biol. Ther. 16, 1397−1406. (32) Faivre, S., Raymond, E., Boige, V., Gatineau, M., Buthaut, X., Rixe, O., Bernareggi, A., Camboni, G., and Armand, J. P. (2001) A phase I and pharmacokinetic study of the novel aza-anthracenedione compound BBR 2778 in patients with advanced solid malignancies. Clin. Cancer Res. 7, 43−50. (33) Utani, K., Kohno, Y., Okamoto, A., and Shimizu, N. (2010) Emergence of micronuclei and their effects on the fate of cells under replication stress. PLoS One 5, e10089. (34) Thompson, S. L., and Compton, D. A. (2011) Chromosome missegregation in human cells arises through specific types of kinetochore-microtubule attachment errors. Proc. Natl. Acad. Sci. U. S. A. 108, 17974−17978. (35) Gewirtz, D. A. (1999) A critical evaluation of the mechanisms of action proposed for the antitumor effects of the anthracycline antibiotics adriamycin and daunorubicin. Biochem. Pharmacol. 57, 727−741. (36) Taatjes, D. J., and Koch, T. H. (2001) Nuclear targeting and retention of anthracycline antitumor drugs in sensitive and resistant tumor cells. Curr. Med. Chem. 8, 15−29. (37) Zeman, S. M., Phillips, D. R., and Crothers, D. M. (1998) Characterization of covalent adriamycin-DNA adducts. Proc. Natl. Acad. Sci. U. S. A. 95, 11561−11565. (38) Evison, B. J., Mansour, O. C., Menta, E., Phillips, D. R., and Cutts, S. M. (2007) Pixantrone can be activated by formaldehyde to generate a potent DNA adduct forming agent. Nucleic Acids Res. 35, 3581−3589. (39) Evison, B. J., Bilardi, R. A., Chiu, F. C., Pezzoni, G., Phillips, D. R., and Cutts, S. M. (2009) CpG methylation potentiates pixantrone and doxorubicin-induced DNA damage and is a marker of drug sensitivity. Nucleic Acids Res. 37, 6355−6370. (40) Mansour, O. C., Evison, B. J., Sleebs, B. E., Watson, K. G., Nudelman, A., Rephaeli, A., Buck, D. P., Collins, J. G., Bilardi, R. A., Phillips, D. R., and Cutts, S. M. (2010) New anthracenedione derivatives with improved biological activity by virtue of stable drugDNA adduct formation. J. Med. Chem. 53, 6851−6866. (41) Pettengell, R., Coiffier, B., Narayanan, G., de Mendoza, F. H., Digumarti, R., Gomez, H., Zinzani, P. L., Schiller, G., Rizzieri, D., Boland, G., Cernohous, P., Wang, L., Kuepfer, C., Gorbatchevsky, I.,

and Singer, J. W. (2012) Pixantrone dimaleate versus other chemotherapeutic agents as a single-agent salvage treatment in patients with relapsed or refractory aggressive non-Hodgkin lymphoma: a phase 3, multicentre, open-label, randomised trial. Lancet Oncol. 13, 696−706. (42) Salvatorelli, E., Menna, P., Cantalupo, E., Chello, M., Covino, E., Wolf, F. I., and Minotti, G. (2015) The concomitant management of cancer therapy and cardiac therapy. Biochim. Biophys. Acta, Biomembr. 1848, 2727−2737. (43) Reszka, K. J., Wagner, B. A., Teesch, L. M., Britigan, B. E., Spitz, D. R., and Burns, C. P. (2005) Inactivation of anthracyclines by cellular peroxidase. Cancer Res. 65, 6346−6353. (44) Cartoni, A., Menna, P., Salvatorelli, E., Braghiroli, D., Giampietro, R., Animati, F., Urbani, A., Del Boccio, P., and Minotti, G. (2004) Oxidative degradation of cardiotoxic anticancer anthracyclines to phthalic acids. Novel function or ferrylmyoglobin. J. Biol. Chem. 279, 5088−5099. (45) Menna, P., Salvatorelli, E., and Minotti, G. (2007) Doxorubicin degradation in cardiomyocytes. J. Pharmacol. Exp. Ther. 322, 408−419. (46) Blanz, J., Mewes, K., Ehninger, G., Proksch, B., Waidelich, D., Greger, B., and Zeller, K. P. (1991) Evidence for oxidative activation of mitoxantrone in human, pig, and rat. Drug Metab. Dispos. 19, 871−880. (47) Salvatorelli, E., Menna, P., Paz, O. G., Chello, M., Covino, E., Singer, J. W., and Minotti, G. (2013) The novel anthracenedione, pixantrone, lacks redox activity and inhibits doxorubicinol formation in human myocardium: insight to explain the cardiac safety of pixantrone in doxorubicin-treated patients. J. Pharmacol. Exp. Ther. 344, 467−478. (48) Lameh, J., Chuang, R. Y., Israel, M., and Chuang, L. F. (1989) Nucleoside uptake and membrane fluidity studies on N-trifluoroacetyladriamycin-14-O-hemiadipate-treated human leukemia and lymphoma cells. Cancer Res. 49, 2905−2908. (49) Minotti, G., Salvatorelli, E., and Menna, P. (2010) Pharmacological foundations of cardio-oncology. J. Pharmacol. Exp. Ther. 334, 2−8. (50) Salvatorelli, E., Menna, P., Cascegna, S., Liberi, G., Calafiore, A. M., Gianni, L., and Minotti, G. (2006) Paclitaxel and docetaxel stimulation of doxorubicinol formation in the human heart: implications for cardiotoxicity of doxorubicin-taxane chemotherapies. J. Pharmacol. Exp. Ther. 318, 424−433. (51) Zhang, S., Liu, X., Bawa-Khalfe, T., Lu, L. S., Lyu, Y. L., Liu, L. F., and Yeh, E. T. (2012) Identification of the molecular basis of doxorubicin-induced cardiotoxicity. Nat. Med. 18, 1639−1642. (52) Vejpongsa, P., and Yeh, E. T. (2014) Prevention of anthracycline-induced cardiotoxicity: challenges and opportunities. J. Am. Coll. Cardiol. 64, 938−945. (53) Cell Therapeutics Biopharma, data on file. (54) National Institute for Health and Care Excellence (2014), Pixantrone monotherapy for treating multiply relapsed or refractory aggressive non-Hodgkin’s B cell lymphoma, NICE technology appraisal guidance [TA306], http://www.nice.org.uk/guidance/ TA306. (55) Muszbek, N., Kadambi, A., Lanitis, T., Hatswell, A. J., Patel, D., Wang, L., Singer, J. W., and Pettengell, R. (2016) The Costeffectiveness of Pixantrone for Third/Fourth-line Treatment of Aggressive Non-Hodgkin’s Lymphoma. Clin. Ther. 38, 503−515. (56) Eyre, T. A., Linton, K. M., Rohman, P., Kothari, J., Cwynarski, K., Ardeshna, K., Bailey, C., Osborne, W. L., Rowntree, C., Eden, D., Shankara, P., Eyre, D. W., Jasani, P., Chaidos, A., Collins, G. P., and Hatton, C. S. (2016) Results of a multicentre UK-wide retrospective study evaluating the efficacy of pixantrone in relapsed, refractory diffuse large B cell lymphoma. Br. J. Haematol. 173, 896−904. (57) Møller, M. B., Gerdes, A. M., Skjødt, K., Mortensen, L. S., and Pedersen, N. T. (1999) Disrupted p53 function as predictor of treatment failure and poor prognosis in B- and T-cell non-Hodgkin’s lymphoma. Clin. Cancer Res. 5, 1085−1091. (58) Leoni, L. M., Bailey, B., Reifert, J., Bendall, H. H., Zeller, R. W., Corbeil, J., Elliott, G., and Niemeyer, C. C. (2008) Bendamustine (Treanda) displays a distinct pattern of cytotoxicity and unique H



Perspective

mechanistic features compared with other alkylating agents. Clin. Cancer Res. 14, 309−317. (59) d’Amore, F., Joergensen, J., Blok Sillesen, I., Segel, E., Bentzen, H., Thorsgaard, M., Jacobsen Pulczynski, E., Pedersen, B. B., Gillström, D., Clausen, M., Kamper, P., Silkjaer, T., Gormsen, L. C., Leppä, S., and Toldbod, H. (2014) Preliminary Clinical Experience on the Efficacy and Feasibility of a New Combination Regimen Consisting of Pixantrone, Etoposide, and Bendamustine with or without the Addition of Rituximab in Patients with Relapsed/Refractory Aggressive Non-Hodgkin Lymphomas. Blood 124, 5435.

I


Pixantrone beyond Anthracyclines - American Chemical Society

Recommend Documents