The Anthracyclines: Will We Ever Find a Better Doxorubicin?
`
`Raymond B. Weiss
`
`The anthracyclines are the class of antitumor drugs
`with the widest spectrum of activity in human cancers,
`and only a few cancers (eg, colon cancer) are unrespon-
`sive to them. The first two anthracyclines were devel-
`oped in the 19605. Doxorubicin (DOX) differs from
`daunorubicin (DNR) only by a single hydroxyl group.
`This fact has spurred researchers worldwide to find
`analogs of DOX that have less acute toxicity, cause
`less cardiomyopathy, can be administered orally,
`and/or have different, or greater, antitumor efficacy.
`Five DOX/DNR analogs are marketed in other coun-
`tries, and one (idarubicin) is available in the United
`States. None of these analogs have stronger antitumor
`efficacy than the original two anthracyclines, but there
`are some differences in toxicity. Methods have been
`fashioned to keep the peak plasma level of DOX muted
`to minimize cardiotoxicity, but the only apparently
`effective method available so far (prolonged drug
`infusion) is cumbersome. The bisoxopiperazine class
`of drugs (especially dexrazoxane) provides protection
`against anthracycline-induced cardiomyopathy and has
`much promise for helping mitigate this major obstacle
`to prolonged use of the anthracyclines. The DOX
`analogs being evaluated in the 19905 have been se-
`lected for their ability to overcome multidrug resis-
`tance in cancer cells. Thirty years after discovery of the
`anticancer activity of the first anthracycline, some
`means of reducing anthracycline toxicity have been
`devised. Current studies are evaluating increased doses
`of epirubicin to improve anthracycline cytotoxicity,
`while limiting cardiotoxicity, but at present DOX still
`reigns in this drug class as the one having the most
`proven cancerocidal effect.
`This is a US government work. There are no restric-
`tions on its use.
`
`HE BIFUNCTIONAL alkylating agents
`were the first drugs proven to have clinical
`anticancer activity. One of them, cyclophospha—
`mide, remains one of the most widely used
`drugs today, even though it was synthesized
`nearly 40 years ago. The other most widely used
`
`
`i From the Departments of Medicine, Walter Reed Army
`Medical Center, Washington, DC; and the Uniformed Services
`University of the Health Sciences, Bethesda, MD.
`Address reprint requests to Raymond B. Weiss, MD, Section
`of Medical Oncology, Walter Reed Army Medical Center,
`Washington, DC 20307.
`The opinions erpressed in this article are solely those of the
`author and are not necessarily those ofany government agency.
`This is a US government work. There are no restrictions on
`its use.
`0093- 7754 /92 / 1906-0007
`
`class of antitumor agents today is the anthracy-
`clines, and few patients with malignancies do
`not receive one of the anthracyclines at some
`point in their clinical course. The history of
`these drugs also goes back to the 19505.
`Farmitalia Research Laboratories of Milan,
`Italy (now part of the Montedison Group con-
`glomerate) began an organized effort in the
`mid—19503 to find anticancer compounds pro-
`duced by soil microbes. In 1958, a Farmitalia
`employee collected a random soil sample from
`the grounds of the Castel del Monte, a 12th-
`century castle that is a local tourist attraction
`near Andria in southeastern Italy. From this soil
`sample was grown a newly recognized species of
`Streptomyces, which produced a bright red pig-
`ment. Di Marco isolated an antibiotic from this
`
`fungus1 that had striking activity against a wide
`spectrum of murine cancers.2 It also had antifun-
`gal and antibacterial properties, but
`it was
`selected for clinical development because of its
`antitumor effects. Di Marco et a1 named this
`
`antibiotic daunomycin, using the name (Daunii)
`of a pro-Roman tribe that once resided in the
`Andria region.
`Also in the early 1960s, Dubost et al at the
`laboratories of Rhéne—Poulenc in suburban Paris
`
`independently isolated a new antibiotic from a
`different species of Streptomyces that also pro-
`duced a red pigment.3 They named their new
`antibiotic rubidomycin, using the French word
`rubis, for ruby. On public presentation of their
`new discoveries, both groups of researchers
`recognized that they had identified the same
`substance. The term daunorubicin (DNR) was
`later coined and adopted as the international
`nonproprietary name (INN) to give equal credit
`to both discoverers. By accepted convention, all
`new anthracyclines are given the suffix of
`“rubicin” in the INN system. The name anthra-
`cycline was created by Brockman in the late ‘
`1950s based on the presence of an anthraqui-
`none chromophore and the polycyclic ring sys—
`tem in the chemical structure, which is similar
`to that of tetracycline.
`Clinical trials of DNR began in 1964 in the
`respective countries of origin and at the Memo-
`
`Genentech 2102
`
`570
`
`Hospira v. Genentech
`Seminars in Once/ogy, Vol 19, No 6 (December), 1992: pp 670-686 IPR2017'00737
`
`Genentech 2102
`Hospira v. Genentech
`IPR2017-00737
`
`1
`
`

`

`THE ANTHRACYCLINES
`
`671
`
`rial Sloan-Kettering Cancer Center in the United
`States. At this time, hematologists treating leu-
`kemia and lymphoma were usually the only
`physicians who used drugs for cancer, and DNR
`was tested mostly in patients with these dis—
`eases. This circumstance was actually serendipi-
`tous because DNR was found by Tan to have
`high activity for acute leukemia,4 and this can-
`cer is still the only one for which it is effective.
`By 1967, French and American investigators
`recognized that DNR could produce fatal car—
`diac toxicity,“ which is still the major obstacle
`for prolonged use of anthracyclines.
`Comparison of the structure of DNR to an
`anthracycline predecessor of DNR (rhodomy-
`cin) showed the Farmitalia investigators that
`minor changes in the chemical structure could
`alter the biological activity of this drug class. A
`colleague and collaborator of Di Marco, Ar—
`camone, then began an effort to develop ana-
`logs of DNR that might also have antitumor
`effects. Arcamone et a1 subjected the Streptomy-
`ces that produced DNR to the mutagenic effects
`of N-nitroso-N-methyl urethane and derived a
`strain that produced a different red-colored
`antibiotic.6 Arcamone named this substance
`
`Adriamycin after the Adriatic Sea, which is only
`a few kilometers from the Castel del Monte.
`
`Because Adriamycin was a registered trade
`name, the INN doxorubicin (DOX) was later
`coined for this new agent. Di Marco et al
`showed that DOX had greater activity than
`DNR against some murine cancers and a better
`therapeutic index.7
`DiMarco, who had moved to the Istituto
`Nazionale Tumori in Milan, then took a puri-
`fied supply of Adriamycin for study to his
`clinical colleague, Bonadonna. Serendipity again
`played a strong role because the first patient
`treated with this new drug by Bonadonna had a
`metastatic fibrosarcoma.8 After one dose of the
`
`drug this patient’s pulmonary metastases re-
`gressed. Twenty-five years later, DOX is still the
`most active drug available for treatment of
`sarcomas. The subsequent high antitumor activ—
`ity of DOX observed by Bonadonna et al9 was
`confirmed by Tan et al10 and many others in the
`United States. In 1974, only 6 years after the
`first patient received this drug, DOX was ap-
`proved for marketing in the United States.
`
`DOX remains today the antitumor agent with
`the widest spectrum of antitumor activity. The
`world of oncology owes a large debt of gratitude
`to the Italian investigators, Di Marco, Ar—
`camone, and Bonadonna, who led the research
`
`with these two anthracyclines in the 19605.
`There is a difference of only a single hydroxyl
`group between the chemical structures of DNR
`and DOX,
`in otherwise complex molecules.
`Despite the minor difference in structure, there
`is a great difference in clinical antitumor activ-
`ity. DNR has little activity in carcinomas and
`sarcomas,11 whereas DOX is one of the most
`effective drugs for these cancers. However,
`there are a few cancers where DOX is ineffec-
`
`tive (eg, colon cancer, melanoma, chronic leuke-
`mias, and renal cancer). In addition, the cumu—
`lative cardiotoxicity limits the duration of DOX
`use to approximately 9 months at usual doses,
`and most cancers will develop resistance to it.
`The singular success in developing DOX and
`its limitations in clinical use have been the basis
`
`for investigators worldwide trying to develop a
`better DOX. This research has followed three
`
`major pathways. One has been the creation of
`new anthracycline analogs, in hopes of emulat-
`ing the success of Arcamone. This process has
`continued from the 1960s to the present, and
`will into the future. No one has kept count of
`the new anthracycline analogs synthesized over
`the past 25 years, but it probably numbers well
`in excess of 2,000, and more are reported every
`month. Over 400 analogs have been synthesized
`in one group (the AD series) alone (personal
`communication, Mervyn Israel, July 1992), and
`553 had been evaluated in the screening pro—
`gram at the National Cancer Institute (NCI) by
`1991 (personal communication, Edward Acton,
`September 1991). Another method has been to
`administer some agent in conjunction With DOX,
`either to protect against cardiotoxicity or to
`overcome drug resistance by the cancer. Finally,
`DOX has been modified “mechanically” (eg, by
`the use of liposomes) to minimize heart expo-
`sure to the drug while maintaining antitumor
`efficacy. The major question to be addressed in
`this article is, have (or will) any of these meth-
`ods really given us a better DOX. Although
`there have been many failures in the attempts to
`improve on DOX, in the past few years some of
`
`2
`
`

`

`672
`
`the developments allow a qualified “yes” to this
`question.
`‘
`
`DOXORUBICIN (DOX) ANALOGS
`
`The scientific and commercial success of
`
`Arcamone in adding a single hydroxyl group to
`an active drug and turning it into DOX has been
`the driving force for developing other new
`anthracycline analogs. Analogs have been ob-
`tained from fungi isolated from soil samples or
`rationally synthesized based on known structure-
`activity relationships. A variation of the fungal
`isolation method is to subject the parent Strepto—
`myces organism to mutagens so that new com-
`pounds are created from genetic code modifica-
`tions. Most of this work has been done in
`
`Europe and Japan, and the only Clinically suc-
`cessful analogs developed so far have come
`from Italy and Japan. The focus of such analog
`development has been to find an anthracycline
`that
`is less cardiotoxic,
`is more able to be
`absorbed orally, has less acute toxicity, or has
`activity in cancers resistant to DOX. In the past
`23 years, such a search has led to creation of
`several dozen anthracyclines with promise for
`clinical advantages that unfortunately were not
`borne out in clinical studies. Table 1 lists some
`
`of these agents and the reasons why they are no
`longer in clinical trial. There are other anthracy-
`cline analogs that either are marketed or in
`current trials and have promise for advantages
`over DOX. These will be discussed individually.
`
`Table 1. Some Anthracyclines Tested Clinically and Found to
`Have No Advantages Over DOX
`Name Findings in Clinical Studies
`
`
`Esorubicin
`
`Ouelamycin
`
`Carminomycin Antitumor activity appears inferior to DOX.
`Detorubicin
`Synthesis difficult. No advantages over
`DOX.
`Antitumor activity appears inferior to DOX.
`No less cardiotoxic than DOX.
`Marcellomycin Myelosuppression erratic in phase Itrials.
`No phase ii trials performed.
`Phase I trials showed both acute and chronic
`iron overloading toxicity resulting in he-
`machromotosis. No phase II trials per-
`formed.
`Phase I trials showed both cardiotoxicity
`and nephrotoxicity. No phase II trials per-
`formed.
`
`Rodorubicin
`
`NOTE. None of thesetanthracyclines are in current clinical
`use.
`Abbreviation: DOX, doxorubicin.
`
`RAYMOND B. werss
`
`Idarubicin
`
`The only anthracycline marketed in the
`United States besides DNR and DOX is idaru—
`bicin (IDA). This agent was synthesized by
`Arcamone12 in 1976 and is actually a DNR
`analog because its only difference from DNR is
`the deletion of the methoxyl group at the C-4
`position on Ring D. The INN was derived from
`its Italian name, 4-demetossidaunorubicina. An-
`imal tumor studies were conducted to compare
`IDA with its parent, DNR, and they showed
`that IDA had greater antitumor activity at lower
`drug concentrations.” IDA was found to have
`higher affinity for lipids than other anthracy—
`clines, which suggested that good oral absorp-
`tion was possible. Because both DNR and DOX
`are poorly absorbed orally and must be adminis-
`tered intravenously, oral administration of IDA
`might provide a distinct advantage.
`Phase I trials of IDA in both intravenous and
`
`oral formulations showed that an approximately
`3.5 X greater dose of IDA orally was necessary
`to produce an equivalent myelotoxic effect to
`that of the intravenous drug. Phase II trials with
`both formulations indicated that IDA was ac-
`
`tive in acute leukemias (both myeloid and
`lymphatic) and some carcinomas, notably breast
`cancer.13 As might be expected, the response
`rates were highest
`in patients who had not
`previously received an anthracycline.l4
`Although oral IDA has been studied in pa—
`tients with acute leukemia, there is no advan-
`"tage to this administration route in a disease
`where insertion of a central venous access
`
`device is standard procedure as soon as the
`diagnosis is made. However,
`the activity of
`intravenous IDA in acute myeloid leukemia
`(AML) naturally raised the question, is it any
`better than the parent compound? Only prospec—
`tive randomized trials comparing IDA with
`DNR, both in conjunction with cytarabine,
`could answer this question. There have been
`four such major randomized trials comparing
`these drugs vis-a—vis, each in a slightly different
`dose and/or age populationfiilé The results in
`these studies are not consistent. Two of the four
`
`studies showed a statistically significant advan—
`tage for IDA in complete remission rate and
`two did not. In the three studies where the data
`
`were provided, none showed any difference in
`median duration of response. Finally, two of the
`
`3
`
`

`

`THE ANTHRACYCLINES
`
`673
`
`four studies showed no difference in surviv-
`
`al.15’16 What is puzzling about these trials is that
`the outcomes (response rate, response dura-
`tion, and overall survival) for the IDA-treated
`groups are not really superior to those achieved
`in a large cooperative study using DNR plus
`cytarabine for induction therapy.l7 Moreover,
`the results of the DNR-treated patients in the
`randomized trials are inferior to those in this
`
`cooperative study.”17 On the other hand, when
`there is any difference in the results of the
`randomized trials, IDA is always superior, so
`the clinician is left with uncertainty about the
`question supposedly to be solved by these stud-
`ies: does IDA have a greater efficacy than DNR,
`or is it a more expensive “me too” drug? The
`one conclusive point in these comparisons is
`that there is no meaningful toxicity difference
`between the two analogs.
`DNR is an essential component of therapy
`for acute lymphocytic leukemia (ALL), as it is in
`AML. IDA has activity in ALL at a rate equiva-
`lent
`to that of DNR,” but no randomized
`comparisons of the two analogs have been
`performed in this disease.
`Oral IDA appears to have efficacy in breast
`cancer,14 in contrast with DNR, although DNR
`has had only cursory study in this disease. 11 If an
`oral anthracycline is effective as a palliative
`therapy for breast cancer,
`it would have a
`clearcut advantage over DOX. Does oral IDA
`represent the hoped—for “better doxorubicin”?
`Unfortunately, the answer seems to be no.
`A randomized comparison of oral IDA versus
`intravenous DOX in 76 patients has shown a
`statistically significant,
`inferior response rate
`for the oral IDA.18 In patients who had received
`no prior chemotherapy, the response rate was
`60% for DOX versus 29% for IDA; in patients
`previously treated,
`the response figures were
`29% versus 12%, respectively. This low re—
`sponse rate of oral IDA has been observed in
`single-arm studies also”:20
`A major problem with oral IDA is the varia—
`tion in bioavailability. Stewart et 21121 showed
`that the oral bioavailability ranged from a low of
`12% to a high of 49% in a series of nine
`patients. These data indicate that there can be
`up to a fourfold difference in ability to absorb
`the drug, with commensurate variations in toxic—
`ity (especially hematologic) and antitumor effi—
`
`cacy. This oral bioavailability issue could be the
`explanation for the response rates in breast
`cancer being lower than DOX.”20
`There are some advantages to oral IDA that
`could make it attractive for use in selected
`
`patients. IDA has activity in indolent lympho-
`mas,22 which is a disease where a convenient
`oral palliative drug could be useful as treatment
`for patients not responding to alkylator therapy.
`Myelodysplastic syndrome is another indolent
`disease that could be treated with oral IDA.
`
`IDA might be useful as part of outpatient
`treatment for chronic myelogenous leukemia or
`as maintenance therapy for AML. The apparent
`lower degree of alopecia associated with oral
`IDA18'20 is also a worthy attribute. There also
`may be a somewhat smaller risk of cardiotoxic-
`ity, but a dose limit for heightened risk of such
`toxicity has not been established. However,
`before such oral
`therapy becomes recom-
`mended, it would be worthwhile to study the
`bioavailability problem in more depth. There
`may be phenotypic metabolic variations that
`could be determined before treatment, with
`commensurate adjustments in drug dose to
`minimize toxicity and maximiZe antitumor effi—
`cacy. The IDA patent is held by Farmitalia
`Carlo Erba, and Adria Laboratories, the Amer—
`ican subsidiary of Farmitalia, does not have
`plans at present to pursue marketing of oral
`IDA in the United States but is interested in
`trials that might find oral IDA a niche in cancer
`treatment.
`
`Epimbicin. (EPI)
`
`In their ongoing search for anthracycline
`analogs in the 1970s, Arcamone et a123 modified
`the aminosugar moiety of DOX and created an
`epimer of the C—4’ hydroxyl group on the
`aminosugar (Table 2). The positional change in
`this hydroxyl group is the sole difference from
`DOX. It was selected for further development
`because its murine and human xenograft antitu-
`mor activity was equivalent to DOX, but it had
`less cardiotoxicity.24 These features suggested
`an improved therapeutic index over DOX.
`When phase I testing of epirubicin (EPI) was
`begun in Milan,25 the same every 3-week sched-
`ule used with DOX was also used for this
`
`analog. However, the tolerable dose range estab—
`lished was 70 to 90 mg/m2, which is equimyelo-
`
`4
`
`

`

`Table 2. Anthracyclines Marketed in the United States and/or Other Countries
`Chemical Structure
`Name
`Where Marketed
`Disease Indication
`
`Daunorublcin
`
`Worldwide
`
`Acute leukemias
`
`Doxorubicin
`
`Worldwide
`
`A wide cross-sectlon ol
`carcinomas, lymphomas,
`and sarcomas
`
`Worldwide
`(intravenous only)
`
`Acute leukemias
`
` ‘
`
`rm,
`
`.
`
`o
`
`0
`ll
`c—cH,—l‘oH.‘
`gage “
`u
`\
`on
`
`
`
`Worldwide
`(except U.S.)
`
`A wide cross-section of
`carcinomas, lymphomas,
`and sarcomas
`
`
`
`Japan, France
`
`Carcinomas, lymphomas,
`and sarcomas
`
`ldarubicin
`
`Epirubicin
`
`Pirarubicin
`
`Aclarublcin
`
`Zorubicln
`
`
`
`Japan, France
`
`Acute leukemias and
`non-Hodgkin's lymphomas
`
`Acute leukemias
`
`
`
`NOTE. The dotted line in each chemical structure encircles the point where a difference from daunorubicin exists. The carbon atoms
`are numbered and the rings are lettered in each structure configuration.
`
`5
`
`

`

`THE ANTHRACYCLINES
`
`675
`
`toxic to the 60 to 75 mg/m2 dose range of DOX.
`Identical to DOX, nadir blood counts occurred
`
`10 to 12 days after each dose with recovery by 21
`days. Pharmacokinetic studies26 showed that
`EPI is more rapidly and extensively metabolized
`than DOX with its alcohol metabolite (epirubici-
`nol) being formed to a greater degree.
`In
`addition, EPI forms more glucuronides than
`DOX because of the positional change of the
`C-4’ hydroxyl group. This glucuronidation facil—
`itates the excretion process.26 As a result, the
`terminal elimination phase of EPI is shorter
`than DOX by an average of 10 hours,
`thus
`producing a higher plasma clearance. These
`features of EPI pharmacokinetics are important
`clinically because they may explain the toxicity
`advantage that EPI has.
`Phase II trials of EPI conducted in the 19803
`
`have established the fact that EPI has activity in
`the human cancers for which DOX is active and
`is inactive in the same tumors.”28 Because
`
`DOX is rarely used as a single agent, EPI also
`was evaluated in combination regimens, particu-
`larly for breast cancer. In two large prospective,
`randomized trials comparing EPI with DOX as
`part of combination therapy for metastatic breast
`cancer
`(both groups received cyclophospha-
`mide and 5-fluorouracil to make CAF or CEF),
`there was no difference in response rate,
`re—
`sponse duration, or survival.”30 Therefore, it is
`clear that EPI has not proved to have an
`antitumor efficacy advantage over DOX and is
`disappointing in this regard.
`EPI was selected for clinical development
`because it appeared to be less cardiotoxic than
`DOX. Does EPI have any toxicity advantages
`over DOX? In considering this point one must
`keep in mind the drug doses used in any
`randomized comparison of these two anthracy-
`clines. Are the doses equimolar or equimyelo—
`toxic (a 15 to 20 mg/m2 greater dose for EPI)?
`This point is very important because if equimo—
`lar doses are compared, EPI will always show
`lower toxicity, both acute and chronic. Such an
`outcome was evident in the two studies compar-
`ing CEF with CAP.”30 Both used equimolar
`doses of EPI and DOX, and both showed less
`acute toxicity and cardiotoxicity for EPI. When
`the drugs are compared at equimyelotoxic doses,
`there is little difference in noncardiac toxicity.31
`EPI clearly produces cardiomyopathy that
`
`can even be fatal sometimes-‘2’33 The morpholog-
`ical features of EPI-induced cardiomyopathy
`are identical to those induced by DOX.33 The
`total dose range in which the cardiotoxicity risk
`increases precipitously (akin to the 550 mg/m2
`cumulative dose limit of DOX) is 900 to 1,000
`mg/m2.32“34 This total dose is ideally reached
`with approximately 11 doses of EPI adminis-
`tered over some 9 months,
`if 85 mg/m2 are
`administered at an every 3—weeks schedule.34 If
`a lower dose of EPI is used (eg, 50 mg/m2) when
`it is incorporated into a combination regimen
`such as CEF,”3O then it will ideally take 17
`doses administered over some 12 months to
`
`reach the toxic range. When one considers that
`the median duration of response to CAF or
`CEF regimens in breast cancer is 8 to 10
`months,29:30’35 one can see that most patients will
`have tumor progression and require a change in
`therapy before they have a serious cardiotoxic-
`ity risk from EPI.
`CEF with a dose of EPI equimolar to that of
`DOX in CAF also produces less acute nausea
`and vomiting,”30 an important factor in patient
`compliance with therapy and quality of life,
`apparently without compromising therapeutic
`efficacy. Therefore, EPI seems to provide a
`marginal toxicity advantage over DOX.
`EPI is now widely marketed in Canada,
`Japan, Australia, and Europe, but not in the
`United States, and has largely supplanted DOX
`in clinical use. A New Drug Application (NDA)
`was submitted to the Food and Drug Adminis-
`tration (FDA) in 1985 but was not approved,
`probably because the cardiotoxicity advantage
`of EPI was modest and applied to only a
`minority of patients whose cancer still had not
`progressed after approximately 10 months of
`therapy. The emesis induced by chemotherapy
`is very distressing to patients. If a CEF regimen
`has equivalent efficacy to CAF, but induces less
`vomiting, EPI should be available for clinical
`use in the United States, even if its therapeutic
`spectrum is not different from DOX.
`DOX is usually administered in doses of 60 to
`75 mg/m2 when it is used alone and at 45 to 50
`mg/m2 when used in combination. Increasing
`the DOX dose to 90 to 135 mg/m2 results in
`more total, and more complete, responses in
`breast cancer.36 However, the cumulative dose
`limit for cardiotoxicity is reached more quickly
`
`6
`
`

`

`676
`
`RAYMOND B. WEISS
`
`with such doses, and congestive heart failure
`can result.36 Symptomatic acute mucositis is also
`a problem.
`The maximum tolerated dose (MTD) of EPI
`in the phase I trials conducted 13 years ago25
`was 70 to 90 mg/m2. Drug dose escalation has
`become a topic of intense research interest in
`oncology in the 1980s, and attempts have been
`made to increase EPI doses. These efforts are
`
`based on the premise that the moderate acute
`and chronic toxicity advantages of EPI over
`DOX might allow escalation of EPI doses with
`commensurate enhanced therapeutic efficacy at
`more moderate toxicity cost than similar stud-
`ies36 with DOX. Moreover, the concomitant use
`of bone marrow protective agents (such as
`filgrastim) and cardiotoxicity protectors (such
`as dexrazoxane [DXZ]) may allow even further
`dose escalations of EPI.
`Initial studies of EPI dose escalation in-
`
`creased the dose to 120 mg/mz, with which
`mucositis became dose—limiting, whereas myelo—
`suppression was not.37 Subsequent studies have
`increased the dose to 180 mg/mz,
`in which
`myelosuppression did become dose—limiting and
`caused neutropenic fevers in 23% of 27 pa—
`tients.38 Complete or partial tumor responses
`occurred in 85% of the patients. A study by
`Bastholt et al39 comparing four EPI doses in
`prospective randomized fashion for breast can-
`cer has been preliminarily reported. The re-
`sponse rate was improved by increasing the EPI
`dose from 40 to 60 mg/m2 to 90 mg/m2, but not
`when it was increased to 135 mg/mz. In addi—
`tion, some patients had their dose increased
`after achieving no tumor regression at a lower
`EPI dose, and about 25% then had a respOnse.39
`The higher number of responses achieved
`with both DOX and EPI administered at height-
`ened dOSes are encouraging, but they will not
`necessarily achieve better survival in patients
`with stage IV breast cancer. On the other hand,
`they could be advantageous for patients with
`locally advanced disease. Further work with
`EPI dose escalation using the toxicity pro-
`tectants is ongoing.
`
`Pirambicin (PRA)
`
`Umezawa, at the Institute of Microbial Chem-
`istry in Tokyo, spent a long, illustrious career
`
`‘ searching for antibiotics, particularly those with
`antitumor activity. In 1966 he isolated bleomy-
`cin, and in the 1970s he turned his attention to
`
`finding new anthracyclines. One of the anthracy-
`clines he discovered was a tetrahydropyranyl
`derivative of DOX, which was initially labeled
`THP-Adriamycin.40 The INN now is pirarubicin
`(PRA; the pronunciation of the “pyra” segment
`was internationalized with different spelling).
`Development of this analog has occurred prima-
`rily in Japan and France.
`Preclinical tumor efficacy studies of PRA by
`Tsuruo et al41 showed general equivalence or
`superiority to DOX. Cardiotoxicity studies per—
`formed in France42 indicated a lesser degree of
`cardiotoxicity than DOX in experimental ani-
`mals, and thus PRA was brought to clinical trial.
`The dose and schedule for this analog used in
`phase II trials has been 50 to 70 mg/m2, admin—
`istered at 3-week intervals. Granulocytopenia is
`dose~limiting. PRA has undergone study in
`acute leukemia, lymphoma, sarcoma, and breast
`cancer, and the response rates have been equiv-
`alent
`to those achieved from DOX.”45 The
`
`spectrum and degree of its antitumor efficacy
`appear similar to DOX. When used in a combi-
`nation regimen for breast cancer,46 the response
`rate is equivalent to that of CAP.
`Preclinical studies suggested a lesser degree
`of cardiotoxicity with PRA, but clinical studies
`have not been performed to establish this point.
`In particular, no randomized prospective com-
`parisons of PRA with any other anthracycline
`(such as has been done with EPI31’34) have been
`reported. Also, the cumulative dose at which
`the cardiotoxicity risk becomes significant
`is
`poorly characterized.
`The one attribute of PRA that could incite
`
`clinical interest is the lower rate of total alope—
`cia. When PRA is used as a single agent, total
`alopecia is uncommon, and it does not develop
`as a cumulative toxicity.”45 Even when PRA is
`administered in combination with 5 ~fluorouracil
`
`and cyclophosphamide,46 total alopecia oc-
`curred in only about half of the patients. DOX
`is well known to produce total alopecia in nearly
`every patient treated. Most oncologists have
`encountered patients (especially women) who
`are extremely reluctant to accept any chemother—
`apy that might cause total alopecia. PRA could
`
`7
`
`

`

`THE ANTHRACYCLINES
`
`677
`
`be an acceptable substitute for DOX or EPI
`(with equivalent antitumor activity) that would ‘
`enhance quality of life.
`The patent on PRA is owned by Meiji Seika
`Kiasha, Ltd. of Tokyo, Japan. It is marketed in
`Japan and, under licensing agreements, in Eu-
`rope. No American company has developed any
`interest in pursuing the clinical trials necessary
`to establish a toxicity superiority of PRA over
`DOX. The one attractive feature of causing less
`alopecia may not be commercially promising
`enough to invest the necessary effort and re-
`sources for seeking marketing approval in the
`United States.
`
`Zorubicin and Aclarubicin
`
`isolated by
`Zorubicin (rubidazone) was
`Rhone-Poulenc investigators in France in 1969
`and was found to have activity in acute leuke-
`mia.47 This analog is marketed in France (Table
`2). It was studied by American investigators and
`found to have no advantage over DNR or DOX,
`so no further development for marketing in the
`United States has been done.
`
`Aclarubicin (aclacinomycin) is another ana-
`log found by Umezawa in the mid-1970s. It also
`has activity in acute leukemias and is marketed
`(Table 2) for this cancer in France and Japan.
`Studies in the United States did not show any
`advantages over DNR or DOX, so all clinical
`trials have been closed.
`
`nia, which was rapidly reversible.50 DOX causes
`severe stomatitis at high doses, and during dose
`escalation. This toxicity feature of iododoxorubi-
`cin (IODO) favors the combining of the anthra—
`cycline with hematologic growth factors, such as
`filgrastim or sargramostim, thus allowing fur—
`ther dose escalation with perhaps greater antitu—
`mor activity.51 This analog might even be useful
`as part of a preparatory regimen for marrow
`transplant.
`Only a few phase II studies of IODO have
`been published, and they have not been full of
`promise for this agent“)52 In a trial of patients
`with advanced breast cancer receiving a 70
`mg/m2 dose, the response rate was only 10%,
`and there was negligible activity in colon and
`lung cancer.52 Another trial51 using a dose of 80
`mg/m2 gave a 35% response rate in advanced
`breast cancer, but this rate was achieved at the
`expense of a 34% incidence of grade 4 granulo-
`cytopenia. This hematologic toxicity was also
`more common with repeated treatment. DOX
`can produce a similar response rate as a single
`agent in breast cancer but without such severe
`hematologic toxicity.
`At this point, IODO appears not to provide
`any clearcut advantage over DOX, although
`testing in a variety of cancers has not been
`performed yet. The only promising path for
`further development may be as a constituent of
`marrow ablative therapy.
`
`Iododoxombicin (IODO)
`
`AD-32
`
`This anthracycline is an analog of DOX
`synthesized by Arcamone and colleagues in the
`mid—1980s.48 It was selected for clinical develop—
`ment because it had activity against DOX-
`resistant P388 leukemia, had greater activity
`than DOX in some preclinical tumor lines, and
`had more rapid cellular uptake than DOX. It
`also had less cardiotoxicity in preclinical test-
`ing.49 The sole structural difference from DOX
`is the presence of an iodine atom at the C—4’
`position, instead of a hydroxyl group.
`Phase I trials were performed in Europe in
`the late 19805, and some phase II trials have
`now been completed. A clinically promising
`feature of the toxicity profile, established in
`phase I
`testing, was the fact
`that
`the sole
`dose-limiting acute toxicity was granulocytope-
`
`A series of DOX analogs have been synthe—
`sized by Israel et al, first at Dana-Farber Cancer
`Institute, and subsequently at the University of
`Tennessee. One of the early compounds in this
`series, AD-32, was created in 1973 and had
`greater antitumor activity,
`less cardiotoxicity,
`and less toxicity in general than DOX in preclin—
`ical testing.53 A phase I trial of intravenously
`administered AD—32 was then performed,54 but
`drug formulation and solubility problems pre—
`vented further clinical development.
`AD—32 has recently been resurrected as an
`intravesical treatment for bladder cancer be-
`
`cause it appears not to cause local tissue injury
`if extravasated and is poorly absorbed systemi—
`cally when administered in the bladder.55 Phase
`I trials have been performed,55 and phase II
`
`8
`
`