`Oral Chemotherapy: Rationale and Future Directions
`
`By Mark D. DeMario and Mark J. Ratain
`
`Purpose and Methods: The expanding role of oral
`chemotherapy in oncology is suggested by the abun-
`dance of orally formulated agents currently in develop-
`ment. The pharmacoeconomic principles that drive oral
`drug formulation are discussed. Patient preference for
`oral therapy is identified as a second major impetus for
`the design of oral cytotoxics. While the rationale for oral
`formulations is apparent, substantial patient compli-
`ance and pharmacokinetic limitations have been identi-
`fied for this route of administration. Specific aspects of
`bioavailability limitations and patient compliance are
`discussed. Relevant pharmacokinetic data for each
`orally formulated chemotherapy agent are compared
`and selected novel oral cytotoxics and cytotoxic modu-
`lators are discussed.
`Results: A review of pharmacokinetic literature sug-
`gests substantial variability in bioavailability for many
`orally formulated cancer cytotoxics. While these find-
`ings are observed for all classes of oral drugs, the issue
`is especially critical for cancer chemotherapy, in which a
`narrow therapeutic index is frequently observed. Im-
`proved bioavailability and reduced interpatient biovari-
`
`AS WITH VIRTUALLY ALL office-based and academic
`
`medicine, the issue of cost-effectiveness must be
`addressed by the medical oncologist. Cost may no longer be
`considered as an aside, but will likely become a central issue
`in therapeutic decision-making, particularly in the limita-
`tions of the palliative setting. At the beginning of the next
`century, it is estimated that health care expenses will amount
`to $1.5
`trillion dollars or 15% of the projected Gross
`National Product.' In 1990, an estimated $35 billion was
`spent on direct costs for cancer care. 2 The direct cost of
`chemotherapy agents represented only a small fraction of
`this sum. A cost survey of one outpatient cancer center noted
`direct chemotherapy charges accounted for only 5.4% of
`total Medicare cost claims.2 In an analysis of adjuvant and
`metastatic breast cancer settings, direct drug charges ac-
`counted for only 19% to 36% of total care costs.3 These
`examples of the relatively small contribution of chemother-
`apy agents to the overall cost of oncology care are represen-
`tative of national health care expenditures as a whole. In
`1991, prescription drugs represented only 6.4% of total
`health care expenditures.4
`While cost-effectiveness has been substantiated for both
`adjuvant and palliative chemotherapy, 5-8 it is likely these
`therapies will continue to be scrutinized in terms of their
`component charges, primarily direct drug costs and adminis-
`
`ability are therefore desirable for new cytotoxic formu-
`lations. Pharmacologic manipulations to improve
`bioavailability and reduce costs are examined.
`Conclusion: Oral chemotherapy represents a funda-
`mental change in contemporary oncology practice,
`driven by pharmacoeconomic
`issues, patient conve-
`nience, and the potential for improved patient quality of
`life. Novel cytostatic therapies that require protracted
`drug administration periods will also favor an oral
`formulation. While the use of oral chemotherapy may
`initially be limited to metastatic disease palliation, dem-
`onstration of equivalent efficacy would allow for its
`subsequent use in adjuvant settings. This efficacy is
`contingent on circumventing bioavailability limitations
`and patient noncompliance. The development of spe-
`cific, low-toxicity inhibitors of CYP3A4, P-glycoprotein
`(P-gp), and other drug metabolizing enzymes such as
`dihydropyrimidine dehydrogenase represents a major
`innovative step in the successful formulation of oral
`chemotherapy.
`J Clin Oncol 16:2557-2567. © 1998 by American
`Society of Clinical Oncology.
`
`tration costs. Administration costs traditionally have incorpo-
`rated some or all of the following charges: hospitalization,
`physician fees, salaries of nursing and technical support
`personnel, infusion equipment supply costs, administration
`supply product fees, and overhead charges (capital equip-
`ment, regulatory compliance costs, and liability insurance
`costs). In the evolving era of managed care, the impact of
`cost-effective strategies on these charges is evident to many
`in practice-based oncology. From 1991 to 1996, the average
`Medicare reimbursement for a chemotherapy administration
`declined 57% from $155.51 to $67.01.9 Additionally, the
`Health Care Finance Administration (HCFA) formally elimi-
`nated payment for all supplies used to administer chemother-
`apy in 1992. In a recent appraisal of cost and reimbursement
`patterns for commonly used infusional and bolus lymphoma,
`
`From the Section ofHematology/Oncology, Department of Medicine,
`Committee on Clinical Pharmacology, and Cancer Research Center
`University of Chicago, Chicago, IL.
`Submitted August 12, 1997; accepted March 24, 1998.
`Address reprint requests to Mark J. Ratain, MD, University of
`Chicago Medical Center, 5841 S Maryland Ave, MC 2115, Chicago, IL
`60637-1470; Email mjratain@mcis.bsd.uchicago.edu.
`© 1998 by American Society of Clinical Oncology.
`0732-183X/98/1607-0002$3.00/0
`
`Journal of Clinical Oncology, Vol 16, No 7 (July), 1998: pp 2557-2567
`
`2557
`
`Downloaded from ascopubs.org by 63.247.71.66 on January 9, 2017 from 063.247.071.066
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`West-Ward Exhibit 1084
`DeMario 1998
`Page 001
`
`
`
`2558
`
`breast, and colon regimens, Lokich et all' noted Medicare
`endorsement averaged just 21%
`to 36 % of total charge
`claims. The larger implications for reimbursement denials of
`novel, off-label, or investigational infusional regimens are
`obvious. Pharmacoeconomics, heralded by Medicare, and,
`increasingly, private-payer capitation constraints will pro-
`mote increasingly cost-effective treatment regimens. From
`the pharmacoeconomic viewpoint, the rapid emergence of
`orally available oncologic agents might be as readily antici-
`pated as the transition from hospital- to outpatient-based
`chemotherapy administration.
`An administrative cost analysis of ganciclovir therapy
`provides an illustration of the economic issues for orally
`formulated agents.,n1 2 A late complication of AIDS, cyto-
`megalovirus retinitis requires a protracted ganciclovir treat-
`ment course. Sullivan et all 2 have examined cost issues of
`ganciclovir administered over an 80-day schedule. While the
`direct drug costs of oral ganciclovir were nearly 170% those
`of the intravenous formulation, mean oral total therapy costs
`averaged just 58% those of intravenous therapy ($4,938 v
`$8,583). Cost variables examined include administration
`and nursing, central intravenous catheter, and infusion pump
`costs, and costs related to adverse effects of intravenous
`ganciclovir administration, such as line sepsis. A significant
`cost differential was noted for induction, adverse-event
`treatment, and combined home care, nursing administration,
`and monitoring charges, for which oral therapy costs repre-
`sented only 54%, 36%, and 14% of intravenous therapy
`costs, respectively. 12 Thus, prescription of oral ganciclovir
`results in increased revenues for its manufacturer and
`decreased revenues to health care providers and hospitals.
`Aside from economic considerations, it seems intuitive
`that patient preferences and quality-of-life issues will be
`another major impetus for the development of orally formu-
`lated chemotherapy. Recognizing the limitations of cyto-
`toxic therapy in many metastatic cancers, patient quality of
`life is increasingly becoming a central consideration in
`palliative treatment regimens. The issue of patients' prefer-
`ences regarding chemotherapy administration has recently
`been examined by Liu et al.' 3 Patients with advanced
`malignancies were asked about their preferred method of
`treatment before initiation of chemotherapy. Of 103 patients,
`more than 90% indicated a preference for orally adminis-
`tered agents, provided significant reductions in efficacy or
`duration of response would not result from this mode of
`treatment. The reasons for patients' preferences included
`convenience (57%), current concerns or previous difficulties
`with intravenous access lines (55%), or preference to control
`the chemotherapy administration environment (33%). If
`equivalent safety and efficacy are demonstrated, an all oral
`regimen would likely first be realized in the palliative
`setting, for example, an oral fluoropyrimidine replacing
`
`DeMARIO AND RATAIN
`
`intravenous fluorouracil (5-FU) in the treatment of meta-
`static colorectal carcinoma.
`
`DRUG ADMINISTRATION CONSIDERATIONS
`To have efficacy in an oral formulation, a chemotherapeu-
`tic drug must be sufficiently bioavailable. Bioavailability
`concerns the rate and extent to which a drug is absorbed into
`the systemic circulation. The bioavailability of oral agents is
`contingent on adequate intestinal absorption and the circum-
`vention of intestinal and, subsequently, hepatic metabolic
`systems. In considering absorption, the limitations of satura-
`bility and structural stability in gastric and intestinal pH
`must be addressed. The importance of saturable absorption
`is exemplified by oral etoposide and oral leucovorin. The
`large bioavailability variation between low- and high-dose
`oral etoposide suggests a saturable absorption system.
`Hande et a114 noted a 76% ± 22% bioavailability of 100 mg
`etoposide, while a 400-mg dose was only 48% ± 18%
`bioavailable. A linear increase of etoposide bioavailability at
`doses up to 200 mg has been demonstrated by other
`investigators, which confirms the saturable absorptive kinet-
`ics. 15 Leucovorin represents a second agent for which
`limited oral bioavailability is clinically relevant. While
`leucovorin's relative bioavailability is 78% in a 40-mg oral
`dose, bioavailability is reduced to only 31% in a 200-mg
`dose.' 6 These constraints would limit the use of oral
`leucovorin in some commonly used adjuvant regimens for
`colon carcinoma.
`The importance of structural stability in gastric and
`intestinal pH is also demonstrated by the etoposide model.
`Etoposide stability is optimal at pH 5. Joel et al' 7 attempted
`to improve oral etoposide bioavailability through coadminis-
`tration of agents that improve stability of etoposide in
`intestinal fluid (ethanol or bile salts) or increase pH (cimeti-
`dine). 17 Although the bioavailability and area under the
`concentration-time curve (AUC) of etoposide were not
`significantly enhanced with the modifications imposed in
`this study, it is nevertheless valuable to consider pH
`manipulations for structural stability in etoposide and,
`potentially, other oral chemotherapeutic drugs.
`An ideal chemotherapeutic drug would have little interpa-
`tient variability (in absorption and AUC) and, more impor-
`tantly, little intrapatient variability with successive doses.
`Comparing bioavailability from 143 studies in several drug
`classes (including antineoplastics), Hellriegel et al' 8 noted a
`significant inverse correlation between decreasing absolute
`bioavailability and intersubject variability in absolute bio-
`availability. It follows that caution must be taken in prescrib-
`ing an oral chemotherapy drug with low bioavailability. The
`generally narrow therapeutic index of these agents means
`significant interpatient variability would predispose some
`
`Downloaded from ascopubs.org by 63.247.71.66 on January 9, 2017 from 063.247.071.066
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`West-Ward Exhibit 1084
`DeMario 1998
`Page 002
`
`
`
`ORAL CHEMOTHERAPY: A REVIEW
`
`individuals to excessive toxicity, or, conversely, inadequate
`efficacy.
`There are numerous examples of clinically significant
`variability in bioavailability for orally formulated chemo-
`therapeutic agents. While etoposide bioavailability averages
`50%, significant interpatient and intrapatient biovariability
`produces a bioavailability range of 25% to 75%.19 For
`busulfan, Hassan et al20 have noted up to a twofold and
`sixfold variability in bioavailability in adult and young
`pediatric populations, respectively. Similarly, Lebbe et al21
`observed a greater than fivefold range of AUC in patients
`who received low-dose oral methotrexate. Noting a fivefold
`interpatient variability in mercaptopurine AUC following
`oral administration, Zimm et al22 questioned the efficacy of
`the standardized oral dose used as maintenance therapy for
`acute lymphoblastic leukemia.
`
`Other Bioavailability Limitations Specific for Oral Drugs
`Intestinal CYP3A4. As mentioned previously, intestinal
`metabolic systems represent a "first-pass" bioavailability
`limitation unique to oral drugs. Principal among these is
`enterocyte CYP3A4. CYP3A4 is an enzyme subtype of
`CYP3, itself one of three major subclasses of cytochrome
`P-450 enzymes (CYPI, CYP2, and CYP3). The P-450
`enzymes are primarily responsible for phase I metabolism,
`facilitating drug excretion through nonconjugation reac-
`tions. CYP3A represents the major P-450 subclass, which
`accounts for as much as 25% of phase I drug metabolism in
`humans. 23 While hepatocytes contain the highest concentra-
`tions of P-450, these enzymes have been isolated from the
`endoplasmic reticulum of several cell types. Four distinct
`intestinal CYP3A P-450s have been characterized in hu-
`mans, CYP3A3, CYP3A4, CYP3A5, and CYP3A7. These
`enterocyte P-450s lie below the surface of the jejunal
`microvillus. In humans, CYP3A4 is the major intestinal
`P-450. 24
`While it is difficult to distinguish between the activity of
`enteric and hepatic CYP3A4, the importance of this entero-
`cyte enzyme in oral bioavailability limitations is suggested
`by clinical studies of cyclosporine metabolism. The brief
`anhepatic phase of orthotopic liver transplantation provides
`a system to analyze isolated metabolic activity of enterocyte
`CYP3A4. Measuring cyclosporine Ml and M2 metabolites
`following jejunal drug delivery, Kolars et al25 concluded that
`as much as 50% of cyclosporine is metabolized by entero-
`cyte P-450 3A4.
`Table 1 lists representative chemotherapeutic substrates
`of CYP3A4. Orally formulated etoposide and cyclophospha-
`mide are subject to substantial first-pass metabolism by
`enterocyte CYP3A4. The exact contribution of enterocyte
`CYP3A4 in the metabolism of oral etoposide is not known;
`however, inhibition studies discussed below suggest the
`
`2559
`
`Table 1. Investigational and Available Oral Chemotherapeutic Agents
`That Are Substrates of CYP3A4 and P-gp
`
`Agent
`
`CYP3A4
`
`P-gp
`
`References
`
`Cyclophosphamide
`Etoposide
`Idarubicin
`Paclitaxel*
`Topotecan*
`Vinorelbine
`
`*Investigational agents.
`
`x
`x
`
`x
`
`x
`
`x
`x
`x
`x
`x
`
`26
`27, 28
`29
`28, 30
`31
`32, 33
`
`possibility for a substantial role.34 Further investigation will
`be required to elucidate the role of enterocyte CYP3A4 in
`the first-pass metabolism of other orally available cytotoxic
`agents.
`In determining the potential pharmacokinetic significance
`of enterocyte CYP3A4 for a given oral agent, it is necessary
`to consider the significant interpatient variability in CYP3A4
`expression. It is known that hepatic CYP3A4 concentrations
`and catalytic activity vary at least 10-fold among individu-
`als. 35 This has practical implications in the metabolism of
`many drugs, including cyclosporine. Likewise, Lown et al 36
`have reported a greater than sixfold interpatient variability in
`enterocyte CYP3A4 metabolic activity. This substantial
`heterogeneity may also have important clinical implications
`in oral drug metabolism. While hepatic CYP3A4 catalytic
`activity may be quantified by the erythromycin breath test
`(ERMBT), this test is not useful for enterocyte CYP3A4.3 6
`Developing a noninvasive, quantitative probe for enterocyte
`CYP3A4 will be necessary to further our understanding of
`this enzyme's contribution in oral drug metabolism.
`Perhaps the most important reason for fully elucidating
`CYP3A4 metabolism is that bioavailability may be substan-
`tially enhanced through pharmacologic manipulations of
`this system. CYP3A4 has numerous inducers, principally
`rifampin, phenobarbital, and dexamethasone. 23,37 Con-
`versely, erythromycin, quinidine, ketoconazole, and cyclos-
`porine serve as inhibitors of CYP3A4.38-40 An appreciation
`of the utility of these agents may be gained by examining the
`cyclosporine model. Herbert et a141 demonstrated a signifi-
`cant reduction in cyclosporine oral bioavailability with the
`concomitant administration of rifampin, whereas Gupta et
`al 42 markedly increased oral cyclosporine bioavailability
`with the coadministration of erythromycin. Similarly, Gomez
`et al43 noted coadministration of ketoconazole increased oral
`cyclosporine bioavailability from 22% to 56%. The cost
`implications of increasing cyclosporine bioavailability
`through CYP3A4 inhibition have been examined. Keogh et
`al44 noted an annual savings of $5,200 per patient for cardiac
`transplant patients who received cyclosporine coadminis-
`tered with ketoconazole.
`Kobayashi et a134 have recently examined the effect of
`ketoconazole modulation on the metabolism of etoposide.
`
`Downloaded from ascopubs.org by 63.247.71.66 on January 9, 2017 from 063.247.071.066
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`West-Ward Exhibit 1084
`DeMario 1998
`Page 003
`
`
`
`2560
`
`Etoposide
`is known
`to be metabolized by CYP3A4-
`mediated O-demethylation.27 With coadministration of the
`potent CYP3A4 inhibitor ketoconazole, a 44% increase in
`plasma etoposide AUC was noted.34 This model suggests
`CYP3A4 modulation may be of great value in improving the
`oral bioavailability of chemotherapeutic agents that are its
`substrates.
`Intestinal P-glycoprotein. P-glycoprotein (P-gp) is a
`drug efflux pump well known to confer chemotherapy
`resistance. Orally formulated cytotoxic drugs known to be
`P-gp substrates include etoposide, idarubicin, and topotecan
`(investigational formulation).29,31 Encoded by MDR1, P-gp
`has been isolated in several human tissues, including the
`intestinal mucosa.45 There are preclinical data that suggest
`P-gp limits the intestinal absorption of paclitaxel, docetaxel,
`and vinblastine.46,47 Moreover, Leu et al48 reported a signifi-
`cant increase in etoposide bioavailability in rat jejunal and
`ileal loops first exposed to intravenous quinidine, a known
`P-gp inhibitor. The effect was realized at quinidine concen-
`trations at or below the therapeutic range. A monoclonal
`antibody against P-gp similarly produced a marked decrease
`in etoposide efflux from this rat jejunal system. Two
`preclinical studies of oral paclitaxel have demonstrated the
`significant bioavailability limitations
`imposed by P-gp.
`Sparreboom et al47 noted a sixfold increase in paclitaxel
`AUC in mdr la (-/-) mice, which lack intestinal P-gp. Using
`a murine model, Van Asperen et al49 have recently examined
`the effects of the P-gp blocker SDZ PSC 833 on the oral
`bioavailability of paclitaxel. The effect of this potent P-gp
`blocker was substantial; SDZ PSC 833 pretreatment resulted
`in a 10-fold increase in paclitaxel AUC. The bioavailability
`limitations imposed by P-gp must therefore be considered in
`the development of orally formulated chemotherapy. Trials
`that use P-gp inhibitors such as quinidine, verapamil, or
`SDZ PSC 833 may be anticipated in the successful formula-
`tion of oral paclitaxel or other agents.
`Table 1 lists representative chemotherapeutic substrates
`of P-gp.28
`
`Patient Compliance Issues
`
`Aside from bioavailability limitations, patient noncompli-
`ance represents a second potential major obstacle for orally
`formulated chemotherapy. Bonadonna and Valagussa 5o un-
`derscored
`the implications of an incomplete treatment
`course in the adjuvant breast setting; markedly inferior
`disease-free survival was experienced in patients who re-
`ceived less than 65% of planned therapy. Various studies
`have examined noncompliance in patient self-administered
`chemotherapy regimens. Lebovits et a151 noted a noncompli-
`ance rate of 43% in 51 breast cancer patients treated over 26
`weeks with an outpatient, oral cyclophosphamide regimen.
`Factors associated with higher rates of noncompliance
`
`DeMARIO AND RATAIN
`
`included lower socioeconomic status or treatment in a
`community-based setting. Levine et al 52 examined outpa-
`tient allopurinol and prednisone compliance in a cohort of
`patients who received concomitant chemotherapy for hema-
`tologic malignancies. With no interventions, a full allopuri-
`nol compliance rate of only 17% was noted in patients
`responsible for one component of a potentially curative
`treatment regimen.52 Moreover, pharmacokinetic analysis
`showed actual compliance was less than half that suggested
`by patient self-report. Importantly, measures designed to
`increase compliance, including patient education, home
`psychologic support, and exercises in pill taking, were able
`to increase compliance nearly threefold. The impact of side
`effects, complexity of treatment regimen, and age on compli-
`ance rates has subsequently been investigated.53 The occur-
`rence, frequency of occurrence, or severity of physical side
`effects correlated with clinic appointment noncompliance,
`but not with compliance of self-administered chemotherapy
`medications.
`For self-administered oral regimens, quantifying compli-
`ance rates will be essential for the accurate determination of
`regimen efficacy. In this regard, a novel electronic model has
`been suggested by Lee et al.54 An electronically activated
`tablet bottle scored opening to indicate daily outpatient
`compliance with an oral chemotherapy agent. A compliance
`rate of 110.6% ± 20.6% was demonstrated in a cohort of 21
`patients responsible for a self-administered component of an
`outpatient lymphoma chemotherapy regimen. It is also
`encouraging that, using the identical electronic model, these
`investigators were able to demonstrate a 93.2% ± 12%
`compliance rate in a cohort of small-cell lung carcinoma
`patients receiving low-dose oral etoposide. 55 The high
`compliance rate was maintained despite the cohort's gener-
`ally poor overall prognosis.
`
`ORALLY AVAILABLE AGENTS AND OVERVIEW OF
`SELECTED NOVEL CYTOTOXICS
`
`Table 2 lists relevant pharmacokinetic and cost data for
`the orally formulated chemotherapeutic agents. For several
`of these agents, bioavailability varies substantially between
`patients, with increasing drug dose, or with food. Given the
`narrow therapeutic index for many of these agents, it is
`desirable to improve bioavailability and reduce intrapatient
`biovariability through mechanisms discussed earlier. Where
`applicable, an oral/intravenous cost ratio is given for equiva-
`lent milligram doses of drug at average wholesale prices.
`These ratios do not consider the extensive costs of intrave-
`nous drug administration. While price data for the investiga-
`tional oral agents are not currently available, it is likely their
`price will be considerably greater than that of the intrave-
`nous preparation, as in the previously discussed case of
`
`Downloaded from ascopubs.org by 63.247.71.66 on January 9, 2017 from 063.247.071.066
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`West-Ward Exhibit 1084
`DeMario 1998
`Page 004
`
`
`
`Table 2. Comparison of Oral Chemotherapeutic Agents
`
`Oral
`Bioavailability
`(%)
`
`0-80
`122
`100
`61
`
`16-50
`14-46
`20-90
`100
`
`48-76
`85
`100
`100
`
`58-95
`100
`70-80
`
`Terminal
`Half-life
`(hrs)
`
`0.22
`4.5
`6-16
`0.80
`
`1.5
`11
`3-15
`2.0-3.0
`
`6.6
`2.8
`0.16
`1.3-2.9
`1.3-2.5
`0.95-1.8
`2.3-3.4
`2.0
`
`6.0-7.7
`
`Cost Ratio
`(oral/
`intravenous)
`
`Comments
`
`122% mean bioavailability and 4.5-hour half-life
`reflect 5-FU modulation with 5-ethynyluracil
`
`Extensive first-pass gut metabolism to thiouric acid
`
`4.56
`
`Bioavailability substantially limited over 30 mg/m2
`
`0.46
`1.53
`
`0.15
`
`Bioavailability reduced at doses >100 mg
`
`Half-lifes are of cis- and trans-4-OH CCNU
`metabolites, respectively
`
`Half-life of platinum in ultra-filtrate with 120
`mg/m2/d x 5 JM216 schedule
`
`Undergoes reversible pH-dependent conversion to
`hydroxy acid
`A colloidal dispersion is poorly bioavailable
`
`2561
`
`References
`
`56-61
`
`62-65
`66
`
`22, 67, 68
`69-71
`72, 73
`70
`
`14
`74
`70,75
`76
`
`77, 78
`71,79, 80
`78
`
`81, 82, 83
`
`84, 85
`
`86-88
`
`89, 90
`
`ORAL CHEMOTHERAPY: A REVIEW
`
`Class
`
`Agent
`
`Pyrimidine antimetabolites
`
`5-FU
`
`Other antimetabolites
`
`Epipodophyllotoxins
`Alkylating agents
`
`UFT
`Capecitabine
`
`Mercaptopurine
`Thioguanine
`Methotrexate
`Hydroxyurea
`
`Etoposide
`Cyclophosphamide
`Procarbazine
`Lomustine
`
`Melphalan
`Busulfan
`Chlorambucil
`
`Platinum derivatives
`
`JM-216
`
`Vinca alkaloids
`
`Vinorelbine
`
`26-45
`
`24-56
`
`Camptothecins
`
`Topotecan
`
`30
`
`2.82
`
`9-aminocamptothecin
`
`27-49
`
`7.5-24.3
`
`ganciclovir. Selected novel oral cytotoxic agents are now
`discussed.
`
`Novel Oral Fluoropyrimidines and Modulators
`
`UFT UFT is an oral preparation of 1-(2-tetrahydrofuryl)-
`5-FU (tegafur) and uracil admixed in a 1:4 molar ratio. UFT
`is a unique oral prodrug that undergoes both hepatic P-450
`and target tissue metabolism to 5-FU. 91 The 1:4 tegafur-to-
`uracil ratio was noted to provide the highest intratumoral
`levels of 5-FU.92 The oral bioavailability of UFT is excel-
`lent; Antilla et a162 noted a comparative oral/intravenous
`AUC of 115% ± 8%. Phase I investigations of 28-
`consecutive-day administration with concomitant high- or
`low-dose leucovorin have suggested feasibility and accept-
`able toxicity at doses less than 350 mg/m 2/d.93,94 UFT has
`been used extensively in Japan. Examining pooled phase II
`of more than 400 patients with cholangiocarcinoma, colorec-
`tal, gastric, and breast cancers, Ota et al95 noted single-agent
`UFT response rates that ranged from 25% to 32%.95 More
`recently, Pazdur et a196 noted a 42% response rate in
`therapy-naive colon carcinoma patients. Comparing oral
`ftorafur and intravenous 5-FU in a small cohort of patients
`with advanced colorectal cancer, Andersen et al97 noted
`equivalent gastrointestinal toxicity with significantly lower
`hematologic toxicity in ftorafur-treated patients. There was
`
`no significant difference in median or overall survival in
`UFT- or 5-FU-treated patients.97 Confirmatory phase III
`trials will be necessary to justify UFT in a metastatic or,
`ultimately, adjuvant colorectal setting. The ability to admin-
`ister UFT over protracted schedules may allow for the
`simulation of continuous infusion 5-FU kinetics, which
`suggests a role as a radiosensitizing agent.
`Capecitabine. Capecitabine represents an orally bioavail-
`able fluoropyrimidine designed with unique tumor selectiv-
`ity. An oral 5'-deoxy-5-fluorocytidine derivative, it is ab-
`sorbed unchanged from intestinal mucosa and converted
`to 5' deoxy-5-fluorocytidine ribonucleotide (5'-DFCR) via
`hepatic acylamidases. The drug is subsequently converted to
`5'-DFUR by cytidine deaminase, an enzyme preferentially
`located in hepatic and tumor tissues. Finally, the conversion
`of 5' deoxy-5-fluorouridine ribonucleotide. (5'-DFUR) to
`5-FU occurs intratumorally via pyrimidine nucleoside phos-
`phorylases (PyNPases), uridine phosphorylase, and thymi-
`dine phosphorylase.98,99 Specificity is achieved with PyN-
`Pase, which is constitutively expressed at high levels within
`tumors.
`Capecitabine phase I data have been determined for
`6-week continuous and intermittent administration sched-
`ules.100 A maximum-tolerated dose (MTD) of 1,657 mg/
`m 2/d was determined for capecitabine on a 6-week daily
`
`Downloaded from ascopubs.org by 63.247.71.66 on January 9, 2017 from 063.247.071.066
`
`Copyright © 2017 American Society of Clinical Oncology. All rights reserved.
`
`West-Ward Exhibit 1084
`DeMario 1998
`Page 005
`
`
`
`2562
`
`administration schedule. At this dose level, plasma 5-FU
`concentrations were similar to those obtained for continuous
`intravenous infusion 5-FU. Direct examination of resected
`colorectal tumor and surrounding tissue showed a tumor-to-
`normal tissue capecitabine ratio of 2.9 in pretreated pa-
`tients. 101 Phase II studies that examined capecitabine on
`12-week continuous and 12-week intermittent schedules
`with and without leucovorin showed response rates of 19%,
`28%, and 24%, respectively. 1 02 Based on these results, phase
`III trials of capecitabine versus single-agent 5-FU might be
`anticipated in the setting of metastatic colon carcinoma.
`Capecitabine recently has been approved by the Food and
`Drug Administration for use in refractory breast carcinoma.
`5-Ethynyluracil (eniluracil). 5-FU has long been recog-
`nized as an active agent in the treatment of a variety of
`human malignancies.
`Improved efficacy for 5-FU has
`previously been achieved with modulatory agents such as
`leucovorin and N-(phosphonacetyl)-L-asparatate
`(PALA).
`5-Ethynyluracil (eniluracil) is a novel modulator shown to
`improve the efficacy of 5-FU in preclinical models through
`the selective, irreversible inhibition of dihydropyrimidine
`dehydrogenase (DPD) metabolism. In individuals with nor-
`mal levels of DPD, 5-FU is quickly metabolized in both its
`intravenous and oral formulations, which results in a parent
`drug half-life of only 5 to 20 minutes.103 Moreover, the
`extremely variable bioavailability (0% to 80%) of oral 5-FU,
`which limits its clinical use, is felt to be a reflection of
`It is
`intrapatient variability in intestinal DPD activity.'"
`therefore anticipated that a DPD inhibitor such as 5-ethynylu-
`racil would increase both 5-FU half-life and oral bioavailabil-
`ity. A recent phase I investigation showed complete absorp-
`tion (average bioavailability, 122%) and a substantially
`prolonged half-life (4.5 hours) for oral 5-FU when it was
`concomitantly administered with 5-ethynyluracil.61 The in-
`terpatient variability of 5-FU bioavailability was also re-
`duced (coefficient of variation [CV], 33%), while systemic
`5-FU clearance was reduced more than 20-fold to a rate
`comparable to that of renal blood flow. This study noted
`neutropenia as the dose-limiting toxicity at a dose of 25
`mg/m 2/d on a daily-times-five every-28-day schedule. The
`impact of improved 5-FU oral bioavailability and prolonged
`half-life has been demonstrated in preclinical models, which
`have shown up to a sixfold increase in the 5-FU therapeutic
`index with coadministration of 5-ethynyluracil.105,106 A
`recent phase II investigation examined the 5-day administra-
`tion of oral 5-FU with 5-ethynyluracil with and without
`leucovorin
`to patients with metastatic colorectal carci-
`noma.107 A 33% response rate was noted in previously
`untreated patients treated with 20 mg/m 2 x 5 oral 5-FU, 50
`mg/d oral leucovorin, and 5-ethynyluracil (eniluracil).10 7
`Additional studies using a 28-day protracted 5-FU schedule
`with 776C85 DPD inhibition are currently underway to
`
`DeMARIO AND RATAIN
`
`clarify the optimal administration schedule for this promis-
`ing new modulatory agent.
`
`Other Oral Cytotoxics in Development
`
`JM-216. Bis-acetato-amine-dichloro-cyclohexyl-amine-
`platinum (IV) or JM216 represents a unique platinum-based
`agent with oral bioavailability. Phase I single-dose data
`suggest linear pharmacokinetics at doses less than 120
`mg/m 2; however, saturability was suggested at doses greater
`than 200 mg/m2.81 Significant interpatient variability (CV)
`to 63%) and AUC
`of the maximum concentration (21%
`(26% to 63%) was demonstrated in this single-dose phase I
`study. JM216 displays a notable biotransformation follow-
`ing oral administration, with six metabolite species. JM-118
`has been identified as the principal cytotoxic species. 83
`JM216 is unique among platinum derivatives for its schedule-
`dependency; in vivo models suggest a maximal therapeutic
`8 An MTD of
`index with 5-consecutive-day administration.
`120 to 140 mg/m 2 daily on a 5-day administration schedule
`has been determined from phase I testing. 82,109 Linear AUC
`pharmacokinetics were observed up to 140 mg/m2/d x 5.82
`Significant toxicities observed for JM216 to date include
`nausea/emesis, diarrhea, neutropenia, and thrombocytope-
`nia. 82,109 A significant correlation between the thrombocyto-
`penia and platinum ultrafiltrate AUC has been observed. 82
`Myelosuppression may be delayed relative to that observed
`for cisplatin or carboplatin.' 10 Importantly, JM216 has not
`demonstrated significant renal or neurotoxicity in preclinical
`or phase I/II trials to date, a significant toxicity advantage
`over cisplatin." 0-112 A recent phase II trial of single-agent
`JM216 for hormone-refractory prostate cancer noted a
`prostate-specific antigen (PSA) reduction in 42% of treated
`patients.110 Phase II 5-day dosing studies have also demon-
`strated responses in small-cell lung cancer.113 A phase II trial
`in previously untreated non-small-cell lung
`of JM-216
`cancer found no objective responses among 13 assessable
`patients treated on a 120 mg/m 2/d x 5 schedule.114
`Vinorelbine. Vinorelbine is