Pharmacokinetic optimisation of anticancer therapy.

PHARMACOKINETICS -THERA PE UT ICS

'----

C1in. Pharmacokinet. 21 (3): 213·231 , 1991 0312·5963/ 91/0009·0213/$09.50/ 0 © Adis International Limited. All rights reserved. CPK1054A

Pharmacokinetic Optimisation of Anticancer Therapy Jan Liliemark and Curt Peterson Department of Clinical Pharmacology, Karolinska Hospital, Stockholm, and Department of Medicine, Karolinska Institute at Huddinge Hospital, Huddinge, Sweden

Contents 213 214 214 214

215 216 216 216 217 217 218 220

225 225 225 225 225 226 226 226 227 227

227

Summary

Summary I. History of Chemotherapy 1.1 Cancer as a Cause of Morbidity and Mortality 1.2 Development of Cancer Chemotherapy 1.3 Dose-Response Relationship in Cancer Chemotherapy 2. Therapeutic Drug Monitoring 2.1 General Aspects 2.2 Special Problems in Cancer Chemotherapy 2.3 Interpretation of Plasma Concentration Data 2.4 Chronopharmacokinetics 2.5 Individualised Treatment to Avoid Adverse Effects 2.6 Correlations Between Pharmacokinetics and Therapeutic Results 3. Dose Adjustments for Reduced Liver or Kidney Function 3.1 Methotrexate 3.2 Cisplatin 3.3 Bleomycin 3.4 Anthracyclines 3.5 Etoposide 3.6 Cyclophosphamide 3.7 Ifosfamide 3.8 Vincristine 3.9 Cytarabine 4. Pharmacologically Directed Phase I Trials

It is obvious that there are great problems with pharmacokinetic individualisation of anticancer therapy. The strong relationship between dose intensity (total dose/unit time) and response revealed in clinical trials with some tumours provides a strong support for studies seeking relationships between the individual plasma pharmacokinetic profile and response to treatment. Unfortunately, studies that define a therapeutic window are sparse, and trials that prospectively test such models are even rarer. Thus, for most cancer drugs, it is not possible to give any definite advice on how to use pharmacokinetic determinations to establish individualised therapy, and there is therefore a definite need for such studies. It is important, however, that attempts to

CUn. Pharmacokinet. 21 (3) 1991

214

establish relationships between drug concentrations and therapeutic effects be founded on a sound theoretical base. When drugs, mainly antimetabolites, are extensively metabolised intracellularly and interact with intracellular processes about which there are data showing a strong interindividual heterogeneity, such data must be considered when designing pharmacokinetic investigations. Cytarabine and fluorouracil are good examples of this. The monitoring of intracellular drug! metabolite concentrations or of the direct biochemical events in the tumour cells seems to be a promising approach with such drugs. It also needs to be emphasised that pharmacokinetically guided individual1sation cannot be achieved before a therapeutic window is _established, i.e. a knowledge of the relationship between drug concentration and clinical effects. The investigators in this field accept a great responsibility when clinical studies are undertaken: a poorly designed study showing no benefit from pharmacokinetically guided individualisation can impair the possibilities of performing more adequate studies in the future.

1. History of Chemotherapy 1.1 Cancer as a Cause of Morbidity

and Mortality Cancer is the second most common cause of death in the Western world, and accounts for 22% of all deaths in the US (Fraumeni et al. 1989). If the tumour is localised, the patient can be cured by surgery or radiotherapy. If it is disseminated, only chemotherapy is curative but unfortunately such cures can be effected in only a few cases. The overall cancer cure rate is about 50%, but only 2% of patients are cured by chemotherapy alone (Fraumeni et al. 1989). In the most common types of solid tumours like cancer of the large bowel, breast and lung, the results of chemotherapy are still discouraging. Haematological malignancies and certain rare solid tumours like testicular teratomas, chorioncarcinomas and some childhood malignancies are the exceptions where modern chemotherapy has been most successful. In childhood acute lymphoblastic leukaemia, the overall cure rate is today well over 50%. 1.2 Development of Cancer Chemotherapy Since its introduction in the 1940s, cancer chemotherapy has improved with the discovery of new agents such as the anthracyclines in the 1960s and the platinum complexes in the 1970s. More importantly, antineoplastic drugs are now used in a more effective way. In the early days, patients were given small daily doses of 1 drug at a time.

After Skipper et al. (1964) showed that normal cells (e.g. in the bone marrow) recovered from druginduced lesions faster than tumour cells, intermittent treatment was introduced in the 1950s. Combination chemotherapy, where the MOPP (mechlorethamine, 'Oncovin', procarbazine, prednisone) treatment of Hodgkin's 'disease (DeVita et al. 1970) remains standard, was introduced in the 1960s. The principal idea was to combine drugs with different mechanisms of action and different dose-limiting toxicities. A prerequisite for the introduction of more effective cancer chemotherapy regimens (accompanied by more severe toxic effects) was the parallel development of improved supportive care techniques such as antimicrobial therapy, supply of blood products, etc. The use of autologous bone marrow rescue and haematological growth factors is of great future potential for eliminating bone marrow toxicity as the major doselimiting factor in cancer chemotherapy. Thus, it might be possible to cure tumours in patients who are not cured by current regimens if a definite relationship has been revealed between dose intensity (see section 1.3) and response rate (DeVita 1986; Hryniuk 1988). The selection of drugs to treat a specific form of cancer is based on the results of previous clinical trials. Enormous amounts of money are spent on the development of new and it is hoped better anticancer agents. To improve the treatment results, we believe that more important than incorporating new drugs in standardised clinical trials is learning how to design therapy against the individual tum-

Optimisation of Anticancer Therapy

our for the individual patient. This involves 2 steps. The first is to select the best drug(s) for the individual patient's tumour and the second is to elaborate and optimise treatment conditions (dose interval, route of administration, infusion time, etc.). A number of experimental systems have been extensively studied for the purpose of predicting the drug sensitivity of the individual patient's tumour. Human tumour fragments can be grown in immunologically incompetent mice or for a few days under the renal capsule of healthy mice (Griffin et al. 1983). Another approach is the human tumour stem cell assay, where single-cell suspensions ofthe tumour are first incubated with a number of drugs in a range of concentrations and then cultured_ in a semisolid medium (Hamburger & Salmon 1977). Dye exclusion assays have also been employed, and have the advantage of being more rapid than the stem cell or subrenal capsule assays (Weisenthal & Lippman 1985). Although a dye exclusion assay has been used successfully (Tidefelt et al. 1989), no system is ready for routine clinical use at present. However, with technical improvements, such systems may in the future become an important aid in selecting antineoplastic therapy for the individual patient. 1.3 Dose-Response Relationship in Cancer Chemotherapy Since most anticancer agents interact with the cellular replication of DNA, growing cells are generally more sensitive to cytotoxic effects than resting cells. This is one possible explanation for the paradoxical finding that some malignancies, such as acute leukaemias which rapidly kill the patients if left untreated, are now curable. One exception is 2-chloro-2'-deoxy-adenosine, which in vitro is equally toxic to resting and dividing cells (Carson et al. 1983). In contrast, modern chemotherapy cannot cure patients with breast cancer, a tumour which grows slowly and where some patients may live many years with advanced disease. A high growth rate may be a prerequisite but it is not a guarantee of curability - exemplified by anaplastic

215

lung cancer which, despite a high growth rate, is curable in few cases. For certain anticancer drugs, a simple doseresponse relationship can be shown in cultured malignant and normal cells. However, for antimetabolites the exposure time is an even more important determinant of biological activity. Antipyrimidines like cytarabine exert their anti tumour effect by incorporation into DNA leading to termination of chain elongation, or by interfering with the incorporation of endogenous nucleotides into DNA. They are therefore theoretically strict S-phase specific agents. Even in very aggressive malignancies such as acute leukaemias only a small fraction of the cells are synthesising DNA. Therefore, the time of exposure is of crucial importance for the therapeutic effect of cytarabine (Skipper et al. 1964). However, scheduling is also of importance for drugs not considered to be phase-specific, such as topoisomerase II-inhibitors. Thus, at identical intracellular areas under the drug concentration-time curve (AVC), daunorubicin is much more cytotoxic after a short exposure to a high concentration than after a long exposure to a low concentration (Andersson et al. 1982). In contrast, in patients with lung cancer, etoposide as a single drug has a much better therapeutic effect after 5 daily bolus administrations compared with administration of the same total dose as a 24h continuous infusion, despite the plasma AVC being identical during the 2 modes of administration (Slevin et al. 1989). This shows that even for drugs believed to act on the same target (topoisomerase II) there is a complex relation between cytotoxicity and concentration/ exposure time (C x t). Most patients with cancer are given combination chemotherapy, which further complicates the evaluation of dose/response relationships. Recently the term 'dose intensity' was introduced (Hryniuk 1988; Hryniuk & Bush 1984). It is defined as the dose given per unit of time (mg/m 2 per week) in comparison with a reference treatment protocol, and assumes that all drugs in a combination contribute equally to the antitumour effect. The dose intensity has been calculated retrospectively in a great number of clinical trials, and

216

the results strongly indicate a good correlation between it and clinical effect. This has led the former director of the National Cancer Institute (USA) to state: 'The most toxic manoeuvre a physician can make when administering chemotherapy is to arbitrarily and unnecessarily reduce the dose' (DeVita 1986). Theoretically, there is an optimal dose intensity in a defined patient population. When this is exceeded, the improved antineoplastic effect will be overshadowed by the increased toxicity. Most cancer chemotherapy regimens used today have been introduced on the basis of results of comparative clinical trials. The drugs are administered in standardised dosages according to bodyweight or (more often) body surface area. The variability in plasma concentrations achieved in a patient population after the administration of a standardised dose is large (Grochow et al. 1990), while the therapeutic window for the individual patient can be narrow. It is therefore probable that a number of patients who would have been cured if they had been treated with a lower dosage have died due to toxicity when treated with a standard dosage. In the same way, other patients who would have tolerated a higher than standard dosage will succumb due to disease progression when treated according to a standard protocol. Thus, it is urgent to develop instruments to individualise drug dosages in cancer chemotherapy. The good correlations found between dose intensity and response in clinical trials suggest that relationships between drug concentration and response in the individual patient can also be found. However, there may also be other important factors besides dosage or drug concentration which play an important role in response but are randomly distributed among patients and therefore less amenable to analyses within clinical trials. Such factors (cellular metabolism, cell growth kinetics, drug resistance, etc.) may severely obscure any individual relationship between drug concentration and response and hamper the possibility of individualising the treatment based on pharmacokinetic determinations. We therefore emphasise that, before monitoring of drug concentrations can be used for pharmacokinetically guided individualis-

Clin. Pharmacokinet. 21 (3) 1991

ation, a good correlation must be shown between pharmacokinetic determinants and response In studies with large numbers of patients.

2. Therapeutic Drug Monitoring 2.1 General Aspects For a number of drugs, i.e. anticonvulsants, digoxin or tricyclic antidepressants, monitoring plasma concentrations for the purpose of individualising drug therapy has been the clinical routine for many years (Sj6qvist et al. 1980). These drugs are generally administered repeatedly and a steady-state situation is achieved where a direct relationship between plasma and tissue drug concentrations can be assumed. Therefore, it is not surprising that the therapeutic and/or toxic effects of these drugs can be related to the plasma concentration and therapeutic windows established. In contrast, due to the interindividual variability in absorption, distribution, metabolism and excretion, there is a poor relationship between the dosage administered and the therapeutic response. 2.2 Special Problems in Cancer Chemotherapy There are a number of problems with therapeutic drug monitoring in cancer chemotherapy. 1. The drugs are generally administered intermittently (e.g. as single injections every 3 weeks) and a steady-state situation is not reached. 2. Many drugs are prodrugs and need to be metabolised before becoming cytotoxic. Most antimetabolites are active only after bioactivation in the target cells, and the active metabolites cannot be detected in plasma. 3. The mechanism of action is often complex or unknown. For drugs exerting irreversible actions, a simple relationship between plasma drug concentration and therapeutic effect cannot be expected. 4. The therapeutic results, in terms of cure rate, can be evaluated only after several years of followup. Establishing a therapeutic interval can take a


very long time (during which the therapeutic tradition might change considerably). 5. The use of multidrug combinations complicates the establishment of therapeutic intervals for a single drug. Thus, there are numerous problems to be coped with when pharmacokinetic monitoring is to be used for individualisation of cancer chemotherapy. The need, however, is obvious. 2.3 Interpretation of Plasma Concentration Data All the factors mentioned above make the prediction of therapeutic and toxic effects from the plasma concentration data difficult in cancer chemotherapy. For example, assume we administer equal doses of 2 closely related drugs, A and B, with identical pharmacodynamics (i.e. identical cytotoxic effects of each molecule at the effector site), but with different pharmacokinetics. At a certain time after administration, we determine the plasma concentration. If we find the plasma concentration of A far exceeds that of B, what should we expect about the cytotoxic effects, assuming that there are no active metabolites? We might think that A has a more pronounced effect than B, because of a slower elimination rate. Alternatively, B might have greater effect than A if its lower plasma concentration is due to a higher tissue (and tumour) uptake. This example illustrates the importance of differences in drug distribution which are not revealed by a single plasma concentration determination unless at steady-state. Recently, the effect of interferon on the pharmacokinetics of melphalan was studied by Ehrsson et al. (1990). It was found that the administration of interferon 5h before melphalan reduces the plasma concentrations of the latter. The authors' interpretation of these results is that the hyperthermia induced by interferon increases the alkylating activity and tissue affinity of melphalan. One way to avoid the problems of interpretation of plasma pharmacokinetic data is, whenever possible, to study the drug concentration not only in plasma but also in tumour cells. This can readily

217

be done in leukaemia patients, since isolation of leukaemic cells from peripheral blood is feasible. Of course, this approach may cause other problems, e.g. how to efficiently extract the drug and metabolites from the cells or tissue, determining the proportion of tumour and normal cells in the specimen to be assayed, etc. With this approach unexpected results have been found. When the pharmacokinetics of equal doses of daunorubicin were compared after infusion for 10 min and 24h in the same patients with acute leukaemia in different treatment courses, it was found that the drug concentration in the leukaemic cells was much higher after prolonged infusion despite similar plasma AVC values (Paul et al. 1989). Plasma pharmacokinetic studies on cytarabine have shown thal -the peak concentration is higher after bolus intravenous injection than with subcutaneous injection, while the AVCs for the 2 modes of administration are identical due to the slow release of the drug from the subcutaneous depot (Slevin et al. 1981). In contrast, the intracellular metabolite cytosine arabinoside 5' -triphosphate (ara-CTP) accumulates to a higher concentration after subcutaneous administration and its AVC is twice as large as that after an identical intravenous dose (Liliemark et al. 1985b) [fig. 1]. This paradoxical relationship between plasma and cellular pharmacokinetics is due to the saturable bioactivation ofcytarabine. Thus, a plasma concentration of that drug above 10 ILmol/L will not lead to an increased formation of the active metabolite in leukaemic cells. It has also been claimed that the short retention of cytarabine in plasma after subcutaneous injection shows that this S-phase specific drug should be administered as a continuous infusion (Spriggs et al. 1985). Studies on the cellular pharmacokinetics, however, suggest that the retention of the active metabolite in the tumour cells is long enough to support the use of subcutaneous injection (Liliemark et al. 1987). 2.4 Chronopharmacokinetics The concept of chronopharmacokinetics, i.e. the diurnal variation of drug concentration, also highlights the problem of how to interpret plasma phar-


218

decrease during daytime has been linked to the increase in activity of the catabolic enzyme dihydropyrimidine dehydrogenase (Harris et al. 1990).

1000

c

2.5 Individualised Treatment to Avoid Adverse Effects

~a.

B

Basically, there are 2 aims for the individualisation of cancer chemotherapy. One is to increase the therapeutic effects (section 2.6) and the other is to minimise the adverse effects of the treatment. If the latter is the desired end-point, only the upper limit of the therapeutic window needs to be defined. For this purpose it is an advantage that the side effects are generally seen shortly (within days to weeks) after, drug administration. On the other hand, depending on the frequency of adverse effects, a study aiming to show a correlation between drug concentration and adverse effects could require large numbers of patients.

Cl

E

:aE .e, c:

! c: ~ c:

100

0

11.

I-

~ 01

5

10

Time (h) Fig. 1. Concentration of the active metabolite ara-CTP in leukaemic cells from I patient with acute myelogenous leukaemia treated with cytarabine 50 mg/m 2 administered intravenously (0) or subcutaneously (e) [from Li1iemark et al. 1985a, with permission].

macokinetic data. It has been shown that the plasma concentration of mercaptopurine is higher after evening administration than in the morning (Langevin et al. 1987). The biochemical mechanism behind this difference has not been elucidated; it may be that the first-pass effect or catabolism of mercaptopurine is lower at night, and/or that the drug uptake and metabolism to active products intracellularly may also be lower. The earlier retrospective finding that children receiving the drug at night had a lower risk of relapse supports the former explanation (Rivard et al. 1985). The diurnal variation of the concentration of fluorouracil has been known for some years (Kawai et aI". 1976). During continuous infusion there is a 5-fold variation of plasma concentrations; the

2.5.1 Methotrexate Methotrexate is an antifolate whose cytotoxic effect is mediated mainly by inhibition of the enzyme dihydrofolate reductase (DHFR). This enzyme reduces dihydrofolate to tetrahydrofolate, which is an important substrate in a number of pathways in the metabolism of mammalian cells. Intracellularly, methotrexate is metabolised to form a polyglutamate by the addition of 1 or more glutamate residues. The methotrexate polyglutamate retains its cytotoxic properties and is also retained intracellularly while the monoglutamate (the parent compound) will pass through the cell membrane when the extracellular drug concentrations are reduced. The polyglutamates of methotrexate have more inhibitory effect on DHFR than the parent drug (Matherly et al. 1983). The dihydrofolate accumulated during methotrexate inhibition of DHFR is also polyglutamated; the polyglutamates of these dihydrofolates, as well as those of the drug, inhibit thymidylate synthetase and enzymes in the purine de novo synthesis such as aminoimidazole carboxamide ribonucleotide (AI CAR) transformylase and glycineamide ribonucleotide (GAR) transformylase (Allegra et al. 1985; Baram

219


protracted treatment with folinic acid, and their plasma methotrexate concentrations should be followed until they drop to < 0.1 ~mo1/L. An earlier mortality rate of 6% with high dose methotrexate was eliminated when concentration monitoring was introduced to guide the administration of folinic acid (Evans et al. 1986). On the basis of these results, the use of plasma drug monitoring is today routine in most centres where high dose methotrexate is used. It has also been shown that pharmacokinetic monitoring after a small test dose of methotrexate can identify patients at risk of toxic effects from a high dose regimen (Kerr et al. 1983).

10- 3

:J

'5

.sc

10- 4

0

'~

!o

c

~

c

0

10- 6

Q)

«i )(

~

~

10-7

E «l

E 10- 8

'"«l

ii: 10 -9~----~~~--------~-------

1

2 3 4

5 6 7

8 9 10 11 12 13

nme (days) Fig. 2. Disappearance of drug from plasma in patients receiving 6h intravenous infusions of methotrexate 50 to 250 mg/ kg. The shaded area shows the disappearance profile in 14 patients who had no toxicity, monitored serially over 72h. Plasma concentrations are shown in (0) patients with evidence of myelosuppression and (e) 6 patients without toxicity with 48h values above 0.9 I'mol/L (from Stroller et al. 1977, with permission).

et al. 1988). Thus, the ability of a cell to form the polyglutamate is very important for the cytotoxic effect of methotrexate. To achieve a cytotoxic concentration in the central nervous system and to overcome moderate drug resistance, this agent is often administered in much higher than conventional doses. Cells treated with methotrexate can be rescued from the cytotoxic effect by treatment with 5-formyl-tetrahydrofolate (folinic acid), which not only restores the depleted tetrahydrofolate pools but also reverses the inhibition ofthymidylate synthetase, AICAR and GAR transformylase by competition with dihydrofolate polyglutamates. Monitoring of plasma methotrexate concentrations can identify individuals who are at risk for life-threatening toxicity after high dose treatment (Stroller et al. 1977) [fig. 2]. Patients with a plasma methotrexate concentration exceeding 0.9 ~mo1/L 48h after a 6h infusion have a very high risk of toxic complications. These patients should receive

2.5.2 Carhop/atin One of the most effective anticancer agents is cisplatin. Treatment with this drug, however, is accompanied by many adverse effects such as nephrotoxicity, ototoxicity and severe nausea and vomiting. Therefore, a number of analogues have been synthesised and tested. Recently introduced was carboplatin, which reduced these adverse effects. In contrast, the dose-limiting toxicity of carboplatin is bone marrow toxicity, in particular thrombocytopenia. Egorin et al. (1986) found that the plasma concentration of non-protein bound carboplatin correlated well with the degree of thrombocytopenia. The main route of elimination of carboplatin is renal excretion; thus, they could also show a close correlation between the renal function and the plasma clearance of the drug. A model was therefore derived for patients with normal liver (bilirubin ~ 20 mg/L) and bone marrow function (WBC ~ 3500/~1, platelets ~ 100 OOO/~l), and a Karnofsky performance status ~ 40%, by which the degree of thrombocytopenia can be predicted taking into consideration the dose, creatinine clearance (CLCR) [measured over 24h twice within a week before treatment] and body surface area (BSA) [Egorin et al. 1985]:

D = {0.091(CLCR/BSA)([P p -P n]/Pp) x 100}

+ 86

where D is the dosage in mg/m 2, Pp is the pretreatment platelet count and Pn is the desired platelet nadir. The platelet nadir can be predicted for patients

220

with both normal and impaired renal function (Egorin et al. 1984) and for patients who have previously received extensive myelotoxic treatment (69 being substituted for 86 in the model). The model is also useful when carboplatin is used in combination with etoposide (Belani et al. 1989). This approach is very promising and, with appropriate basic knowledge about the relationship between dose, pharmacokinetics, renal and liver function and toxic effects, the concept will probably be useful for a number of drugs. The strategy for such studies has been discussed by Egorin (1990). 2.5.3 Vinblastine It has been shown that myelosuppression after

vinblastine can be predicted not by the dose but by plasma drug monitoring. It is therefore recommended that plasma concentrations above 1.5 to 2.0 J.Lg/L are avoided during prolonged (12-week) intravenous infusions (Retain et al. 1987). 2.5.4 Other Drugs A number of drugs like menagaril and hexamethylene bisacetamide show a good correlation between plasma AVC or concentration at steadystate (CSS) and toxicity but only weak correlation between dose, renal and hepatic function and plasma pharmacokinetics (Egorin et al. 1986, 1987). In these cases, individual pharmacokinetic monitoring can be used to avoid toxicity. It must be emphasised, however, that the relationship between pharmacokinetics and pharmacodynamics may be highly dependent on the group of patients studied (Retain et al. 1989). The degree of toxicity found in 1 group of patients may be quite different in another group of younger patients with less advanced disease and a higher performance status at the same systemic exposure to the drug (AVC).

2.6 Correlations Between Pharmacokinetics and Therapeutic Results Attempts to correlate drug concentrations to anticancer effects have not been very successful to date. The therapeutic end-points of cure or remis-


sion can be evaluated only after a considerable period of time, in contrast to acute toxicity. This is a major obstacle in the attempts to establish a therapeutic interval for an anticancer drug. For statistical reasons, the number of failures and cures (or remissions) must be large enough to allow the determination of the therapeutic interval. The pharmacological failures also need to be clearly distinguished from other failure reasons, e.g. supportive care failure. Nevertheless, there are examples of studies where therapeutic intervals have been defined for anticancer drugs and used prospectively to individualise therapy. 2.6.1 Cytarabine Cytarabine is the main drug used to treat acute myelogenous leukaemia. It must be phosphorylated to the triphosphate, ara-CTP, in the target cells to exert cytotoxicity. In patients with refractory acute leukaemia treated with high dose cytarabine 3 g/m 2 alone every 12h for 6 to 12 doses, a marked variability of ara-CTP concentrations in the leukaemic cells was found (Liliemark et al. 1985a). Furthermore, there was no apparent relationship between the plasma concentration of the parent drug and the intracellular concentration of the active metabolite. It was shown that the likelihood of entering complete remission was very low in patients refractory to or relapsed after standard cytarabine treatment with a figure of less than 75 J.LmoljL as the trough concentration of ara-CTP in their leukaemic cells (Plunkett et al. 1985) [fig. 3]. This finding was in line with earlier in vitro data showing a correlation between the retention of araCTP in leukaemic cells after incubation with cytarabine and the duration of complete remission (Rustum & Preisler 1979). In a subsequent study it was shown that when the intracellular concentration of ara-CTP was elevated above this critical level, the remission rate in this poor prognosis group of patients could be raised from 3 of23 (13%) to 4 of 8 (50%) [Plunkett et al. 1985]. The intracellular concentration of ara-CTP in leukaemic cells during continuous infusion is directly proportional to the dose delivered in the individual patient (Heinemann et al. 1989) and the desired target concen-


•

700

::J

'is E .=.

300

c 0

ec 8c 8

•

1•

500

200

a.

t-

•

• • •

• •

•

01 OJ

e t-

100

••

------

o

•

•

Complete remission

.. • ••

~., .r:.

221

-

•••

•

__ 1__ - - ---•

•

t

•

Resistant disease

Supportive care failure

Fig. 3. Relationship between concentration of ara-CTP in circulating leukaemic blasts 12h after infusion of cytarabine (3 g(m 2, 6 to 12 x 2h infusion every 12h) and response. The broken rule at 75 Itmol(L indicates the lowest concentration discriminating significantly between complete remission and treatment failure (resistant disease plus supportive care failure) [from Plunkett et al. 1985, with permission).

tration (75 I-Imol/L) can thus be achieved by adjusting the dosage according to the ara-CTP concentration obtained with a test dose (administered by continuous infusion). One limitation to this approach to individualising treatment of acute myelogenous leukaemia is the requirement of a leukaemic cell count> 5000/1-11 in the peripheral blood. It must be emphasised, however, that there is no direct proportionality between the dose of cytarabine and the cellular concentration of ara-CTP at plasma concentrations above 10 I-Imol/L, which corresponds to a dose rate of about 250 mg/m 2 per hour (Plunkett et al. 1987). However, when similar studies were undertaken in previously untreated patients, no correlation was found between cellular ara-CTP pharmacokinetics

and duration of response (Estey et al. 1990). The mechanism behind this difference between refractory and previously untreated patients is not clear. The pharmacological basis for cytarabine treatment has recently been carefully reviewed by Heinemann and Jehn (1990). 2.6.2 Methotrexate In the mid 1980s, a relationship was reported between the plasma pharmacokinetics of methotrexate and the outcome of maintenance treatment of children with acute lymphoblastic leukaemia (Borsi et al. 1987; Evans et al. 1984). However, when these results were followed up it was apparent that a low clearance of methotrexate did not prevent relapses but only delayed them (Evans et al. 1989). Recently, data have been presented suggesting that the ability of leukaemic cells (taken from newly diagnosed patients) to form methotrexate polyglutamates in vitro correlates with the outcome of maintenance treatment (Whitehead et al. 1989). This study focuses on the difficulty in finding a correlation between drug or metabolite concentrations and therapeutic outcome with drugs that need bioactivation in the target cells and which are administered when there are no target cells available (during maintenance treatment). Due to these early and somewhat conflicting results, it is impossible to give advice on individualised administration of methotrexate, in order to increase the therapeutic effect. 2.6.3 Mercaptopurine The maintenance treatment of acute lymphoblastic leukaemia with mercaptopurine differs from most other antineoplastic regimens in that the drug is orally administered daily for 2 to 3 years. Thus, steady-state is achieved and the plasma concentration may therefore reflect the concentration of parent drug and metabolites in other compartments as well as in bone marrow cells. Early studies on the urinary excretion of radiolabelled drug showed that about 50% of the ingested drug could be recovered. Later studies using specific high performance liquid chromatography (HPLC) showed, however, that the bioavailability of mercaptopurine is

222


low and highly variable between individuals (Lafolie et al. 1986; Lbnnerholm et al. 1986; Zimm et al. 1983). Recent studies also show that there is a large intraindividual variation, making repeated determinations necessary to characterise the pharmacokinetics in an individual (Lafolie et al. 1991). Two independent investigators have found a good correlation between the plasma AVe of mercaptopurine during maintenance treatment and the relapse risk in children with acute lymphoblastic leukaemia (Hayder et al. 1989; Koren et al. 1990) [fig. 4]. In the former study, each child was monitored between 2 and 12 times during the maintenance therapy. The lower ends of the therapeutic interval in these 2 studies are in the same range ("" 300 /-Lg/L· h). It has also been shown by several investigators that the relapse risk of this disease is lower in patients treated until neutropenia occurs than in patients treated with a standard dosage (Hayder 1989; Schmiegelow et al. 1988). Mercaptopurine is a prod rug and intracellular bioactivation (fig. 5) is required for its clinical ef-

DNA

RNA

i

1

TdGTP

TGTP

i

TdGOP

'"/

/

TGDP

Fig. 5. Cellular metabolism of mercaptopurine. Abbreviations: PRPP = phosphoribosylpyrophosphate; MPR = mercaptopurineribose; Me-MPR = methylmercaptopurineribose; TIMP = thioinosine 5'-monophosphate; TXMP = thioxantine 5'monophosphate; TGMP, TGDP and TGTP = thioguanosine 5'-mono, di- and triphosphate; TdGDP and TdGTP = thiodeoxyguanosine 5'-di and triphosphate; ATP = adenosine triphosphate; TG = thioguanine.

700I -

•

600I -

I

500,-

400I 300

-

200 100

-

••" Relapse

.•

• • I

I

••• Event free

..

TOXICity

Fig. 4. Relationship between mean plasma area under the concentration-time curve from time zero to 4h (AUC(O_4) of mercaptopurine. and outcome of maintenance therapy in 22 children with acute lymphoblastic leukaemia (from Hayder et al. 1989, with permission).

fects. Therefore, it would be a great advantage if the pharmacokinetics of these intracellular metabolites in leukaemic cells could be studied during treatment. However, when patients with acute lymphoblastic leukaemia in complete remission are treated with methotrexate and mercaptopurine, by definition, no leukaemic cells can be obtained for pharmacokinetic studies. Lennard and co-workers therefore made the assumption that the intracellular concentration of mercaptopurine metabolites in erythrocytes would bear some relationship to that in leukaemic and normal progenitor cells. In agreement with earlier in vitro results (Tidd et al. 1972) a relationship was found between the concentration of thioguanosine nucleotides (TGN) in erythrocytes and the degree of neutropenia 2 weeks later (Lennard et al. 1983). These authors also showed that patients with low concentrations of TGN in their erythrocytes during maintenance treatment with mercaptopurine run a high relapse risk (Len-


223

nard & Lilleyman 1989; Lennard et al. 1990) [fig. 6]. Thus, the determination of TGN in erythrocytes has the potential to become an important instrument for individualisation of maintenance treatment with mercaptopurine in acute lymphoblastic leukaemia. In summary, prospective studies investigating the benefit of pharmacokinetically guided mercaptopurine therapy are highly warranted. Until such studies have been done, no definite recommendation on how to individualise treatment can be given, i.e. it is still an open question whether plasma pharmacokinetics, erythrocyte TGN concentrations or the degree of neutropenia best reflect the outcome in children with acute lymphoblastic leukaemia. However, if monitoring of plasma concentrations is used, it is definitely advisable to repeat this analysis several times in each patient because of the great intraindividual variability of mercaptopurine pharmacokinetics shown to occur (Lafolie et al. 1991). 2.6.4 Fluorouracil Fluorouracil is a fluorinated pyrimidine antimetabolite. Its mechanism of action is complicated and under debate, but 3 mechanisms have been

29

100

9

~

I ••••••• :

_._......-........._....._....._-B

~1

>

.~

13

:::>

-;{!.

e....

1ii >

.~

:::>

III

~

100

A

\.., ~28

1ii

Q)

proposed (fig. 7): first, after phosphodeoxyribosylation, the 5' monophosphate of the deoxyribonucleoside, FdUMP (see fig. 7 for abbreviations) inhibits the enzyme thymidylate synthetase (TS). TS methylates dUMP to dTMP which is an important step in the de novo synthesis of deoxythymidylate, triphosphate (dTTP). Methyl-tetra-hydrofolate (MeFH4) is the methyl donor and together with TS and dUMP or FdUMP it forms a ternary complex. The degree of the inhibition of TS by fluorouracil and the stability of the complex depend on the amount of enzyme and the cellular concentrations of MeFH4, dUMP and FdUMP. Thus, there are a number of intracellular events which need to be considered when studies are designed to correlate drug! metabolite pharmacokinetics and response to fluorouracil. The cytotoxic effect can also tre mediated secondly through the incorporation of 5-FUTP into the RNA, or thirdly by the incorporation of FdUTP into DNA. There is no definite proof that anyone of these mechanisms is more important than the others. It is possible that the activity of key enzymes in the different pathways determines the different mech-

III

50

Q)

&;

a.

x2 = 11.9

as II:

p

III

w

< 0.001

p

< 0.01

O+-~~~~~~r-~~~~

0 0

50

o

100

50

100

Time (months)

Fig. 6. Actuarial relapse-free and event-free survival from the start of mercaptopurine therapy for 120 children in remission when divided at the median RBC thioguanosine nucleotides concentrations (175 pmo\f8 x 108 RBC) [A: n = 60 above median; B: n = 60 below median). Numbers on the curves indicate the proportion remaining at risk at the time of analysis (from Lennard et al. 1990, with permission).

224


anisms of action. These may vary between tumours, and between tumours and normal tissue. There are a number of studies showing a relationship between plasma fluorouracil concentrations and toxicity (Milano et al. 1988; Thyss et al. 1986). This relationship appears to be stronger than that between dosage and toxicity (Milano et al. 1988). There are also studies suggesting a relationship between the plasma clearance or AUC offluorouracil and antitumour response. These studies, however, contain only a small number of patients and the results have not been confirmed in larger series (Hillcoat et al. 1978; Seitz et al. 1983). It has also been shown that there is no relationship between the plasma pharmacokinetics of fluorouracil and the cellular concentration of its metabolites (Finan et al. 1987). Considering the complicated cellular events determining the pharmacodynamics of fluorouracil, from a theoretical point of view, it is apparent that measurement of its plasma concentrations can have only a limited applicability for individualisation of treatment. Fluorouracil is used in the treatment of solid

FUrd

3

~

tumours. The studies of cellular pharmacokinetics and pharmacodynamics meet considerable practical difficulties as surgical biopsies need to be taken before and after administration of the drug (Spears et al. 1984). Recently, however, methods for determination ofTS activity and FdUMP and dUMP concentrations from tissue samples have improved since fine needle aspiration techniques have come into use (Spears et al. 1989). It has been shown that the degree of TS inhibition after fluorouracil treatment correlates well with response (Spears et al. 1990). The folate concentration in tumour tissue has been shown to be important for the inhibitory effect of the drug on TS (Spears et al. 1989). This data will be useful in the development of methods to improve and individualise the treatment of solid tumours with fluorouracil. However, it is as yet too early to provide detailed information on using these determinations for individualisation of treatment. The intracellular concentration of fluorouracil can also be determined without invasive techniques by nuclear magnetic resonance spectroscopy (Wolf et al. 1990). This technique currently lacks the appropriate sensitivity for determination

FUDP

~

FUMP

~

-------+

RNA

6

5-FU

~

FUTP

~

t FdUrd

~

~

FdUMP

FdUDP

~

FdUTP ---------.. DNA

I I I

I I

I I

I I

I

I I I I

dUMP

te 5

•

dTMP

Fig. 7. Cellular metabolism of fluorouracil. Abbreviations: 5-FU = fluorouracil; FUrd = 5-fluorouridine; FdUrd = 5-fluorodeoxyuridine; FUMP, FUDP and FUTP = 5-fluorouridine 5'-mono-, di and triphosphate; FdUMP, FdUDP and FdUTP = 5-fluorodeoxyuridine 5'-mono-, di- and triphosphate; dUMP = deoxyuridylate, dTMP = deoxythymidylate. Key enzymes: I uridine phosphorylase; 2 orotate phosphoribosyltransferase; 3 uridine kinase; 4 thymidine phosphorylase; 5 thymidylate synthetase; 6 ribonucleoside diphosphate reductase.


of intracellular metabolites but is of great future potential for intracellular pharmacokinetic studies in solid tumours.

225

the plasma concentration of methotrexate itself be used to identify patients at high risk. 3.2 Cisplatin

2.6.5 Teniposide It has been reported that the css of teniposide

during continuous infusion correlates to 'oncolytic response' in patients with relapsed childhood malignances, especially acute leukaemias (Rodman et al. 1987). This information has been used for pharmacokinetic individualisation of therapy with this agent in maintenance treatment of children with acute lymphoblastic leukaemia in remission (Evans et al. 1990).

3. Dosage Adjustments for Reduced Liver or Kidney Function It is a well known clinical experience that patients with impaired liver or kidney function may react to drugs in unexpected ways. In most cases this is due to a reduced elimination rate, but other pharrnacokinetic parameters such as protein binding may also be affected. Precautions should be taken when giving cancer chemotherapy to patients with impaired renal or liver function but we still lack reliable information allowing generalised recommendations on dosage reductions for most drugs.

3.1 Methotrexate Most methotrexate is excreted in the urine unchanged or as 7-hydroxy-methotrexate. It has been recommended that, if renal function is impaired, the dose of this drug should be reduced in proportion to the decrease in CLcR. Methotrexate may also cause nephrotoxicity, probably by precipitation of 7-hydroxy-methotrexate which has poor water solubility at an acid or neutral pH. The rate of elimination increases considerably if the urine is alkalinised. A postinfusion increase in serum creatinine may not be a reliable predictor of toxicity, since it can take several days for serum creatinine to adjust to changes in renal function. As described above, it is therefore recommended that

Cisplatin binds extensively to plasma proteins; its optimal use is limited by nephrotoxicity. Elimination of free platinum species by the kidneys has been reported to exceed the glomerular filtration rate, suggesting active secretion (Jacobs et al. 1980). The risk of nephrotoxicity is enhanced if other nephrotoxic drugs like aminoglycosides or methotrexate are administered simultaneously. Because of the increase in toxicity associated with a decreased elimination of free platinum species, the dosage of cisplatin should be reduced in patients with impaired renal function in proportion to the reduction in CLcR (Chabner 1990). The risk of nephrotoxicity is less with the novel cisplatin analogue carboplatin (see also section 2.5.2). There is an increased risk of ototoxicity with cumulative dose and there also appears to be a correlation between the ototoxic effects of cisplatin and the plasma concentration: patients with a cisplatin concentration of > 1 mgfL during infusion have an increased risk of suffering impaired hearing (Laurell & Jungnelius 1990). 3.3 Bleomycin Bleomycin is a mixture of glucopeptides with high affinity for squamous epithelium cells. About 70% of administered bleomycin is excreted unchanged in the urine and there is a strong association between renal failure and pulmonary toxicity after treatment with this agent (Dalgleish et al. 1984). Therefore, a reduction to 50% ofthe normal dose is recommended in cases of renal failure (CLcR < 25 ml/min/m2). 3.4 Anthracyclines Liver metabolism plays an important role in the elimination of anthracyclines (doxorubicin, epirubicin, daunorubicin). In contrast to doxorubicin and daunorubicin, epirubicin and its reduced me-

226


tabolite epirubicinol are conjugated to glucuronic acid. There are recommendations of a dose reduction of 25 to 50% when there is reduced liver function (Benjamin 1975), but these are not supported by convincing pharmacokinetic data. A number of studies have shown that there is also nonhepatic metabolism of anthracyclines, e.g. in blood cells. 3.5 Etoposide About 50% of the administered dose of etoposide is excreted unchanged or as metabolites in the urine (D'Incalci et al. 1986; Sinkule et al. 1984) and there is a direct relationship between its clearance and CLcR. A rough guideline for modifying the etoposide dosage in patients with renal dysfunction is therefore to reduce the dose (D) in proportion to the CLcR down to 50% of the standard dosage: D = Standard dose/2

X

(CLcR/normal CLcR + 1)

However, etoposide is highly bound to plasma proteins, mainly albumin (,.,; 95%). A subnormal serum albumin will therefore increase its unbound fraction and clerance (Sinkule et al. 1984). Similarly, as bilirubin is also bound to albumin, an increased serum bilirubin concentration can displace etoposide and increase the unbound fraction. This means that the total clearance of etoposide will increase (Clark et al. 1988) or remain unchanged (Hande et al. 1990). As an increased bilirubin is often associated with impairment of the liver function, the clearance of the pharmacologically active unbound fraction will decrease and the AVC of unbound etoposide will increase (Stewart et al. 1990). An impaired liver function as such, without increased bilirubin, will decrease the clearance of total as well as unbound drug. Thus, the impact of altered liver function on the pharmacokinetics of etoposide is very complex (Hande 1990; Retain 1990). It could therefore be argued that routine monitoring of plasma concentrations of patients with liver function impairment will be useful. It remains to be shown, however, that these pharmacokinetic alterations due to organ dysfunction have an impact on the toxic effects of etoposide.

3.6 Cyclophosphamide Cyclophosphamide is a prodrug which requires microsomal hydroxylation (mainly iIi the liver) to form the active metabolite 4-hydroxy-cyclophosphamide, which is further converted in the target tissue to phosphoramide mustard and acrolein via aldophosphamide. These microsomal enzymes can be induced by a number of substances, e.g. drugs such as phenobarbital or alcohol, and also by cyclophosphamide itself. The half-life of cyclophosphamide therefore becomes progressively shorter with repeated administration and patients receiving phenobarbital also have a shorter cyclophosphamide half-life (D'Incalci et al. 1979). The peak of the alkylating activity, on the other hand, increases if patients are given drugs inducing hydroxylation before cyclophosphamide is administered, but the elimination of the alkylating activity is faster (Bagley et al. 1973). This means that both the bioactivation and elimination of cyclophosphamide depend on the liver function. In patients with severe liver failure, a reduced clearance was consequently seen (Juma 1984). It was found, however, that there were fewer adverse effects in these patients than in those with normal liver function, suggesting that the amount of unchanged cyclophosphamide excreted in the urine is increased in patients with liver failure. A dosage reduction in patients with liver failure can therefore not be recommended at present. The effect of renal failure on the pharmacokinetics of cyclophosphamide is also of minor importance unless severe, and no dose alteration is indicated in renal failure (Juma et al. 1981). 3.7 Ifosfamide High dose ifosfamide has been associated with a high frequency of CNS adverse effects. Although the metabolite chloracetaldehyde has come under suspicion (Goren et al. 1986), the exact mechanism behind this is uncertain. However, the occurrence of adverse effects is strongly related to hypoalbuminaemia and renal function and a useful nomo-

227


gram for assessing the risk has been presented by Meanwell et al. (1986). 3.8 Vincristine Although there are no definite data supporting a reduction of vincristine dosage in cases of liver dysfunction, the extensive liver metabolism and biliary excretion of this agent support a reduction in the case of hyperbilirubinaemia (50% reduction in bilirubin> 30 mg/L) [Chabner & Myers 1989]. 3.9 Cytarabine Only a small fraction of the cytarabine administered is excreted unchanged in the urine. The major route of inactivation is through deamination to the inactive metabolite uracilarabinoside (ara-U). The major site for deamination in humans is the liver, but information on the impact ofliver failure on the pharmacokinetics and toxicity of cytarabine is unfortunately lacking. However, the large interindividual variability in intracellular bioactivation reduces the importance of minor differences in the plasma drug concentration due to organ failure (Liliemark et al. 1985a). When high dose cytarabine (2 to 3 g/m2, 1 to 3h infusions twice daily for 4 to 6 days) is used, one of the- major problems is the cerebellar dysfunction which is apparently unpredictable. It has been stated that high age is the major determinant of the risk of this adverse effect (Herzig et al. 1987). It has recently been shown, however, that renal function and not age is an independent risk-factor for this adverse effect, and that a dosage reduction is recommended in patients with CLCR < 60 ml/min [calculated from serum creatinine (Cockcroft & Gault 1976)] (Damon et al. 1989). It also seems clear that the mode of administration, i.e. the rate of infusion, is important for the spectrum of adverse effects. When it is administered as a continuous infusion (Estey et al. 1987) or in an intermediate dose (0.5 g/m2, 2h infusion) for a prolonged time (Estey et al. 1988), the dose-limiting toxicity of high dose cytarabine is gastrointestinal, and the cerebellar dysfunction is rarely seen.

4. Pharmacologically Directed Phase I Trials Thousands of chemicals are screened every year for anticancer activity in preclinical test systems. Some of these drugs have interesting properties which warrant further testing in clinical trials. Due to the severe and sometimes life~threatening adverse effects of these drugs, only cancer patients who have failed standard treatments are used in phase I clinical trials, and as a result the number of patients available for such trials is limited. To determine the maximum tolerated dose (MTD) in humans, the dosage is escalated during the study and 3 to 5 patients, depending on whether toxicity is seen or not, are treated at each dose level. The choice of the starting level depends on the toxicity in animals, generally mice. The controversial question is, how rapid should the dosage escalation be? If it is too fast, some patients might suffer or die from undue toxicity. On the other hand, if it is too slow a greater number of patients will receive an ineffective treatment, more patients than necessary will enter the study and the development of this new drug (and other drugs) will be slowed. Therefore, to optimise dosage escalation a so-called modified Fibonaci scheme has been employed to date (Goldsmith et al. 1975). This means that the dosage escalation is made with larger steps (100%) early in the study and the subsequent steps are smaller (67 to 33%). Even so, up to 15 escalation steps need to be made in some studies, requiring more than 50 patients in a single phase I study. To decrease the number of dosage increments and patients in phase I trials Collins (1988) has suggested that pharmacokinetic monitoring can be used throughout the trial to direct the velocity of the dosage escalation. It was assumed that there is a better relationship between the plasma AUC and toxicity in humans and mice than between the MTD in humans and the lethal dose in 10% of the population (LD 10) in mice. This assumption is based on a retrospective analysis of pharmacokinetic and toxicity data from human phase I and preclinical mouse studies (Collins et al. 1986). The concept is that the MTD can be predicted when

228

the AVC at LDIO in mice and that in humans at the starting dose level are compared. The dosage escalation can then be made in greater steps (100%, or the square root of the ratio between the mouse AVC at LDIO and the AVC in humans at the starting dose) until 40% of the AVC at LDIO in mice is reached. This strategy has been used with some success in a newly published phase I study of 4'iodo-4'-deoxy-doxorubicin (Gianni et al. 1990). In this study, the species differences in drug metabolism complicated the study and limited the usefulness of the strategy. It is, however, clear that this method can be used to minimise the number of patients required for phase I trials and also to limit the number of patients treated at inefficient dosage levels in such studies (Collins et al. 1990). The preclinical data and the outline for a pharmacologically directed phase I study for CI-941, an anthrapyrazole drug, has been presented by Graham et al. (1989). The potential problems with and the strategy for pharmacokinetically guided dosage escalation of phase I trials have been carefully outlined by the European Organisation for Research on Treatment of Cancer (EORTC) pharmacokinetics and metabolism group (1987).

Addendum Recently, a large study (Egorin et al. 1991) showed that there is a correlation between the plasma carboplatin AUe and the antitumoural response in patients with ovarian cancer. It was shown that, although the myelotoxic effect of the treatment increased with increasing AUe, there was a plateau with regard to the antitumoural effect.

References Allegra CJ, Drake JC, Joliver J, Chabner BA. Inhibition of phosphoribosylaminoimidazolecarboxamide transformylase by methotrexate and dihydrofolic acid polyglutamates. Proceedings of the National Academy of the Sciences of the United States of America 82: 4881-4885, 1985 Andersson B, Beran M, Peterson C, Tribukait B. Significance of cellular pharmacokinetics for the cytotoxic effects of daunorubicin. Cancer Research 42: 178-183, 1982 Bagley CM, Bostick FW, DeVita Jr VT. Clinical pharmacology of cyclophosphamide. Cancer Research 33: 226-233, 1973 Baram J, Chabner B, Drake Jc, Fitzhugh AL, Sholar PW, et al.


Identification and biochemical properties of 10-formyl dihydrofolate, a novel folate found in methotrexate-treated cells. Journal of Biological Chemistry 263: 7105-7111, 1988 Belani C, Egorin MJ, Abrams JS. Hiponia D, Eisenberger M, et al. A novel pharmacodynamically based approach to dose optimization of carboplatin when used in combination with etoposide. Journal of Clinical Oncology 7: 1896-1902, 1989 Benjamin RS. A practical approach to adriamycin toxicity. Cancer Chemotherapy Report 6: 191-194, 1975 Borsi JD, Revesz T, Schuler D. Prognostic importance of systemic clearance of methotrexate in childhood acute lymphoblastic leukemia. Cancer Chemotherapy and Pharmacology 19: 261-264, 1987 Carson DA, Wasson DB, Teatle R, Yu A. Specific toxicity of 2chlorodeoxyadenosine towards resting and proliferating human lymphocytes. Blood 62: 737-743, 1983 Chabner BA. Clinical strategies for cancer treatment: the role of drugs. In Chabner & Collins (Eds) Cancer chemotherapy, pp. 1-15, J.B. Lippincott Company, Philadelphia, 1990 Chabner BA, Myers CEo Clinical pharmacology of cancer chemotherapy. In DeVita et al. (Eds) Cancer, pp. 382-384, J.B. Lippincott Company, Philadelphia, 1989 Clark PI, Joel SP, Houston S, Gregory WM, Slevin ML. Predictors of etoposide pharmacokinetics in man. British Journal of Cancer 58 (Suppl. 2): 258-259, 1988 Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 16: 31-41, 1976 Collins 1M. Pharmacology and drug development. Journal of the National Cancer Institute 80: 790-792, 1988 Collins JM, Grieshaber CK, Chabner BA. Pharmacologically guided phase I clinical trials based upon preclinical drug development. Journal of the National Cancer Institute 82: 13211326, 1990 Collins JM, Zaharko DS, Dedrick RL. Chabner BA. Potential roles for preclinical pharmacology in phase I clinical trials. Cancer Treatment Reports 70: 73-80. 1986 D'lncalci M. Bolis G, Facchinetti T, Mangioni C, Morasca L, et al. Decreased half-life of cyclophosphamide in patients under continual treatment. European Journal of Cancer 19: 7-10,1979 D'lncalci M, Rossi C, Zucchetti M, Urso R, Cavalli F, et al. Pharmacokinetics of etoposide in patients with abnormal renal and hepatic function. Cancer Research 46: 2566-2571, 1986 Dalgleish AG, Woods RL, Levi JA. Bleomycin pulmonary toxicity: its relationShip to renal dysfunction. Medical and Pediatric Oncology 12: 313-317, 1984 Damon LE, Mass R, Linker CA. The association between highdose cytarabine neurotoxicity and renal insufficiency. Journal of Clinical Oncology 7: 1563-1568, 1989 DeVita Jr VT. Dose-response is alive and well. Editorial. Journal of Clinical Oncology 4: 1157-1159, 1986 DeVita Jr VT, Serpick AA, Carbone PP. Combination chemotherapy in the treatment of advanced Hodgkin's disease. Annals of Internal Medicine 73: 881-895,1970 Egorin MJ. Phase I trials: a strategy of ongoing refinement. Journal of the National Cancer Institute 82: 446-447, 1990 Egorin MJ, Jodrell D, Canetta R, Langenberg P, Goldbloom E, et al. Tumor response and toxicity in ovarian cancer correlates with carboplatin (CBDCA) area under the curve (AUC). Proceedings of the Annual Meeting of the American Society of Clinical Oncology 10: 184, 1991 Egorin MJ, Sigman LM, vanEcho DA, Forrest A, Whitacre MY, et al. Phase I clinical and pharmacokinetic study of hexamethylene bisacetamide (NSC 95580) administered as a five-day continuous infusion. Cancer Research 47: 617-623, 1987


Egorin MJ, vanEcho DA, Olman EA, Whitacre MY, Forrest A, et al. Prospective validation of a pharmacologically based dosing scheme for the cis-diaminedichloroplatinum (II) analog diaminecyclobutanedicarboxylatoplatinum. Cancer Research 45: 6502-6506, 1985 Egorin MJ, vanEcho DA, Tippling SJ, Olman EA, Whitacre MY, et al. Pharmacokinetics and dosage reduction of cisDiamino( I, I-cyclobutanedicarboxylato) platinum in patients with impaired renal function. Cancer Research 44: 5432-5438, 1984 Egorin MJ, vanEcho DA, Whitacre MY, Forrest A, Sigman LM, et al. Human pharmacokinetics, excretion and metabolism of the anthracykline antibiotic menagaril (7-OMEN, NSC 269148) and their correlation with clinical toxicities. Cancer Research 46: 1513-1520, 1986 Ehrsson H, Eksborg S, Wallin I, Dsterborg A, Mellstedt H. Oral melphalan pharmacokinetics: influence of interferon-induced fever. Oinical Pharmacology and Therapeutics 47: 86-90, 1990 EORTC Pharmacokinetics and Metabolism Group. Pharmacokinetically guided dose escalation in phase I clinical trials: commentary and proposed guidelines. European Journal of Cancer and Oinical Oncology 23: 1083-1087, 1987 Estey EH, Keating MJ, McCredie KB, Freireich EJ, Plunkett W. Cellular ara-CTP pharmacokinetics, response and karyotype in newly diagnosed acute myelogenous leukemia. Leukemia 4: 9599, 1990 Estey E, Keating MJ, Plunkett W, McCredie KB, Freireich EJ. Continuous infusion high dose cytosine arabinoside without anthracyclines as induction and intensification therapy in adults under age 50 with newly diagnosed acute myelogenous leukemia. Seminars in Oncology 14 (Suppl. I): 58-63, 1987 Estey EH, Plunkett WK, Keating MJ, McCredie KB, Freireich EJ. Cytosine arabinoside (ara-C) in intermediate doses (IDAC) as therapy for patients (Pts) with acute myelogenous leukemia (AML). Proceedings of the American Association for Cancer Research 29: 209, 1988 Evans WE, Crom WR, Schell MJ, Kalwinsky DK, Rivera GK.. Reappraisal of methotrexate clearance as prognostic factor in childhood acute lymphocytic leukemia. Proceedings of the Annual Meeting of the American Association for Cancer Research 30: 241, 1989 Evans WE, Crom WR, Stewart CF, Bowman WP, Chen CH, et al. Methotrexate systemic clearance influences probability of relapse in children with standard-risk acute lymphocytic leukemia. Lancet I: 359-362, 1984 Evans WE, Crom WR, Yalowich J. Methotrexate. In Evans et al. (Eds) Applied pharmacokinetics: principles of therapeutic drug monitoring; pp. 1009-1056, Applied Therapeutics Inc., Spokane, 1986 Evans WE, Rodman JH, Petros WP, Madden T, Crom WR, et al. Individualized doses of chemotherapy for children with acute lymphocytic leukemia (ALL). Proceedings of the Annual Meeting of the American Society of Clinical Oncology 9: 69, 1990 Finan PJ, Chisholm EM, Woodhouse L, Giles GR. The relationship between plasma pharmacokinetics and tissue metabolites of 5-fluorouracil (5-FU) in patients with colorectal cancer. European Journal of Surgical Oncology 13: 349-353, 1987 Fraumeni JJ, Hoover R, Devesa S, Kinlen L. Epidemiology of Cancer. In DeVita et al. (Eds) Cancer: principles and practice of oncology, 3rd ed. pp. 196-235, J.B. Lippincott Company, Philadelphia, 1989 Gianni L, Vigiano L, Surbone A, Ballinari 0, Casali P, et al. Pharmacology and clinical toxicity of 4'-iodo-4'-deoxydoxo-

229

rubicin: an example of successful application of pharmacokinetics to dose escalation in phase I trials. Journal of the National Cancer Institute 82: 469-477, 1990 Goldsmith RF, Slavik M, Carter SK. Quantitative prediction of drug toxicity in humans from toxicology in small and large animals. Cancer Research 43: 1354-1364, 1975 Goren MP, Wright RK, Pratt CB, Pell FE. Dechloroethylation of ifosfamide and neurotoxicity. Lancet 2: 1219-1220, 1986 Graham MA, Newell DR, Foster BJ, Calvert AH. The pharmacokinetics and toxicity of the anthrapyrazole anti-cancer drug CI-941 in the mouse: a guide for rational dose esca1ation in patients. Cancer Chemotherapy and Pharmacology 23: 8-14, 1989 Griffin TW, Bogden AE, Reich SD, Antonelli D, Hunter RE, et al. Initial clinical trials of the subrenal capsule assay as a predictor of tumor response to chemotherapy. Cancer 52: 185192, 1983 Grochow LB, Baraldi C, Noe D. Is dose normalization to weight or body surface area useful in adults? Journal of National Cancer Institute 82: 323-325, 1990 Hamburger A W, Salmon SE. Primary bioassay of human tumor stem cells. Science 197: 461-463, 1977 Hande KR. Reply. Journal of Oinical Oncology 8: 2088-2089, 1990 Hande KR, WolffSN, Greco FA, Hainsworth JD, Reed G, et al. Etoposide kinetics in patients with obstructive jaundice. Journal of Clinical Oncology 8: 1101-1108, 1990 Harris BE, Song R, Soong S-J, Diasio RB. Relationship between dihydropyrimidine dehydrogenase activity and plasma 5-fluorouracillevels with evidence for circadian variation of enzyme activity and plasma drug levels in cancer patients receiving 5fluorouracil by protracted continuous infusion. Cancer Research 50: 197-201, 1990 Hayder S. Maintenance therapy in childhood acute lymphoblastic leukemia, Thesis, Karolinska Institute, Stockholm, 1989 Hayder S, Lafolie P, Bjork 0, Peterson C. 6-Mercaptopurine plasma levels in children with acute lymphoblastic leukemia: relation to relapse risk and myelotoxicity. Therapeutic Drug Monitoring 11: 617-622, 1989 Heinemann V, Estey E, Keating MJ, Plunkett W. Patient-specific dose rate for continuous infusion high-dose cytarabine in relapsed acute myelogenous leukemia. Journal of Clinical Oncology 7: 622-628, 1989 Heinemann V, Jehn U. Rationals for a pharmacologically optimized treatment of acute nonlymphocytic leukemia with cytosine arabinoside. Leukemia 4: 790-796, 1990 Herzig RH, Hines JD, Herzig GP, WolffSN, Cassileth PA, et al. Cerebellar toxicity with high-dose cytosine arabinoside. Journal of Clinical Oncology 6: 927-932, 1987 Hillcoat BL, McCulloch PB, Figueredo AT, Ehsan MH, Rosenfeld JM. Oinical response and plasma levels of 5-fluorouracil in patients with colonic cancer treated by drug infusion. British Journal of Cancer 38: 719-724,1978 Hryniuk WM. More is better. Journal of Oinical Oncology 6: 1365-1367, 1988 Hryniuk W, Bush H. The importance of dose intensity in chemotherapy of metastatic breast cancer. Journal of Oinical Oncology 2: 1281-1288, 1984 Jacobs C, Kalman SM, Tretton M, Weiner MW. Renal handling of cis-diaminedichloroplatinum (II). Cancer Treatment Reports 64: 1223-1226, 1980 Juma FD. Effect of liver function on the pharmacokinetics of

230

cyclofosfamide. European Journal of Clinical Pharmacology 26: 591-593, 1984 Juma FD, Rogers HJ, Trounce JR. EtTect of renal insufficiency on the pharmacokinetics of cyclophosphamide and some of its metabolites. European Journal of Clinical Pharmacology 19: 443-451, 1981 Kawai M, Rosenfeld J, McCulloch P, Hillcoat B. Blood levels of 5-fluorouracil during a five day continuous infusion. British Journal of Cancer 36: 346-347, 1976 Kerr IG, Jolivet J, Collins JM, Drake J, Chabner BA. Test dose for predicting high-dose methotrexate infusions. Clinical Pharmacology and Therapeutics 33: 44-51, 1983 Koren G, Ferrazini G, Sulh H, Langevin AM, Kapelushnik J, et al. Systemic exposure to mercaptopurine as a prognostic factor in acute lymphocytic leukemia in children. New England Journal of Medicine 323: 17-21, 1990 Lafolie P, Hayder S, Bjork 0, Astrom L, Liliemark J, et al. Large interindividual variations in the pharmacokinetics of oral 6mercaptopurine in maintenance therapy of children with acute leukemia and non-Hodgkin lymphoma. Acta Paediatrica Scandinavica 75: 797-803, 1986 Lafolie P, Hayder S, Bjork 0, Peterson C. Intra-individual variations in 6-mercaptopurine concentrations in plasma and red blood cells during oral maintenance therapy of leukemic children. European Journal of Clinical Pharmacology, in press, 1991 Langevin AM, Koren G, Soldin SJ, Greenberg M. Pharmacokinetic case for giving 6-mercaptopurine maintenance doses at night. Lancet 2: 505-506, 1987 Laurell G, Jungnelius U. High-dose cisplatin treatment. Hearing loss and plasma concentrations. Laryngoscope 100: 724-734, 1990 Lennard L, Lilleyman JS. Variable mercaptopurine metabolism and treatment outcome in childhood lymphoblastic leukemia. Journal of Clinical Oncology 7: 1816-1823, 1989 Lennard L, Lilleyman JS, Loon JV, Weinshilboum RM. Genetic variation in response to 6-mercaptopurine for childhood acute lymphoblastic leukemia. Lancet 336: 225-229, 1990 Lennard L, Rees CA, Lilleyman JS, Maddocks JL. Childhood leukemia: a relationship between intracellular 6-mercaptopurine metabolites and neutropenia. British Journal of Clinical Pharmacology 16: 359-363, 1983 Liliemark J, Dixon DO, Plunkett W. The relationship between I-~-D arabinofuranosyl cytosine in plasma to I-~-D arabi nofuranosyl cytosine 5-triphosphate in leukemic cells during treatment with high dose I-~-D arabinofuranosyl cytosine. Cancer Research 45: 5952-5957, 1985a Liliemark JO, Gahrton G, Paul CY, Peterson CO. AraC in plasma and araCTP in leukemic cells after subcutaneous injection and continuous infusion of araC in patients with acute nonlymphoblastic leukemia. Seminars in Oncology 14 (Suppl. I): 167171, 1987 Liliemark JO, Paul CY, Gahrton CG, Peterson CO. Pharmacokinetics of I-~-D-arabinofuranosylcytosine 5' -triphosphate in leukemic cells after intravenous and subcutaneous administration of I-~-D-arabinofuranosylcytosine. Cancer Research 45: 2373-2375, 1985b Lonnerholm G, Lindstrom B, Ludvigsson J, Myrdal U. Plasma and erythrocyte concentrations of mercaptopurine after oral administration in children. Pediatric and Hematological Oncology 3: 27-35, 1986 Matherly LH, Fry DW, Goldman ID. Role of methotrexate poIyglutamation and cellular energy metabolism in inhibition of methotrexate binding to dihydrofolate reductase by 5-formyltetrahydrofolate in Ehrlich ascites tumor cells in vitro. Cancer


Research 43: 1313-1318, 1983 Meanwell CS, Blake AE, Kelly KA, Honigsberger L, Blackledge G. Prediction of ifosfamide/mesna associated encephalopathy. European Journal of Cancer and Clinical Oncology 22: 815819,1986 Milano G, Roman P, Khater R, Frenay M, Renee N, et al. Dose versus pharmacokinetics for predicting tolerance to 5-day continuous infusion of 5-FU. International Journal of Cancer 41: 537-541, 1988 Paul C, Tidefelt U, Liliemark J, Peterson C. Increasing the accumulation of daunorubicin in human leukemic cells by prolonging the infusion time. Leukemia Research 13: 191-196, 1989 Plunkett W, lacoboni S, Estey E, Dannhauser L, Liliemark JO, et al. Pharmacologically directed ara-C therapy for refractory leukemia. Seminars in Oncology 2 (Suppl. I): 20-30, 1985 Plunkett W, Liliemark JO, Adams TM, Nowak B, Estey E, et al. Saturation of I-~-D-arabinofuranosylcytosine 5-triphosphate accumulation in leukemia cells during high-dose I-~-D-arabi nofuranosylcytosine therapy. Cancer Research 47: 3005-3011, 1987 Retain MJ. Dose reduction of etoposide in jaundiced patients. Journal of Clinical Oncology 8: 2088, 1990 Retain MJ, Schilsky RL, Choi KE, Guarnieri C, Grimmer D, et al. Adaptive control of etoposide administration: impact on interpatient pharmacodynamic variability. Clinical Pharmacology and Therapeutics 45: 226-233, 1989 Retain MJ, Vogelzang NJ, Sinkule JA. Interpatient and intrapatient variability in vinblastine pharmacokinetics. Clinical Pharmacology and Therapeutics 41: 61-67, 1987 Rivard GE, Hoyoux C, Infante-Rivard C, Champagne J. Maintenance chemotherapy of childhood acute lymphoblastic leukemia: better in the evening. Lancet 2: 1264-1266, 1985 Rodman JH, Abromowith M, Sinkule JA, Hayes FA, Rivera GK, et al. Clinical pharmacodynamics of continuous infusion teniposide: systemic exposure as a determinant of response in a phase I trial. Journal of Clinical Oncology 5: 1007-1014, 1987 Rustum YM, Preisler HD. Correlation between leukemic cell retention of I-~-D-arabinofuranosylcytosine 5' -triphosphate and response to therapy. Cancer Research 39: 42-49, 1979 Schmiegelow K, Pulczynska M, Seip M. White cell count during maintenance chemotherapy for standard-risk childhood acute lymphoblastic leukemia: relation to relapse rate. Pediatric Hematology and Oncology 5: 259-267, 1988 Seitz JF, Cano JP, Rigault JP, Aubert C, Carcassonne Y. Chimotherapie des cancers digestifs etendus par Ie fluorouracil: relations entre la reponse clinique et la clairance plasmatique du medicament. Gastroenterologie C1inique et Biologie 7: 374380, 1983 Sinkule JA, Hutson P, Hayes FA, Etcubanas E, Evans W. Pharmacokinetics of etoposide (VP-16) in children and adolescents with refractory solid tumors. Cancer Research 44: 3109-3113, 1984 Sjoqvist F, Borga, 0, Orme MLE. Fundamentals of clinical pharmacology. In Avery (Ed.) Drug treatment: principles and practice of clinical pharmacology and therapeutics, pp. 1-61, Adis Press, Sydney, 1980 Skipper HE, Schabel J, Wilcox WS. Experimental evaluation of potential anticancer agents XJI[: on the criteria and kinetics associated with 'curability' of experimental leukemia. Cancer Treatment Reports 55: I-Ill, 1964 Slevin ML, Clark PI, Joel SP, Malik S, Osborne RJ, et al. A randomized trial to evaluate the etTect of schedule on the activity of etoposide in small-cell lung cancer. Journal of Clinical On-


231

cology 7: 1333- 1340, 1989 Slevin ML, Piall EM, Aherne GW, Johnstone A, Sweatman MC, et al. The pharmacokinetics of subcutaneous cytosine arabinoside in patients with acute myelogenous leukemia. British Journal of Clinical Pharmacology 12: 507-510, 1981 Spears CP, Gustavsson BG, Frosing R. Folinic acid modulation of fluorouracil: tissue kinetics of bolus administration. Investigational New Drugs 7: 27-36, 1989 Spears C, Gustavsson B, Mitchel MS, Spicer 0, Berne M, et al. Thymidylate synthetase inhibition in malignant tumours and normal liver in patients given bolus intravenous 5-fluorouracil. Cancer Research 44: 4144-4150, 1984 Spears CP, Waugh W, Leichman L, Leichman CG, Jeffers S, et al. Salvage therapy of breast cancer with fluorouracil (5-FU) and high-dose leukovorin (LV): response correlations with tumor pharmacodynamics. Proceedings of the Annual Meeting of the American Society of Clinical Oncology 9: 73, 1990 Spriggs DR, Griffin J, Wisch J, Kufe OW. Clinical pharmacology of low-dose cytosine arabinoside. Blood 65: 1087-1089, 1985 Stewart CF, Arbuck SG, Fleming RA, Evans WE. Changes in the clearance of total and unbound etoposide in patients with liver dysfunction. Journal of Clinical Oncology 8: 1874-1879, 1990 Stroller RG, Hande KR, Jacobs SA, Rosenberg SA, Chabner BA, et al. Use of plasma pharmacokinetics to predict and prevent methotrexate toxicity. New England Journal of Medicine 297: 630-634, 1977 Thyss A, Milano G, Renee N, Vallicioni J, Schneider M, et al. Clinical pharmacokinetic study of 5-FU in continuous 5 day infusion for head and neck cancer. Cancer Chemotherapy and Pharmacology 16: 64-66, 1986 Tidd OM, Kim SC, Horakova K, Morivaki A, Paterson ARP. A

delayed cytotoxic reaction for 6-mercaptopurine. Cancer Research 32: 317-322, 1972 Tidefelt U, Sundman-Engberg B, Rhedin A-S, Paul C. In vitro drug testing in patients with acute leukemia with incubations mimicking in vivo intracellular drug concentrations. European Journal of Hematology 43: 374-384, 1989 Weisenthal LM , Lippman ME. Clonogenic and nonclonogenic in vitro chemosensitivity assays. Cancer Treatment Reports 69: 615-632, 1985 Whitehead WM, Vuchich MJ, Rosenblatt OS, Shuster J, Witte A, et al. Children with acute lymphoblastic leukemia (ALL) experience better event free survival (EFS) if their Iymphoblasts (LY) at diagnosis accumulate high levels of methotrexate polyglutamates (MTXPG). Proceedings of the Annual Meeting of the American Association for Cancer Research 30: 246, 1989 Wolf W, Present CA, Servis KL, EI-Tahtawry A, Albright MJ, et al. Tumor trapping of 5-fluorouracil: in vivo 19F NMR spectroscopic pharmacokinetics in tumor bearing humans and rabbits. Proceedings of the National Academy of the Sciences of the United States of America 87: 492-496, 1990 Zimm S, Collins J, Riccardi R, O'Neill 0 , Narang P, et al. Variable bioavailability of oral mercaptopurine: is maintenance chemotherapy in acute lymphoblastic leukemia being optimally delivered? New England Journal of Medicine 308: 10051009, 1983 Correspondence and reprints: Dr Curt Peterson, Department of Clinical Pharmacology, Karolinska Hospital, PO Box 60 500, S-104 01 Stockholm, Sweden.

2nd Jerusalem Conference on

Pharmaceutical Sciences and Clinical Pharmacology Date: 24-29 May, 1992 Venue: Jerusalem, Israel For further information, please contact: Professor Meir Bialer 2nd Jerusalem Conference on Pharmaceutical Sciences and Clinical Pharmacology P.O. Box 50006 Tel Aviv 61500 ISRAEL

Pharmacokinetic optimisation of antiemetic therapy.

Pharmacokinetic optimisation of anticonvulsant therapy.

Pharmacokinetic optimisation of oral hypoglycaemic therapy.

Optimisation of immunosuppressive therapy using pharmacokinetic principles.

Practical optimisation of antiarrhythmic drug therapy using pharmacokinetic principles.

Pharmacokinetic optimisation of angiotensin converting enzyme (ACE) inhibitor therapy.

Pharmacokinetic optimisation of histamine H1-receptor antagonist therapy.

[Mecanisms of pharmacokinetic interactions involving oral anticancer agents].

PP2A-Mediated Anticancer Therapy.

tazobactam in patients with severe infections: A possible pharmacokinetic optimisation?

Targeting apoptosis for anticancer therapy.

PARP inhibitors for anticancer therapy.

Preclinical pharmacokinetic studies of 3-deazaneplanocin A, a potent epigenetic anticancer agent, and its human pharmacokinetic prediction using GastroPlus™.

The oral complications of specific anticancer therapy.

Nicotine therapy in smoking cessation. Pharmacokinetic considerations.

Trial watch: Dendritic cell-based anticancer therapy.

Energy disruptors: rising stars in anticancer therapy?

Exploiting tumor cell senescence in anticancer therapy.

Cell Death Identification in Anticancer Therapy-Letter.

Trial Watch: Proteasomal inhibitors for anticancer therapy.

Targeted anticancer therapy: overexpressed receptors and nanotechnology.

Human gene therapy: anticancer trial's surprise approval.

Pharmacokinetic considerations on Orgaran (Org 10172) therapy.

Anticancer properties of nimbolide and pharmacokinetic considerations to accelerate its development.