Pharmacokinetics of ethylenediaminemalonatoplatinum(II) (JM-40) during phase I trial.

Pharmacokinetics of the cis-platin analog ethylenediaminemalonatoplatinum(II) (JM-410) was studied in 28 cycles of 19 patients during the phase I study of this drug. The drug was administered intravenously by short-term (10-60 min) infusion. Doses ranged from 20 to 1,200mg m-2. JM-40 was determined in plasma ultrafiltrate and urine by HPLC. Platinum (Pt) concentrations were determined in plasma, plasma ultrafiltrate, urine and red blood cells by atomic absorption spectrometry up to 5 days after administration of the drug. Ultrafilterable Pt could be determined up to 45 days after the infusion in one patient sampled over such a long period. Pharmacokinetics of JM-40 showed a linear behaviour. The final half-life of total Pt in plasma was 4.1 +/- 0.9 days. The disposition of JM-40 was similar to that of ultrafilterable Pt in respect to t1/2 alpha (10 and 13 min), t1/2 beta (44 and 57 min), volumes of distribution Vc (11 and 121) and Vss (17 and 201), systemic clearance (256 and 223 ml min-1), renal clearance (69 and 73 ml min-1) and metabolic clearance (183 and 154 ml min-1). During the first 6 h 27 +/- 9% of the administered dose was excreted as JM-40. Cumulative platinum excretion in the urine amounted to 29 +/- 13% and 60 +/- 13% over the first 6 h, 24 h and 5 days, respectively. The uptake of platinum in red blood cells was limited, comprising only 0.24 +/- 0.12% of the administered dose. Although JM-40 and carboplatin are structurally closely related, pharmocokinetics and toxicity of JM-40 were more similar to cis-platin than to carboplatin.

Ethylenediaminemalonatoplatinum(II) (JM-40, Figure 1) is one of the second generation platinum complexes developed with the aim to achieve a better therapeutic index than that of the antitumour drug cis-diammine-dichloroplatinum(II) (cis-platin) (Cleare et al., 1978). JM-40 was selected for clinical evaluation on the basis of comparable activity (Rose & Bradner 1984;Boven et al., 1985) and a favorable toxicity profile (Schurig et al., 1984;Lelieveld et al., 1984) compared to cis-platin in preclinical studies. In particular the low emetogenic potential as determined in the ferret (Schurig et al., 1984) and the limited nephrotoxicity observed in dogs (Lelieveld et al., 1984) were reasons to conduct a phase I trial. In this trial the maximum tolerable dose was reached at 1,200 mg m -2 (Winograd et al., 1986). Nephrotoxicity, nausea and vomiting were dose-limiting toxicities.
Investigation of the pharmacokinetics during phase I clinical trials is important because it may help to design an optimal therapeutic regimen in phase II trials (Kovach, 1983). In the case of an analogue, pharmacokinetics can be compared with that of the parent compound. Furthermore, when preclinical information on animal pharmacokinetics is available human pharmacokinetics at low doses of the phase I trial may aid in escalating the dose as quickly and safely as possible (Collins et al., 1986, Van Hennik et al. 1987. Thus, clinical pharmacokinetics of JM-40 and platinum were investigated in plasma, plasma ultrafiltrate, urine and red blood cells.

Patients and methods
Patients and materials Pharmacokinetic studies were performed in 19 patients who received 28 cycles of JM-40 within a dose range of 20-1,200mgm-2. The median age of the patients (6 females, 13 males) was 57 yrs (range 37-74 yrs). All patients had a normal liver function. Renal function was decreased (creatinine clearance <60mlmin-1) at the start of only 3 treatment cycles. Six patients had previously received other platinum compounds (cis-platin, spiroplatin).
JM-40 was supplied by Johnson Matthey, Reading, Berkshire, UK. It was formulated as an aqueous solution (5 mg ml -1) (T.J., Schoemaker, Slotervaart Hospital, Amsterdam, The Netherlands) and diluted 1: 1 in 10% glucose prior to administration. The solution was given i.v. with an infusion time (T) of 10 min up to a dose of 300 mg m -2. T increased to 60 min because of increasing volumes of the solubilized drug at higher dose levels.
Sampling and analysis Blood samples (5 ml) were collected in heparinized tubes prior to administration of JM-40, at the end of the infusion and at 10,20,30,60,90,120,150,180,210,240,360,480,1,220, 1,440 min as well as 2, 3, 4, and 5 days thereafter. Samples were processed immediately after collection. Blood was centrifuged and 2 portions of I ml plasma were ultrafiltrated in the MPS-I micropartition system provided with YMT filters (cut-off 30,000 dalton, Amicon, Oosterhout, The Netherlands) ( Van der Vijgh et al., 1986). Red blood cells (RBCs) were washed twice with an equal volume of normal saline and centrifuged. Urine was collected in successive portions up to 2, 4, 6, 24 h and 2, 3, 4, and 5 days after JM-40 administration. All samples for platinum analysis were stored at -25°C. Platinum concentrations in plasma (total Pt), plasma ultrafiltrate (free Pt), RBCs and urine were determined for all 28 courses by atomic absorption spectrometry (AAS) as described before .
Not all patients were sampled completely. Therefore different groups of patients were used to calculate the various pharmacokinetic parameters. In 10 courses JM-40 was determined in plasma ultrafiltrate and urine by high performance liquid chromatography (HPLC) with UV detection at 214 nm . Because of the limited stability of JM-40 in body fluids  all JM-40 determinations were performed immediately after collection of the samples. Stability of JM-40 in plasma was determined by incubating JM-40 in plasma of healthy volunteers at 37°C for 5 h. The initial concentration was 100pyM. At 0, 0.5, 1, 2, 3 and 5h, samples of lml were taken, ultrafiltrated, and analyzed for JM-40 as outlined above.
Pharmacokinetic data analysis Plasma concentration vs. time curves were fitted to a polyexponential equation n Cp Yi exp (-1it) by the computer program NONLIN (Metzler et al., 1974).  (Wagner, 1976). AUC values were corrected if total Pt levels were not zero at the start of administration to pretreated patients.
Renal clearance (CLR) was determined from the cumulative urinary excretion (CUE) divided by the A UC both measured over the same time-interval (6 h, including the infusion time). The AUC 0-6h was calculated by means of the linear trapezoidal rule. Metabolic clearance (CLM) was calculated from CL-CLR. Half-life of total Pt over day 1-5 was additionally determined by use of the least squares method.
The volume of distribution at steady state (V1') of total Pt was calculated as CL x AUMC/AUC (Wagner, 1976). VI', of JM-40 and free Pt was calculated according to Collier (1983)  The first order rate constant of metabolic elimination, KM, represents the overall reactivity of the drug towards body constituents (plasma as well as tissue components). It was calculated by analogy with the overall first order rate constant of elimination at steady state K,,=CL/IV,' (Benet et al., 1979). Thus Table I).
Pt and total Pt in plasma as well as Pt in RBCs after administration of an intermediate dose of JM-40 to a patient sampled over 45 days. The curves of other patients, followed for 5 days, had the same appearance as those in Figure 2 up to day 5. Free Pt concentrations could only be measured over at least 5 days in patients who received JM-40 at a dose of 300mgm 2 or higher. JM-40 could be measured over the first 3.5-7h after infusion of 300-1200mgm-2 (detection limit of the HPLC assay was 1 tiM). The curves of total Pt and free Pt over the first 5 days showed a triphasic decline, while two phases could be observed for JM-40. Plasma levels of free Pt were higher than that of JM-40, indicating reaction of JM-40 with low molecular weight endogenous compounds. The presence of the long lasting third phase of free Pt is thought to be due to platinum containing breakdown products of protein-JM-40 complexes. A small temporal rise of the total Pt concentration somewhere between 3-8 h after administration was observed in 10 of the 28 concentration-time curves (36%), suggesting an enterohepatic recirculation as has been described for cis-platin before (Vermorken et al., 1984b. Peak concentrations and basic pharmacokinetic parameters of JM-40 in individual courses as calculated by NONLIN are given in Table I. The best number of exponents as decided by NONLIN was 2 except for patient No. 2, where the distribution phase could not be distinguished from the elimination phase. Mean values of common pharmacokinetic parameters are listed in Table II. In general, the curves of free Pt were best fitted by three exponential terms, except in the lower dose range (<120 mg m 2) where the free Pt concentrations were too low to observe the third phase. The half-life of the third phase of free  Pt (1.9 + 0.7 days) was comparable with that of total Pt, suggesting that this phase represents free Pt released by degradation of macromolecules reacted with JM-40. All pharmacokinetic parameters of free Pt were calculated using the exponents and coefficients of the first two phases only (Table II). This allowed a comparison of the pharmacokinetics of free platinum originating from JM-40 to those originating from other platinum compounds, because (a) a third phase could not be observed for the other compounds due to either a lower dose or the detection limit of the assay, (b) kinetic parameters calculated this way refer to probably comparable species between the compounds. Total Pt concentration vs. time curves were also best fitted by 3exponential equations in most cases. The goodness of fit parameter r2 (Wagner et al., 1977) ranged from 0.98-0.9999.
Half-lives of free and total Pt (over the first two phases) were higher than that of JM-40 due to metabolism and protein binding, respectively. The high standard deviation observed for t, may be due to the observed increase in halflife with increasing infusion time. The half-life for total Pt calculated by means of the least squares method over the discrete time interval of day 1-5 was 4.1 + 0.9 days, being slightly higher than the value of the terminal half-life calculated by NONLIN. In figure 2 a fourth phase in the curves of total Pt and free Pt was observed starting from day 16 onwards. In this patient, half-lives were determined by use of the least squares method being 15.5d and 14.4d for total Pt and free Pt, respectively. Linear correlations were observed between dose m 2 and A UC (P<0.05 for JM-40 (n = 10) and P<0.01 for free Pt (n=28) and total Pt (n=28)), indicating linear pharmacokinetics over the dose range studied (20-1,200 mg m 2).
At 6 and 24 h after administration protein binding of platinum was 91 + 2 and 93 + 1%, respectively. Due to protein binding not only in plasma but also in tissues, the systemic clearance of total Pt was low and the volume of distribution at steady state was high compared to JM-40 and free Pt. The volume of distribution and the clearance of free Pt determined from the first two phases were comparable with those of JM-40. As expected, the volume of distribution in the central compartment, VI, was similar for all three species. The in vitro degradation rate constant of JM-40 in plasma, kinvitro, was found to be 0.004min-1. For free Pt a value of 0.002min-1 was determined earlier . From these values and the mean values of CLR and CLM the fractions eliminated from the central and peripheral compartments were calculated to be fi = 0.44 and f2=0.56 for JM-40 and f1=0.43 and f2= 0.57 for free Pt. V., values calculated with these parameters were about three times higher for total Pt than for JM-40 and free Pt. The resultant values of V,, for JM-40 and free Pt listed in Table II were 17% and 20% higher, respectively, than when calculated in the conventional way (Wagner, 1976), ignoring elimination from the peripheral compartment. The metabolic clearance, CLM, and the overall first order rate constant of metabolic elimination, KM, were comparable for JM-40 and free Pt.
Appreciable amounts of JM-40 were excreted in urine. The cumulative urinary excretion over the first 6h was 27+9%D (n=5) and 29+13%D (n=19) for JM-40 and free Pt, respectively. After 24h 42+14% (n=19) and after 5 days 60 + 13% (n = 12) of the administered dose was excreted as free Pt. The mean values of renal clearance of JM-40 and free Pt as measured over the first 6h after administration, were comparable. Both values were similar to the mean creatinine clearance (71+22, n = 23) as measured a day before the start of therapy. However, individual values of creatinine clearance and renal clearance of JM-40 (n= 5) and free Pt (n= 10) did not show statistically significant correlations.
Pt was rapidly taken up by RBCs during the first 20 min of exposure. Maximum levels were reached between 2 and 8 h after the start of infusion. The half-life of Pt in RBCs as measured over day 1-5, was 14 + 3 days (n = 6). At the maximum concentration only 0.24 + 0.12% of the dose (n=25) was bound to RBCs (Long et al., 1981). Therefore, platinum uptake by RBCs is not a site of drug accumulation.

Discussion
Our HPLC method (Van der Vijgh et al., 1984) was sensitive enough to determine JM-40 for at least 5 final half-life times. Total and free Pt in plasma were determined up to 5 days following infusion, allowing a reliable fit of a triexponential equation through the concentation-time curves. A fourth phase was observed in one patient sampled up to 45 days. The half-lives of the third and fourth phase of total Pt (4.1 and 15.5 days) were comparable to those of cis-platin (5.3 and 12.0 days) (Vermorken et al., 1984b), suggesting binding to the same plasma proteins as cis-platin and equivalent turn-over rates of the Pt-labeled proteins as in the case of cis-platin.
The pharmacokinetic parameters of free Pt were calculated with exclusion of the third (and fourth) phase for the following reasons. Protein binding is regarded as irreversible (Repta et al., 1980). Therefore, the long terminal phase of free Pt and the parallelism of the fourth phase of free Pt with that of total Pt (Figure 2) suggests that the third phase of the free Pt curves (if detectable) represents platinum containing degradation products of high molecular weight compounds (i.e. proteins and tissue components) (King et al., 1986). Since protein bound platinum has no toxic or antitumour activity (Repta et al., 1980) this will probably also hold for the platinum containing degradation products of those proteins. Besides, pharmacokinetic parameters of free Pt calculated without taking into consideration an eventually present third phase allows intercomparison of free platinum species originating from most platinum compounds for which it was not possible to detect a third phase. Therefore, we decided to omit these secondary free platinum species from the calculation of clearance and volume parameters by using only the first two phases, mainly referring to JM-40 and its low molecular weight metabolites. The pharmacokinetic parameters of free Pt were similar to those of JM-40. This similarity suggests that the formation of low molecular weight (<30,000 dalton) metabolites accounts for only a small part of the total metabolic elimination of JM-40.
The major part of metabolic elimination of platinum compounds is due to irreversible ligand exchange reactions. It is very likely that this takes place not only in the central compartment but also in the peripheral compartments (lumped together as compartment 2). This was taken into account by the way VSS was calculated (Collier, 1983). Accurate estimation of VSS was desirable, because it was used to calculate KM. The rate constants KM of JM-40 and free Pt were lower than that of free Pt after administration of cis-platin (15x10-3min-1) due to a lower CLM and a comparable VI' (Vermorken et al., 1984bElferink et al., submitted). The lower value of CLM is in agreement with the lower values of (a) the in vitro rate constants of plasma protein binding (0.0020 vs. 0.0068min-1 ( ) and (b) degradation of the intact drugs in plasma (0.0040 min-1, this study vs. 0.0077 min-1 (Repta et al., 1980)) compared to cis-platin, respectively. The difference in CLM is also reflected by the difference in tissue binding (% D) as observed in animals (Boven et al., 1985;Van der Vijgh et al., 1983) being 2 to 4 times higher for cis-platin than for JM-40.
As mentioned before, platinum complexes like cis-platin (Vermorken et al., 1984b or carboplatin (Elferink et al., submitted;Harland et al., 1984) did not show a third phase in the free Pt vs. time curves. Probably a third phase of free Pt is also present after administration of cis-platin and carboplatin, but with concentrations of secondary free Pt below the usual limit of detection, due to a lower dose (cis-platin) or a smaller rate constant of protein binding (carboplatin) . Indeed, very recently one study reported a very long terminal phase for free Pt following administration of cis-platin, determined with a very sensitive assay for free Pt (Reece et al., 1986). Half-lives of free Pt were lower after cis-platin than after JM-40. This is principally due to the higher protein binding of cis-platin compared to JM-40, which is also reflected by a lower CUE of platinum after cis-platin than after  Since the urinary excretion of proteins is generally very limited excreted Pt may be regarded as free Pt. Renal clearance of free Pt after administration of JM-40 was similar to that after cis-platin . Therefore, the higher cumulative urinary excretion of free Pt after JM-40 compared to that after cis-platin (CUEO-6h= 24+5%D (Vermorken et al., 1984b), CUEO_24h=28+4%D ) is due to the lower rate of reaction with proteins after  KM of JM-40 was higher than that of carboplatin (diammine( 1, l-cyclobutanedicarboxylato)platinum(II), KM = 1.5 x 10-3min-1 for free Pt) (Elferink et al., submitted), which is in agreement with the difference of their in vitro reaction rates with plasma proteins Harland et al., 1984). This means that the in vivo reactivity of JM-40 is higher than that of carboplatin, although they are structurally closely related. X-ray structure analysis revealed, however, that the geometry around the platinum atom of JM-40 shows more deviations from bond angle ideality than that of carboplatin (Cutbush et al., 1983;Neidle et al., 1980). Furthermore, the cyclobutane group of carboplatin may sterically hinder nucleophilic attack of the platinum atom (Neidle et al., 1980). These facts may explain the higher reactivity of JM-40 compared to carboplatin, and also why, from a pharmacokinetic point of view, JM-40 seems to be more related to cis-platin than to carboplatin (Vermorken et al., 1984bElferink et al., submitted;Harland et al., 1984). Besides, the dose-limiting toxicities of JM-40 (nephro and gastro-intestinal toxicity, (Winograd et al., 1986)) are similar to that of cis-platin, whereas the dose limiting toxicity of carboplatin is myelotoxicity (Joss et al., 1984). Vermorken et al. (1985) indicated a relationship between the nephrotoxic properties of 6 platinum complexes and their stability in aqueous solution. Therefore, the intermediate in vivo reactivity of JM-40 may account for its dose limiting nephrotoxicity (in contrast to carboplatin), but at a much higher dose than cis-platin. JM-40 was not entered into phase II trials because of its toxicity profile. Nevertheless, JM-40 has contributed to a better understanding of possible relationships between chemical, pharmacokinetic and clinical properties of platinum compounds.