Embed Size (px)
Citation preview
(CANCER RESEARCH 45, 464-469, January 1985]
Pharmacokinetics of Vindesine Given as an Intravenous Bolus and 24-HourInfusion in Humans1
Takao Ohnuma,2 Larry Norton, Alicja Andrejczuk, and James F. Holland
Department oÃNeoplastia Diseases, Mount Sinai School of Medicine, New York, New York 10029
ABSTRACT MATERIALS AND METHODS
The pharmacokinetics of vindesine was examined after thedetermination of serum drug levels by radioimmunoassay inpatients who received the drug either as an i.v. bolus or a 24-hr
infusion. After i.v. bolus, vindesine was eliminated from the serumby triphasic decay. The central compartment was approximately6 times the serum volume. The peak serum level achieved byi.v. bolus was approximately 16 times that achieved by the 24-hr infusion. The post-24-hr-infusion serum decay followed bi-
phasic decay. Pharmacokinetic modeling, assuming a prolongedinfusion period, resulted in a triphasic decay curve, with anextremely short distribution phase which would not be clinicallydetectable. This was due to the incorporation of the distributionphase into the infusion period. This explains the experimentaldata of a biphasic decay curve observed after 24-hr infusion.
Pharmacokinetic parameters for the two phases observed after24-hr infusion were similar to values calculated from i.v. bolusdata. The c x t for 24-hr infusion was identical to that after i.v.
bolus; theoretically, the c x f appears constant regardless ofinfusion time. It is concluded that the rate of elimination and/orthe c x f, rather than the peak serum level, played a role in thedegree of hematological toxicity.
INTRODUCTION
VDS3 is a new synthetic antineoplastic agent derived from the
Vinca alkaloid vinblastine sulfate. Clinical studies have beencarried out in various institutions in the United States and Europe(1,3,6, 7,10,14,19). In our Phase I study of VDS, human dosefindings were done in 2 schedules, i.v. bolus and 24-hr infusion
administered weekly. Major clinical side effects observed afteri.v. bolus and 24-hr infusion were myelosuppression and neurc-
toxicity. At identical dosage levels, degrees and parameters oftoxicity were indistinguishable (14).
During our Phase I study, we measured the serum drugconcentration by radioimmunoassay and compared the phar-
macokinetic parameters of the 2 schedules. Attempts were madeto explain clinical toxicity in terms of pharmacokinetic parameters.
1Supported in part by Contract N01-CM-97275 from the Division of Cancer
Treatment, National Cancer Institute, NIH, Bethesda, MD; by USPHS ResearchGrant CA 15936-03; by the United Leukemia Fund, Inc., New York, NY; and by T.J. Marteli Foundation for Leukemia and Cancer Research, New York, NY.
' Chemotherapy Foundation Senior Clinical Investigator. To whom requests for
reprints should be addressed.3The abbreviations used are: VDS. vindesine sulfate (NSC 245467, CAS Regis
try No. 59917-39-4; 23-amino-O'-deacetyldemethoxyvincaleukoblastine, deacetyl-
vinWastine amidosulfate; Lilly Compound No. 99094, Eldisine); VLB, vinblastine;VCR, vincristine.
Received January 11, 1982: accepted September 27, 1984.
VDS was supplied by Dr. R. Dyke of the Lilly Laboratories for ClinicalResearch, Indianapolis, IN, as a lyophilized powder of 10 mg in 10-ml
ampuls. The ampuls were refrigerated during storage. Just prior to use,the drug was diluted with bacteriostatic sodium chloride injection solutionand was then administered either as a slow i.v. bolus (approximately 1min) through the side arm of a running infusion, or as a 24-hr infusion
after further diluting in one liter of 5% dextrose.Serum VDS levels were measured in a total of 11 patients: 6 after i.v.
bolus, 4 after 24-hr infusion, and 1 after bolus and 24-hr infusions givenat a 3-week interval. None of the patients had preexisting hepatic or
renal impairment, but one (Patient 1) had a moderate third space fluids.All patients had WBC counts more than 4,000//il and platelets more than100,000//il. All but one patient (Patient 11) (Table 3) received 4 mg/sq mof the drug. Patient 11 received 5 mg/sq m. One patient (Patient 6)(Table 1) developed transient elevation of transaminases after each of 2courses of i.v. bolus treatment.
After administration of VDS, either by i.v. bolus or 24-hr infusion,
blood samples were collected at selected time intervals (after bolus at 0,5, 15, and 30 min and then 1, 2, 4, 6, 12, 24, 48, and 72 hr; with theinfusion schedule at 0 and 12 hr during infusion and after the infusion at5 and 30 min and then 1, 2, 4, 8, 24, 48, and 72 hr) from an armcontralateral to the site of the drug administration. Serum was separatedfrom coagulated blood samples by centrifugation and stored at -75°
until assay. Informed consent was obtained from each patient prior tothe study.
VDS serum levels were measured by radioimmunoassay (12,13,16,17). The reaction tubes (12- x 75-mm polystyrene tubes No. 2054;
Falcon Plastics, Oxnard, CA) consisted of 200 n\ of 0.2 M glycine buffer,pH 8.8, which contained 0.25% human serum albumin (Plasmanate,Cutter Laboratories, Berkeley, CA), 1.0% normal sheep serum (Antibodies, Davis, CA), and 0.0242% methiolate; 100 /il standard or unknownserum (appropriately diluted with the glycine buffer, ¡fnecessary), 100 ^l[3H]VLB solution (Amersham/Searie Corp., Arlington Heights, IL; 0.27
lid or 0.69 ng/tube diluted with 0.02 M acetate buffer, pH 4.4) and 100n\ rabbit anti-VLB antiserum (Lot No. 24-245-2-G, Eli Lilly Research
Laboratories, Indianapolis, IN: diluted 1:2500 with glycine buffer). Thecontents of the tubes were mixed gently, and the tubes were cappedtightly and incubated for 4 days at 4°.At the end of the incubation
period, 500 n\ of dextran-coated charcoal suspension [1 % Nor it-A (Sigma
Chemical, St. Louis, MO) and 0.5% dextran 70 (Pharmacia, Upsala,Sweden)] in glycine buffer were added to each tube, and incubation wascontinued at room temperature for 30 min with gentle shaking twice.The tubes were centrifuged at 1000 x g for 10 min, and the supernatantwas decanted directly into scintillation vials. One-half ml of NCS (Amer-sham) was added and heated at 50°for 20 min. The material was cooled
to room temperature. Ten ml of PCS (Amersham) were added, and thevials were then counted in a liquid scintillation counter (Model LS-355,
Beckman Instruments, Palo Alto, CA). Standard solutions were made bydiluting VDS in acetate buffer with concentrations ranging from 0.2 to100 Mg/liter. Control tubes included those containing glycine buffer and[3H]VLB only ("total"), 100 ^l of nonimmune rabbit serum substituted forthe antiserum ("blank") and 100 fi\ of glycine buffer in place of thestandard or unknown ("bound"). The "blank" cpm were subtracted fromthose of the "bound" to give the actual bound radioactivity, B. The "total"
CANCER RESEARCH VOL. 45 JANUARY 1985
464
Research. on February 1, 2021. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from
PHARMACOKINETICS OF VOS IN HUMANS
cpm minus the "bound" cpm gave the actual free radioactivity, F. TheB/F based on "total," "blank," and "bound" should be close to 1.0. B/F
for the standard and unknown was calculated as:
8 _ Sample cpm - "blank" cpm
F "Total" cpm - sample cpm (A)
B/F for the standard curve was plotted on a logit-log scale, and unknown
concentrations were read from the linear portion of the graph. The assayswere run in duplicate and repeated at least twice. All the serum drugvalues were graphed semilogarithmically and curve fit by the method ofnonlinear least-square using the Meeter-Marquardt-Wood algorithm (4).
Using serum VDS levels after i.v. bolus, the first-order mass transferrate constant was calculated on the basis of a "first-pass" 3-compartrnent
model (Chart 1) (11,20). This model was chosen because animal studieshad indicated that the biliary excretion was the major route of VDSelimination (2). Thus, the second compartment (Chart 1) is assumed tobe in equilibrium with the nepatobiliary system, and the elimination occursprincipally from the second compartment. The third compartment isunidentified physiologically.
RESULTS
The VDS standard curve is shown in Chart 2. On the logit-log
scale, radioimmunoassay values of VDS were linear in the rangeof 1.0 to 100 /¿g/liter.VDS is stable in 0.9% sodium chloride or5% dextrose for more than 24 hr at room temperature.4 The
drug was also stable in frozen serum when repeatedly defrostedand measured 2 to 3 times during a 6-month time span.
Serum VDS levels after i.v. bolus followed triexponential decay, as confirmed by the precise curve-fit of equation B:
= Pe- + Be-(B)
where c,t,(f ) is the serum VDS concentration at time t after bolusdrug administration; P, A, and B are constants representingintercepts on the ordinate at time zero; and ir, «,and ßare thefirst-order disposition constants with ir > «> ß> 0. Pharmaco
kinetic parameters obtained from a single i.v. bolus administration in 7 patients are shown in Tables 1 and 2. An example of
Dot t=o
2k2ik|2'k|3k3l3
2k2ik,2D¡koIk|3*k3l3
<20
Chart 1. 'First-pass" 3-compartment open model used to calculate the phar-
macokinetic parameters presented In Tables 1 to 3. Upper scheme, i.v. bolus; lowerscheme, i.v infusion.
.90
.80
.70
.60
.50
.40
.30¡.20
.10
.02
0.2 1.0 10.0VINDESINE (>ig/liter)
100.0
Chart 2. VDS radioimmunoassay standard curve. See "Materials and Methods"for the assay. The "bound" (B) cpm/"free" (F) cpm ratio was plotted on a logit
scale on the ordinate.
the least-squares curve-fit of Equation B to the data is illustratedin Chart 3. From these data, the initial half-life of serum drugdecay was approximately 9 min. The second serum half-life wasapproximately 4 hr, and the third half-life was approximately 35
hr. Patient 6 had pharmacokinetic parameters essentially withinthe values obtained from others, except ßof 0.0085 or the thirdhalf-life of approximately 80 hr. The volume of distribution of the
central compartment (Chart 1, Compartment 1) was calculatedto be approximately 18 liters. VDS showed a high value for exitrate constants (k,2 and ki3) from the central compartment, whilethe elimination rate constant was quite low. This implied thattissue distribution or inactivation rather than elimination wasprimarily responsible for the drug clearance from the serum.
In contrast to i.v. bolus, the serum drug levels after 24-hr
infusion followed biexponential decay, as shown in Chart 4.Least-squares regression fits precisely Equation C:
c„(f)= A'e-°'v- " +(C)
where cs,(f ) is the serum VDS concentration after 24-hr infusion,
7 is the infusion time (24 hr), and f is the time after the start ofinfusion. A' and B' are constants representing y-axis values att = 7, and a' and ß'are the first-order disposition rate constantswith a' > ß'> 0. Pharmacokinetic parameters obtained after
24-hr infusion in 5 patients are shown in Table 3. Comparison of
pharmacokinetic parameters in Tables 1 and 3 shows that thepeak serum VDS concentration after the 24-hr infusion was
approximately 6% of the peak concentration obtained by i.v.bolus. Noting that a and a' as well as ßand ,)" are equivalent
within the range of experimental variability, it is observed that,after a 24-hr infusion of VDS, the ir phase of the serum is not
seen.For the determination of c x t, there were insufficient data
points to allow curve-fitting during the infusion period. However,
the VDS serum level at the 12th hr after the start of the infusionwas 88% of the value within 15 min after the end of infusion in4 patients. Using this observation plus the shape of the decaycurve, with the absence of the ir phase, an overall serum VDScurve may be estimated which can then be used to approximatec x i. Theoretically (20), for an infusion of length 7, the serumconcentration cs/(f) for f > 7 is given by:
4 R. L. Nelson, personal communication. P(7)e- + A(T)e- ' + B(T)e-(D)
CANCER RESEARCH VOL. 45 JANUARY 1985
465
Research. on February 1, 2021. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from
PHARMACOKINETICS OF VOS IN HUMANS
CQ :
I-
tO CO CO ^~ O N«
8 ißS S S
CM CO O»<D (D^ ^- CM O CM IO CÑIr^i ö^" öScj cöW
O *- O i- O r- O
»- CM * f~ m r- 00en evi c\j co «-;co po o o ö o ö ö
CO CO CO 00 in OD <Oi- C\J OJ *- CNJ O CNJ
p p p p pp po o ö ö ö ö ö
ooo tg <o. . . . *~.*R "Õto ID cn ro ^ CNJco
T- OO CM i- IO IO tO
inÎ3i-cg^cocnoor*; »-
ÌÌ2
SaSKSts °^Cgt^r^CVIi-;^ +1
o o ööo o ö in
ö
CD
CNiinoipoioicn ™oöinoSr-'cxjco^ "Hh-min^rininco ^
pcncnooincxicn <-a' IA tf> v- (Q CVÎO +1o> r^ p œ O) inTT CO CO LO CM Öl N.
O CD CO CO (D CM CM<D U) (0 (O N K N
o o o ooo o
»-CMco^-incoi^ +|
where P(T), A(T), and B(T) are pharmacokinetic "constants"which are functions of the infusion time 7":
•(«- 7r)(ir - 0)
«(0- «)(«-
ifo [(£2- |8)(^31
V, \ ß(a-
-1)
- ir)
(E)
(F)
(G)
(Dose)/T and
The pharmacokinetic values /fzo, ta, ta, T, a, and ßfromTables 1 and 2 were calculated, not from the "true" bolus
administration but, rather, from a T = Vfeo-hrinfusion. Therefore,
it is preferred to calculate P(T), A(T), and B(T) for all 7 > Veotirfrom P, A, and B as follows:
P(7) = P x |P(7-)/P('/6o)|= (1 - e-4437) (H)
559 1A(T) = AX \A(T)IA(V«,)\ = — - (1 - e
422 28(7) = S x jB(7)/B(Veo)| = — - (1 -
- -°1757) (I)
7) (J)
The values of P(24), /A(24). and 8(24), calculated directly fromEquations H, I, and J, were 4.23, 22.9 and 6.68, respectively. Ofnote is the observation that the value of P(24) = 4.23 is sufficiently small, and the decay e~*<'~ 24)sufficiently rapid, so as to
reduce the impact of the w phase on the serum concentration ofVDS after termination of the infusion. It is also of confirmatoryimportance that calculated -4(24) = 22.9 and 3(24) = 6.68 areclose and proportional to the measured values ¿'= 26.4 ±3.32and B' = 7.5 ± 1.06 from the infusion (7 = 24) data (from
Equation C) as summarized in Table 3. The measured values areprobably higher than the estimates by virtue of the incorporationof the "hidden" P(24). In this regard, the calculated value of C.X24)
or P(24) + A(24) + 8(24) = 33.8 corresponds nicely with themeasured value of 33.9 (Table 3, Cs,(f = 7)).
At time 7, by reference to Equation D, the heights of components IT,a, and ßare given by P\T), A(T), and B(T), respectively.Thus, at time Veo< t < T during an infusion of length 7, theserum concentration should equal P(t ) + A(t) + B(t), where thesevalues are calculated from Equations K, L, and M below, whichare modifications of Equations H, I, and J.
P(t)
A(t)
101.6
559.1
(1 - e-44»)
B(t) =
(K)
(L)
(M)
These equations are valid because the total dose administered
by time f is —=r—-so that
Dose-f Dose
CANCER RESEARCH VOL. 45 JANUARY 1985
466
Research. on February 1, 2021. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from
PHARMACOKINETICS OF VOS IN HUMANS
Table 2First-order mass transfer rate constant (see Chan 1) to lit serum vindesine levels after i.v. bo/us
Patient1234567Mean*«(hr1)0.6016.761.451.543.644.054.923.28
±0.83"*2,(hr1)0.8440.7440.3210.2351.620.8131.010.799±0.174ft«(hr1)0.09711.310.2720.1570.1410.2170.3190.359+ 0.161ft*
(hr1)0.02580.05260.03770.02470.03340.01350.04720.0336
±0.00513*M(hr1)0.2240.1710.08210.1220.2600.0920.1240.154 ±0.0255Clearance
(liters/hr/kg)0.1200.2790.0980.1940.06170.1280.04490.1
32 ±0.03
"Mean±S.E.
This concept allows for the simulation of the serum concentrationboth during the infusion and after f = T.
24HOURS
48AFTER START
56 72OF INFUSION
Chart 3. Serum clearance of VDS after 4 mg/sq m i.v. "bolus" (V«o-hrinfusion)
as measured by radioimmunoassay (Patient 4). The ordinate is expressed as anatural logarithm of VDS concentration in >ig/titer.Each point represents a meanof the assay run in duplicate and repeated at least twice. The line representscomputer-generatednonlinear least-squaresto fit Equation B.
7
e
5
4
3
2 2
e
e 24 4e 66 72HOURS AFTER START OF INFUSION
Chart 4. Postinfusion serum clearance of VOS, 5 mg/sq m given as a 24-hrinfusion i.v. (Patient 11). The ordinate is expressed as a natural logarithm of VDSconcentration in ¿ig/liter.Eachpoint representsa meanof the assay run in duplicateand repeated at least twice. The line represents computer-generated nonlinearleast-squaresto fit Equation C.
ÕIf Veo< t < T, then P(f) + A(t) + B(t).10therwise, PfTJe-*' " n + ^(TJe-^' ~ n + S(T)e,-«' - (N)
Simulations of Equation N are illustrated in Charts 5 and 6 forT = Veoand 24 hr, respectively. Confirming the validity of thismodel is the calculation that the ratio cs,(12)/cs/(f) varies between0.84 and 0.95 for values of / between 0 and 15 min after the endof the infusion at T = 24. This corresponds with the measured
value of 0.88.The c x f for the infusion data can now be estimated by
integrating Equation N (see "Appendix"). This has been per
formed for the mean infusion data (see Table 3, footnotes) andthe bolus data (see Table 1, footnotes). It is to be noted that thec x f for the 24-hr infusion was identical to that for bolus
administration.
DISCUSSION
The radioimmunoassay used in the present study is a sensitiveone, but it also cross-reacts with other Vinca alkaloids (12,16).This implies that the method might cross-react not only with the
parent VDS but probably with some metabolite(s) present in thebiological material. In this context, what we expressed as VDSshould more accurately be called VDS equivalents. We haveused the term VDS in this treatise in order to avoid confusionwhen one compares the data produced from different institutions; most of the earlier Vinca alkaloid pharmacology data usingradioimmunoassay were derived by the use of a method developed by Root ef a/. (17).
Nelson ef a/. (12, 13) and Owellen ef al. (16) reported phar-
macokinetic parameters of VDS in humans obtained by similarradioimmunoassay. They have shown that the central compartment was compatible with the total blood volume. In contrast,the central compartment obtained in this report is approximately4-fold greater. The reason for this discrepancy is unclear. Inorder to avoid extravasation of the drug, we administered it in aslow i.v. bolus (over approximately 1 min) through the side armof running infusion. This might have caused a low P + A + Bvalue (by Equations K, L, and M) and, consequently, a high V,.On the other hand, we noticed that our V, value for VDS issimilar to the reported V, for VCR (12) and VLB (12,15). Our V2and V3 values are in accord with those reported by Owellen efal. (16). The mean terminal serum half-lives in our study are
greater than the reported values (12, 16) but clearly less than
CANCER RESEARCH VOL. 45 JANUARY 1985
467
Research. on February 1, 2021. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from
PHARMACOKINETICS OF VOS IN HUMANS
Table 3Pharmacokinetic parameters obtained from a 24-hr infusion of vindesine
Biexponentia! serum decay curve followed Equation C.
DosePatient7*8
91011(mg/sq
m)4.04.0
4.04.05.0(mg)7.2
6.06.08.4
9.0A'
(/•g/Kter)19.2
35.024.027.418.6a'
(HO0.268
0.1630.1820.1570.242B'
Oig/liter)5.85
8.335.69
10.14.14»'
(hr1)0.0201
0.02450.003420.019370.00616c-(f
= T)to/liter)"25.0
43.329.737.522.7
Mean 26.4 ±3.32a'c 0.202 ±0.025 7.50 ±1.06" 0.0147 ±0.00417 33.9 ±4.06e
* Patient 7 in Table 1 was also studied for 24-hr infusion." Mean ±S.E.c Data from Patient 11, who receiveda different dose from the others, were not used for calculation.
The c x r. estimated by integration of Equation N (see "Appendix"), is 1083 (pg/hr/liter).
8 24 43 56 72HOURS AFTER START OF INFUSION
Chart 5. Simulation of the VOS serum concentration after i.v. "bolus" (V«o-hr
infusion) based on Equation N. The ordinate is expressed as a natural logarithm ofVDS concentration in fig/liter. The values of P, A, and B are intercepts of each ofthe 3 straight lines to the x-axis.
those reported for VCR. The discrepancy of the half-life obtained
from this study and the earlier report is not clear. Our data arepartially influenced by a value in Patient 6, who had a terminalhalf-life of 80 hr.
The reasons why Patient 6 had a prolonged terminal half-life
are unclear. The only difference in clinical course of this patientfrom the others is the transient elevation of transaminases aftereach course of VDS administration. He might have had preexisting subclimcal hepatic impairment. The dose we used (4 mg/sqm) is larger than those reported by Nelson et al. (13); thus, thiscould possibly be a dose-dependent phenomenon. Owellen et
al. used the doses ranging from 1.5 to 4.0 mg/sq m (16), but thenumber of patients in each dose level is too small to concludethe dose dependence. All the patients we studied had plateletsmore than 100,000/^1, whereas the reported study did not mention the platelet value. A possibility of platelet binding as a factorinfluencing the half-life could not be excluded.
Somewhat wide fluctuations of pharmacokinetic parametersfrom one patient to the other were noted for both bolus and 24-
hr infusion. We noted that the P value of Patient 1 is smaller thanare the P values of the other patients. This patient had a largeretroperitoneal sarcoma and a moderate amount of leg edema.
24 40 56 72HOURS AFTER START OF INFUSION
Charte. Simulation of the VDS serum concentration during and after 24-hrinfusion based on Equation N. The ordinate is expressed as a natural logarithm ofVDS concentration in ng/liter. P. A, and S are the values on the y-axis at t = 24 hrfor each of the 3 lines.
The small P value may be related to this. The low P value didnot influence the c x f value, however. Determinants for pharmacokinetic parameters for VDS of possible importance includetumor load and earner protein concentration (5); these and othersuch factors as nutritional status and concurrent medicationsmight have affected the pharmacokinetic parameters.
Pharmacokinetic parameters from postinfusional blood curveshave been assessed for other drugs (8). Although a mathematicalequation which enables one to determine the parameters identical to an i.v. bolus by utilizing the postinfusion curve has beenpresented (8), the present study clearly indicates the limitationof such an approach because of the likelihood that, after aprolonged infusion, distribution phase constants P and w maynot be recognizable from experimental data.
Increased toxicity in patients with major hepatic impairment(14) and increased terminal half-life in Patient 6 suggested that
the liver plays a major role in VDS excretion or metabolism.Determination of the influence of preexisting hepatic dysfunctionon serum VDS pharmacokinetics would be of interest. Neurological toxicity was more insidious and difficult to quantify. Obviously, pharmacokinetic parameters obtained after the firstcourse of chemotherapy are inadequate to explain the delayed
CANCER RESEARCH VOL. 45 JANUARY 1985
468
Research. on February 1, 2021. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from
PHARMACOKINETICS OF VOS IN HUMANS
neurological toxicities.We noted that the comparison of the 2 schedules is a practi
cable method to correlate the relationship between pharmaco-
kinetic data and clinical toxicity. Thus, our pharmacokinetic evaluation revealed that the 24-hr infusion of VDS produced onlyone-sixteenth of the peak serum level achieved by i.v. bolus,whereas the c x t value after 24-hr infusion was identical to that
after i.v. bolus. Since the observed hematological toxicities wereessentially similar for the 2 schedules (14), it may be concludedthat the toxicity is less related to the peak serum levels than thatimmunoassayable c x f and/or the elimination half-life.
The observation that the c x i for the bolus administration isidentical to that after 24-hr infusion should be of interest. This
was mathematically confirmed. Theoretically, c x f would beconstant, regardless of infusion time. One might be tempted tocarry out prolonged infusion in attempts to correlate c x t andbiological effects.
Implications of these observations on the therapeutic effectsremain to be elucidated. In our Phase I study, none of the patientswent into partial or complete responses (14). It is noteworthythat remissions were reported to be induced in patients withacute leukemia with a 48-hr infusion, whereas a twice-daily i.v.
bolus regimen did not produce a response (10). Similarly, continuous infusion of VDS for 5 days appeared to be more efficaciousthan was the bolus schedule in the treatment of patients withrefractory breast carcinoma (21). The improved therapeutic efficacy of VDS infusion over bolus cannot be explained fromavailable pharmacokinetic parameters alone. In this context, thefollowing considerations may be offered. First, it is possible thatpharmacokinetic parameters (c x f, half-lives, etc.) from 48-hrinfusion or 5-day continuous infusion are entirely different fromthose of 24-hr infusion. This possibility is, however, unlikely
because c x t appears constant regardless of infusion time.Second, therapeutic efficacy may be related to a biologicalconcentration-effect window (biologically effective drug concentration x exposure time) rather than mere c x i. Third, N-desformyl-VCR, an inactive metabolite, was identified in the urine
of a patient receiving VCR (18). It is possible that such ametabolite may be present in radioimmunoassayable VDS, andintracellular metabolism of VDS may be independent of serumc x f. In-depth pharmacokinetic studies of VDS for both bolusand infusion in patients with VDS-sensitive tumors might give
insight into the relationship between pharmacokinetic parameters and therapeutic efficacy.
ACKNOWLEDGMENTS
We thank Dr. Mary Root of Eli Lilly Research Laboratories, Indianapolis,IN, forthe gift of rabbit anti-VLB antibodies.
REFERENCES
1. Blum, R. H., and Dawson, D. M. Vindesine: Phase I study of a Vinca alkaloid.Proc. Am. Assoc. Cancer Res., 17:108, 1976.
2. Gulp, H. W.. Daniels,W. D., and McMahon, R. E. Dispositionand tissue levelsof [3H]vindesinein rats. Cancer Res., 37:3053-3056,1977.
3. Currie, V., Wong, O., Krakoff, I. H., and Young, C. W. PhaseI trial of vindesinein patients with advanced cancer. Cancer Treat. Rep.. 62: 1333-1336,1978.
4. Daniel,C., and Wood, F. S. Fitting Equationsto Data, pp. 320-333. New York:Wuey-lntersoence, 1971.
5. Domgian,D. W., and Owellen, R. T. Interaction of vinblastine, vincristine, andcolchicine with serum proteins. Biochem. Pharmacol.,22: 2113-2119,1973.
6. Dyke, R. W., and Nelson, R. L. Phase I anticancer agent vindesine (deacetyl
vinblastineamide sulfate).Cancer Treat. Rev., 4:135-142,1977.7. Gralla, R. J., Tan, C. T. C., and Young, C. W. Vindesine,a review of phase II
trials. Cancer Chemother. Pharmacol.,2: 271-274,1979.8. Loo, J. C. K., and Riegelman,S. Assessment of pharmacokinetic constants
from post-infusionblood curves obtained after i.v. infusion. J. Pharm.Sci., 59:53-55,1970.
9. Lyness, J. N. Simpson quadrature used adaptively (noise killed), SQUANK.Commun. Assoc. Computing Machinery, 73: 260-263,1970.
10. Mathe, G., Misset, J. L., DeVassal,F., Goureia, J., Hayat, M., Machover, D.,Belpomme, D., Pico, J. L., Schwartzenberg, L., Ribaud, P., Jasmin, U., andDeLuca,L. PhaseIIclinicaltrials with vindesinefor remissioninduction in acuteleukemia, blastic crisis of chronic myeloid leukemia, lymphosarcoma andHodgkin's disease:absenceof cross-resistancewith vincristine.CancerTreat.Rep., 62: 805-809,1978.
11. Nagashima, R., Levy, G., and O'Reilly, R. A. Comparative pharmacokineticsof coumarine anticoagulants IV. Application of a three-compartment model tothe analysisof the dose-responsekinetics of bishydroxycoumarinelimination.J. Pharm.Sci., 57:1888-1895,1968.
12. Nelson, R. L., Dyke, R. W., and Root, M. A. Comparative pharmacokineticsfor vinca alkaloids in man. Proceedingsof the 78th Annual Meeting A. Soc.Clin. Pharmacol.Ther., 1977.
13. Nelson, R. L., Dyke, R. W., and Root, M. A. Clinical pharmacokinetics ofvindesine.Cancer Chemother. Pharmacol.,2: 243-246,1979.
14. Ohnuma, T., Greenspan, E. M., and Holland, J. F. Initial clinical studies withvindesine: tolerance to weekly i.v. bolus and 24-hr infusion. Cancer Treat.Rep., 64. 25-30,1980.
15. Owellen, R. J., Hartke, C. A., and Hains, F. 0. Pharmacokineticsand metabolism of vinblastine in humans.Cancer Res., 37: 2597-2602,1977.
16. Owellen, R. J., Root, M. A., and Hains, F. O. Pharmacokineticsof vindesineand vincristine in humans.Cancer Res., 37:2603-2607,1977.
17. Root, M. A., Gerzon, K., and Dyke, R. W. A radioimmunoassayfor vinblastineand vincristine.Federationof AnalyticalChemistryand SpectroscopySocieties.Abstract 183, p. 125, October, 1975.
18. Sethi, V. S., Castle, M. C., Surratt, P.,Jackson, D. V., and Spurr, C. L. Isolationand partial characterization of human urinary metabolites of vincristine sulphate. Proc. Am. Assoc. Cancer Res., 22:173,1981.
19. Valdivieso,M., Richamn.S., Burgess, A. M., Bodey, G. P., and Freinch.E. J.Initial clinicalstudies of vindesine.CancerTreat. Rep., 65: 873-875,1981.
20. Wagner, J. G. Fundamentalsof clinical pharmacokinetics,pp 114-117. Hamilton, IL: Drug IntelligencePublications,1975.
21. Yap, H. Y., Blumenschein,G. R., Bodey, G. P., Hortobagyi, G. N., Buzdar, A.U, and DiStefano, A. Vindesine in the treatment of refractory breast cancer:improvement in the therapeutic index with continuous 5-day infusion. CancerTreat. Rep., 65: 775-779,1981.
Appendix: Integration of c x (
The c x f for infusion data are defined as the definite integral of c„(f) (EquationN) over the interval from zero to infinity. Let P(t) = K,(i), A(t) = KJ(t), and B(f) =K¿t),where a, = T, «2= a, and «3= ß.Specify z-,= 101.6, Z2= 559.1, and z3 =422.4. Then, Equations K, L, and M may be written:
and Equation N reduces to
Õ3If V«o< f < 24, then £K,(i).
3 M
Otherwise, I K,(24)e^" ~ r).
(O)
(P)
I
Since
and
then
JL ÕK.<0*=•f ÕK.(0* = i {z, + (-^)<e-" -•slfiO1-1 **) |_1 b-1 I \*l ' /
1) (Q)
r!K,(De-«--"*=i^ =Ã]-(^y«'-D•X M i-1 /ti 1-1 I \IC|77
f- 3
i c„(f)cff= £z, «1083, which is independent of T for all T > Veo (S)** >-i
These results may be confirmed by numerical integration of Equation N (9).
CANCER RESEARCH VOL. 45 JANUARY 1985
469
Research. on February 1, 2021. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from
1985;45:464-469. Cancer Res Takao Ohnuma, Larry Norton, Alicja Andrejczuk, et al. and 24-Hour Infusion in HumansPharmacokinetics of Vindesine Given as an Intravenous Bolus
Updated version
http://cancerres.aacrjournals.org/content/45/1/464
Access the most recent version of this article at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
Subscriptions
Reprints and
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Permissions
Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)
.http://cancerres.aacrjournals.org/content/45/1/464To request permission to re-use all or part of this article, use this link
Research. on February 1, 2021. © 1985 American Association for Cancercancerres.aacrjournals.org Downloaded from
