Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Prescription costs wikipedia , lookup
Electronic prescribing wikipedia , lookup
Discovery and development of antiandrogens wikipedia , lookup
Adherence (medicine) wikipedia , lookup
Pharmacogenomics wikipedia , lookup
Intravenous therapy wikipedia , lookup
List of comic book drugs wikipedia , lookup
Theralizumab wikipedia , lookup
(CANCER RESEARCH 45, 464-469, January 1985] Pharmacokinetics of Vindesine Given as an Intravenous Bolus and 24-Hour Infusion in Humans1 Takao Ohnuma,2 Larry Norton, Alicja Andrejczuk, and James F. Holland Department oÃ-Neoplastia Diseases, Mount Sinai School of Medicine, New York, New York 10029 ABSTRACT MATERIALS AND METHODS The pharmacokinetics of vindesine was examined after the determination of serum drug levels by radioimmunoassay in patients who received the drug either as an i.v. bolus or a 24-hr VDS was supplied by Dr. R. Dyke of the Lilly Laboratories for Clinical Research, Indianapolis, IN, as a lyophilized powder of 10 mg in 10-ml infusion. After i.v. bolus, vindesine was eliminated from the serum by triphasic decay. The central compartment was approximately 6 times the serum volume. The peak serum level achieved by i.v. bolus was approximately 16 times that achieved by the 24hr infusion. The post-24-hr-infusion serum decay followed biphasic decay. Pharmacokinetic modeling, assuming a prolonged infusion period, resulted in a triphasic decay curve, with an extremely short distribution phase which would not be clinically detectable. This was due to the incorporation of the distribution phase into the infusion period. This explains the experimental data of a biphasic decay curve observed after 24-hr infusion. Pharmacokinetic parameters for the two phases observed after 24-hr infusion were similar to values calculated from i.v. bolus data. The c x t for 24-hr infusion was identical to that after i.v. bolus; theoretically, the c x f appears constant regardless of infusion time. It is concluded that the rate of elimination and/or the c x f, rather than the peak serum level, played a role in the degree of hematological toxicity. Vinca alkaloid vinblastine sulfate. Clinical studies have been carried 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 dose findings were done in 2 schedules, i.v. bolus and 24-hr infusion lid or 0.69 ng/tube diluted with 0.02 M acetate buffer, pH 4.4) and 100 n\ rabbit anti-VLB antiserum (Lot No. 24-245-2-G, Eli Lilly Research administered weekly. Major clinical side effects observed after i.v. bolus and 24-hr infusion were myelosuppression and neurctoxicity. At identical dosage levels, degrees and parameters of toxicity were indistinguishable (14). During our Phase I study, we measured the serum drug concentration by radioimmunoassay and compared the pharmacokinetic parameters of the 2 schedules. Attempts were made to explain clinical toxicity in terms of pharmacokinetic parame ters. 1Supported in part by Contract N01-CM-97275 from the Division of Cancer Treatment, National Cancer Institute, NIH, Bethesda, MD; by USPHS Research Grant 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 CANCER 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 than 100,000//il. All but one patient (Patient 11) (Table 3) received 4 mg/sq m of the drug. Patient 11 received 5 mg/sq m. One patient (Patient 6) (Table 1) developed transient elevation of transaminases after each of 2 courses of i.v. bolus treatment. After administration of VDS, either by i.v. bolus or 24-hr infusion, until assay. Informed consent was obtained from each patient prior to the 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 (Antibod ies, Davis, CA), and 0.0242% methiolate; 100 /il standard or unknown serum (appropriately diluted with the glycine buffer, ¡fnecessary), 100 ^l [3H]VLB solution (Amersham/Searie Corp., Arlington Heights, IL; 0.27 VDS3 is a new synthetic antineoplastic agent derived from the vinWastine amidosulfate; Lilly Compound No. 99094, Eldisine); VLB, vinblastine; VCR, vincristine. Received January 11, 1982: accepted September 27, 1984. 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 given at a 3-week interval. None of the patients had preexisting hepatic or 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 the infusion schedule at 0 and 12 hr during infusion and after the infusion at 5 and 30 min and then 1, 2, 4, 8, 24, 48, and 72 hr) from an arm contralateral to the site of the drug administration. Serum was separated from coagulated blood samples by centrifugation and stored at -75° INTRODUCTION reprints should be addressed. 3 The abbreviations used are: VDS. vindesine sulfate (NSC 245467, CAS Regis try No. 59917-39-4; 23-amino-O'-deacetyldemethoxyvincaleukoblastine, deacetyl- ampuls. The ampuls were refrigerated during storage. Just prior to use, the drug was diluted with bacteriostatic sodium chloride injection solution and was then administered either as a slow i.v. bolus (approximately 1 min) through the side arm of a running infusion, or as a 24-hr infusion Laboratories, Indianapolis, IN: diluted 1:2500 with glycine buffer). The contents of the tubes were mixed gently, and the tubes were capped tightly 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 was continued at room temperature for 30 min with gentle shaking twice. The tubes were centrifuged at 1000 x g for 10 min, and the supernatant was decanted directly into scintillation vials. One-half ml of NCS (Amersham) was added and heated at 50°for 20 min. The material was cooled to room temperature. Ten ml of PCS (Amersham) were added, and the vials were then counted in a liquid scintillation counter (Model LS-355, Beckman Instruments, Palo Alto, CA). Standard solutions were made by diluting VDS in acetate buffer with concentrations ranging from 0.2 to 100 Mg/liter. Control tubes included those containing glycine buffer and [3H]VLB only ("total"), 100 ^l of nonimmune rabbit serum substituted for the antiserum ("blank") and 100 fi\ of glycine buffer in place of the standard or unknown ("bound"). The "blank" cpm were subtracted from those of the "bound" to give the actual bound radioactivity, B. The "total" RESEARCH VOL. 45 JANUARY 1985 464 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1985 American Association for Cancer Research. PHARMACOKINETICS OF VOS IN HUMANS cpm minus the "bound" cpm gave the actual free radioactivity, F. The B/F based on "total," "blank," and "bound" should be close to 1.0. B/F .80 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 assays were run in duplicate and repeated at least twice. All the serum drug values were graphed semilogarithmically and curve fit by the method of nonlinear least-square using the Meeter-Marquardt-Wood algorithm (4). Using serum VDS levels after i.v. bolus, the first-order mass transfer rate constant was calculated on the basis of a "first-pass" 3-compartrnent model (Chart 1) (11,20). This model was chosen because animal studies had indicated that the biliary excretion was the major route of VDS elimination (2). Thus, the second compartment (Chart 1) is assumed to be in equilibrium with the nepatobiliary system, and the elimination occurs principally from the second compartment. The third compartment is unidentified physiologically. RESULTS The VDS standard curve is shown in Chart 2. On the logit-log scale, radioimmunoassay values of VDS were linear in the range of 1.0 to 100 /¿g/liter.VDS is stable in 0.9% sodium chloride or 5% dextrose for more than 24 hr at room temperature.4 The drug was also stable in frozen serum when repeatedly defrosted and measured 2 to 3 times during a 6-month time span. Serum VDS levels after i.v. bolus followed triexponential de cay, as confirmed by the precise curve-fit of equation B: + Be- = Pe- (B) where c,t,(f ) is the serum VDS concentration at time t after bolus drug administration; P, A, and B are constants representing intercepts on the ordinate at time zero; and ir, «,and ßare the first-order disposition constants with ir > «> ß> 0. Pharmaco kinetic parameters obtained from a single i.v. bolus administra tion in 7 patients are shown in Tables 1 and 2. An example of Dot t=o 2k2ik,2D¡koIk|3*k3l3 <20 3-compartment .70 .60 .50 .40 .30 ¡.20 .10 .02 1.0 100.0 10.0 VINDESINE (>ig/liter) 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 0.2 scale on the ordinate. the least-squares curve-fit of Equation B to the data is illustrated in Chart 3. From these data, the initial half-life of serum drug decay was approximately 9 min. The second serum half-life was approximately 4 hr, and the third half-life was approximately 35 hr. Patient 6 had pharmacokinetic parameters essentially within the values obtained from others, except ßof 0.0085 or the third half-life of approximately 80 hr. The volume of distribution of the central compartment (Chart 1, Compartment 1) was calculated to be approximately 18 liters. VDS showed a high value for exit rate constants (k,2 and ki3) from the central compartment, while the elimination rate constant was quite low. This implied that tissue distribution or inactivation rather than elimination was primarily 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 of infusion. A' and B' are constants representing y-axis values at t = 7, and a' and ß'are the first-order disposition rate constants with 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 the peak 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 2k2ik|2'k|3k3l3 Chart 1. 'First-pass" .90 open model used to calculate the phar- macokinetic parameters presented In Tables 1 to 3. Upper scheme, i.v. bolus; lower scheme, i.v infusion. 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 infusion was 88% of the value within 15 min after the end of infusion in 4 patients. Using this observation plus the shape of the decay curve, with the absence of the ir phase, an overall serum VDS curve may be estimated which can then be used to approximate c x i. Theoretically (20), for an infusion of length 7, the serum concentration cs/(f) for f > 7 is given by: P(7)e- 4 R. L. Nelson, personal communication. CANCER + A(T)e- ' + B(T)e- RESEARCH VOL. 45 JANUARY 1985 465 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1985 American Association for Cancer Research. (D) PHARMACOKINETICS OF VOS IN HUMANS where P(T), A(T), and B(T) are pharmacokinetic which are functions of the infusion time 7": "constants" tO CO CO ^~ O N« 8 iß SSS (E) •(« - 7r)(ir - 0) CM CO O»<D (D ^ ^- CM O CM IO CÑI r^i ö^" öS cj cöW (F) «(0- «)(« ifo [(£2- |8)(^31 O *- O »- CM * i- f~ O r- m V, \ ß(a- (Dose)/T and r- The pharmacokinetic values /fzo, ta, ta, T, a, and ßfrom Tables 1 and 2 were calculated, not from the "true" bolus 00 administration but, rather, from a T = Vfeo-hrinfusion. Therefore, it is preferred to calculate P(T), A(T), and B(T) for all 7 > Veotir from P, A, and B as follows: in ÃŽ3 »- i-cg^cocnoor*; P(7) = P x |P(7-)/P('/6o)|= ÌÌ 2 (1 - e-4437) 559 1- (1 -- e -°1757) \A(T)IA(V«,)\ = — A(T) = AX 8(7) = S x jB(7)/B(Veo)| = 422 — -2 (1 - in OD <O CNJ O CNJ p p p p pp (G) O en evi c\j co «-;co p o o o ö o ö ö CO CO CO 00 i- C\J OJ *- -1) - ir) p 7) (H) (I) (J) o o ö ö ö ö ö CQ : The values of P(24), /A(24). and 8(24), calculated directly from Equations H, I, and J, were 4.23, 22.9 and 6.68, respectively. Of note is the observation that the value of P(24) = 4.23 is suffi ciently small, and the decay e~*<' ~ 24)sufficiently rapid, so as to ooo <o "Õ . . . tg. *~.*R to ID cn ro ^ CNJco reduce the impact of the w phase on the serum concentration of VDS after termination of the infusion. It is also of confirmatory importance that calculated -4(24) = 22.9 and 3(24) = 6.68 are close and proportional to the measured values ¿'= 26.4 ±3.32 and B' = 7.5 ± 1.06 from the infusion (7 = 24) data (from SaSKSts ° ^Cgt^r^CVIi-;^ +1 o o ööo o ö in Equation C) as summarized in Table 3. The measured values are probably higher than the estimates by virtue of the incorporation of the "hidden" P(24). In this regard, the calculated value of C.X24) ö CD CNiinoipoioicn oöinoSr-'cxjco^ h-min^rininco T- OO CM i- â„¢ "H ^ or P(24) + A(24) + 8(24) = 33.8 corresponds nicely with the measured value of 33.9 (Table 3, Cs,(f = 7)). At time 7, by reference to Equation D, the heights of compo nents 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, the serum concentration should equal P(t ) + A(t) + B(t), where these values are calculated from Equations K, L, and M below, which are modifications of Equations H, I, and J. IO IO tO pcncnooincxicn a' IA tf> v- (Q CVÃŽ O o> r^ p Å“ O) in TT CO CO LO CM Öl <+1 P(t) N. A(t) 101.6 (1 - e-44») 559.1 (K) (L) O CD CO CO (D CM CM <D U) (0 (O N K N B(t) = I- o o o ooo (M) o These equations are valid because the total dose administered by time f is —=r—-so that »-CMco^-incoi^ +| Dose-f CANCER RESEARCH VOL. 45 JANUARY Dose 1985 466 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1985 American Association for Cancer Research. PHARMACOKINETICS OF VOS IN HUMANS Table 2 First-order mass transfer rate constant (see Chan 1) to lit serum vindesine levels after i.v. bo/us (hr1)0.6016.761.451.543.644.054.923.28 Patient1234567Mean*« (hr1)0.02580.05260.03770.02470.03340.01350.04720.0336 (liters/hr/kg)0.1200.2790.0980.1940.06170.1280.04490.1 ±0.83"*2,(hr1)0.8440.7440.3210.2351.620.8131.010.799 32 ±0.03 + 0.161ft* 54 ±0.0255Clearance ±0.00513*M(hr1)0.2240.1710.08210.1220.2600.0920.1240.1 ±0.174ft«(hr1)0.09711.310.2720.1570.1410.2170.3190.359 "Mean±S.E. This concept allows for the simulation of the serum concentration both during the infusion and after f = T. Õ If 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 for T = Veoand 24 hr, respectively. Confirming the validity of this model is the calculation that the ratio cs,(12)/cs/(f) varies between 0.84 and 0.95 for values of / between 0 and 15 min after the end of 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) and the bolus data (see Table 1, footnotes). It is to be noted that the c x f for the 24-hr infusion was identical to that for bolus 24 48 56 72 HOURS AFTER START OF INFUSION administration. Chart 3. Serum clearance of VDS after 4 mg/sq m i.v. "bolus" (V«o-hr infusion) as measured by radioimmunoassay (Patient 4). The ordinate is expressed as a natural logarithm of VDS concentration in >ig/titer. Each point represents a mean of the assay run in duplicate and repeated at least twice. The line represents computer-generatednonlinear least-squares to fit Equation B. 7 e 5 4 3 2 DISCUSSION The radioimmunoassay used in the present study is a sensitive one, 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 the biological material. In this context, what we expressed as VDS should more accurately be called VDS equivalents. We have used the term VDS in this treatise in order to avoid confusion when one compares the data produced from different institu tions; most of the earlier Vinca alkaloid pharmacology data using radioimmunoassay were derived by the use of a method devel oped by Root ef a/. (17). Nelson ef a/. (12, 13) and Owellen ef al. (16) reported pharmacokinetic parameters of VDS in humans obtained by similar radioimmunoassay. They have shown that the central compart ment was compatible with the total blood volume. In contrast, the central compartment obtained in this report is approximately 4-fold greater. The reason for this discrepancy is unclear. In order to avoid extravasation of the drug, we administered it in a slow i.v. bolus (over approximately 1 min) through the side arm of running infusion. This might have caused a low P + A + B value (by Equations K, L, and M) and, consequently, a high V,. On the other hand, we noticed that our V, value for VDS is similar to the reported V, for VCR (12) and VLB (12,15). Our V2 and V3 values are in accord with those reported by Owellen ef al. (16). The mean terminal serum half-lives in our study are greater than the reported values (12, 16) but clearly less than 2 e e 24 4e 66 72 HOURS AFTER START OF INFUSION Chart 4. Postinfusion serum clearance of VOS, 5 mg/sq m given as a 24-hr infusion i.v. (Patient 11). The ordinate is expressed as a natural logarithm of VDS concentration in ¿ig/liter. Eachpoint represents a mean of the assay run in duplicate and repeated at least twice. The line represents computer-generated nonlinear least-squares to fit Equation C. CANCER RESEARCH VOL. 45 JANUARY 1985 467 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1985 American Association for Cancer Research. PHARMACOKINETICS OF VOS IN HUMANS Table 3 Pharmacokinetic parameters obtained from a 24-hr infusion of vindesine Biexponentia! serum decay curve followed Equation C. DosePatient7*8 m)4.04.0 (/•g/Kter)19.2 (HO0.268 6.0 9 10 11(mg/sq Mean 4.0 6.08.4 4.0 5.0(mg)7.2 9.0A' 35.0 24.0 27.4 18.6a' 26.4 ±3.32a'c 0.163 0.182 0.157 0.242B' Oig/liter)5.85 (hr1)0.0201 = T) to/liter)"25.0 8.33 5.69 10.1 4.14»' 0.0245 0.00342 0.01937 0.00616c-(f 43.3 29.7 37.5 22.7 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 received a 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 72 HOURS AFTER START OF INFUSION 24 40 56 72 HOURS 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 of VDS concentration in fig/liter. The values of P, A, and B are intercepts of each of the 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 are partially influenced by a value in Patient 6, who had a terminal half-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 patient from the others is the transient elevation of transaminases after each course of VDS administration. He might have had preexist ing subclimcal hepatic impairment. The dose we used (4 mg/sq m) is larger than those reported by Nelson et al. (13); thus, this could possibly be a dose-dependent phenomenon. Owellen et al. used the doses ranging from 1.5 to 4.0 mg/sq m (16), but the number of patients in each dose level is too small to conclude the dose dependence. All the patients we studied had platelets more than 100,000/^1, whereas the reported study did not men tion the platelet value. A possibility of platelet binding as a factor influencing the half-life could not be excluded. Somewhat wide fluctuations of pharmacokinetic parameters from one patient to the other were noted for both bolus and 24hr infusion. We noted that the P value of Patient 1 is smaller than are the P values of the other patients. This patient had a large retroperitoneal sarcoma and a moderate amount of leg edema. CANCER Charte. Simulation of the VDS serum concentration during and after 24-hr infusion based on Equation N. The ordinate is expressed as a natural logarithm of VDS concentration in ng/liter. P. A, and S are the values on the y-axis at t = 24 hr for each of the 3 lines. The small P value may be related to this. The low P value did not influence the c x f value, however. Determinants for phar macokinetic parameters for VDS of possible importance include tumor load and earner protein concentration (5); these and other such factors as nutritional status and concurrent medications might have affected the pharmacokinetic parameters. Pharmacokinetic parameters from postinfusional blood curves have been assessed for other drugs (8). Although a mathematical equation which enables one to determine the parameters iden tical to an i.v. bolus by utilizing the postinfusion curve has been presented (8), the present study clearly indicates the limitation of such an approach because of the likelihood that, after a prolonged infusion, distribution phase constants P and w may not 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 dysfunction on serum VDS pharmacokinetics would be of interest. Neurolog ical toxicity was more insidious and difficult to quantify. Ob viously, pharmacokinetic parameters obtained after the first course of chemotherapy are inadequate to explain the delayed RESEARCH VOL. 45 JANUARY 1985 468 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1985 American Association for Cancer Research. PHARMACOKINETICS OF VOS IN HUMANS vinblastine amide 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-infusion blood 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., and DeLuca, L. PhaseII clinicaltrials with vindesinefor remissioninduction in acute leukemia, blastic crisis of chronic myeloid leukemia, lymphosarcoma and Hodgkin's disease: absenceof cross-resistancewith vincristine.CancerTreat. Rep., 62: 805-809,1978. 11. Nagashima, R., Levy, G., and O'Reilly, R. A. Comparative pharmacokinetics of coumarine anticoagulants IV. Application of a three-compartment model to the analysis of the dose-response kinetics of bishydroxycoumarinelimination. J. Pharm. Sci., 57:1888-1895,1968. 12. Nelson, R. L., Dyke, R. W., and Root, M. A. Comparative pharmacokinetics for vinca alkaloids in man. Proceedings of the 78th Annual Meeting A. Soc. Clin. Pharmacol.Ther., 1977. 13. Nelson, R. L., Dyke, R. W., and Root, M. A. Clinical pharmacokinetics of vindesine. Cancer Chemother. Pharmacol.,2: 243-246,1979. 14. Ohnuma, T., Greenspan, E. M., and Holland, J. F. Initial clinical studies with vindesine: 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 metab olism of vinblastine in humans. Cancer Res., 37: 2597-2602,1977. 16. Owellen, R. J., Root, M. A., and Hains, F. O. Pharmacokineticsof vindesine and vincristine in humans. Cancer Res., 37:2603-2607,1977. 17. Root, M. A., Gerzon, K., and Dyke, R. W. A radioimmunoassayfor vinblastine and 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. Isolation and partial characterization of human urinary metabolites of vincristine sul phate. Proc. Am. Assoc. Cancer Res., 22:173,1981. 19. Valdivieso, M., Richamn.S., Burgess, A. M., Bodey, G. P., and Freinch. E. J. Initial clinical studies of vindesine. Cancer Treat. Rep., 65: 873-875,1981. 20. Wagner, J. G. Fundamentalsof clinical pharmacokinetics, pp 114-117. Ham ilton, 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. Cancer Treat. Rep., 65: 775-779,1981. neurological toxicities. We noted that the comparison of the 2 schedules is a practi cable method to correlate the relationship between pharmacokinetic data and clinical toxicity. Thus, our pharmacokinetic eval uation revealed that the 24-hr infusion of VDS produced only one-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 were essentially similar for the 2 schedules (14), it may be concluded that the toxicity is less related to the peak serum levels than that immunoassayable c x f and/or the elimination half-life. The observation that the c x i for the bolus administration is identical to that after 24-hr infusion should be of interest. This was mathematically confirmed. Theoretically, c x f would be constant, regardless of infusion time. One might be tempted to carry out prolonged infusion in attempts to correlate c x t and biological effects. Implications of these observations on the therapeutic effects remain to be elucidated. In our Phase I study, none of the patients went into partial or complete responses (14). It is noteworthy that remissions were reported to be induced in patients with acute leukemia with a 48-hr infusion, whereas a twice-daily i.v. bolus regimen did not produce a response (10). Similarly, contin uous infusion of VDS for 5 days appeared to be more efficacious than was the bolus schedule in the treatment of patients with refractory breast carcinoma (21). The improved therapeutic effi cacy of VDS infusion over bolus cannot be explained from available pharmacokinetic parameters alone. In this context, the following considerations may be offered. First, it is possible that pharmacokinetic parameters (c x f, half-lives, etc.) from 48-hr infusion or 5-day continuous infusion are entirely different from those 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 biological concentration-effect window (biologically effective drug concen tration x exposure time) rather than mere c x i. Third, Ndesformyl-VCR, an inactive metabolite, was identified in the urine Appendix: Integration of c x ( The c x f for infusion data are defined as the definite integral of c„(f ) (Equation N) 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: of a patient receiving VCR (18). It is possible that such a metabolite may be present in radioimmunoassayable VDS, and intracellular metabolism of VDS may be independent of serum c x f. In-depth pharmacokinetic studies of VDS for both bolus and infusion in patients with VDS-sensitive tumors might give insight into the relationship between pharmacokinetic parame ters and therapeutic efficacy. (O) and Equation N reduces to Õ3 If V«o < f < 24, then £K,(i). 3 M Otherwise, I K,(24)e^" ACKNOWLEDGMENTS (P) ~ r). I We thank Dr. Mary Root of Eli Lilly Research Laboratories, Indianapolis,IN, for the gift of rabbit anti-VLB antibodies. Since JL ÕK.<0*=• f ÕK.(0* = i {z, + (-^)<e-" •slfiO1-1 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 levels of [3H]vindesinein rats. Cancer Res., 37:3053-3056,1977. 3. Currie, V., Wong, O., Krakoff, I. H., and Young, C. W. Phase I trial of vindesine in 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, and colchicine with serum proteins. Biochem. Pharmacol.,22: 2113-2119,1973. 6. Dyke, R. W., and Nelson, R. L. Phase I anticancer agent vindesine (deacetyl CANCER RESEARCH and **) |_1 b-1 I \*l ' / 1) (Q) r!K,(De-«-"*=i^ =Ã]-(^y«' -D •X M i-1 /ti 1-1 I \IC|77 then f- i ** 3 c„(f)cff = £z, «1083, which is independent of T for all T > Veo >-i These results may be confirmed by numerical integration of Equation N (9). VOL. 45 JANUARY 1985 469 Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1985 American Association for Cancer Research. (S) Pharmacokinetics of Vindesine Given as an Intravenous Bolus and 24-Hour Infusion in Humans Takao Ohnuma, Larry Norton, Alicja Andrejczuk, et al. Cancer Res 1985;45:464-469. Updated version E-mail alerts Reprints and Subscriptions Permissions Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/45/1/464 Sign up to receive free email-alerts related to this article or journal. To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at [email protected]. To request permission to re-use all or part of this article, contact the AACR Publications Department at [email protected]. Downloaded from cancerres.aacrjournals.org on June 16, 2017. © 1985 American Association for Cancer Research.