- Email: [email protected]
Chronic effects of fractionated renal irradiation on the pharmacokinetics of intravenous methotrexate
0360-3016/87 $3.00 + .OO Copyright Q 1987 Pergamon Journals Ltd.
In:. J. Radiolion Oncology Bid. Phys.. Vol. 13, PP. 759-764 Printed in the U.S.A. All righIs reserved.
??Original Contribution
CHRONIC EFFECTS OF FRACTIONATED RENAL IRRADIATION ON THE PHARMACOKINETICS OF INTRAVENOUS METHOTREXATE JOHNS.HOLCENBERG, M.D.,’ JoHNE.MouLDER,PH.D.,~~' M.D., PH.D.,~ MARK D. KRAILO,PH.D.,~ BRIAN L. FISH, B.S.,2
BARTON A. KARMEN,
BARBARA J. RING, B.S.’ AND SUSAN ADAMS, B.S.’ ’Pediatrics and Biochemistry, University of Southern California, Los Angeles, CA; 2Radiation Oncology Medical College
Wisconsin, Milwaukee, WI; 3Pediatrics, University of Texas Health Sciences Center, Dallas, TX; 4Preventive Medicine, University of Southern California; and ‘Pharmacology, Medical College Wisconsin The chronic effects of renal irradiation on the pharmacology of methotrexate was studied in a rat model. Unanesthetized rats received 2 doses of bilateral fractionated kidney irradiation (16.2 Gy or 19.8 Gy in 9 fractions). Alterations in renal function were first seen at 3 months in the 19.8 Gy group and 12 months in the 16.2 Gy groups. Life table analysis showed a shift in the survival curve of about 3 months between the 2 radintion doses. The pharmacokinetics of i.v. methotrexate showed an increase in the area under the plasma curve beghming at 9 months in the 19.8 Gy group and at 15 months in the 16.2 Gy group. The volume of distribution of methotrexate was smaller in the irradiited rats than in unirradiated controls. Multiple linear regression models showed significant correlations between parameters of methotrexate &arance and certain renal function tests. Nevertheless, no set of renaI function tests consistently predicted alteration in methotrexate clearance in the 2 radiition groups. Furthermore, time after irradiition remnined a highly significant variable indicating that renal irradiation causes time dependent change in methotrexate pharmacokinetics that can not be accounted for by the usual tests of renal function. Renal irradiation, Methotrexate, Pharmacokinetics. INTRODUCTION
ment of many cancers. A preliminary report of this work was presented at a symposium.‘*
Certain antitumor agents have potentiated local reactions to concurrent radiation therapy.” Much less is known about late interactions between these therapeutic modalilities.’ Rubin16 postulated that both radiation therapy and chemotherapy deplete parencbymal stem cell populations while radiation also affects microvasculature and connective tissue. Thereby, either therapeutic modality may cause late potentation of the toxic effects of the other. In addition to local toxic effects, radiation therapy could alter drug absorption, distribution, metabolism, or excretion. The kidney is one of the most radiosensitive of the critical organs and it is the major excretory path for many anticancer drugs.” In this paper we report the effects of the development of chronic radiation nephritis in a rat model on the pharmacokinetics and excretion of methotrexate. Methotrexate was studied because it is eliminated by glomerular filtration and renal tubular secretion, it has a small margin of safety and it is used in combination with radiation therapy in the treat-
METHODS
AND MATERIALS
Female rats from a WAG/RijMCW breeding colony of defined microbiologic floral3 were irradiated, treated and followed throughout their life span in a moderatesecurity barrier facility.* The animals were 2 months old at the start of each of three experiments. Rats were randomly assigned to sham irradiation, 16.2 Gy or 19.8 Gy. Rats were irradiated without anesthesia, using 250 kV X rays (HVL of 0.5 mm Cu) and Plexiglass jigs adapted from those used by Sheldon et al. ‘* Parallel-opposed lateral fields encompassed both kidneys with a 5 mm margin. Sham-irradiated animals were placed in the jigs for the same time as the irradiated animals. The irradiation was delivered at 1.5 Gy/min in 9 equal fractions over 11 days. Two additional control groups were followed: (a) unirradiated age matched controls that were not placed in the jig, who received one dose of methotrexate at 15
Reprint requests to: Dr. John S. Holcenberg, Division of Hematology-Oncology, Childrens Hospital of Los Angeles, P.O. Box 54700, Los Angeles, CA 90054-0700. Acknowledgement-This work was supported by National
Cancer Institute Grants CA24652 and CA34840 and the T. J. Martell Foundation for Leukemia and Cancer Research. Accepted for publication 8 December 1986. 759
760
I. J. Radiation Oncology 0 Biology 0 Physics
months; (b) age matched controls that received 19.8 Gy and one dose of methotrexateat 15 months after the irradiation. In the fkst experiment, 16 rats received 19.8 Gy, 8 were sham-irradiated, and 8 were age matched controls. In the second experiment, 9 rats received 19.8 Gy, 16 received 16.2 Gy, 9 were sham-irradiated and 7 were age matched controls. In the third experiment, 8 rats received 19.8 Gy, 10 received 16.2 Gy, 7 were sham-irradiated and 5 were age matched controls. Selected animals received tail vein injections of methotrexate before irradiation and at 1, 3,6, 9, 12, 15, and 18 months after irradiation. The irradiated and shamirradiated animals received no more than seven injections of methotrexate (mean f SD is 5.6 + 1.2). These animals had methotrexate measured in either the plasma or urine on alternate occasions. Blood samples were obtained from a tail vein or the retro-orbital plexus before and at 5 to 10 minutes, 0.5, 1, 1S, 3, 6, and 24 hours after the injection. Urine was collected in a metabolic cage between O-4,4-24 and 24-48 hours after an injection. The animals were monitored every l-3 months for physical appearance, weight, hematocrit, white blood count, blood urea nitrogen (BUN), serum creatinine, plasma protein and r-glutamyl-transpeptidase, blood glucose and urinary creatinine, urine volume per 24 hours, and urine N-acetyl-glucosarm ine (NAG). Urine and serum creatinine, BUN and protein were determined by commercial kits or a clinical analyzer. NAG was assayed as descrkd4 Methotrexate was assayed in urine and serum using a radioligand assay.” Intravenous injections were graded by the amount that extravasatated. Only injections with > 90% of the dose injected in the vein were used for analysis. The methotrexate plasma data were fitted to the following equation by the nonlinear least-squares analysis of Autoan-Nonlin with a weighting factor of l/C”: C = Ae-“’+ Be-” + Gee7’.
Area under the serum curve (AUC) was calculated from the parameters:
The data was also analyzed by noncompartmental methods. l4V, is the volume of distribution at steady state and MRT is the mean resident time of methotrexate in the serum, where: V,=Totaldose--*
AUMC MRT=s. AUC: ’
The total area under the serum concentration curve (AUC,) and the first moment ofthe serum concentration
May 1987, Volume 13, Number 5
time curve (AUMC) was calculated by the trapazoidal rule extrapolating the last sampling time to infinity. The survival of each animal was taken to be the time from irradiation or sham-irradiation to death. Animals who were lolled were considered censored at that time; in all other cases an event was considered to have occurred. The survival experience of the treatment groups was compared using the Savage scores test.’ Life tables were constructed using the method of Kaplan and Meier.’ Weight, renal function tests and parameters of methotrexate (MTX) pharmacokinetics were compared with the treatment groups by Student’s t-test at times of MTX administration. The value of various measures of renal function for predicting the parameters of MTX pharmacokinetics was assessed using multiple linear regression analysis.’ Best subset regression using Mallow’s C statistic as selection criterion was used to determine the minimal number of renal function tests required to predict the various MTX parameters5 For the analysis, only the last MTX determination was considered. Separate models were fitted for each treatment group because preliminary analysis indicated that a single method did not adequately describe these data. RESULTS
The two radiation schedules produced chronic nephritis with alternations in renal function beginning at three months in the 19.8 Gy group and at 12 months in the 16.2 Gy group. Death was first seen in the 19.8 Gy group at nine months after irradiation and in the 16.2 Gy group at 12 months. A life table analysis shows a shift in the survival curve of about 3 months between the high and low radiation groups (Fig. I). These survival curves show a significant difference (p = .0038) between each radiation dose and the sham-irradiated control. There was no significant difference between the sham-irradiated and the age-matched controls. At death or sacrifice, tumors were noted in 6 of 26 of the 16.2 Gy group, 5 of 30 of the 19.8 Gy group and 0 of 44 of the control groups. Blood tests showed elevation in serum creatinine and blood urea nitrogen beginning at 10 months after irradiation in the 19.8 Gy group, but not in the 16.2 Gy group (Fig. 1). No consistent changes were seen in hematocrit, white blood count, protein, glucose, or y-glutamyl transpeptidase. The rats irradiated with 19.8 Gy gained significantly less weight than the sham-irradiated controls (p < .O1 at 6 months and later). The weight of rats in the 16.2 Gy group became significantly less than the control group at 15 months. The results of urine tests are shown in Figure 2. Beginning at 3 months after radiation, the 19.8 Gy group showed elevation in median urine protein concentration and mean NAG activity per mg creatinine. The mean urinary creatinine concentration showed a progressive
761
Renal irradiation and methotrexate 0 J. S. HOLCENBERG etal.
and by nonlinear least squares fit are significantly greater than the sham-irradiated group (p < .05 by Student ttest). There was no significant difference between the 2 irradiation groups. Figure 4 shows the changes with time in the mean areas under the curves by the non linear least squares fit and the trapazoidal rule and in the volume of distribution by noncompartmental analysis. Significant differences are noted by asterisks. The AUC by both methods of calculation is larger in the 19.8 Gy than the sham irradiated rats beginning at 9 months after radiation. In the 16.2 Gy group, AUC becomes larger than sham-irradiated animals at 15 and 18 months. Because the total body clearance is equal to the dose of methotrexate divided by the AUC, these larger values for AUC in the irradiated rats indicate that methotrexate is cleared at a slower rate in these animals. There is more variation in the volume of distribution but the radiated groups tend to have smaller values than control animals at 15 and 18 months following radiation.
100
ae _ 90 3 .> 80 $
70
*
60 100
F
80
E” 60 “;
40
m
20 0
5
1.8
E
1.4
5
1.0
E 3 z 0
200 150 100
.6
50
.2
0 0
3
6
9
12
15
18
Months Fig. 1. Effects of Renal Irradiation. A. Survival by Life Table Analysis., B. Bun, C. Serum Creatinine. Symbols are 0 ?? , 0, 16.2 Gy; and A A, 19.8 Gy. Sham-irradiated; 0 -
20 15 10 5 0
decrease after 4 months; the mean urine volume per 24
800
hours began to increase after 10 months in this group. The product of these two measurements, the urinary creatinine excreted per 24 hr, did not differ from controls when normalized to body weight. The group of rats irradiated with 16.2 Gy showed less effect on nrinary protein excretion and delayed alterations in NAG/mg creatinine, urinary creatinine concentration and volume. The control animals showed no changes in these parameters. There were no consistent changes in urine pH and no ketones, blood or glucose in the urine of irradiated or unirradiated rats. Figure 3 shows the plasma disappearance curves for methotrexate after intravenous injections 15 months after sham-irradiated, 16.2 Gy and 19.8 Gy groups. The mean f SD. (number) for areas under the curves by trapazoidalruleare30.6+ 11.9(9),44.3+ 15.0(8),and59.1 t- 18.9(8) rM.hr, respectively. The mean f SD (number) areas under the curves by a polynomial least squares fit are 38.9 f 13.3(9), 58.0 f 15.3(8), and 78.5 + 23.9(8) PM - hr, respectively. The areas under the curve for the 16.2 Gy and 19.8 Gy groups by both the trapazoidal rule
600 400 200 0
3
6
9 12 Months
15
18
Fig. 2. Changes in Urine Measurements with Time After Radiation. Values are means of all measurements for sham-irradiated plus age matched control (O), 16.6 Gy (0) and 19.8 Gy (A) animals except for urine protein concentrations which are median values.
762
I. J. Radiation Oncology 0 Biology 0 Physics
May 1987, Volume 13, Number 5
ation jig. There were no significant differences in any of the parameters. SimiIarly, there were no significant differences in the mean fraction of methotrexate excreted in the urine between 11 sham-irradiated rats and 7 age-matched controls at 15 months. Analysis of the averaged data showed that total body methotrexate clearance is delayed in rats irradiated with 16.2 or 19.8 Gy when renal dysfunction appears. Temporally, alternations in urine NAG to creatinine ratios and protein excretion proceeded the changes in methotrexate clearance. Increases in urine volume and de-
100 5 I a -
I
I
1~1
I
80
8 I z”
60
f
40
a
k
20
10
15
25
Hours
Fig. 3. Methotrexate Concentration in Plasma after Intravenous Injection of 10 mg/kg Sham-irradiated (O), 16.6 Gy (0) and 19.8 Gy (A) Groups. Values are the means of 8- 10 rats per point.
In the 19.8 Gy group at 12 and 15 months, there was a tendency towards higher zero time intercepts B and G and the sum of these intercepts by the nonlinear least squares analyses. The differences from control were significant for B at 12 months and for G and the sum of A + B + G at 15 months. The exponent p and y and the mean resident time did not differ among the groups. There was not sufficient data to accurately determine cy. The mean values for the percentage of the injected dose of methotrexate that appeared in the urine did not differ significantly among the irradiated or control animals at any of the times following radiation. The mean t SD. for all animals are 35.7 + 16.3, 40.0 + 15.6, and 40.7 f 15.4% (N = 160) for O-4,0-24, and O-48 hr after the injection, respectively. The plasma methotrexate concentrations and renal function of 9 sham irradiated and 8 age-matched control rats were compared at 15 months following sham-irradiation. The age matched controls received only 1 injection of methotrexate and had not been placed in the radi-
2
60
; +
40
4 a
20
IL
1.2 g -I
0.8
> : 0.4 0 3
6
9
12
15
18
Months Fig. 4. Effect of Renal Radiation on Methotrexate Pharmacokinetics. Area under the curve was calculated by a nonlinear least squares method (AUC) 0--41or trapazoidal rule (AUCJ. Volume of distribution was calculated by a non compartmerital analysis. (See Methods). Symbols are same as in Figure 1. Values that are significantly different from the control rats are shown by asterisks.
763
Renal irradiation and methotrexate 0 J.S.HOLCENBERG etnl.
creases in urinary creatinine concentration correlated with the methotrexate clearance for both irradiation doses. BUN and serum creatinine abnormalities were seen only in the 19.8 Gy group. Plots of AUC for methotrexate vs urine creatinine concentration show that the 2 radiation groups segregate significantly from the sham-irradiated group at urine concentrations < 100 mg/dl. (p c .O1 for the 19.8 Gy group and p < .05 for the 16.2 Gy group by Student ttest.) The AUC for 19.8 Gy group was significantly different from the controls at BUN values > 25 mg/dl. (p < .Ol.)
The clinically important question is whether any of the commonly used tests for renal function correlate with the pharmacokinetics of intravenously administered methotrexate after renal irradiation. To answer this question we analyzed these parameters with multiple linear regression models. Plots of the parameters of renal function and methotrexate pharamacokinetics for the last observation in each rat were very similar to those shown for the whole data set in Figures 1,2, and 4. Table 1 shows the significant correlations between the 3 major parameters of methotrexate plasma pharmacokinetics, AU& Vd/kg, and MRT, and the renal function tests for the last observations. Arrows indicate the direction of the correlations with significance by a two-sided Student ttest for the 16.2 Gy and 19.8 Gy groups. This table shows that significant correlations exist between the parameters of methotrexate pharmacokinetics and renal function. Nevertheless, no set of renal function parameters consistently predicts alteration in the 2 radiation groups or in the 3 methotrexate parameters. Furthermore, time after irradiation is a highly significant factor in 5 of the 6 analyses indicating that the renal radiation causes time dependent changes in methotrexate pharamacokinetics that cannot be accounted for by the functional parameters measured. Similar analyses were performed with the sham-inzliated group, with all groups combined and with the irradiated groups combined. None of these analyses showed a consistent pattern of tests that correlated with methotrexate pharmacokinetics. Although the mean percentage of the dose of methotrexate excreted in the urine did not differ among the groups of animals, significant correlations were seen in these analyses. Methotrexate excretion correlated with urine protein concentration in the 16.2 Gy group and with urine volume plus BUN in the 19.8 Gy group. DISCUSSION These experiments have shown that renal irradiation can affect the clearance of methotrexate, confirming current hypotheses on the consequences of late organ damage.‘*16Doses of radiation were selected that reproducibly produce chronic nephritis.‘* These doses are similar to those received by patients in some combined modality
Table 1. Correlations of methotrexate pharmacokinetics renal function tests for last observation AUC
Time Weight U Cr/ml U Vo1/24 hr U Cr/24 hr/ 1OOgm CrCl/ 1OOgm SCr BUN U Protein U NAG/Cr
t
t t
-
1
_
-
-
t
_ 1
-
T T t
1
-
1
19.8 GY
t
-
-
_
-
_
1 1
16.2 GY
1 -
-
-
19.8 GY
1
-
-
-
-
1 1
1 t
1
1
17 .72
-
-
1 18 .60
17 .40
AL_-_-__
UPH
S Protein N R2
16.2 GY
19.8 GY
1
MRT
Vd/Kg
16.2 GY
with
-
-
.lg87
.lg73
.::
Note: Correlations are shown with values > 1.70 by student t-test. Arrows indicate direction of the correlation. Dash indicates no significant change. N is the number of rats analyzed. R2 is the squared multiple correlation of the model indicated by the arrows. AUC is the area under the serum curve, Vd is the volume of distribution of steady state per kg of body weight and MRT is the mean resident time of methotrexate in the seNrn.
treatment, such as, that for Wilm’s tumor, lower hemibody irradiation, and total body irradiation with bone marrow reconstitution. Methotrexate represents a class of antitumor agents that is excreted largely unchanged in the urine by glomular filtration and tubular excretion. Alterations in renal function did not correlate closely in time with changes in methotrexate pharmacokinetics. The first changes in renal function (increased urine protein concentration and NAG/cr) occurred at 3 months after irradiation; changes in methotrexate clearance were first seen at 9 months. Furthermore, serum creatinine and BUN increased at about the same time as the changes in methotrexate in the 19.8 Gy group but did not change in the rats receiving 16.2 Gy. Temporally, urine creatinine concentration and urine volume ap peared to correlate with the mean area under the methotrexate plasma concentrations and its volume of distribution. Analyses undertaken to determine whether a single or combination of renal functional measures correlated with the parameters of methotrexate pharmacokinetics showed that different combinations of factors significantly correlated for each radiation dose. Furthermore, time after irradiation remained a highly significant factor in most of the analyses indicating that the irradiation was producing effects that were not measured by our tests. This is not surprising because a major change was noted in the volume of distribution of methotrexate in the irra-
764
I.
J. RadiationOncology0 Biology0 Physics
diated animals. This change could be caused by alteration in renal blood flow, damage to the intestine in the radiation field or changes in fluid balance from effects on renin, angiotensin or other renal hormones6 The administered methotrexate dose of 10 mg/kg (approximately 30 mg/m’) produced no toxicity in any of the rats. Much higher doses (> 1,000 mg/m2) are administered in some clinical protocols. The duration of methotrexate concentrations > lo-* M appear to be most related to toxicity.2 At these doses, alterations in creatinine clearance and liver function (SGPT) are associated with delayed clearance and higher blood levels3 Methotrexate toxicity might be expected when higher doses are given to animals or patients with radiation nephritis. The interactions can be very complex when the chemotherapeutic agent also produces renal damage.‘We
May 1987,Volume 13,Number5 have recently observed increased c&platinum toxicity after renal radiation and increased chronic radiation nephritis after c&-platinum treatment.12 What tests of renal function should be monitored in patients treated with methotrexate following radiation to the kidneys? The results of our rat experiments suggest that both measurements of glomular 6ltration and tubular function are important. Thus, creatinine clearance, BUN, urine protein and urine concentrating ability should be determined. Recently Hot-i et al.’ has shown that measurement of tubular function by phenolsulfonphtbalein greatly aids in predictions of correct doses of an antibiotic in patients with renaI failure. This more direct assessment of tubular function should be considered in patients receiving drugs that are secreted by the tubules after renal irradiation.
REFERENCES 1. Brown, J.M.: Drug or radiation changes to the host which
2. 3.
4.
5. 6.
7. 8.
9.
could a&ct the outcome of combined modality therapy. Znt.J. Radiat. Oncol. Biol. Phys. 5: 1151-I 163, 1979. Chabner, B.A. Methotrexate: In PharmacoZogicPrinciples of Cancer Treatment, Chabner, B. (Ed.). Philadelphia, W.B. Saunders. 1982, pp. 229-225. Crom, W.R., Glynn, A.M., Abramowitch, M., Pui, C-H., Dodge, R., Evans, W.E.: Use of automatic interaction detector method to identify patient characteristics t&ted to methotrexate clearance. CZin Pharmacol. Ther. 39: 592597,1986. Dance, N., Price, R.G., Robings, D., Stirling, J.L.: Betagalactosidase, beta-glucosidam, and N-acetyl-betaglucosamine in human kidney. Clin. Chem. Acta 24: 189-197, 1969. Daniel, C., Wood, F.: Fitting Equations to Data. New York, John Wiley and Sons. 1980, pp. 86. Greenberger, J.S., Weichselbaum, R.R., Cassady, J.R.: Radiation Nephritis in Cancer and the Kidney, Reiselbach, R.E., andGarnick,M.B. (Ed.). Philadelphia, LeaandFebiger. 1982, pp. 8 14-823. Hori, R., Okumura, K., Kamiya, A., Nihara, H., Nakano, H.: Ampicillin and cephalexin in renal insufficency. Clin. Pharmacol. Ther. 34: 792-798,1983. Jonas, A.M., Besch, E.L., Brennan, P.C., Foster, H.L., Knapka, J.J., Kruckenberg, SM., Serrano, L.J.: Longterm holding of laboratory rodents. Inst. Lab. Animal Res. News 29: l-26,1976. Kalbfleisch, J.D., Prentice, R.L.: The Statistical Analysis of Failure Time Data. NW York, John Wiley and Sons. 1980, pp. 10-19.
10. Kamen, B.A., Takach, P.L., Vatev, R., Caston, J.D.: A rapid, radiochemical-ligand binding assay for methotmxate. Anal. Biochem. 70: Y-63,1976. 11. Major, P., Kufe, D., Frei, E.: Role of the kidney in the pharmacokinetics of anticancer agents. In Cancer and the Kidney Rieselbach, R.E. and Gamick, M.B. (Ed.). Philadelphia, Lee and Febiger. 1982, pp. 263-274. 12. Moulder, J.E., Holcenberg, J.S., Kamen, B.A., Cheng, M., Fish, B.L.: Renal irradiation and the phamtacology and toxicity of methotrexate and c&platinum. Znt. J. Radiat. Oncol. Biol. Phys. 12: 1415-1418, 1986. 13. Moulder, J.E., Martin, D.F.: Hypoxic fraction determinations in the BA 1112 rat sarcoma: Variations within and among assay techniques. Radiat. Res. 98: 536-548,1984. 14. Perrier, D., Mayersohn, M.: Non-compartmental determination of the steady-state volume of distribution for any mode of administration. J. Pharm. Sci. 71: 372-373,1983. 15. Phillips, T.L., Fu, K.K.: Quantification ofcombined radiation therapy and chemotherapy effects on critical normal tissues cancer. 37: 1186- 1200,1976. 16. Rubin, P.: Late effects of chemotherapy and radiation therapy: A new hypotheses. Znt. J. Radiat. Oncol. Biol. Phys. lo: 5-34,1984. 17. Sedman, A.J., Wagner, J.G.: Auto an a discussion-making pharmacokinetic computer program. Ann Arbor, M.I. Publication Distribution Service, 1976. 18. Sheldon, P.W., Hill, S.A., and Moulder, J.E.: Radioprotection by pentobarbitone sodium of a murine tumor in vivo. Int. J. Radiat. Biol. 32: 571-575, 1977.
In:. J. Radiolion Oncology Bid. Phys.. Vol. 13, PP. 759-764 Printed in the U.S.A. All righIs reserved.
??Original Contribution
CHRONIC EFFECTS OF FRACTIONATED RENAL IRRADIATION ON THE PHARMACOKINETICS OF INTRAVENOUS METHOTREXATE JOHNS.HOLCENBERG, M.D.,’ JoHNE.MouLDER,PH.D.,~~' M.D., PH.D.,~ MARK D. KRAILO,PH.D.,~ BRIAN L. FISH, B.S.,2
BARTON A. KARMEN,
BARBARA J. RING, B.S.’ AND SUSAN ADAMS, B.S.’ ’Pediatrics and Biochemistry, University of Southern California, Los Angeles, CA; 2Radiation Oncology Medical College
Wisconsin, Milwaukee, WI; 3Pediatrics, University of Texas Health Sciences Center, Dallas, TX; 4Preventive Medicine, University of Southern California; and ‘Pharmacology, Medical College Wisconsin The chronic effects of renal irradiation on the pharmacology of methotrexate was studied in a rat model. Unanesthetized rats received 2 doses of bilateral fractionated kidney irradiation (16.2 Gy or 19.8 Gy in 9 fractions). Alterations in renal function were first seen at 3 months in the 19.8 Gy group and 12 months in the 16.2 Gy groups. Life table analysis showed a shift in the survival curve of about 3 months between the 2 radintion doses. The pharmacokinetics of i.v. methotrexate showed an increase in the area under the plasma curve beghming at 9 months in the 19.8 Gy group and at 15 months in the 16.2 Gy group. The volume of distribution of methotrexate was smaller in the irradiited rats than in unirradiated controls. Multiple linear regression models showed significant correlations between parameters of methotrexate &arance and certain renal function tests. Nevertheless, no set of renaI function tests consistently predicted alteration in methotrexate clearance in the 2 radiition groups. Furthermore, time after irradiition remnined a highly significant variable indicating that renal irradiation causes time dependent change in methotrexate pharmacokinetics that can not be accounted for by the usual tests of renal function. Renal irradiation, Methotrexate, Pharmacokinetics. INTRODUCTION
ment of many cancers. A preliminary report of this work was presented at a symposium.‘*
Certain antitumor agents have potentiated local reactions to concurrent radiation therapy.” Much less is known about late interactions between these therapeutic modalilities.’ Rubin16 postulated that both radiation therapy and chemotherapy deplete parencbymal stem cell populations while radiation also affects microvasculature and connective tissue. Thereby, either therapeutic modality may cause late potentation of the toxic effects of the other. In addition to local toxic effects, radiation therapy could alter drug absorption, distribution, metabolism, or excretion. The kidney is one of the most radiosensitive of the critical organs and it is the major excretory path for many anticancer drugs.” In this paper we report the effects of the development of chronic radiation nephritis in a rat model on the pharmacokinetics and excretion of methotrexate. Methotrexate was studied because it is eliminated by glomerular filtration and renal tubular secretion, it has a small margin of safety and it is used in combination with radiation therapy in the treat-
METHODS
AND MATERIALS
Female rats from a WAG/RijMCW breeding colony of defined microbiologic floral3 were irradiated, treated and followed throughout their life span in a moderatesecurity barrier facility.* The animals were 2 months old at the start of each of three experiments. Rats were randomly assigned to sham irradiation, 16.2 Gy or 19.8 Gy. Rats were irradiated without anesthesia, using 250 kV X rays (HVL of 0.5 mm Cu) and Plexiglass jigs adapted from those used by Sheldon et al. ‘* Parallel-opposed lateral fields encompassed both kidneys with a 5 mm margin. Sham-irradiated animals were placed in the jigs for the same time as the irradiated animals. The irradiation was delivered at 1.5 Gy/min in 9 equal fractions over 11 days. Two additional control groups were followed: (a) unirradiated age matched controls that were not placed in the jig, who received one dose of methotrexate at 15
Reprint requests to: Dr. John S. Holcenberg, Division of Hematology-Oncology, Childrens Hospital of Los Angeles, P.O. Box 54700, Los Angeles, CA 90054-0700. Acknowledgement-This work was supported by National
Cancer Institute Grants CA24652 and CA34840 and the T. J. Martell Foundation for Leukemia and Cancer Research. Accepted for publication 8 December 1986. 759
760
I. J. Radiation Oncology 0 Biology 0 Physics
months; (b) age matched controls that received 19.8 Gy and one dose of methotrexateat 15 months after the irradiation. In the fkst experiment, 16 rats received 19.8 Gy, 8 were sham-irradiated, and 8 were age matched controls. In the second experiment, 9 rats received 19.8 Gy, 16 received 16.2 Gy, 9 were sham-irradiated and 7 were age matched controls. In the third experiment, 8 rats received 19.8 Gy, 10 received 16.2 Gy, 7 were sham-irradiated and 5 were age matched controls. Selected animals received tail vein injections of methotrexate before irradiation and at 1, 3,6, 9, 12, 15, and 18 months after irradiation. The irradiated and shamirradiated animals received no more than seven injections of methotrexate (mean f SD is 5.6 + 1.2). These animals had methotrexate measured in either the plasma or urine on alternate occasions. Blood samples were obtained from a tail vein or the retro-orbital plexus before and at 5 to 10 minutes, 0.5, 1, 1S, 3, 6, and 24 hours after the injection. Urine was collected in a metabolic cage between O-4,4-24 and 24-48 hours after an injection. The animals were monitored every l-3 months for physical appearance, weight, hematocrit, white blood count, blood urea nitrogen (BUN), serum creatinine, plasma protein and r-glutamyl-transpeptidase, blood glucose and urinary creatinine, urine volume per 24 hours, and urine N-acetyl-glucosarm ine (NAG). Urine and serum creatinine, BUN and protein were determined by commercial kits or a clinical analyzer. NAG was assayed as descrkd4 Methotrexate was assayed in urine and serum using a radioligand assay.” Intravenous injections were graded by the amount that extravasatated. Only injections with > 90% of the dose injected in the vein were used for analysis. The methotrexate plasma data were fitted to the following equation by the nonlinear least-squares analysis of Autoan-Nonlin with a weighting factor of l/C”: C = Ae-“’+ Be-” + Gee7’.
Area under the serum curve (AUC) was calculated from the parameters:
The data was also analyzed by noncompartmental methods. l4V, is the volume of distribution at steady state and MRT is the mean resident time of methotrexate in the serum, where: V,=Totaldose--*
AUMC MRT=s. AUC: ’
The total area under the serum concentration curve (AUC,) and the first moment ofthe serum concentration
May 1987, Volume 13, Number 5
time curve (AUMC) was calculated by the trapazoidal rule extrapolating the last sampling time to infinity. The survival of each animal was taken to be the time from irradiation or sham-irradiation to death. Animals who were lolled were considered censored at that time; in all other cases an event was considered to have occurred. The survival experience of the treatment groups was compared using the Savage scores test.’ Life tables were constructed using the method of Kaplan and Meier.’ Weight, renal function tests and parameters of methotrexate (MTX) pharmacokinetics were compared with the treatment groups by Student’s t-test at times of MTX administration. The value of various measures of renal function for predicting the parameters of MTX pharmacokinetics was assessed using multiple linear regression analysis.’ Best subset regression using Mallow’s C statistic as selection criterion was used to determine the minimal number of renal function tests required to predict the various MTX parameters5 For the analysis, only the last MTX determination was considered. Separate models were fitted for each treatment group because preliminary analysis indicated that a single method did not adequately describe these data. RESULTS
The two radiation schedules produced chronic nephritis with alternations in renal function beginning at three months in the 19.8 Gy group and at 12 months in the 16.2 Gy group. Death was first seen in the 19.8 Gy group at nine months after irradiation and in the 16.2 Gy group at 12 months. A life table analysis shows a shift in the survival curve of about 3 months between the high and low radiation groups (Fig. I). These survival curves show a significant difference (p = .0038) between each radiation dose and the sham-irradiated control. There was no significant difference between the sham-irradiated and the age-matched controls. At death or sacrifice, tumors were noted in 6 of 26 of the 16.2 Gy group, 5 of 30 of the 19.8 Gy group and 0 of 44 of the control groups. Blood tests showed elevation in serum creatinine and blood urea nitrogen beginning at 10 months after irradiation in the 19.8 Gy group, but not in the 16.2 Gy group (Fig. 1). No consistent changes were seen in hematocrit, white blood count, protein, glucose, or y-glutamyl transpeptidase. The rats irradiated with 19.8 Gy gained significantly less weight than the sham-irradiated controls (p < .O1 at 6 months and later). The weight of rats in the 16.2 Gy group became significantly less than the control group at 15 months. The results of urine tests are shown in Figure 2. Beginning at 3 months after radiation, the 19.8 Gy group showed elevation in median urine protein concentration and mean NAG activity per mg creatinine. The mean urinary creatinine concentration showed a progressive
761
Renal irradiation and methotrexate 0 J. S. HOLCENBERG etal.
and by nonlinear least squares fit are significantly greater than the sham-irradiated group (p < .05 by Student ttest). There was no significant difference between the 2 irradiation groups. Figure 4 shows the changes with time in the mean areas under the curves by the non linear least squares fit and the trapazoidal rule and in the volume of distribution by noncompartmental analysis. Significant differences are noted by asterisks. The AUC by both methods of calculation is larger in the 19.8 Gy than the sham irradiated rats beginning at 9 months after radiation. In the 16.2 Gy group, AUC becomes larger than sham-irradiated animals at 15 and 18 months. Because the total body clearance is equal to the dose of methotrexate divided by the AUC, these larger values for AUC in the irradiated rats indicate that methotrexate is cleared at a slower rate in these animals. There is more variation in the volume of distribution but the radiated groups tend to have smaller values than control animals at 15 and 18 months following radiation.
100
ae _ 90 3 .> 80 $
70
*
60 100
F
80
E” 60 “;
40
m
20 0
5
1.8
E
1.4
5
1.0
E 3 z 0
200 150 100
.6
50
.2
0 0
3
6
9
12
15
18
Months Fig. 1. Effects of Renal Irradiation. A. Survival by Life Table Analysis., B. Bun, C. Serum Creatinine. Symbols are 0 ?? , 0, 16.2 Gy; and A A, 19.8 Gy. Sham-irradiated; 0 -
20 15 10 5 0
decrease after 4 months; the mean urine volume per 24
800
hours began to increase after 10 months in this group. The product of these two measurements, the urinary creatinine excreted per 24 hr, did not differ from controls when normalized to body weight. The group of rats irradiated with 16.2 Gy showed less effect on nrinary protein excretion and delayed alterations in NAG/mg creatinine, urinary creatinine concentration and volume. The control animals showed no changes in these parameters. There were no consistent changes in urine pH and no ketones, blood or glucose in the urine of irradiated or unirradiated rats. Figure 3 shows the plasma disappearance curves for methotrexate after intravenous injections 15 months after sham-irradiated, 16.2 Gy and 19.8 Gy groups. The mean f SD. (number) for areas under the curves by trapazoidalruleare30.6+ 11.9(9),44.3+ 15.0(8),and59.1 t- 18.9(8) rM.hr, respectively. The mean f SD (number) areas under the curves by a polynomial least squares fit are 38.9 f 13.3(9), 58.0 f 15.3(8), and 78.5 + 23.9(8) PM - hr, respectively. The areas under the curve for the 16.2 Gy and 19.8 Gy groups by both the trapazoidal rule
600 400 200 0
3
6
9 12 Months
15
18
Fig. 2. Changes in Urine Measurements with Time After Radiation. Values are means of all measurements for sham-irradiated plus age matched control (O), 16.6 Gy (0) and 19.8 Gy (A) animals except for urine protein concentrations which are median values.
762
I. J. Radiation Oncology 0 Biology 0 Physics
May 1987, Volume 13, Number 5
ation jig. There were no significant differences in any of the parameters. SimiIarly, there were no significant differences in the mean fraction of methotrexate excreted in the urine between 11 sham-irradiated rats and 7 age-matched controls at 15 months. Analysis of the averaged data showed that total body methotrexate clearance is delayed in rats irradiated with 16.2 or 19.8 Gy when renal dysfunction appears. Temporally, alternations in urine NAG to creatinine ratios and protein excretion proceeded the changes in methotrexate clearance. Increases in urine volume and de-
100 5 I a -
I
I
1~1
I
80
8 I z”
60
f
40
a
k
20
10
15
25
Hours
Fig. 3. Methotrexate Concentration in Plasma after Intravenous Injection of 10 mg/kg Sham-irradiated (O), 16.6 Gy (0) and 19.8 Gy (A) Groups. Values are the means of 8- 10 rats per point.
In the 19.8 Gy group at 12 and 15 months, there was a tendency towards higher zero time intercepts B and G and the sum of these intercepts by the nonlinear least squares analyses. The differences from control were significant for B at 12 months and for G and the sum of A + B + G at 15 months. The exponent p and y and the mean resident time did not differ among the groups. There was not sufficient data to accurately determine cy. The mean values for the percentage of the injected dose of methotrexate that appeared in the urine did not differ significantly among the irradiated or control animals at any of the times following radiation. The mean t SD. for all animals are 35.7 + 16.3, 40.0 + 15.6, and 40.7 f 15.4% (N = 160) for O-4,0-24, and O-48 hr after the injection, respectively. The plasma methotrexate concentrations and renal function of 9 sham irradiated and 8 age-matched control rats were compared at 15 months following sham-irradiation. The age matched controls received only 1 injection of methotrexate and had not been placed in the radi-
2
60
; +
40
4 a
20
IL
1.2 g -I
0.8
> : 0.4 0 3
6
9
12
15
18
Months Fig. 4. Effect of Renal Radiation on Methotrexate Pharmacokinetics. Area under the curve was calculated by a nonlinear least squares method (AUC) 0--41or trapazoidal rule (AUCJ. Volume of distribution was calculated by a non compartmerital analysis. (See Methods). Symbols are same as in Figure 1. Values that are significantly different from the control rats are shown by asterisks.
763
Renal irradiation and methotrexate 0 J.S.HOLCENBERG etnl.
creases in urinary creatinine concentration correlated with the methotrexate clearance for both irradiation doses. BUN and serum creatinine abnormalities were seen only in the 19.8 Gy group. Plots of AUC for methotrexate vs urine creatinine concentration show that the 2 radiation groups segregate significantly from the sham-irradiated group at urine concentrations < 100 mg/dl. (p c .O1 for the 19.8 Gy group and p < .05 for the 16.2 Gy group by Student ttest.) The AUC for 19.8 Gy group was significantly different from the controls at BUN values > 25 mg/dl. (p < .Ol.)
The clinically important question is whether any of the commonly used tests for renal function correlate with the pharmacokinetics of intravenously administered methotrexate after renal irradiation. To answer this question we analyzed these parameters with multiple linear regression models. Plots of the parameters of renal function and methotrexate pharamacokinetics for the last observation in each rat were very similar to those shown for the whole data set in Figures 1,2, and 4. Table 1 shows the significant correlations between the 3 major parameters of methotrexate plasma pharmacokinetics, AU& Vd/kg, and MRT, and the renal function tests for the last observations. Arrows indicate the direction of the correlations with significance by a two-sided Student ttest for the 16.2 Gy and 19.8 Gy groups. This table shows that significant correlations exist between the parameters of methotrexate pharmacokinetics and renal function. Nevertheless, no set of renal function parameters consistently predicts alteration in the 2 radiation groups or in the 3 methotrexate parameters. Furthermore, time after irradiation is a highly significant factor in 5 of the 6 analyses indicating that the renal radiation causes time dependent changes in methotrexate pharamacokinetics that cannot be accounted for by the functional parameters measured. Similar analyses were performed with the sham-inzliated group, with all groups combined and with the irradiated groups combined. None of these analyses showed a consistent pattern of tests that correlated with methotrexate pharmacokinetics. Although the mean percentage of the dose of methotrexate excreted in the urine did not differ among the groups of animals, significant correlations were seen in these analyses. Methotrexate excretion correlated with urine protein concentration in the 16.2 Gy group and with urine volume plus BUN in the 19.8 Gy group. DISCUSSION These experiments have shown that renal irradiation can affect the clearance of methotrexate, confirming current hypotheses on the consequences of late organ damage.‘*16Doses of radiation were selected that reproducibly produce chronic nephritis.‘* These doses are similar to those received by patients in some combined modality
Table 1. Correlations of methotrexate pharmacokinetics renal function tests for last observation AUC
Time Weight U Cr/ml U Vo1/24 hr U Cr/24 hr/ 1OOgm CrCl/ 1OOgm SCr BUN U Protein U NAG/Cr
t
t t
-
1
_
-
-
t
_ 1
-
T T t
1
-
1
19.8 GY
t
-
-
_
-
_
1 1
16.2 GY
1 -
-
-
19.8 GY
1
-
-
-
-
1 1
1 t
1
1
17 .72
-
-
1 18 .60
17 .40
AL_-_-__
UPH
S Protein N R2
16.2 GY
19.8 GY
1
MRT
Vd/Kg
16.2 GY
with
-
-
.lg87
.lg73
.::
Note: Correlations are shown with values > 1.70 by student t-test. Arrows indicate direction of the correlation. Dash indicates no significant change. N is the number of rats analyzed. R2 is the squared multiple correlation of the model indicated by the arrows. AUC is the area under the serum curve, Vd is the volume of distribution of steady state per kg of body weight and MRT is the mean resident time of methotrexate in the seNrn.
treatment, such as, that for Wilm’s tumor, lower hemibody irradiation, and total body irradiation with bone marrow reconstitution. Methotrexate represents a class of antitumor agents that is excreted largely unchanged in the urine by glomular filtration and tubular excretion. Alterations in renal function did not correlate closely in time with changes in methotrexate pharmacokinetics. The first changes in renal function (increased urine protein concentration and NAG/cr) occurred at 3 months after irradiation; changes in methotrexate clearance were first seen at 9 months. Furthermore, serum creatinine and BUN increased at about the same time as the changes in methotrexate in the 19.8 Gy group but did not change in the rats receiving 16.2 Gy. Temporally, urine creatinine concentration and urine volume ap peared to correlate with the mean area under the methotrexate plasma concentrations and its volume of distribution. Analyses undertaken to determine whether a single or combination of renal functional measures correlated with the parameters of methotrexate pharmacokinetics showed that different combinations of factors significantly correlated for each radiation dose. Furthermore, time after irradiation remained a highly significant factor in most of the analyses indicating that the irradiation was producing effects that were not measured by our tests. This is not surprising because a major change was noted in the volume of distribution of methotrexate in the irra-
764
I.
J. RadiationOncology0 Biology0 Physics
diated animals. This change could be caused by alteration in renal blood flow, damage to the intestine in the radiation field or changes in fluid balance from effects on renin, angiotensin or other renal hormones6 The administered methotrexate dose of 10 mg/kg (approximately 30 mg/m’) produced no toxicity in any of the rats. Much higher doses (> 1,000 mg/m2) are administered in some clinical protocols. The duration of methotrexate concentrations > lo-* M appear to be most related to toxicity.2 At these doses, alterations in creatinine clearance and liver function (SGPT) are associated with delayed clearance and higher blood levels3 Methotrexate toxicity might be expected when higher doses are given to animals or patients with radiation nephritis. The interactions can be very complex when the chemotherapeutic agent also produces renal damage.‘We
May 1987,Volume 13,Number5 have recently observed increased c&platinum toxicity after renal radiation and increased chronic radiation nephritis after c&-platinum treatment.12 What tests of renal function should be monitored in patients treated with methotrexate following radiation to the kidneys? The results of our rat experiments suggest that both measurements of glomular 6ltration and tubular function are important. Thus, creatinine clearance, BUN, urine protein and urine concentrating ability should be determined. Recently Hot-i et al.’ has shown that measurement of tubular function by phenolsulfonphtbalein greatly aids in predictions of correct doses of an antibiotic in patients with renaI failure. This more direct assessment of tubular function should be considered in patients receiving drugs that are secreted by the tubules after renal irradiation.
REFERENCES 1. Brown, J.M.: Drug or radiation changes to the host which
2. 3.
4.
5. 6.
7. 8.
9.
could a&ct the outcome of combined modality therapy. Znt.J. Radiat. Oncol. Biol. Phys. 5: 1151-I 163, 1979. Chabner, B.A. Methotrexate: In PharmacoZogicPrinciples of Cancer Treatment, Chabner, B. (Ed.). Philadelphia, W.B. Saunders. 1982, pp. 229-225. Crom, W.R., Glynn, A.M., Abramowitch, M., Pui, C-H., Dodge, R., Evans, W.E.: Use of automatic interaction detector method to identify patient characteristics t&ted to methotrexate clearance. CZin Pharmacol. Ther. 39: 592597,1986. Dance, N., Price, R.G., Robings, D., Stirling, J.L.: Betagalactosidase, beta-glucosidam, and N-acetyl-betaglucosamine in human kidney. Clin. Chem. Acta 24: 189-197, 1969. Daniel, C., Wood, F.: Fitting Equations to Data. New York, John Wiley and Sons. 1980, pp. 86. Greenberger, J.S., Weichselbaum, R.R., Cassady, J.R.: Radiation Nephritis in Cancer and the Kidney, Reiselbach, R.E., andGarnick,M.B. (Ed.). Philadelphia, LeaandFebiger. 1982, pp. 8 14-823. Hori, R., Okumura, K., Kamiya, A., Nihara, H., Nakano, H.: Ampicillin and cephalexin in renal insufficency. Clin. Pharmacol. Ther. 34: 792-798,1983. Jonas, A.M., Besch, E.L., Brennan, P.C., Foster, H.L., Knapka, J.J., Kruckenberg, SM., Serrano, L.J.: Longterm holding of laboratory rodents. Inst. Lab. Animal Res. News 29: l-26,1976. Kalbfleisch, J.D., Prentice, R.L.: The Statistical Analysis of Failure Time Data. NW York, John Wiley and Sons. 1980, pp. 10-19.
10. Kamen, B.A., Takach, P.L., Vatev, R., Caston, J.D.: A rapid, radiochemical-ligand binding assay for methotmxate. Anal. Biochem. 70: Y-63,1976. 11. Major, P., Kufe, D., Frei, E.: Role of the kidney in the pharmacokinetics of anticancer agents. In Cancer and the Kidney Rieselbach, R.E. and Gamick, M.B. (Ed.). Philadelphia, Lee and Febiger. 1982, pp. 263-274. 12. Moulder, J.E., Holcenberg, J.S., Kamen, B.A., Cheng, M., Fish, B.L.: Renal irradiation and the phamtacology and toxicity of methotrexate and c&platinum. Znt. J. Radiat. Oncol. Biol. Phys. 12: 1415-1418, 1986. 13. Moulder, J.E., Martin, D.F.: Hypoxic fraction determinations in the BA 1112 rat sarcoma: Variations within and among assay techniques. Radiat. Res. 98: 536-548,1984. 14. Perrier, D., Mayersohn, M.: Non-compartmental determination of the steady-state volume of distribution for any mode of administration. J. Pharm. Sci. 71: 372-373,1983. 15. Phillips, T.L., Fu, K.K.: Quantification ofcombined radiation therapy and chemotherapy effects on critical normal tissues cancer. 37: 1186- 1200,1976. 16. Rubin, P.: Late effects of chemotherapy and radiation therapy: A new hypotheses. Znt. J. Radiat. Oncol. Biol. Phys. lo: 5-34,1984. 17. Sedman, A.J., Wagner, J.G.: Auto an a discussion-making pharmacokinetic computer program. Ann Arbor, M.I. Publication Distribution Service, 1976. 18. Sheldon, P.W., Hill, S.A., and Moulder, J.E.: Radioprotection by pentobarbitone sodium of a murine tumor in vivo. Int. J. Radiat. Biol. 32: 571-575, 1977.