- Email: [email protected]
Circadian time dependence of murine tolerance for carboplatin
~fO?;KOLO<;Y
AND APPLIED PHARMACOLOGY
96.233-247
Circadian Time Dependence NACEURA.BOUGHATTAS,*-~~FRANCIS ANNIE
ROULON,~
CHARLES
(1988)
of Murine Tolerance Lt%r,*.t FOURNIER,$
for Carboplatin’
BERNARDHECQUET,$GUYLEMAIGRE,§ AND
ALAIN
REINBERG*
*L/.-l CNRS 581 “Chronobiologie-Chronopharmacologie” Fondatiorz ‘4. de Rothschild, 75019 Paris, France, ?Pharrnacologie 2, SMST and ICIG (CNRS UA 04-l 163). H&pita1 Paul-Brousse. 94800 r ‘illejuif: France; $Lahorutoire de Pharmuc~odynamie, Centre Oscar Lambret. 59000 Lille, France: and $Service dI4natomie et de C)fologie Pathologiyue. H6pital.4ntoine Bkl&e. 92000 Clarnart. France
Received December 22. 1987; accepted June 18. 1988
Circadian Time Dependence of Murine Tolerance for Carboplatin. BOUGHA~TAS, N. A.. L&, F., HECQUET, B., LEMAIGRE, G., ROULON, A., FOURNIER. C.. AND REINBERG, A. (1988). Tosicol. Appl. Phermacol. 96,233-247. A large amplitude circadian rhythm in murine tolerance for the anticancer agent, carboplatin (cyclobutane dicarboxylatoplatinum 11, CBDCA) was demonstrated. Two studies were performed in a total of 266 male B6D2Fl mice standardized by LD 12: 12. In the first experiment CBDCA (80 mg/kg/day) was administered intravenously (iv) daily for three consecutive days at all six circadian stages (3, 7. 11, 15, 19, or 23 hr after light onset, HALO). CBDCA dosing at I5 HALO resulted in 58% long-term survivors as compared to 0% after treatment at 3 or 23 HALO (x’ = 28; p < 0.001). In the second experiment. CBDCA (72 or 80 mg/kg/day X 3 days, iv) was administered at any of three circadian stages (0, 8, or 16 HALO). Mice were killed. blood was collected, and seven tissues were obtained 5 and 10 days after the first dose. in order to determine serum urea and creatinine concentrations, leukocyte and red blood cell counts. and to evaluate histologic lesions. No renal toxicity was encountered. Bone marrow and colon mucosa were the major target tissues of CBDCA in these dosages and schedules. CBDCA dosing at 16 HALO was least toxic to the bone marrow as assessed by peripheral leukocyte count and histologic score (p from ANOVA ~0.05). Histologically assessed lesions of the colon mucosa were less severe after CBDCA dosing at I6 HALO as compared to those at 8 HALO, and significantly so for the lowest dosage tested (p - 0.05). Uptake of CBDCA 24 hr after the third dose ranged from 23 @g/g of dry tissue in the colon to 7 pg/g in the duodenum. Mean tissue concentrations increased between Day 4 and Day 10 for the liver and spleen. and remained similar for the kidney. No consistent circadian dependence was found with regard to Day 4 mean Pt uptake in different tissues. whereas the lowest Day IO Pt concentrations corresponded to CBDCA dosing at 16 HALO for all tissues investigated. Toxicity did not appear to be directly related to the total platinum concentration in these tissues. s 19x8 ACKIC~IC P~CSS.I~C.
Murine tolerance for over 20 antineoplastic drugs depends upon their dosing time in the 24-hr scale(L&i er al., 1988a). This applies in particular to ci.c-dichlorodiammineplatinum ’ Supported in part by Grant 6 180 from the Association pour la Recherche sur le Cancer, Villejuif and by Bristol Laboratories, Paris, France. ‘To whom correspondence should be addressed at Pharmacologic 2. I.C.I.G.. H&pita1 Paul-Brousse, 14, av. Paul-Vaillant-Couturier. 94804 Villejuif C&dex. France. 233
II (CDDP) (Hrushesky et al.. 1982) and to the alkylating agents, cyclophosphamide (Haus et al., 1974) and peptichemio (L&vi et ul.. 1987). CDDP-induced hematologic and renal toxicities were lessenedby administering this drug in the activity span of rats (L&i et al., 1982a; 1982b). Carboplatin (cyclobutane dicarboxylatoplatinum II, CBDCA) is a new platinum analog, which lacks renal toxicity but exhibits similar antitumor effects as CDDP, mainly 0041-008x/88 Copvr&t C
$3.00
1988 hy Academic Press. Inc All rights of reproduction in am form reserved.
234
BOUGHATTAS
ET
TABLE
AL.
I
MAIN CHARACTERISTICS OF THE Two STUDIES INVESTIGAT [NC; THE CIRCDIAN ANCE FOR CBDCA (70 OR 80 mg/kg/day iv X 3 days) IN MAL.E B6D2FI MICE AGED FOR3WEEKSByLDl?l?
Study
Dates of study 10/7/86
219187
Nof mice Cont. treat.
Time points studied (HALO”)
VARIABILITY IO WEEKS.AND
Endpoint
IN HOST TOL.FRSYUCHROUIZEU
Day of assessment
6
I14
3. 7. I I. 15. 19.23
Day 40 survival Body weight loss Rectal temperature
Each day for 40 days Every 3rd day for 4 weeks Days I, IO. 18. and 28
36
107
0. 8. 16
Body weight loss RBC, WBC count Serum urea & creatinine Histology of 7 tissues” Platinum concentrations 6 tissues
Day 5 and Day 5 and Day 5 and Day5and Dayiland
” Hours after light onset. ’ Kidney. liver. spleen. bone marrow.
duodenum.
jejunum.
through interactions with DNA (Harrap ef al., 1980; Mong et ul.. 1980a. 1980b: Knox rt nl.. 1986). Its dose-limiting toxicity is bone marrow suppression (Calvert et al.. 1982; Curt et al., 1983: Evans et ul., 1983: Van Echo rf al., 1984: Koeller ef al.. 1985). The present investigation is aimed at assessing the role of dosing time upon CBDCAinduced toxicity. and at isolating possible
in
IO 10 10 IO IO
and colon
mechanisms of such rhythms. Results were subsequently expected to guide the chronopharmacologic optimization of the use of CBDCA in cancer patients, as was already achieved for cisplatin and other anticancer agents (Hrushevsky, 1985: Levi, 1987). Mortality, body weight loss, rectal temperature, leucopenia, and histologic lesions in seven tissues were used as criteria for toxicity.
TABLE
2
CIRCADIAN RHYTHM IN TOL.FRANCE OF MICE FOR CBDCA: RESULTS FROM COSINOR ANALYSIS. WITH A PERIOD T = 24 HR
Variable Synchronization Rectal temperature (“C) Day I Toxicity Rectal temperature (“C) Day 10 Body wt loss (aI) Day 10 + Survival at 50% overall mortalitv $ Long-term survivors Survival time (davs)
h’ data
120
87 87 6 6 114
Mesor
P”
? SEM
Double
amplitude
+ SD
Acrophase (HALO) * SD (min)
37.8 + 0.1
1.5 2 0.5
10.03 10.00
I
35.8 -t 0.2 20.7 k 0.8
1.X k 1.6 18.5 -+ 6.0
15.00 t 155 13.10* 70
54.2 -t 7.2 21.7 k4.7 17.8 t I.0
XI.0 55.5 17.4 + 7.1
13.20 15.20 14.50 k 90
,Ll’ot~. HALO. hours after light onset. ” From an b’test of the rejection of the null amplitude
hypothesis.
19.50 k
80
CIRCADIAN
TOLERANCE
FOR
235
CARBOPLATIN Key
ln,ectm
Tvne
Logrank
(Halo) 15 19 11
p
36 05 d 125
co
005
07 03
. .. . .. ..
71
test
x’
23
I
3
6
11
14 TIME
16 FOLLOWING
23 CBDCA
TREATMENT
26
31
34
39
(days,
FIG. I. Survival curves of B6DZFl mice over the 40 days following the iv administration of 80 mg/kg/ day x 3 days of carboplatin (CBDCA). Each survival curve corresponds to a different dosing time of CBDCA along the 24-hr scale (study 1). Survival rate differed as a function of dosing time both when 50%) overall mortality was reached (x2 = 45: p < 0.00 I ) and on Day 40 (x’ = 18: (I< 0.00 I ).
Platinum uptake in six of these tissues was also examined as a function of treatment time. MATERIALS
AND
METHODS
Two studies were carried out on a total of 166 1O-weekold male B6D2FI mice (IFFA-CREDO. L’Arbresle. France) in October 1986 and February 1987. Mice were housed three per cage, with free access to food and water. and were synchronized at least 3 weeks prior to each study (Reinbergand Smolensky. 1983) in a chronobiologic animal facility (ES1 Flufrance, Arcueil, France). The lighting regimen was 12 hr of light (L) and 12 hr of darkness (D) (LD 11: 12) with light intensity equal to 300 + 50 Ix. Each facility had six sound proof, temperature-controlled (‘73 * 1°C) compartments with liltered air (100 -+ IO liter/min), and each compartment had its own programmable lighting regimen. This design allowed the performance of six circadian stages of CBDCA dosing (3, 7, I I. 15. 19. and 23 hr after light onset. HALO) in study 1 and three circadian stages in
study 2 (8. 16. and 34 HALO). All drug administrations were between 9:00 and 1200 hr. Synchronization of circadian rhythms of mice was checked prior to drug dosing by measuring the rectal temperature of each mouse. 2. I)nr!: CBDCA was kindly supplied by Bristol-Myers (France) in vials containing 150 mg of lyophilized (‘w diammine ( I. I -cyclobutane dicarboxylato)platinum and 150 mg of mannitol. The CBDCA solution was freshly prepared on each study day by adding an adequate volume of distilled water in order to obtain the desired concentration. Two doses were administered to mice (71 or 80 mg/kg/day *: 3 day) in a fixed fluid volume (5 ml/kg body wt) which corresponded to the CBDCA concentrations of 14.4 and 16 mg/ml. respectively. These doses are the LDIO and LD50. respectively (unpublished results). Each mouse was given an intravenous injection ofeither the drug solution or the 80 mg/kg of mannitol into the retro-orbital sinus daily for three consecutive days.
The study designs rized in Table I.
and toxicity
endpoints
are summa-
236
BOUGHATTAS
ET AL.
A
12 DOSE:
9.
72
mglkgldx3d
DOSE:00mglkg
i.v
/d=3d
i.v
T
6.
3.
TIME
OF
(hours
INJECTION
after
light
onset
1
ANOVA
. . . _ . _ _.
CEDCA
72mg/kg
CEDCA
80mglkgid
/d I v.3d I v.3d
COSINOR
F
P
P
106
0001
0001
46
032
002
t +25
. CONTROL
-25.
3 03
07
TIME
(hours
II
15 after
19 light
onset
23 )
MESOR
( 72 m g /kg.3)
MESOR
(80 m g lkg.3)
CIRCADIAN
TOLERANCE
In study I. the survival rate was computed for each group at 50% overall mortality and at study completion on Day 40. Body weight loss was computed as percentage change for each mouse. In study 2, leukocyte and red blood cell counts were determined in blood (-500 ~1) obtained from the retroorbital sinus. These cells were quantified in a Z.M. Coultronics Coulter counter according to standard procedures. Serum urea and creatinine concentrations were determined by spectrophotometry (Uree enzymatique UV H.P. Biotrol; and Creatinine cinetique Biomerieux. France). H~.rrc’(~urhr,lo~j~, srlrdj,. Immediately after being bled, the mice were killed. Liver and spleen were immediately fixed into Carnoy’s solution; kidney. duodenum, jejunum. colon, and bone marrow (femur) were fixed into Bouin’s picroformol solution. Twenty-four hours later. these organs were dehydrated and embedded into parahin. Sections were stained with hematein-eosin. Each encoded slide was examined by the same histopathologist who was unaware of the treatment groups. Lesions were graded between 0 (normal) and 4. The highest grade (4) corresponded to complete necrosis for the bone marrow, to necrosis of the superficial part of the villi for the small howel, and to a deep disorganization of epithelial structures of the colon with dedifferentiation. glandular. dilatation. and inflammatory infiltrate. The reliability of such a scoring system has previously been documented (Levi (‘I ul.. 1987). T;F.YI(Y Ji.~~~ihu(iolr of p/utirrrr/?r Tissue distribution of (‘BDCA was studied in 36 mice (3 mice/time point/dose) on Day 4 or IO following the first dose of 73 or 80 mg/ kg/day at 8. 16. or 24 HALO. After mice were killed, liver. spleen, kidney. duodenum. jejunum. and colon were sampled and immediately weighed. then air-dried to a constant weight. Dry tissues were digested in 1 ml of nitric acid for I hr at 100°C. After acid evaporation Pt tissue content was measured by tlameless atomic absorption spectrophotometry (FAAS) with electrothermal atomization (ETA) using a Perkin-Elmer Model 1280 IHecquet (ll al.. 1983). The following time temperature program was used: drying for 30 set at I XC, ashing for 30 set at 14Oo”C, and atomization for 3 set at 2700°C. .kuri~ical anu/wis. Means and standard errors of the mean (SEM) were calculated for each variable. dose. and time point. The statistical significance of differences bctween groups was validated for parametric data by one-
FOR
337
CARBOPLATIN
or two-way analysis of variance (ANOVA). Taking into account the non-normal distribution, platinum concentration was also estimated in each tissue using medians (hl) and coefficients of dispersion (CD) according to the following formulas (De Prins
/ Ct - .If ( /0.6745
CD = ICti - I%/1/[(0.798(h’
- l)]
when
,I 2 6
when
/I < 6,
where Ct is tissue concentration. Differences in tissue concentrations were then analyzed by nonparametric tests (Mann-Whitney test and/or Kruskall-Wallis ANOVA). Differences in survival rates were analyzed by x2 test and survival curves by logrank test (Peto rf al.. 1977). Time series were analyzed for circadian rhythmicit! (with a period, T. 214 hr) by the cosinor method (Nelson (‘1 al.. 1979). A rhythm was characterized by three parameters: the mesor, M (24 hr-adjusted mean), the double amplitude, 2A (difference between the minimum and the maximum of the fitted cosine function). and the acrophase. 4 (time of maximum, with time of light onset as 9 reference). M, ?A. and + were obtained with their 95% confidence limits, if a rhythm was detected. This was achieved when A differed from zero (non-null amplitude rtest) with p 4 0.05.
RESULTS StlKiJ I Results from cosinor are shown in Table 3.
analysis
of the data
Rectal tempc~ratm~. A circadian rhythm was statistically validated by cosinor on Day 1. Thus, mice were well synchronized prior to dosing with CBDCA. The acrophase was localized near the middle of the dark span as is usually the case. Treatment resulted in hypothermia. which was maximal on Day 18 (mean, - 1 1%). Temperature decrease varied as a function of dosing time (-20% at 3 HALO: -9%’ at 15
FIG. 2. (A) Mean circulating leukocyte count in control and CBDCA-treated mice 5 and 10 days after the first of three consecutive doses (72 or 80 mg/kg/day X 3 days). Mice were treated with CBDCA and sampled at the same circadian stage. Three circadian stages (0, 8, and 16 hr after light onset, HALO) were compared with regard to the effect of CBDCA upon such endpoint. (B) Leukopenia index. percentage change in circulating WBC count (relative to time-qualitied controls) on Day 5 and 10 after treatment with CBDCA (72 or 80 mg/kg/day .i 3 days iv). Dose and dosing-time dependence of leukopenia index as modeled by cosinor analysis.
238
BOUGHATTAS
ET AL.
CIRCADIAN
TOLERANCE
FOR
CARBOPLATIN
339
Ftc;. 3. Transverse section of femoral bone marrow 5 days after first dose of mannitol (control) or 72 mg/kg of CBDCA at either 08.00 or 16.00 HALO. Hematein-eosin staining. (4) Control-histologic score = 0 (magnitication x85). (B) CBDCA at 8 HALO-important medullar necrosis score = 3 (~85). (C) CBDCA at 16 HALO moderate decrease ofcellular population. histologic score = I (K X5).
HALO). The changes in rectal temperature followed a circadian rhythm. Rectal temperature of survivors returned back to pretreatment values 4 weeks after the start of treatment (on Day 28, mean change - I %). &Y/J* ~tx4ghf c$arz,~~~.Mean maximal body weight loss was achieved on Day 6 after the tirst dose of CBDCA and remained unchanged until Day 18 (- 19% of the initial body weight). On Day 10 mean body weight loss varied between 1 I %, in mice dosed at 15 HALO and 34% in those injected at 23 HALO (F = 20.2; 11< 0.01). Among the 25 mice surviving on Day 38. those who had received CBDCA at I5 or 19 HALO had recovered their initial body weight (respective mean body weight change, -8% and +%), whereas those
treated at 7 HALO had not (-20%‘) (F from ANOVA = 3.6; p = 0.07). Survivd. Deaths were observed between Day 5 and 25. Survival rate remained subsequently unchanged. Survival curves ofthe six studied groups were computed according to Kaplan-Meyer method (Kaplan and Meyer. 1958). Percentage of survival differed significantly as a function of both CBDCA dosing time (circadian stage) and time span following treatment (days) (Fig. 1). Fifty percent overall mortality was reached on Day 12. On this day, survival rate varied between 95% in those mice dosed at 15 HALO and 0%’ in those treated at 23 HALO. The long-term survival rate varied between 58% ( 15 HALO) and 0% (3 and 23 HALO). Circadian rhythms in survival rates and survival times were also detected by cosinor with acrophases localized
240
BOUGHATTAS
O-
06
TIME
16 (hours
after
24 light
onset
I
FIG. 4. Histologic scores of bone marrow 5 and IO days after the first iv dose of CBDCA. Three circadian stages of CBDCA injection are compared. A dosing time-related effect was statistically validated on Day 5. Minimal lesions corresponded to dosing at I6 HALO. On Day 10, maximal lesions were found (mean score = 3.9) whatever the circadian stage.
near 15 HALO (Table 2). The difference between mean longest and shortest survival times was 19 days, with optimal dosing time being 15 HALO (F from ANOVA = 9.0; p < 0.001).
Study ,7 Bonl,) Mw+ght and survival. On Day 5, body weight loss was similar to that observed in study 1 (19% of the initial body wt). On Day 10. body weight loss varied between 27% in mice dosed at 16 HALO and 33% in those injected at 8 HALO (F = 4.1, p = 0.02). Thirteen percent of those mice receiving CBDCA at 0 or 8 HALO had died from toxicity before Day 10. Inlmunohnnatolo~ic toxicit?: In controls circulating RBC count remained similar whatever its sampling time or sampling day (mean f SEM = 10.6 + 0.3 X lo6 cells/mm’). CBDCA resulted in a moderate decrease in circulating RBC count on Day 10 (F = 9.6; p < 0.005) irrespective of the treatment dose and timing (8.5 + 1.1 X IO6 cells/mm3).
ET AL.
Mean circulating leukocyte count varied as a function of sampling time in controls, and irrespective of sampling day. A marked circadian rhythm was detected by cosinor (y < 0.001; 4 = 5.30 f 1.20 hr). Following CBDCA treatment, the circulating WBC count decreased, more so on Day 10 as compared to Day 5 (Fig. 2A). Because of the physiologic circadian rhythm in circulating total WBC count, data from treated mice were also expressed as percentages of the corresponding mean time-qualified control values irrespective of sampling day (leucopenia index). Mean leucopenia ranged from 37% after dosing with CBDCA (80 mg/kg/day X 3 days) at 16 HALO to 77% after drug dosing at 8 HALO (Fig. 2B). A dose response was documented both for raw data and for transformed data (Fig. 2). A severe bone marrow necrosis was observed on Day 5 in treated mice. The extent of such necrosis varied as a function of both dose (1; = 3.1: p from ANOVA -0.05) and dosing time (Fig. 3). The lowest histologic score occurred at 16 HALO and the highest at 8 HALO (Fig. 4). A circadian rhythm in the bone marrow score was further demonstrated by cosinor. On Day 10, a complete necrosis was apparent in all specimens whatever the dosing time (Table 3). Ir2tedinul to.uicitj! Intestinal lesions mainly affected the colon mucosa. Two days after the last CBDCA injection (Day 5). a deep disorganization of epithelial structures was observed with dedifferentiation. inflammatory infiltrate of chorion, and kryptic abscess (Fig. 5). Five days later (Day 10) most colic lesions were repaired, since a partial recovery of mucous secretion was observed. A dose-response relationship was statistically validated both on Day 5 and on Day 10. Whereas no effect of dosing time was apparent on Day 5. recovery of colic lesions appeared to be faster after CBDCA dosing at 16 HALO, and significantly so for the lower dose used (Table 3).
CIRCADIAN
TOLERANCE
FOR
TABLE Toxrcir~
OF CBDCA
241
CARBOPLATIN
3
INMICEFORBONEMARROWANDCOLONMUCOSA,ASASSESSEDBYHISTOLOGICSCORES Two-way Time
Tissue
Dose” bWc&W
Bone marrow
72
80
Day
72
intravenously
00
08
16
F
P
F
2.2 f 0.3 3.5 + 0.3
3.8
0.03
3.00
-co.00
1
1
2.6 k 0.2
3.2 * 0.2
4.0 + 0
3.8 k 0.2
5
2.8 k 0.2
3.3 -+ 0.2
3.9iIO.l
4.0 + 0
2.8 f 0.2 4.0,o
3.8
0.03
8.6
2.8 f 0.5 1.x+0.7
2.7 f 0.3 1.2 + 0.4
3.0 i 0.6 0.3 f 0.3
3.1
0.06
16.2
3.0 f 0.4 1.2 f 0.4
3.2 +- 0.4 2.4 to.1
3.0 f 0.6 1.9 kO.5
0.8
0.5
5
5 I0
” Injected
Day
5
IO
80
Time
IO
10 Colon
(hr after light onset)
ANOVA
for three consecutive
7.0
P
10.00 1 0.01
days.
Minor lesions were observed in the duodenum or the jejunum. None was found in liver, spleen. or kidney. Serum urea and creatinine were similar in controls and treated mice. Platinum distribution in tissues. Twentyfour hours after the third CBDCA dose (Day 4), mean highest platinum concentrations were observed in the colon and kidney. Platinum uptake was about two times less in liver, spleen. and jejunum and three times less in duodenum (Fig. 6). On Day 10, mean tissue platinum concentration decreased three- to sixfold in duodenum, jejunum, and colon as compared to their respective values on Day 4. Conversely platinum concentrations increased over this 6-day span both in the liver and in the spleen, but were unchanged in the kidney. An effect of CBDCA dosing time was apparent on the platinum concentration in all tissues on Day 10, with lowest tissue concentrations corresponding to drug dosing at 16 HALO (Fig. 7). With respect to tissue platinum concentration, results of three-way ANOVA, (dose, sampling time, sampling day) and nonparametric Kruskall-Wallis ANOVA validated an effect of sampling day for all tissues except the kidney, an effect of
CBDCA dose for spleen, duodenum, jejunum, and colon, and an effect of dosing time for spleen, duodenum, and jejunum.
Correlation hettileen platinum distribution and tissue toxicit)). Median Pt uptake on Day 4 correlated with the median colon toxicity score on Day 5 (r = 0.79; df= 5; p < 0.05). No correlation was found between the median Pt concentration and the median colon score on Day 10 (r = 0.38; df= 5;~ > 0.10). However, the correlation between Pt uptake on Day 4 and Day 10 histologic score was close to statistical significance (r = 0.70; df = 5: p - 0.07). DISCUSSION Optimal tolerance for CBDCA was achieved when this drug was given during the activity span of mice near 15 HALO, whatever the endpoint studied (changes in rectal temperature or in body weight, survival). The adequate animal synchronization by the six different LD 12: 12 schedules was demonstrated by circadian rhythms in rectal temperature and circulating leukocytes. as previously reported (Haus et al., 1983; Levi et al.. 1987; Levi et [II.. 1988a). Necrosis of both
CIRCADIAN
TOLERANCE
FOR
243
CARBOPLATIN
Sampling 4
colon P (Mann-whitney)
kidney 0.2
liver
*pII?=?”
ieimum
duodenum
< 0.01
(24hr
day pos.1
3’ddose)
(NS)
FIG. 6. Histogram of Pt concentration (median * coefficient of dispersion, CD) after intravenous CBDCA dosing (72 or 80 mg/kg/day > 3 days). Six tissues were obtained 24 hr after the third dose of CBDCA (Day 4) or 6 days later (Day IO).
bone marrow and colon mucosa constituted the main mechanisms ofthe lethal toxicity of CBDCA. since no significant lesion was found in any other organ studied. Irrespective of CBDCA dose or dosing time, a greater decrease in total WBC count was observed on Day 10 as compared to that of Day 5. in accordance with previous results (Siddik PI al.. 1987a). Bone marrow necrosis was also more severe on this day. Similar results were reported in mice using the survival of hematopoietic stem cells (CFU) in the spleen colony assay (Lelieveld rl al.. 1984). Conversely, colic lesions were more pronounced on Day 5 as compared to those of Day 10. The lowest damage in colon and bone marrow and the least hematologic toxicity corresponded to CBDCA dosing at 16 HALO. Drug-induced leucopenia was 50%) less severe after treatment at this circadian time, as compared to that of 0 or 8 HALO. This may partly be accounted for by the circadian rhythm which characterizes cellular proliferation. More spe-
cifically, the proliferative ability of bone marrow granulomonocytic precursors is highest in the dark (activity span) and DNA synthesis is increased in the light (rest span) of mice (Sheving et al.. 1983; Levi et al., 1988b). In addition, cell proliferation in mouse colon displays a marked circadian rhythm. The circadian peaks of both M and S phasesof the cell division cycle occurred during the rest (light) span (Scheving et al.. 1983) with the mitotic index being maximal near 8 HALO (Chang, 1971; Chang and Nadler. 1975: Hamilton, 1979). Maximum Pt concentrations were measured in the colon and kidney 24 hr after the last CBDCA dose, as previously observed (Litterst, 1984: Boven et al.. 1985). On Day 10 as compared to Day 4, a significant increaseof Pt content occurred in the liver and in the spleen. This may be due to irreversible Pt binding both to nucleophilic sites of macromolecules and to plasma proteins (Harland c/t nl.. 1984: Gaver cf N/., 1986: Laznickova
FIG. 5. Transverse section ofcolon 5 days after first dose of 80 mg/kg of mannitol or CBDCA. Hemateineosin staining. (A) Control. histologic score = 0 (magnification X85). (B) CBDCA at 8 HALO, deep disorganization ofepithelial structures with dedifferentiation. glandular dilatation. inflammatory infiltrate ofchorion. and kryptic abscess. Score = 4 (/85).
244
BOUGHATTAS
ET AI
34
30
26
16
2:
TIME of CBCCA (Hours Aftor
INJECTION Light 0nm.t)
FIG. 7. Mean Pt concentration measured in six tissues. IO days after the first dose of intravenous (72 or 80 mg/kg/day x 3 days) at any ofthree circadian stages.
rt ul.. 1986; Siddik et al.. 1987b). Kidney Pt concentration remained similar on Day 10 as compared to that of Day 4. Such sustained concentration may result from covalent
CBDCA
binding of Pt to intracellular components in the kidney. Conversely, the marked decrement of Pt concentration in the colon from Day 4 to 10 paralleled the recovery of muco-
CIRCADIAN
TOLERANCE
sal lesions. This may result from the renewal of damaged mucosal cells. Such a relationship between the Pt concentration and the extent of mucosal damage was also suggested along the circadian time scale. Time-dependent tissue distribution was apparent on Day 10 for all tissues investigated with 16 HALO corresponding to the lowest mean Pt concentration. On Day 10, colic lesions were also less after CBDCA dosing at 16 HALO as compared to those at 8 HALO. Furthermore a statistically significant correlation was observed between Pt uptake on Day 4 and colic lesions on Day 5. Nonetheless. no apparent correlation was documented between Pt uptake and tissue toxicity in the other tissues studied, since the highest Pt uptake was found in the kidney. where no lesion was documented. The lack of any active urine secretion or tubular reabsorptive process (Daley-Yates and McBrien. 1983: Siddik et nl.. 1987b) for CBDCA and its greater chemical stability (Cleare et (I/.. 1980: Howe-Grant and Lippard, 1980) may explain its apparent lack of nephrotoxicity (Daskal (‘t cl/., 1980; Harrap rt ~1.. 1980: Cal\.ert et al., 1982: Egorin et ~1.. 1984; Batzer and Aggarwal, 1986) as compared to that of CDDP. Thus glomerular filtration is the main mechanism of CBDCA excretion in the rat (Harland et uI.. 1984: Siddik ~1u/.. 1987h). This function exhibits a well-known circadian rhythm with maximal values occurring near the middle of the activity span of rats or mice (Cal et ~1.. 1986). This may account for an increased excretion of this drug near I6 HALO. Nonetheless. platinum accumulation in the kidney following repeated CBDCA dosages may lead to some moderate degree of renal toxicity. as was reported both in rats (Levine et al.. 198 1; Lelieveld t’t u/.. 1984) and recently h vitro (Phelps et ~1.. 1987). CBDCA and CDDP exhibit different toxic and pharmacologic properties. However, the circadian dosing time associated with optimal tolerance appears to be similar for both platinum analogs. Further studies are needed
FOR
CARBOPLATIN
245
to assess whether this also applies to their antitumor effectiveness. Let us emphasize that CDDP was significantly more effective against a transplantable rat immunocytoma when dosed at the circadian time when it was best tolerated (Levi (‘I al.. 198 1). The relevance of the mouse model to predict the least toxic dosing time of anticancer agents in patients has already been documented for adriamycin (Hrushevsky, 1985). 4’-tetrahydropyranyl-adriamycin (L&i, 1987). and CDDP (Hrushevsky, 1985; Levi, 1987). CBDCA was best tolerated by mice after dosing it near 16 HALO. This may correspond to 15.00- 16.00 hr in cancer patients.
ACKNOWLEDGMENTS The authors express their thanks to G. Metzger. M. Mechkouri. and M. ltzhaki for-statistical assistance. to E. Co& for secretarial assistance in the preparation of the manuscript. and to Z. Khalouche for iconographic world.
REFERENCES BATZER. M. A.. AND AC;CARWAL. S. K. (1986). An in vitro screening system for the nephrotoxicity of vatious platinum coordination complexes. C~rrc~cr C%ctnc~ther. Ptwmac~~d 17, 209-2 1 I. BOVEN, E.. VAN DER VIJGH. W. J. F.. NAUTA. M. M.. SCHLUPER. H. M. M.. AND PINEDO. H. M. (1985). Comparative activity and distribution studies of five platinum analogues in nude mice bearing human ovarian carcinoma xenografts. C’upr(,er Rex 45, 86-90. CAL. J. C., DORIAN, C.. AND CAMBAR. J. (1986). Circadian and circannual changes in nephrotoxic effects of heavy metals and antibiotics. In .lnr~ru/ Kurirlc- o/ (‘hronophurrnacol~~~~~~ (A. Reinberg. M. Smolensky and G. Labrecque Eds.), Vol. 2. pp. 343-176. Pergamon. Oxford. New York. CALVERT. A. M.. HARLAND. S. J.. NEWELL. D. R.. SIDDIK, Z. H.. JONES, A. C.. McEr WAIN, T. J.. RAJU. S.. WILTSHAW. E.. SMITH. I. E.. BAKER. J. M.. PECKHAM. M. J.. AND HARRAP, K. R. (1982). Early clinical studies with cis-diammineI. I cyclobutane dicarboxylate platinum II. C‘unwr C’twmorhzr. Ptmrrmm~i. 9, 140147.
CHANG. W. W. L. ( 197 I ). Renewal of the epithelium in the descending colon of the mouse. III. Diurnal variation in the proliferative activity of epithelial cells. .Irnw J. :lnat. 131. 1 I I-120.
246
BOUGHATTAS
CHANG, W. W. L.. AND NADLER. N. J. ( 1975). Renewal of the epithelium in the descending colon in the mouse. IV. Cell population kinetics of vacuolated-columnar and mucous cells. .-lrner. J .Jnal. 144, 39-56. CLEARE. M. J.. HYDES. P. C.. HEPBURN. D. R.. AND MALERBI. B. W. (1980). Antitumor platinum complexes: Structure-activity relationships. In C’is&zt~rl. C’urrmr Stutm und NW Devcloptnctm (A. W. Prestayko. S. T. Crooke, and S. K. Carter Eds.), pp. 149170. Academic Press. New York. CURT, G. A.. GRYGIEL. J. J.. CORDEN. B. J., OZOI s. R. F., WEISS, R. B., TELL. D. T.. MYERS. C. E.. AND COLLINS. J. M. (1983). A phase I and pharmacokinetic study of diammine cyclobutane dicarboxylato platinum (NSC 24 1340). Cuttcrr Rev. 43, 4470-4473. DALEY-YATES. P. T., AND MCBRIEN. D. C. H. (1983). The mechanism of renal clearance of cisplatin (cisdichlorodiammine platinum II) and its modification by furosemide and probenecid. Biohtx I-‘llurrr~ac,ol. 31.2243-1246. DAXAL, Y.. PRESTAYKO. A. W.. AND CROOKE. S. T. (1980). Morphological manifestations ofcisplatin analogs in rats. And ultrastructural study. In C’i.sp/ufin (A. W. Prestayho. S. T. Crooke. and S. K. Carter Eds.). pp. 249-269. Academic Press. New York. DE PRINS. J.. CORNELISSEN. G.. AND MALBECQ, W. (1986). Statistical procedures in chronobiology and chronopharmacology. .-innu. Km,. C%rott~,~/turtttut,f)/ 2,27-141. EGORIN. M. J.. VAN ECHO, D. A.. TIPPING. S. T., 01.. MAN. E. A.. WHITACRE. M. Y., THOMPSON. B. W., AND AISNER. J. ( 1984). Pharmacokinetics and dosage reduction of cisdiammine ( 1.1 -cyclobutane dicarboxylato) platinum in patients with impaired renal functions. Cuttwr Rcs. 44. 5432-5438. EVANS. B. D.. RAJLI. K. S.. CALVERT, A. H., HARLAND. S. J.. AND WILTSHAW. E. (1983). Phase II study of JM8. a new analog. in advanced ovarian carcinoma. Chnwr 7’rcwf. Rep. 67, 997- 1000. HAMILTON, E. ( 1979). Diurnal variation in proliferative compartments and their relation to ctyptogenic cells in the mouse colon. Cc,// Tir.slrr> Kine/. 12. 9 I-100. HARL.&ND. S. J.. NE~EI.L, D. R.. SIDDIK. 2. H., CHADWICK. R., CALVERT. .A. H.. AND HARRAP. K. R. ( 1984). Pharmacokinetics of ci.r-diammineI,1 -cyclebutane dicarboxylate platinum (II) in patients with normal and impaired renal function. Cunccr Rev. 44, 1693-1697. HARRAP. K. R.. JONES. M.. WILKINSOIU‘. C. R.. CLINK. H. M.. SPARROW. S.. MITCHLE\. B.. CLARKE, S.. AND VEASEY. A. (1980). Antitumor, toxic. and biochemical properties of cisplatin and eight other platinum complexes. In Cisplutit?. CWTWZ/ Stuizrs und NW Dewlopmwt.s (A. W. Prestayko. S. T. Crooke. and S. K. Carter Eds.). Vol. 1’. pp. 193-112. Academic Press, New York.
ET AL. HAUS, E.. FERNANDES. G., KUHL. J. F. W.. YUNIS, E. J.. LEE, J. D.. AND HALBERG. F. ( 1974). Murine circadian susceptibility rhythm to cyclophosphamide. C/tronohio/o,~iu 1. 270-277. HAUS, E.. LAKATUA. D. J.. SWOYFR, J.. AND SA~XEI-~LIJNDEEN, L. (1983). Chronobiology in hematolog> and immunology. :lmcr. J. .-lnu!. 168.467-5 17. HECQIIET. B.. ADENIS, L.. AND DEMAILLF. A. (1983). /II vitro interactions of TN06 with human plasma. (‘un(‘CT (%rtno~her. Pharmacol. I I, I77- 18 1. HOWE-GRAN’I. M. E., AND LIPPARD, S. J. ( 19X0). Aqueos platinum (II) chemistry: hinding to biological molecules. In die/a/ lon.7 in RIo/o~~I& S~:strtn.c(H. Sigel Ed.). Vol. 1 1. pp. 63-125. Dekker, New York HRUSHES~;~‘. W.. Live, F.. AND HALBERG, F. ( 1981). Circadian-stage dependence of (‘,r-diammine dichloro platinum lethal toxicity in rats. C’mwr Rcs. 42. 945949. HRLISHESK\I, W. J. M. ( 1985). Circadian timing of cancer chemotherapy. Smwc~ 228, 73-75. GAVFR. R. C.. SWIGOR. J. E.. G~ORC;E. A. M.. HA>.NE:s. Ll. S.. ANU DEEB. G. (1986). Protein binding and degradation of ‘“C-carboplatin in human plasma and urine in virrzJ. Prw .Attwr. Ls.~~tc C’utmr Rm. [Abstract No. I I3 I] KAPI AN. E. L.. AND MF\ TV. P. ( 1958). Non parametric estimation from incomplete observations. J .,lt?lcpr-. Stur. 1.s.s~~~53. 457-48 1 KOEI I.ER. J. M., TRUMP, D. L.. T~XCH, K. D.. EAR11~~1. R. H.. DAVIS. T. E.. AND TORME~. D. C. ( 1985). Phase I clinical trial and pharmacokinetics of carboplatin (NSC 241240) by single monthly 3%minutes infusion. C‘unwr57, X-X5. KNOX. R. J.. FRIEDI.OS, F.. IL>.DAI I., D. A.. AND RUBERTS. J. J. ( 1986). Mechanism ofcytotoxicity of anticancer platinum drugs: Evidence that cir-diammine dichloroplatinum (II) and c,ic-diammine-( I. l-cyclobutane dicarboxylato) platinum (II) differ only in the kinetics of their interaction with DNA. Cirtww Rec. 46, 1972-1979. LALNI~ROV~, A.. LAZNICEK, M.. Kvt-TINA. J.. 4~11 DROBNIK, J. (I 986). Pharmacokinetics and plasma protein binding oftwo platinum cytostatics CHIP and CBDC‘A in rats. Chttcw Chcvttotltcr. f’hurttw~~rtl. 17. 133-336. LFI IEVELD. P.. VAN DER VIJL;~I. W. J. F.. VELDML:IZEN, R. W., VA~V VEI.ZEN, D., VAU PII.I-‘IXN, L. M.. A r.4ss1. G.. AND DANGLI’Y. A. ( 1984). Preclinical studies on toxicity. antitumor activity. and pharmacokinetics of cisplatin and three recently developed derivatives. Eur. J. <‘utwr C’Ittr. Otmtl. 8, 1087- 1 104. L&I. F.. HALBERG. F.. NESBIT. M.. HAIIS, E.. AND LtVINE. H. ( I98 1 ). Chrononcology. In .Veoplustn.c C‘rwpurutiw Pu’Lltl~~ht~~~~0l‘Gtwwllt it] .hitmh. Plutl1.v. utzd ,lfu?f (H. E. Kaiser Ed.). Chap. 14. pp. 267-3 16. Williams & Wilkins, Baltimore. London.
CIRCADIAN
TOLERANCE
LEVI. F.. HRUSHESKY. W., BLOMQUIST,J.. LAKATUA, D.. HAUS, E.. HALBERG, F.. AND KENNEDY. 9. J. ( 1989a). Reduction of cis-diammine dichloroplatinum nephrotoxicity in rats by optimal circadian drug timing. Cancer Rev. 42, 950-955. LEVI. F.. HRUSHEXY, W.. HAUS, E.. HALBERC;, F.. AND KENNEDY. 9. J. (1989b). Circadian rhythm in lethal. renal. and hematologic tolerance for cis-diammine dichloroplatinum (II). E~rr. .I. Chnccr C‘lin. Om~tl. 18, 471-477.
LEVI. F. (1987). Chronopharmacology of anticancer agents and cancer chronotherapy. In Klinischr Plrurttzuko/o,qie (H. Kuemmerle Ed.). pp. l-17. Ecomed Press. Landsberg. Lfvl. F.. HORVATH. C.. MECHKOURI. M.. ROULON. A.. BAILLFUI.. F.. LEMAIGRE. G.. REINBERG. A.. AND MUHE, G. (I 987). Circadian time dependence of murine tolerance for the alhylating agent peptichemio. Fur- .I. C’unccr C/in. Oncol. 23(5), 487-497. LEVI. F.. BOUGF~A~~AS, N. A.. AND BUZSEK, 1. (1988a). Comparative murine chronotoxicity of anticancer agents and related mechanisms. In .drzn~~l Review r!/ C’krorlc)~/2unl?u(.ol~!~~. (A. Reinberg. M. Smolensky. and G. Labrecque. Eds.). Vol. 4. pp. 383-33 I, Pergamon. Oxford. Livi, F.. BLAZSEK. 1.. AND FERLE-VIDOVIC’, A. (1988b). Circadian and seasonal rhythms in murine bone marrow colony forming cells affect tolerance for the anticancer agent 4’.tetrahydropyranyladriamycin (THP). E.xp IIcmiIoI in press. I.~vIN~:. 9. S.. HENRY. M. C.. PoKr. C. D.. RI(.HIER, W. R.. AND URwNth, M. A. ( 198 1). Nephrotoxic potential ofcisdiammine dichloroplatinum and four analogs in male Fisher 344 rats. .7. Su//. (hnc,er 11~~1 67, 20 I-106. 1.11’1 ERST. C. L. ( 1984). Plasma pharmacokinetics. urinap excretion. and tissue distribution ofplatinum following i.v. administration of cyclobutane dicarboxylatoplatinum II and cis-platinum to rabbits. In Pkulit1utt1 C‘oorditlu/iott it? C’LUICW C’/wttwthoqt~~ (M. P. Hacker, E. B. Douple, and I. H. Krakoff Ed%). pp. 71XI. Martinus-Nijhoff. Hingham. MA. MONA;. S.. HCIANC;. C. H.. PRESTA~XO, A. W.. AI\;I) CROO~F. S. T. ( 1980a). Effects of second-generation
FOR
247
CARBOPLATIN
platinum analogs on isolated PM-2 DNA and their cytotoxicity irl ,)itro and in viro. (htlcer Res. 40, 33 I83324. MONG, S.. PRESTAYKO, A. W.. AND CROOKE. S. T. (1980b). 112 ,)itro interaction of covalently linked closed circular DNA with the second generation platinum compounds. In C’isplutitz. C’urretzf Status and h’e+l D~velopmolrs (A. W. Prestayko. S. T. Crooke. and S. K. Carter. Ed%), Vol. I?. pp. 213-226. Academic Press. New York. NELSON. W.. TONC;. Y.. LEE, J. K.. AND HALBERC;, F. ( 1979). Methods for cosinor rhythmometry. C7tnvw hiohgiu 6, 305-323. PETO, R.. PIKE. M. C., ARMIIAGE, P.. BRESLOW. N. E.. Cox. D. R.. HOWARD. V.. MANTEL. N., MCPHERSON. K.. PETO, J.. AND SOUTH. P. G. (1977). Design and analysis of randomized clinical trials requiring prolonged observation of each patient. Brif. J C‘unwr 35, l-39. PHELPS. J. J.. GANDOI.FI. A. J.. BRENDEL. K.. AND DORR. R. T. ( 1987). Cisplatin nephrotoxicity: In taifro studies with precision-cut rabbit renal cortical slices. 7‘~~.~ico/.
REINBERG,
.-lpp/.
Pharttzu~~o/.
90. 50 I-5
A.. AND SMOLENXY.
logical Rlr~~thtt~.~ Plf!,.sio~par/2o/t!~ic..
uttd und
12.
M. (Eds.)
(1983).
Bio-
,2frdic~rtw Ccll~lur. Mcabtlic~, P/lurtt2ac,o/~t,~i~ .-I TJWC[~. p. 305.
Springer-Verlag. New York. SCHEVINC;. L. E.. PA~!LEY, J. E.. TSAI. T. H.. SCHEVINC;. L. A. (1983). Chronobiology of cell proliferation implications for cancer chemotherapy. In Bio/o~~icul Rh~fhttn putholo+jc..
uttd
:2/c,dic~itw
Ccll&w.
:\letuholic,.
Ph~~.sio-
u?lcl Phurtnuwlqpic :lsj~,c~t.c (A. Reinberg and M. H. Smolenshy Eds.). pp. 79-l 30. SpringerVerlag. New York. SIDDlh. Z. H., BOXALL. F. E.. AUD HARRAP. K. R. ( 1987a). Haematological toxicity of carboplatin in rats. Urir. .I. C’UWCY 55. 375-379. SIDDIK. Z. H.. NEWEL-I. D. R., BOXALL. F. E., HARRAP. K. R. (1987b). The comparative pharmacokinetics of carboplatin and cisplatin in mice and rats. Biochcw Phurtwcol.
36.
1925-
1937.
VAN ECHO. D.A.. EGORI. M. J.. WHITACRE, M. Y.. OLMAN. E. A.. AND AISNER. J. (1984). Phase I clinical and pharmacologic trial ofcarboplatin daily for 5 days. C’unc~c~r Twut Rep. 68, 1 103- I 104,
AND APPLIED PHARMACOLOGY
96.233-247
Circadian Time Dependence NACEURA.BOUGHATTAS,*-~~FRANCIS ANNIE
ROULON,~
CHARLES
(1988)
of Murine Tolerance Lt%r,*.t FOURNIER,$
for Carboplatin’
BERNARDHECQUET,$GUYLEMAIGRE,§ AND
ALAIN
REINBERG*
*L/.-l CNRS 581 “Chronobiologie-Chronopharmacologie” Fondatiorz ‘4. de Rothschild, 75019 Paris, France, ?Pharrnacologie 2, SMST and ICIG (CNRS UA 04-l 163). H&pita1 Paul-Brousse. 94800 r ‘illejuif: France; $Lahorutoire de Pharmuc~odynamie, Centre Oscar Lambret. 59000 Lille, France: and $Service dI4natomie et de C)fologie Pathologiyue. H6pital.4ntoine Bkl&e. 92000 Clarnart. France
Received December 22. 1987; accepted June 18. 1988
Circadian Time Dependence of Murine Tolerance for Carboplatin. BOUGHA~TAS, N. A.. L&, F., HECQUET, B., LEMAIGRE, G., ROULON, A., FOURNIER. C.. AND REINBERG, A. (1988). Tosicol. Appl. Phermacol. 96,233-247. A large amplitude circadian rhythm in murine tolerance for the anticancer agent, carboplatin (cyclobutane dicarboxylatoplatinum 11, CBDCA) was demonstrated. Two studies were performed in a total of 266 male B6D2Fl mice standardized by LD 12: 12. In the first experiment CBDCA (80 mg/kg/day) was administered intravenously (iv) daily for three consecutive days at all six circadian stages (3, 7. 11, 15, 19, or 23 hr after light onset, HALO). CBDCA dosing at I5 HALO resulted in 58% long-term survivors as compared to 0% after treatment at 3 or 23 HALO (x’ = 28; p < 0.001). In the second experiment. CBDCA (72 or 80 mg/kg/day X 3 days, iv) was administered at any of three circadian stages (0, 8, or 16 HALO). Mice were killed. blood was collected, and seven tissues were obtained 5 and 10 days after the first dose. in order to determine serum urea and creatinine concentrations, leukocyte and red blood cell counts. and to evaluate histologic lesions. No renal toxicity was encountered. Bone marrow and colon mucosa were the major target tissues of CBDCA in these dosages and schedules. CBDCA dosing at 16 HALO was least toxic to the bone marrow as assessed by peripheral leukocyte count and histologic score (p from ANOVA ~0.05). Histologically assessed lesions of the colon mucosa were less severe after CBDCA dosing at I6 HALO as compared to those at 8 HALO, and significantly so for the lowest dosage tested (p - 0.05). Uptake of CBDCA 24 hr after the third dose ranged from 23 @g/g of dry tissue in the colon to 7 pg/g in the duodenum. Mean tissue concentrations increased between Day 4 and Day 10 for the liver and spleen. and remained similar for the kidney. No consistent circadian dependence was found with regard to Day 4 mean Pt uptake in different tissues. whereas the lowest Day IO Pt concentrations corresponded to CBDCA dosing at 16 HALO for all tissues investigated. Toxicity did not appear to be directly related to the total platinum concentration in these tissues. s 19x8 ACKIC~IC P~CSS.I~C.
Murine tolerance for over 20 antineoplastic drugs depends upon their dosing time in the 24-hr scale(L&i er al., 1988a). This applies in particular to ci.c-dichlorodiammineplatinum ’ Supported in part by Grant 6 180 from the Association pour la Recherche sur le Cancer, Villejuif and by Bristol Laboratories, Paris, France. ‘To whom correspondence should be addressed at Pharmacologic 2. I.C.I.G.. H&pita1 Paul-Brousse, 14, av. Paul-Vaillant-Couturier. 94804 Villejuif C&dex. France. 233
II (CDDP) (Hrushesky et al.. 1982) and to the alkylating agents, cyclophosphamide (Haus et al., 1974) and peptichemio (L&vi et ul.. 1987). CDDP-induced hematologic and renal toxicities were lessenedby administering this drug in the activity span of rats (L&i et al., 1982a; 1982b). Carboplatin (cyclobutane dicarboxylatoplatinum II, CBDCA) is a new platinum analog, which lacks renal toxicity but exhibits similar antitumor effects as CDDP, mainly 0041-008x/88 Copvr&t C
$3.00
1988 hy Academic Press. Inc All rights of reproduction in am form reserved.
234
BOUGHATTAS
ET
TABLE
AL.
I
MAIN CHARACTERISTICS OF THE Two STUDIES INVESTIGAT [NC; THE CIRCDIAN ANCE FOR CBDCA (70 OR 80 mg/kg/day iv X 3 days) IN MAL.E B6D2FI MICE AGED FOR3WEEKSByLDl?l?
Study
Dates of study 10/7/86
219187
Nof mice Cont. treat.
Time points studied (HALO”)
VARIABILITY IO WEEKS.AND
Endpoint
IN HOST TOL.FRSYUCHROUIZEU
Day of assessment
6
I14
3. 7. I I. 15. 19.23
Day 40 survival Body weight loss Rectal temperature
Each day for 40 days Every 3rd day for 4 weeks Days I, IO. 18. and 28
36
107
0. 8. 16
Body weight loss RBC, WBC count Serum urea & creatinine Histology of 7 tissues” Platinum concentrations 6 tissues
Day 5 and Day 5 and Day 5 and Day5and Dayiland
” Hours after light onset. ’ Kidney. liver. spleen. bone marrow.
duodenum.
jejunum.
through interactions with DNA (Harrap ef al., 1980; Mong et ul.. 1980a. 1980b: Knox rt nl.. 1986). Its dose-limiting toxicity is bone marrow suppression (Calvert et al.. 1982; Curt et al., 1983: Evans et ul., 1983: Van Echo rf al., 1984: Koeller ef al.. 1985). The present investigation is aimed at assessing the role of dosing time upon CBDCAinduced toxicity. and at isolating possible
in
IO 10 10 IO IO
and colon
mechanisms of such rhythms. Results were subsequently expected to guide the chronopharmacologic optimization of the use of CBDCA in cancer patients, as was already achieved for cisplatin and other anticancer agents (Hrushevsky, 1985: Levi, 1987). Mortality, body weight loss, rectal temperature, leucopenia, and histologic lesions in seven tissues were used as criteria for toxicity.
TABLE
2
CIRCADIAN RHYTHM IN TOL.FRANCE OF MICE FOR CBDCA: RESULTS FROM COSINOR ANALYSIS. WITH A PERIOD T = 24 HR
Variable Synchronization Rectal temperature (“C) Day I Toxicity Rectal temperature (“C) Day 10 Body wt loss (aI) Day 10 + Survival at 50% overall mortalitv $ Long-term survivors Survival time (davs)
h’ data
120
87 87 6 6 114
Mesor
P”
? SEM
Double
amplitude
+ SD
Acrophase (HALO) * SD (min)
37.8 + 0.1
1.5 2 0.5
10.03 10.00
I
35.8 -t 0.2 20.7 k 0.8
1.X k 1.6 18.5 -+ 6.0
15.00 t 155 13.10* 70
54.2 -t 7.2 21.7 k4.7 17.8 t I.0
XI.0 55.5 17.4 + 7.1
13.20 15.20 14.50 k 90
,Ll’ot~. HALO. hours after light onset. ” From an b’test of the rejection of the null amplitude
hypothesis.
19.50 k
80
CIRCADIAN
TOLERANCE
FOR
235
CARBOPLATIN Key
ln,ectm
Tvne
Logrank
(Halo) 15 19 11
p
36 05 d 125
co
005
07 03
. .. . .. ..
71
test
x’
23
I
3
6
11
14 TIME
16 FOLLOWING
23 CBDCA
TREATMENT
26
31
34
39
(days,
FIG. I. Survival curves of B6DZFl mice over the 40 days following the iv administration of 80 mg/kg/ day x 3 days of carboplatin (CBDCA). Each survival curve corresponds to a different dosing time of CBDCA along the 24-hr scale (study 1). Survival rate differed as a function of dosing time both when 50%) overall mortality was reached (x2 = 45: p < 0.00 I ) and on Day 40 (x’ = 18: (I< 0.00 I ).
Platinum uptake in six of these tissues was also examined as a function of treatment time. MATERIALS
AND
METHODS
Two studies were carried out on a total of 166 1O-weekold male B6D2FI mice (IFFA-CREDO. L’Arbresle. France) in October 1986 and February 1987. Mice were housed three per cage, with free access to food and water. and were synchronized at least 3 weeks prior to each study (Reinbergand Smolensky. 1983) in a chronobiologic animal facility (ES1 Flufrance, Arcueil, France). The lighting regimen was 12 hr of light (L) and 12 hr of darkness (D) (LD 11: 12) with light intensity equal to 300 + 50 Ix. Each facility had six sound proof, temperature-controlled (‘73 * 1°C) compartments with liltered air (100 -+ IO liter/min), and each compartment had its own programmable lighting regimen. This design allowed the performance of six circadian stages of CBDCA dosing (3, 7, I I. 15. 19. and 23 hr after light onset. HALO) in study 1 and three circadian stages in
study 2 (8. 16. and 34 HALO). All drug administrations were between 9:00 and 1200 hr. Synchronization of circadian rhythms of mice was checked prior to drug dosing by measuring the rectal temperature of each mouse. 2. I)nr!: CBDCA was kindly supplied by Bristol-Myers (France) in vials containing 150 mg of lyophilized (‘w diammine ( I. I -cyclobutane dicarboxylato)platinum and 150 mg of mannitol. The CBDCA solution was freshly prepared on each study day by adding an adequate volume of distilled water in order to obtain the desired concentration. Two doses were administered to mice (71 or 80 mg/kg/day *: 3 day) in a fixed fluid volume (5 ml/kg body wt) which corresponded to the CBDCA concentrations of 14.4 and 16 mg/ml. respectively. These doses are the LDIO and LD50. respectively (unpublished results). Each mouse was given an intravenous injection ofeither the drug solution or the 80 mg/kg of mannitol into the retro-orbital sinus daily for three consecutive days.
The study designs rized in Table I.
and toxicity
endpoints
are summa-
236
BOUGHATTAS
ET AL.
A
12 DOSE:
9.
72
mglkgldx3d
DOSE:00mglkg
i.v
/d=3d
i.v
T
6.
3.
TIME
OF
(hours
INJECTION
after
light
onset
1
ANOVA
. . . _ . _ _.
CEDCA
72mg/kg
CEDCA
80mglkgid
/d I v.3d I v.3d
COSINOR
F
P
P
106
0001
0001
46
032
002
t +25
. CONTROL
-25.
3 03
07
TIME
(hours
II
15 after
19 light
onset
23 )
MESOR
( 72 m g /kg.3)
MESOR
(80 m g lkg.3)
CIRCADIAN
TOLERANCE
In study I. the survival rate was computed for each group at 50% overall mortality and at study completion on Day 40. Body weight loss was computed as percentage change for each mouse. In study 2, leukocyte and red blood cell counts were determined in blood (-500 ~1) obtained from the retroorbital sinus. These cells were quantified in a Z.M. Coultronics Coulter counter according to standard procedures. Serum urea and creatinine concentrations were determined by spectrophotometry (Uree enzymatique UV H.P. Biotrol; and Creatinine cinetique Biomerieux. France). H~.rrc’(~urhr,lo~j~, srlrdj,. Immediately after being bled, the mice were killed. Liver and spleen were immediately fixed into Carnoy’s solution; kidney. duodenum, jejunum. colon, and bone marrow (femur) were fixed into Bouin’s picroformol solution. Twenty-four hours later. these organs were dehydrated and embedded into parahin. Sections were stained with hematein-eosin. Each encoded slide was examined by the same histopathologist who was unaware of the treatment groups. Lesions were graded between 0 (normal) and 4. The highest grade (4) corresponded to complete necrosis for the bone marrow, to necrosis of the superficial part of the villi for the small howel, and to a deep disorganization of epithelial structures of the colon with dedifferentiation. glandular. dilatation. and inflammatory infiltrate. The reliability of such a scoring system has previously been documented (Levi (‘I ul.. 1987). T;F.YI(Y Ji.~~~ihu(iolr of p/utirrrr/?r Tissue distribution of (‘BDCA was studied in 36 mice (3 mice/time point/dose) on Day 4 or IO following the first dose of 73 or 80 mg/ kg/day at 8. 16. or 24 HALO. After mice were killed, liver. spleen, kidney. duodenum. jejunum. and colon were sampled and immediately weighed. then air-dried to a constant weight. Dry tissues were digested in 1 ml of nitric acid for I hr at 100°C. After acid evaporation Pt tissue content was measured by tlameless atomic absorption spectrophotometry (FAAS) with electrothermal atomization (ETA) using a Perkin-Elmer Model 1280 IHecquet (ll al.. 1983). The following time temperature program was used: drying for 30 set at I XC, ashing for 30 set at 14Oo”C, and atomization for 3 set at 2700°C. .kuri~ical anu/wis. Means and standard errors of the mean (SEM) were calculated for each variable. dose. and time point. The statistical significance of differences bctween groups was validated for parametric data by one-
FOR
337
CARBOPLATIN
or two-way analysis of variance (ANOVA). Taking into account the non-normal distribution, platinum concentration was also estimated in each tissue using medians (hl) and coefficients of dispersion (CD) according to the following formulas (De Prins
/ Ct - .If ( /0.6745
CD = ICti - I%/1/[(0.798(h’
- l)]
when
,I 2 6
when
/I < 6,
where Ct is tissue concentration. Differences in tissue concentrations were then analyzed by nonparametric tests (Mann-Whitney test and/or Kruskall-Wallis ANOVA). Differences in survival rates were analyzed by x2 test and survival curves by logrank test (Peto rf al.. 1977). Time series were analyzed for circadian rhythmicit! (with a period, T. 214 hr) by the cosinor method (Nelson (‘1 al.. 1979). A rhythm was characterized by three parameters: the mesor, M (24 hr-adjusted mean), the double amplitude, 2A (difference between the minimum and the maximum of the fitted cosine function). and the acrophase. 4 (time of maximum, with time of light onset as 9 reference). M, ?A. and + were obtained with their 95% confidence limits, if a rhythm was detected. This was achieved when A differed from zero (non-null amplitude rtest) with p 4 0.05.
RESULTS StlKiJ I Results from cosinor are shown in Table 3.
analysis
of the data
Rectal tempc~ratm~. A circadian rhythm was statistically validated by cosinor on Day 1. Thus, mice were well synchronized prior to dosing with CBDCA. The acrophase was localized near the middle of the dark span as is usually the case. Treatment resulted in hypothermia. which was maximal on Day 18 (mean, - 1 1%). Temperature decrease varied as a function of dosing time (-20% at 3 HALO: -9%’ at 15
FIG. 2. (A) Mean circulating leukocyte count in control and CBDCA-treated mice 5 and 10 days after the first of three consecutive doses (72 or 80 mg/kg/day X 3 days). Mice were treated with CBDCA and sampled at the same circadian stage. Three circadian stages (0, 8, and 16 hr after light onset, HALO) were compared with regard to the effect of CBDCA upon such endpoint. (B) Leukopenia index. percentage change in circulating WBC count (relative to time-qualitied controls) on Day 5 and 10 after treatment with CBDCA (72 or 80 mg/kg/day .i 3 days iv). Dose and dosing-time dependence of leukopenia index as modeled by cosinor analysis.
238
BOUGHATTAS
ET AL.
CIRCADIAN
TOLERANCE
FOR
CARBOPLATIN
339
Ftc;. 3. Transverse section of femoral bone marrow 5 days after first dose of mannitol (control) or 72 mg/kg of CBDCA at either 08.00 or 16.00 HALO. Hematein-eosin staining. (4) Control-histologic score = 0 (magnitication x85). (B) CBDCA at 8 HALO-important medullar necrosis score = 3 (~85). (C) CBDCA at 16 HALO moderate decrease ofcellular population. histologic score = I (K X5).
HALO). The changes in rectal temperature followed a circadian rhythm. Rectal temperature of survivors returned back to pretreatment values 4 weeks after the start of treatment (on Day 28, mean change - I %). &Y/J* ~tx4ghf c$arz,~~~.Mean maximal body weight loss was achieved on Day 6 after the tirst dose of CBDCA and remained unchanged until Day 18 (- 19% of the initial body weight). On Day 10 mean body weight loss varied between 1 I %, in mice dosed at 15 HALO and 34% in those injected at 23 HALO (F = 20.2; 11< 0.01). Among the 25 mice surviving on Day 38. those who had received CBDCA at I5 or 19 HALO had recovered their initial body weight (respective mean body weight change, -8% and +%), whereas those
treated at 7 HALO had not (-20%‘) (F from ANOVA = 3.6; p = 0.07). Survivd. Deaths were observed between Day 5 and 25. Survival rate remained subsequently unchanged. Survival curves ofthe six studied groups were computed according to Kaplan-Meyer method (Kaplan and Meyer. 1958). Percentage of survival differed significantly as a function of both CBDCA dosing time (circadian stage) and time span following treatment (days) (Fig. 1). Fifty percent overall mortality was reached on Day 12. On this day, survival rate varied between 95% in those mice dosed at 15 HALO and 0%’ in those treated at 23 HALO. The long-term survival rate varied between 58% ( 15 HALO) and 0% (3 and 23 HALO). Circadian rhythms in survival rates and survival times were also detected by cosinor with acrophases localized
240
BOUGHATTAS
O-
06
TIME
16 (hours
after
24 light
onset
I
FIG. 4. Histologic scores of bone marrow 5 and IO days after the first iv dose of CBDCA. Three circadian stages of CBDCA injection are compared. A dosing time-related effect was statistically validated on Day 5. Minimal lesions corresponded to dosing at I6 HALO. On Day 10, maximal lesions were found (mean score = 3.9) whatever the circadian stage.
near 15 HALO (Table 2). The difference between mean longest and shortest survival times was 19 days, with optimal dosing time being 15 HALO (F from ANOVA = 9.0; p < 0.001).
Study ,7 Bonl,) Mw+ght and survival. On Day 5, body weight loss was similar to that observed in study 1 (19% of the initial body wt). On Day 10. body weight loss varied between 27% in mice dosed at 16 HALO and 33% in those injected at 8 HALO (F = 4.1, p = 0.02). Thirteen percent of those mice receiving CBDCA at 0 or 8 HALO had died from toxicity before Day 10. Inlmunohnnatolo~ic toxicit?: In controls circulating RBC count remained similar whatever its sampling time or sampling day (mean f SEM = 10.6 + 0.3 X lo6 cells/mm’). CBDCA resulted in a moderate decrease in circulating RBC count on Day 10 (F = 9.6; p < 0.005) irrespective of the treatment dose and timing (8.5 + 1.1 X IO6 cells/mm3).
ET AL.
Mean circulating leukocyte count varied as a function of sampling time in controls, and irrespective of sampling day. A marked circadian rhythm was detected by cosinor (y < 0.001; 4 = 5.30 f 1.20 hr). Following CBDCA treatment, the circulating WBC count decreased, more so on Day 10 as compared to Day 5 (Fig. 2A). Because of the physiologic circadian rhythm in circulating total WBC count, data from treated mice were also expressed as percentages of the corresponding mean time-qualified control values irrespective of sampling day (leucopenia index). Mean leucopenia ranged from 37% after dosing with CBDCA (80 mg/kg/day X 3 days) at 16 HALO to 77% after drug dosing at 8 HALO (Fig. 2B). A dose response was documented both for raw data and for transformed data (Fig. 2). A severe bone marrow necrosis was observed on Day 5 in treated mice. The extent of such necrosis varied as a function of both dose (1; = 3.1: p from ANOVA -0.05) and dosing time (Fig. 3). The lowest histologic score occurred at 16 HALO and the highest at 8 HALO (Fig. 4). A circadian rhythm in the bone marrow score was further demonstrated by cosinor. On Day 10, a complete necrosis was apparent in all specimens whatever the dosing time (Table 3). Ir2tedinul to.uicitj! Intestinal lesions mainly affected the colon mucosa. Two days after the last CBDCA injection (Day 5). a deep disorganization of epithelial structures was observed with dedifferentiation. inflammatory infiltrate of chorion, and kryptic abscess (Fig. 5). Five days later (Day 10) most colic lesions were repaired, since a partial recovery of mucous secretion was observed. A dose-response relationship was statistically validated both on Day 5 and on Day 10. Whereas no effect of dosing time was apparent on Day 5. recovery of colic lesions appeared to be faster after CBDCA dosing at 16 HALO, and significantly so for the lower dose used (Table 3).
CIRCADIAN
TOLERANCE
FOR
TABLE Toxrcir~
OF CBDCA
241
CARBOPLATIN
3
INMICEFORBONEMARROWANDCOLONMUCOSA,ASASSESSEDBYHISTOLOGICSCORES Two-way Time
Tissue
Dose” bWc&W
Bone marrow
72
80
Day
72
intravenously
00
08
16
F
P
F
2.2 f 0.3 3.5 + 0.3
3.8
0.03
3.00
-co.00
1
1
2.6 k 0.2
3.2 * 0.2
4.0 + 0
3.8 k 0.2
5
2.8 k 0.2
3.3 -+ 0.2
3.9iIO.l
4.0 + 0
2.8 f 0.2 4.0,o
3.8
0.03
8.6
2.8 f 0.5 1.x+0.7
2.7 f 0.3 1.2 + 0.4
3.0 i 0.6 0.3 f 0.3
3.1
0.06
16.2
3.0 f 0.4 1.2 f 0.4
3.2 +- 0.4 2.4 to.1
3.0 f 0.6 1.9 kO.5
0.8
0.5
5
5 I0
” Injected
Day
5
IO
80
Time
IO
10 Colon
(hr after light onset)
ANOVA
for three consecutive
7.0
P
10.00 1 0.01
days.
Minor lesions were observed in the duodenum or the jejunum. None was found in liver, spleen. or kidney. Serum urea and creatinine were similar in controls and treated mice. Platinum distribution in tissues. Twentyfour hours after the third CBDCA dose (Day 4), mean highest platinum concentrations were observed in the colon and kidney. Platinum uptake was about two times less in liver, spleen. and jejunum and three times less in duodenum (Fig. 6). On Day 10, mean tissue platinum concentration decreased three- to sixfold in duodenum, jejunum, and colon as compared to their respective values on Day 4. Conversely platinum concentrations increased over this 6-day span both in the liver and in the spleen, but were unchanged in the kidney. An effect of CBDCA dosing time was apparent on the platinum concentration in all tissues on Day 10, with lowest tissue concentrations corresponding to drug dosing at 16 HALO (Fig. 7). With respect to tissue platinum concentration, results of three-way ANOVA, (dose, sampling time, sampling day) and nonparametric Kruskall-Wallis ANOVA validated an effect of sampling day for all tissues except the kidney, an effect of
CBDCA dose for spleen, duodenum, jejunum, and colon, and an effect of dosing time for spleen, duodenum, and jejunum.
Correlation hettileen platinum distribution and tissue toxicit)). Median Pt uptake on Day 4 correlated with the median colon toxicity score on Day 5 (r = 0.79; df= 5; p < 0.05). No correlation was found between the median Pt concentration and the median colon score on Day 10 (r = 0.38; df= 5;~ > 0.10). However, the correlation between Pt uptake on Day 4 and Day 10 histologic score was close to statistical significance (r = 0.70; df = 5: p - 0.07). DISCUSSION Optimal tolerance for CBDCA was achieved when this drug was given during the activity span of mice near 15 HALO, whatever the endpoint studied (changes in rectal temperature or in body weight, survival). The adequate animal synchronization by the six different LD 12: 12 schedules was demonstrated by circadian rhythms in rectal temperature and circulating leukocytes. as previously reported (Haus et al., 1983; Levi et al.. 1987; Levi et [II.. 1988a). Necrosis of both
CIRCADIAN
TOLERANCE
FOR
243
CARBOPLATIN
Sampling 4
colon P (Mann-whitney)
kidney 0.2
liver
*pII?=?”
ieimum
duodenum
< 0.01
(24hr
day pos.1
3’ddose)
(NS)
FIG. 6. Histogram of Pt concentration (median * coefficient of dispersion, CD) after intravenous CBDCA dosing (72 or 80 mg/kg/day > 3 days). Six tissues were obtained 24 hr after the third dose of CBDCA (Day 4) or 6 days later (Day IO).
bone marrow and colon mucosa constituted the main mechanisms ofthe lethal toxicity of CBDCA. since no significant lesion was found in any other organ studied. Irrespective of CBDCA dose or dosing time, a greater decrease in total WBC count was observed on Day 10 as compared to that of Day 5. in accordance with previous results (Siddik PI al.. 1987a). Bone marrow necrosis was also more severe on this day. Similar results were reported in mice using the survival of hematopoietic stem cells (CFU) in the spleen colony assay (Lelieveld rl al.. 1984). Conversely, colic lesions were more pronounced on Day 5 as compared to those of Day 10. The lowest damage in colon and bone marrow and the least hematologic toxicity corresponded to CBDCA dosing at 16 HALO. Drug-induced leucopenia was 50%) less severe after treatment at this circadian time, as compared to that of 0 or 8 HALO. This may partly be accounted for by the circadian rhythm which characterizes cellular proliferation. More spe-
cifically, the proliferative ability of bone marrow granulomonocytic precursors is highest in the dark (activity span) and DNA synthesis is increased in the light (rest span) of mice (Sheving et al.. 1983; Levi et al., 1988b). In addition, cell proliferation in mouse colon displays a marked circadian rhythm. The circadian peaks of both M and S phasesof the cell division cycle occurred during the rest (light) span (Scheving et al.. 1983) with the mitotic index being maximal near 8 HALO (Chang, 1971; Chang and Nadler. 1975: Hamilton, 1979). Maximum Pt concentrations were measured in the colon and kidney 24 hr after the last CBDCA dose, as previously observed (Litterst, 1984: Boven et al.. 1985). On Day 10 as compared to Day 4, a significant increaseof Pt content occurred in the liver and in the spleen. This may be due to irreversible Pt binding both to nucleophilic sites of macromolecules and to plasma proteins (Harland c/t nl.. 1984: Gaver cf N/., 1986: Laznickova
FIG. 5. Transverse section ofcolon 5 days after first dose of 80 mg/kg of mannitol or CBDCA. Hemateineosin staining. (A) Control. histologic score = 0 (magnification X85). (B) CBDCA at 8 HALO, deep disorganization ofepithelial structures with dedifferentiation. glandular dilatation. inflammatory infiltrate ofchorion. and kryptic abscess. Score = 4 (/85).
244
BOUGHATTAS
ET AI
34
30
26
16
2:
TIME of CBCCA (Hours Aftor
INJECTION Light 0nm.t)
FIG. 7. Mean Pt concentration measured in six tissues. IO days after the first dose of intravenous (72 or 80 mg/kg/day x 3 days) at any ofthree circadian stages.
rt ul.. 1986; Siddik et al.. 1987b). Kidney Pt concentration remained similar on Day 10 as compared to that of Day 4. Such sustained concentration may result from covalent
CBDCA
binding of Pt to intracellular components in the kidney. Conversely, the marked decrement of Pt concentration in the colon from Day 4 to 10 paralleled the recovery of muco-
CIRCADIAN
TOLERANCE
sal lesions. This may result from the renewal of damaged mucosal cells. Such a relationship between the Pt concentration and the extent of mucosal damage was also suggested along the circadian time scale. Time-dependent tissue distribution was apparent on Day 10 for all tissues investigated with 16 HALO corresponding to the lowest mean Pt concentration. On Day 10, colic lesions were also less after CBDCA dosing at 16 HALO as compared to those at 8 HALO. Furthermore a statistically significant correlation was observed between Pt uptake on Day 4 and colic lesions on Day 5. Nonetheless. no apparent correlation was documented between Pt uptake and tissue toxicity in the other tissues studied, since the highest Pt uptake was found in the kidney. where no lesion was documented. The lack of any active urine secretion or tubular reabsorptive process (Daley-Yates and McBrien. 1983: Siddik et nl.. 1987b) for CBDCA and its greater chemical stability (Cleare et (I/.. 1980: Howe-Grant and Lippard, 1980) may explain its apparent lack of nephrotoxicity (Daskal (‘t cl/., 1980; Harrap rt ~1.. 1980: Cal\.ert et al., 1982: Egorin et ~1.. 1984; Batzer and Aggarwal, 1986) as compared to that of CDDP. Thus glomerular filtration is the main mechanism of CBDCA excretion in the rat (Harland et uI.. 1984: Siddik ~1u/.. 1987h). This function exhibits a well-known circadian rhythm with maximal values occurring near the middle of the activity span of rats or mice (Cal et ~1.. 1986). This may account for an increased excretion of this drug near I6 HALO. Nonetheless. platinum accumulation in the kidney following repeated CBDCA dosages may lead to some moderate degree of renal toxicity. as was reported both in rats (Levine et al.. 198 1; Lelieveld t’t u/.. 1984) and recently h vitro (Phelps et ~1.. 1987). CBDCA and CDDP exhibit different toxic and pharmacologic properties. However, the circadian dosing time associated with optimal tolerance appears to be similar for both platinum analogs. Further studies are needed
FOR
CARBOPLATIN
245
to assess whether this also applies to their antitumor effectiveness. Let us emphasize that CDDP was significantly more effective against a transplantable rat immunocytoma when dosed at the circadian time when it was best tolerated (Levi (‘I al.. 198 1). The relevance of the mouse model to predict the least toxic dosing time of anticancer agents in patients has already been documented for adriamycin (Hrushevsky, 1985). 4’-tetrahydropyranyl-adriamycin (L&i, 1987). and CDDP (Hrushevsky, 1985; Levi, 1987). CBDCA was best tolerated by mice after dosing it near 16 HALO. This may correspond to 15.00- 16.00 hr in cancer patients.
ACKNOWLEDGMENTS The authors express their thanks to G. Metzger. M. Mechkouri. and M. ltzhaki for-statistical assistance. to E. Co& for secretarial assistance in the preparation of the manuscript. and to Z. Khalouche for iconographic world.
REFERENCES BATZER. M. A.. AND AC;CARWAL. S. K. (1986). An in vitro screening system for the nephrotoxicity of vatious platinum coordination complexes. C~rrc~cr C%ctnc~ther. Ptwmac~~d 17, 209-2 1 I. BOVEN, E.. VAN DER VIJGH. W. J. F.. NAUTA. M. M.. SCHLUPER. H. M. M.. AND PINEDO. H. M. (1985). Comparative activity and distribution studies of five platinum analogues in nude mice bearing human ovarian carcinoma xenografts. C’upr(,er Rex 45, 86-90. CAL. J. C., DORIAN, C.. AND CAMBAR. J. (1986). Circadian and circannual changes in nephrotoxic effects of heavy metals and antibiotics. In .lnr~ru/ Kurirlc- o/ (‘hronophurrnacol~~~~~~ (A. Reinberg. M. Smolensky and G. Labrecque Eds.), Vol. 2. pp. 343-176. Pergamon. Oxford. New York. CALVERT. A. M.. HARLAND. S. J.. NEWELL. D. R.. SIDDIK, Z. H.. JONES, A. C.. McEr WAIN, T. J.. RAJU. S.. WILTSHAW. E.. SMITH. I. E.. BAKER. J. M.. PECKHAM. M. J.. AND HARRAP, K. R. (1982). Early clinical studies with cis-diammineI. I cyclobutane dicarboxylate platinum II. C‘unwr C’twmorhzr. Ptmrrmm~i. 9, 140147.
CHANG. W. W. L. ( 197 I ). Renewal of the epithelium in the descending colon of the mouse. III. Diurnal variation in the proliferative activity of epithelial cells. .Irnw J. :lnat. 131. 1 I I-120.
246
BOUGHATTAS
CHANG, W. W. L.. AND NADLER. N. J. ( 1975). Renewal of the epithelium in the descending colon in the mouse. IV. Cell population kinetics of vacuolated-columnar and mucous cells. .-lrner. J .Jnal. 144, 39-56. CLEARE. M. J.. HYDES. P. C.. HEPBURN. D. R.. AND MALERBI. B. W. (1980). Antitumor platinum complexes: Structure-activity relationships. In C’is&zt~rl. C’urrmr Stutm und NW Devcloptnctm (A. W. Prestayko. S. T. Crooke, and S. K. Carter Eds.), pp. 149170. Academic Press. New York. CURT, G. A.. GRYGIEL. J. J.. CORDEN. B. J., OZOI s. R. F., WEISS, R. B., TELL. D. T.. MYERS. C. E.. AND COLLINS. J. M. (1983). A phase I and pharmacokinetic study of diammine cyclobutane dicarboxylato platinum (NSC 24 1340). Cuttcrr Rev. 43, 4470-4473. DALEY-YATES. P. T., AND MCBRIEN. D. C. H. (1983). The mechanism of renal clearance of cisplatin (cisdichlorodiammine platinum II) and its modification by furosemide and probenecid. Biohtx I-‘llurrr~ac,ol. 31.2243-1246. DAXAL, Y.. PRESTAYKO. A. W.. AND CROOKE. S. T. (1980). Morphological manifestations ofcisplatin analogs in rats. And ultrastructural study. In C’i.sp/ufin (A. W. Prestayho. S. T. Crooke. and S. K. Carter Eds.). pp. 249-269. Academic Press. New York. DE PRINS. J.. CORNELISSEN. G.. AND MALBECQ, W. (1986). Statistical procedures in chronobiology and chronopharmacology. .-innu. Km,. C%rott~,~/turtttut,f)/ 2,27-141. EGORIN. M. J.. VAN ECHO, D. A.. TIPPING. S. T., 01.. MAN. E. A.. WHITACRE. M. Y., THOMPSON. B. W., AND AISNER. J. ( 1984). Pharmacokinetics and dosage reduction of cisdiammine ( 1.1 -cyclobutane dicarboxylato) platinum in patients with impaired renal functions. Cuttwr Rcs. 44. 5432-5438. EVANS. B. D.. RAJLI. K. S.. CALVERT, A. H., HARLAND. S. J.. AND WILTSHAW. E. (1983). Phase II study of JM8. a new analog. in advanced ovarian carcinoma. Chnwr 7’rcwf. Rep. 67, 997- 1000. HAMILTON, E. ( 1979). Diurnal variation in proliferative compartments and their relation to ctyptogenic cells in the mouse colon. Cc,// Tir.slrr> Kine/. 12. 9 I-100. HARL.&ND. S. J.. NE~EI.L, D. R.. SIDDIK. 2. H., CHADWICK. R., CALVERT. .A. H.. AND HARRAP. K. R. ( 1984). Pharmacokinetics of ci.r-diammineI,1 -cyclebutane dicarboxylate platinum (II) in patients with normal and impaired renal function. Cunccr Rev. 44, 1693-1697. HARRAP. K. R.. JONES. M.. WILKINSOIU‘. C. R.. CLINK. H. M.. SPARROW. S.. MITCHLE\. B.. CLARKE, S.. AND VEASEY. A. (1980). Antitumor, toxic. and biochemical properties of cisplatin and eight other platinum complexes. In Cisplutit?. CWTWZ/ Stuizrs und NW Dewlopmwt.s (A. W. Prestayko. S. T. Crooke. and S. K. Carter Eds.). Vol. 1’. pp. 193-112. Academic Press, New York.
ET AL. HAUS, E.. FERNANDES. G., KUHL. J. F. W.. YUNIS, E. J.. LEE, J. D.. AND HALBERG. F. ( 1974). Murine circadian susceptibility rhythm to cyclophosphamide. C/tronohio/o,~iu 1. 270-277. HAUS, E.. LAKATUA. D. J.. SWOYFR, J.. AND SA~XEI-~LIJNDEEN, L. (1983). Chronobiology in hematolog> and immunology. :lmcr. J. .-lnu!. 168.467-5 17. HECQIIET. B.. ADENIS, L.. AND DEMAILLF. A. (1983). /II vitro interactions of TN06 with human plasma. (‘un(‘CT (%rtno~her. Pharmacol. I I, I77- 18 1. HOWE-GRAN’I. M. E., AND LIPPARD, S. J. ( 19X0). Aqueos platinum (II) chemistry: hinding to biological molecules. In die/a/ lon.7 in RIo/o~~I& S~:strtn.c(H. Sigel Ed.). Vol. 1 1. pp. 63-125. Dekker, New York HRUSHES~;~‘. W.. Live, F.. AND HALBERG, F. ( 1981). Circadian-stage dependence of (‘,r-diammine dichloro platinum lethal toxicity in rats. C’mwr Rcs. 42. 945949. HRLISHESK\I, W. J. M. ( 1985). Circadian timing of cancer chemotherapy. Smwc~ 228, 73-75. GAVFR. R. C.. SWIGOR. J. E.. G~ORC;E. A. M.. HA>.NE:s. Ll. S.. ANU DEEB. G. (1986). Protein binding and degradation of ‘“C-carboplatin in human plasma and urine in virrzJ. Prw .Attwr. Ls.~~tc C’utmr Rm. [Abstract No. I I3 I] KAPI AN. E. L.. AND MF\ TV. P. ( 1958). Non parametric estimation from incomplete observations. J .,lt?lcpr-. Stur. 1.s.s~~~53. 457-48 1 KOEI I.ER. J. M., TRUMP, D. L.. T~XCH, K. D.. EAR11~~1. R. H.. DAVIS. T. E.. AND TORME~. D. C. ( 1985). Phase I clinical trial and pharmacokinetics of carboplatin (NSC 241240) by single monthly 3%minutes infusion. C‘unwr57, X-X5. KNOX. R. J.. FRIEDI.OS, F.. IL>.DAI I., D. A.. AND RUBERTS. J. J. ( 1986). Mechanism ofcytotoxicity of anticancer platinum drugs: Evidence that cir-diammine dichloroplatinum (II) and c,ic-diammine-( I. l-cyclobutane dicarboxylato) platinum (II) differ only in the kinetics of their interaction with DNA. Cirtww Rec. 46, 1972-1979. LALNI~ROV~, A.. LAZNICEK, M.. Kvt-TINA. J.. 4~11 DROBNIK, J. (I 986). Pharmacokinetics and plasma protein binding oftwo platinum cytostatics CHIP and CBDC‘A in rats. Chttcw Chcvttotltcr. f’hurttw~~rtl. 17. 133-336. LFI IEVELD. P.. VAN DER VIJL;~I. W. J. F.. VELDML:IZEN, R. W., VA~V VEI.ZEN, D., VAU PII.I-‘IXN, L. M.. A r.4ss1. G.. AND DANGLI’Y. A. ( 1984). Preclinical studies on toxicity. antitumor activity. and pharmacokinetics of cisplatin and three recently developed derivatives. Eur. J. <‘utwr C’Ittr. Otmtl. 8, 1087- 1 104. L&I. F.. HALBERG. F.. NESBIT. M.. HAIIS, E.. AND LtVINE. H. ( I98 1 ). Chrononcology. In .Veoplustn.c C‘rwpurutiw Pu’Lltl~~ht~~~~0l‘Gtwwllt it] .hitmh. Plutl1.v. utzd ,lfu?f (H. E. Kaiser Ed.). Chap. 14. pp. 267-3 16. Williams & Wilkins, Baltimore. London.
CIRCADIAN
TOLERANCE
LEVI. F.. HRUSHESKY. W., BLOMQUIST,J.. LAKATUA, D.. HAUS, E.. HALBERG, F.. AND KENNEDY. 9. J. ( 1989a). Reduction of cis-diammine dichloroplatinum nephrotoxicity in rats by optimal circadian drug timing. Cancer Rev. 42, 950-955. LEVI. F.. HRUSHEXY, W.. HAUS, E.. HALBERC;, F.. AND KENNEDY. 9. J. (1989b). Circadian rhythm in lethal. renal. and hematologic tolerance for cis-diammine dichloroplatinum (II). E~rr. .I. Chnccr C‘lin. Om~tl. 18, 471-477.
LEVI. F. (1987). Chronopharmacology of anticancer agents and cancer chronotherapy. In Klinischr Plrurttzuko/o,qie (H. Kuemmerle Ed.). pp. l-17. Ecomed Press. Landsberg. Lfvl. F.. HORVATH. C.. MECHKOURI. M.. ROULON. A.. BAILLFUI.. F.. LEMAIGRE. G.. REINBERG. A.. AND MUHE, G. (I 987). Circadian time dependence of murine tolerance for the alhylating agent peptichemio. Fur- .I. C’unccr C/in. Oncol. 23(5), 487-497. LEVI. F.. BOUGF~A~~AS, N. A.. AND BUZSEK, 1. (1988a). Comparative murine chronotoxicity of anticancer agents and related mechanisms. In .drzn~~l Review r!/ C’krorlc)~/2unl?u(.ol~!~~. (A. Reinberg. M. Smolensky. and G. Labrecque. Eds.). Vol. 4. pp. 383-33 I, Pergamon. Oxford. Livi, F.. BLAZSEK. 1.. AND FERLE-VIDOVIC’, A. (1988b). Circadian and seasonal rhythms in murine bone marrow colony forming cells affect tolerance for the anticancer agent 4’.tetrahydropyranyladriamycin (THP). E.xp IIcmiIoI in press. I.~vIN~:. 9. S.. HENRY. M. C.. PoKr. C. D.. RI(.HIER, W. R.. AND URwNth, M. A. ( 198 1). Nephrotoxic potential ofcisdiammine dichloroplatinum and four analogs in male Fisher 344 rats. .7. Su//. (hnc,er 11~~1 67, 20 I-106. 1.11’1 ERST. C. L. ( 1984). Plasma pharmacokinetics. urinap excretion. and tissue distribution ofplatinum following i.v. administration of cyclobutane dicarboxylatoplatinum II and cis-platinum to rabbits. In Pkulit1utt1 C‘oorditlu/iott it? C’LUICW C’/wttwthoqt~~ (M. P. Hacker, E. B. Douple, and I. H. Krakoff Ed%). pp. 71XI. Martinus-Nijhoff. Hingham. MA. MONA;. S.. HCIANC;. C. H.. PRESTA~XO, A. W.. AI\;I) CROO~F. S. T. ( 1980a). Effects of second-generation
FOR
247
CARBOPLATIN
platinum analogs on isolated PM-2 DNA and their cytotoxicity irl ,)itro and in viro. (htlcer Res. 40, 33 I83324. MONG, S.. PRESTAYKO, A. W.. AND CROOKE. S. T. (1980b). 112 ,)itro interaction of covalently linked closed circular DNA with the second generation platinum compounds. In C’isplutitz. C’urretzf Status and h’e+l D~velopmolrs (A. W. Prestayko. S. T. Crooke. and S. K. Carter. Ed%), Vol. I?. pp. 213-226. Academic Press. New York. NELSON. W.. TONC;. Y.. LEE, J. K.. AND HALBERC;, F. ( 1979). Methods for cosinor rhythmometry. C7tnvw hiohgiu 6, 305-323. PETO, R.. PIKE. M. C., ARMIIAGE, P.. BRESLOW. N. E.. Cox. D. R.. HOWARD. V.. MANTEL. N., MCPHERSON. K.. PETO, J.. AND SOUTH. P. G. (1977). Design and analysis of randomized clinical trials requiring prolonged observation of each patient. Brif. J C‘unwr 35, l-39. PHELPS. J. J.. GANDOI.FI. A. J.. BRENDEL. K.. AND DORR. R. T. ( 1987). Cisplatin nephrotoxicity: In taifro studies with precision-cut rabbit renal cortical slices. 7‘~~.~ico/.
REINBERG,
.-lpp/.
Pharttzu~~o/.
90. 50 I-5
A.. AND SMOLENXY.
logical Rlr~~thtt~.~ Plf!,.sio~par/2o/t!~ic..
uttd und
12.
M. (Eds.)
(1983).
Bio-
,2frdic~rtw Ccll~lur. Mcabtlic~, P/lurtt2ac,o/~t,~i~ .-I TJWC[~. p. 305.
Springer-Verlag. New York. SCHEVINC;. L. E.. PA~!LEY, J. E.. TSAI. T. H.. SCHEVINC;. L. A. (1983). Chronobiology of cell proliferation implications for cancer chemotherapy. In Bio/o~~icul Rh~fhttn putholo+jc..
uttd
:2/c,dic~itw
Ccll&w.
:\letuholic,.
Ph~~.sio-
u?lcl Phurtnuwlqpic :lsj~,c~t.c (A. Reinberg and M. H. Smolenshy Eds.). pp. 79-l 30. SpringerVerlag. New York. SIDDlh. Z. H., BOXALL. F. E.. AUD HARRAP. K. R. ( 1987a). Haematological toxicity of carboplatin in rats. Urir. .I. C’UWCY 55. 375-379. SIDDIK. Z. H.. NEWEL-I. D. R., BOXALL. F. E., HARRAP. K. R. (1987b). The comparative pharmacokinetics of carboplatin and cisplatin in mice and rats. Biochcw Phurtwcol.
36.
1925-
1937.
VAN ECHO. D.A.. EGORI. M. J.. WHITACRE, M. Y.. OLMAN. E. A.. AND AISNER. J. (1984). Phase I clinical and pharmacologic trial ofcarboplatin daily for 5 days. C’unc~c~r Twut Rep. 68, 1 103- I 104,