[CANCER RESEARCH 43, 1426-1431, March 1983]0008-5472/83/0043-0000$02.00

Intraperitoneal c/s-Diamminedichloroplatinum with Systemic ThiosulfateProtection1

Stephen B. Howell,2 Craig E. Pfeifle, Wally E. Wung, and Richard A. Olshen

Departments of Medicine [S. B. H., C. E. P., W. E. W.] and Mathematics [R. A. O.], the General Clinical Research Center, and the Cancer Center, University ofCalifornia, San Diego, La Jol/a, California 92093

ABSTRACT

The toxicity and pharmacokinetics of c/s-diamminedichloro-platinum (cisplatin) (90 mg/sq m) administered as a single 4-hr peritoneal dialysis, with or without concurrent i.v. infusion ofsodium thiosulfate, 0.43 or 2.13 g/sq m/hr for 12 hr, werestudied on 20 courses of treatment. When given without thiosulfate, the toxicity of cisplatin was systemic rather than local,and the peritoneal cavity/plasma ratio of the area under thecurve was 12. Addition of i.v. thiosulfate significantly reducedthe nephrotoxicity. The concentration of cisplatin in the peritoneal cavity was sufficiently greater than that in the plasma toprevent thiosulfate, which equilibrated into the cavity, frominterfering with the antitumor activity of cisplatin in the peritoneum. This study demonstrates a pharmacokinetic advantageof i.p. chemotherapy with cisplatin.

INTRODUCTION

It has been amply demonstrated both in experimental murinesystems and in humans that some tumors which are sensitiveto a drug in vitro do not respond when treated in vivo (32, 33).Inadequacy of drug delivery is probably the most importantreason for the failure of chemotherapy in these tumors (20).Intracavitary chemotherapy has the potential of markedly increasing total drug delivery for at least 2 major groups ofhuman tumors, ovarian carcinoma and mesothelioma, and possibly for others. The usefulness of intracavitary chemotherapyis already well established for the treatment of meningealleukemia (1, 16), and the intravesicular route is used routinelyin the management of early-stage bladder carcinoma (22, 31).

When drugs are administered by peritoneal dialysis, the ratioof the total drug exposure (area under the concentration xtime curve, AUC3) for the peritoneal cavity to that for the

systemic circulation is determined by the relative clearances ofthe drug from these 2 compartments (9). Pharmacokineticmodeling predicts that, when the drug is slowly removed fromthe peritoneal cavity and rapidly eliminated from the rest of thebody, or when the drug is extensively metabolized duringpassage from the cavity to plasma, the concentration in thecavity can be maintained at a level several orders of magnitudehigher than in plasma (9). This prediction has been confirmedin initial studies of 5-fluorouracil (27-29), methotrexate (13,

17, 18), and doxorubicin (23, 24). Under conditions where the

' Supported by Grant CA 23100 from the National Cancer Institute, by GrantRR-00827 from the NIH/Division of Research Resources. This work was conducted in part by the Clayton Foundation for Research—California Division.

2 Clayton Foundation investigator. To whom requests for reprints should beaddressed, at the Department of Medicine T-012, University of California, SanDiego, La Jolla, Calif. 92093.

3 The abbreviations used are: AUC, area under the concentration x timecurve; cisplatin. c/s-diamminedichloroplatinum; DDTC, diethyldithiocarbamate.

Received July 28. 1982; accepted December 2, 1982.

cavity/plasma concentration ratio is sufficiently high, the i.v.infusion of a competitive neutralizing agent may potentiallyfurther improve the therapeutic index by blocking the toxicityof the drug reaching the plasma. The success of this approachdepends both on the rate at which the neutralizing agentequilibrates into the peritoneal cavity and on how the degree ofneutralization changes as a function of the concentration of thechemotherapeutic agent. If the slope of this interaction is steepenough, then, when just enough neutralizing agent is infusedi.v. to block systemic toxicity, even if the agent equilibrates intothe cavity it will not be able to interfere with the cytotoxicactivity of the higher concentration of chemotherapeutic agentat the surface of the tumor. We have successfully used thisapproach to increase the duration of peritoneal exposure tovery high concentrations of methotrexate to as long as 5 days(13).

Cisplatin is one of the most effective drugs currently availablefor the treatment of ovarian carcinoma (26, 30, 34, 35). Whenit is administered i.v., its dose is limited to 100 to 120 mg/sqm by nephrotoxicity. We have previously demonstrated thatconcurrent administration of sodium thiosulfate can protectagainst nephrotoxicity in mice (15). In this clinical trial, we havestudied the pharmacokinetics and toxicity of cisplatin administered i.p. and have examined the effect of concurrent i.v.infusion of sodium thiosulfate on these parameters. We havealso investigated the relationship between drug exposure andthe survival of clonogenic human ovarian carcinoma cells invitro over a range of total drug exposure encompassing thatwhich can be achieved with i.p. therapy.

PATIENTS AND METHODS

The subjects were patients with malignant ascites and a positivecytology or patients whose tumor was entirely confined to the peritonealcavity. Seven patients received a total of 20 courses of ¡.p.cisplatin.There were 2 males and 5 females with a median age of 51 (range, 31to 60). Four patients had ovarian carcinoma and one each had mesothelioma, melanoma, and malignant carcinoid. Entrance criteria included a serum creatinine of less than 1.5 mg/dl, a blood urea nitrogenof less than 40 mg/dl, a leukocyte count of greater than 3000/cu mm,a platelet count of greater than 100,000/cu mm, a life expectancy ofgreater than 1 month, and recovery from all toxicities due to priortreatment.

Patients were hydrated with 0.9% NaCI solution i.v. for 12 hr priorto treatment. The abdomen was drained completely through either apreviously placed Tenchkoff catheter or a percutaneously insertedperitoneal dialysis catheter. The requisite dose of cisplatin (BristolLaboratories, Syracuse, N. Y.) was diluted to isotonicity with sterilewater, suspended in 2 liters of warm 0.9% NaCI solution, and instilledinto the peritoneal cavity by gravity flow over 10 min. After a single 4-

hr dwell, the cavity was again drained as completely as possible.Concurrent with the instillation of cisplatin, a bolus of 12.5 g mannitolwas given i.v. followed by 30 g of mannitol in 1 liter 0.9% NaCI solutioninfused over 6 hr. If urine flow was less than 100 ml/hr prior to

1426 CANCER RESEARCH VOL. 43

Research. on January 19, 2021. © 1983 American Association for Cancercancerres.aacrjournals.org Downloaded from

Cisplatin i.p.

treatment, 20 mg furosemide was also administered i.v. concurrentwith cisplatin administration. When sodium thiosulfate(Eli Lilly and Co..Indianapolis, Ind.) was administered, a loading dose of either 0.8 or4.0 g/sq m was given at the start of the cisplatin instillation, and thiswas followed by either 0.43 or 2.13 g/sq m/hr by continuous infusionfor 12 hr. The entire thiosulfate dose was mixed in 2 liters of H2O andadministered at 167 ml/hr. Hydration i.v. at a minimum of 167 ml/hrwas also continued for 12 hr after the start of cisplatin exposure.Premedication with phenothiazine was used for control of vomiting.Retreatment was every 3 weeks. The cisplatin dose was 90 mg/sq mon all courses.

On the first course of treatment, patients had 1.8 to 2.1 mCi 99mTc

sulfur colloid instilled into the peritoneal cavity along with the cisplatin,and the abdomen was scanned to assess adequacy of distribution. Theprotocol called for pretreatment and follow-up audiograms as well as

weekly complete blood counts, serum creatinine, blood urea nitrogen,and liver function tests. Peritoneal fluid cell counts, cytology, andcultures were obtained at each treatment.

Species of non-protein-bound cisplatin still capable of reacting withnucleophilic sites were measured in plasma by high-pressure liquidchromatography (2). Blood was drawn into ice-cold heparinized tubes,the plasma was separated at 4°,and then the protein was removed by

centrifugation through CF 25 A filter cones (Amicon Corp., Lexington,Mass.). Reactive forms of cisplatin were immediately derivatized byaddition of excess DDTC (2) and promptly subjected to Chromatographie separation and quantitation. Thiosulfate was measured by theiodine titration method (10).

Samples of peritoneal dialysate were obtained every 60 min followingcisplatin instillation, and venous blood samples were obtained at aminimum of every 30 min until cisplatin was no longer detectable. Insome patients, samples were obtained at more frequent intervals.Pharmacokinetic parameters for the peritoneal cavity were estimatedby computer-assisted fitting of the data to a one-compartment model.

It was assumed that drug exposure for the peritoneal cavity ceasedfollowing removal of the cisplatin at 4 hr. The area under the plasmacisplatin elimination curve was calculated by numerical integration fromtime zero to infinity.

WI-L2 is a B-lymphoblastoid cell line (21). WI-L2 cells were main

tained in log phase growth by serial passage in vitro in Roswell ParkMemorial Institute media containing 10% fetal calf serum (Flow Laboratories, Inglewood, Calif.). The effects of cisplatin and thiosulfate weremeasured as changes in growth rate over a 3-day period of time. Thesensitivity of ovarian carcinoma colony-forming cells to cisplatin was

measured in vitro using a drug exposure of 1 hr followed by culture insoft agar according to the technique of Hamburger et al. (11 ). Cells forthese assays were obtained from patients failing primary chemotherapyand from the T17 and T428 human ovarian xenografts maintained atthe University of California, San Diego, Nude Mouse Research Center.

RESULTS

Dose Rate Considerations. Chart 1 shows the ability ofthiosulfate to protect a human lymphoblast cell line (WI-L2)

against the antiproliferative effect of cisplatin. The concentrations of cisplatin were chosen to encompass the range of peakplasma cisplatin levels reported following i.v. injection of amaximum tolerated dose (100 mg/sq m) (12, 25). Molar thio-sulfate/cisplatin ratios of 280 to 950 were required to restorethe growth rate of these representative nonmalignant cells to50% of control, indicating that thiosulfate is not a very potentantagonist of cisplatin. However, it is apparent from the data inChart 1 that, whereas a plasma thiosulfate concentration of500 /¿g/mlwould be sufficient to protect against the cytotox-

icity of 1 (UMcisplatin, this concentration of thiosulfate wouldbe unable to protect against the toxicity of a 10-fold higherconcentration of cisplatin. Thiosulfate is excreted primarily by

the kidney, and its clearance closely approximates that ofcreatinine (3, 10). Based on calculations using Equation A,which relates the plasma concentration at steady state (Css) tothe infusion rate and clearance, it was estimated that continuous sodium thiosulfate infusions of 0.43 and 2.13 g/sq m/hrwould yield steady-state plasma concentrations of approxi

mately 94 and 471 ftg/ml, respectively, in patients with normalrenal function, and these dose rates were selected for clinicaltrial.

Infusion rate

Clearance(A)

Effect of Thiosulfate on Cisplatin Toxicity. Chart 2 showsthe effect of thiosulfate on the nephrotoxicity of cisplatin asmeasured by the maximum percentage of change in serumcreatinine during the 3-week period following treatment on 19

évaluablecourses. When cisplatin was administered by the i.p.route without thiosulfate, the mean increase in serum creatininewas 42 ±41 % (S.D.). The addition of an i.v. thiosulfate infusionof 0.43 g/sq m/hr produced a peak creatinine change of -11

±31% (p < 0.05) and, when a thiosulfate dose rate of 2.13g/sq m/hr was used, the maximum change was 11 ±19% ( p< 0.1). Thus, the lower thiosulfate dose rate significantly reduced the nephrotoxicity of cisplatin. Although the effect of thehigher dose rate of thiosulfate was not statistically significant

70

60

50

40

30

20

l 10 100 1000

SODIUM THIOSULFATE CONCENTRATION, (jg/ml

Chart 1. Dose-response curves for the antagonism by sodium thiosulfate ofcisplatin-induced decrease in the growth rate of WI-L2 cells in vitro. Cisplatinconcentrations: 1 /IM (•);4 JIM(O); and 10 JIM(x). Antagonism was quantitatedby growth rate. Point, mean of triplicate cultures.

bJe>Z

UJ<zI-^o

—Ht-<ZUJUJ(EOUirUJ

S0-=)i

S5X

?Z100806040200-20-40•,1

••".j

T ,

O 0.43 2.13SODIUM THIOSULFATE, g/mz/hr

Chart 2. Nephrotoxicity of cisplatin, 90 mg/sq m i.p., as a function of sodiumthiosulfate dose. Nephrotoxicity was quantitated as thémaximum percentage ofchange in sérumcreatinine during the 3 weeks following treatment. Bars, geometric mean.

MARCH 1983 1427

Research. on January 19, 2021. © 1983 American Association for Cancercancerres.aacrjournals.org Downloaded from

S. e. Howe// ef al.

by f test, using 2.13 g/sq m/hr of thiosulfate it has beenpossible to escalate the cisplatin dose to 270 mg/sq m withoutany significant nephrotoxicity (14) confirming that this doserate of thiosulfate is in fact protective.

Table 1 presents data on the hematological toxicity of i.v.cisplatin alone and in combination with i.v. thiosulfate on 18évaluablecourses. Cisplatin at a dose of 90 mg/sq m i.p. byitself caused a decrease in the leukocyte count which averagedonly 2% but produced an average 31 % decrease in the plateletcount (p < 0.05, t test for paired observations). The additionof thiosulfate had no effect on the changes in the leukocytecount, but at a dose rate of 0.43 g/sq m/hr it reduced, and ata dose rate of 2.13 g/sq m/hr it completely eliminated, thereduction in platelet count (p > 0.05, f test for paired observations).

Instillation of cisplatin i.p. did not produce any local toxicitywithin the peritoneal cavity. No patient in this study developedclinical symptoms or signs of peritonitis. Pretreatment peritoneal fluid cell counts were often abnormal, but cell counts werenot increased when measured at 3 weeks after treatment, andthey showed no tendency to increase with serial courses oftherapy. Since initiation of the i.p. cisplatin treatment programat this institution, we have had the opportunity to directlyexamine the serosal surfaces in 5 patients either at laparoscopyor autopsy within 4 weeks of the last course of treatment, andin no case were there any changes in the serosal surfaces thatcould be attributed to treatment. In patients with documentedi.p. adhesions, none of whom were included in this study, wehave observed mild-to-moderate abdominal pain during filling

of the abdomen, and this has been attributed to mesenterictraction. The amount of nausea and vomiting produced by i.p.instillation of cisplatin was less intense and of shorter durationthan that produced by equivalent i.v. doses; however, nauseaand vomiting were still significant and universal.

Effect of Thiosulfate on Cisplatin Pharmacokinetics. Non-protein-bound cisplatin was measured with an assay that detects only those forms of cisplatin in the plasma that are stillcapable of reacting with nucleophilic sites on other molecules,in this case DDTC (2). Data were sufficient for pharmacokineticanalysis on 18 of the 20 courses. Chart 3 shows the DDTC-reactive cisplatin kinetics in a representative patient with ovarian carcinoma receiving 90 mg/sq m. There was a nearlyexponential decrease in the peritoneal cisplatin concentrationwith time. The pharmacokinetic data is presented in Table 2.The i.p. instillation of 90 mg/sq m of cisplatin produced anAUC for the peritoneal cavity of 97.1 fig-hr/ml and a plasmaAUC of 8.1 jug-hr/ml when measured using the DDTC assay.Thus, the AUC for the peritoneal cavity was 12-fold greater

than that for the plasma following i.p. instillation of cisplatin.The data in Table 2 also demonstrate the impact of the

concomitant i.v. infusion of thiosulfate at 0.43 and 2.13 g/sqm/hr on the cisplatin kinetics. The lower dose rate producedno reductions in any of the pharmacokinetic parameters foreither the peritoneal cavity or plasma. The higher thiosulfatedose rate produced a 23 to 32% decrease in the mean valuesfor the peak peritoneal cisplatin concentration, (p < 0.05) andthe peritoneal (p < 0.05) and plasma (p < 0.20) AUC, butchanges in peritoneal AUC were matched by proportionalchanges in plasma AUC so that the mean AUC ratio did notvary with thiosulfate dose rate. Thiosulfate did not change theperitoneal clearance of the cisplatin, reflecting the fact thatneither dose rate of thiosulfate affected the half-life of 0.85 ±0.26 hr for DDTC-reactive cisplatin in the peritoneal cavity.

Pharmacokinetics of Sodium Thiosulfate. Serial plasmaand peritoneal measurements of thiosulfate concentration wereavailable from 13 courses of treatment in 8 patients receivingthiosulfate at a dose rate of 2.13 g/sq m/hr. The insensitivityof the thiosulfate assay precluded accurate measurements inpatients receiving 0.43 g/sq m/hr. Chart 4 shows that the i.v.loading dose of 4 g/sq m followed by a constant infusion of2.13 g/sq m/hr produced a steady-state plasma thiosulfate

concentration which averaged 410 jug/ml, reflecting a totalbody clearance of 85 ml/min/sq m. Even at the time of thefirst measurement immediately after instillation of the cisplatin,measurable concentrations of thiosulfate were already presentin the peritoneal cavity, and within 2 hrthe peritoneal thiosulfate

lOO

2345

TIME, hours

Chart 3. DDTC-reactive cisplatin concentrations in the peritoneal cavity (O)and plasma (•)in a representative patient, cisplatin, 90 mg/sq m was instilled in2 liters of 0.9% NaCI solution into the peritoneal and allowed to remain in theperitoneal cavity for 4 hr.

Table 1Influence of i.v. administered sodium thiosulfate on the hematological toxicity of cisplatin instilled into the peritoneal cavity

No. ofévaluablecourses66

6Thiosul

fate dose(g/sq m/

hr)00.43

2.33Pretreatment

mean67.5

(3.4-1 8.9)d6.7(3.6-11.5)5.2 (3.0-7.4)WBC/cu

mm"Nadir

mean"6.4(2.9-11.5)

5.2(2.3-16.4)5.0 (3.0-9.9)Platelet

count/cummaMean0

percentage of

decrease-2.0±63.4e

-18.2 ±31.34.8 ±53.8Pretreatment

mean"407(249-1160)

393 (287-702)317(243-547)Nadir

mean"275(210-585)

324 (228-580)327 (260-547)Mean0

percentage

ofchange-30.5±17.1

- 14.3 ±28.23.8 ±13.0

b Geometric mean.0 Arithmetic mean.

Numbers in parentheses, range.8 Mean ±S.D.

1428 CANCER RESEARCH VOL. 43

Research. on January 19, 2021. © 1983 American Association for Cancercancerres.aacrjournals.org Downloaded from

Cisplatin i.p.

Table 2

Influence of i.v. administered sodium thiosu/fate on the pharmacokinetics of cisplatin instilled into the peritoneal cavity

No. ofévaluablecourses7

65Thiosul-

fate dose(g/sq m/

hr)0

0.432.13Mean

peak cisplatin concentration(/ig/ml)Peritoneal

cavity46.0±2.1 8b

54.0 ±12.834.6 ±21.3Plasma2.4

±1.42.7 ±1.72.4 ±2.2Mean

AUCPeritoneal

cavity97.1

±64.9106.6 ±66.0

74.7 ±60.8Plasma8.1

±5.18.3 ±4.65.5 ±4.8Mean

AUC ratio"

(fig.hr/ml)15.2

±11.013.4 ± 5.014.7 ± 6.5Mean

peritonealclearance (liters/

sqm/hr)0.77

±0.270.81 ±0.360.83 ±0.28

Arithmetic mean of the ratios calculated individually for each course.' Mean ±S.D.

concentration had reached steady state at a value that was81% of the plasma level. Thus, thiosulfate introduced into thesystemic circulation entered the peritoneal cavity rapidly, andsignificant differences in the availability of thiosulfate for theneutralization of cisplatin in the 2 compartments existed onlyduring the first hr after the start of i.v. thiosulfate infusion.

In Vitro Cisplatin Dose-Response Curve. The pharmacoki-

netic results presented above indicate that, by instilling cisplatin via the i.p. route, the peritoneal cavity received a total drugexposure that was approximately 12-fold greater than the

systemic circulation. In order to determine the significance ofthis greater AUC for the killing of tumor cells, primary tumorsfrom 3 patients with ovarian carcinoma and samples of 2 humanovarian carcinoma xenografts growing in nude mice weretested in vitro for their sensitivity to cisplatin using the soft agarcolony assay of Hamburger et al. (11). All the tumors testedhad been exposed to prior therapy with cisplatin. Chart 5indicates that the logarithm of the surviving fraction is linearlyrelated to the logarithm of the AUC (r = 0.93; p = 0.005). AnAUC of 8.1 fig-hr/ml for DDTC-reactive cisplatin, which is

close to the maximum tolerated plasma exposure in the absence of thiosulfate, yielded a surviving fraction of 0.27. A 12-

fold greater AUC, such as was achieved in the peritoneal cavity,produced a surviving fraction of 0.17. Thus, it is apparent that,because of the logarithmic nature of the dose-response curve,

even the large increase in peritoneal AUC that was achievablewith i.p. instillation of cisplatin would not necessarily be expected to increase tumor cell kill by more than approximately0.1 log unit over that attainable with AUCs associated withconventional routes of cisplatin infusion.

Responses. Table 3 presents information on the 6 patientswho were évaluable for response, including 4 with ovariancarcinoma, one with mesothelioma, and 1 with malignant car-

cinoid. All patients had had extensive prior therapy, and theovarian carcinoma patients had failed combination chemotherapy programs. All évaluablepatients received treatment withcisplatin alone on some courses and cisplatin with thiosulfateon others. In some patients, the dose of cisplatin was escalatedto tolerance after treatment at the 90-mg/sq m level (14). The

patient with carcinoid responded with a >50% decrease in thevolume of his ascites and normalization of a markedly elevated5HIAA lasting 5 months. The patient with mesothelioma had adecrease in his ascites to the point where it was not clinicallydetectable, and disappearance of an abdominal wall mass thatlasted until he was lost to follow-up after 4 months. One of the

patients with ovarian carcinoma who had residual micronodulardisease and a positive peritoneal fluid cytology after debulkingsurgery converted her cytology to negative and has remainedfree of disease for 12+ months while receiving continuedtherapy.

TIME, hours

Chart 4. Geometric mean sodium thiosulfate concentrations in the plasma(O) and the peritoneum (•)produced by administration of 4.0 g/sq m i.v. bolusfollowed by 2.13 g/sq m/hr by constant i.v. infusion. Bars, S.D.

I 0

O I

O.OIO.I I IO IOO IOOO

AUC, pg-hr/ml

Chart 5. Mean surviving fraction of human ovarian carcinoma colony-formingcells as a function of the area under the concentration x time curve. Cells wereexposed to various concentrations of cisplatin for 1 hr. Bars, S.D. Cells wereobtained from tumors previously treated with cisplatin.

DISCUSSION

When administered by the i.v. route, cisplatin is rapidlydelivered by capillary flow to the bone marrow and the kidneysbut is much less effectively delivered to free-floating tumorcells in the peritoneal cavity or poorly vascularized small se-rosal nodules. The results of this study suggest that, for thesecomponents of tumors confined to the peritoneal cavity, the i.p.route of cisplatin administration appears to be much moreadvantageous than the i.v. route, even when used withoutthiosulfate. Instillation of cisplatin via the i.p. route resulted ina 12-fold greater exposure for the peritoneal cavity than forthe plasma. In addition, at the relatively large 90-mg/sq m dose

of cisplatin, sufficient drug leaked into the plasma so that eventhose components of the tumor supplied only by capillary flowreceived a cytotoxic exposure to cisplatin. Thus, by utilizingthe i.p. route, one can markedly increase the total drug expo-

MARCH 1983 1429

Research. on January 19, 2021. © 1983 American Association for Cancercancerres.aacrjournals.org Downloaded from

S. B. Howell et al.

Table 3Characteristics of patients évaluablefor response and nature of response to treatment with i.p. cisplatin and i. v. thiosulfate

PatientH.

M.F.

A.W.

D.L.

M.C.

H.B.

M.DiagnosisMalignant

carcinoidMesotheliomaOvarian

carcinomaOvarian

carcinomaOvarian

carcinomaOvarian

carcinomaPrior

chemotherapyp-ChlorophenylalanineMethotrexateMelphalan,

doxorubicin,hexamethyl-melamine,5-fluorouracil,methotrex-ate.

vinblastine.progesteronecisplatin.doxorubicin,cyclophospha-mide,

hexamethylmelamine,metho-trexate,5-fluorouracilcisplatin,

doxorubicin,cyclophospha-mide,hexamethylmelamine,metho-trexate,

5-fluorouracilcisplatin,doxorubicin,cyclophospha-midePresentationMassive

ascites, macronodulari.p.disease,5H1AA530.Massive

ascites, 2-cm abdominalwallmassAscites,

cytologypositiveMinimal

residual disease afterthird-looklaparotomyMinimal

residual diseasefollowingsecond-looksurgery, cytology pos

itiveMassiveascites, macronodular dis

ease, cytology positiveResponse>50%

decrease in ascites, normalization of 5H1 AA for 5mos.Complete

resolution of ascites and disappearance of abdominal wallmasslasting

4 mos. until lost tofollow-upCompleteresolution of asciteslasting1

1 mos., persistently positive cytologyPersistent

minimal residual diseaseatfourth-looklaparotomy after3coursesConverted

cytology to negative, no evidence of disease after localhepaticirradiation

for 12+mos.Decreasedascites, no response of pal

pable tumor masses

sure for free-floating tumor cells and small nodules for whichthere is no limitation of drug diffusion, while at the same timedelivering significant amounts of drug by capillary flow. However, it is apparent that by itself the use of the i.p. route wouldnot permit much escalation in the total amount of cisplatinactually administered to the patient since renal damage wasstill the major toxicity.

The addition of i.v. sodium thiosulfate to the program of i.p.cisplatin administration had the effect of reducing nephrotox-

icity, and at the higher dose rate it also reduced the degree ofthrombocytopenia. In a separate study, we have shown thatthe degree of protection against nephrotoxicity was quite dramatic, and that even cisplatin doses of 270 mg/sq m did notproduce nephrotoxicity when thiosulfate was administered concurrently (6, 14). The mechanism of this protective effectremains uncertain. Cisplatin reacts readily with nucleophilicsites on a variety of sulfur-containing molecules (4, 36), and inall probability thiosulfate can react covalently with cisplatin toform a complex that has no toxicity for either normal or malignant tissues (15). This inactive cisplatin/thiosulfate complex isnot measured by the assay used for cisplatin in this study.However, despite the presence of thiosulfate in the plasma,DDTC-reactive cisplatin was paradoxically readily detected inamounts equivalent to those present in the absence of thiosulfate infusion. Thus, the protection against nephrotoxicity wasnot due to a major decrease in the delivery of active drug tothe kidneys. Thiosulfate is not a very potent neutralizing agentfor cisplatin, as evidenced by the fact that the thiosulfateconcentration must be 280 to 950 times higher than that of thecisplatin in order to reverse toxicity to a lymphoblastoid cellline, and the fact that the rate constant for the reaction is low.4

One likely explanation for the ability of thiosulfate to protectwithout decreasing the plasma ADC is that thiosulfate may beconcentrated extensively in the kidneys (3, 10), possibly producing sufficiently high local concentrations of thiosulfate toneutralize cisplatin reaching the renal tubules.

The average clearance of DDTC-reactive cisplatin from the

peritoneal cavity of 0.77 liters/sq m/hr was somewhat greaterthan that predicted on the basis of molecular weight alone(approximately 0.2 liters/sq m/hr) (8). This was probably duein part to the fact that cisplatin can be cleared not only bydiffusion into capillaries and lymphatics but also by reactionwith nucleophilic sites on proteins and other compounds pres-

4 W. E. Wung and S. B. Howell, unpublished observations.

ent in peritoneal fluid to form inactive complexes (5, 12,19). Itis noteworthy that the peritoneal clearance of cisplatin did notvary with serial courses of treatment, which is in agreementwith the observation that it produced no evidence of chemicalperitonitis, and no changes in the serosal membranes wereobserved in 5 patients by direct visualization.

The steady-state concentration of thiosulfate in the plasma

(410 jug/ml) was very close to that predicted on the basis ofEquation A. Thiosulfate distributes in the total extracellular fluid(10), and a significant differential in peritoneal and plasmathiosulfate concentration was observed only for the first hr ofinfusion. However, the concentration differential for cisplatinbetween the 2 compartments was well above the 10-fold difference that, based on in vitro data, should be sufficient toprevent thiosulfate from interfering with the antitumor activityof cisplatin in the peritoneal cavity. Several other compoundsin addition to sodium thiosulfate can reduce the systemictoxicity of cisplatin in animals (4, 34), and thiosulfate may notbe the ideal neutralizing agent. However, the facts that it isreadily available for use in humans, that it produces littletoxicity, and that there has been a long period of experiencewith the clinical use of the agent for the treatment of cyanidepoisoning (7, 8) recommends it for further study.

The adequacy of drug distribution in the abdomen, the penetration of the drug into the tumor from its free surface, andthe intrinsic sensitivity of the tumor are all important considerations that bear on the potential clinical utility of i.p. cisplatinwith or without thiosulfate. Little information is available onwhat fraction of the tumor surface can be exposed to drugwhen the abdomen is distended with 2 liters of fluid or on theability of cisplatin to penetrate the tumor by diffusion. One cananticipate that, in patients with ovarian carcinoma, mesotheli-oma, or other tumors, loculation of fluid may be a frequentproblem, and that because of its reactivity the diffusion coefficient of cisplatin in tissue may be quite low. These difficultiescan be circumvented in part by using a large enough doseadministered into the peritoneal cavity that the amount of drugthat leaks into the plasma is equivalent to a maximum tolerateddose of drug given i.v.

The problem of intrinsic drug resistance on the part of thetumor may not be as easily surmounted. Our data indicate thattumors obtained from patients previously treated with cisplatinare highly resistant when tested in vitro. The dose-responsecurve for the killing of colony-forming cells in these tumors wasvery shallow, so that for each 1-log increase in AUC there was

1430 CANCER RESEARCH VOL. 43

Research. on January 19, 2021. © 1983 American Association for Cancercancerres.aacrjournals.org Downloaded from

Cisplatin i.p.

only a 0.2-log additional tumor cell kill. Untreated tumors canbe expected to have steeper dose-response curves (11) than

were observed for previously treated tumors in this study, andthis suggests that the increased AUC achievable with the i.p.route should be of much greater value to the patient with apreviously untreated tumor than for patients with prior cisplatinexposure. Nevertheless, at present there is little information onhow many logs of in vitro clonogenic cell kill are required percourse for a treatment program to be effective in vivo, and wehave observed excellent responses in several patients evenwhen macronodular disease was present (14). Thus, we believethat at this point in its development intracavitary chemotherapymerits further investigation for both early and untreated, andlate and previously treated disease.

ACKNOWLEDGMENTS

The authors would like to thank Bristol Laboratories and Eli Lilly and Co. forproviding the drugs used in this study. In addition, they would like to thank KathyPrice for preparation of the manuscript.

REFERENCES

1. Band, P. R., Holland, J. F., and Bernard, J. Treatment of central nervoussystem leukemia with intrathecal cytosine arabinoside. Cancer (Phila.), 32:744-748, 1973.

2. Bannister, S. J., Sternson, L. A., and Repta, A. J. Urine analysis of platinumspecies derived from c/s-dichlorodiammineplatinum (II) by high performanceliquid chromatography following derivatization with sodium diethyldithiocar-bamate. J. Chromatogr., ) 73: 333-342, 1979.

3. Biegen, E., Orning, K., and Aas, K. The renal clearance of thiosulfate in man.Scand. J. Clin. Lab. Invest., 1: 102-108, 1949.

4. Borch, R. F., Katz, J. C., Lieder, P. H.. and Pleasants, M. E. Effect ofdiethyldithiocarbamate rescue on tumor response to cisplatinum in a ratmodel. Proc. Nati. Acad. Sei. U. S. A. 77: 5441-5444, 1980.

5. Burchenal, J. H., Kalaher, K., Dew, K., Lokys, L., and Gale, G. Studies ofcross-resistance, synergistic combinations and blocking of activity of platinum dérivâtes.Biochimie (Paris), 60. 961-965. 1978.

6. Campbell, T. N., Howell, S. B., Pfeifle, C. E., Wung, W. E., and Bookstein,J. Pharmacokinetics of infra-arterial cisplatin in man. Proc. Am. Soc. Clin.Oncol., 1: 27, 1982.

7. Chen, K. K., and Rose, C. L. Nurite and thiosulfate therapy in cyanidepoisoning. J. Am. Med. Assoc., 149: 113-119, 1952.

8. Chen, K. K., and Rose, C. L. Treatment of acute cyanide poisoning. J. Am.Med. Assoc., 159: 1154-1155, 1956.

9. Dedrick, R. L., Myers, C. E., Bungay, P. M., and DeVita, V. T. Pharmacoki-netic rationale for peritoneal drug administration in the treatment of ovariancancer. Cancer Treat. Rep., 62: 1-11, 1978.

10. Oilman, A., Phlips, F. S., and Koelle, E. S. The renal clearance of thiosulfatewith observations on its volume distribution. Am. J. Physiol., 146: 348-357,1946.

11. Hamburger, A., Salmon, S. E., Kim, M. B., Trent, J. M., Soehnien, B. J.,Alberts, D. S., and Schmidt, H. J. Direct cloning of human ovarian carcinomacells in agar. Cancer Res., 38: 3438-3444, 1978.

12. Himmelstein, K. M., Patton, T. F., Belt, R. J., Taylor, S., Repta, A. J., andSternson, L. A. Clinical kinetics of intact cisplatin and some related species.Clin. Pharmacol., 29: 658-664, 1981.

13. Howell, S. B., Chu, B. C. F., Wung, W. E., Mehta, B. M., and Mendelsohn,J. Long-duration intracavitary infusion of methotrexate with systemic leuco-

vorin protection in patients with malignant effusions. J. Clin. Invest.. 67:1161-1170, 1981.

14. Howell, S. B., Pfeifle, C. E., Wung, W. E., and Olshen, R. A. Intraperitonealchemotherapy with cisplatin. Proc. Am. Soc. Clin. Oncol., 1: 27, 1982.

15. Howell, S. B., and Taetle, R. The effect of sodium thiosulfate on cis-

dichlorodiammineplatinum(ll) nephrotoxicity and antitumor activity in theL1210 leukemia. Cancer Treat. Rep., 64: 611-616, 1980.

16. Hyman, C. B., Bogle, J. M., Brubaker, C. A. Williams, K., and Hammond, D.Central nervous system involvement by leukemia in children. II. Therapy withintrathecal methotrexate. Blood, 25: 13-22, 1965.

17. Jones, R. B., Collins, J. M., Myers, C. E., Brooks, A. E., Hubbard, S. M.,Balow, J. E., Brennan, M. R., Dedrick, R. L., and DeVita, V. T. High-volumeintraperitoneal chemotherapy with methotrexate in patients with cancer.Cancer Res., 41: 55-59, 1981.

18. Jones, R. B., Myers, C. E., Guarino, A. M., Dedrick, R. L., Hubbard, S. M.,and DeVita, V. T. High volume intraperitoneal chemotherapy ("belly bath")

for ovarian cancer. Cancer Chemother. Pharmacol., 1: 161-166, 1978.19. LeRoy, A. F., Lutz, R. J., Dedrick, R. L., Litters!, C. L., and Guarino, A. M.

Pharmacokinetic study of c/s-dichlorodiammineplatinum(ll) (DDP) in the beagle dog: thermodynamic and kinetic behavior of DDP in a biologic milieu.Cancer Treat. Rep., 63. 59-70, 1979.

20. Levin, V. A., Patlak, C. S., and Landahl, H. D. Heuristic modeling of drugdelivery to malignant brain tumors. J. Pharmacokinet. Biopharm., 8: 257-

296, 1980.21. Levy, J. A., Buell, D. N., Creech, C., Hirshaut, Y., and Silverberg, H. Further

characterization of the WI-L1 and WI-L2 lymphoblastoid lines. J. Nati.Cancer Inst., 46: 647-652, 1971.

22. Nieh, P. F., Daly, J. J., Heaney, J. A., Prout, G. R., Jr. Effect of intravesicalthio-TEPA on normal and tumor urothelium. J. Urol., 119: 59-61, 1978.

23. Ozols, R., Grotzinger, K., Myers, C.. Young, R. C. Ovarian cancer (OC):intraperitoneal (I.P.) therapy with Adriamycin (ADR). Proc. Am. Assoc.Cancer Res., 20: 242, 1979.

24. Ozols, R. F., Young, R. C., Speyer, J. L., Weltz, M., Collins, J. M.. Dedrick,R. L., and Myers, C. E. Intraperitoneal (I.P.) Adriamycin (ADR) in ovariancarcinoma (O.C.). Proc. Am. Assoc. Cancer Res., 21: 425, 1980.

25. Patton, T. F., Repta, A. J., Sternson, L. A., and R. J. Pharmacokinetics ofintact cisplatin in plasma. Infusion versus bolus dosing. Int. J. Pharm.(Amst.), T0:77-85, 1982.

26. Pesando, J. M., Come, S. E., Parker, L. M., Griffiths, C. T., and Canellos,C. P. c/s-Diamminedichloroplatinum (CDDP) therapy of advanced ovariancancer. Proc. Am. Assoc. Cancer Res., 20: 328, 1979.

27. Speyer, J. L., Collins, J. M.. Dedrick, R. L., Brennan, M. F., Buckpitt, A. R.,Londer, H., DeVita, V. T., Jr.. and Myers. C. E. Phase I pharmacologicalstudies of 5-fluorouracil administered intraperitoneally. Cancer Res., 40:567-672, 1980.

28. Speyer, J. L., and Myers, C. E. The use of peritoneal dialysis for delivery ofchemotherapy to intraperitoneal malignancies. Recent Results Cancer Res.,74: 264-269, 1980.

29. Speyer, J. L., Sugarbaker, P. H., Collins, J. M., Dedrick, R. L.. Klecker.R. W., Jr.. and Myers, C. E. Portal levels and hepatic clearance of 5-fluorouracil after intraperitoneal administration in humans. Cancer Res., 41:1916-1922, 1981.

30. Thigpen, T., LaGasse, L., and Bundy, B. Phase II trial of cisplatinum intreatment of advanced ovarian adenocarcinoma. Proc. Am. Assoc. CancerRes., 20:84, 1979.

31. Veenema, R. J., Dean. A. I Ir., Roberts, M., Finergut, B., Showhury. B. K.,and Tarassoly. H. Bladder carcinoma treated by direct instillation of thio-TEPA. J. Urol., 88: 60-63, 1962.

32. Von Hoff, A. A., Casper, J., Bradley, E., Sandback, J., Jones, D., andMakuch, R. Association between human tumor colony-forming assay resultsand response of an individual patient's tumor to chemotherapy. Am. J. Med.,

70: 1027-1032, 1981.33. Von Hoff, D. D., Page, C., Harris, G.. Clark, G., Cowan, J., and Coliman,

C. A., Jr. Prospective clinical trial of a human tumor cloning system. Proc.Am. Assoc. Cancer Res., 22: 154, 1981.

34. Wiltshaw, E., Subramarian, S., Alexopoulos, C., and Barker, G. H. Cancerof the ovary: a summary of experience with c/s-dichlorodiammineplatinum(ll)at the Royal Marsden Hospital. Cancer Treat. Rep., 62: 555-558, 1979.

35. Young, R. C., Von Hoff, D. D., and Gormely, P. cis-Dichlorodiamminepla-tinum(ll) for the treatment of advanced ovarian cancer. Cancer Treat. Rep.,63: 1539-1544, 1979.

36. Yuhas, J. M., Active versus passive absorption kinetics as the basis forselective protection of normal tissues by S-2-(3-aminopropylamino)-ethyl-phosphorothioic acid. Cancer Res., 40: 1519-1524, 1980.

1431

Research. on January 19, 2021. © 1983 American Association for Cancercancerres.aacrjournals.org Downloaded from

1983;43:1426-1431. Cancer Res   Stephen B. Howell, Craig E. Pfeifle, Wally E. Wung, et al.   Thiosulfate Protection

-Diamminedichloroplatinum with SystemiccisIntraperitoneal

  Updated version

  http://cancerres.aacrjournals.org/content/43/3/1426

Access the most recent version of this article at:

   

   

   

  E-mail alerts related to this article or journal.Sign up to receive free email-alerts

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/43/3/1426To request permission to re-use all or part of this article, use this link

Research. on January 19, 2021. © 1983 American Association for Cancercancerres.aacrjournals.org Downloaded from