Intraperitoneal cisplatin and regional hyperthermia for ovarian carcinoma

0360.3016/93 $6.00 + .Xl Copyright Q 1993 Pergamon Press Ltd.

Im J Radmron Oncology Bid Phyv., Vol. 27. pp. 1245-1251 Printed I” the U.S.A. All rights reserved.

0 Oncology Intelligence INTRAPERITONEAL

CISPLATIN AND REGIONAL HYPERTHERMIA FOR OVARIAN CARCINOMA

KENNETHA. LEOPOLD,M.D.,* JAMESR. OLESON,M.D., PH.D.,* DANIEL CLARKE-PEARSON, M.D.,? JOHN SOPER,M.D.,+ ANDREW BERCHUCK,M.D.,+ THADDEUSV. SAMULSKI,PH.D.,* ROD L. PAGE, D.V.M.,* JANYEBLIVIN, R.N., M.S.N.,* JANICEK. TOMBERLIN,M.D.* AND MARK W. DEWHIRST,D.V.M., PH.D.* *Department

of Radiation Oncology and +Division of Gynecologic Oncology, Duke University Medical Center, Durham, NC 277 10; and the ‘North Carolina State School of Veterinary Medicine, Raleigh, NC

Purpose: To review the theoretical basis and results of a Phase I study of concurrent intraperitoneal cisplatin and hyperthermia in the treatment of ovarian carcinoma. Methods and Materials: Previously treated patients with epithelial ovarian carcinoma received intraperitoneal instillation of cisplatin and 60 minutes of regional hyperthermia, with a goal temperature of 41S’C. Cisplatin dose started at 20 mg/m* with escalation to the maximally tolerated dose. Six such cycles given every 3 weeks were planned. Pharmacokinetic studies with and without hyperthermia were performed. Results: Fifteen patients receiving 17 courses of treatment were evaluable. The maximally tolerated dose of cisplatin was between 80 and 120 mg/m*. The dose limiting toxicity was nephrotoxicity in all but one course. The median intraperitoneal temperature was 40.7”C, the majority of treatments in which the goal temperature was not reached had power limited by patient discomfort. No major toxicities attributable to hyperthermia were noted. Pharmacokinetic studies noted no significant differences between treatments with vs. without hyperthermia, with intraperitoneal to plasma area under the curve ratios being 30-35. Ten patients had a decline in their CA-125 count during treatment, although in only two patients did this response persist beyond their course of treatment. Conclusion: Intraperitoneal cisplatin and regional hyperthermia can be performed with reasonable toxicity. The maximally tolerated dose of 80-120 mg/m* in pretreated patients (which is similar to those reported with cisplatin alone) and median intraperitoneal temperatures of 40.7”C, however, are felt to be too low to be efficacious in a significant percentage of women with bulky recurrent disease. Further study using intravenous thiosulfate and controlled analgesia is being performed. Hyperthermia,

Intraperitoneal chemotherapy,

Cisplatin, Ovarian carcinoma.

INTRODUCTION

some drugs may be lower than that of the plasma, allowing cancer in

use of higher drugconcentrations within the peritoneal cavity than can be used in serum. Concentration gradients

women in the United States (23). About 70% of women with ovarian cancer die from their disease ( 18) with failure or progression occurring predominantly within the peritoneal cavity (21). Conventional adjuvant or palliative treatment includes intravenous (IV) chemotherapy with a platinum-based compound. The large majority of patients relapse after this therapy, however, and salvage with other IV chemotherapy is dismal with most patients dying within 1 year (18). Dedrich et al. (2) proposed the use of intraperitoneal (IP) drug delivery for ovarian carcinoma. The rationale of this approach is that the peritoneal clearance rate of

can be calculated and depend on lipid solubility (increasing solubility lowers the IP/plasma concentration gradient) and molecular weight (increasing MW increases the gradient). IP drug delivery has potential applications for therapy of gastrointestinal carcinomas as well as ovarian carcinoma, and can also be extended to use of monoclonal antibodies ( 15). Cisplatin is known to be one of the most active single agents for ovarian carcinoma when administered systemically ( 15). Clinical pharmacokinetic studies ( 1, 13, 20) have shown that IP cisplatin is locally nontoxic and can result in 20-30 times greater area under the concentration

Reprint requests to: Kenneth A. Leopold, M.D., Radiation Oncology, Dartmouth-Hitchcock Medical Center, 2 Maynard Street, Hanover, NH 03756. Acknowledgements-Dr. Leopold is a recipient of the American Cancer Society Clinical Oncology Career Development

Award. This work was supported in part by NIH-NC1 Grant 5POl CA42745-06. Presented at “Radiation and Platinum Agents Combined in Cancer Management,” sponsored by Bristol-Myers Squibb. Accepted for publication 6 July 1993.

Ovarian

carcinoma

is the fourth most common

1245

1246

I. J. Radiation Oncology 0 Biology 0 Physics

X time curve (AUC) within peritoneal fluid than in plasma, compared to IV administration. This result agrees with calculated advantages (2). The concentration of cisplatin has major therapeutic importance according to dose-effect analysis (20). Also, acquired cellular resistance to cisplatin may be partially overcome with sufficiently high drug concentrations ( 13). Many trials of IP cisplatin-based chemotherapy have now been reported (59, 19,2 1). Piver et al. ( 19) recently reported on 3 1 patients with Stage III and IV disease treated with second-line IP cisplatin, cytarabine, and bleomycin. Patients were selected who had failed initial IV chemotherapy (lack of complete response (CR) or recurrence after CR). In these 31 patients there were five surgically confirmed complete responses and three partial responses. Patients with Stage IV disease and patients with > 2 cm residual disease failed to respond. All responders had originally responded to IV cisplatin-based therapy. The median survival from initial therapy was 3 1 months; the median survival from salvage IP therapy was 16 months. Piver et al. argue that equivalent results may be possible with IP cisplatin alone. These results suggest that the IP therapy approach may be particularly useful as an adjuvant or consolidation in initial treatment. The concept of delivery of IP chemotherapy within a heated perfusate to potentiate drug activity within the peritoneal cavity was proposed at least 10 years ago (2426). Theoretically, however, in situ heating of instilled fluid is advantageous to extracorporeal heating of perfusate, since in situ heating should improve temperature homogeneity. The choice of chemotherapeutic agents to be used with hyperthermia would, ideally, include those known to have efficacy in treating ovarian carcinoma at normothermia, and to have demonstrated potentiation of effect at elevated temperatures. The usefulness of cisplatin delivered either systemically or IP has been discussed above. Herman et al. have shown in vitro (7) that cisplatin cytotoxicity is increased by two logs at 42°C with clinically achievable concentrations (27) of cisplatin in tissue (corresponding to about 7 microgram/ml in vitro). Cisplatin has a relatively greater degree of thermal potentiation at 42°C than carboplatin or other platinum complexes (8). The effect of heat may be to increase the cisplatin-DNA reaction rate (8) although mechanisms are incompletely understood. Potentiation of cisplatin cytotoxicity at elevated temperatures is marked only for simultaneous heat and drug (28). This potentiation persists in some cisplatinresistant cell lines (3). Moreover, elevated temperatures reversed acquired cisplatin resistance in a murine tumor line (14). There is thus a strong rationale for selecting cisplatin as the first agent in investigating toxicity and efficacy of an IP drug combined with hyperthermia. Previous work at Duke/North Carolina State University demonstrated increased uptake and retention of cisplatin within the IP cavity when fluid within the cavity was heated to 41.5”C (29). This uptake was greatest at the peritoneal surfaces which may be the areas at greatest risk

Volume 27, Number 5, 1993

for recurrence. At temperatures of 42-43°C for exposure times of about 1 h, direct thermal cytotoxicity also becomes significant (6). Although direct electromagnetic heating may preferentially raise the internal temperatures of gross tumor nodules, such temperatures may be difficult to maintain throughout the entire cavity. Accordingly, the effectiveness of this approach may still be limited to patients with nodules less than 2 cm because the predominant heating mechanism is thermal conduction from fluid into tissue. In summary, addition of hyperthermia to IP cisplatin may markedly potentiate the cytotoxic effect of cisplatin in regions of high risk, and may partially reverse acquired cellular resistance to cisplatin. Direct thermal cytotoxicity may also enhance regional disease control. Given the potential gains cited above with the use of IP cisplatin and concurrent regional HT, we decided to undertake a Phase I study of this approach in women with ovarian cancer refractory to IV cisplatin. The results of this study are presented here. METHODS

AND MATERIALS

Experimental design and methods Eligibility criteria. The study was designed as a Phase I study of increasing doses of IP cisplatin delivered concurrently with regional HT. Patients were eligible if they had histologically confirmed epithelial ovarian cancer resistant to conventional cisplatin based IV therapy. Resistance was defined as disease progression while receiving cisplatin based IV chemotherapy (patients must have received at least three cycles of cisplatin at 2 40 mg/m* per cycle) or persistent macroscopic disease within 1 year of cisplatin based IV therapy (persistence demonstrated by exploratory laparotomy alone was acceptable). Other eligibility criteria included the absence of serious co-morbidity, performance status (Gynecologic Oncology Group (GOG) criteria) of O-2, normal renal function (serum creatinine < 2.0 mg/dl), WBC > 3,000, absolute granulocyte count > 1,500, platelet count > 100,000, and PT/PTT within normal limits. Written informed consent approved by the Institutional Review Board was obtained for all patients. The study was approved by the Federal Food and Drug Administration which supplied an investigational device exemption for the HT device and an IND for cisplatin. Treatment plan. A Port-a-cath or Tenckoff catheter was surgically placed into the peritoneal cavity. After correction of any serum electrolyte abnormalities, patients were hydrated with 2 liters of 5% dextrose in l/2 normal saline solution given IV over the 4 h prior to therapy. Early in the study, loperamide (0.5-2.0 mg), metachlopromide (23 mg/kg), diphenhydramine ( 10 mg), and dexamethasone ( 10 mg) were given IV just prior to treatment and 2 h after cisplatin administration. When ondansetron became available, it was given at 0.15 mg/kg with dexamethasone 10 mg IV. Ondansetron was repeated at the same dose 2 and 4 h following the initial dose. One liter of normal saline preheated to 37°C was instilled IP followed im-

Intraoperative cisplatin and regional hyperthermia 0 K. A. LEOPOLDet al.

mediately by instillation of cisplatin in 1 liter of normal saline given as rapidly as possible. A 4-sensor fluoroptic temperature sensor was subsequently placed within the Tenckoff catheter. Hydration was maintained through the course of treatment and for at least 4 h after treatment with 05-l/2 normal saline solution at 2 150 cc/hr IV with furosemide used as needed to keep urine output 2 100 ml/hr. No attempt was made to remove IP fluid at any time after it was instilled. Regional hyperthermia was delivered using an annular phased array of radiofrequency (RF) waves operating at 82 MHz.’ The power was incrementally increased until the median IP temperature measurement was - 4 1.5”C or to patient tolerance. The phase and amplitude of the array of antennas were adjusted to maximize IP temperatures while minimizing discomfort. Oral temperature was not allowed to exceed 38.5”C. Power application continued for 60 min after an IP temperature first reached 41.5”C or after 30 min of power application if 41.5”C was not reached by that time. Temperatures were monitored IP using fluoroptic temperature probes.* These fiberoptic probes determine temperature via the luminescent decay of light emmitted by a fluorescent material (sensor) attached to the end of optical fibers. Four probes were bundled together. The first three sensors (starting at the probe tip) were spaced three cm apart. The fourth sensor was spaced four cm from the third sensor. Vaginal and rectal temperatures were monitored along the paths of catheters placed within these cavities using high resistance thermistors. These thermistors were translated along catheters within these cavities by means of an automated device. Temperatures in the peritoneum, rectum and vagina were recorded every 5 s. Oral temperatures were monitored approximately every ten minutes using a high resistance thermistor. Sensors were checked prior to each treatment and calibrated as needed. The first three patients’ first treatment consisted of a HT treatment after the IP instillation of two liters of normal saline (without the administration of cisplatin) to ensure the safety and feasibility of this part of the treatment regimen. Cisplatin dosing. The starting dose of cisplatin was 20 mg/m’. The dose was then designed to be increased according to a modified Fibonacci series in which dose was increased by 100% for each succeeding cohort of patients until biological effects were observed. Dose was increased by 50% after that point until the maximal tolerated dose (MTD) was reached. The MTD was defined as the lowest dose at which > 216 patients had any 2 grade 3 toxicity which did not decrease to I grade 1 toxicity by 6 weeks following treatment, or grade 2 renal toxicity or serum creatinine 2 2.0 mg% which did not decrease to I grade 1 or < 2.0 mg%, respectively, by 6 weeks following treatment.

’ Sigma 60, BSD Medical Corp., Salt Lake City, UT. ’ Luxtron Corp., Santa Clara, CA.

1247

Three patients were treated at each dose level until toxicity was observed; at that point, 6 patients were treated per dose level. Patients were scheduled to receive 6 cycles of treatment with treatments given every 3 weeks. Treatment was delayed for serum creatinine r 2.0 mg%, 2 grade 2 renal toxicity, or any 2 grade 3 toxicity until serum creatinine fell to < 2.0 mg% and other toxicities were I grade 1. If there was a toxicity causing a treatment delay of 2 two weeks or if there was a renal toxicity then all further treatments for that patient were given with a cisplatin dose that was reduced one dose level (i.e., to the previous cohorts dose level). Gynecologic Oncology Group toxicity criteria were used. Treatment was discontinued for 2 grade 2 toxicity or serum creatinine r 2.0 mg% which did not decrease to I grade 1 or serum creatinine < 2.0 mg%, respectively, within 6 weeks of the last treatment. Treatment was also discontinued for disease progression or deterioration in the patient’s general status. Pharmacokinetic studies. Pharmacokinetic studies were performed on the first 2 patients treated at cisplatin dose levels of 20, 40, 80, and 120 mg/m*. Hyperthermia was not given during either the first or second cisplatin treatment (determined randomly) in these patients. Samples of blood and peritoneal instillate were obtained prior to cisplatin infusion, at the end of infusion, at 5, 15, 30,45, 60, 120, and 240 min postinfusion, and at 24 h postinfusion. Plasma and instillate were centrifuged at 5000 rpm, 5°C for 30 min through a 30,000 molecular weight separation filter. Unfiltered and filtered samples were kept at -20°C until assayed. Urine volume and urine samples were obtained at 4 and 24 h after cisplatin infusion and stored at -20°C until assayed. Platinum concentration was determined in duplicate for each sample by atomic absorption spectroscopy, using a method previously developed by our group (17). Briefly, samples were analyzed using a model 4000 atomic absorption spectrophotometer with a HGA400 graphite furnace.3 Samples were pipetted onto L’vov platforms and stabilized temperature platform furnace methods were used. The platinum lamp used a wavelength of 265.9 nm with a slit bandwidth of 0.7 nm. Cisplatin statistical moment estimates were calculated as follows on total and filtered plasma and intraperitoneal platinum concentration profiles (4): Area Under the Curve (AUC) was approximated by use of the trapezoidal rule for numerical integration. Area Under the first Moment Curve (AUMC) was estimated using an integrated form of the trapezoidal rule. Half-life (T l/2) was estimated by the formula: (AUMC)(ln 2)/AUC, where In 2 is the natural logarithm of 2. Systemic Clearance (Cls) of free cisplatin was estimated by the formula: Dose/AUC. Apparent Volume of Distribution (Vdss) was calculated by the formula: (Cls) (MRT) where MRT is the mean residence time. Follow-up. Patients were seen in follow-up at least every

3 Perkin-Elmer,

Norwalk,

CT.

1248

I. J. Radiation Oncology 0 Biology 0 Physics

2 months following completion of therapy. Exploratory laparotomy was performed when required for patient management. CA-125 levels were measured at each follow-up visit. Other diagnostic procedures were performed as indicated by the clinical situation. Thermal analysis. Thermometry data acquired during the “steady state” period of each treatment was computed and analyzed for the determination of treatment temperature indices. The “steady state” period (typically 60 min) was defined to be the time interval starting when an IP temperature reached 4 1.5 “C or 30 min after RF power was initiated (whichever occurred first) until the time when power was terminated. The temperature indices calculated included TgO,Tso, and T10 (the temperatures which 90%, 50%, and 1O%,respectively, of the measurements equaled or exceeded) averaged over the “steady state” period (12). The median and average temperature index values were also determined for each patient’s entire course of treatments. The maximum oral temperature and the time interval that the oral temperature exceeded 385°C was noted for each treatment. RESULTS

Patient accrual Eighteen patients were entered onto the protocol. Two of these patients received two courses of treatment. One patient deteriorated prior to therapy and one patient could not tolerate the administration of the required volume of intraperitoneal fluid and therefore received no therapy. One patient who was part of the pharmacokinetic study was withdrawn from the study because of disease progression after HT alone. Seventeen treatment courses were therefore evaluable for toxicity to chemotherapy. Three of these patients were withdrawn from the study because of toxicity after chemotherapy alone, never having received chemotherapy with HT (the specifics of the toxicities are detailed in the “Dose Escalation” section). No patient with extensive ascites was entered and no patient required therapeutic pericentesis while undergoing treatment to allow for the instillation of fluid for treatment. Dose escalation Cisplatin doses of 20 and 40 mg/m’ with HT were tolerated without life-threatening or persistent morbidity. Dose reduction was needed in only one of the 10 patients treated at these levels. The one patient who experienced toxicity developed a small bowel obstruction which led to her demise. Following the successful treatment at 40 mg/m2, the dose was escalated to 80 mg/m2. Four patients began treatment at a cisplatin dose of 80 mg/m2. One patient, treated as part of the pharmacokinetic study, was withdrawn from study after tolerating the first cycle of chemotherapy without significant toxicity. Withdrawal occurred secondary to acute renal failure precipitated by an IV contrast reaction which occurred on the day prior to

Volume 27, Number 5, 1993

the next scheduled treatment. The other three patients required cisplatin dose reductions because of prolonged elevations of serum creatinine. These three were able to complete four cycles of treatment. Further treatment was not delivered in one of three patients because of persistently elevated serum creatinine and in the other two patients because of disease progression. After three patients had tolerated at least three cycles of treatment at 80 mg/m2, the dose was escalated to 120 mg/m2. Three patients began therapy at a cisplatin dose of 120 mg/m2. The first two patients received their first treatment with drug alone (i.e., no HT) as part of the pharmacokinetic study. Both of these patients developed persistently elevated serum creatinine which necessitated withdrawal from the study before concurrent cisplatin and HT could be delivered. One further patient was entered at a cisplatin dose of 120 mg/m2 and received one cycle of drug at this dose along with HT. She experienced a transient increase in serum creatinine (maximum level of 2.3 mg/dl). When the toxicities of the other two patients treated with a dose of cisplatin of 120 mg/m2 became apparent, however, it was decided to reduce the cisplatin dose of the third patient to 40 mg/m2 for future treatments. This dose was subsequently tolerated in this patient without life-threatening or dose-limiting toxicity. Following the toxicities associated with 120 mg/m2 and since it became apparent that there were toxicities associated with 80 mg/m2, it was decided to enter the next cohort of patients at 60 mg/m2. Elevations in serum creatinine were seen in all three patients treated at a cisplatin dose of 60 mg/m2, but they decreased to < 2.0 mg/dl in less than 2 weeks following treatment. These three patients were able to complete their planned courses of treatment. It was clear following this cohort of patients that the MTD was between 80 and 120 mg/m2. At this point, given the limited duration of responses that were seen in the majority of patients, it was felt that a more precise determination of the MTD was not warranted. All but one patient treated at a cisplatin starting dose of 2 40 mg/m2 had at least a transient rise in serum creatinine. GI toxicities, predominantly nausea and vomiting, occurred in all but 1 patient at r 40 mg/m2. This nausea was generally not observed while the patients remained hospitalized and on IV antiemetics, but rather was noted 7-10 days following treatment corresponding to the immediate post hospital discharge period. One patient had profound vomiting leading to rapid deterioration, including seizures, acute renal failure, and death. Four patients developed small bowel obstructions during the course of treatment with one patient requiring surgical management (obstruction was found, intraoperatively, to be secondary to tumor). Most patients experienced hematologic toxicity which included anemia, thrombocytopenia, and leukopenia. Although these toxicities delayed treatment occasionally, they did not require cessation of treatment and were not life-threatening. Only one neuropathy (grade 1) was noted.

1249

Intraoperative cisplatin and regional hyperthermia 0 K. A. LEOPOLDer al. Thermal analysis

Pharmacokinetic

The median and average of all of the patients median IP temperatures was 40.7”C. The median T90 and Tlo were 39.7”C and 41.6”C, respectively. The corresponding median TgO, TsO and T10 vaginal temperatures were 38.8’C, 40.4”C, and 41.O”C, respectively. Median T90, TsOand T10 rectal temperatures were 39.7”C, 40.4”C, and 4 1.1“C, respectively. The median of all of the patients’ oral temperatures was 37.9”C. The single highest IP temperature measured was 44.2”C. Five of the 5 1 treatments achieved the goal IP temperature of median TSo 2 4 1.5“C. This occurred in four different patients. Four of these five treatments were the fourth (three treatments) or fifth (one treatment) in the patients’ course (the other was the first treatment). In seven of the treatments in which the goal temperature was not reached power was limited because of high nontarget temperatures (five oral, two superficial). Four of the treatments limited by high Toral were in the same patient. Power was noted to be limited in 28 treatments due to patient discomfort. This discomfort was generally described as a burning sensation in SC sites generally of the thighs and/or buttocks. Power was occasionally limited because of dyspnea (l), dysethesias (l), claustrophobia (2), or cramps or pressure (3). No specific cause for limiting power was noted in four treatments, although excessively high temperatures were not noted in any of these treatments. Four patients had toxicities noted which were directly attributable to HT. Two patients experienced soreness (1 hip, 1 thigh) persisting after HT treatment, 1 patient had skin erythema, and 1 patient had blisters. In addition to these toxicities, several obese patients were noted to develop SC anterior thigh nodules associated with tenderness consistent with focal fat necrosis. These were presumably secondary to focal high power deposition. These abnormalities resolved spontaneously in all cases within 3 weeks.

Table 1 describes the pharmacokinetic parameters from the treatment groups. Cisplatin pharmacokinetics were not different in patients treated with either IP cisplatin alone or combined with regional hyperthermia. The relative exposure of cisplatin to the peritoneal cavity and serum, AUCip/AUCpl, was 30-35 for both groups.

data

Tumor response

The median pretreatment CA- 125 level in the 14 courses of treatment in which concurrent chemotherapy and HT were given was 84 units/ml (range: 1l- 1238). A response to treatment as indicated by a reduction in CA125 levels was noted in 10 of these 14 treatment courses (median percent reduction: 30.5; range: 6-67%). CA-125 levels stabilized at a low level (around 20 units/ml) in an additional patient. Two patients had responses that were sustained beyond the course of treatment. The remainder of the responses, however, did not persist through the course of treatment. The median duration of time from the initiation of treatment until CA- 125 levels were noted to be rising (in responders) was 4.0 months. Tumor progression was clearly noted in several instances to be related to limited access of the cisplatin to all tumor-bearing regions. In one patient (JW), the exposed peritoneal surfaces, which had diffuse involvement with nodules up to 0.8 cm in diameter pretreatment, were found at posttreatment second-look laparotomy to have had a marked response. A 4-5 cm in diameter pelvic mass also responded well, with only a 2 cm in diameter mass found at reexploration, much of which, histologically, contained desmoplastic tissue. Within a region of small bowel containing dense adhesions, however, extensive tumor involvement was found at reexploration. Two other patients (SH and JW) seemed to respond well except for lesions on the undersurface of the diaphragm. Two further

Table 1. Statistical moments of pharmacokinetic data for ultrafilterable cisplatin. Standard errors of means are given in parentheses and the number (N) of contributing evaluations are given in brackets. Calculation formulae are described under the experimental design and methods section Normothermia

Hyperthermia

Normothermia

Plasma AUC (% dose-min/ml) Half-life (hrs) CLs (ml/min/M’) Vdss (L/M2)

0.39 (.27) 181 6.6 (.4) 171 177 (107) 171 101 (59) 171

Peritoneal Instillate 0.36 (.22) 191 6.8 (.7) [81 176 (88) 181 104 (57) 181

12.7 (8.8) 171 6.4 (1) 171 9.4 (9.5) [7j 4.6 (3.9) 171

Urine (W dose elim @ 4 hr)

8.3 (2.6) 151

Hyperthermia

7.1 (3.6) 171

AUC = Area under the curve; Cls = Systemic clearance; Vdss = Apparent volume of distribution.

12.2 (12.8) 171 6.7 (I) 171 7.9 (3.7) rjl ’ 4.4 (1.9) [71

1250

I. J. Radiation Oncology 0 Biology 0 Physics

patients (CR and MF) had progression of liver metastases as their predominant site of failure. Five patients who received concurrent cisplatin and HT have died from progression of their disease at a median of 9 months (average 9.8 months) after the initiation of treatment. Seven patients were alive with disease at the time of analysis with a median follow-up of 14.0 months. DISCUSSION

The cisplatin MTD of 80- 120 mg/m2 found in this study is similar to tolerable doses of IP cisplatin reported in other studies not using hyperthermia (16, 22). Prior studies using cisplatin either IV or IP have found renal toxicity to be the dose-limiting toxicity. This was also found to be the case in our study. With regional HT, the temperature of the kidneys would be expected to be similar to the systemic temperature since the kidneys receive a high blood flow. Accordingly, renal potentiation of the cisplatin effect would not be expected to be as great as intraperitoneal potentiation. We had hypothesized that the addition of hyperthermia would increase uptake and retention of cisplatin within the peritoneal cavity and thus allow for equivalent or greater doses of IP cisplatin with a concomitant increase in efficacy (29). We were only able, however, to deliver doses similar to what others have reported. Since the peritoneal cavity presumably was at a higher temperature than the kidneys, however, there presumably would be a therapeutic gain from the addition of HT to similar doses of cisplatin alone. Howell et al. have demonstrated the ability to deliver substantially higher doses of IP cisplatin (200 mg/m*) by using concomitant IV thiosulfate (10, 11). They further claim that the thiosulfate selectively protects the kidneys thus preserving the systemic utility of the drug. Their clinical data suggests a therapeutic gain from adding thiosulfate. This should be a safe combination to use with regional hyperthermia and should allow for higher IP doses. One aspect of the rationale for the use of regional hyperthermia and IP cisplatin is the pharmacological advantage identified by our group in a preclinical study using beagles (29). The pharmacokinetic advantage seen in the preliminary study was not observed in the present study. Several differences exist between the preliminary and cur-

Volume 27, Number 5, 1993

rent studies that may explain this finding. The most important difference between the two studies was that, in the preclinical study, the instillate was drained and the abdominal cavity rinsed with normal saline. No attempt was made to remove cisplatin in the present study. This could account for the prolonged half-life and relatively large AUC values in the present study. In addition, differences in pretreatment medications, the level of anesthesia produced in the dogs, and differences in regional/ whole body temperatures between the two populations could have influenced drug distribution and kinetics. Intraperitoneal temperatures were, in general, lower than planned. The power limiting factor (when identified) was usually patient discomfort. Significant toxicities directly attributable to hyperthermia were unusual. It is reasonable, given this scenario, to increase power deposition by increasing patient analgesia. The risks of this approach, however, are significant. This is true for any hyperthermia situation since the pain threshold is generally close to the toxicity threshold (given treatment times of 45-60 min). This risk is higher in the setting of regional power deposition since monitoring maximum normal tissue temperatures can be problematic. The difficulty in monitoring temperatures is a result of sparse temperature monitoring within the treatment region as well as from potential electromagnetic field concentrations outside of the treatment regions (particularly in thighs). Accordingly, analgesia and power increase should be done only in a controlled fashion. Most patients had at least a transient tumor response. This was generally noted as a decline in serum CA- 125 levels. Homogeneous fluid distribution likely did not occur in our patients predominantly as a result of adhesions. It is likely, in the setting of bulky disease, therefore, that both IP doses of cisplatin and IP temperatures will need to be increased to improve response. Therapeutic gain might also be derived from adding other chemotherapeutic agents. In conclusion, we have shown the feasibility of the simultaneous administration of IP cisplatin and regional hyperthermia. The MTD of 80-120 mg/m’ is felt to be too low to be of significant clinical use in patients with bulky disease. Studies are in progress using IV thiosulfate to improve IP cisplatin doses, and well-controlled analgesia to improve IP temperatures.

REFERENCES 1. Casper, E. S.; Kelsen, D. P.; Alcock, N. W.; Lewis, Jr., J. L. IP cisplatin in patients with malignant ascites: Pharmacokinetic evaluation and comparison with the IV route. Cancer Treat. Rep. 67:235-238; 1983. 2. Dedrich, R. L.; Myers, C. E.; Bungay, P. M.; DeVita, Jr., V. T. Pharrnacokinetic rationale for peritoneal drug administration in the treatment of ovarian cancer. Cancer Treat. Rep. 62:1-9; 1978. 3. Eichholtz-Wirth, H.; Hietel, B. Heat sensitization to cisplatin in two cell lines with different drug sensitivity. Int. J. Hyperther. 6:47-55; 1990.

4. Gibaldi, M.; Perrier, D. Pharmacokinetics, York: Marcel Dekker; 1982.

2nd ed. New

5. Hacker, N. F.; Berek, J. S.; Pretorius, R. G.; Zuckerman, J.; Eisenkop, S.; Lagasse, L. D. Intraperitoneal cis-platinum as salvage therapy for refractory epithelial ovarian cancer. Obst. Gyn. 70:759-764; 1987. 6. Hahn, G. M. Hypertherrnia Press; 1982.

and cancer. New York Plenum

7. Herman, T. S. Temperature diamminedichloroplatinum,

dependence of adriamycin, cisbleomycin, and 1,3-bis(2-

Intraoperative cisplatin and regional hyperthermia 0 K. A. LEOPOLD elal.

8.

9. 10.

11.

12.

13.

14.

15.

16.

17.

18.

chloroethyl)- 1-nitrosourea cytotoxicity in vitro. Cancer Res. 43:5 17-520; 1983. Herman, T. S.; Teicher, B. A.; Chan, V.; Collins, L. S.; Abrams, M. J. Effect of heat on the cytotoxicity and interaction with DNA of a series of platinum complexes. Int. J. Radiat. Oncol. Biol. Phys. 16:443-449; 1989. Howell, S. B. Intraperitoneal chemotherapy for ovarian carcinoma. J. Clin. Oncol. 6:1673-1675; 1988. Howell, S. B.; Pfeifle, C. E.; Wung, W. E.; Olshen, R. A. Intraperitoneal cisplatin with systemic sodium thiosufate protection. Cancer Res. 43: 1426-1431; 1983. Howell, S. B.; Taetle, R. Effect of sodium thiosulfate on cisDichlorodiammineplatinum (II) toxicity and antitumor activity in L 12 10 leukemia. Cancer Treat. Rep. 64:6 1 l-6 16; 1980. Leopold, K. A.; Dewhirst, M.; Samulski, T.; Harrelson, J.; Tucker, J. A.; George, S. L.; Dodge, R. K.; Grant, W.; Clegg, S.; Prosnitz, L. R; Oleson, J. R. Relationships among tumor temperature, treatment time, and histopathological outcome using preoperative hyperthennia with radiation in soft tissue sarcomas. Int. J. Radiat. Oncol. Biol. Phys. 22:989-998; 1992. Levin, L.; Hryniuk, W. M. Dose intensity analysis of chemotherapy of advanced ovarian carcinoma. Proc. Am. Sot. Clin. Oncol. 5:1 12; 1986. Mansouri, A.; Henle, K. J.; Benson, A. M.; Moss, A. J.; Nagle, W. A. Characterization of a cisplatin-resistant subline of murine RIF-1 cells and reversal of drug resistance by hyperthermia. Ca Res. 49:2674-2678, 1989. Markman, M. Intraperitoneal antineoplastic agents for tumors principally confined to the peritoneal cavity. Cancer Treat. Rev. 13:219-242; 1986. Markman, M.; Hakes, T.; Reichman, B.; Hoskins, W.; Rubin, S.; Jones, W.; Almadrones, L.; Yordan, E. L.; Eriksson, J.; Lewis, J. L. Intraperitoneal cisplatin and cytarabine in the treatment of refractory or recurrent ovarian carcinoma. J. Clin. One. 9(2):204-210; 199 1. McGahan, M. C.; Tyczkowska, K. The determination of platinum in biological materials by electrothermal atomic absorption spectroscopy. Spectrochimica Acta 42B:665-668; 1987. Ozols, R. F.; Young, R. C. Chemotherapy of ovarian cancer. Sem. Oncol. 11:251-263; 1984.

1251

19. Piver, M. S.; Lele, S. B.; Marchetti, D. L.; Baker, T. R.,; Emrich, L. J.; Hartman, A. B. Surgically documented response to intraperitoneal cisplatin, cytarabine, and bleomycin after intravenous cisplatin-based chemotherapy in advanced ovarian adenocarcinoma. J. Clin. Oncol. 6: 16791684; 1988. 20. Pretorius, R. G.; Hacker, N. F.; Berek, J. S.; Ford, L. C.; Hoeschele, J. D.; Butler, T. A.; Lagasse, L. D. Pharmacokinetics of IP cisplatin in refractory ovarian carcinoma. Cancer Treat. Rep. 67:1085-1092; 1983. 21. Richardson, G. S.; Scully, R. E.; Nikrui, N.; Nelson, J. H. Common epithelial cancer of the ovary. New Eng. J. Med. 312:415-483; 1985. 22. Schilsky, R. L.; Choi, K. E.; Grayhack, J.; Grimmer, D.; Guarnieri, C.; Fullem, L. Phase I clinical and pharmacological study of intrapeitoneal cisplatin and fluorouracil in patients with advanced intraabdominal cancer. J. Clin. Oncol. 8( 12):2054-2061; 1990. 23. Silverberg, E.; Seidman, H.; Mushinski, M. H.; Gelb, S. K. Cancer statistics, 1985. New York, NY: Professional Education Publications, American Cancer Society; 1985. 24. Smith, G. V.; MacMillan, R.; Stribling, J. Intraperitoneal approach to regional hyperthermia-possible anticancer applications. World J. Surg. 7:732-737; 1983. 25. Spratt, J. S.; Adcock, R. A.; Muskovin, M.; Sherrill, W.; McKeown, J. Clinical delivery system for intraperitoneal hyperthermic chemotherapy. Cancer Res. 40:256-260; 1980. 26. Spratt, J. S.; Adcock, R. A.; Sherrill, W.; Travathen, S. Hyperthermic peritoneal perfusion system in canines. Cancer Res. 40:253-255; 1980. 27. Stewart, D. J.; Benjamin, R. S.; Luna, M.; Feun, L.; Caprioli, R; Seifert, W.; Loo, T. L. Human tissue distribution of platinum after cis-diamminedichloroplatinum. Cancer Chemother. Pharmacol. 10:5 l-54; 1982. 28. Wallner, K. E.; Li, G. C. Effect of drug exposure duration and sequencing on hyperthermic potentiation of mitomycinC and cisplatin. Cancer Res. 47:493-495; 1987. 29. Zakris, E. L.; Dewhirst, M. W.; Riviere, J. E.; Hoops, P. J.; Page, R. L. Pharmacokinetics and toxicity of intraperitoneal cisplatin combined with regional hyperthermia. J. Clin. Oncol. 5:1613-1620; 1987.