Consolidation With Intraperitoneal Cisplatin in First-Line
Therapy of Advanced Ovarian Cancer By U. Beller, J. Speyer, N. Colombo, J. Sorich, J. Wernz, H. Hochster, A. Zeleniuch-Jacquotte, R. Porges, and E.M. Beckmant Seventy-five patients with advanced epithelial ovarian cancer were treated with a combined modality regimen of systemic, induction chemotherapy followed by intraperitoneal therapy (IPT). All patients underwent initial surgery for staging and/or cytoreduction followed by cisplatin 20 mg/m2 intravenously (IV) for 5 days and cyclophosphamide 600 mg/m2 on day 4 every 3 to 4 weeks for two to four cycles. Patients were then evaluated for IPT and, if eligible, had an intraperitoneal (IP) catheter placed. IPT consisted of cisplatin 60 mg/m2 in 2 L on day 1 and IV cyclophosphamide 600 mg/m2 on day 2 every 3 weeks for three to six cycles. Patients who demonstrated a clinical complete response (CCR) were then referred for second-look laparotomy (SLL). Of 71 patients who completed the induction phase, 53 (75%) were eligible for IPT, and 49 patients entered the therapy phase. Toxicity of the combined modality approach was acceptable and did not differ from our previous experience using the same drugs systemically. Thirty-two of the 49 patients who
THE
PHARMACOLOGIC advantage of in-
traperitoneal (IP) instillation of chemothera-
peutic agents has been carefully studied and confirmed in numerous clinical trials. Pharmacologic
studies on a variety of compounds have shown that a large concentration gradient exists between the
peritoneal fluid and the plasma following IP administration.1-3 As a result, the peritoneal surfaces are
completed IPT achieved a CCR, which was confirmed by SLL in 20 patients. Twenty recurrences were documented in the 32 CCR patients, 13 occurred in patients after SLL. Projected median survival of all patients is 38 months. Median survival correlated with amount of residual disease following initial surgery (23 months for bulky v 45 months for minimal residual; P < .001) and with performance status ([PS]; 24 months for PS 2, 3 v > 46 months for PS 0; P < .001). Patients who presented with bulky tumors were less likely to reach the consolidation IPT phase. Incorporation of IP cisplatin into the first-line regimen for treatment of ovarian cancer does not appear to have major impact on the survival of all treated patients when compared with our historical control series. Combined IV and IPT cisplatin and cyclophosphamide is feasible with acceptable toxicity. Its impact on response and survival may be limited to only "good-prognosis" patients. J Clin Oncol 9:809-817. © 1991 by American Society of Clinical Oncology. the disease locally, using a high concentration of an active agent administered directly to the area at
risk.l 3 '5 The drug then enters the tumor by diffusion as well as through the bloodstream after systemic absorption. These principles have been the basis for a number of phase I-II trials in which ovarian cancer patients who failed or recurred following initial
exposed to higher drug concentrations for longer periods of time when compared with plasma concentrations, producing favorable area under the
curve (AUC) ratios for peritoneal cavity to blood exposures. For two main reasons, ovarian cancer presents a particularly good model in which to use this
pharmacologic advantage of the peritoneal cavity. A dose-response relationship appears to exist in the chemotherapeutic treatment of ovarian cancer, thus justifying the search for ways to administer higher doses of cytotoxic agents without expos-
ing the patient to the potential life-threatening toxicities of systemic therapy.4' 5 A relationship between dose intensity and response also appears to exist. 6 Since ovarian cancer is primarily confined
to the peritoneal cavity, it is advantageous to treat
From the Kaplan Cancer Center, Division of Gynecologic Oncology, Departmentof Obstetrics and Gynecology, Division of Medical Oncology, and Departmentof Medicine, New York University Medical Center, New York, NY SubmittedJune 19, 1990; acceptedNovember 2, 1990. fDeceased. Supported in part by a grant from the Blinken Ovarian CancerResearch Fundand Public Health Service grantno. CA 16087. Dr Belier is now Director of Gynecologic Oncology at the Shaare Zedek Medical Center,Jerusalem, Israel. N. Colombo, MD, is a recipient of a fellowship from the American ItalianFoundationforCancerResearch. Address reprint requests to J. Speyer, MD, New York University Medical Center, 530 FirstAve, Suite 4J, New York, NY10016. © 1991 by American Society of Clinical Oncology. 0732-183X9110905-0010$3.00/0
Journalof Clinical Oncology, Vol 9, No 5 (May), 1991: pp 809-817
Downloaded from ascopubs.org by UNIVERSITY of OTAGO on April 21, 2019 from 139.080.135.089 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
809
810
BELLER ET AL
systemic therapy were managed by IP therapy (IPT). 1-3,7-10 In one published report summarizing the results of a number of trials with IP cisplatin, some patients achieved prolonged survival, especially when treated for small (< 2 cm) residual disease (median survival, > 49 months).8 Cisplatin has been used because of its antitumor activity in ovarian cancer, proven pharmacologic advantage in IPT, and the minimal local peritoneal toxicity it causes. The better results seen with small residual disease may be explained by better tumor penetration of cisplatin administered into the peritoneal cavity. 1 Interestingly, the published trials have described successful IPT even in patients who had shown complete responses after primary cisplatincontaining regimens. Patients responded to the same drug when given IP even though these recurrent or residual tumors were often resistant 7 to further systemic therapy with cisplatin. '12 Our experience and that of others has shown that despite good initial responses to combination chemotherapy, most patients eventually recur in the peritoneal cavity. Therefore one way of reducing the incidence of this phenomenon may be through the use of early local treatment via the IP route. It was our hypothesis that IPT could potentially consolidate the initial tumor response and improve the overall survival of patients if introduced earlier in the course of therapy, before drug
resistance developed, and as part of the initial treatment regimen that followed surgical debulking. This is one of the first reports of a clinical trial in which IP chemotherapy has been used to consolidate systemic chemotherapy in the planned first-line treatment following surgery for advanced 13 ovarian cancer. ',14
MATERIALS AND METHODS From January 1984 through December 1988, patients with the diagnosis of stages II-IV epithelial ovarian cancer at New York University-Bellevue Medical Center, were candidates for this phase II trial. Patients were required to have histologically proven adenocarcinoma of the ovary and surgical staging. Patients were excluded on the basis of the following: if they could not be treated with platinum combinations due to renal or neurologic dysfunction, if they had expressed unwillingness to have a catheter placed or to have IPT, the presence of another malignancy, and previous treatment with cisplatin. All other patients were considered potential participants in this trial, regardless of tumor bulk following initial surgery, tumor histologic type, grade of tumor, and performance status (PS). Patients were treated according to the regimen described in Fig 1. Preoperative evaluation also included an abdominalpelvic computed tomographic (CT) scan and PS evaluation. In essence, patients had initial surgery performed for staging and tumor reduction followed by two phases of chemotherapy. Surgery consisted of total abdominal hysterectomy, bilateral salpingo-oophorectomy, omentectomy, and cytoreductive surgery. In patients with no gross evidence of peritoneal spread, staging laparotomy included multiple biopsies of peritoneal surfaces as well as retroperitoneal lymph node sampling as previously described."5 The first
SYSTEMIC CHEMOTHERAPY 2
q 21-28 days 3-4 cycles
cisplatin 20 mg/m x 5 days 2 cyclophosphamide 600 mg/m day 4
iv-
NON INVASIVE EVALUATION FOR INTRAPERITONEAL THERAPY
INELIGIBLE FOR IP PHASE
ELIGIBLE FOR IP PHASE
continue systemic chemotherapy
IP cisplatin 60 mg/m2 in 2 L IV cyclophosphamide 600 mg/m2
CLINICAL EVALUATION
PERSISTENT DISEASE
CCR Second-look laparotomy
PCR
RD
Second-line chemotherapy
Fig 1. Schema of protocol design: Initial surgery for ovarian cancer. PCR, pathologic complete response; RD, residual disease.
Downloaded from ascopubs.org by UNIVERSITY of OTAGO on April 21, 2019 from 139.080.135.089 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
811
FIRST-LINE IV AND IP CISPLATIN phase of chemotherapy consisted of systemic chemotherapy with cisplatin 20 mg/m2 in 1,000 cc dextrose 5% (D5) one half 0.45% normal saline given over 2 hours daily for 5 days, and cyclophosphamide 600 mg/m 2 in 500 cc D5/water on day 4. Cycles were repeated every 3 to 4 weeks (depending on hematologic toxicity) for a total of three to four planned cycles.1' 617 We allowed some flexibility in the number of intravenous (IV) cycles based on the amount of disease at the initiation of treatment. Patients were reassessed for response after two cycles by physical examination and/or CT scan. The decision whether to administer up to two additional cycles was based on persistence of bulk residual disease. Thus, patients with bulky disease were allowed up to four cycles to maximize tumor reduction before IPT. On completion of this first phase, patients were evaluated by noninvasive modalities for eligibility for the second phase with IPT. For IPT, patients were to receive a planned six courses of IP cisplatin and IV cyclophosphamide. Patients who demonstrated disease progression or persistent bulky peritoneal carcinomatosis, those who had developed severe medical contraindications (eg, renal dysfunction, peripheral neuropathy) during the first phase of treatment, or those who had known IP fluid distribution problems related to adhesions were excluded from the second (IP) phase. These excluded patients were then further treated with systemic IV chemotherapy. Catheters were introduced under general anesthesia through a small (< 5 cm) left paramedian incision at the level of the umbilicus. Five patients who had had an incomplete initial surgery underwent a full exploratory laparotomy at this time with complete removal of tumor and genital organs. We preferred the use of a catheter attached to a subcutaneous port (Infusaport; Intermedics Infusaid, Inc, Norwood, MA), although some patients (including the earliest patients entered and the very obese patients) had a standard Tenckhoff catheter (Quinton Instrument Co, Seattle, WA)."1 Immediately following insertion of the catheter, a series of peritoneal lavages were performed using 1.5% dialysate solution supplemented with potassium chloride (KCI) (4 mEq/L) and heparin (500 U/L), with a gradual increase of fluid volume reaching 2,000 cc on the fourth postoperative day. An abdominal-pelvic CT scan was then performed with IP contrast 9 to document proper distribution of fluid throughout the peritoneal cavity. Documentation of adequate fluid distribution throughout the peritoneal cavity, including the under surface of both diaphragms and the pelvic floor, was a requisite for entry into the IPT phase. IPT was then begun in accordance with the dose and schedule of drug administration described in Fig 1. Cisplatin (60 mg/m2 ) was administered in 2 L of sodium chloride supplemented with KCl (4 mEq/l) and heparin (500 U/L), and allowed to dwell for 4 hours. A dwell time of 4 hours was chosen to try to make the duration of drug exposure uniform in all patients at a time that most cisplatin would be absorbed.3 In some cases, when catheter outflow stopped this was not possible. Fluid was drained out when possible and followed by a 1 L wash with supplemented dialysate solution. Cyclophosphamide was given IV at a dose of 600 mg/m 2 on day 2. Antiemetic therapy was given to all patients prophylactically and consisted of lorazepam 1 to 2 mg intramuscularly (IM), diphenhydramine 25 to 50 mg IM,
prochlorperazine 10 mg orally and/or IM, or metoclopramide 2 to 3 mg/kg IV every 2 hours x three daily doses, and dexamethasone 20 mg IV before cisplatin. Additional prochlorperazine and lorazepam were given every 4 hours as needed. Patients were to receive six courses of cisplatin IP and cyclophosphamide IV. Reasons for attenuation of this therapy included the development of significant grade 3 peripheral neuropathy, rise in serum creatinine > 2.0, catheter-related pain or blockage, infection, or patient refusal. At the completion of the planned number of treatments, noninvasive evaluation of tumor response was carried out by physical examination, chest x-ray, abdominal-pelvic CT scan,'2 peritoneal fluid cytology, and tumor marker studies. If negative, patients were considered to have a clinical complete response (CCR) and a second-look laparotomy (SLL) was performedl5 2 • if the patient consented. Patients with only minimal residual disease at presentation were considered in CCR if tumor markers and noninvasive evaluation showed no progression of disease. These patients were then offered SLL. It was determined that patients with a negative SLL had a pathologic complete response (PCR). They received no further treatment and were observed closely. If residual disease was found at SLL, patients received further therapy with other regimens.
StatisticalMethods Fisher's exact test was used in the analysis of contingency tables. The method of Kaplan and Meier22 was used to estimate survival curves. All patients were included in the analysis. The patients were censored at the time they were last seen alive, whether they completed the protocol or not. The log-rank test," and log-rank test for trend 24 were used to assess differences in survival distributions. Unadjusted median survivals were estimated from Kaplan-Meier curves. Cox's proportional hazards multivariate model25 was used to assess the prognostic value of residual tumor bulk and PS when considered simultaneously. All P values are two-sided. RESULTS Patientsand Treatment
Seventy-five patients were entered on this protocol. Their characteristics are presented in Table 1. Ninety-two percent of the patients had stage III or IV disease. Forty-nine percent started induction chemotherapy with residual disease larger than 2 cm in diameter. Seventy-one percent had a PS of 0 to 1, 29% had a PS of 2 to 3. The applicability and compliance of this protocol are described in the
flow chart (Fig 2). The treatment data are presented in Table 2. Four patients refused to complete the induction IV phase and are not assessable for response. The toxicities they demonstrated during the first treatment were not different from those of other patients. The median number of IV cycles administered was four; one patient was
Downloaded from ascopubs.org by UNIVERSITY of OTAGO on April 21, 2019 from 139.080.135.089 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
812
BELLER ET AL Table 1. Patient Characteristics (N
Age (years) Median Mean Range Stage II III IV Histologic grade 1 2 3 Residual disease > 2 cm
Therapy of Advanced Ovarian Cancer By U. Beller, J. Speyer, N. Colombo, J. Sorich, J. Wernz, H. Hochster, A. Zeleniuch-Jacquotte, R. Porges, and E.M. Beckmant Seventy-five patients with advanced epithelial ovarian cancer were treated with a combined modality regimen of systemic, induction chemotherapy followed by intraperitoneal therapy (IPT). All patients underwent initial surgery for staging and/or cytoreduction followed by cisplatin 20 mg/m2 intravenously (IV) for 5 days and cyclophosphamide 600 mg/m2 on day 4 every 3 to 4 weeks for two to four cycles. Patients were then evaluated for IPT and, if eligible, had an intraperitoneal (IP) catheter placed. IPT consisted of cisplatin 60 mg/m2 in 2 L on day 1 and IV cyclophosphamide 600 mg/m2 on day 2 every 3 weeks for three to six cycles. Patients who demonstrated a clinical complete response (CCR) were then referred for second-look laparotomy (SLL). Of 71 patients who completed the induction phase, 53 (75%) were eligible for IPT, and 49 patients entered the therapy phase. Toxicity of the combined modality approach was acceptable and did not differ from our previous experience using the same drugs systemically. Thirty-two of the 49 patients who
THE
PHARMACOLOGIC advantage of in-
traperitoneal (IP) instillation of chemothera-
peutic agents has been carefully studied and confirmed in numerous clinical trials. Pharmacologic
studies on a variety of compounds have shown that a large concentration gradient exists between the
peritoneal fluid and the plasma following IP administration.1-3 As a result, the peritoneal surfaces are
completed IPT achieved a CCR, which was confirmed by SLL in 20 patients. Twenty recurrences were documented in the 32 CCR patients, 13 occurred in patients after SLL. Projected median survival of all patients is 38 months. Median survival correlated with amount of residual disease following initial surgery (23 months for bulky v 45 months for minimal residual; P < .001) and with performance status ([PS]; 24 months for PS 2, 3 v > 46 months for PS 0; P < .001). Patients who presented with bulky tumors were less likely to reach the consolidation IPT phase. Incorporation of IP cisplatin into the first-line regimen for treatment of ovarian cancer does not appear to have major impact on the survival of all treated patients when compared with our historical control series. Combined IV and IPT cisplatin and cyclophosphamide is feasible with acceptable toxicity. Its impact on response and survival may be limited to only "good-prognosis" patients. J Clin Oncol 9:809-817. © 1991 by American Society of Clinical Oncology. the disease locally, using a high concentration of an active agent administered directly to the area at
risk.l 3 '5 The drug then enters the tumor by diffusion as well as through the bloodstream after systemic absorption. These principles have been the basis for a number of phase I-II trials in which ovarian cancer patients who failed or recurred following initial
exposed to higher drug concentrations for longer periods of time when compared with plasma concentrations, producing favorable area under the
curve (AUC) ratios for peritoneal cavity to blood exposures. For two main reasons, ovarian cancer presents a particularly good model in which to use this
pharmacologic advantage of the peritoneal cavity. A dose-response relationship appears to exist in the chemotherapeutic treatment of ovarian cancer, thus justifying the search for ways to administer higher doses of cytotoxic agents without expos-
ing the patient to the potential life-threatening toxicities of systemic therapy.4' 5 A relationship between dose intensity and response also appears to exist. 6 Since ovarian cancer is primarily confined
to the peritoneal cavity, it is advantageous to treat
From the Kaplan Cancer Center, Division of Gynecologic Oncology, Departmentof Obstetrics and Gynecology, Division of Medical Oncology, and Departmentof Medicine, New York University Medical Center, New York, NY SubmittedJune 19, 1990; acceptedNovember 2, 1990. fDeceased. Supported in part by a grant from the Blinken Ovarian CancerResearch Fundand Public Health Service grantno. CA 16087. Dr Belier is now Director of Gynecologic Oncology at the Shaare Zedek Medical Center,Jerusalem, Israel. N. Colombo, MD, is a recipient of a fellowship from the American ItalianFoundationforCancerResearch. Address reprint requests to J. Speyer, MD, New York University Medical Center, 530 FirstAve, Suite 4J, New York, NY10016. © 1991 by American Society of Clinical Oncology. 0732-183X9110905-0010$3.00/0
Journalof Clinical Oncology, Vol 9, No 5 (May), 1991: pp 809-817
Downloaded from ascopubs.org by UNIVERSITY of OTAGO on April 21, 2019 from 139.080.135.089 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
809
810
BELLER ET AL
systemic therapy were managed by IP therapy (IPT). 1-3,7-10 In one published report summarizing the results of a number of trials with IP cisplatin, some patients achieved prolonged survival, especially when treated for small (< 2 cm) residual disease (median survival, > 49 months).8 Cisplatin has been used because of its antitumor activity in ovarian cancer, proven pharmacologic advantage in IPT, and the minimal local peritoneal toxicity it causes. The better results seen with small residual disease may be explained by better tumor penetration of cisplatin administered into the peritoneal cavity. 1 Interestingly, the published trials have described successful IPT even in patients who had shown complete responses after primary cisplatincontaining regimens. Patients responded to the same drug when given IP even though these recurrent or residual tumors were often resistant 7 to further systemic therapy with cisplatin. '12 Our experience and that of others has shown that despite good initial responses to combination chemotherapy, most patients eventually recur in the peritoneal cavity. Therefore one way of reducing the incidence of this phenomenon may be through the use of early local treatment via the IP route. It was our hypothesis that IPT could potentially consolidate the initial tumor response and improve the overall survival of patients if introduced earlier in the course of therapy, before drug
resistance developed, and as part of the initial treatment regimen that followed surgical debulking. This is one of the first reports of a clinical trial in which IP chemotherapy has been used to consolidate systemic chemotherapy in the planned first-line treatment following surgery for advanced 13 ovarian cancer. ',14
MATERIALS AND METHODS From January 1984 through December 1988, patients with the diagnosis of stages II-IV epithelial ovarian cancer at New York University-Bellevue Medical Center, were candidates for this phase II trial. Patients were required to have histologically proven adenocarcinoma of the ovary and surgical staging. Patients were excluded on the basis of the following: if they could not be treated with platinum combinations due to renal or neurologic dysfunction, if they had expressed unwillingness to have a catheter placed or to have IPT, the presence of another malignancy, and previous treatment with cisplatin. All other patients were considered potential participants in this trial, regardless of tumor bulk following initial surgery, tumor histologic type, grade of tumor, and performance status (PS). Patients were treated according to the regimen described in Fig 1. Preoperative evaluation also included an abdominalpelvic computed tomographic (CT) scan and PS evaluation. In essence, patients had initial surgery performed for staging and tumor reduction followed by two phases of chemotherapy. Surgery consisted of total abdominal hysterectomy, bilateral salpingo-oophorectomy, omentectomy, and cytoreductive surgery. In patients with no gross evidence of peritoneal spread, staging laparotomy included multiple biopsies of peritoneal surfaces as well as retroperitoneal lymph node sampling as previously described."5 The first
SYSTEMIC CHEMOTHERAPY 2
q 21-28 days 3-4 cycles
cisplatin 20 mg/m x 5 days 2 cyclophosphamide 600 mg/m day 4
iv-
NON INVASIVE EVALUATION FOR INTRAPERITONEAL THERAPY
INELIGIBLE FOR IP PHASE
ELIGIBLE FOR IP PHASE
continue systemic chemotherapy
IP cisplatin 60 mg/m2 in 2 L IV cyclophosphamide 600 mg/m2
CLINICAL EVALUATION
PERSISTENT DISEASE
CCR Second-look laparotomy
PCR
RD
Second-line chemotherapy
Fig 1. Schema of protocol design: Initial surgery for ovarian cancer. PCR, pathologic complete response; RD, residual disease.
Downloaded from ascopubs.org by UNIVERSITY of OTAGO on April 21, 2019 from 139.080.135.089 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
811
FIRST-LINE IV AND IP CISPLATIN phase of chemotherapy consisted of systemic chemotherapy with cisplatin 20 mg/m2 in 1,000 cc dextrose 5% (D5) one half 0.45% normal saline given over 2 hours daily for 5 days, and cyclophosphamide 600 mg/m 2 in 500 cc D5/water on day 4. Cycles were repeated every 3 to 4 weeks (depending on hematologic toxicity) for a total of three to four planned cycles.1' 617 We allowed some flexibility in the number of intravenous (IV) cycles based on the amount of disease at the initiation of treatment. Patients were reassessed for response after two cycles by physical examination and/or CT scan. The decision whether to administer up to two additional cycles was based on persistence of bulk residual disease. Thus, patients with bulky disease were allowed up to four cycles to maximize tumor reduction before IPT. On completion of this first phase, patients were evaluated by noninvasive modalities for eligibility for the second phase with IPT. For IPT, patients were to receive a planned six courses of IP cisplatin and IV cyclophosphamide. Patients who demonstrated disease progression or persistent bulky peritoneal carcinomatosis, those who had developed severe medical contraindications (eg, renal dysfunction, peripheral neuropathy) during the first phase of treatment, or those who had known IP fluid distribution problems related to adhesions were excluded from the second (IP) phase. These excluded patients were then further treated with systemic IV chemotherapy. Catheters were introduced under general anesthesia through a small (< 5 cm) left paramedian incision at the level of the umbilicus. Five patients who had had an incomplete initial surgery underwent a full exploratory laparotomy at this time with complete removal of tumor and genital organs. We preferred the use of a catheter attached to a subcutaneous port (Infusaport; Intermedics Infusaid, Inc, Norwood, MA), although some patients (including the earliest patients entered and the very obese patients) had a standard Tenckhoff catheter (Quinton Instrument Co, Seattle, WA)."1 Immediately following insertion of the catheter, a series of peritoneal lavages were performed using 1.5% dialysate solution supplemented with potassium chloride (KCI) (4 mEq/L) and heparin (500 U/L), with a gradual increase of fluid volume reaching 2,000 cc on the fourth postoperative day. An abdominal-pelvic CT scan was then performed with IP contrast 9 to document proper distribution of fluid throughout the peritoneal cavity. Documentation of adequate fluid distribution throughout the peritoneal cavity, including the under surface of both diaphragms and the pelvic floor, was a requisite for entry into the IPT phase. IPT was then begun in accordance with the dose and schedule of drug administration described in Fig 1. Cisplatin (60 mg/m2 ) was administered in 2 L of sodium chloride supplemented with KCl (4 mEq/l) and heparin (500 U/L), and allowed to dwell for 4 hours. A dwell time of 4 hours was chosen to try to make the duration of drug exposure uniform in all patients at a time that most cisplatin would be absorbed.3 In some cases, when catheter outflow stopped this was not possible. Fluid was drained out when possible and followed by a 1 L wash with supplemented dialysate solution. Cyclophosphamide was given IV at a dose of 600 mg/m 2 on day 2. Antiemetic therapy was given to all patients prophylactically and consisted of lorazepam 1 to 2 mg intramuscularly (IM), diphenhydramine 25 to 50 mg IM,
prochlorperazine 10 mg orally and/or IM, or metoclopramide 2 to 3 mg/kg IV every 2 hours x three daily doses, and dexamethasone 20 mg IV before cisplatin. Additional prochlorperazine and lorazepam were given every 4 hours as needed. Patients were to receive six courses of cisplatin IP and cyclophosphamide IV. Reasons for attenuation of this therapy included the development of significant grade 3 peripheral neuropathy, rise in serum creatinine > 2.0, catheter-related pain or blockage, infection, or patient refusal. At the completion of the planned number of treatments, noninvasive evaluation of tumor response was carried out by physical examination, chest x-ray, abdominal-pelvic CT scan,'2 peritoneal fluid cytology, and tumor marker studies. If negative, patients were considered to have a clinical complete response (CCR) and a second-look laparotomy (SLL) was performedl5 2 • if the patient consented. Patients with only minimal residual disease at presentation were considered in CCR if tumor markers and noninvasive evaluation showed no progression of disease. These patients were then offered SLL. It was determined that patients with a negative SLL had a pathologic complete response (PCR). They received no further treatment and were observed closely. If residual disease was found at SLL, patients received further therapy with other regimens.
StatisticalMethods Fisher's exact test was used in the analysis of contingency tables. The method of Kaplan and Meier22 was used to estimate survival curves. All patients were included in the analysis. The patients were censored at the time they were last seen alive, whether they completed the protocol or not. The log-rank test," and log-rank test for trend 24 were used to assess differences in survival distributions. Unadjusted median survivals were estimated from Kaplan-Meier curves. Cox's proportional hazards multivariate model25 was used to assess the prognostic value of residual tumor bulk and PS when considered simultaneously. All P values are two-sided. RESULTS Patientsand Treatment
Seventy-five patients were entered on this protocol. Their characteristics are presented in Table 1. Ninety-two percent of the patients had stage III or IV disease. Forty-nine percent started induction chemotherapy with residual disease larger than 2 cm in diameter. Seventy-one percent had a PS of 0 to 1, 29% had a PS of 2 to 3. The applicability and compliance of this protocol are described in the
flow chart (Fig 2). The treatment data are presented in Table 2. Four patients refused to complete the induction IV phase and are not assessable for response. The toxicities they demonstrated during the first treatment were not different from those of other patients. The median number of IV cycles administered was four; one patient was
Downloaded from ascopubs.org by UNIVERSITY of OTAGO on April 21, 2019 from 139.080.135.089 Copyright © 2019 American Society of Clinical Oncology. All rights reserved.
812
BELLER ET AL Table 1. Patient Characteristics (N
Age (years) Median Mean Range Stage II III IV Histologic grade 1 2 3 Residual disease > 2 cm
