High-dose,.pdf

High-Dose, Short-Duration Cisplatin/ Doxorubicin Combination Chemotherapy for Advanced Ovarian Epithelial Cancer Richard E. Hunter, MD,* Thomas W. Griffin, MD,t Sarah Stevens, MD,t Lynda D. Roman, MD,* Faran Bokhari, PhD,t Frank R. Reale, MD,$ Won K. Tak, MD,§ T. J. Fitzgerald, MD,§ Michael B. Dillon, MD,* and Peter G. Rose, MD* Sixty-one patients with epithelial ovarian cancer were treated with intensive highdose, short-course chemotherapy that consisted of cisplatin (120mg/m') and doxorubicin (70 mg/m') every 3 weeks for four cycles. Patients in complete clinical remission were offered second-look laparotomy (SLL). Patients with minimal or no residual disease at SLL were randomized to either cyclophosphamide (1000 mg/m2 every 21 days for six cycles) or whole-abdominal radiation therapy. All patients completed therapy with a median leukocyte nadir 1,3/p1 and platelet nadir of 9O/p1. Forty-five patients (74%) had a complete clinical response. Results of twenty-two of 36 second-look procedures (64%) showed no evidence of disease (NED). After SLL, 19 patients received six courses of cyclophosphamide and 16 patients received whole-abdominal radiation. Nine patient who refused SLL and one patient with negative SLL findings refused additional treatment. The median survival time for all patients was 51.3 months. High-dose intensive chemotherapy regimens have high response rates, but survival needs to be compared with traditional low-dose regimens. Although high-dose cisplatin and doxorubicin were myelosuppressive, the resulting complications were manageable. There was no significant difference between the mean survival times of patients receiving Cytoxan, abdominal radiation, or no treatment as second-line therapy. Cancer 68:1890-1894,1991.

D

epithelial Carcinoma O f the ovary remains the leading cause of death among gynecologic malignancies. The introduction of cisplatinbased combination chemotherapy has led to increases in response rates, complete response rates, and median duration of survival in patients with advanced (Stages I11 and IV) ovarian cancer. However, the number of diseaseESPITE RECENT ADVANCES,

'

From the Departments of *Obstetrics and Gynecology, Division of Gynecologic Oncology, ?Medicine, Division of Oncology, $Pathology, and §Radiation Therapy, University of Massachusetts Medical Center. Worcester, Massachusetts. The authors thank Joyce Sirois, RN; Janet H. Hagstrom for coordinating and typing the manuscript: JefY Collins for technical assistance; and all investigators and patients who participated in this study. Address for reprints: Richard E. Hunter, MD, Department of Obstetrics and Gynecology, University of Massachusetts Medical Center, 55 Lake Avenue North, Worcester, MA 01655. Accepted for publication August I , 1991.

free long-term survivors has remained disappointingly smalL2 Therefore, new therapeutic strategies for this disease are urgently needed. The optimal agent or agents, timing of administration, optimum dose, and optimal duration of therapy is yet to be determined. The importance of drug dose in determining tumor response to chemotherapy has been frequently ~tressed.~-~ Clinical trials have shown a clear dose-response effect with cisplatin.6 However, few studies have been performed to determine the effect of high-dose intensity and short-duration chemotherapy as a possible determinant of outcome. To evaluate such a dose-response in advanced ovarian cancer, this study was designed to investigate highdose, short-duration cisplatin/doxorubicin combination chemotherapy with early second-look laparotomy (SLL). The patients with minimal or no residual disease after SLL were randomized to high-dose cyclophosphamide or whole-abdominal radiation.

1890

HIGH-DOSE ClSPLATIN/DOXORUBlCININ OVARIAN CA

No. 9

BASIC SCHEMA OF THE TREATMENT PLAN Maximal Surgery

4

\

\ / \' /\ Second-look Laparotomv

Cyclophospharnide (1.000 rng/m2 x 6)

Residual Disease >2 cm Oft Study

Total Abdominal Radiation

FIG. 1. Basic scheme of the treatment plan

6't

d.

1891

61 32-73 53 56

No. of patients Age range (yr) Mean age (yr) Median age (yr) Stage 1 Stage II Stage 111 Stage IV Total

1

6 41

7 61

Results

Putient Prof2le Sixty-one previously untreated patients with epithelial ovarian cancer were originally entered in the study. Patient eligibility required a histologically documented epithelial carcinoma of the ovary, no previous or concomitant malignancy, Eastern Cooperative Oncology Group ( ECOG) performance status of 0 to 2, and informed consent. Patient profile and and pathologic condition are shown in Tables 1 and 2, respectively. Two patients were evaluable only for toxicity because one died of renal failure during the first course of chemotherapy and the second died of septicemia after the fourth course of chemotherapy. The majority of patients had a poorly differentiated serous cystadenocarcinorna. There were 47 Stage 111 patients and 7 Stage IV patients. The majority of Stage IV patients were classified in that stage because of cytologically malignant pleural effusions. Forty-two patients (69%) had their tumors cytoreduced to less than 2 cm, and 19 patients ( 3 1%) had residual tumors greater than 2 cni after the initial operation.

Toxicity The myelosuppression and toxicity associated with the cisplatin/doxorubicin regimen is shown in Tables 3 and 4. Thirty-eight patients (60%) experienced nausea and vomiting. The median nadir leukocyte count was 1.3/pl for all the cycles administered. A mean nadir leukocyte

TABLE2.

Cycles of Cisplatinum (120 mum2) and Adriamycin (100 mum2)

Residual Disease <2 cm

Hwzter

TABLEI. Patient Profle

Materials and Methods From 1985 to 1988,6 I patients with epithelial ovarian cancer Stages IC to IV who treated at the University of Massachusetts and affiliated hospitals were entered in a prospective study. All patients initially had maximum cytoreductive surgery. After pathologic confirmation of an invasive epithelial ovarian carcinoma, patients were invited to participate in this institutionally approved study. Within 14 days of surgery, chemotherapy consisting of cisplatin ( I20 mg/m2) and doxorubicin (70 mg/m') was administered every 2 1 days for four cycles. Because of the dose-intensity of cisplatin, patients were carefully hydrated. Patients were prehydrated with 6 1 of normal saline over 24 hours. Cisplatin was then administered in 6 1 of normal saline over the second 24-hour period. Patients then received an additional 6 1 of normal saline posthydration over the third 24-hour period. Doxorubicin was given as a bolus injection. Antiemetic coverage was routinely provided with haloperidol, phenobarbitol, and dexamethasone. Patients received full supportive care including transfusion of blood products, antibiotics, and nutritional support when appropriate. Doses were not modified for myelosuppression or gastrointestinal toxicity. All patients without clinical evidence of disease at the completion of chemotherapy were offered SLL to determine the extent of response. Patients with residual disease less than 2 cm or no disease at SLL were offered randomization to either high-dose cyclophosphamide or total-abdominal radiation therapy. Patients with residual disease of 2 cm or greater at SLL were removed from study. Cyclophosphamide was administered as outpatient therapy at 1 g/m' every 2 I days for six cycles. Patients randomized to radiation therapy received 3000 cGy to the whole abdomen followed by a pelvic boost of an additional 2000 cGy. Figure I outlines the basic scheme of the treatment plan. Patient survivals were compared by the MantelCox statistic.

-

Pathologic Condition

Histologic grade Well differentiated Moderately differentiated Poorly differentiated Total T u m o r histologic condition Mucinous cystadenocarcinoma Serous cystadenocarcinoma Endometrioid adenocarcinorna Clear cell Mixed tumor Undifferentiated adenocarcinoma Total

6 9 46 61 3 41

4 4 2 7

61

CANCER November 1 1991

1892

TABLE3. Myelosuppression and Nephrotoxicity of the Cisplatin/Doxorubicin Regimen

Vol. 68

TABLE5.

Long-Term Survivor Characteristics* Initial surgery

Nadirs

Mean

Medians

Leuk ( ~ 1 ) Plt (PI) HGB (g/dl) HCT (%) Peak creat (mg/dl)

1.7 k 1.2 122 t 118 9.7 i 1.5 28.8 t 4.7 1.1 k 0.5

I .3 90.0 9.8 28.8 0.9

Residual disease

<2cm >2cm Total

No. of patients

No. of long-term survivors

(%)

(”/.)

42 (69) 19 (31) 61

23 (54) 6 (31) 29 (47)

Leuk: leukocyte count X 100 units. Plt: platelet count X 1000 units; HGB: hemoglobin: HCT. hematocrit; Creat: creatinine.

count of 1.9/pl or less was associated with 69% of the courses. Thirty-two patients were admitted to the hospital with documented leukopenia, and seven patients had documented septicemia. There was one death due to septicemia. Significant thrombocytopenia was seen with a median nadir platelet count of 9O/pl. Forty-four percent of the courses were associated with nadir platelet counts of less than 25/p1. The thrombocytopenia was successfully managed with platelet transfusions and was not associated with any mortality. Twenty-six patients were admitted to the hospital for packed erythrocyte transfusions. Nephropathy was evident from serum creatinine levels of greater than 1.5 mg/dl associated with 10%ofthe cisplatin/doxorubicir. courses. One patient died of renal failure. Peripheral neuropathy was a problem with morbidity. Twenty-seven patients (46%)had symptoms of peripheral neuropathy. The majority were mild to moderate, with limitations in buttoning buttons and some limitations of patient gait. Two patients (3%)were severely affected and required a walker or cane.

Tumor Response Forty-five patients (74%) had a complete clinical response. Thirty-six patients underwent second-look procedures (34 laparotomies and 2 laparoscopies). Both patients who underwent a second-look laparoscopy showed no evidence of disease (NED). Nine patients refused a second-look operation. Twenty-two patients (63%) who underwent second-look procedures showed NED. Of these, 19 (86%) had their tumors cytoreduced to less than 2 cm and 3 (14%) had residual tumors greater than 2 cm

Second-look surgervt NED$ Microscopic Macroscopic (cm) < 2-6 > 2-5 Total

22 (61) 3 (8)

16 (73) 0 (0)

1 1 (31) 36

1 (9) 17

NED: no evidence of disease. * Survival z 40 months. t Forty-five eligible for second-look laparotomy, 9 refused operations. $ Two laparoscopies.

after initial surgery (Table 5). Of the positive second-look results, three were microscopic, six had less than 2 cm of tumor, and five had greater than 2 cm of tumor.

Patient Survival There were 29 long-term (greater than 40 months) survivors (49%). One was Stage I (3%), four were Stage I1 (14%), 20 Stage 111 (68%), and 4 Stage IV (1 3%) (Table 6). The long-term survival rate of Stages I and I1 patients was 7 1%, and the long-term survival rate of advanced cases (Stages 111 and IV) was 46%. The effect of initial tumor debulking on patient survival is shown in Figure 2. The median survival time for patients debulked to less than 2 cm disease is 46.9 -t 3.7 months compare with 22.5 t- 4.1 months for patients debulked to greater than 2 cm residual disease at initial surgery. A Mantel-Cox statistical test shows these survival times to be significant to P = 0.006. Sixteen of the patients with negative SLL findings were long-term survivors, and this constitutes 55% of all long-term survivors. The patient survival curve for

TABLE6. Long-Term* Survivor Characteristics ~~

TABLE4.

Myelosuppression and Nephrotoxicity of the Cisplatin/ Doxorubicin Regimen: Patient Distribution

Nadir leukocyte count

Nadir platelet count

Peak creatinine level

> 4.0 (8)

> 100 (44)

> 1.2 (15)

3.0-3.9 (6) 2.0-2.9 ( I 8) 1.0-1.9 (35) < 1.0 (33)

75-99 (12) 50-74 ( 1 4) 25-49 (16) < 25 (14)

> 1.5 (10)

mb)

No. of Stage patients

I I1 111

IV Total

1

6 46 6 59

CCR SLL 1 5 34 5 45

1 4 28 3 36

Refused Neg Mic Mac SLL SLL SLL SLL

0

1

1 6 2 9

3 18

0 0 2

1

1

22

3

0 1 9 1 11

~

LTS (%)

__

l(38) 4(14) 20(68) 4(13) 29(49)

CCR: complete clinical response; SLL: second-look laparotomy-laparoscopy; Neg: no evidence of disease; Mic: microscooic disease: Mac: macroscopic disease; LTS: long-term survivors. * Survival 1 40 months.

HIGH-DOSE CISPLATIN/DOXORUBICIN IN OVARIAN CA

No. 9

1893

Hurztev el a/.

--2cm >2cm p c 0.006

---_

Median Suwival:51.3 Months 0

0

, 0

J

10

20

'

J

,

40

30

J

,

50

l

'

I

60

'

70

I

'

I

'

90

80

,

I

0

100

10

,

20

,

~

30

!

40

~

50

3

60

-

70

\

80

~

5

90

~

100

Months

Months

FIG.2. Patient survival by initial tumor debulking.

FIG.4. Patient survival by second-look laparotomy results.

all patients is shown in Figure 3. The median survival time for the patients was 5 1.3 months. Survival as a function of second-look procedure results are shown in Figure 4 (P= 0.0012).

months after the start of chemotherapy. Cisplatin/doxorubicin was chosen for the combination because of our previous successful experience with these drugs and because we wished to reserve cyclophosphamide, an agent with significant activity in ovarian cancer, for use after SLL. To preserve dose-intensity, the chemotherapeuticdrugs were repeated at full doses despite the severe myelosuppression produced (Tables 3 and 4). Thirty-two patients (48%)required at least one hospitalization for leukopenia and fever, and seven patients had documented septicemia. One patient died of septicemia after the fourth course of chemotherapy. The degree of myelosuppression seen with this drug combination (cisplatin [ 120 mg/m2] and doxorubicin [70 g/m2]) exceeded the myelosuppression seen with our previous combination (cisplatin [ 100 mg/m2] and doxorubicin [70 g/m2]) despite the minor difference in platinum dose.* The cisplatin was given as a 24-hour continuous infusion, sandwiched in 72 hours of aggressive saline hydration. Nausea and vomiting, although frequent (62%), were well controlled with aggressive sedation and antiemesis during the period of infusion. No patient withdrew from therapy because of this side effect. Similar to our previous experience with this drug schedule,' the nephrotoxicity of high-dose cisplatin given with this degree of hydration was minimal with the exception of one patient who died during her first course of chemotherapy of renal failure (Table 4). In contrast, the greatest morbidity was seen with a sensory neuropathy that developed in 46% of patients. Severe proprioceptive loss leading to difficulties in ambulation occurred in a small percentage of patients (3%). In all patients the sensory neuropathy improved over time from cessation of cisplatin chemotherapy, although recovery was frequently partial and slow and taking from months to years to improve. The most severely affected patients had permanent symptoms.

Discussion

Levin and Hryniuk7 recently emphasized the importance of dose-intensity in the outcome of therapy in patients with ovarian cancer. The average relative dose-intensity (mg/m2/week)correlated significantly with clinical response and median survival time. Our group had previously noticed the ability of a high-dose combination of cisplatin and doxorubicin' to induce rapid clinical remissions in a large fraction of patients with advanced ovarian cancer. Therefore, in this current trial, we examined the ability of cisplatin/doxorubicin given at highdose intensity for four cycles to produce surgically confirmed tumor regression. Tumor response was assessed by an early SLL that was performed approximately 4

.-

r

0.6-

xe

a

p 0.4-

-

.-c

---- NED MICRO

-3

_.-.- c 2 c m MACRO ..,........ > 2 c m MACRO

5 0.2-

0

p c 0.00005

oi

,

0

10

20

'

,

30

'

1

40

~

1

50

'

60

1

,

70

1

80

Months

FIG.3. Survival of the total patient population.

~

1

90

~

I

100

1

'

1

!

'

CANCER November 1 199 1

1894

In terms of tumor response, this treatment schedule was extremely effective in the production of pathologic complete responses (PCR) at the 4 months. For example, in our previous study,' only six of the 50 entered patients ( I 2%) obtained a PCR confirmed at laparotomy. In the current study of a comparable patient population in terms of tumor stage, 22 of the 59 patients (37%) obtained a PCR at SLL. This increase reflects both an increase in PCR in patients undergoing SLL (22 of 36 patients versus 6 of 20 patients) and the percentage of entered patients eligjble for SLL (44 of 59 patients [75%] ver.sus 29 of 50 patients [58%]). This highly toxic chemotherapeutic regimen was successful in increasing the median survival time of all the patients from 28 months in our previous study' to 40 months in this study. However, it is important to notice that approximately two-thirds of the patients in the current study were debulked to less than 2 cm residual disease, and one-third to greater than 2 cm disease. In our previous study, two-thirds of the patients were debulked to greater than 2 cm disease and only one-third to less than 2 cm of disease. As in other clinical trials of a similar nature, the increased survival of our patient population may be partly reflective of the reduced residual disease after initial surgery. Complete pathologic remission was induced in 3790 of patients as opposed to 13% in our previous protocol. These patients showed a significantly increased mean survival time of 55.5 k 3.2 months. This confirms the importance of negative SLL findings as an important predictor of survival"" because patients with macroscopic second-look laparotomies showed a mean survival time of only 24 months. This conclusion is corroborated by the observation that approximately 55% of the long-term survivors (survival greater than 40 months) had NED at SLL. Treatment after SLL did not appear to have a major effect on patient outcome. This is reflected in the prognostic significance of PCR at SLL ( P = 0.001) (Fig. 4). Because almost 60% of the patients treated after SLL had minimal (less than 2 cm) or no residual disease, the results obtained with external beam radiation are disappointing. Dembo' reported excellent tumor control in patients with this minimal intraabdominal tumor bulk treated de nuvo with external beam radiation. Our poor results with wholeabdominal radiation after SLL are consistent with the experimental data that tumor cells resistant to chemotherapy are also resistant to irradiation." No significant differences in survival times were obtained among patients who received either cyclophospharnide or abdominal radiation or no treatment at all. The varied effects of these treatments on patient survival may have been masked by the small number of patients available in each of the subgroups.

'

Vol. 68

Other groups have recently reported their results with short-course chemotherapy. Hainsworth et ~ 1 .reported '~ a response rate of 72% and median survival time of 45 months in patients treated with intensive cisplatin-based chemotherapy of brief duration. Rothenberg et reported preliminary results with high-dose, short-course chemotherapy. Their clinical complete response rates were 78% in patients with Stage 111 disease and 29% in patients with Stage IV disease. The surgically documented complete response rate after chemotherapy was 2490. The toxicity of the intensive short-induction approach was substantial, with peripheral neuropathy being dose limiting. These results are consistent with the current trial and suggest that prolonged survival may be obtained in advanced ovarian cancer with only three to four cycles of high-dose chemotherapy. In conclusion, the survival time of patients with advanced epithelial ovarian cancer can be increased to a median of 5 1.3 months with high-dose cisplatin/doxorubicin chemotherapy with severe, but manageable, myelosuppression. NED at SLL may be considered a reliable predictor of long-term survival. REFERENCES I . Piver SM, Baker TR. Lack of substantial five year disease-free survival by primary aggressive surgery and cisplatin-based chemotherapy or by salvage intraperitoneal cisplatin-based chemotherapy. Eur J Gynaecol Oncol 1990; 1:243-250. 2. Barker GH. Wiltshaw E. Use of high dose cisdichlorodiammine platinumum (11) (NSC-I 19875) following failure on previous chemotherapy for advanced carcinoma of the ovary. Br J Ohsfei G j ~ n a m ~ l 1981; 88:1192-1199. 3. Ozols RF, Behrens BC, Ostchaga Y ef a/. High dose cisplatin and high dose carboplatin in refractory ovarian cancer. Cancer Treut Rev 1985; 12:59-65. 4. Ozols RF, Ostchega Y, Curt G ef a/. High-dose carboplatin in refractory ovarian cancer patients. J Clin Oncol 1987; 5: 197-20 1. 5. Hryniuk WM, Levine MN. Analysis of dose intensity for adjuvant chemotherapy trials. J Clin Oncol 1986; 4: I 162- 1 170. 6. Bruckner HW, Wallach R. High-dose cisplatinum for refractory ovarian cancer. Gynecol Oncol 1984; 12:64-67. 7. Levin L. Hryniuk WW. Dose intensity analysis of chemotherapy regimens in ovarian carcinoma. J Clin Oncol 1987; 51756-767. 8. Griffin TW, Hunter RE, Cederbaurn A1 el a/. Treatment of advanced ovarian cancer with sequential combination Chemotherapy. Cancer 1987; 602150-2155. 9. Ho GA, Beller U, Speyer JL, Colombo N, Wernz J , Beckman EM. A reassessment of the role of second-look laparotomy in advanced ovarian cancer. JClin Oncol 1987; 5:1316-1321. 10. Lipprnan SM, Alberts DS, Slymen DJ ef al. Second-look laparotomy in epithelial ovarian carcinoma: Prognostic factors associated with survival duration. Cancer 1988; 619571-2577. I I . Dembo AJ. Radiation therapy in the management of ovarian cancer. Clin Obstet Gynecol 1983; 10:109-126. 12. Behrens BC, Hamilton TC, Masuda H el a/. Characterization of a cis-diammine-dichloroplatinurn (11)-resistant human ovarian cancer cell line and its use in evaluation of platinum analogues. Cancer Research 1987; 47:414-418. 13. Hainsworth JD, Grosh WW. Burnett LS. Jones 111 WH. Wolff SN, Greco FA. Advanced ovarian cancer: Long-term results of treatment with intensive cisplatin-based chemotherapy of brief duration. Ann Infern Med 1988; 108:165-170. 14. Rothenberg ML, Ozols RF, Glatstein E, Myers CE, Young RC. Dose-intensive induction therapy for advanced epithelial ovarian cancer (OvCa): Cyclophosphamide (C), high dose cisplatin (P) and abdominal radiation (R) (Abstr). Proc Am Soc Clin Oncol 1990; 9:1969.