Combination of Ifosfamide, Paclitaxel, and Cisplatin for the Treatment of Metastatic and Recurrent Carcinoma of the Uterine Cervix: A Phase II Study of the Hellenic Cooperative Oncology Group

Gynecologic Oncology 85, 476 – 482 (2002) doi:10.1006/gyno.2002.6649

Combination of Ifosfamide, Paclitaxel, and Cisplatin for the Treatment of Metastatic and Recurrent Carcinoma of the Uterine Cervix: A Phase II Study of the Hellenic Cooperative Oncology Group Meletios A. Dimopoulos,* ,1 Christos A. Papadimitriou,* Kyrillos Sarris,* Gerassimos Aravantinos,† Charalambos Kalofonos,‡ Dimitra Gika,* Georgios M. Gourgoulis,* Eleni Efstathiou,* Dimosthenis Skarlos,§ and Dimitrios Bafaloukos ¶ *Department of Clinical Therapeutics and Radiotherapy, Alexandra Hospital, Athens University School of Medicine, Athens, Greece; †Department of Medical Oncology, Agii Anargiri Hospital, Athens, Greece; ‡Department of Internal Medicine, University of Patra School of Medicine, Rio, Greece; §Department of Medical Oncology, E. Dunant Hospital, Athens, Greece; and ¶Department of Medical Oncology, Metropolitan Hospital, Piraus, Greece Received November 7, 2001; published online April 24, 2002

Objectives. Ifosfamide, paclitaxel, and cisplatin have moderate single-agent activity in patients with metastatic or recurrent cancer of the uterine cervix. We administered a combination of these three agents to a large number of patients with metastatic or recurrent cervical cancer to evaluate its activity. Methods. Sixty patients were treated on an outpatient basis with Ifosfamide (I) 1500 mg/m 2 intravenously over 1 h on Days 1–3, paclitaxel (T) 175 mg/m 2 as a 3-h intravenous infusion on Day 1, and cisplatin (P) 75 mg/m 2 intravenously over 2 h on Day 2 with granulocyte colony-stimulating factor (G-CSF) support. The chemotherapy was repeated every 4 weeks for a maximum of six courses. Results. Fifty-seven patients received at least two courses of treatment and are evaluable for response. Twenty-six patients (46%) achieved an objective response, including 19% complete and 27% partial responses. The median duration of response was 11.5 months and the median time to progression and survival for all patients were 8.3 and 18.6 months, respectively. Patients with excellent performance status, with disease recurrence outside the radiation field, and with nonsquamous tumors had the highest response rate and best survival. Some degree of neurotoxicity occurred in 44% of patients. Grade 3 or 4 toxicity included granulocytopenia in 26% of patients, anemia in 13%, thrombocytopenia in 7%, and neurotoxicity in 3%. Conclusion. The ITP regimen is relatively well tolerated and moderately active in patients with metastatic carcinoma of the uterine cervix. Patients more likely to benefit are those with nonsquamous histology, with excellent performance status, and with disease recurrence outside the radiation field. © 2002 Elsevier Science (USA)

Key Words: cervical cancer; ifosfamide; cisplatin; paclitaxel.

1 To whom correspondence should be addressed at 227 Kifissias Avenue, Kifissia, Athens 14561, Greece. Fax: ⫹(301) 08131383. E-mail: mdimop@ med.uoa.gr.

0090-8258/02 $35.00 © 2002 Elsevier Science (USA) All rights reserved.

476

INTRODUCTION Carcinoma of the uterine cervix is one of the most common malignant neoplasms among women and remains the leading female malignancy in populations of low socioeconomic level [1]. The disease is usually radioresponsive and highly curable in the early stages. Either surgery or radiotherapy alone for stage IB and IIA tumors has resulted in 5-year survival rates of 75 to 90%. Nevertheless, for patients who present with stage IV disease or for those with recurrent disease after radiotherapy, no consistent improvement in survival has been observed during the last 30 years [2]. For patients who have advanced or recurrent disease that is not curable by surgery or irradiation, treatment with cytotoxic agents, alone or in combination, can be considered. Rates of response to the most active single agents vary between 20 and 30%, with a median response duration of 3 to 6 months and a 5- to 9-month survival rate [2– 4]. A number of combination regimens have also been explored during the last two decades, most in uncontrolled trials. Although high response rates have been reported in some of these studies, it is difficult to interpret the relative merits of the combination regimens in these selected patient populations. At this point, no agent or multiagent regimen has been shown to be significantly more efficacious in terms of survival than single-agent cisplatin. However, the combination of cisplatin and ifosfamide is associated with a higher response rate and longer progression-free survival time compared with cisplatin alone [5]. Furthermore, paclitaxel has shown moderate activity in squamous cell carcinoma of the uterine cervix and our preliminary data with the combination of paclitaxel and cisplatin indicate activity in this disease [6, 7]. Thus, in September 1997 we initiated a prospective, multicenter phase II study in which we administered the combination of ifosfamide, paclitaxel (Taxol), and cisplatin (ITP) to patients with primary stage IV or recurrent carcinoma of the uterine cervix.

METASTATIC AND RECURRENT CARCINOMA OF UTERINE CERVIX

PATIENTS AND METHODS Patient Selection Patients were eligible for the study if they had documented primary stage IV or recurrent carcinoma of the uterine cervix, were no longer candidates for curative surgery or radiation therapy, and had not received chemotherapy for advanced disease. Prior neoadjuvant chemotherapy or chemotherapy given as radiation sensitizer was allowed, provided that tumor recurrence occurred at least 6 months after completion of this treatment. Other eligibility requirements included an Eastern Cooperative Oncology Group (ECOG) performance status of 0 to 3, one or more lesions measurable in perpendicular diameters by physical examination or imaging modalities, granulocyte count of 1500/␮L or greater, platelet count of 100,000/␮L or greater, serum creatinine concentration of 1.5 mg/dL or less, serum bilirubin less than 2.0 mg/dL, 3 weeks from prior surgery, and 4 weeks from radiation. Our study was approved by the hospital ethics committees, and informed consent was obtained from patients before study entry. Patients were excluded if they had brain metastases, active infection, serious concurrent medical illnesses, and preexisting clinically significant peripheral neuropathy. Evaluation Before study entry, all patients underwent a complete physical and gynecologic examination, assessment of performance and pain status, complete blood cell (CBC) count and differential, liver and kidney function tests, urinalysis, ECG, pelvic and abdominal computed tomography, and chest X ray with computed tomography of the chest in cases of lung or mediastinal metastases. Imaging of the brain and the bones was performed when clinically indicated. During treatment, CBC, urinalysis, creatinine and liver function tests, performance status, and pain and toxicity evaluations were conducted before each cycle. Response to treatment was assessed after the third and sixth cycles by repeating all abnormal imaging modalities. Treatment Plan All patients received a pretreatment regimen, designed to abrogate allergic reactions, which consisted of dexamethasone 20 mg orally, 12 and 6 h prior to paclitaxel, and diphenhydramine 25 mg and ranitidine 50 mg both given by intravenous injection 30 min before paclitaxel. The chemotherapy was administered on an outpatient basis and consisted of paclitaxel 175 mg/m 2 diluted in 500 mL of 0.9% saline and infused intravenously over 3 h on Day 1. Ifosfamide was given at a dose of 1500 mg/m 2 intravenously over 1 h on Days 1–3, diluted in 1000 mL of 0.9% saline. On Day 1, ifosfamide infusion followed paclitaxel administration. MESNA was given at a dose of 300 mg/m 2 intravenously over 15 min, before and after each dose of ifosfamide. The same dose of MESNA was given orally at home 4 and 8 h after administra-

477

tion of ifosfamide. On Day 2, before ifosfamide administration, patients received 900 mL of 0.9% saline with 100 mL of 20% mannitol over 1 h, immediately followed by cisplatin 75 mg/m 2 diluted in 1000 mL of 0.9% saline and infused over 2 h. Subsequently, the patients received an additional liter of normal saline with potassium and magnesium before ifosfamide administration to minimize cisplatin-induced renal toxicily. Appropriate antiemetics were used before and after the administration of chemotherapy. Courses of ifosfamide, paclitaxel, and cisplatin were administered every 28 days for a maximum of six cycles. Chemotherapy was discontinued in cases of progressive disease or unacceptable toxicity. Granulocyte colony-stimulating factor (G-CSF) was systematically administered subcutaneously from Days 7 to 11 (5 days) on each cycle at a dose of 5 ␮g/kg/day. Definition of Response and Toxicity WHO criteria for response and toxicity were used [8]. All patients who started the treatment were evaluable for toxicity and patients who received at least two courses of treatment were evaluable for response. Assessment of response was based on the evaluation of all lesions present in a patient. A complete response (CR) required disappearance of all clinically or radiographically detectable disease for at least 4 weeks. A partial response (PR) required a greater than 50% reduction in the sum of the products of the two largest perpendicular dimensions of bidimensionally measurable lesions for at least 4 weeks. Stable disease was defined as regression not meeting the aforementioned criteria for objective response, with no progression for at least 3 months. All other cases were considered to have progressive disease. Dose-Delivery Analysis Dose intensity is a measurement of the dose received as a function of time. We used the method described by Hryniuk and Goodyear [9]. A value for received dose intensity was calculated by dividing the cumulative dose treatment given to each patient. One dose interval was added to the treatment period of each patient to adjust for methodologic problems in dealing with those patients who received fewer than six cycles [10]. The received dose intensity was calculated from the beginning of chemotherapy. Statistical Analysis The study was a nonrandomized, phase II study in which an objective response rate (CR ⫹ PR) of at least 40% would be clinically significant; thus, a sample size of 33 patients would provide adequate precision (95% confidence interval (CI), 23– 58%). An early stopping rule was in place according to Simon’s two-stage “minimax” design [11] for type 1 and type 2 error levels of 5 and 20%, respectively. At least five responses in the first 18 patients were required to conclude that the true response rate is more than 20% and to proceed to the next

478

DIMOPOULOS ET AL.

TABLE 1 Patient Characteristics Patients Characteristic

No.

Age, years Median Range ECOG performance status 0 1 2 3 Histology Squamous cell carcinoma Adenocarcinoma Adenosquamous cell carcinoma Clear-cell carcinoma Previous treatment Radiation Surgery plus radiation Surgery Chemotherapy plus radiation None Sites of metastatic tumor involvement Only inside of radiation field Only outside of radiation field Inside and outside or radiation field No. of sites of tumor involvement 1 2 ⱖ3

%

national Federation of Gynecology and Obstetrics (FIGO) stage (III or IV), 75% of patients had received pelvic irradiation, 32% of patients had tumors with histology other than pure squamous cell carcinoma, and 53% of patients had tumors that were poorly differentiated. Response

52 (29–74) 28 16 14 2

47 27 23 3

41 9 8 2

68 15 14 3

32 13 3 2 10

53 22 5 3 17

22 31 7

37 51 12

37 17 6

62 28 10

stage. At the second stage, if at least 11 responses were observed in the 33 patients we would be able to reject the hypothesis that the response rate is less than 40%. Response duration was defined as the time from PR or CR to the appearance of progressive disease. Time to progression was measured from the time of initiation of treatment to the time of last patient contact or documented progressive disease. Survival was measured from the time of initiation of therapy to the last patient contact or death. Time to progression and survival curves were constructed using the Kaplan–Meier product-limit method [12]. Differences in survival were compared with the log-rank statistical test using a microcomputer-assisted program [13]. Statistical analyses of frequency data were performed with the ␹ 2 test.

A total of 296 ITP cycles (median, 6 cycles, range, 1– 6 cycles) were administered. Three patients received only one course of treatment (two because of early disease progression and one because of denial to continue the treatment), leaving 57 patients evaluable for response. Twenty-six patients (46%) achieved an objective clinical response including 11 CRs (19%) and 15 PRs (27%) (Table 2). Based on the analysis of Brader et al. [14], we divided our patients into two groups by site of disease—patients with disease in a previously irradiated site and patients with disease outside a previously irradiated field, whether or not they also had disease within the irradiated field (Table 2). When these two groups were analyzed with respect to response, a higher response rate was observed for patients with disease outside a radiation field (57% vs 21%, P ⫽ 0.02). We observed responses in most metastatic sites outside of a previously irradiated field. Complete responses occurred in patients with assessable primary tumors and with lymph node metastases. A partial response was noted in 3 of 8 patients with bone metastases; however, none of the 4 patients with liver metastases responded to ITP. Ten patients had no prior therapy because they presented with metastatic disease (stage IVA or IVB). Among these patients we observed 3 CRs and 3 PRs. Four of these patients subsequently received radiotherapy. Several other variables were assessed for their possible prognostic value for objective response, such as age, performance status, hemoglobin, histology, tumor grade, mode of primary treatment, and time from primary diagnosis to disease recurrence. We observed responses in 36% of patients with pure squamous cell carcinoma and in 67% of patients with other histologies (0.01). Patients with performance status of 0 on the ECOG scale had a higher response rate than did patients with more impaired performance (67% vs 27%, P ⫽ 0.001). TABLE 2 Response to ITP

Total patients

RESULTS Between September 1997 and December 1999, 60 patients with metastatic or recurrent cervical cancer were accrued to this study. Final data analysis was performed in January 2002, 25 months after accrual of the last patient. The main characteristics of the patients are summarized in Table 1. Forty-eight percent of patients initially presented with an advanced Inter-

Patients with disease in radiation field only

Patients with disease in other site

Response

No.

%

No.

%

No.

%

CR PR SD PD Total

11 15 19 12 57

19 27 33 21 100

2 3 10 6 21

9 14 48 29 100

9 12 9 6 36

25 33 25 17 100

479

METASTATIC AND RECURRENT CARCINOMA OF UTERINE CERVIX

FIG. 1.

Overall survival of patients with squamous (—) versus nonsquamous (- -) tumors.

Thirty-two patients had impaired performance status before treatment; performance status improved in 15 patients, 7 patients maintained their performance status, and performance status deteriorated in 10 patients. Among the 28 patients with an ECOG performance status of zero, 7 showed deterioration of their performance status. Before treatment 30 patients were complaining of moderate or severe pain and were taking opioid analgesics regularly. After treatment pain improved in 13 patients: 8 patients were not receiving any pain treatment and 5 patients were taking NSAIDs only. The pain remained stable in 5 patients and became worse in 12 patients. Before treatment 12 patients had a lower extremity edema; this complication improved in 4 patients. Weight loss was prominent in 11 patients before treatment and improved appetite with weight gain was noted in 3 patients.

vs 10.5 months, P ⫽ 0.01). Patients without impaired performance status survived longer than the other patients (21 months vs 11 months, P ⫽ 0.04). Toxicity Detailed toxicity data according to the WHO Scale for all 60 patients are summarized in Table 3. Toxic nonhematologic reactions consisted primarily of grade 2 or 3 nausea and emesis in 50% of patients and alopecia in 95% of patients. Grade 1 and 2 renal toxicity occurred in 2 patients but was completely reversible without discontinuing the treatment. Forty-five percent of patients had some degree of peripheral neurotoxicity, but 2 patients developed debilitating grade 3 peripheral neu-

TABLE 3 Toxicity According to WHO Criteria

Progression and Survival The median time to progression for all patients was 8.3 months (range, 0.85 to 40.2⫹ months). The median response duration for the 27 responding patients was 11.5 months (range, 3.7 to 34.7⫹ months). The median overall survival for all patients was 18.6 months (range, 0.85 to 53⫹ months). Several parameters were analyzed for their possible association with survival including age, performance status, tumor histology, tumor differentiation, prior treatment with radiation, and time from primary diagnosis to disease recurrence. The median survival of patients with pure squamous cancers was inferior when compared with that of patients with other histologies (11.5 months vs 24.7 months, P ⫽ 0.006) (Fig. 1). Patients relapsing outside a previously irradiated site survived longer than patients relapsing inside a radiotherapy field (20 months

% of patients Toxicity

0

1

2

3

4

Granulocytopenia Anemia Thrombocytopenia Alopecia Nausea and emesis Renal Hepatic Stomatitis Diarrhea Neurotoxicity Allergy

45 17 83 3 26 96 95 85 89 54 96

16 26 7 2 24 2 5 12 3 31 2

13 42 3 7 36 2 0 3 3 10 0

10 10 2 88 14 0 0 0 5 3 2

16 3 5 0 0 0 0 0 0 0 0

480

DIMOPOULOS ET AL.

TABLE 4 Relative Dose Intensity of ITP Relative dose intensity Drug delivery

⬎90%

81–90%

71–80%

⬍70%

Ifosfamide Paclitaxel Cisplatin

73 82 80

15 8 10

4 2 2

8 8 8

ropathy. Generalized fatigue occurred in 12 patients (20%), because of which patients were temporarily unable to perform normal activities and were confined to bed for 50% or more of the day. Two patients developed grade 3 diarrhea which required the administration of intravenous fluids. Grade 3 or 4 granulocytopenia occurred in 26% of patients: 10 episodes of neutropenic fever were documented, and 1 patient died of neutropenic sepsis. Grade 3 or 4 anemia and thrombocytopenia occurred in 13 and 7% of patients, respectively. Another patient died of pulmonary embolism after the third course of ITP. No patient developed symptoms or signs that could be related to ifosmamide-induced encephalopathy. Dose Intensity The relative dose intensities are listed in Table 4. For ifosfamide, 88% of patients had a dose intensity of more than 80% of the intended dose. For both pactitaxel and cisplatin this figure was 90%. DISCUSSION The prognosis of patients who have recurrent or metastatic cervical cancer, which is not amenable to curative surgery or radiation therapy, is dismal. Systemic chemotherapy is frequently administered to such patients in an attempt to improve symptoms and possibly to prolong survival. Among several chemotherapeutic agents that have been used, cisplatin and ifosfamide have reproducible activity in at least 20% of patients. Several combination chemotherapy regimens that contain cisplatin and ifosfamide have been tested in phase II studies and objective responses have been documented in 30 to 70% of patients [15–17]. A large, prospective, randomized trial reported by the Gynecologic Oncology Group (GOG) indicated that the combination of cisplatin and ifosfamide had a higher response rate and a longer progression-free survival compared with cisplatin alone. However, the combination chemotherapy regimen was more toxic and there was no difference in overall survival [5]. A recently, published EORTC study indicated that BEMP combination chemotherapy is more active than singleagent cisplatin at the expense of a higher incidence of toxicity. However, the combination chemotherapy is not associated with longer progression-free survival or overall survival when com-

pared with single-agent cisplatin [18]. During recent years single-agent paclitaxel has been shown to be active in 17 to 25% of patients with squamous cell cervical cancer and the combination of paclitaxel and cisplatin has shown activity in 45 to 46.3% of patients [6, 7, 19, 20]. Based on these observations we designed a prospective multicenter phase II study in which ifosfamide was added to our previously reported combination of paclitaxel and cisplatin [7]. We included patients with metastatic or recurrent cervical cancer, regardless of histology. We observed objective responses in 46% of patients, including complete responses in 19% of patients. The response rate in patients relapsing within radiotherapy fields was 21%, whereas the response rate in patients with disease in other sites was 57%. Our regimen included the same agents that Zanetta et al. have previously reported. In their study ifosfamide was administered at a total dose of 5 g/m 2 IV over 24 h and chemotherapy courses were repeated every 3 weeks. They reported objective responses in 67% of patients and the response rate was 52% in irradiated and 75% in nonirradiated areas [21]. Although comparisons among different phase II studies may be hazardous, we observed that our results appeared inferior to those reported previously by Zanetta et al. These differences are most likely due to different patients characteristics, to the larger number of patients accrued in our study, and to the fact that our study was multicenter. However, we cannot rule out the possibility that the administration of our chemotherapy regimen every 4 weeks instead of every 3 weeks might have interfered with the response rate [21]. We observed that our regimen was more active in patients with excellent performance status, with disease recurrence outside a radiotherapy field, and with nonsquamous histologies. Six of ten patients presenting with de novo metastatic disease responded to ITP. Several previous studies have indicated that performance status and site of disease are important prognostic factors for response [14, 22, 23]. We also noted that the same parameters were associated with longer survival in our patient population, indicating again that among patients with metastatic or recurrent cervical cancer those more likely to benefit from systemic, platinum-based chemotherapy are those with excellent performance status and with disease outside radiotherapy fields. Thus, our findings confirm the reports of previous studies that used cisplatin-based chemotherapy for the treatment of advanced cervical cancer. These studies have indicated that performance status is the most important predictive factor for survival [5, 18]. Our study indicated that the ITP regimen was particularly active in patients with nonsquamous cell cancer. Few trials have focused on these tumors. Single-agent cisplatin and ifosfamide have been associated with objective responses in 20 and 15% of patients, respectively [24, 25]. A recent, large prospective study examined the activity of single-agent paclitaxel in these patients. An overall response rate of 31% was noted with a median response duration of 4.8 months. This study indicated

METASTATIC AND RECURRENT CARCINOMA OF UTERINE CERVIX

that paclitaxel is probably the most active agent for the treatment of nonsquamous cancer of the cervix [26]. Our ITP regimen induced objective responses in 68% of our patients with nonsquamous cell tumors. These data appear promising and further evaluation of ITP in this subset of patients is warranted to confirm its activity. Most studies of patients with recurrent or metastatic cervical cancer report a median time to progression of 4 to 6 months and median survival of 8 to 10 months. In Zanetta and colleagues’ TIP trial the median follow-up was 8 months for surviving patients and thus no conclusions regarding time to progression and survival could be made [21]. In our study these parameters appeared longer than expected. We believe that these differences are more likely due to different patients characteristics and that the appropriate way to address this question is a prospective randomized trial. We were able to administer the chemotherapy on an outpatient basis, and the treatment was relatively well tolerated. There were no incidents of significant renal impairment. The administration of ifosfamide over 3 days probably resulted in the absence of encephalopathy. We decided to administer G-CSF routinely to reduce the degree of myelotoxicity and the incidence of neutropenic infections. Despite the routine administration of G-CSF for a median of 5 days, the incidence of significant neutropenia was 26%, including 10 episodes of neutropenic fever and 1 neutropenic death. In Zanetta and co-workers’ TIP trial G-CSF was not administered routinely; grade 3 or 4 neutropenia occurred in 91% of patients, treatment delays were necessary in 22% of patients, and G-CSF was used in 20% of patients [21]. Thus, we have no proof that G-CSF was necessary in all our patients and its use could have been restricted to patients who had previously experienced neutropenic fever or to those with protracted neutropenia necessitating a delay in the administration of chemotherapy. Forty-five percent of our patients developed some degree of neurotoxicity, including grade 3 neurotoxicity in 3%. The regimen of 3-h paclitaxel and cisplatin has been used in prior studies and is associated with a higher incidence of neurotoxicity compared with the 24-h dosing schedule of paclitaxel combined with cisplatin [27]. It seems that the incidence and severity of neurotoxicity are the result of the high peak levels of paclitaxel achieved in the systemic compartment during and after the 3-h infusion at the same time that cisplatin, a known neurotoxic agent, is administered. To reduce the severity of this side effect we administered paclitaxel 24 h before cisplatin [27]. Our data show that the severity of neurotoxicity may be less than expected when these two neurotoxic agents are not administered concomitantly. In summary, our combination of ifosfamide, paclitaxel, and cisplatin, administered to outpatients, was relatively well tolerated. Objective responses were noted in almost one-half of the treated patients. However, our regimen did not appear to be more active than the previously reported combination of paclitaxel with cisplatin. Our data confirm previous observations

481

that patients with excellent performance status and with disease recurrence outside the radiation field are more likely to benefit from systemic chemotherapy. The ITP regimen appeared more active in patients with nonsquamous tumors. REFERENCES 1. Thigpen T, Vance R, Khansur T. Carcinoma of the uterine cervix: current status and future directions. Semin Oncol 1994;21:43–56. 2. Thigpen T, Vance RB, Khansur T. The platinum compounds and paclitaxel in the management of carcinomas of the endometrium and uterine cervix. Semin Oncol 1995;22(suppl. 12):67–75. 3. Park RC, Thigpen T. Chemotherapy in advanced and recurrent cervical cancer. Cancer 1993;71:1446 –50. 4. Vermorken JB. The role of chemotherapy in squamous cell carcinoma of the uterine cervix: a review. Int J Gynecol Cancer 1993;3:129 – 42. 5. Omura GA, Blessing JA, Vaccarello L, Berman ML, Clarke-Pearson DL, Mutch DG, Anderson B. Randomized trial of cisplatin versus cisplatin plus mitolactol versus cisplatin plus ifosfamide in advanced squamous carcinoma of the cervix: a Gynecologic Oncology Group study. J Clin Oncol 1997;15:165–71. 6. McGuire WP, Blessing JA, Moore D, Lentz SS, Photopulos G. Paclitaxel has moderate activity in squamous cervix cancer: a Gynecologic Oncology Group study. J Clin Oncol 1996;14:792–5. 7. Papadimitriou CA, Sarris K, Moulopoulos LA, Fountzilas G, Anagnostopoulos A, Voulgaris Z, et al. Phase II trial of paclitaxel and cisplatin in metastatic and recurrent carcinoma of the uterine cervix. J Clin Oncol 1999;17:761– 6. 8. Miller AB. Reporting results of cancer treatment. Cancer 1981;47:207–14. 9. Hryniuk KWM, Goodyear M. The calculation of received dose intensity. J Clin Oncol 1990;8:1935–7. 10. Coppin CM. The description of chemotherapy delivery options and pitfalls. Semin Oncol 1987;14(suppl. 14):32– 42. 11. Simon R. Optimal two-stage designs for phase II clinical trials. Controlled Clin Trials 1989;10:1–10. 12. Kaplan E, Meier F. Nonparametric estimation from incomplete observations. J Am Stat Assoc 1958;58:457– 81. 13. Campos-Filho N, Franco ELF. Micro-computer-assisted univariate survival data analysis using Kaplan–Meier life table estimators. Comput Methods Programs Biomed 1988;27:223– 8. 14. Brader KR, Morris M, Levenback C, Levy L, Lucas KR, Gershenson DM. Chemotherapy for cervical carcinoma: factors determining response and implications for clinical trial design. J Clin Oncol 1998;16:1879 – 84. 15. Buxton EJ, Meanwell CA, Hilton C, Mould J, Spooner D, Lateif T, et al. Combination bleomycin, ifosfamide, and cisplatin chemotherapy in cervical cancer. J Natl Cancer Inst 1989;81:359 – 61. 16. Ramm K, Vergote IB, Kaern J, Troppe CG. Bleomycin, ifosfamide, cisplatinum in pelvic recurrence of previously irradiated cervical carcinoma: a second look. Gynecol Oncol 1992;46:203–7. 17. Cervellino JC, Araujo CE, Sanchez O, Nishihama A. Cisplatin and Ifosfamide in patients with advanced squamous cell carcinoma of the uterine cervix. Acta Oncol 1995;34:257–9. 18. Vermorken JB, Zanetta G, DeOliveira CF, Van der Burg MEL, Lacave AJ, Teodorovic I, et al. Randomized phase III trial of bleomycin, vindesine, mitomycin-C and cisplatin (BEMP) versus cisplatin in disseminated squamous-cell carcinoma of the uterine cervix. Ann Oncol 2001;12:967–74. 19. Rose PG, Blessing JA, Gershenson DM, McGehee R. Paclitaxel and cisplatin as first-line therapy in recurrent or advanced squamous cell carcinoma of the cervix: a Gynecologic Oncology Group study. J Clin Oncol 1999;17:1676 – 80.

482

DIMOPOULOS ET AL.

20. Piver MS, Ghamande SA, Eltabbakh GH, O’Neill-Coppola C. First-line chemotherapy with paclitaxel and platinum for advanced and recurrent cancer of the cervix. Gynecol Oncol 1999;75:334 –7. 21. Zanetta G, Fei F, Parma G, Balestrino M, Lissoni A, Gabriele A, Mangioni G. Paclitaxel, ifosfamide and cisplatin chemotherapy for recurrent or persistent squamous-cell cervical cancer. Ann Oncol 1999;10:1171– 4. 22. Potter ME, Hatch KD, Potter MY, Shingleton HM, Baker VV. Factors affecting the response of recurrent squamous cell carcinoma of the cervix to cisplatin. Cancer 1989;63:1283– 6. 23. Zanetta G, Torri W, Bocciolone L, Lucchini V, Mangioni C. Factors predicting response to chemotherapy and survival in patients with metastatic or recurrent squamous cell cervical carcinoma: a multivariate analysis. Gynecol Oncol 1995;58:58 – 63. 24. Thigpen JT, Blessing JA, Fowler WC, Hatch K. Phase II trials of cisplatin

and piperazinedione as single agents in the treatment of advanced or recurrent non-squamous cell carcinoma of the cervix: a Gynecologic Oncology Group study. Cancer Treat Rep 1986;70:1097–1100. 25. Sutton GP, Blessing JA, Disaia PJ, McGuire WP. Phase II study of ifosfamide and mesna in onsquamous carcinoma of the cervix: a Gynecologic Oncology Group study. Gynecol Oncol 1993;47:48 –50. 26. Curtin JP, Blessing JA, Webster KD, Rose PG, Mayer AR, Fowler WC, et al. Paclitaxel, an active agent in nonsquamous carcinomas of the uterine cervix: a Gynecologic Oncology Group study. J Clin Oncol 2001;19: 1275– 8. 27. Connely E, Markman M, Kennedy A, Webster K, Kyle B, Peterson G, Belinson J. Paclitaxel delivered as a 3-hour infusion with cisplatin in patients with gynecologic cancers: unexpected incidence of neurotoxicity. Genecol Oncol 1996;62:166 – 8.