Introduction

Recent clinical trials showed that long-term prognosis of patients with platinum-sensitive recurrent epithelial ovarian cancer was markedly improved by treatment with platinum-based combination chemotherapy, for example pegylated liposomal doxorubicin/carboplatin [1] or paclitaxel/carboplatin [2, 3]. However, although improved patient survival with single-agent chemotherapy has been reported after several phase III studies [4–6], the results have been unsatisfactory. We performed a phase II study on single-agent gemcitabine treatment of heavily pretreated Japanese patients with recurrent epithelial ovarian cancer and reported a median time to progression of 8.8 months and median survival of 11.2 months [7]. Further, because not only treatment effects but also patients’ quality of life (QOL) should be considered when indicating salvage therapy, oral chemotherapy would be a good alternative choice considering the convenience of patients. Only etoposide has been recommended in the NCCN Practice Guidelines in Oncology (v.2.2010) [8] as an oral agent for treatment of patients with platinum-resistant recurrent ovarian cancer, but etoposide treatment is not covered by health insurance for ovarian cancer in Japan. Therefore, we retrospectively evaluated the treatment efficacy and toxicity of oral cyclophosphamide (CPA) salvage therapy for a future phase II study.

Patients and methods

We primarily planned to perform a single-institute noncomparative prospective phase II study to evaluate the efficacy and toxicity of oral CPA therapy and started the study in June 2007 with the approval of the institutional review board of the Kinki University School of Medicine. Fully informed consent was obtained from all the patients before enrolment in this study. The study was designed with a null hypothesis that true response probability is less than the clinically significant level of 5% for salvage therapy. If this hypothesis is rejected, we will accept the specified alternative hypothesis that the true response probability is at least a target level of 20% with reference to our present phase II study of single-agent gemcitabine [7]. The sample size was calculated as 38 patients, and a one-sided alpha level of 0.05 and 90% power were determined by using the Southwest Oncology Group Statistical One Arm Binomial Tool [9]. However, because enrollment of patients was unsuccessful, we decided to discontinue this study and perform a feasibility study for a future multicenter phase II study.

The treatment effects were determined according to the new guidelines to evaluate the response to treatment in solid tumors (RECIST) [10], and toxicity was determined by Common Terminology Criteria for Adverse Events (CTCAE) v3.0 [11].

The treatment regimen was as follows: 100 mg/day CPA (Endoxan®, Shionogi, Osaka, Japan) was administered 50 mg orally twice a day for 3 weeks with a 1-week break until the patient was diagnosed with progressive disease. Furthermore, prophylactic treatment with 5-hydroxytryptamine-3 (5-HT3) receptor antagonists was not permitted. The treatment was stopped when patients showed treatment resistance over 5 days with grade 4 hematologic toxicity, febrile neutropenia, and grade 3/4 non-hematologic toxicity.

Results

Clinicopathological characteristics of enrolled patients

We enrolled 14 patients in this study; the clinicopathological characteristics of these patients are shown in Table 1. Principal recurrent sites were pelvic cavity (85.7%) and liver (42.9%), and 7 patients (50.0%) had 2 target lesions. All the patients received a median of 3 (range 3–5) regimens of chemotherapy, including paclitaxel and carboplatin combination therapy.

Table 1 Characteristics of patients
Full size table

Treatment results

Median dose of 6300 mg (range 2000–26800 mg) of CPA was orally administered to patients; 1 patient showed partial response and 8 patients showed stable disease. Median time to progression was 3 months (range 1–13 months) and median survival period was 7 months (range 2–28 months). Although 13 patients died because of disease progression, 1 patient survived with the disease (Fig. 1; Table 2).

Fig. 1
figure 1

Survival of subjected patients

Full size image
Table 2 Treatment results
Full size table

Treatment toxicities

Toxicity profiles are shown in Table 3. CTCAE grade 3/4 hematological toxicity was observed in 4 patients (leukopenia in 1, neutropenia in 2, and thrombocytopenia in 1), and only 1 patient was treated with granulocyte-colony stimulating factor (G-CSF). Moreover, although gastrointestinal toxicity, especially anorexia and nausea, was frequently observed as non-hematological toxicity response, no treatment was discontinued because of treatment-related toxicity or patient refusal.

Table 3 Treatment toxicity
Full size table

Discussion

Management of recurrent epithelial cancer patients, especially treatment of platinum-refractory or heavily pretreated patients, remains controversial. Both NCCN Guidelines and Ovarian Cancer Guidelines of Japan Society of Gynecologic Oncology (2007) [12] recommend single-agent salvage chemotherapy or palliative treatment. Sehouli et al. [13] conducted a comparative phase 3 study on single-agent topotecan therapy and nonplatinum topotecan combination (with oral etoposide or gemcitabine) therapy on 502 patients with relapsed epithelial ovarian cancer. The results of this study, which included 174 patients (34.7%) with platinum-resistance (treatment-free interval <12 months), revealed that rather than nonplatinum topotecan combination therapy, monotherapy with topotecan improved the survival of patients with relapsed epithelial ovarian cancer. This study also suggested that single-agent therapy would be more suitable than combination chemotherapy for platinum-refractory or heavily pretreated patients. Several recent clinical trials have suggested the use of pegylated liposomal doxorubicin for treatment of patients with relapsed ovarian cancer [14, 15]. However, because the purpose of salvage chemotherapy is achievement of prolonged survival by realizing tumor dormancy while maintaining patients’ QOL, oral chemotherapy would be a useful treatment strategy, especially for heavily pretreated patients.

Regarding salvage chemotherapy for relapsed epithelial ovarian cancer patients, etoposide is the only agent recommended in the NCCN Guidelines. Previous studies [16–19] have reported that oral administration of etoposide to patients with relapsed ovarian cancer resulted in 16–26.8% overall response rate, 4.3–8.7 months median response duration, and approximately 11 months overall survival. However, Rose et al. [17] reported that grade 3 or 4 hematologic toxicity was common in oral etoposide therapy, with leukopenia occurring in 41.2%, neutropenia in 45.4%, thrombocytopenia 9%, and anemia in 13.4% patients. Moreover, they reported 3 treatment-related deaths (neutropenic sepsis and thrombocytopenic bleeding) and secondary leukemia in 1 patient. Therefore, although oral etoposide therapy would be determined as effective salvage therapy for patients with relapsed epithelial ovarian cancer, the drawback of severe hematologic toxicity is still unsolved.

Cyclophosphamide has long been widely used as an effective alkylating agent for hematological malignancy [20] and solid cancer [21]. CPA has also been known to be effective in combination with adriamycin and cisplatin (CAP) [22] or with cisplatin (CP) [23] for epithelial ovarian cancer. However, few studies were reported on oral CPA treatment for patients with ovarian cancer until the time when phase III studies recommended the use of paclitaxel in place of CPA. Furthermore, no previous studies have been conducted on oral CPA treatment for patients with both platinum-resistant and paclitaxel-resistant relapsed epithelial ovarian cancer. Beck et al. reported retrospective treatment results for 126 advanced ovarian carcinomas. Although the usual dosage of CPA varied between 50 and 150 mg daily, 78 patients were treated with oral CPA (48 patients were treated with 200 mg intravenous CPA daily for 10 days). Beck et al. concluded that there was no apparent influence on efficacy of CPA, regardless of whether the initial treatment was given orally (response rate 48%) or intravenously (response rate 52%). Although the prescribed dosage of oral CPA was 100–200 mg/day, we adopted the dosage of 100 mg/day (50 mg twice a day), because previous studies showed that the minimum effective dosage of CPA was 50 mg/day or 2 mg/kg day [20]. We were concerned that our study would not comply with treatment guidelines, because the pretreated patients were heavily subjected to CPA dosages. Although our planned prospective phase II study could not be completed, because of the slow accrual of patients, in this retrospective evaluation moderate gastrointestinal toxicity, for example anorexia and nausea, was observed for 1 responder only, and we achieved median survival of 7 months. This study also revealed that although oral CPA therapy led to more frequent gastrointestinal toxicity than oral etoposide therapy, hematological toxicity was clearly mild even for the heavily pretreated patients. It is necessary to design a multicenter phase II study to investigate the effects of oral CPA therapy on heavily pretreated patients with relapsed epithelial ovarian cancer, because although this study was feasible with regard to the small number of cases, we believe that oral CPA would be an effective therapy with low clinical expense. Moreover, we believe that routine combined application of an oral 5-HT3 receptor antagonist might improve patients’ survival to enable participation in future phase II trials.