Bleomycin has been used for ESCC since the 1970s and 1980s. Bleomycin as a single agent for ESCC has been reported to have a response rate of 15–20 % [2–4]. A randomized trial comparing chemotherapy with bleomycin and best supportive care did not show a survival benefit [25]. Bleomycin is no longer used because of its pulmonary toxicity in combination with other drugs or radiotherapy.

5-FU, in combination with other drugs and/or radiation therapy, is the most commonly used chemotherapeutic drug for ESCC. When 5-FU was used as a single agent, a 15 % response rate was observed in previously treated patients administered intermittent bolus of 5-FU in an Eastern Cooperative Oncology Group trial [5]. S-1, an oral fluoropyrimidine preparation combining tegafur, gimeracil, and oteracil potassium in a molar ratio of 1:0.4:1, has been used for gastric, head and neck, lung, colon, and other cancers. The response rate of S-1 for pretreated patients with ESCC was reported to be 25 % [6].

Platinum agents have been used mostly in combination with 5-FU, topoisomerase inhibitors, and taxanes. As monotherapy, CDDP is administered at doses of 50–120 mg/m² every 3–4 weeks; the cumulative response rate for ESCC was 21 % [7, 26, 27]. In a randomized phase II trial, addition of 5-FU to CDDP was compared to CDDP monotherapy administered at 100 mg/m² every 3 weeks in 92 patients with ESCC. Although the response rate was higher in the combination group (35 vs. 19 %), survival was similar in both groups (monotherapy vs. combination: 33 vs. 28 weeks) [7]. Carboplatin, a second-generation platinum analogue, was developed to maintain the antitumor activity of CDDP and to reduce toxicity. Carboplatin has also been used mostly in combination; its single agent activity is limited, with response rates of 0–14 % [8, 9]. Oxaliplatin, a platinum derivative with less emetogenic, nephrotoxic, and ototoxic effects compared to CDDP, has been evaluated mainly in combination regimens for esophageal cancer. Nedaplatin, a second-generation platinum derivative, is ten times as water soluble as CDDP with less gastrointestinal and renal toxicity. In a phase II study of nedaplatin monotherapy at 100 mg/m² via intravenous infusion every 4 weeks, five partial responses (55.6 %) were observed in nine patients with ESCC who had received prior chemotherapy, including two partial responses in four patients previously treated with CDDP [10].

Taxanes have shown activity against not only adenocarcinoma but also squamous cell carcinoma of the esophagus. Paclitaxel promotes the stabilization of microtubules, and it is a cell cycle-specific agent affecting cells in the G2/M phase [28]. Paclitaxel as monotherapy has been evaluated as first-line chemotherapy at a dose of 250 mg/m² administered via 24-h intravenous infusion every 3 weeks. Of 52 patients who enrolled to this study, 18 patients had ESCC. Five (25 %) partial responses and five (28 %) minor responses were observed among the patients with squamous cell carcinoma. On the other hand, complete and partial responses were observed in 34 % of the adenocarcinoma group [11]. The efficacy of a weekly schedule of paclitaxel was also evaluated in 86 patients with esophageal cancer, including 32 cases of squamous cell carcinoma, at a dose of 80 mg/m² weekly over a 1-h infusion. Of 15 squamous cell cancer patients who did not receive prior chemotherapy, 2 (13 %) achieved a partial response, and there were no responses among patients who received prior chemotherapy [12]. A weekly schedule at a dose of 100 mg/m² administered via 1-h intravenous infusion on days 1, 8, 15, 22, 29, and 36 every 7 weeks was also evaluated for previously treated patients with esophageal cancer in a phase II trial. The overall response rate of patients with squamous cell cancer was 43.1 %, with four patients (7.8 %) achieving a complete response. Although grade 3 or 4 neutropenia (52.8 %), leukopenia (45.3 %), anorexia (9.4 %), and fatigue (9.4 %) were observed, weekly paclitaxel was highly active and well tolerated [13]. Docetaxel at doses of 75–100 mg/m² administered every 3 weeks has also shown activity against adenocarcinoma of the esophagus, with a response rate of approximately 20 % in previously untreated patients [14, 15]. For squamous cell carcinoma, a phase II trial of docetaxel at a dose of 70 mg/m² administered every 3 weeks was conducted. The majority of patients (94 %) had squamous cell carcinoma in this trial. The response rate was reported to be 16 % for pretreated patients and 36 % for untreated patients [16]. However, careful management of infection is needed because grade 3/4 neutropenia (88 %) and febrile neutropenia (18 %) were observed in this trial.

The vinca alkaloid vindesine was evaluated in several phase II trials; it demonstrated reproducible antitumor activity, with a response rate of 20 % in cases of squamous cell carcinoma [17–19]. Vinorelbine, which has less neurotoxicity compared with vincristine and vindesine, was evaluated in patients with ESCC in a phase II trial by the European Organization for Research and Treatment of Cancer. Vinorelbine was administered weekly as a 25 mg/m² short intravenous infusion. Response rates of 20 and 6 % were observed in untreated patients and pretreated patients, respectively [20].

There have been reports on the use of topoisomerase inhibitors for the treatment of ESCC. Etoposide, an inhibitor of type II topoisomerase, demonstrated a response rate of 19 % in one trial [21]. In contrast, other trials showed response rates of less than 5 % [29, 30]. Irinotecan, a type I topoisomerase inhibitor, has shown modest activity in ten previously treated patients with ESCC, with a 10 % response rate [22].

Other drugs have been tested for ESCC as single agents, and they have demonstrated antitumor activity, with response rates of 0–42 %; these include methotrexate [5], ifosfamide [23], gemcitabine [24], mitomycin-C [26], and doxorubicin [5].

Because of the limited activity of single-agent chemotherapy, most of the drugs described above have also been tested in combination regimens. Two randomized trials compared best supportive care to combination chemotherapy. 5-FU and CDDP combination therapy failed to show an advantage in overall survival (both 12 months) compared to best supportive care in patients with ESCC, with or without prior surgery [31]. Another randomized trial in 24 patients with esophageal cancer, including 19 patients with squamous cell cancer, also failed to show a meaningful survival benefit with cyclophosphamide and doxorubicin combination therapy. Although these randomized trials did not show a survival benefit with combination therapy, combination regimens with 5-FU, CDDP, and newer agents have been used for the treatment of ESCC (Table 12.2).