Combined-Modality Treatment of Esophageal Cancer

Introduction

Esophageal carcinoma is relatively rare in the United States, accounting for about 1% of cancers diagnosed each year. However, there are markedly different patterns of prevalence in other parts of the world.[1] High-incidence areas include India, southern Africa, northern Iran, and northern China.

Until recently, squamous-cell carcinoma comprised the vast majority of cases of esophageal cancer. However, the incidence of adenocarcinoma in white males has been rapidly increasing by approximately 10% per year for the past 10 years. This percentage increase in frequency exceeds that of all other cancers for that time period.[2,3] The prognosis for patients with either type of tumor histology appears to be poor.

The two predominant risk factors for the development of esophageal squamous-cell carcinoma are alcohol(Drug information on alcohol) and tobacco abuse.[4] Adenocarcinoma has been linked to Barrett's metaplasia of the esophageal mucosa as a consequence of chronic gastroesophageal reflux disease. The prevalence of adenocarcinoma in association with Barrett's metaplasia has been found to range from 3% to 16% in various reported series.[5-7]

Esophageal cancer patients who have clinically metastatic disease at the time of diagnosis have a median survival of 6 months. Even patients with potentially localized, resectable disease who are treated with surgery have a median survival of approximately 12 months and a 5-year survival of 11% to 14%.[1]

The poor prognosis of esophageal cancer relates to its tendency for early spread into the submucosal lymphatic network, as well as to the absence of the serosal layer of the esophagus, which potentially limits local disease extension. Microscopic and/or gross disease is often present in lymph node or distant metastatic sites by the time the primary tumor becomes large enough to cause symptoms of dysphagia. Because of the frequency of disease dissemination at presentation, recent investigations have used systemic therapy combined with locoregional therapies to gain better local control and manage the systemic component of esophageal malignancy.

Preoperative Chemotherapy

The potential benefits from adding preoperative chemotherapy include downstaging the disease to facilitate surgical resection, preventing local recurrence, and eradicating micrometastatic disease. At the time of surgery, response to chemotherapy can be assessed directly. The disadvantages of preoperative chemotherapy include the potential emergence of drug-resistant clones and the delay in definitive treatment with the risk of disease progression.

Randomized Trials

Four randomized trials have compared preoperative chemotherapy to surgery alone.

1. Roth et al[8]—reported a small study from The M.D. Anderson Cancer Center in which 39 patients were randomized to treatment with either surgery alone or preoperative chemotherapy with cisplatin(Drug information on cisplatin) (Platinol), vindesine, and bleomycin(Drug information on bleomycin) (Blenoxane), followed by surgery and postoperative chemotherapy with cisplatin and vindesine(Drug information on vindesine) for 6 months. Despite a good chemotherapy response of 47%, the two arms had identical median survival rates of 9 months.
2. Schlag—In a second randomized study, reported by Schlag,[9] 46 patients were randomized to either surgery or three cycles of preoperative chemotherapy with cisplatin and fluorouracil(Drug information on fluorouracil) (5-FU) followed by surgery. At a median follow-up of 20 months, an interim analysis was done. Resectability rates were similar: 79% for the chemotherapy/surgery group and 77% for the surgery group. Median survival was 10 months for each group. The study was stopped, and the authors concluded that preoperative chemotherapy did not increase surgical resectability or survival for patients with esophageal carcinoma.
3. Norwegian Study—A study by Nygaard et al[10] from Norway randomized 187 patients to one of four treatments: surgery alone, preoperative chemotherapy and surgery, preoperative radiation and surgery, or preoperative chemotherapy followed by radiation and then surgery. There was no significant difference in survival rates between the preoperative chemotherapy group and the surgery alone group (3-year survival rates of 3% and 9%, respectively).
4. Intergroup Trial—The United States Intergroup completed a two-arm randomized trial, which compared three cycles of preoperative cisplatin/5-FU followed by surgery and postoperative cisplatin/5-FU vs treatment with surgery alone. Over 400 patients were enrolled. Until the mature results of this trial are known, use of preoperative chemotherapy must be considered investigational, since all completely reported randomized trials have been negative.

Preoperative Chemoradiation

Patterns of relapse for esophageal carcinoma include both locoregional recurrence and distant metastases. Therefore, the combination of radiation and chemotherapy before surgery has been used in an attempt to improve both local and distant control.

Nonrandomized Studies

Several studies have used preoperative cisplatin and 5-FU plus radiation:

RTOG/SWOG Studies—Two studies independently conducted by the Radiation Therapy Oncology Group (RTOG)[11] and the Southwest Oncology Group (SWOG)[12] tested a nearly identical treatment regimen of cisplatin and 5-FU combined with 30 Gy of radiation. Operability was 66% and 63%, respectively. Potentially curative resection was not reported in the RTOG study and was only considered to be 49% in the SWOG study.

Median survival in the RTOG study was 13 months, and 2- and 3-year survival rates were 15%, and 8%, respectively. In the SWOG study, median survival was 14 months, and 2- and 3-yer survival rates were 28%, and 16%. In both studies, patients who achieved a histologic complete response (CR) to the preoperative treatment regimen (27% in the RTOG study and 25% in the SWOG study) had improved overall survival after surgical resection.

St. Louis University Study—Naunheim et al[13] reported a study from the St. Louis University Medical Center in which 47 patients (29 squamous-cell carcinoma and 18 adenocarcinoma) were treated with simultaneous cisplatin, 5-FU, and 3,000 to 3,600 cGy of radiation preoperatively. Thirty-nine of these patients went to surgical exploration, and 83% (72% of the total group) were resectable. Overall median survival was 23 months, and 3-year actuarial survival was 40%. These survival results were significantly better than those of a group of historical controls from the same institution treated with either surgical resection or radiation, who had a median survival of 10.5 months and 0% with a 3-year survival.

University of Michigan Study—One of the most promising results with combined preoperative chemoradiotherapy came from a trial conducted at the University of Michigan Medical Center. Forastiere et al[14] treated 43 patients with an intensive 21-day in-patient regimen of chemoradiation followed by surgical resection. Fluorouracil (5-FU) was administered as a continuous infusion for 21 days, and cisplatin and vinblastine(Drug information on vinblastine) (Velban) were given on days 1 to 5 and 17 to 21. Radiation was delivered concomitantly to a total dose of 3,750 to 4,500 cGy, followed by surgical resection 3 to 4 weeks later. Of the 43 patients, 22 had squamous-cell carcinoma and 21 had adenocarcinoma. The vast majority (95%) of patients completed the preoperative treatment even though toxicity was substantial, and 84% had a potentially curative resection. In 24% of the patients, no residual tumor was found in the resected esophagus.

The final report of this study was recently published, with a median follow-up of 78.7 months.[15] The median survival for all patients was 29 months, which was a major improvement from the 12-month median survival of a group of historical control patients treated with surgery alone at the same institution. The rate of 5-year survival was 34%.

Patients who achieved a histologic CR from the preoperative regimen had a median survival of 70 months, and 60% were alive at 5 years. However, patients with residual cancer in the esophagus at time of resection also achieved an improved median survival of 26 months, and 32% were alive at 5 years. This substantial increase in long-term survival among the patients with residual tumor in the resected esophagus suggests a continued benefit from esophagectomy. These pilot data served as the rationale for a randomized study recently completed at the University of Michigan Medical Center. The preliminary results of this randomized study will be discussed further in the next section on randomized trials.

Fox Chase Study—Sauter and colleagues[16] at Fox Chase Cancer Center used a preoperative regimen of 5-FU/mitomycin (Mutamycin)/radiotherapy in adenocarcinoma patients only.[16] They treated 30 patients with stage I or II disease with two cycles of continuous- infusion 5-FU, one cycle of mitomycin(Drug information on mitomycin), and concurrent radiation to a total dose of 6,000 cGy. Of the 30 patients, 10 were unable to tolerate the entire regimen due to severe esophagitis. Surgical exploration was done in 21 patients; 18 patients underwent a curative resection, and, of these patients, 39% had a pathologic complete response.

There was a trend for improved survival in patients who had no residual tumor in the resected specimen. Rates of survival at 2 and 3 years for all patients were 33% and 25%, respectively. Median survival was better for those patients who underwent esophagectomy than for those who did not (23 vs 9 months), although the study was not designed to test this question in a randomized fashion.

Role of Preresection Esophagogastroduodenoscopy—A study by Bates et al[17] at the University of North Carolina attempted to determine the role of preresection esophagogastroduodenoscopy in predicting which patients treated with chemoradiation could potentially forgo surgery. Thirty-five patients were treated with 5-FU, cisplatin, and radiation to a total dose of 4,500 cGy. The pathologic complete response was 51%.

Of the 35 patients, 22 (63%) underwent a preresection upper endoscopy with brushings and biopsies 1 to 3 days before the planned esophagectomy. Although 17 (77%) of the 22 patients had negative pathology from this evaluation, 7 (41%) of these 17 patients had residual tumor in the resected specimen. Therefore, the authors concluded that preresection endoscopy was not a reliable predictor of pathologic complete response. Esophagectomy appeared to contribute to the survival of this group, because 25% of patients with residual tumor in the resected esophagus were long-term survivors.

These nonrandomized trials show that patients who achieve a histologic complete response generally have better survival than those who have persistent disease in the esophagus after chemoradiation, and a complete response rate of at least 25% is a typical result from these regimens. Survival benefit from chemoradiation can only be demonstrated through randomized trials.