What is the most efficient current multimodality therapy to be applied?

T.N. Walsh (Dublin)

At least 50% of esophageal cancers have overt metastases at presentation and the majority of the rest will die despite apparent curative surgery, suggesting occult systemic metastases [1]. The need for adjuvant systemic therapy is not disputed, just the most effective protocol. There are few randomized trials of multimodal therapy for adenocarcinoma and none comparing regimens. Information from non-randomized studies and squamous carcinoma must be interpreted with caution as the two tumors differ in aetiology, distributions, stage at presentation and survival for each stage. The response rate and outcome following response also differs [2] but there is no evidence to suggest that a regimen effective for one will not benefit the other.

Realistically the best regimen can only improve outcome by single or low double percentage figured as with breast and colon cancer. Large trials are needed to prove such differences significant. Most trials on esophageal cancer are underpowered. Hence, in one an increase in survival from 15% to 32% is not statistically significant [3]. Because of the paucity of adenocarcinoma data this paper examines some randomized trials of squamous carcinoma as well as non-randomized trials of esophageal adenocarcinoma for evidence of efficacy.

What works for squamous cell carcinoma? Evidence from randomized trials

Chemotherapy

The most promising agents studied are 5-fluorouracil (5-FU), cisplatin, mitomycin and paclitaxel. Single agent therapy is more effective than combination therapy.

Preoperative chemotherapy

Early randomized trials with smaller numbers comparing preoperative chemotherapy with surgery alone found no benefit for chemotherapy with considerable toxicity [4, 5]. Recent larger trials have been more optimistic. Law et al. [6] reported on 147 patients with esophageal cancer randomized to preoperative cisplatin and 5-FU or surgery alone. Median survival figures (16.8 versus 13 months) were not statistically different. The Medical Research Council [7] randomized 802 patients with operable squamous and adenocarcinoma to two cycles of cisplatin and 5-FU or surgery alone. Median survivals of 17.6 versus 13.6 months were statistically significantly different. The Rotterdam Esophageal Tumour Study Group [8] randomized 148 patients with operable squamous cancer to preoperative cisplatin and etoposide or surgery alone. Early results indicate an enhanced median survival of 18.7 months versus 11 months. Thus, differences in survival become significant in larger trials.

Postoperative chemotherapy

Pouliquen et al. [9] found no benefit for postoperative chemotherapy with low-dose 5-FU and cisplatin. Recently Ando et al. [10] presented preliminary data on a randomized trial of postoperative cisplatin (80 mg/m2) and 5-FU (800 mg/m2 x 5 days) versus surgery alone in 242 patients who underwent transthoracic esophagectomy with lymphadenectomy. Fiveyear disease free survival was 58% versus 46% favouring postoperative chemotherapy. Overall survival was similar but a trend favoured therapy in the node-positive subgroup.

Pre- and postoperative chemotherapy

Roth et al. [11] compared preoperative and postoperative cisplatin, vindesine and bleomycin with surgery alone in 36 patients and found no survival difference between the limbs. Saltz and Kelsen [12] randomized patients to receive pre- and postoperative cisplatin, vinblastin and bleomycin or surgery alone and again found no survival advantage. More recently Kelsen et al. [13] reported on 467 patients with operable squamous or adenocarcinoma of the esophagus randomized to 3 cycles of preoperative cisplatin and 5-FU followed by 2 cycles postoperatively. Median and 3-year survival figures were similar.

Radiotherapy

Preoperative radiotherapy

Seven randomized trials have compared preoperative radiotherapy with surgery for squamous cell carcinoma, of which six have found no survival advantage for neoadjuvant radiotherapy [14-19]. Only one describe a survival advantage for preoperative radiotherapy [5] but patients receiving preoperative chemotherapy and radiotherapy were pooled with those who had radiotherapy only.

Postoperative radiotherapy

Three trials comparing postoperative radiotherapy with surgery alone found no survival advantage for adjuvant radiotherapy, which was associated with significant morbidity. Following curative resection Zieren et al. [20] randomized 68 patients to 55.8 Gy radiotherapy or surgery alone, a French group [21] randomized 221 patients to 45-55 Gy postoperative radiotherapy or surgery and Fok et al. [22] randomized 60 patients to 49 Gy radiotherapy or surgery alone. None found an advantage and one [21] found a significantly disadvantage for postoperative therapy.

Chemoradiotherapy

Chemoradiotherapy versus radiotherapy

Four randomized trials compared radiotherapy and chemotherapy with radiotherapy alone. Araujo et al. [23] reported no difference in outcome in 59 patients with squamous carcinoma randomized to 5-FU, mitomycin C and bleomycin and radiotherapy or radiotherapy alone. Roussel et al. [24] reported no survival advantage in 144 patients with advanced carcinoma randomized to methotrexate and 56 Gy radiotherapy over radiotherapy alone. Sischy et al. [25] treated 135 patients with 5-FU and mitomycin C and 40 Gy radiotherapy or radiotherapy alone. The median survival for the multimodality group was significantly longer at 14 months than 9 months for radiotherapy alone. Herskovic et al. [26] randomized 121 potentially curable patients to 64 Gy radiation alone or 50 Gy concurrent with two courses of cisplatin and 5-FU. Long-term follow-up confirms a 3-year survival of 30% versus 0% favouring multimodal therapy [27]. These trials support concurrent chemotherapy and radiotherapy over radiotherapy alone.

Chemoradiotherapy plus surgery versus surgery

Five randomized trials have compared preoperative chemotherapy and radiotherapy with surgery alone for squamous cancer and one trial pooled adenocarcinoma (75%) and squamous cancer. Half used consecutive, the rest concurrent chemoradiotherapy. Nygaard et al. [5] randomized 186 patients into four subgroups, two of which were to receive sequential chemotherapy and radiotherapy or surgery alone. They found a survival advantage for neoadjuvant radiotherapy over surgery alone but no further advantage for added chemotherapy. LePrise et al. [28] reported no survival advantage in 86 patients randomized to receive sequential preoperative chemotherapy (5-FU and cisplatin) and low dose radiotherapy (20 Gy) over surgery alone. Bosset et al. [29] reported on the EORTC trial of 282 patients randomized to sequential 37 Gy together with single-agent cisplatin chemotherapy. Disease-free survival was 40% versus 28% favouring multimodal therapy but there was no overall survival difference (36% versus 34%) at 3 years. Apinop et al. [30] reported on 69 patients with squamous cell carcinoma randomized to preoperative concurrent CRT with two cycles of cisplatin and 5-FU plus 40 Gy radiotherapy. The difference in survival of 10 months versus 7 months was not significant. Urba et al. [3] randomized 100 patients with squamous or adenocarcinoma (70%) to concurrent cisplatin, vinblastin and 5-FU plus 45 Gy radiotherapy or surgery only. There was a trend in 3 years survival of 15% versus 32% favouring multimodal therapy (p = 0.07) not reaching significance. Walsh et al. [31] randomized 98 patients with squamous cell carcinoma of the esophagus to preoperative concurrent cisplatin and 5-FU and 40 Gy radiotherapy or surgery alone. Survival at 3 years was 32% for the multimodal limb and 16% for surgery alone while the 5-year survival was 36% versus 11% for surgery. In summary, trials support a trend or significant difference especially for concurrent and multi-agent treatment.

What works for adenocarcinoma? Evidence from non-randomized trials

Preoperative chemotherapy

The only trials on preoperative chemotherapy have been single arm and uncontrolled [32, 33] showing some promise but poorly tolerated.

Pre- and postoperative chemotherapy

Ajani et al. [34] reported a median survival of 24 months in 35 patients with resectable adenocarcinoma treated with preoperative and postoperative etoposide, cisplatin and 5-FU. In a further trial they treated 26 potentially resectable patients with preoperative etoposide, doxorubicin, cisplatin and GM-CSF and a further three postoperative doses to responding patients [35]. They concluded that less toxic and more effective regimens were required.

Combined chemotherapy and radiotherapy

Chemotherapy and radiotherapy without surgery

Coia et al. [36], in an uncontrolled study, treated 20 patients with adenocarcinoma with combined 5-FU, mitomycin and radiation therapy. The median relapse-free survival was 10 months and the median survival was 15 months.

Chemotherapy and radiotherapy plus surgery

Hoff et al. [37] treated 39 patients with adenocarcinoma with two cycles of cisplatin, 5-FU, etoposide, leucovorin and 30 Gy radiation therapy followed by resection [40]. A complete response to preoperative therapy was seen in 19%. Actuarial survival at 12, and 24 months was 72 and 51%. Stewart et al. [38] reported an uncontrolled trial of cisplatin, 5-FU and leucovorin, and concomitant 30 Gy radiation in 24 patients with adenocarcinoma. The median survival was more than 26 months. Wolfe et al. [2] used preoperative multi-agent chemotherapy and radiotherapy in 93 patients with adenocarcinoma. The 5-year survival amongst resected patients was 25%. Twenty percent of those resected had a sterilised specimen with a 5-year survival of 60%. Urba et al. [39] reported on 5-FU as single agent therapy with 49 Gy radiotherapy followed by transhiatal esophagectomy in 24 patients. The median survival was only 11 months. Naunheim et al. [40] treated 28 patients preoperatively with two courses of cisplatin and 5-FU and 30 or 36 Gy radiation. Seventeen percent had a complete response. Actuarial survival at 2, and 3 years was 28 and 20% respectively. Algan et al. [41] compared high-dose chemoradiation with high-dose chemoradiation followed by esophagectomy in a study of patients with potentially resectable adenocarcinoma and found no survival advantage for esophagectomy after treatment. Keller et al. [42] studied 46 patients with stage I or II adenocarcinoma of the esophagus or gastroesophageal junction treated with concomitant 60 Gy radiation and 5-FU and mitomycin (10 mg/m2 bolus). Overall median survival was 16.6 months and 2-year survival 27%.

What works for adenocarcinoma? Evidence from randomized trials

There are no randomized trials on chemotherapy versus surgery or radiotherapy versus surgery. Neither are there randomized trials comparing different agents. Urba et al. pooled adenocarcinoma (75%) with squamous cell carcinoma and found a survival trend favouring multimodal therapy as reported above. Walsh et al. [43] reported the only randomized trial comparing preoperative 5-FU and cisplatin and 40 Gy radiotherapy with surgery alone in 113 patients. Chemotherapy consisted of two cycle of. Radiation therapy was given over 15 days, the first five concurrent with the first course of chemotherapy. The median survival based on intention to treat, was 17 versus 12 months favoring multimodality therapy (p = 0.003). The actual five-year survival rates were 34% and 2% respectively. Multimodal treatment was superior to surgery alone for resectable adenocarcinoma.

Summary

Preoperative but not postoperative chemotherapy has shown promise in large randomized trials. Neither preoperative nor postoperative radiotherapy has proven efficacy. Concurrent combined treatment given preoperatively using multiple agents has shown a complete pathological response of 20-30% for adenocarcinoma with a survival advantage over surgery alone. Refinements of current agents may add a few percentage points of difference. For a radical difference new strategies are needed.

References

1. O'Sullivan GC, Sheehan D, Clarke A, et al. Micrometastases in esophagogastric cancer: high detection rate in resected rib segments. Gastroenterology 1999;116 (3):543-548.

2. Wolfe WG, Vaughn AL, Seigler HF, et al. Survival of patients with carcinoma of the esophagus treated with combinedmodality therapy. J Thorac Cardiovas Surg 1993;105:749-755.

3. Urba S, Orringer M, Turris A, et al. A randomized trial comparing surgery to preoperative concomitant chemoradiation plus surgery in patients with resectable esophageal cancer: updated analysis. Proc ASCO 1997;16:227a.

4. Schlag PM. Randomized trial preoperative chemotherapy for squamous cell cancer Arch Surg 1992;127:1446-1450.

5. Nygaard K, Hagen S, Hansen HS, et al. Pre-operative radiotherapy prolongs survival in operable esophageal carcinoma: a randomized, muilticenter study of pre-operative radiotherapy and chemotherapy. World J Surg 1992;16:1104-1110.

6. Law SY, Fok M, Chow S, et al. Preoperative chemotherapy versus surgical therapy alone for squamous cell carcinoma of the esophagus. A prospective randomized trial. J Thorac Cardiovasc Surg 1997;114:210-217.

7. Medical Research Council. MRC/UKCCCR Upper GI Tract Cancer National Meeting, London, UK. 10th Feburary 2000.

8. Kok TC, Lanschot JV, Siersema PD, et al. Neoadjuvant chemotherapy in operable esophageal squamous cell cancer:final report of a phase III multicentre randomized controlled trial. Proc Am Soc Clin Oncol 1997;16:277a.

9. Pouliquen X, Levard H, Hay JM, et al. 5-Fluouracil and cisplatin therapy after palliative surgical resection of squamous cell carcinoma of the esophagus. A multicenter randomized trial. French Association for Surgical Research. Ann Surg 1996;223:127-133.

10. Ando N, Iizuka T, Ide H, et al. A randomized trial of surgery alone vs surgery plus postoperative chemotherapy with cisplatin and 5-fluorouracil for localized squamous carcinoma of the thoracic esophagus: the Japan Clinical Oncology Group Study (JCOG 9204). Presented at ASCO, Atlanta, 1999.

11. Roth JA, Pass HI, Flanagan MM, et al. Randomized clinical trial of preoperative and postoperative adjuvant chemotherapy with cisplatin, vindesine and bleomycin for carcinoma of the esophagus. J Thorac Cardiovasc Surg 1988;96:242-246.