Most of the series which will be discussed in the upcoming sections are based on populations of squamous cell carcinoma of the esophagus. There is a clear benefit in local control with the addition of radiation and possibly a survival advantage. However, many of these studies were conducted prior to the advent of PET staging by which we now can identify 10-15% of patients with occult metastatic disease which may change their management and survival outcomes.
The largest of these series is by Xiao and included 495 patients with squamous cell carcinoma of the esophagus who received postoperative radiation therapy (n=220) or surgery alone (n=275) (24)
. Radiation portals encompassed the bilateral supraclavicular areas and entire mediastinum to a total of 60 Gy (40 Gy prescribed to midplane and 20 Gy from horizontal portals, treated over 6 weeks). Survival was improved non-significantly with the addition of RT from 32% to 41% (p=0.45). Stage III patients had a distinct, significant overall survival improvement with the addition of RT from 13% to 35% at 5 years (p=0.003). This trial has been criticized for not employing an intention-to-treat analysis, since it excluded 54 patients who did not complete the planned course of treatment. The lack of informed patient consent called into question the ethical standards of this trial (25)
In a separate retrospective analysis by Xiao et al by extent of lymph node status, 549 patients were classified into three groups: Group 1: no lymph node involvement, Group 2: one-two positive lymph nodes, Group 3: three or more positive lymph nodes. The 5-year survival rate of patients with positive lymph nodes (Groups 2 and 3) was 18% with surgery alone compared to 34% with the addition of RT (p=0.038) (26)
. Also, for similar stage III patients, the number of lymph nodes predicted survival outcomes with 5-year survival at 58% for group 1, 31% for Group 2, and 14% for Group 3. Although there was no survival benefit for lymph node negative patients, those with one to two positive lymph nodes had an improvement in 5-year overall survival with the addition of RT from 24% to 45%. For patients with 3 or more positive lymph nodes, 5-year survival outcomes were 21% with RT versus no survivors with surgery alone. Not only is number of metastatic lymph nodes prognostic, but the addition of RT improved survival in patients with positive lymph nodes.
An analysis of the Surveillance Epidemiology and End Results (SEER) database evaluated the impact of adjuvant radiation in 1046 patients, who received surgery alone (65%) or postoperative radiation (35%) (27)
. For Stage III patients there was significant improvement in median (15 to 19 months), 3-year overall survival (18 to 29%) (p< 0.001), and disease specific survival (18 to 24 months) (p< 0.001) which was present for both adenocarcinoma and squamous cell carcinomas. No improvement in survival was seen with Stage II esophageal cancer (AJCC 6th
edition) with the addition of postoperative RT. Multivariate analysis also confirmed that the addition of adjuvant RT was associated with an improved survival (HR 0.70, 95% CI 0.59-0.83, p<0.001). This analysis is limited by the lack of information about chemotherapy, radiation fields and doses, and margin status.
Teniere et al evaluated patients with squamous cell carcinoma of the middle to lower third of the esophagus and randomized them to observation (n =102) or postoperative RT (n=119) (45-55 Gy in 1.8 Gy per fraction to the bilateral supraclavicular regions, mediastinum, and involved celiac lymph nodes) (28)
. Patients were stratified by nodal involvement extent. Five-year survival in node negative patients was 38% versus 7% with involved nodes. Postoperative RT did not confer a survival benefit (5-year survival of 19% in both arms). Rates of local regional recurrence were lower in patients receiving postoperative radiation versus surgery alone (85% vs
70%) but not statistically significant. Patients without nodal involvement did have significant improvement in local regional recurrence with the addition of radiation therapy (90% vs
Fok et al included both squamous cell carcinoma and adenocarcinoma histologies in their study and stratified patients based on palliative (n=70) versus curative (n=60) resection prior to randomization to postoperative RT versus observation (29)
. Prescribed radiation doses of 49 Gy for curative resection and 52.5 Gy for palliative resection in 3.5 Gy per fraction were used, delivered to a 5 cm margin both proximal and distal to the initial tumor extent as delineated by barium swallow. Although they demonstrated a decline in local recurrence rates for those who underwent palliative resection followed by adjuvant RT (20% postoperative RT, 46% no RT, p=0.04), there was no statistical difference in local recurrence for those who had complete resection (15% with RT versus 31% with surgery alone, p=0.06). The overall median survival was significantly shorter for patients receiving postoperative RT (8.7 months) versus control (15.2 months). In patients with residual tumor in the mediastinum after resection, two died of tracheobronchial obstruction compared to nine in the control group. The authors concluded that the shorter survival of patients who underwent postoperative radiotherapy was the result of irradiation-related death and the early appearance of metastatic disease, although patients were less likely to have a recurrence obstructing the tracheobronchial tree. The major criticism of this trial has been the large fraction sizes and total dose delivered which may have contributed to the increased mortality rates and resulted in substantially higher gastric pull-up complications (37% with RT versus 6% with surgery alone) and six fatal bleeding events in the RT group. Similarly, Zieren et al evaluated 68 squamous cell carcinoma patients who were randomized to either observation or postoperative RT, finding no difference in overall or disease-free survivals, but an increase in fibrotic esophageal strictures in the RT arm (30)
Trials postoperative radiotherapy versus surgery alone
In a meta-analysis of postoperative radiotherapy trials, no significant difference in the risk of mortality with postoperative radiotherapy and surgery at one year compared with surgery alone was detected (RR, 1.23; 95% CI, 0.95 to 1.59; p = 0.11) (31)
. The rate of local recurrence with radiotherapy was lower in the tirals of Xiao and Fok (24)
, but the two trials of Teniere and Zieren (28)
noted this benefit was achieved at the expense of increased morbidity.
Given modern day techniques, improved treatment planning with strict dose volume histogram data, postoperative RT is expected to be safer with less toxicity than previous studies. Based on the aforementioned studies, improvements in local control can be expected and is particularly important in the setting of nodal positivity or R1/R2 resection.