Local endoscopic resection ± thermal ablation (PDT, BARRX) is indicated for tumors restricted to the mucosa with a size <2 cm and Barrett esophagus with low-grade or high-grade dysplasia. Endoscopic mucosal resection (EMR) is a procedure where the inner lining of the esophagus is removed with instruments attached to the endoscope in “piece meal technique”. Endoscopic submucosal dissection (ESD) of tumors has improved the success rate of “en bloc resection” but is still technically difficult for large lesions. After the abnormal tissue is removed, patients take drugs called proton pump inhibitors to suppress acid production in the stomach. This can help keep the disease from returning. Endoscopic therapy is highly effective and safe for patients with mucosal AC, with excellent long-term results. In an almost 5 year follow up of 1,000 patients treated by endoscopic resection, there was no mortality and less than 2 % had major complications [16]. Photodynamic therapy (PDT) is a method that can be used to treat tumor remnants after local endoscopic resection. PDT alone for early and late stage esophageal cancer has been abandoned due to a lack of acquisition of histology samples, limitation of effectivity to the tumor surface, and a high rate of postprocedural strictures. For this technique, a light-activated photosensitizer (porfimer sodium (Photofrin^™) is injected into a vein. Over the next couple of days, the drug is more likely to enrich in cancer cells than in normal cells. A special type of laser light is then focused on the cancer through an endoscope. This light causes apoptosis inside the cancer cells. Instead of PDT radiofrequency ablation (RFA) can be used which rarely causes strictures or bleeding in the esophagus and induces a high rate of complete remission [17–19]. In this procedure, a balloon containing many small electrodes (BARRX^™) is passed into an area of tumor through an endoscope. The balloon is then inflated so that the electrodes are in contact with the inner lining of the esophagus. Then an electrical current is passed through it, which kills the cells in the lining by heating them. Over time, normal cells will grow in to replace the tumor cells. Endoscopy (with biopsies) then is done periodically to watch for any further changes in the lining of the esophagus. In case of a non-operable tumor situation endoscopic metal stent placement is a proper solution to resolve dysphagia (Fig. 10.4). Stent placement is also important to treat certain surgical complications (see Sect. 10.9.2.1).

Surgery is regarded as standard treatment only in carefully selected operable patients with localized tumors. T1/T2-tumors without metastases are suitable for primary surgery, in most of the cases subtotal en-bloc-esophagectomy with two field lymphadenectomy is preferred. For localized disease with suspected lymph node involvement (N1-3) preoperative therapy is recommended for AC. Transthoracic esophagectomy with two-field lymph node resection and a gastric tube anastomosized in the left neck is recommended for intrathoracic SCC. If the stomach has been removed in a previous operation a colon segment can be interposed. Debates continue about minimally invasive techniques that should reduce postoperative complication rates and recovery times. No standard treatment can be identified for carcinomas of the cervical esophagus. The extent of surgery in AC of the mid-to-distal esophagus or AC of the gastric cardia is still a matter of debate since one randomized study showed no significant improvement in long-term survival for extended transthoracic compared with transhiatal resection (in the case of a transhiatal resection lymph nodes of the middle and lower mediastinum are spared) [20, 21]. However, compared with limited transhiatal resection extended transthoracic esophagectomy for AC of the mid-to-distal esophagus showed an ongoing trend towards better 5-year survival. Moreover, patients with a limited number of positive lymph nodes in the resection specimen seem to benefit from an extended transthoracic esophagectomy. Another technique for AC of the gastric cardia is the Merendino procedure. This approach is suitable for patients without suspicion of lymph node involvement since lymph nodes of the middle and lower mediastinum and several lymph nodes of the lesser curvature of the stomach are spared. Best results are achieved by surgeons who have the highest numbers of esophageal cancer patients. A nationwide Swedish population-based study figured out a statistically significant reduction in 3-months mortality for surgeons with high annual or cumulative operation volume [22]. These results were similar to an earlier American study by Birkmeyer et al. [23]. Interestingly, there was no independent association between annual hospital volume and overall survival, and hospital volume was not associated with short-term mortality after adjustment for hospital clustering effects. This was different to the earlier American study by Birkmeyer et al. [24]. Patients who do respond poorly to neoadjuvant chemotherapy may benefit from a salvage operation [25], especially if R0-resection can be achieved [26]. In general, the most important prognostic factors of esophageal surgery seem to be curative (R0-) resection and extended lymphadenectomy.

T3/T4 and T1-2 N+ AC tumors should be primarily treated with neoadjuvant chemotherapy or chemoradiation (see Sect. 10.9.5) in order to increase chance of curative (R0)-resection. A recent meta-analysis including ten studies and 2,062 randomized patients showed a significant improvement in overall survival (OS) after neoadjuvant chemotherapy, with a relative risk reduction of 13 % (HR 0.87; 95 % CI 0.79–0.96; p = 0.005), resulting in a 2 year survival benefit of 5.1 %. Whereas this difference was not significant for patients with SCC (HR 0.92; 95 % CI 0.81–1.04; p = 0.18) it was highly significant for patients with AC (HR 0.83; 95 % CI 0.71–0.95; p = 0.01) [27]. However, in Japan, neoadjuvant chemotherapy for SCC with cisplatin/5–fluorouracil is still regarded as standard treatment and can not be replaced by adjuvant chemotherapy with cisplatin/5–fluorouracil [28]. Perioperative chemotherapy of distal esophageal and EGJ cancer was first established with the phase III UK-MAGIC-study, which was published in 2006 in the New England Journal of Medicine and primarily designed for stomach cancer patients [29]. In this trial a perioperative ECF (epirubicin/cisplatin/5–fluorouracil) regimen (n = 250) decreased tumor size and stage and significantly improved progression-free survival (PFS) (HR 0.66; 95 % CI 0.53–0.81; p < 0.001) and overall survival (OS) (HR 0.75; 95 % CI 0.6–0.93; p = 0.009) in comparison to surgery alone (n = 253). Cisplatin/5-fluorouracil regimen as an alternative in this setting (distal esophageal, EGJ and stomach cancer) was published 5 years later derived from the results of the French FFCD multicenter phase III trial (n = 113 for perioperative chemotherapy and n = 111 for surgery alone) [30]. This trial showed a significantly increased curative resection rate, disease-free survival (DFS) (HR 0.65; 95 % CI 0.48–0.89; p = 0.003) and OS (HR 0.69; 95 % CI 0.5–0.95; p = 0.02). Interestingly, in these trials only 49.5 % respectively 50 % of patients who completed preoperative chemotherapy and surgery underwent postoperative chemotherapy as planned in the treatment protocol. This was mainly due to disease progression or early death, patient choice, postoperative complications, problems with the Hickman catheter, previous toxic effects, lack of response to preoperative treatment, and worsening coexisting diseases underlining the higher impact of preoperative chemotherapy on survival data.

Neoadjuvant radiation was evaluated in six randomized fully published studies. Clinical response was detected in about one third of patients, however there was no significant advantage in survival. Two studies even reported a decreased OS after neoadjuvant radiation. A meta-analysis comprising 1,147 patients with mostly SCC from five randomized studies concludes that neoadjuvant radiation results in a 11 % relative risk reduction for the endpoint death (HR 0.89; 95 % CI 0.78–1.01) [31]. Difference in survival was 3 % after 2 years and 4 % after 5 years. This result was statistically non-significant (p = 0.062).

Preoperative chemoradiation with cisplatin/5–fluorouracil regimen and 41.4–45 Gy in 1.8 Gy fractions is recommended in T3-4 AC and SCC tumors and T1-2 N+ AC tumors. It is suggested, however, that preoperative chemoradiation will also increase post-operative mortality rates. In cases of response to neoadjuvant chemoradiation (SCC) further continuation results in equivalent OS compared with surgery alone, albeit that the non-operative strategy is associated with a higher local tumor recurrence (see Sect. 10.9.7). Bimonthly FOLFOX instead of the cisplatin/5-fluorouracil regimen does not improve PFS and has similar toxicities according to a large randomized phase III study (AC and SCC, PRODIGE 5/ACCORD 17 trial, ASCO 2012, LBA 4003 and Lancet Oncol 2014; 15: 305-314). The large phase III CROSS study from the Netherlands established a neoadjuvant carboplatin/paclitaxel/radiation regimen [32]. Two-hundred-seventy five patients (75 %) had AC, 84 (23 %) had SCC, and 7 (2 %) had large-cell undifferentiated carcinoma. Complete resection (R0) was achieved in 92 % of patients in the chemoradiation-surgery group versus 69 % in the surgery group (p < 0.001). Postoperative complication rate was similar in the two treatment groups, and in-hospital mortality was 4 % in both. Median OS was 49.4 months in the chemoradiation-surgery group versus 24.0 months in the surgery group (HR 0.657; 95 % CI 0.495–0.871; p = 0.003). After 24 months chemoradiation reduced local recurrence rate from 34% to 14% (p<0.001; Oppedijk V., JCO 2014; 32: 385-91). In contrast, the FFCD9901 phase III study did not show any benefit for neoadjuvant chemoradiation with 5-fluorouracil and cisplatin for early tumor stages UICC I and II (Mariette C., JCO 2014; 32: 2416-22). Finally, two meta-analyses of older randomized controlled trials for neoadjuvant chemoradiation showed a clear benefit in terms of OS in comparison to surgery alone, especially for patients with AC [27, 33]. In detail, the meta-analysis by Jin et al. comprised 11 randomized controlled trials from 1992 to 2008 including 1,308 patients [34–44]. The meta-analysis by Sjoquist et al. included 17 randomized controlled trials from their previous meta-analysis and 7 further studies. Twelve were randomized comparisons of neoadjuvant chemoradiation versus surgery alone (n = 1,854) [32, 34–42, 45, 46], nine were randomized comparisons of neoadjuvant chemotherapy versus surgery alone (n = 1,981), and two compared neoadjuvant chemoradiation with neoadjuvant chemotherapy (n = 194) in patients with resectable esophageal carcinoma. One factorial trial included two comparisons and was included in analyses of both neoadjuvant chemoradiation (n = 78) and neoadjuvant chemotherapy (n = 81).

In the pivotal Intergroup-0116 phase III trial by Macdonald et al. adjuvant chemoradiation (without preoperative chemotherapy) improved both DFS (HR 1.52; 95 % CI 1.23–1.86; p < 0.001) and OS (HR 1.35; 95 % CI 1.09–1.66; p = 0.005) in curatively resected patients with mainly gastric and EGJ AC [47]. Updated results from last year, confirmed that adjuvant chemoradiation (45 Gy radiation dose) remains a rational standard therapy for curatively resected gastric and EGJ cancer with primaries T3 or greater and/or positive nodes (n = 559 in the study) at least in the United States where D2 resection is less common than in Europe or Japan [48]. For this reason, the Intergroup-0116 study was criticized in Asia and Europe because a majority of patients received less than a D1 lymph node dissection at surgery, whereas fewer than 10 % underwent the more extensive D2 resection. This gave way to speculation that postoperative chemoradiation simply compensated for inadequate surgery. Although significantly fewer local and regional recurrences were found in the chemoradiation group the absolute number of local recurrences was to small to draw definitive conclusions. However, a Danish phase II study examining only patients with EGJ AC recently confirmed the Intergroup-0166 results (116 patients were treated with adjuvant chemoradiation) [49]. Median time of survival was prolonged by 10 months in favour of those who received chemoradiation.

Selected unfit patients with localized tumors not considered for surgery can be treated with curative intent by combined chemoradiation. Otherwise, principles of palliative therapy are recommended for these patients. A randomized controlled phase III study from the United States (RTOG trial 85-01) clearly demonstrated superiority of chemoradiation in comparison to radiation alone in patients with SCC and AC [50]. However, chemotherapy could be administered as planned in only 89 (68 %) of 130 patients (10 % had life-threatening toxic effects with combined therapy vs. 2 % in the radiation only group). Four courses of cisplatin/5–Fluorouracil combined with radiation doses of 50.4 Gy in fractions of 1.8 Gy are regarded as standard in the USA. Increased radiation doses up to 60 Gy in fractions of 1.8–2.0 Gy are recommended in parts of Europe and Japan. Michael Stahl and colleagues compared chemoradiation (etoposide and cisplatin, 40 Gy) followed by surgery (arm A, n = 86) with definitive chemoradiation (60 Gy) (arm B, n = 86). OS was equivalent in both SCC groups, local PFS was better in arm A (HR 2.1; 95 % CI 1.3–3.5; p = 0.003), but treatment related mortality less in arm B (3.5 % vs. 12.8 %, p = 0.03) [51]. These results were confirmed by a similar randomized French trial (259 patients were randomly assigned) using 5–fluorouracil and cisplatin as combination partners for radiation (only SCC patients) [52]. Median survival time was 17.7 months in the surgery group vs. 19.3 months in the definitive chemoradiation group. A third prospectively randomized study from Hong Kong (81 patients were randomly assigned) demonstrated a remarkable 5-year survival rate of 48.6 % for the definitive chemoradiation (5–fluorouracil/cisplatin/50–60 Gy) group and a trend to improved 5-year survival in node-positive disease (only SCC patients) [53]. In a recent study presented at ASC0 2012 (PRODIGE 5/ACCORD 17 trial; Hong TS et al.; LBA 4003 and Conroy T., Lancet Oncol 2014; 15: 305-314) patients with non-operable localized esophageal carcinoma (85 % SCC, 15 % AC) were randomized to two different chemoradiation protocols. Radiation dose was 50 Gy in both arms. Patients in A received six cycles of FOLFOX (5–fluorouracil/leucovorin/oxaliplatin) every 2 weeks and patients in arm B four cycles of 5–fluorouracil/leucovorin/cisplatin every 3 weeks. PFS (9.7 month vs. 9.4 month), the primary study endpoint, and OS survival (20.2 months vs. 17.5 months) were similar in both arms. Therefore, a radiation protocol with FOLFOX may be a good alternative in patients with renal insufficiency or reduced general condition. Addition of epithelial derived growth factor receptor (EGFR) 1 inhibitor cetuximab to a capecitabine/cisplatin/radiation backbone did result in greater toxicity, lower rate of completion of standard therapy and significantly worse survival (22 months vs. 25 months; p = 0.043) in patients with locally advanced SCC (73 %) or AC (27 %) as demonstrated by a recent large UK study (SCOPE-1, NCT00509561) [54]. Docetaxel/cisplatin/radiation combination is feasible too, as demonstrated in a Korean phase II study (36 SCC patients) [55]. Moreover, carboplatin/paclitaxel/radiation as in the neoadjuvant setting (see above) is another option (Honing J., Ann Oncol 2014; 25: 638-643). In a recent meta-analysis of three randomized studies definitive chemoradiation in patients with SCC did not demonstrate any survival benefit over other curative strategies, but treatment-related mortality rates were lower (HR 7.60, p = 0.007) [56]. In addition, a recent analysis from the American National Cancer Database unveiled that OS was lower for patients with stage II/III disease of either histologic subtype treated with chemoradiation alone as compared with surgery plus chemoradiation (p < 0.001) [57]. A study from Korea suggested vascular endothelial growth factor (VEGF) as positive predictive factor and cyclooxygense-2 (COX-2) as negative prognostic factor for OS in patients with SCC after definitive chemoradiation [58]. Improvement of definite chemoradiation for locally advanced disease is a focus of current research. Proton-beam therapy (PBT) and intensity modulated radiation therapy (IMRT) are both forms of radiation therapy that are designed to treat a specific area of the body while affecting as little of the surrounding normal tissue as possible. PBT is a newer technology that is designed to further reduce the amount of radiation that affects the surrounding normal tissue. A particle accelerator is used during treatment to hit the tumor with a beam of protons. As a result, DNA damage of cells is induced by these charged particles, ultimately resulting in cell death or decrease of cell proliferation. Since tumors show a high rate of cells division and a reduced rate of cell repair they are particularly vulnerable to attacks on DNA. Protons have little lateral side scatter in the tissue due to their relatively large mass. The beam stays focused on the tumor shape, does not broaden much, and causes only low-dose side-effects to surrounding tissue. IMRT, which is less expensive, comprises an advanced mode of high-precision radiotherapy that uses computer-controlled linear accelerators (3-D computed tomography (CT) or magnetic resonance images (MRI) are used for planning) to deliver precise radiation doses to a malignant tumor or specific areas within the tumor. The radiation dose can be more precisely adjusted to the three-dimensional shape of the tumor by modulating—or controlling—the intensity of the radiation beam in multiple small volumes. Using IMRT, higher radiation doses and combinations of multiple intensity-modulated fields coming from different beam directions can be focused to regions within the tumor while the dose to surrounding normal critical structures can be minimized.