Neoadjuvant Chemoradiotherapy for Resectable Oesophageal and Gastro-oesophageal Junction Cancer—Do We Need Another Randomised Trial?

Clinical Oncology 23 (2011) 696e705 Contents lists available at ScienceDirect

Clinical Oncology journal homepage: www.elsevier.com/locate/clon

Overview

Neoadjuvant Chemoradiotherapy for Resectable Oesophageal and Gastro-oesophageal Junction CancerdDo We Need Another Randomised Trial? M. Hingorani *, T. Crosby y, A. Maraveyas *, S. Dixit *, A. Bateman z, R. Roy * * Queen’s

Centre for Oncology and Haematology, Castle Hill Hospital, Cottingham, UK Velindre Cancer Centre, Cardiff, UK z Southampton University Hospitals NHS Trust, Southampton, UK y

Received 16 February 2011; received in revised form 22 March 2011; accepted 18 May 2011

Abstract Aims: The optimal neoadjuvant therapy option for locally advanced oesophageal cancer remains elusive. Neoadjuvant chemoradiotherapy (CRT) is the preferred modality of choice in the USA. In contrast, neoadjuvant chemotherapy is commonly used in the UK. We provide a comprehensive overview of the available evidence for defining the ideal neoadjuvant treatment algorithm. Materials and methods: The PubMed database combined with American Society of Clinical Oncology and American Society for Therapeutic Radiology and Oncology websites were searched online to identify randomised studies and published meta-analyses that have compared these modalities compared with surgery alone. In particular, we searched for randomised trials that may have directly compared outcomes after neoadjuvant CRT or chemotherapy. Results: We identified 17 published randomised studies of neoadjuvant CRT (n ¼ 9) and chemotherapy (n ¼ 8) compared with surgery alone and one prospective series that compared the above modalities against each other. Studies evaluating CRT have reported pathological complete response rates of 15e40% and no increase in postoperative mortality was observed, except in one study that used a hypofractionated radiation schedule. Two randomised studies showed significant survival benefit and the remaining (n ¼ 7) were negative, but showed a trend towards improved survival. Furthermore, at least four meta-analyses have shown improved survival in favour of CRT extending up to an absolute benefit of 13% at 2 years. In comparison, five studies of neoadjuvant chemotherapy showed no survival difference and two of the remaining studies that showed significant benefit included gastric adenocarcinomas and used peri-operative chemotherapy. All the above studies have shown uniformly poor pathological complete response rates of less than 10 percent. Moreover, three metaanalyses were negative, but two showed up to 7% absolute survival benefit at 2 years in favour of chemotherapy. The trial comparing the above modalities showed a trend towards improved survival in favour of CRT, but closed early due to poor recruitment. Conclusion: Data from the above studies are potentially conflicting and inconclusive for defining the optimal neoadjuvant treatment schedule. In our opinion, the above question can only be answered within the context of a randomised control trial. We have included a proposal for a trial design for direct comparison of these modalities. Ó 2011 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. Key words: Chemoradiotherapy; neoadjuvant chemotherapy; oesophageal cancer

Statement of Search Strategies Used and Sources of Information The searched published literature using the PubMed online database with the following search terms: oesophageal cancer, and neoadjuvant therapy (n ¼ 135), neoadjuvant chemotherapy (n ¼ 128), or chemoradiotherapy (n ¼ 161). The following limits were used for the search:

Author for correspondence: M. Hingorani, Queen’s Centre for Oncology and Haematology, Castle Hill Hospital, Cottingham HU16 5JQ, UK. Tel: þ44-101482-461309. E-mail address: [email protected] (M. Hingorani).

TYPE: clinical trials, meta-analysis, LANGUAGE: English. In addition, we searched the American Society of Clinical Oncology and American Society for Therapeutic Radiology and Oncology websites for any published randomised studies in last 5 years.

Introduction Cancers of the oesophagus or gastro-oesophageal junction (GOJ) have poor prognosis and are often locally advanced or metastatic at presentation [1,2]. Most recent series have reported 5 year survival rates of 15e20% for surgery alone [3e5]. Naturally, the importance of a multidisciplinary

0936-6555/$36.00 Ó 2011 The Royal College of Radiologists. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.clon.2011.05.005

M. Hingorani et al. / Clinical Oncology 23 (2011) 696e705

approach in the management of these tumours cannot be overemphasised. The last decade has witnessed significant developments, with general acceptance of the beneficial effects of neoadjuvant therapy, including the use of systemic chemotherapy or preoperative chemoradiotherapy (CRT). The therapeutic rationale for the use of neoadjuvant strategies relates to potential down-staging of the primary tumour to facilitate complete surgical resection (R0) and the targeting of occult micro-metastatic disease. However, the trials investigating these modalities have generated conflicting results and currently there seems to be no worldwide consensus as to the optimal neoadjuvant approach. This is reflected in the more frequent use of preoperative CRT in the USA compared with the UK practice of using neoadjuvant chemotherapy. In the UK there has been a general reluctance to adopt the practice of preoperative CRT, primarily due to concerns about surgical morbidity and complications. Here we provide a comprehensive overview of the available evidence on the role of neoadjuvant CRT. We present a balanced argument against the blanket approach of using neoadjuvant chemotherapy for all patients in the UK and emphasise the important need for revisiting the case for a tri-modality approach in the preoperative management of oesophageal and GOJ cancer. Finally, we propose a randomised controlled study design to provide us with a more definitive answer.

Materials and Methods We searched published literature using the PubMed online database with the following search terms: oesophageal cancer, and neoadjuvant therapy (n ¼ 135), neoadjuvant chemotherapy (n ¼ 128) or chemoradiotherapy (n ¼ 161). The following limits were used for the search: TYPE: clinical trials, meta-analysis, LANGUAGE: English. In addition, we searched the American Society of Clinical Oncology and American Society for Therapeutic Radiology and Oncology websites for any published studies in last 5 years. The primary aim was to carry out a comparison of neoadjuvant CRT and chemotherapy with respect to treatment-related outcomes (see below). We aimed to identify randomised studies that had evaluated the effects of neoadjuvant CRT or chemotherapy compared with surgery alone. In particular we searched for any studies that had compared these modalities directly against each other. Studies evaluating definitive CRT or effects of neoadjuvant CRT in early stage (I or II) oesophageal cancer were excluded. In addition, we excluded studies that had used a suboptimal dose of radiotherapy (<30 Gy) or an unexplained non-conventional dose of chemotherapy drugs. The final search revealed 17 randomised studies of neoadjuvant CRT (n ¼ 9) or chemotherapy (n ¼ 8) compared with surgery alone in patients with potentially operable oesophageal or GOJ cancer that had reported on long-term treatment-related outcomes. These included the quality of surgical resection, the incidence of pathological complete response (PCR), postoperative mortality and overall

697

survival. We only identified one prospective randomised series that was designed to compare these modalities directly against each other. The search also revealed six published meta-analyses for neoadjuvant CRT and five meta-analyses for chemotherapy, which have been included for discussion.

Results Neoadjuvant Chemoradiotherapy The use of neoadjuvant CRT has several potential advantages. Patients will probably tolerate CRT much better preoperatively due to the prolonged physiological recovery after an oesophagectomy and the radiation target volume is easier to define because of visible disease. In contrast, postoperative CRT may be associated with radioresistance caused by inadequate oxygenation of post-surgical tissues and the presence of a gastric conduit within the thorax, which may increase radiotherapy-induced toxicity [6]. Most conventional CRT protocols for oesophageal cancer have involved the use of a platinum agent and fluoropyrimidine (5fluorouracil; 5-FU). These studies are summarised in Table 1. In one of the earliest trials published by Apinop et al. [7] 69 patients with squamous oesophageal cancer were randomised to CRT followed by surgery or surgery alone. The CRT (n ¼ 35) arm received two cycles of cisplatin (100 mg/m2) and 5-FU (1000 mg/m2 days 1e4) delivered concurrently with radiotherapy (40 Gy in 20 fractions). The trial reported a PCR rate of 20% and 5 year survival of 24% in the CRT arm compared with 10% in the surgery alone arm, which was not statistically significant. Postoperative morbidity and mortality were similar in the two arms [7]. Subsequently, Walsh et al. [8] reported on a randomised trial of 113 patients with oesophageal or GOJ adenocarcinomas who were randomly assigned to surgery alone or preceded by CRT. Patients received two cycles of chemotherapy (5-FU 15 mg/kg days 1e5 and cisplatin 75 mg/m2 on day 7) with radiotherapy (40 Gy in 15 fractions over 3 weeks) delivered concurrently with the first cycle of chemotherapy. The study showed a significantly reduced frequency of pathological nodal involvement (42% versus 82%) and a 25% incidence of PCR after CRT. Neoadjuvant CRT significantly improved the median (16 months versus 11 months) and 3 year survival (32% versus 6%) (P ¼ 0.01) [8]. These results were criticised because of the lower than expected survival with surgery alone, the lack of diagnostic computed tomography as an essential staging investigation, and the presence of poor statistics, including the publication of incongruent survival plots [9,10]. The European Organization for Research and Treatment of Cancer (EORTC) multicentre study randomised 297 patients with resectable oesophageal squamous cell carcinoma to immediate surgery or CRT (two cycles of cisplatin 80 mg/m2 weeks 1 and 4) followed by split-course hypofractionated radiotherapy (two courses of 18.5 Gy in five fractions administered during weeks 1 and 4). Patients receiving CRT had a significant improvement in local

698

M. Hingorani et al. / Clinical Oncology 23 (2011) 696e705

Table 1 Randomised trials of neoadjuvant chemoradiotherapy (CRT) in potentially operable oesophageal cancer. Patients were randomised to neoadjuvant CRT followed by surgery or surgery alone Reference

Protocol

Outcome

Survival

[7] SCC of oesophagus (n ¼ 69)

Cisplatin, 5-FU 40 Gy in 20 fractions

No increase in postoperative complications PCR 20%

5 year survival 24% (CRT þ S) compared with 10% (S) Median survival of 9.7 months (CRT þ S) compared with 7.4 months (S)

[8] AC of oesophagus and GOJ (n ¼ 113)

Cisplatin, 5-FU 40 Gy in 15 fractions

No increase in postoperative complications PCR 25%

3 year survival 32% (CRT þ S) compared with 6% (S) (P ¼ 0.01)

[11] Oesophageal SCC (n ¼ 297)

Cisplatin split course hypo-fractionated radiotherapy (37 Gy)

16.7% patients died in CRT þ S arm compared with 5% in S arm PCR 26%

No difference in median survival (18.6 months in both groups)

[12,13] Oesophageal and GOJ cancer (AC ¼ 75) (n ¼ 100)

Cisplatin, 5-FU, vinblastine 45 Gy in 15 fractions

No increase in postoperative complications PCR 28%

3 year survival 30% (CRT þ S) compared with 16% (S)

[14] Oesophageal SCC (n ¼ 101)

Cisplatin, 5-FU 45.6 Gy in 1.2 Gy BD

No increase in postoperative complications PCR 43%

Median survival 28.2 months (CRT þ S) and 27.3 months (S)

[15] Oesophageal cancer (AC ¼ 158) (n ¼ 256)

Cisplatin, 5-FU 35 Gy in 15 fractions

No increase in postoperative complications PCR 16%

[16] Oesophageal cancer (AC ¼ 45) (n ¼ 91)

Cisplatin, 5-FU 64 Gy in 32 fractions

No increase in postoperative complications

Median survival of 22.2 months (CRT þ S) compared with 19.3 months (S) 4 year survival 29% (CRT þ S) compared with 23% (S)

[17] Oesophageal cancer (AC ¼ 42) (n ¼ 56)

Cisplatin, 5-FU 50.4 Gy in 28 fractions

No increase in postoperative complications PCR 40%

5 year survival 36% (CRT þ S) compared with 16% (S) (P ¼ 0.002)

[25] Oesophageal and GOJ cancer (AC ¼ 273) (n ¼ 363)

Paclitaxel, carboplatin 41.4 Gy in 23 fractions

No increase in postoperative complications PCR 32.6%

3 year survival 59% (CRT þ S) compared with 48% (S)

PCR, pathological complete response; S, surgery; AC, adenocarcinoma; SCC, squamous cell carcinoma; GOJ, gastro-oesophageal junction; 5-FU, 5-fluorouracil.

control (hazard ratio ¼ 0.6; 0.4e0.9) and progression-free survival (PFS), but no long-term survival benefit was observed with a median survival of 18.6 months in both arms. CRT was associated with a higher incidence of curative surgical (R0) resection, but there were more treatmentrelated deaths in the CRT arm, with 17 patients (16.7%) dying from postoperative complications compared with five patients (5%) in the control arm, which may have negated any probable survival benefit from CRT [11]. Urba et al. [12] reported on a randomised trial of 100 patients with localised oesophageal or GOJ cancer (squamous cell carcinoma ¼ 25; adenocarcinoma ¼ 75) who were randomly assigned to surgery with or without induction CRT. The CRT schedule involved cisplatin (20 mg/m2 per day, days 1e5 and 17e21), infusional 5-FU (300 mg/m2 per day, days 1e4 and 17e20) and vinblastine (1 mg/m2 per day on days 1e4 and 17e20) plus concurrent hyperfractionated radiotherapy (45 Gy in 1.5 Gy twice daily fractions for 3 weeks). Surgery was carried out on day 42 after a 3 week rest. The use of CRT was associated with a PCR rate of 28%, which transformed into a clinically relevant but statistically non-significant improvement in 3 year survival favouring CRT (30% versus 16%). In a subsequent subgroup analysis, complete responders had significantly better 3 year survival compared with those with residual tumour in the resected specimen (64% versus 19%) [13]. The combined therapy group had a significantly lower locoregional recurrence rate

(19% versus 42%), but the distant metastatic rate was similar (65% versus 60%). After a follow-up of 8.2 years, the median survival was similar in both arms (16.9 months for CRT and surgery; 17.6 months for surgery alone). The lack of survival benefit has been attributed to small patient numbers and the inadequate power of the study. Lee et al. [14] reported on a randomised controlled trial of 101 patients with squamous cell carcinoma who were randomised to CRT using cisplatin (60 mg/m2) and 5-FU (1000 mg/m2 days 2e4) combined with radiotherapy (45.6 Gy in 1.2 Gy twice a day on days 1e28) followed by surgery or surgery alone. The use of neoadjuvant CRT was not associated with an increase in postoperative mortality. The trial reported a PCR rate of 43% after CRT and at a median follow-up of 25 months the median survival was 28.2 months in the CRT arm compared with 27.3 months in the surgery alone arm. The trial was characterised by an unexpectedly high drop-out rate for oesophagectomy in the CRT arm and had to close early without recruiting the number of patients required to show a survival difference between the two arms [14]. The Australasian Gastrointestinal Trials Group reported on a study of 256 (squamous cell carcinoma ¼ 95; adenocarcinoma ¼ 158) patients with resectable oesophageal tumours who were randomised to neoadjuvant CRT followed by surgery or surgery alone [15]. The CRT schedule involved 35 Gy given in 15 fractions delivered concomitantly

M. Hingorani et al. / Clinical Oncology 23 (2011) 696e705

with one cycle of cisplatin (80 mg/m2) and a 4 day infusion of 5-FU (800 mg/m2/day). Surgery was carried out at 3e6 weeks after CRT and the use of CRT did not increase the risk of postoperative morbidity (49% in the CRT arm versus 55% in the surgery alone arm). The use of CRT was associated with a higher probability of R0 resection (80% versus 59%), much less incidence of lymph node involvement (43% versus 67%) and a PCR rate of 16%. The intention to treat analysis of 256 randomised patients showed a small trend for improved PFS and overall survival after CRT, but this did not reach statistical significance. The median survival was 22.2 months for patients receiving CRT compared with 19.3 months for those allocated surgery alone. The above study reported an unusually large number of patients (n ¼ 28) in the CRT arm not proceeding to surgery primarily due to clinical deterioration or disease progression. Further subgroup analysis revealed improved outcomes in patients developing a PCR after CRT who had median PFS of 26.2 months and 3 year survival of 49%. Similarly, there was a significant increase in PFS in patients with squamous cell carcinoma tumours (hazard ratio 0.47; 0.25e0.86), but the trial was underpowered for detecting meaningful differences in the overall survival of these tumours. Carstens et al. [16] reported on the results of a randomised controlled trial in 91 patients with resectable oesophageal cancer (squamous cell carcinoma ¼ 50%; adenocarcinoma ¼ 50%). Patients were randomised to surgery or preoperative CRT involving one cycle of induction chemotherapy (cisplatin 100 mg/m2 on day 1 plus infusional 5-FU 750 mg/m2 daily on days 1e5) followed by two further cycles of chemotherapy administered concurrently with radiotherapy (64 Gy in 2 Gy per fraction). In the CRT group, only 50% of the patients completed treatment according to protocol, and 40% required modification of the protocol. However, there were no treatment-related deaths and the postoperative morbidity was not increased after CRT, with 76% of patients undergoing complete R0 resection. The study reported a trend for improved survival at 2 years (37% versus 25%) and 4 years (29% versus 23%) that was not statistically significant [16]. Tepper et al. [17] reported on the results of the CALGB 9781 study, which was originally designed as a randomised trial for evaluating the effects of neoadjuvant CRT compared with surgery in 500 patients with stage IeIII oesophageal or GOJ cancer. Because of poor accrual, the study was closed prematurely after only 56 patients were enrolled (adenocarcinoma ¼ 42; squamous cell carcinoma ¼ 14). The treatment schedule comprised of two cycles of cisplatin (80 mg/m2) and infusional 5-FU (1000 mg/m2/day for 4 days) in weeks 1 and 5 combined with 50.4 Gy in 28 fractions. PCR was achieved in 10 of 25 assessable patients in the tri-modality arm (40%). Surgical morbidity or mortality was not increased in patients receiving CRT. The intent to treat analysis revealed a significant survival benefit in favour of the tri-modality arm with a median survival of 4.48 years compared with 1.78 years, and a 5 year overall survival of 36% compared with 16% (P ¼ 0.002). Kaklamanos et al. [18] reported on an initial metaanalysis of 669 patients from five randomised controlled

699

trials and showed a non-significant absolute survival benefit of 6.4% in favour of neoadjuvant CRT and a 3.4% increase in postoperative mortality. Fiorica et al. [19] reported on a meta-analysis of 764 patients from six randomised controlled trials and showed that neoadjuvant CRT significantly reduced the 3 year mortality rate (odds ratio 0.53; 95% confidence interval 0.31e0.93; P ¼ 0.03). However, the risk for postoperative mortality was higher in the CRT plus surgery group (odds ratio 2.10; 95% confidence interval 1.18e3.73; P ¼ 0.01) [19]. In addition, four other different meta-analyses have shown a beneficial effect of neoadjuvant CRT compared with surgery alone. The first meta-analysis included 1116 patients from nine randomised controlled trials, which showed a statistically significantly improved 3 year survival benefit in patients treated with CRT (hazard ratio 0.45; 0.26e0.79) [20]. The second metaanalysis of 10 randomised controlled trials with 1209 patients was associated with a significantly better 2 year allcause mortality (hazard ratio ¼ 0.81; 0.70e0.93), which transformed into a 13% absolute difference in survival at 2 years. The survival benefit was independent of the histological subtype and the results were similar for squamous cell carcinomas and adenocarcinomas [21]. A recent metaanalysis of 1308 patients from 13 randomised controlled trials showed improved 1, 3 (odds ratio ¼ 1.78; 1.20e2.66) and 5 year (odds ratio ¼ 1.46; 1.07e1.99) survival after neoadjuvant CRT. The incidence of postoperative complications was similar, but the overview reported an increase in mortality after CRT (odds ratio ¼ 1.68; 1.03e2.73) [22]. Finally, a recent systematic overview of 3840 patients from 38 trials reported an average R0 resection and PCR rate of 88.4 and 25.8%, respectively, with the use of CRT. The postoperative mortality was reported as 5.2% and the 5 year survival rates ranged from 16 to 59% for all patients and from 34 to 62% in those with PCR [23]. More recently, at the American Society of Clinical Oncology 2010 proceedings in Chicago, the Dutch Trials Group reported on the initial results of the phase III CROSS study, which randomised patients (n ¼ 363) to neoadjuvant CRT with 5 weekly cycles of paclitaxel (50 mg/m2) and carboplatin (AUC2) combined with concurrent radiotherapy (41.4 Gy in 23 fractions) or surgery alone [24]. Most cancers were adenocarcinomas (n ¼ 273) and the reported R0 resection rate was 92.3% in the CRT arm versus 64.9% in the surgery alone arm; the PCR rate was 32.6%. Postoperative mortality was 3.7% in the surgery alone arm compared with 3.8% in the CRT arm. The overall survival was significantly better in the group of patients treated with CRT (hazard ratio ¼ 0.67; 0.50e0.92) and the median survival was 49 months in the CRT arm versus 26 months in the surgery alone arm. Similarly, the 1, 2 and 3 year survival rates were 82, 67 and 59% in the CRT arm and 70, 52 and 48% in the surgery alone arm [25].

Neoadjuvant Chemotherapy Randomised trials of neoadjuvant chemotherapy compared with surgery alone are summarised in Table 2.

700

M. Hingorani et al. / Clinical Oncology 23 (2011) 696e705

Table 2 Randomised trials of neoadjuvant chemotherapy in potentially operable oesophageal cancer. Patients were randomised to neoadjuvant chemotherapy followed by surgery or surgery alone Reference

Protocol

Outcome

Survival

[26] Oesophageal SCC (n ¼ 46)

2 cycles bleomycin, cisplatin, vinblastine

No PCR

Median survival 17 months in both groups

[27] Oesophageal SCC (n ¼ 147)

2e3 cycles cisplatin, 5-FU

PCR 6.7%

Median survival 16.8 months (C þ S) compared with 13 months (S) (P ¼ 0.17)

[28] Oesophageal SCC (n ¼ 96)

2e3 cycles cisplatin, 5-FU

PCR 12.8%

Median survival 25 months (C þ S) compared with 24 months (S)

[30] Oesophageal or GOJ cancer (n ¼ 802)

2 cycles cisplatin, 5-FU

R0 resection 60% (C þ S) arm compared with 54.4% (S) No PCR reported

5 year survival 23% (C þ S) compared with 17% (S) (P ¼ 0.03)

[31] Oesophagus, GOJ, gastric AC (n ¼ 503)

3 cycles pre- and postoperative epirubicin, cisplatin, 5-FU

R0 resection similar in both arms No PCR reported

5 year survival 36.3% (C þ S) compared with 23% (S) (P ¼ 0.009)

[32] Oesophagus, GOJ, gastric AC (n ¼ 224)

2e3 cycles pre- and post-operative cisplatin, 5-FU

R0 in 87% (CT þ S) arm compared with 74% (S) No PCR reported

5 year DFS 38% (CT þ S) compared with 24% (S) (P ¼ 0.02)

[2] Oesophageal or GOJ cancer (n ¼ 467)

3 cycles cisplatin, 5-FU

R0 resection similar in both arms PCR 2.5%

3 year survival 26% (C þ S) compared with 23% (S)

[33] Gastric and GOJ AC (n ¼ 340)

2 cycles, 5-FU, cisplatin

R0 resection C þ S (81.9%) compared with S (66.7%) No PCR reported

No survival benefit (trial closed early with 144 patients). Median survival 36 months in both arms

PCR, pathological complete response; C, chemotherapy; S, surgery; AC, adenocarcinoma; SCC, squamous cell carcinoma; GOJ, gastrooesophageal junction; 5-FU, 5-fluorouracil.

Initial studies of neoadjuvant chemotherapy were limited to patients with oesophageal squamous cell carcinoma and these studies failed to show a survival difference between the two arms [26e28]. The current UK practice of treating patients with neoadjuvant chemotherapy is based on the results of two important randomised studies d the MRCOEO2 and MAGIC trials. The MRC-OEO2 study reported significant prolongation of the 2 year survival (43% versus 34%) and median survival (16.8 months versus 13.3 months) in favour of the chemotherapy arm [29]. In a recent trial update, the survival benefit was reported to be maintained at 5 years with 23% overall survival in the chemotherapy arm compared with 17% in the surgery alone arm, which was independent of histological subtype (P ¼ 0.03) [30]. Similarly, the MAGIC study reported a 16% reduction in the risk of death and significant prolongation of 5 year survival (36% versus 23%) in the chemotherapy arm (P ¼ 0.009) [31]. The French FNLCC study showed very similar results to those observed in the MAGIC trial, with an overall survival of 38% compared with 24% in favour of chemotherapy (P ¼ 0.02) [32]. Both the MAGIC and French studies included gastric adenocarcinomas and used a perioperative chemotherapy protocol, but less than 50% of patients received the full intended cycles of adjuvant chemotherapy. In contrast, the Intergroup-0113 study involving 467 patients with potentially resectable oesophageal or GOJ tumours failed to show any beneficial effect from the use of neoadjuvant chemotherapy [2]. However, a longer median time (63 days) to surgery and possible

high-quality surgery resulting from explicit and rigid selection criteria were cited as possible causes for the negative outcome. Similarly, the EORTC phase III study presented in an abstract form at the American Society of Clinical Oncology 2009 proceedings reported no survival advantage with the use of neoadjuvant chemotherapy in patients with operable tumours of the stomach and GOJ, but again this may have been influenced by the surgery alone arm doing exceptionally well (median survival more than 36 months) combined with relatively small patient numbers due to poor trial accrual [33]. The conflicting opinion of the above studies is reinforced by the results of at least three different meta-analyses that failed to confirm the survival benefit of neoadjuvant chemotherapy [21,34,35]. In contrast, the meta-analysis by Kaklamanos et al. [18] of seven randomised controlled trials showed a 4.4% absolute survival benefit at 2 years in favour of neoadjuvant chemotherapy. More recently, Gebski et al. [36] reported a review of eight randomised trials of surgery alone or chemotherapy followed by surgery in patients with potentially operable oesophageal cancer (n ¼ 1724). The hazard ratio for all-cause survival at 2 years favoured the use of chemotherapy followed by surgery (hazard ratio ¼ 0.90; 0.81e1.0) with an absolute survival benefit of 7%. However, no survival advantage was observed in patients with squamous cell carcinoma compared with patients with adenocarcinomas [36]. It is important to note that some of the recent trials of peri-operative chemotherapy were excluded from the meta-analysis.

M. Hingorani et al. / Clinical Oncology 23 (2011) 696e705

NeoAdjuvant Chemotherapy Versus Chemoradiotherapy Despite the extensive published literature on neoadjuvant chemotherapy and CRT, there are hardly any randomised studies that have compared one modality against the other. Our search revealed only one prospective series by Stahl et al. [37], which reported on the multicentre German POET (Pre-operative Chemotherapy or Radiochemotherapy in Oesophago-gastric Adenocarcinoma) trial in patients with GOJ adenocarcinomas. Patients were randomly assigned to 16 weeks of chemotherapy alone (cisplatin and leucovorin-modulated infusional 5-FU) or 12 weeks of the same chemotherapy regimen followed by radiotherapy (30 Gy over 3 weeks) concurrent with cisplatin and etoposide. The trial was closed because of poor accrual after only 126 of the planned 394 patients were recruited. Complete (R0) resection was possible in a similar proportion of each group (70%), although the PCR rate was significantly higher after CRT (16% versus 2%). At a median follow-up of 46 months, patients undergoing CRT had trend towards a superior median (33 months versus 21 months) and 3 year survival (47% versus 28%). The trial used a relatively low dose of radiation that may have masked the absolute benefit of CRT. Furthermore, the trial used a prolonged non-conventional schedule of induction and concurrent chemotherapy, which is not standard practice. In another retrospective series of 122 patients, the use of CRT was associated with higher PCR (11% versus 4%), but the recurrence rate and overall survival were broadly similar in both groups [38].

Discussion The role of CRT in the management of oesophageal cancer has evolved in last few years, including its possible use as a definitive treatment modality in selected patients with localised disease [39]. However, in the absence of randomised data, surgery is still considered to be the treatment of choice for patients fit enough for radical resection. The last decade has established the important role of multi-modality therapy for improving long-term outcomes in patients proceeding to radical surgery. However, we are still far from developing an ideal evidencebased neoadjuvant treatment algorithm, despite numerous previous reviews and meta-analyses, some of which show a clear preference in favour of a specific approach, depending on their geographical origin. Recently, the Gastrointestinal Cancer Disease Site Group of Cancer Care Ontario’s Program in Evidence-Based Care published updated clinical guidelines on the adjuvant management of oesophageal cancer and recommended neoadjuvant CRT as the preferred modality of choice and neoadjuvant chemotherapy as the possible second alternative option in selected patients [40]. We do not have specific biological markers to indicate which subgroup of patients would probably benefit from using a particular approach and this is reflected in the significant worldwide variation in clinical practice.

701

The current UK approach favours the use of neoadjuvant chemotherapy and one of the main deterrents against the use of neoadjuvant CRT in the UK is related to the possible risk of increased postoperative complications. Most of the studies with conventional CRT regimens have not shown an increase in the risk of surgical morbidity or mortality. The incidence of surgical mortality has remained fairly similar in these studies (5e7%), apart from the EORTC study by Bossett et al. [11], which showed an unacceptably high (16.7%) risk of postoperative death. However, the latter study used an unconventional split-course of hypo-fractionated radiotherapy that may have explained the above results. The increased risk of surgical mortality identified in a recent meta-analysis was reported to be due to the inclusion of the above study and resolved when this was excluded from the analysis [23]. Based on the above observation, it also seems likely that the results from two other meta-analyses that showed an increase in postoperative mortality after neoadjuvant CRT may have been influenced by the inclusion of the above study [18,19]. In the recent past, the importance of optimal downstaging has been widely recognised and it is now accepted that tumour response and PCR are surrogate markers of long-term outcome [41,42]. Both the MAGIC (69.3% versus 64.4%) and OEO2 (60% versus 54%) trials reported marginally higher rates of R0 resection after chemotherapy, but there were no reports of any PCR. In the French FNLCC study, patients undergoing neoadjuvant chemotherapy had significantly higher rates of curative resection (R0) (87% versus 74%), but there were no reports of any PCR. The study showed a direct linear relationship between the probability of R0 resection and overall survival. The Intergroup-0113 study showed no differences in the rates of complete (R0) resection in the two arms, and PCR was observed only in five (2.5%) of 202 patients. In a later update, the study reported on the prognostic significance of complete (R0) resection (39% 3 year survival compared with 12% for R1 [microscopic] and 4% for R2 [macroscopic] resection). In direct contrast to the above results, most neoadjuvant CRT studies using conventional doublet regimens have consistently shown PCR rates of 15e25%. Furthermore, unlike neoadjuvant chemotherapy, several studies of CRT have shown improvement in local control [10e17] and previous meta-analyses have validated the significant reduction in local failure rates after CRT (odds ratio ¼ 0.38) (0.23e0.63) and 0.64 (0.410.99) [20e23]. Paradoxically, despite these impressive results, most of the randomised trials with neoadjuvant CRT have only shown a trend towards improved survival, compared with the significant survival benefit observed with neoadjuvant chemotherapy in the MRC-OEO2, MAGIC and the French FNLCC/FNDD trials. The primary reason for this striking discrepancy reflects the fact that all these chemotherapy trials were fairly large welldesigned studies conducted mainly in the UK and Europe and among them recruited more than 1500 patients. In contrast, most of the randomised studies involving neoadjuvant CRT have suffered from poor recruitment and low patient numbers with inadequate statistical power to detect a meaningful long-term survival benefit. Furthermore, these

702

M. Hingorani et al. / Clinical Oncology 23 (2011) 696e705

trials have evaluated heterogeneous chemoradiation schedules, making cross-trial comparison difficult. Indeed, the largest study before the recent CROSS trial was the EORTC study with 296 patients [10]. The CALGB9781 study intended to recruit 500 patients, but suffered from poor accrual and closed after only 56 patients [13]. The cause of poor accrual in these trials was probably multifactorial, including the personal preference and opinion of the oncologist, who may not be amenable to the idea of surgery alone. In our opinion, the superiority of one modality over the other d neoadjuvant chemotherapy or CRT d in the management of locally advanced oesophageal and GOJ cancer can be only comprehensively answered in the context of large multicentre randomised controlled trials with adequate statistical power. We propose a randomised controlled trial design that should provide a prospective comparative analysis of standard treatment including initial chemotherapy followed by surgery and postoperative CRT in the event of R1 resection, and an experimental arm of neoadjuvant CRT followed by surgery. Patients with lower oesophageal and GOJ adenocarcinoma may receive further adjuvant chemotherapy. We have included postoperative CRT for R1 positive margins in the standard treatment algorithm and although its use remains contentious in the UK it is commonly used in many countries and certainly included in the National Comprehensive Cancer Network

recommendations and guidelines [43]. In addition, the US Intergroup trial reported long-term survival in only those patients with R1 disease who had received postoperative CRT [44]. The trial design is illustrated in Figure 1. The optimal CRT regimen remains open for discussion. Most of the randomised evidence on the role of neoadjuvant CRT relates to the use of cisplatin and 5-FU-based regimens. However, more robust outcomes have been observed with CRT protocols using novel triplet regimens, which have included higher PCR rates and also improved long-term survival. Van de Schoot et al. [45] reported on a trial of 50 patients with stage II or III oesophageal cancer (adenocarcinoma ¼ 42) who were treated with paclitaxel (175 mg/m2 on days 1 and 22), carboplatin (AUC 5 on days 1 and 22) and 5-FU (200 mg/m2 daily, days 1e42) concurrent with radiotherapy (45 Gy) followed by surgery. The regimen was well tolerated and 84% of patients completed treatment as planned. Forty-five (96%) patients underwent complete R0 resection with a postoperative mortality rate of 8.5%. The PCR rate was 38% with estimated 3 and 5 year survival rates of 56 and 48%, respectively. The estimated 3 year survival in responders was 61% compared with 33% in non-responders. Spigel et al. [46] recently reported on a phase I/II study involving 59 patients with resectable tumours of the oesophagus and GOJ (adenocarcinoma ¼ 69%; squamous cell carcinoma ¼ 18%) who received weekly oxaliplatin

Fig 1. Schematic illustration of the proposed design for the randomised controlled trial. Patients are divided into two groups based on the separation of lower oesophageal and gastro-oesophageal junction adenocarcinomas from tumours at other sites and squamous cell cancers. **Patients with R1 resection in the chemotherapy arm should be considered for postoperative chemoradiotherapy. PFS, progression-free survival; OS, overall survival; QOL, quality of life.

M. Hingorani et al. / Clinical Oncology 23 (2011) 696e705

(40 mg/m2), docetaxel (20 mg/m2) and capecitabine (1000 mg/m2 orally twice daily on days 1e7, 15e21 and 29e35) and concurrent radiotherapy (45 Gy) [46]. There were no dose-limiting toxicities in the initial cohort of 10 patients. In the subsequent phase II cohort of 49 patients, planned interruptions of chemotherapy drugs occurred in about 50% of patients. Nearly all (98%) patients received the intended schedule of radiation. The most commonly observed grade 3e4 toxicities were oesophagitis (20%), dehydration (16%), fatigue (12%) and nausea (16%). Sixtynine per cent of patients proceeded to surgery and the postoperative mortality was 8.5%. Based on intention to treat, the PCR rate was 49%; 71% if only the surgical patients were included. The median PFS and overall survival were 16.3 and 24.1 months, respectively. The 2 year PFS and overall survival were 45.1 and 52.2%, respectively. Despite the excellent pathological outcomes there remain appropriate concerns about the risk of excessive toxicity from the above triplet regimens. In addition, the postoperative mortality of 8.5% does not compare favourably with other CRT regimes. In our opinion, the CALGB regimen of 50.4 Gy in 28 fractions with concurrent cisplatin and 5-FU (PCR of 40%) or the CROSS regimen of 41.4 Gy in 23 fractions with concurrent docetaxel and carboplatin (PCR of 32.6%) would be appropriate options to consider in the CRT arm of the proposed study. The effect of the underlying histological subtype in determining the response to neoadjuvant therapy remains controversial. Based on the understanding that squamous cell carcinoma tumours are locally infiltrative, whereas adenocarcinoma primarily recurs distally with metastatic progression, it was proposed that neoadjuvant CRT should be the treatment of choice for squamous cell carcinoma and chemotherapy may suffice for adenocarcinoma [47]. However, this seems to be far more complex and at least two trials have failed to show any benefit from CRT followed by surgery in patients with squamous cell carcinoma of the oesophagus [11,14]. The EORTC trial of locally advanced squamous cell carcinoma tumours alone was negative for survival benefit, but this may have been influenced by the high postoperative mortality rates [11]. The limited benefit of neoadjuvant CRT in patients with squamous cell carcinoma of the oesophagus was also highlighted in the FFCD 9102 study, which evaluated the effect of surgery or further consolidation chemoradiation in patients responding to the initial induction CRT. The study reported a 3 month mortality rate of 9.3% in the combined arm compared with 0.8% in the CRT arm, and there was no overall survival difference in the two arms [48]. Another trial based on a similar design reported higher rates of local failure after CRT, but there was still no difference in overall survival [49]. Furthermore, a previous meta-analysis failed to show a significant benefit from neoadjuvant chemotherapy in patients with squamous cell carcinoma tumours [42]. Based on the above results several authorities are of the opinion that multi-modality therapy may have a limited role in the management of patients with oesophageal squamous cell carcinoma [50,51]. However, the situation remains far from clear and

703

the recent CROSS trial reported a significant reduction in observed hazard ratios for patients with squamous cell carcinoma tumours after CRT [25]. Although the data for the outcomes of patients with oesophageal squamous cell carcinoma support the case for stratification of such patients in prospective trials, we do not feel there is enough evidence to exclude these patients from any potential neoadjuvant CRT trial. The primary end points for the trial should naturally be tumour regression and PCR incidence, PFS and overall survival, with secondary end points of surgical morbidity and mortality and also overall quality of life. One of the criticisms about the previous trials in oesophageal cancer has been the relative lack of comprehensive staging protocols, which may have underestimated the beneficial role of neoadjuvant treatments because of the inclusion of patients with metastatic disease. Future trials for oesophageal and GOJ cancer should routinely incorporate these investigations (e.g. endoscopic ultrasound and positron emission tomography) within their respective staging protocols.

Conclusion The management of oesophageal cancer with combined modality neoadjuvant strategies is complex and the available evidence conflicting. We have discussed some of these controversies and recommend attempting to resolve them within the context of a well-designed randomised controlled trial. We have made initial recommendations for the trial design, but this remains open for discussion and scrutiny. The data emerging from novel neoadjuvant CRT regimens is exciting, but needs further investigation prior to their routine incorporation in the treatment protocols. Most importantly, in the UK we have to first agree on the benefits of neoadjuvant CRT before working towards finding the optimum regimen.

References [1] Daly JM, Karnell LH, Menck HR. National cancer database report on oesophageal carcinoma. Cancer 1996;78: 1820. [2] Kelsen DP, Ginsberg R, Pajak TF, et al. Chemotherapy followed by surgery compared with surgery alone for localized oesophageal cancer. N Engl J Med 1998;339:1979. [3] Hulscher JB, van Sandick JW, de Boer AG, et al. Extended transthoracic resection compared with limited transhiatal resection for adenocarcinoma of the oesophagus. N Engl J Med 2002;347:1662. [4] Altorki N, Kent M, Ferrara C, Port J. Three-field lymph node dissection for squamous cell and adenocarcinoma of the oesophagus. Ann Surg 2002;236:177. [5] Orringer MB, Marshall B, Iannettoni MD. Transhiatal esophagectomy: clinical experience and refinements. Ann Surg 1999;230:392. [6] Berger B, Belka C. Evidence-based radiation oncology: oesophagus. Radiother Oncol 2009;92:276e290. [7] Apinop C, Puttisak P, Preecha N. A prospective study of combined therapy in esophageal cancer. Hepatogastroenterology 1994;41:391e393.

704

M. Hingorani et al. / Clinical Oncology 23 (2011) 696e705

[8] Walsh TN, Noonan N, Hollywood D, Kelly A, Keeling N, Hennessy TP. A comparison of multimodal therapy and surgery for oesophageal adenocarcinoma. N Engl J Med 1996; 335:462e467. [9] DeMeester SR. Adenocarcinoma of the oesophagus and cardia: a review of the disease and its treatment. Ann Surg Oncol 2006;13:12e30. [10] Funk EM, Witte J. Multimodal therapy for oesophageal adenocarcinoma. N Engl J Med 1997;336:375. [11] Bosset JF, Gignoux M, Triboulet JP, et al. Chemoradiotherapy followed by surgery compared with surgery alone in squamous-cell cancer of the oesophagus. N Engl J Med 1997; 337:161e167. [12] Urba SG, Orringer MB, Turrisi A, et al. Randomized trial of preoperative chemoradiation versus surgery alone in patients with locoregional oesophageal carcinoma. J Clin Oncol 2001;19:305. [13] Kleinberg L, Forastiere A. Chemoradiation in the management of oesophageal cancer. J Clin Oncol 2007;25:4110e4117. [14] Lee JL, Park SI, Kim SB, et al. A single institutional phase III trial of preoperative chemotherapy with hyperfractionation radiotherapy plus surgery versus surgery alone for resectable esophageal squamous cell carcinoma. Ann Oncol 2004;15:947e954. [15] Burmeister BH, Smithers BM, Gebski V, et al. Trans-Tasman Radiation Oncology Group; Australasian Gastro-Intestinal Trials Group. Surgery alone versus chemoradiotherapy followed by surgery for resectable cancer of the oesophagus: a randomised controlled phase III trial. Lancet Oncol 2005;6:659e668. [16] Carstens H, Albertsson M, Friesland S, et al. A randomized trial of chemoradiotherapy versus surgery alone in patients with resectable oesophageal cancer (abstract). J Clin Oncol 2007;25:205. [17] Tepper J, Krasna MJ, Niedzwiecki D, et al. Phase III trial of trimodality therapy with cisplatin, fluorouracil, radiotherapy, and surgery compared with surgery alone for esophageal cancer: CALGB 9781. J Clin Oncol 2008;26(7):1086e1092. [18] Kaklamanos IG, Walker GR, Ferry K, Franceschi D, Livingstone AS. Neoadjuvant treatment for resectable cancer of the esophagus and the gastroesophageal junction: a meta-analysis of randomized clinical trials. Ann Surg Oncol 2003; 10:754e761. [19] Fiorica F, Di Bona D, Schepis F, et al. Preoperative chemoradiotherapy for oesophageal cancer: a systematic review and meta-analysis. Gut 2004;53:925e930. [20] Urschel JD, Vasan H. A meta-analysis of randomized controlled trials that compared neoadjuvant chemoradiation and surgery to surgery alone for resectable oesophageal cancer. Am J Surg 2003;185:538. [21] Malthaner RA, Wong RK, Rumble RB, Zuraw L. Members of the Gastrointestinal Cancer Disease Site Group of Cancer Care Ontario’s Program in Evidence-based Care. Neo-adjuvant or adjuvant therapy for resectable oesophageal cancer: a systematic review and meta-analysis. BMC Med 2004;2:35. [22] Jin Hai-Lin, Zhu Hong, Ling Ting-Sheng, Zhang Hong-Jie, Shi Rui-Hua. Neo-adjuvant chemoradiotherapy for resectable oesophageal carcinoma: a meta-analysis. World J Gastroenterol 2009;15(47):5983e5991. [23] Courrech Staal EF, Aleman BM, Boot H, van Velthuysen ML, van Tinteren H, van Sandick JW. Systematic review of the benefits and risks of neoadjuvant chemoradiation for oesophageal cancer. Br J Surg 2010;97(10):1482e1496. [24] van Heijl M, van Lanschot JJ, Koppert LB, et al. Neo-adjuvant chemoradiation followed by surgery versus surgery alone for patients with adenocarcinoma or squamous cell carcinoma of the oesophagus (CROSS). BMC Surg 2008;8:21.

[25] Gaast AV, van hagen P, Hulshof M, et al. Effect of preoperative concurrent chemoradiotherapy on survival of patients with resectable esophageal or esophagogastric junction cancer: results from a multicenter randomized phase III study. J Clin Oncol 2010;28, 15s;abstract 4004. [26] Maipang T, Vasinanukorn P, Petpichetchian C, et al. Induction chemotherapy in the treatment of patients with carcinoma of the esophagus. J Surg Oncol 1994;56:191e197. [27] Law S, Fok M, Chow S, Chu KM, Wong J. Preoperative chemotherapy versus surgical therapy alone for squamous cell carcinoma of the esophagus: a prospective randomized trial. J Thorac Cardiovasc Surg 1997;114:210e217. [28] Ancona E, Ruol A, Santi S, et al. Only pathologic complete response to neoadjuvant chemotherapy improves significantly the long term survival of patients with resectable esophageal squamous cell carcinoma: final report of a randomized, controlled trial of preoperative chemotherapy versus surgery alone. Cancer 2001;91:2165e2174. [29] Medical Research Council Oesophageal Cancer Working Group. Surgical resection with or without preoperative chemotherapy in oesophageal cancer: a randomised controlled trial. Lancet 2002;359:1727. [30] Allum WH, Stenning SP, Bancewicz J, et al. Long-term results of a randomized trial of surgery with or without preoperative chemotherapy in oesophageal cancer. J Clin Oncol 2009;27:5062. [31] Cunningham D, Allum WH, Stenning SP, et al. MAGIC Trial Participants. Perioperative chemotherapy versus surgery alone for resectable gastroesophageal cancer. N Engl J Med 2006;355:11e20. [32] Boige V, Pignon J, Saint-Aubert B, et al. Final results of a randomized trial comparing preoperative 5-fluorouracil/ cisplatin to surgery alone in adenocarcinoma of stomach and lower oesophagus (ASLE): FNLCC ACCORD07-FFCD 9703 trial (abstract). J Clin Oncol 2007;25:200. [33] Schuhmacher C, Schlag P, Lordick F, et al. Neo-adjuvant chemotherapy versus surgery alone for locally advanced adenocarcinoma of the stomach and cardia: randomized EORTC phase III trial #40954. (Abstract #4510). J Clin Oncol 2009;27:204. [34] Urschel JD, Vasan H, Blewett CJ. A meta-analysis of randomized controlled trials that compared neo-adjuvant chemotherapy and surgery to surgery alone for resectable oesophageal cancer. Am J Surg 2002;183:274e279. [35] Graham AJ, Shrive FM, Ghali WA, et al. Defining the optimal treatment of locally advanced esophageal cancer: a systematic review and decision analysis. Ann Thorac Surg 2007;83: 1257e1264. [36] Gebski V, Burmeister B, Smithers BM, Foo K, Zalcberg J, Simes J, Australasian Gastro-Intestinal Trials Group. Survival benefits from neo-adjuvant chemoradiotherapy or chemotherapy in esophageal cancer e a meta-analysis. Lancet Oncol 2007;8:226e234. [37] Stahl M, Walz MK, Stuschke M, et al. Phase III comparison of preoperative chemotherapy compared with chemoradiotherapy in patients with locally advanced adenocarcinoma of the esophagogastric junction. J Clin Oncol 2009;27:851. [38] Luu TD, Gaur P, Force SD, et al. Neo-adjuvant chemoradiation versus chemotherapy for patients undergoing esophagectomy for oesophageal cancer. Ann Thorac Surg 2008;85(4): 1217e1223. discussion 1223e1224. [39] Crosby TD, Evans M. The management of locally advanced carcinoma of the oesophagus d SCOPE for improvement? Clin Oncol (R Coll Radiol) 2009;21(4):281e284. [40] Malthaner R, Wong RK, Spithoff K. Gastrointestinal Cancer Disease Site Group of Cancer Care Ontario’s Program in

M. Hingorani et al. / Clinical Oncology 23 (2011) 696e705

[41]

[42]

[43]

[44]

[45]

Evidence-based Care. Preoperative or postoperative therapy for resectable oesophageal cancer: an updated practice guideline. Clin Oncol (R Coll Radiol) 2010;22(4):250e256. Berger AC, Farma J, Scott WJ, et al. Complete response to neo-adjuvant chemoradiotherapy for resectable oesophageal carcinoma: a meta-analysis. J Clin Oncol 2005;23:4330e4337. Rizk NP, Venkatraman E, Bains MS, et al. American Joint Committee on Cancer staging system does not accurately predict survival in patients receiving multimodality therapy for oesophageal adenocarcinoma. J Clin Oncol 2007;25:507. NCCN Clinical Practice Guidelines in Oncology. Oesophageal Cancer. Available at: http://www.nccn.org/professionals/ physician_gls/PDF/esophageal.pdf; 2010. V1. Kelsen DP, Winter KA, Gunderson LL, et al. Long-term results of RTOG trial 8911 (USA Intergroup 113): a random assignment trial comparison of chemotherapy followed by surgery compared with surgery alone for oesophageal cancer. J Clin Oncol 2007;25:3719. van de Schoot L, Romme EA, van der Sangen MJ, et al. A highly active and tolerable neoadjuvant regimen combining paclitaxel, carboplatin, 5-FU, and radiation

[46]

[47] [48]

[49]

[50]

[51]

705

therapy in patients with stage II and III oesophageal cancer. Ann Surg Oncol 2008;15:88. Spigel DR, Greco FA, Meluch AA, et al. Phase I/II trial of preoperative oxaliplatin, docetaxel, and capecitabine with concurrent radiation therapy in localized carcinoma of the oesophagus or gastroesophageal junction. J Clin Oncol 2010; 28(13):2213e2219. Siewert JR, Ott K. Are squamous and adenocarcinomas of the oesophagus the same disease? Semin Radiat Oncol 2007;17:38.  O, et al. Chemoradiation folBedenne L, Michel P, Bouche lowed by surgery compared with chemoradiation alone in squamous cancer of the esophagus: FFCD 9102. J Clin Oncol 2007;25(10):1160e1168. Stahl M, Stuschke M, Lehmann N, et al. Chemoradiation with and without surgery in patients with locally advanced squamous cell carcinoma of the oesophagus. J Clin Oncol 2005;23:2310. Hyngstrom JR, Posner MC. Neo-adjuvant strategies for the treatment of locally advanced oesophageal cancer. J Surg Oncol 2010;101(4):299e304. Barnett SA, Rizk NP. Randomized clinical trials in oesophageal carcinoma. Surg Oncol Clin North Am 2010;19(1):59e80.