Evaluation of progression prior to surgery after neoadjuvant chemoradiotherapy with computed tomography in esophageal cancer patients

The American Journal of Surgery (2014) -, -–-

Evaluation of progression prior to surgery after neoadjuvant chemoradiotherapy with computed tomography in esophageal cancer patients Jan B. Hulshoff, B.Sc.a,1, Justin K. Smit, M.D., Ph.D.a, Eric J. van der Jagt, M.D., Ph.D.b, John T. Plukker, M.D., Ph.D.a,* Departments of aSurgical Oncology, bRadiology, University Medical Center Groningen, Groningen, The Netherlands KEYWORDS: Computed tomography; Progression esophageal cancer; Neoadjuvant CRT

Abstract BACKGROUND: The risk of tumor progression during neoadjuvant chemoradiotherapy (CRT) in esophageal cancer (EC) is around 8% to 17%. We assessed the efficacy of computed tomography (CT) to identify these patients before esophagectomy. METHODS: Ninety-seven patients with locally advanced EC treated with Carboplatin/Paclitaxel and 41.4 Gy neoadjuvantly were restaged with CT. Two radiologists reviewed pre- and post-CRT CT images. The primary outcome was detection of clinically relevant progressive disease. Missed metastases were defined as metastatic disease found during surgery or within 3 months after post-CRT CT. RESULTS: Progressive disease was detected in 9 patients (9%). Both radiologists detected 5 patients with distant metastases (liver, n 5 4; lung metastasis, n 5 1), but missed progressive disease in 4 cases. One radiologist falsely assessed 2 metastatic lesions, but after agreement progressive disease was detected with sensitivity and specificity of 56% and 100%, respectively. CONCLUSION: CT is effective in detecting clinically relevant progressive disease in EC patients, after neoadjuvant treatment. Ó 2014 Elsevier Inc. All rights reserved.

Esophageal cancer (EC) is now the 8th most commonly diagnosed malignancy worldwide.1 Currently, the primary curative treatment for EC consists of neoadjuvant chemoradiotherapy (CRT) followed by surgery. However, the prognosis of patients with locally advanced EC remains poor even after subsequent curative-intended resection, with an overall survival rate of 35% to 38%.2,3 There were no relevant financial relationships or any sources of support in the form of grants, equipment, or drugs. The authors declare no conflicts of interest. * Corresponding author. Tel.: 131-503612317; fax: 131-503614873. E-mail address: [email protected] Manuscript received May 29, 2013; revised manuscript August 25, 2013 1 Control over data. 0002-9610/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2013.08.048

A recently performed meta-analysis and outcome of the latest published randomized controlled Chemoradiotherapy for Oesophageal Cancer Followed by Surgery Study trial underline the significant positive effect of neoadjuvant CRT on survival and local recurrence rate.4,5 Even then, only approximately 50% of the patients benefit from neoadjuvant CRT, while 8% to 17% of the patients develop distant metastases in the time interval between CRT and surgery.6–9 Surgery in these patients is futile and should be avoided. Instead, maintenance of quality of life with palliative support on guidance of complaints should be the principal treatment. Guidelines regarding restaging after neoadjuvant CRT are lacking. Conventional staging methods, including endoscopic ultrasonography (EUS) and computed tomography (CT), are based on disturbance of anatomical

2 structures. Limitations of these techniques include difficulty with distinguishing vital tumor tissue from necrotic or fibrotic tissue and the delay between cell death and tumor shrinkage.10 On the other hand, CT has a low false positivity with a high availability rate and is associated with relatively low costs. Furthermore, the spatial resolution of CT has improved over the last decennium. Detection of viable tumor tissue with 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) is based on increased metabolic activity of the tumor, leading to a high FDG uptake.11 Integrated PET-CT scans combine the metabolic information from the PET with the anatomical location found with the CT scan. Three studies showed a better detection of distant metastases with integrated PET-CT (3 out of 4 patients) compared with CT alone (2 out of 4 patients).6–8 PET-CT, however, also has a high false positivity rate, which has been reported to be as high as 8%.8 In our esophageal center, all patients received a postCRT CT to assess whether the patient has developed progressive disease during the time interval. In this study, we hypothesize that CT is effective in detecting distant metastasis and clinically relevant lymph node progression to restrict futile surgery after neoadjuvant CRT.

Patients and Methods Patients Between 2006 and 2011, 116 patients with potentially resectable locally advanced EC who received a restaging CT within 3 weeks after neoadjuvant CRT were eligible for inclusion. In total, 19 patients were excluded because of missing post-CRT data (n 5 7), missing pre-CRT CT data (n 5 1), treatment in different medical centers (n 5 3), a premature post-CRT CT (n 5 6) made during the neoadjuvant treatment, and clinically proven progressive disease during CRT treatment (n 5 2). Finally, 97 of the 116 patients could be included in this study. This study was performed in accordance with the guidelines of the national and our own Ethical Board. We created a database containing patient and tumor characteristics, treatment-related data and follow-up data. Patient characteristics are summarized in Table 1. The male/female ratio is 3.3:1 with a median age of 65 years (interquartile range 60 to 70 years) and tumors staged clinically from stage IIa (T2N0M0) to stage IIIc (T4aN1-2M0/ T1-4aN3M0) at the time of diagnosis.

The American Journal of Surgery, Vol -, No -, - 2014 Table 1

Patients, tumor, and treatment characteristics Number of patients

Male 75 (77%) Age (y), median (IQR) 65 (60–70) BMI, median (IQR) 25.6 (23–28.4) Histology Adenocarcinoma 80 (82%) Squamous cell carcinoma 16 (17%) Adenosqaumous cell carcinoma 1 (1%) Tumor location Middle esophagus 5 (5%) Distal esophagus 92 (95%) Initial stage IIa (T2N0M0) 4 (4%) IIb (T3N0M0, T1-2N1M0) 20 (21%) IIIa (T1-2N2M0, T3N1M0, T4aN0M0) 38 (39%) IIIb (T3N2M0) 20 (21%) IIIc (T4aN1-2M0, Tall,N3M0) 15 (15%) Complete CRT 85 (88%) Time in weeks, median (IQR) Pathology and surgery 17 (15–19) Pre- and post-CRT CT 14 (12–15) End neoadjuvant CRT and surgery 6 (5–8) Surgery (including diagnostic laparotomy) 92 (95%) Complete resection 85 (88%) Stage after neoadjuvant CRT 0 (pCR, TisN0M0) 15 (15%) Ia (T1N0M0) 11 (11%) Ib (T1-2N0M0) 2 (2%) IIa (T2N0M0) 9 (9%) IIb (T3N0M0, T0-2N1M0) 31 (32%) IIIa (T1-2N2M0, T3N1M0, T4aN0M0) 11 (11%) IIIb (T3N2M0) 5 (5%) IIIc (T4aN1-2M0, T4bNallM0, TallN3M0) 4 (4%) IV (TallNallM1) 9 (9%) Progressive disease between CRT and 9 (9%) surgery* Progression within 3 months after surgery 0 (0%) Follow-up in weeks (IQR) 36 (20–52) CRT 5 chemoradiotherapy; CT 5 computed tomography; IQR 5 interquartile range; pCR 5 pathologically complete response. *Pathologically or clinically proven metastatic disease.

meeting participated by at least one of the two experienced radiologist, an oncologist, a radiotherapeutic oncologist, a nuclear medicine physician, pathologist, a gastroenterologist, and a surgical oncologist involved in the treatment of EC patients. All participants of this meeting agreed that the patients included in this study should be treated with neoadjuvant CRT.

Initial staging procedure Initial staging consisted of EUS with fine needle aspiration on indication and a CT of the chest and abdomen. A full-body FDG-PET or PET/CT was employed in all T2-T4a and/or nodal-positive patients (N1). The outcomes of all staging methods were discussed in a multi-disciplinary

Restaging preoperative post-chemoradiotherapy computed tomography All patients received both pre- and post-CRT 64-multisliced CT scan (PhilipsMX 8000, Philips Medical Systems, Best,

J.B. Hulshoff et al.

Detection of esophageal cancer progression with CT after neoadjuvant CRT

The Netherlands; Siemens Somatom Sensation, Siemens, Erlanger, Germany or GE medical systems, Milwaukee, WI), according to the EC protocol used in our center. The CT includes the cervical, thoracic, and abdominal region and was performed with adequate oral and intravenous contrast. PreCRT CT images with a maximum slice thickness of 5 mm from other medical centers were allowed. All included patients had their preoperative post-CRT CT within at least 3 to 4 weeks after the end of the neoadjuvant treatment. All post-CRT CT scans were performed in our center, with a Somatom Sensation 64 CT (Siemens, Erlanger, Germany) with slices of 2/1.5 mm. In case of suspicious progressive disease, additional examinations (either PET, magnetic resonance imaging, ultrasonography, cytology, or a combination) were carried out to strengthen or prove progressive disease.

3

Treatment Neoadjuvant treatment consisted of intravenous Paclitaxel (50 mg/m2) and Carboplatin (area under the curve 5 2), administered 5 times during a 4-week radiotherapy period with a total dose of 41.4 Gy given in 23 fractions of 1.8 Gy.5 As standard surgical procedure, we performed a radical transthoracic esophagectomy through a left or right thoracolaparotomy combined with a 2-field lymphadenectomy of mediastinal and abdominal lymph nodes, including nodes at the celiac trunk and those along the common hepatic and splenic artery at the upper border of the pancreas.

Pathology Reassessment of progressive disease Two experienced radiologists, including a senior radiologist specialized in gastroesophageal cancer, reviewed both the pre- and post-CRT CT images and restaged these images according to the 7th tumor node metastasis (TNM) edition.12 Both radiologists were blinded to the outcome of the CT and further treatments. The primary outcome was treatment change because of progressive disease (yes or no) detected by comparing the pre- and postCRT CT. Progressive disease was defined as any radiological visible increase in tumor volume, including both nodal progression and distant metastasis. Locoregional progression without signs of invasion in surrounding organs was not included as clinically relevant progressive disease in this study, because it would not alter the course of surgical treatment. As a reference standard, we used either the cytological or histological confirmation of progressive disease detected before and during surgery, or based on pathological examination of the resected specimens. In addition, we considered all suspect lesions that occurred within 3 months after the post-CRT CT as missed with CT. All patients received a PET/CT scan 3 months after surgical treatment for other research purposes, which was also used to detect progressive disease. Patients suspected of progressive disease were discussed in a multidisciplinary board and received additional examinations, such as magnetic resonance imaging, ultrasonography, or EUS-guided fine needle aspiration on indication. In defining true progressive disease, we included all cytologically/histhologically proven progressive disease detected before or during surgery and pathological proven or clinical evidence of recurrent/persistent disease found within 3 months after the post neoadjuvant CRT CT. We determined the sensitivity, specificity, negative predictive value, and positive predictive value of both radiologists in detecting progressive disease by comparing the pre- and post-CRT CT. Patients assessed differently by the 2 radiologists were re-evaluated until agreement was obtained.

Pathological processing of all resected specimens and lymph nodes has been standardized according to a detailed protocol on radicality, including the proximal, the distal, and the circumferential margin. The extension of the primary tumor, localization, number of nodal involvement, and L/N ratio was accurately described.

Statistical analysis SPSS Version 17 (SPSS, Inc, Chicago, IL) was used to perform statistical analyses. The distribution of all patient characteristics was assessed with a QQ- and PP-plot. Normally distributed variables were reported as mean 6 standard deviation, non-normally distributed variables as median [interquartile range], and categorical variables were reported as amount [percentage].

Results In Table 1, the patient, tumor, and treatment-related characteristics are summarized. Of the 97 patients, 9 (9%) patients had progressive disease after neoadjuvant CRT. The initial post-CRT CT prevented futile surgery in 5 (5/97; 5%) patients: 4 patients developed liver metastases and 1 patient developed lung metastases. Three patients had progressive disease identified during surgery: 1 patient had a previous undetected peritoneal metastasis, 1 patient had a metastatic lesion on the top of the bladder, and 1 patient had both liver and omental metastases. In addition, cervical lymph node metastasis was observed in 1 patient with PET-CT performed because of a suspect vertebral lesion which was assessed as benign on PET-CT. More information on these lesions and the imaging techniques used to strengthen or prove metastatic disease is revealed in Table 2. Pathologic proof of progressive disease was obtained in 6 patients. There were no patients with a metastasis found within 3 months after the postCRT CT.

The American Journal of Surgery, Vol -, No -, - 2014

4 Table 2

Characteristics of patients with progressive disease relevant for further treatment

cTNM

Histological tumor type

Tumor location

Metastasis location

Initially detected with CT

Additional techniques

T4aN1 T3N1 T3N0 T3N1 T3N1 T3N2

Adeno Adeno Adeno Adeno Sqaumous Adeno

Distal Distal Distal Distal Distal Distal

Liver Liver Liver Liver Lung Nodal,* Liver, Omental

PET Biopsy Biopsy

T3N1 T3N1 T4aN1

Adeno Adeno Adeno

Distal Distal Distal

Bladder top Peritoneal Cervical lymph node

Yes Yes Yes Yes Yes Yes No No No No No

No

Surgery No No No No No Yes

Yes Yes PET 1 Biopsy

Pathologically proven No Yes Yes No No No Yes Yes Yes Yes Yes

cTNM 5 clinical tumor node metastasis; CT 5 computed tomography; PET 5 positron emission tomography. *Resectable lymph node progression.

Efficacy in detecting progressive disease Both radiologists detected 4 patients with liver metastases and 1 patient with lung metastases. CT images are depicted in Fig. 1. Progressive disease was missed by the radiologists in 4 patients: 1st patient with a bladder top metastasis, the 2nd had a peritoneal metastasis, the 3rd had a cervical lymph node metastasis, and the 4th had both liver and omental metastases. During reassessment of the CT images of patients with missed progressive disease, the radiologists agreed that the missed cervical node metastasis, which was detected on PET/CT, should have been found on the post-CRT CT. However, because both radiologists missed this cervical node involvement, it was scored as missed progressive disease. Fig. 2 depicts both the CT and PET-CT images in this patient. Resectable nodal progression was detected by both radiologists in 2 patients, but this was not relevant for subsequent treatment. However, one of these patients also had a liver and omental metastasis identified during surgery. Radiologist 1 falsely suggested 2 metastases – a lung and vertebral metastasis (Fig. 3) – identifying progressive disease with a sensitivity and specificity of 56% and 98% and an negative predictive value (NPV) and Positive Predictive Value (PPV) of 96% and 71%, respectively. Radiologist 2 correctly assessed the vertebral lesion as nonspecific based on the osteoblastic character, identifying progressive disease with a sensitivity and specificity of 58% and 100% and an NPV and PPV of 96% and 100%, respectively. After the differently assessed CT images were discussed, the radiologists agreed that the vertebral lesion was nonspecific for metastasis and the lung lesion was benign. After agreement, both radiologists detected clinically relevant progressive disease, with a sensitivity and specificity of 56% and 100% and an NPV and PPV of 96% and 100%, respectively. Retrospectively, with knowledge of the metastases and its location, a suspicious hepatic lesion of approximately

1 cm could be seen in the patients with both liver and omental metastases. Even after re-evaluation, the CTmissed omental metastases could not be detected. The CT-missed bladder top metastasis was located outside of the range of the abdominal CT.

Comments In this study, CT after neoadjuvant CRT (post-CRT CT) showed to be effective in the detection of clinically relevant progressive disease which leads to a change in treatment in these EC patients. Detection of progressive disease before surgery on post-CRT CT prevented futile surgery in 5 (5%) of these patients, but missed progressive disease in 4 (4%) patients. Detection of progressive disease that alters the initial treatment strategy is of great importance, as the majority of these patients is beyond curative treatment and should be refrained from surgery. In this study, we used only small-slice (2/1.5 mm) postCRT CT scans for adequate radiological assessment of suspect lymph nodes and/or metastatic lesions by 2 radiologists. Most previously published studies used thicker slice (5/5 mm) CT scans to detect tumor progression. The study by Bruzzi et al6 had distant metastases as primary outcome, comparing CTwith PET-CT in the detection of distant metastasis after neoadjuvant CRT. With PET-CT, it was possible to detect distant metastasis in 7 patients (8%) after neoadjuvant CRT, while CT alone detected distant metastasis in 5 (6%) of these patients. Both missed metastases were located outside the range of the routinely performed CT imaging of the chest and abdomen. This was also observed in our study with a bladder top metastasis that was located outside of the range of routine CT imaging. Cerfolio et al8 compared CT with EUS and PET-CT in restaging patients with EC after neoadjuvant CRT. The additional value of a PET-CT was limited; only one additional metastasis was found and both PET-CT and CT still

J.B. Hulshoff et al.

Detection of esophageal cancer progression with CT after neoadjuvant CRT

5

Figure 1 CT images of the patients with progressive disease detected by both radiologists. (A) Pre-CRT CT and (B) post-CRT CT. Images 1 to 4 are of patients with liver metastases, and image 5 is of a patient with lung metastasis.

missed an omental metastasis. PET-CT also falsely suggested one more distant metastasis.

A more recent study, Blom et al,7 compared CT with PET-CT in detecting distant metastases after neoadjuvant

6

The American Journal of Surgery, Vol -, No -, - 2014

Figure 2 Images of the patient with cervical lymph node progression after neoadjuvant CRT that was missed by both radiologists. (A) Pre-CRT CT, (B) post-CRT CT, and (C) PET-CT was made to assess the suspect vertebral lesions in this patient.

CRT. With PET-CT, 4 out of 5 patients with a distant interval metastasis were identified, while CT alone detected only 2 of these lesions. One of the metastases that were missed with CT was located in the iliac crest, which is usually located outside the range of the CT. The other metastasis missed with CT was located in the right scapula. Both CT and PET-CT missed 1 patient with peritonitis carcinomatosa and PET-CT falsely suggested a pulmonary metastasis. One of the reasons most studies advise a PET-CT is that EC tends to metastasize to uncommon and unusual distant locations after CRT, such as skeletal muscles, brain, peritoneum, subcutaneous soft tissue, pleura, pancreas, and thyroid gland.13 These distant metastases can be missed with CT, because only thoracic and abdominal CT images are made. PET scans, in contrast, are full-body scans. However, most metastases after neoadjuvant CRT were located in either the liver or the lungs.6,7,14 In our study, 6 out of 9 metastases were located in either the liver or the lungs, of which CT detected 5 metastases. Most of

the metastases located in the liver and the lungs can be found with CT.15 In one of the 4 patients with missed tumor progression on CT in this study, re-assessment of a PET-CT for suspected vertebral metastases showed progression of cervical lymph node, which was initially not been detected on post-CRT CT. In addition, after postoperative reexamination of the CT, a suspect lesion could be detected in the liver of the patient with both liver and omental metastases. Omental metastases are generally hardly detected with CT and the missed peritoneal metastases in this study could not be detected even after revision. It is known that CT is inaccurate in detecting peritoneal metastases of ,1 cm.16,17 However, the additional value of PET/CT is also limited in depicting small-volume metastatic lesions of 5 to 7 mm.18 Currently, accurate detection of peritoneal metastasis is only possible by surgical intervention, including laparoscopy or laparotomy.19,20 In our center, we therefore start with a laparotomy at the first phase of the esophagectomy followed by subsequent curative resection through a thoracotomy.

Figure 3 CT images of the patient falsely assessed by radiologist 1 as patient with progressive disease. (A) Pre-CRT CT and (B) post-CRT CT. Image 1 was falsely assessed as lung metastasis and image 2 was falsely assessed as vertebral metastasis.

J.B. Hulshoff et al.

Detection of esophageal cancer progression with CT after neoadjuvant CRT

One of the disadvantages of this retrospective study is that we did not compare CT with PET-CT. Future research should compare a thin-slice thickness CT with PET-CT, in the detection of patients with progressive disease after neoadjuvant CRT. According to the retrospective design of our study, we were unable to obtain pathological prove of metastatic disease in 3 patients with distant metastasis. Because of small numbers it was not possible to accurately determine the interobserver variability between both, a skilled and a specialized radiologist on EC in reviewing CT images in this study. The different findings between them, 1 radiologist had 2 false-positive results, might be caused by a difference in experiences. The potentially extended interval caused by the neoadjuvant treatment and the rate of progression during neoadjuvant treatment emphasizes the importance of imaging after neoadjuvant CRT. Different studies assessed PET-CT to be more effective in the detection of distant metastasis. However, compared with CT, PET-CT has a lower availability, is more expensive, and has a high amount of false-positive outcomes. CT is therefore a useful method to detect patients with clinically relevant progressive disease in EC patients treated with neoadjuvant CRT.

References 1. Jemal A, Bray F, Center MM, et al. Global cancer statistics. CA Cancer J Clin 2011;61:69–90. 2. Allum WH, Stenning SP, Bancewicz J, et al. Long-term results of a randomized trial of surgery with or without preoperative chemotherapy in esophageal cancer. J Clin Oncol 2009;27:5062–7. 3. Omloo JM, Lagarde SM, Hulscher JB, et al. Extended transthoracic resection compared with limited transhiatal resection for adenocarcinoma of the mid/distal esophagus: five-year survival of a randomized clinical trial. Ann Surg 2007;246:992–1000; discussion, 1000–1. 4. Sjoquist KM, Burmeister BH, Smithers BM, et al. Survival after neoadjuvant chemotherapy or chemoradiotherapy for resectable oesophageal carcinoma: an updated meta-analysis. Lancet Oncol 2011;12:681–92. 5. van Hagen P, Hulshof MC, van Lanschot JJ, et al. Preoperative chemoradiotherapy for esophageal or junctional cancer. N Engl J Med 2012; 366:2074–84. 6. Bruzzi JF, Swisher SG, Truong MT, et al. Detection of interval distant metastases: clinical utility of integrated CT-PET imaging in patients

7.

8.

9.

10.

11. 12.

13.

14.

15.

16.

17.

18.

19.

20.

7

with esophageal carcinoma after neoadjuvant therapy. Cancer 2007; 109:125–34. Blom RL, Schreurs WM, Belgers HJ, et al. The value of postneoadjuvant therapy PET-CT in the detection of interval metastases in esophageal carcinoma. Eur J Surg Oncol 2011;37:774–8. Cerfolio RJ, Bryant AS, Ohja B, et al. The accuracy of endoscopic ultrasonography with fine-needle aspiration, integrated positron emission tomography with computed tomography, and computed tomography in restaging patients with esophageal cancer after neoadjuvant chemoradiotherapy. J Thorac Cardiovasc Surg 2005;129: 1232–41. Flamen P, Van Cutsem E, Lerut A, et al. Positron emission tomography for assessment of the response to induction radiochemotherapy in locally advanced oesophageal cancer. Ann Oncol 2002;13: 361–8. Westerterp M, van Westreenen HL, Reitsma JB, et al. Esophageal cancer: CT, endoscopic US, and FDG PET for assessment of response to neoadjuvant therapy–systematic review. Radiology 2005;236:841–51. Bar-Shalom R, Valdivia AY, Blaufox MD. PET imaging in oncology. Semin Nucl Med 2000;30:150–85. Sobin LH, Compton CC. TNM seventh edition: what’s new, what’s changed: communication from the International Union Against Cancer and the American Joint Committee on Cancer. Cancer 2010;116: 5336–9. Bruzzi JF, Truong MT, Macapinlac H, et al. Integrated CT-PET imaging of esophageal cancer: unexpected and unusual distribution of distant organ metastases. Curr Probl Diagn Radiol 2007;36:21–9. Watt I, Stewart I, Anderson D, et al. Laparoscopy, ultrasound and computed tomography in cancer of the oesophagus and gastric cardia: a prospective comparison for detecting intra-abdominal metastases. Br J Surg 1989;76:1036–9. van Vliet EP, Steyerberg EW, Eijkemans MJ, et al. Detection of distant metastases in patients with oesophageal or gastric cardia cancer: a diagnostic decision analysis. Br J Cancer 2007;97:868–76. Metser U, Jones C, Jacks LM, et al. Identification and quantification of peritoneal metastases in patients with ovarian cancer with multidetector computed tomography: correlation with surgery and surgical outcome. Int J Gynecol Cancer 2011;21:1391–8. Wang Z, Chen JQ. Imaging in assessing hepatic and peritoneal metastases of gastric cancer: a systematic review. BMC Gastroenterol 2011; 11:19. Sironi S, Messa C, Mangili G, et al. Integrated FDG PET/CT in patients with persistent ovarian cancer: correlation with histologic findings. Radiology 2004;233:433–40. Bonavina L, Incarbone R, Lattuada E, et al. Preoperative laparoscopy in management of patients with carcinoma of the esophagus and of the esophagogastric junction. J Surg Oncol 1997;65:171–4. de Graaf GW, Ayantunde AA, Parsons SL, et al. The role of staging laparoscopy in oesophagogastric cancers. Eur J Surg Oncol 2007;33: 988–92.