The role of endoscopic ultrasound in staging oesophageal cancer

Transcription

1 The role of endoscopic ultrasound in staging oesophageal cancer Poster No.: C-1509 Congress: ECR 2010 Type: Educational Exhibit Topic: GI Tract Authors: J. K. Bell, D. Awad, D. Kasir, J. Vickers, L. Williams; Salford/UK Keywords: Oesophageal cancer, Cancer staging, Endoscopic ultrasonography DOI: /ecr2010/C-1509 Any information contained in this pdf file is automatically generated from digital material submitted to EPOS by third parties in the form of scientific presentations. References to any names, marks, products, or services of third parties or hypertext links to thirdparty sites or information are provided solely as a convenience to you and do not in any way constitute or imply ECR's endorsement, sponsorship or recommendation of the third party, information, product or service. ECR is not responsible for the content of these pages and does not make any representations regarding the content or accuracy of material in this file. As per copyright regulations, any unauthorised use of the material or parts thereof as well as commercial reproduction or multiple distribution by any traditional or electronically based reproduction/publication method ist strictly prohibited. You agree to defend, indemnify, and hold ECR harmless from and against any and all claims, damages, costs, and expenses, including attorneys' fees, arising from or related to your use of these pages. Please note: Links to movies, ppt slideshows and any other multimedia files are not available in the pdf version of presentations. Page 1 of 66

2 Learning objectives To explore the role and limitations of diagnostic endoscopic ultrasound (EUS) in the local staging of oesophageal cancer. To describe the various types of specialised endoscopes used. To discuss the role of EUS guided fine needle aspiration (FNA) and core biopsies. Description of the relative merits and limitations of the various competing imaging modalities, including MDCT and PET-CT. Page 2 of 66

3 Background Role of endoscopic ultrasound and competing modalities in staging oesophageal malignancy Various imaging modalities are utilised to stage oesophageal cancer. These include endoscopic ultrasound (EUS), MDCT and PET-CT. The main emphasis of this poster will be on the role and technique of EUS. Laparoscopy also has an important role in staging peritoneal involvement. Endoscopic Ultrasound The primary role of EUS in oesophageal cancer is local tumour and lymph node staging. It is of limited value in detecting distant metastases, although it may potentially detect lesions in the left lobe of the liver and adrenal lesions. It is also a sensitive tool in detecting Page 3 of 66

4 small volume ascites, which may reflect peritoneal disease. It therefore complements other cross-sectional modalities and laparoscopy in staging oesophageal and junctional tumours. EUS also allows accurate assessment of tumour length and position to allow radiotherapy planning. EUS is performed by endoscopists from different specialities, but all with a common interest in upper gastroinestinal tract pathology. In the United Kingdom they are usually from the background of Gastroenterology, Surgery or Gastrointestinal Radiology. Fig.: Radial endoscopic ultrasound image of an eccentric left sided T1 tumour. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 4 of 66

5 EUS guided FNA and core biopsies EUS allows ultrasound guided fine needle aspiration and core biopsies to be performed on accessible nodes and occasionally primary oesophageal lesions (inaccesible to endoscpic biopsies). MDCT MDCT of the thorax, abdomen and pelvis is the usual initial staging tool following endoscopic detection of malignancy. If unequivocal inoperable disease (local invasion of major structures or distal metastases) is detected on MDCT, the tumour is considered inoperable and the patient referred for palliative options. Page 5 of 66

6 Fig.: MDCT: axial image of lower oesophageal / junctional tumour. Impossible to determine if left diaphragmatic crus involved. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM PET-CT In many institutions where operative intervention is indicated, PET-CT is requested to exclude distal metastatic disease. It is especially helpful for characterisation of equivocal lymph nodes and focal liver lesions detected on conventional MDCT. It is not uncommon to detect incidental lesions such as colonic polyps or neoplasms. PET-CT has little role in local tumour staging and locally involved lymph nodes are commonly masked by prominent F18-FDG uptake in the adjacent primary tumour. Page 6 of 66

7 Fig.: Coronal fused PET-CT image of mid-oesophageal lesion with avid uptake of F18FDG. Page 7 of 66

8 References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Laparoscopy Laparoscopy provides additional information regarding peritoneal disease as small volume peritoneal metastases are commonly not detected on cross-sectional imaging modalities. Multi-disciplinary teams (MDT) It is vital that the choice of staging and subsequent management of oesophageal malgnancies are discussed by teams of medical professionals including Radiologists, Surgeons, Gastroenterologists, Pathologists, Oncologists and Specialist Nurses. This allows for the best outcomes and use of resources. Often there will be agreed pathways regarding staging. The results of staging allow choices regarding further treatment options including surgery, chemotherapy, radiotherapy, endoscopic mucosal resection, as well as purely palliative options such as endoscopic stenting. Endoscopists performing EUS are an important part of that team. Endoscopic Ultrasound - Endoscopes Endoscopic ultrasound is performed with specialised endoscopes with ultrasound transducers integrated onto the end. These are referred to as echoendoscopes. There are several types of echoendoscope available. They can have a probe that produces a radial image at right angles to the long axis of the endoscope. Alternatively a curvilinear probe is utilised that produces an image parallel to the endoscope. The latter Page 8 of 66

9 allows for ultrasound guided punctures to be made. In all cases a small disposable balloon is placed over the transducer. These are inflated with water to allow better contact / sound transmission. Gastroscope Usually a standard gastroscope is placed first to assess the malignant stricture. It is useful to observe the extent of the tumour below the gastro-oesophageal junction following retroflexion of the endoscope. The presence of Barrett's disease should be documented. Endoscopic biopsies are occasionally repeated if indicated. If the gastroscope does not pass through the tumour, then an ERCP type wire is passed through the stricture, allowing subsequent passage of a slim probe. Fig.: Conventional gastroscope used to assess extent of malignant stricture and place wires prior to utilisation of "slim probe". References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Slim Probe Radial Echoendoscope If a tight malignant stricture is encountered a thin caliber (8.5mm) slim probe echoendoscope is passed over a wire. These have a 7.5 MHz probe (Olympus MH-908). The latter endoscope lacks an endoscopic image and therefore is referred to as "blind". In the event that the stricture is too tight to allow passage of the slim probe, the stricture can potentially be dilated. This is avoided at our institution due to the risk of perforation. Page 9 of 66

10 Fig.: 8.5 mm slim probe. It lacks an endoscopic image and is therefore "blind". These are placed over an ERCP type wire. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Conventional Radial Echoendoscope In the event that the gastroscope passes easily through the malignant stricture, a conventional radial echoendoscope is subsequently used. The latter gives an oblique endoscopic view, has a maximal diameter of 13.8mm and has a 10 MHz probe (Olympus GF-UE260-AL5). Page 10 of 66

11 Fig.: Standard radial echoendoscope(olympus GF-UE260-AL5). The ultrasound probe lies at the site of the red band at the end of the endoscope. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 11 of 66

12 Fig.: Example of balloon placed over the end of an ultrasound probe. These are inflated with water to allow better contact. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Curvilinear Echoendoscope and Image Guided FNA / Core Biopsy Conventionally a radial echoendoscope is used to stage oesophageal malignancy, but there is no reason why a curvilinear probe cannot fulfill this role. These are commonly referred to as "linear probes", but produce a curvilinear image familiar to most Radiologists. They produce an image parallel to the long axis of the endoscope. These require repeated rotations of the endoscope to obtain a 360 degrees appreciation of the tumour however. Their great advantage is that they allow image guided fine needle aspiration (FNA) and core biopsies to be performed. Echoendoscopes such as the Olympus UCT 240-AL5 are available (diameter 14.6mm with 3.7mm accessory channel). Page 12 of 66

13 Page 13 of 66

14 Fig.: Photographic images demonstrating the technical stages of EUS FNA. A: a curvilinear echoendoscope; B: the protective needle sheath protruding from the scope, which allows atraumatic appreciation of the needle prior to puncture and prevents damage to the accessory channel of the endoscope; C: finally, the needle protruding out of the sheath. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Fig.: The needle handle at the proximal end of the scope. This allows control of how far the protective sheath protrudes from the end of the echoendoscope (usually fixed) and the extent to which the actual needle protrudes. Limits on the maximal needle extent can be controlled. The inner stylet is inserted into the the distal end of the handle. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 14 of 66

15 Fig.: Ultrasound image showing an echogenic FNA needle sampling a malignant lymph node. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 15 of 66

16 Images for this section: Page 16 of 66

17 Fig. 1: Radial endoscopic ultrasound image of an eccentric left sided T1 tumour. Page 17 of 66

18 Fig. 2: MDCT: axial image of lower oesophageal / junctional tumour. Impossible to determine if left diaphragmatic crus involved. Page 18 of 66

19 Fig. 3: Coronal fused PET-CT image of mid-oesophageal lesion with avid uptake of F18FDG. Page 19 of 66

20 Fig. 4: Conventional gastroscope used to assess extent of malignant stricture and place wires prior to utilisation of "slim probe". Fig. 5: 8.5 mm slim probe. It lacks an endoscopic image and is therefore "blind". These are placed over an ERCP type wire. Page 20 of 66

21 Fig. 6: Standard radial echoendoscope(olympus GF-UE260-AL5). The ultrasound probe lies at the site of the red band at the end of the endoscope. Fig. 7: Example of balloon placed over the end of an ultrasound probe. These are inflated with water to allow better contact. Page 21 of 66

22 Page 22 of 66

23 Fig. 8: Photographic images demonstrating the technical stages of EUS FNA. A: a curvilinear echoendoscope; B: the protective needle sheath protruding from the scope, which allows atraumatic appreciation of the needle prior to puncture and prevents damage to the accessory channel of the endoscope; C: finally, the needle protruding out of the sheath. Fig. 9: The needle handle at the proximal end of the scope. This allows control of how far the protective sheath protrudes from the end of the echoendoscope (usually fixed) and the extent to which the actual needle protrudes. Limits on the maximal needle extent can be controlled. The inner stylet is inserted into the the distal end of the handle. Page 23 of 66

24 Fig. 10: Ultrasound image showing an echogenic FNA needle sampling a malignant lymph node. Page 24 of 66

25 Imaging findings OR Procedure details Imaging Technique With a radial probe a circumferential image is obtained at ninety degrees to the long axis of the scope, which is ideal for assessing the tubular structure of the oesophagus. Doppler assessment is possible with electronic ultrasound probes, but usually adds little diagnostic information. Linear echoendoscopes are required for image guided sampling, but as discussed can potentially allow diagnostic staging. In order to achieve optimal contact with the wall of the oesophagus the ultrasound transducer is covered with a balloon that is inflated with water to allow optimal sound transmission. Excessive insufflation of the balloon will result in resistance of movement or inability to traverse the malignant stricture. Care has to be taken remove as much air as is possible to improve the ultrasound image. This unfortunately results in poor endoscopic views. If the slim probe is used, no suction is possible and it is therefore imperative not to excessively insufflate with the prior use of the gastroscope. Once the oesophagus is intubated, the image is manipulated such that the aorta is approximately in the 5 o'clock position. This results in an image that is orientated in much the same way as a conventional axial CT image i.e. the left side of the screen represents right sided structures of the body with posterior structures seen on the lower aspect of the screen. The endoscope is usually first passed into the stomach and any pathological coeliac axis or left gastric nodes are noted. Ascites is excluded. The left lobe of the liver is normally well visualised allowing detection of hepatic metastases. The pancreatic body, spleen, left kidney and left adrenal gland are commonly visualised. Adrenal metastases will on occasion be detected. Withdrawing into the oesophagus, the position of the diaphragmatic crura are noted and the presence of a hiatus hernia is documented. The tumour length / extent are assessed and pathological mediastinal nodes excluded. It is important to document the position of the tumour relative to the gastro-oesophageal junction. The position of the carina and aortic arch are documented. When constructing a report it is important to produce a standardised report easily understood by all those involved in the patient's management. Anatomical layers Page 25 of 66

26 Five layers of the oesophageal wall are seen on EUS, which correspond with the histological layers. Other imaging modalities lack the spatial resolution to allow stratification of these layers. Sonographic layer Echogenicity Histological layer 1 Inner hyperechoic Mucosal interface 2 Hypoechoic Deep mucosa 3 Hyperechoic Submucosa 4 Hypoechoic Muscularis propria 5 Outer hyperechoic Interface with Adventitia Numbered sonographic layers identified on conventional EUS Page 26 of 66

27 Fig.: EUS image illustrating the five sonographic layers of the oesophageal wall. 1: mucosa; 2: muscularis mucosae; 3: submucosa; 4: muscularis propria; 5: adventitia. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM TNM Staging Classification TNM staging is designated by the International Union against Cancer (UICC). 1 This has recently been updated to a seventh edition with major changes relevant to oesophageal tumour staging. T (local tumour) staging: Tis Carcinoma in situ/high grade dysplasia T1a Lamina propria or muscularis mucosae T1b Submucosa (figure 3 on page 48) T2 Muscularis propria (figure 4 on page 49) T3 Adventitia (figure 6 on page 51) T4a Pleura, pericardium, diaphragm or adjacent peritoneum T4b Other adjacent structures e.g. aorta, vertebral body, trachea (figure 10 on page 55) It must be documented whether pleura, diaphragm, azygous vein, pulmonary vein, pericardium, aorta, bronchus, carina, trachea or other structures are involved in a T4 tumour, as not all will preclude surgery. This is the reasoning behind the new T4a classification. A tumour the epicenter of which is within 5 cm of the esophagogastric junction and also extends into the oesophagus is classified and staged according to the oesophageal scheme Page 27 of 66

28 All other tumours with an epicenter in the stomach greater than 5 cm from the oesophagogastric junction or those within 5 cm of the EGJ without extension into the oesophagus are staged using the gastric carcinoma scheme 4 It is important to be familiar with the Siewart classification of junctional tumours : Type I: carcinoma of the distal oesophagus, which may infiltrate the oesophagogastric junction from above. Type II: true carcinoma of the cardia arising immediately at the oesophagogastric junction. Type III: subcardial gastric carcinoma that infiltrates the oesophagogastric junction and distal oesophagus from below. (depending on where the epicentre of the tumour is this could still be considered an oesophageal tumour for TNM staging). N (regional nodal) staging 1 N0 No regional lymph nodes N1 1 to 2 regional lymph nodes (figure 13 on page 58) N2 3 to 6 regional lymph nodes N3 Greater than 6 regional lymph nodes It should be noted that the presence of a malignant coeliac axis node is no longer considered M1a disease which may impact on the considered importance of lymph node EUS guided FNA at this site. M (metastases) staging M1 Metastases Includes non-regional nodes. Page 28 of 66

29 Imaging appearances of malignant disease T Stage T1 disease Page 29 of 66

30 Fig.: T1 tumour. EUS image demonstrating the hyperechoic layers of the oesophageal wall (mucosa, submucosa and adventitia). A hypoechoic tumour is identified, which does not extend through the submucosa, although possibly invades the structure (T1b). References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 30 of 66

31 T2 disease Fig.: T2 tumour extending through the submucosa and invading the muscularis propria, which has an intact outer margin. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM T3 disease Page 31 of 66

32 Fig.: Note irregular margin to outer margin of the 4th sonographic layer (muscularis propria) denoting invasion into the adventia. Note lymph node. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 32 of 66

33 Fig.: Junctional tumour at level of left crus. The outer margin of the muscularis propria is irregular in keeping with T3 disease. The tumour appears to abut, but not invade the left crus. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 33 of 66

34 Fig.: MDCT: axial image of lower oesophageal / junctional tumour. Impossible to determine if left diaphragmatic crus involved. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM T4 disease Page 34 of 66

35 Fig.: Bulky eccentric lower oesophageal tumour. Tumour invades the anterior aspect of the descending throacic aorta. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 35 of 66

36 Fig.: A T4 junctional tumour invading the left lobe of the liver at o'clock. This was confirmed at laparoscopy. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM EUS is the imaging modality with the highest sensitivity and specificity for the assessment of T stage. In a meta-analysis by Puli et al., they reported that EUS performs better with advanced (T4) over early disease (T1). The pooled sensitivity of EUS to detect T1 disease was 81.6% and 91.4% in T4 disease. 2 The normal oesophageal wall has a thickness of less than 3mm and abnormal thickening is often easily seen on CT. However, a collapsed oesophagus or hiatus hernia may cause apparent thickening of the distal oesophagus resulting in a false positive diagnosis of malignancy. Mural stratification is not demonstrated on CT and it is impossible to distinguish between early T stages. CT appearances such as degree of contact with thoracic aorta and loss of fat planes allow some assessment of local tumour extent, but not possible to obtain the high resolution images seen with endoscopic ultrasound. Page 36 of 66

37 The role of FDG-PET in T staging is very limited given its poor spatial resolution. Tracer uptake also occurs in benign conditions, such as inflammation, and mild FDG activity may result from normal muscular contractions. The primary application of PET in oesophageal cancer staging is to detect distant disease. Fig.: Axial PET-CT image demonstrating increased tracer uptake in an oesophageal tumour. The resolution of PET-CT is insufficient to have significant value in determining T stage. This lesion was T4 on EUS (invasion of descending thoracic aorta). References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM N (nodal disease) Page 37 of 66

38 Lymph node involvement is common at the time of diagnosis of oeseophageal carcinoma and is present in 60-80% of cases. Malignant nodes are suggested by an increase in size. However, this is not absolute as lymph nodes may be increased secondary to a coexisting benign condition, such as chronic lung disease. In the case of adenocarcinoma, on a background of Barrett's oesophagus, nodes may be enlarged due to either or both conditions. It is also possible for small nodes to contain malignant microdeposits. Enlarged lymph nodes in the abdomen are more indicative of malignant nodal disease than mediastinal nodes, which are often reactive. Nodes over 10mm are generally considered to be malignant. Smaller size criteria are applied to supraclavicular and retrocrural nodes with an established lower limit of normal of 6mm. Regional lymph nodes are more accurately assessed with EUS. As with CT, the size of the node is used to differentiate benign from malignant enlargement, but similarly, this is not conclusive. The added benefit of EUS is the ability to assess internal structure. In the presence of the following four criteria, the probability of nodal malignancy is approximately 80%. Size greater than 1cm Hypoechoic with loss of the normal hyperechoic fatty hilum Well demarcated Round rather than kidney-bean in shape Page 38 of 66

39 Fig.: Probable benign node. It has a short axis diameter if 6mm, is ovoid in shape and has a preserved echogenic (fatty) hilum. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 39 of 66

40 Fig.: Sub-centimetre rounded hypoechoic node with well defined margins. It has no central echogenic hilum. It is likely malignant. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 40 of 66

41 Fig.: Significantly enlarged hypoechoic node. It is almost certainly malignant. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM FNA substantially improves the sensitivity and specificity of EUS in evaluating N stage 6 disease, with a reported increase in sensitivity of 96.7% compared to 84.7% without. Curvilinear echoendoscopes are used to perform FNA or core biopsy. The ability to obtain a definitive cytological or histological sample is the greatest advantage of EUS guided FNA. Care should be taken to avoid traversing the primary tumour as it may not be th possible to ascertain the origin of the sample. In the former TNM classification (6 edition) tumours were upstaged to M1a in the presence of coeliac axis disease and therefore nodal sampling below the coeliac axis trifurcation was given added emphasis. Under the latest classification coeliac nodes are considered local. Most locally involved nodes will be excised will be excised at surgery. There are low risks associated with FNA (bleeding Page 41 of 66

42 and infection). Close working relationships with cytopathology is essential to optimise cytology results. The choice to perform EUS guided FNA should be a multi-disciplinary decision. Fig.: Echogenic needle identified within a hypoechoic malignant node. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 42 of 66

43 Fig.: Malignant cells identified from a cytology specimen. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM M Stage Endoscopic ultrasound has a limited role in detecting metastatic disease. It may potentially detect low volume ascites not detected on MDCT, indicating peritoneal disease. Adrenal and liver lesions will usually have been detected on prior staging MDCT. The major advantage of PET-CT is in the detection of distant metastases, especially distant nodal disease which is equivocal on MDCT. FDG-PET has been shown to be more accurate than CT in identifying distant metastases. In their meta-analysis, van Vliet et al. reported sensitivities and specificities of 71% and 93% for FDG-PET and 52% and 7 92% for CT, respectively. Cross-sectional imaging may also detect incidental primary lesions. Page 43 of 66

44 Fig.: MDCT: Bone windows demonstrate lytic bone metastasis on anterior aspect of vertebral body. Note adjacent thickening of oesophagus at site of primary tumour and large volume ascites. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 44 of 66

45 Fig.: PET-CT demonstrates incidental avid FDG uptake in sigmoid colon. Patient initially presented with mid-oesophageal tumour. Colonoscopy demonstrated colonic adenocarcinoma. References: J. K. Bell; Radiology, Salford Royal NHS Foundation Trust, Salford, UNITED KINGDOM Page 45 of 66

46 Images for this section: Page 46 of 66

47 Fig. 1: EUS image illustrating the five sonographic layers of the oesophageal wall. 1: mucosa; 2: muscularis mucosae; 3: submucosa; 4: muscularis propria; 5: adventitia. Page 47 of 66

48 Fig. 2: T1 tumour. EUS image demonstrating the hyperechoic layers of the oesophageal wall (mucosa, submucosa and adventitia). A hypoechoic tumour is identified, which does not extend through the submucosa, although possibly invades the structure (T1b). Page 48 of 66

49 Fig. 3: T1 tumour: an eccentric left sided lower oesophageal tumour with probable invasion into the submucosa (the echogenic layer, also referred to as sonographic layer 3). Appearances suggestive of T1b disease. Page 49 of 66

50 Fig. 4: T2 tumour extending through the submucosa and invading the muscularis propria, which has an intact outer margin. Page 50 of 66

51 Fig. 5: T3 tumour. The outer margin of the 4th sonographic layer (hypoechoic muscularis propria) is indistinct suggestive of invasion into adventitia. Probable maligant node. Page 51 of 66

52 Fig. 6: T3 junctional tumour at the level of the left crus. EUS illustrating the normal sonographic appearance of the left crus (white arrow) and pleura (red arrow). If the lesion were to invade the crus it would be classified T4a. Page 52 of 66

53 Fig. 7: Axial CT image in the portal venous phase showing direct invasion of the left lobe of the liver (red arrow). Page 53 of 66

54 Fig. 8: A T4 junctional tumour invading the left lobe of the liver at o'clock. This was confirmed at laparoscopy. Page 54 of 66

55 Fig. 9: Axial PET-CT image demonstrating increased tracer uptake in an oesophageal tumour. The resolution of PET-CT is insufficient to have significant value in determining T stage. This lesion was T4 on EUS (invasion of descending thoracic aorta). Page 55 of 66

56 Fig. 10: Invasion of the thoracic aorta is confirmed on EUS. Page 56 of 66

57 Fig. 11: Probable benign node. It has a short axis diameter if 6mm, is ovoid in shape and has a preserved echogenic (fatty) hilum. Page 57 of 66

58 Fig. 12: Sub-centimetre rounded hypoechoic node with well defined margins. It has no central echogenic hilum. It is likely malignant. Page 58 of 66

59 Fig. 13: Significantly enlarged hypoechoic node. It is almost certainly malignant. Page 59 of 66

60 Fig. 14: Cytology: malignant cells from a sample obtained by FNA of a malignant node. Page 60 of 66

61 Fig. 15: MDCT: Bone windows demonstrate lytic bone metastasis on anterior aspect of vertebral body. Note adjacent thickening of oesophagus at site of primary tumour and large volume ascites. Page 61 of 66

62 Conclusion EUS has excellent sensitivity and specificity in accurately diagnosing the T and N stage of oesophageal cancer due to the application of high frequency probes. Understanding its complementary role is vital to radiologists involved in oesophageal cancer staging. It is the most sensitive imaging modality for the detection of regional lymph node metastases, which is increased with guided FNA or core biopsy. MDCT is the primary imaging modality in the staging of oesophageal carcinoma. In the absence of distant disease on MDCT further staging with with EUS and FDG-PET should be considered. The diagnostic performance of PET-CT for detection of distant metastases is significantly higher than that of MDCT. It should be considered routinely in patients considered for curative surgery where resources allow. CT, EUS and PET play a crucial role in staging oesophageal carcinoma. Each has benefits and limitations that should be understood by the members of a multidisciplinary team. Page 62 of 66

63 Personal Information Page 63 of 66

64 Declaration No conflict of interest to declare. Dr Jon Bell Specialist Registrar Manchester Radiology Training Scheme, UK Dr D Awad Specialist Registrar Manchester Radiology Training Scheme, UK Dr D Kasir Consultant GI Radiologist Salford Royal NHS Foundation Trust, UK Mr J Vickers Consultant Upper GI Surgeon Salford Royal NHS Foundation Trust, UK Dr Luke Williams Consultant GI Radiologist Salford Royal NHS Foundation Trust, UK Page 64 of 66

65 References 1. Sobin L, Gospodarowicz M, Wittekind C. International union against cancer TNM classification of malignant tumours (7th edition). Blackwell December Puli SR, Reddy JBK, Bechtold ML, et al. Staging accuracy of esophageal cancer by endoscopic ultrasound: A meta-analysis and systematic review. World J Gastroenterology. 2008; 14(10): Penman I, Norton S, Harris K. Staging of oesophago-gastric carcinoma by endoscopic ultrasonography: guidance and minimum standards. UK EUS Users Group in association with the British Society of Gastroenterology. September Siewart JR, Feith M, Werner M, Stein HJ. Adenocarcinoma of the esophagogastric junction results of surgical therapy based on anatomical/topographic classification in 1,002 consecutive patients. Ann Surg. 2000; 232(3): Skehan SJ, Brown AL, Thompson M, et al. Imaging features of primary and recurrent esophageal cancer at FDG PET. Radiographics. 2000;20: Pfau PR, Perlman SB, Stanko P, et al. The role and clinical value of EUS in a multimodality esophageal carcinoma staging program with CT and positron emission tomography. Gastrointestinal Endoscopy. 2007; 65: Van Vliet EP, Heijenbrok-Kal MH, Hunink MG, Kuipers EJ et al. Staging investigations for oesophageal cancer: a meta-analysis. Br J Cancer 2008 Feb 12;98(3): Page 65 of 66

66 Page 66 of 66